dissenting:
I respectfully dissent.
This ease is the arbitration counterpart to Wal-Mart Stores, Inc. v. Dukes, 546 U.S. -, 131 S.Ct. 2541, 180 L.Ed.2d 374, 2011 WL 2437013 (June 20, 2011). As in Wal-Mart, the class action complaint filed *128by appellants in the Southern District, followed by their request for class arbitration, the subject of this proceeding, asserted a company-wide gender discrimination claim against a national employer based on a statistical study finding disparities in pay and promotions between male and female employees where the company has delegated relevant decision-making power to local executives. Wal-Mart, 546 U.S. at -, 131 S.Ct. 2541, 2011 WL 2437013, at *9-10. I need not pause to consider the Wal-Mart decision’s effect on the present matter because Stolt-Nielsen v. AnimalFeeds International Corp., — U.S. —, 130 S.Ct. 1758, 176 L.Ed.2d 605 (2010), is a binding precedent on all fours.
The issue in Stoltr-Nielsen, which reversed a decision of this court, was, in the Supreme Court’s words, “whether imposing class arbitration on parties whose arbitration clauses are ‘silent’ on that issue is consistent with the Federal Arbitration Act (FAA).” Id. at 1764. The Court answered that question in the negative. Id. at 1775.
My colleagues attempt to distinguish Stolt-Nielsen on several grounds. First, they note the stipulation in Stolt-Nielsen that the arbitration agreement there was “silent” as to class arbitration. Maj. Op. at 119-20, 123-24. Second, they rely on Stolt-Nielsen’s recognition of the possibility of implied agreements to class arbitration. Maj. Op. at 120-21, 123-24. Third, they make reference to the arbitrator’s reliance on Ohio law in this case. Maj. Op. at 125-26. Fourth, they rely on provisions of the various arbitration agreements at issue here empowering arbitrators to award generally available types of legal and equitable relief. Maj. Op. at 126-27. Fifth, and finally, they invoke the limited scope of judicial review of arbitration agreements. Maj. Op. at 123-25. None of these grounds, which I address seriatim, suffices to distinguish Stolt-Nielsen.
I.
My colleagues emphasize the existence of a stipulation between the parties in Stolt-Nielsen that the pertinent arbitration clause was “silent” as to class arbitration. Maj. Op. at 119-20, 123-24. This stipulation simply reflected the fact that each party recognized that the arbitration clause neither specifically authorized nor specifically prohibited class arbitration. Brief for Respondent at 26, Stolt-Nielsen, — U.S. —, 130 S.Ct. 1758, 176 L.Ed.2d 605 (2010) (No. 08-1198), 2009 WL 3404244 at *26. The disputed issue was, as it was in the Second Circuit, the effect of that silence. Stolt-Nielsen SA v. AnimalFeeds Int’l Corp., 548 F.3d 85, 86 (2d Cir.2008) (“The question presented on this appeal is whether the arbitration panel, in issuing a clause construction award construing that silence to permit class arbitration, acted in manifest disregard of the law.”). That is precisely the issue in this matter.
Neither party here claims that any of the various arbitration clauses at issue1 either specifically authorizes or specifically precludes class arbitration, the same facts that constituted the “silence” that existed in Stolt-Nielsen. Indeed, that is the reason my colleagues seek refuge in a nonexistent finding of the arbitrator, discussed infra Part II, that an agreement to class arbitration was implied.
If the parties’ arguments here and the face of the agreements are insufficient to establish their “silence,” then the arbitrator’s decision certainly was when she explicitly found “there is no mention of class *129claims.” Clause Construction Award (Ex. F to Sterling Mem. of Law in Support of Mot. to Vacate) at 3, Jock v. Sterling Jewelers, Inc., 677 F.Supp.2d 661 (2009) (No. 08 Civ. 2875). That being the case, and not having the benefit of the Supreme Court’s views in Stolt-Nielsen, the arbitrator framed the issue before her by specifically relying on the Second Circuit’s decision in Stolt-Nielsen. Following our now-reversed decision, she found only that the arbitration clauses here did not “prohibit class arbitrations,” and therefore that they allowed class arbitrations. Clause Construction Award, supra, at 4. In short, the arbitrator viewed the arbitration clause here as, in all pertinent respects, identical to that in Stolt-Nielsen.2
II.
