With the decline and perhaps eventual demise of “manifest disregard of the law” after Hall Street Associates v. Mattel, Inc., more attention has been devoted by counsel and the courts to what it means for an arbitrator to “exceed [her] powers” (FAA Section 10 (a) (4)) such that vacatur is warranted of an award made at a place of arbitration in the United States.
In the Stolt-Nielsen case, the Supreme Court held that an arbitrator exceeded powers in ordering class arbitration when the parties had made no agreement about class arbitration. That conclusion was reached for essentially two reasons. First, the choice between class and individual arbitration could not be considered as merely as procedural issue arising in the course of the arbitration, as the fundamental nature of the proceeding turned on the decision. Second, when the arbitrator’s rationale is not even an arguable application of contract language or governing law, but instead is a policy judgment by the arbitrator (her “own brand of industrial justice”), the arbitrator, having a mandate only to apply the contract and the law, cannot make a decision on another basis — even of a question expressly submitted to her by the parties.
In a decision last week that may help to calm fears that Stolt-Nielsen signals a new wave of uncertainty for award enforcement and a shift in momentum toward more exacting judicial review of awards, the US Fourth Circuit Court of Appeals held that an arbitral tribunal did not exceed its powers by ruling on the validity of an arbitration agreement whose validity was also at issue before a different arbitral tribunal hearing essentially the same dispute between the same parties, the latter tribunal having been constituted under the challenged agreement. (Central West Virginia Energy, Inc. v. Bayer Cropscience LP, 2011 WL 2725819 (4th Cir. July 14, 2011.))
This was a duelling contracts/duelling arbitrations case arising from coal supply agreements. A 1997 agreement, renewed several times, provided for AAA commercial arbitration in Charleston, West Virginia. A purported 2008 replacement contract, whose validity was contested, provided for AAA arbitration in Richmond, Virginia. Buyer commenced arbitration (in Charleston) under the 1997 contract and claimed it remained in force. Seller commenced arbitration (in Richmond) under the purported 2008 contract. And in the case started by buyer, seller invoked the 2008 agreement as a defense, putting its validity in issue. But seller insisted that the first tribunal defer to the second tribunal on the validity issue, or allow a court to decide which of the tribunals should rule on validity. The first tribunal refused these requests, declared the 2008 contract invalid, and entered an award for buyer. The second tribunal followed the ruling of the first. And district courts in both cities refused to vacate the awards.
In the Fourth Circuit, seller invoked Stolt-Nielsen and the Supreme Court’s 2002 decision in Howsam v. Dean Witter Reynolds, Inc., to argue that the question of which tribunal should rule on validity of the challenged contract was a “gateway” issue for a district court. The Court rejected this, on the basis that the question whether one arbitral tribunal or another resolves an issue, unlike the question whether an issue is decided in arbitration at all, is not an issue of arbitrability but rather an issue of arbitral procedure.
Seller tried to take Stolt-Nielsen even a step further. Seller argued that the first arbitral tribunal had wrongly relied on a theory that, by placing contract validity in issue, seller had waived the objection to having the first tribunal not the second decide that issue. The waiver theory, seller argued, was on a par with the Stolt-Nielsen tribunal’s policy-based preference for class arbitration, and was impermissibly the arbitrators’ “own brand of industrial justice.” The Fourth Circuit rejected this as well, and found that in fact the point of departure of the tribunal’s analysis was the broad “all disputes” language of the arbitration clause, making the tribunal’s decision that it should resolve the contract validity dispute an arguable, and hence permissible, construction of the agreement.
What lessons can we draw from this arguably sui generis duelling arbitrations case? One lesson, I believe, is that American arbitration jurisprudence provides the answer to some criticisms, from abroad, that the US needs to “modernize” or at least refine the text of the Federal Arbitration Act. Certainly it is true that an arbitration statute that refers in general terms to “exceed[ing] powers” as a ground for vacatur does not give arbitrators clear guidance on what to do in a case such as Central West Virginia. But what effort to catalog the manifold ways that an arbitrator might exceed her powers would be a useful replacement for the jurisprudence of US courts on the subject? When one reads the Fourth Circuit’s decision, one gains a measure of confidence that “common law” adjudication of arbitration law issues has yielded principles that are relatively consistent, or at least reconcilable, from case to case. Arbitrators who read the arbitrability and vacatur case law can proceed with a large measure of confidence, and it is to be doubted that statutory reform would increase that confidence level, at least not without an unacceptable cost in the creation of more detailed, but possibly underinclusive and rigid, categories of prohibited arbitral conduct.