Center for International Commercial and Investment Arbitration
Supreme Court case law teaches us that the federal interest in arbitration does not consist of enforcing agreements to arbitrate according to some sort of abstract or ideal arbitral model, but rather according to the particular arbitral model upon which the parties had agreed. This body of law is driven by the same notions of party autonomy that underlie the law of arbitration generally. That parties may agree to forego access to national courts in favor of arbitration is an initial manifestation of that attitude. By logical extension, the parties also enjoy extraordinary latitude in determining the features that "their" eventual arbitration should display.
A legal system that elevates party autonomy in arbitration to such a high level of importance ought, at least in principle, to develop some consistent methodology for identifying the arbitral features that the parties intended for their arbitration. United States arbitration law simply has not done so, thus leaving arbitrators and courts alike somewhat at sea. For the arbitrators, there is more at stake than merely respecting party preferences, important though that is. A tribunal that deviates substantially from the procedures agreed upon by the parties, and insisted upon by one of them, risks finding its award vacated or denied recognition or enforcement. Of course, for any number of reasons, an award resulting from an arbitral process that deviated from party intentions may nevertheless survive and earn recognition or enforcement: the deviation may be regarded as insubstantial; the complaining party may be deemed to have waived its objections or otherwise acquiesced; or the general latitude that arbitrators enjoy in matters of procedure may operate to shield the award.
George A. Bermann,
Ascertaining the Parties' Intentions in Arbitral Design,
Penn. St. L. Rev.
Available at: https://scholarship.law.columbia.edu/faculty_scholarship/712