My colleagues also seek to distinguish Stolt-Nielsen on the ground that the decision recognized the possibility of implied agreements to class arbitration. Maj. Op. at 120-21, 123-24. True. However, the Supreme Court explicitly held that, in an FAA proceeding, an implied agreement could not be inferred from an arbitration clause’s failure to “preclude class arbitrations.” 130 S.Ct. at 1775 (emphasis omitted). Had it not decided that question, the Court would have had either to address whether such an agreement might be implied from the arbitration clause in that case or to remand for such a determination. Instead, it held that “imposing class arbitration on parties whose arbitration clauses are ‘silent’ on that issue is [injconsistent with the Federal Arbitration Act.” Id. at 1764, 1775-76.
An “implicit” agreement to class arbitration cannot, therefore, be inferred from an arbitration agreement’s “silence” or “failure to preclude” class arbitrations, much less from thin air. Indeed, the arbitrator in this case did not purport to find an implied agreement. Following our decision in Stolt-Nielsen, where we speculated that “the arbitration panel may have concluded that even though the arbitration clauses are silent ... their silence bespeaks an intent not to preclude class arbitration,” 548 F.3d at 99, she inferred the class arbitration agreement entirely from the failure of the various agreements to “prohibit class arbitration,” precisely what the Supreme Court said cannot be legitimately done. 130 S.Ct. at 1775.
Nowhere in her opinion does she purport to identify any provision of the agreement supporting the existence of an implied agreement. Far from it. Some of the agreement’s provisions are in fact inconsistent with class arbitration, and the arbitrator had to manifestly disregard key provisions of the arbitration clauses in order to hold that class arbitration was authorized.
The members of the purported class each signed a provision that was often different from that signed by other mem*130bers of the purported class. The arbitration clauses contain significant differences as to: (i) waiver of rights to seek relief in other tribunals; (ii) the kinds of disputes subject to the RESOLVE program; (iii) the particular laws under which claims may arise; (iv) contribution amounts by employees to the costs of arbitration; (v) descriptions of the relief available; (vi) limitations periods; (vii) the role of a court before which litigation has already begun; (viii) where the arbitration is to take place; and (ix) the source of the law under which the arbitrable claim is to be decided. See Spagnola Affidavit (Ex. C to Sterling’s Mem. of Law in Support of Mot. to Vacate) at Tab 2, Jock, 677 F.Supp.2d 661 (No. 08 Civ.2875).
When the various agreements are read in the aggregate, or even separately, the differences preclude class arbitrations. The arbitrator herself acknowledged that many of the arbitration agreements contained “unique contractual provisions for local venues, the application of local laws, and the selection of locally-licensed arbitrators,” Clause Construction Award, supra, at 4, all of which are inconsistent with class arbitration. Pertinent common themes in the arbitration clauses are (i) the need to follow Steps 1 and 2 of the RESOLVE program — Step (1): a complaint with reference to supporting evidence; and Step (2): resort to mediation if Sterling does not remedy the complaint— and (ii) the failure to mention class arbitration — silence. These common provisions are inconsistent with an implied agreement to class arbitration. As a result, the arbitrator found it necessary to hold that Steps 1 and 2 of the RESOLVE program were to be ignored in class arbitrations. Then, having framed the issue according to this court’s Stoltr-Nielsen decision, she went on to find that the failure to “prohibit class arbitrations” prevailed. Clause Construction Award, supra, at 5.
My colleagues dismiss as “disingenuous” Sterling’s argument “that permitting class arbitration to proceed would lose sight of the requirement that ‘parties may specify with whom they choose to arbitrate their disputes,’ Stolt-Nielsen, 130 S.Ct. at 1774, because the putative class in this case would presumably include only employees who signed the RESOLVE agreement.” Maj. Op. at 127 n. 3. However, Sterling agreed to arbitrate only with employees who complete steps 1 and 2 of the RESOLVE procedures; it has not agreed to arbitrate with each and every employee that signed the RESOLVE program agreement, whether or not they assert a claim pursuant to its provisions.
Far from implying class arbitration from the agreement, therefore, the arbitrator manifestly disregarded both unique and common features of all the agreements based on the view of silence as to class arbitration upheld by the Second Circuit in Stolt-Nielsen but thereafter rejected by the Supreme Court. Having now been enlightened by the Supreme Court’s views, we do not have the arbitrator’s excuse.
III.
My colleagues also believe that the arbitrator purported to be acting under Ohio law. Maj. Op. at 125-26. That is not the case. All she stated about Ohio law was that “the question of whether class claims are permitted or prohibited by an agreement that does not expressly address the issue ... has apparently not been addressed in any reported decision by an Ohio court.” Clause Construction Award, supra, at 4.
Moreover, Ohio law does not govern the issue before us. Stolt-Nielsen held that consent to class arbitration cannot be inferred from an agreement’s failure to preclude it. The Supreme Court did not base *131its decision on the law of a particular state or federal maritime law. Rather, it relied on principles of law derived solely from the Federal Arbitration Act. Stolt-Nielsen, 130 S.Ct. at 1773 (“While the interpretation of an arbitration agreement is generally a matter of state law, the FAA imposes certain rules of fundamental importance, including the basic precept that arbitration ‘is a matter of consent, not coercion.’ ”) (internal citations omitted); id. at 1775 (“The [arbitration] panel’s conclusion is fundamentally at war with the foundational FAA principle that arbitration is a matter of consent.”) (emphasis added). The Court’s exclusive reliance on the FAA as a source of governing law was in contrast to the Second Circuit’s extensive discussion of New York and federal maritime law in its Stolh-Nielsen decision, 548 F.3d at 96-101, and to the arguments before the Supreme Court that there were considerable precedents in New York and maritime law authorizing class arbitration in the face of contractual silence. Brief for Respondent at 31-33 & 38, Stolt-Nielsen, 130 S.Ct. 1758 (No. 08-1198).
The Supreme Court treated New York and federal maritime law as irrelevant. Rather, it framed the question as “whether imposing class arbitration on parties whose arbitration clauses are ‘silent’ on that issue is consistent with the Federal Arbitration Act,” and answered that question in the negative. Stolt-Nielsen, 130 S.Ct. at 1764. The exclusive role of the FAA in providing substantive law to determine the requisite consent to class arbitration was recently confirmed by the Supreme Court’s decision in AT&T Mobility LLC v. Concepcion, — U.S. —, 131 S.Ct. 1740, 179 L.Ed.2d 742 (2011). That case held that the FAA substantive law “preempted” state contract law rendering unenforceable as unconscionable an explicit contractual preclusion of class arbitration. Id. at 1753.
The Supreme Court’s concerns under the FAA with inferring consent to class arbitration are based on the fundamental differences between bilateral arbitration and class arbitration. As Stolt-Nielsen stated:
An implicit agreement to authorize class-action arbitration, however, is not a term that the arbitrator may infer solely from the fact of the parties’ agreement to arbitrate. This is so because class-action arbitration changes the nature of arbitration to such a degree that it cannot be presumed the parties consented to it by simply agreeing to submit their disputes to an arbitrator____ Consider just some of the fundamental changes brought about by the shift from bilateral arbitration to class-action arbitration. An arbitrator chosen according to an agreed-upon procedure no longer resolves a single dispute between the parties to a single agreement, but instead resolves many disputes between hundreds or perhaps even thousands of parties.... And the commercial stakes of class-action arbitration are comparable to those of class-action litigation, even though the scope of judicial review is much more limited. We think that the differences between bilateral and class-action arbitration are too great for arbitrators to presume, consistent with their limited powers under the FAA, that the parties’ mere silence on the issue of class-action arbitration constitutes consent to resolve their disputes in class proceedings.
130 S.Ct. at 1775-76 (internal citations omitted) (emphasis added).
Moreover, as the Supreme Court has recently noted at some length, even where a designated arbitration organization has established rules for class arbitration, as has the American Arbitration Association (“AAA”) in the present case, the shortcom*132ings of class arbitration are substantial. Concepcion, 131 S.Ct. at 1750-52 (“[W]hile it is theoretically possible to select an arbitrator with some expertise relevant to class-certification question, arbitrators are not generally knowledgeable in the often-dominant procedural aspects of certification, such as the protection of absent parties .... The absence of multilayered review makes it more likely that errors will go uncorrected.”). Class proceedings carry significant risks for plaintiffs as well as defendants. A class proceeding may well be driven by a quest for a substantial award of counsel fees alone, but the award will bind the members of the class who may pay little attention to the proceedings and are generally unaware of the consequences. Even where, as under the AAA rules, any settlement must be approved by the arbitrator, the critical safeguard of fairness review by an independent Article III judge, present in judicial class actions, is lacking. See id. at 1752; Fed.R.Civ.P. 23. While approval of a settlement by the arbitrator may sometimes be required, counsel will have an influence on both the selection of an arbitrator and even his or her fees. See Hans Smit, Contractual Modification of the Arbitral Process, 113 Penn. St. L.Rev. 995, 1009 (2009) (noting that because the tribunal is normally composed of arbitrators designated by the parties, it will normally agree with the parties once they have agreed upon a proper class and settlement). An arbitrator who disapproves a settlement may do so only at a risk of not being selected in future cases.
IV.
My colleagues also rely upon the provision in the present agreement that the arbitrators may award any legal or equitable relief generally available in courts. Maj. Op. at 126-27. Clearly, this provision refers only to relief in the form of an award based on a violation of law or contract — damages, injunctions, etc. — and not to the availability of procedures used to pursue such relief. A class can be certified and yet not get “relief,” i.e. it may lose. Significantly, one form of this provision found in some of the arbitration agreements here states that “the Arbitrator shall have the power to award any types of legal or equitable relief that would be available in a court of competent jurisdiction including, but not limited to, the costs of arbitration, attorney fees and punitive damages for causes of action when such damages are available under law.” Spagnola Affidavit at Tab 2, supra, at 1. Applying the canon “ejusdem generis” — a general item in a list of specific items of a particular genre must be construed to be limited to that genre, see Rajah v. Mukasey, 544 F.3d 427, 435 (2d Cir.2008) — the word “relief’ is a reference to remedies compensating or punishing for harm done.
V.
Finally, my colleagues note the limited review of arbitration awards exercised by federal courts. Maj. Op. at 123-25. While that is surely a relevant point, this concern was not only as fully applicable to the award in Stolt-Nielsen as it is here but was also discussed extensively by the Supreme Court in that case. 130 S.Ct. at 1767-68. I will rely on the Supreme Court’s discussion.
Moreover, my colleagues’ deference here is not to the arbitrator’s actual decision, which framed the issue under the Second Circuit’s decision in Stolt-Nielsen, found that the agreement made “no mention of class” arbitration, allowed class arbitration because the agreement did not “prohibit” it, and exceeded her authority by manifestly disregarding explicit provisions of the agreement inconsistent with class arbitration. See Yusuf Ahmed Alghanim & Sons *133v. Toys “R” Us, Inc., 126 F.3d 15, 23 (2d Cir.1997) (“[Arbitral] awards may be vacated ... where the arbitrator’s award is in manifest disregard of the terms of the agreement.”). Rather their deference is to a decision that was, in fact, never made by the arbitrator — namely that the arbitration agreement implied an agreement to class arbitration.
I therefore respectfully dissent.
. Many arbitration agreements, with significantly different provisions, were signed by the individual Sterling employees, see infra, p. 129-30.
. Moreover, my colleagues’ narrow reading of Stolt-Nielsen — limiting it to cases with supposed formal stipulations as to contractual silence — casts grave doubt over the reasons for the Court's even granting a writ of certiorari in that case. Under Supreme Court Rule 10, "a writ of certiorari will be granted only for compelling reasons.” The Supreme Court grants such writs only where "a state court or a United States court of appeals has decided an important question of federal law that has not been, but should be, settled by this Court, or has decided an important federal question in a way that conflicts with relevant decisions of this Court.” Sup.Ct. R. 10(c). If Stolt-Nielsen resolves only the effect of a sui generis and idiosyncratic stipulation of the parties, the case hardly meets those criteria. Given my colleagues' narrow reading of the decision and their reliance on an analysis indistinguishable from the decision of the Second Circuit reversed by the Supreme Court, Stolt-Nielsen has been rendered an insignificant precedent in this circuit.