It is the first day of the new hockey season in North America, a suitable occasion for Arbitration Commentaries to bring you content inspired by our neighbors in the Great White North.

By now the news will probably have reached you that the highest appellate court of the province of Ontario, the Ontario Court of Appeal, earlier this week affirmed a first instance court’s ruling that denied a motion to vacate in part a NAFTA arbitral tribunal’s award against the Government of Mexico and in favor of the US multinational Cargill, Inc. (Mexico v. Cargill, Inc., 2011 ONCA 622 (Oct. 3, 2011)). In case you are learning about the case for this first time here, I report very briefly: Mexico established trade barriers to protect its cane sugar industry from competition from Cargill and other US suppliers of high fructose corn syrup. In Cargill’s NAFTA arbitration case under the UNCITRAL Rules, Cargill claimed and was awarded damages in two categories: lost profits of its Mexican affiliate, and lost profits on trans-border sales from Cargill in the US to the Cargill Mexican affiliate. The Tribunal rejected Mexico’s position that the second category of damages were not sustained by Cargill as an investor in relation to its investment; the Ontario Supreme Court (first instance) agreed with the Tribunal, and in this decision the Ontario Court of Appeal also agreed. Because the arbitration took place in Toronto under the UNCITRAL Rules, rather than under the ICSID Convention as NAFTA also permits, there was the opportunity for a motion to vacate under Ontario’s version of the UNCTIRAL Model Law on International Commercial Arbitration. Mexico took the position that the award of damages for Cargill’s lost US sales from the US to its Mexico affiliate were “beyond the scope of the submission to arbitration” (Model Law Article 34(2)(a)(iii)).

The issue to which the Court of Appeal devotes the most time and energy, and which will be of greatest transnational interest, is: what is the scope of review in a national court, applying the Model Law, when presented with a motion to vacate the award on jurisdiction of an UNCITRAL Rules investment treaty arbitral tribunal? The Ontario Court of Appeal held that the scope of review is “correctness” but precisely what this means and what it transnational implications are is open to debate and is the subject of this commentary.

Under American arbitration law, an arbitral tribunal’s decision concerning its own jurisdiction is reviewed de novo, without deference, unless the parties clearly and unmistakably agreed to submit the jurisdiction issues to arbitration. These American law rules are grounded in the notion that arbitration agreements are simple bilateral contracts between commercial parties, and that their existence and validity are suitable for courts to decide independently (without deference) because the decisions involve ordinary state law contract principles that courts apply every day. Equally, issues of the scope of disputes that are arbitrable fall normally within the province of courts to resolve independently, albeit with the benefit of a presumption based on the “pro-arbitration policy” that the US Supreme Court has engrafted on the FAA. But one reason for permitting courts to decide scope issues independently, even on review of an award, is that construing an arbitration clause normally should not be that different from construing other contract language, and with the benefit of contract interpretation jurisprudence this exercise usually will be well within the judicial “wheelhouse.”

But how would or should these rules apply when and if the Government of Mexico comes before a US court to vacate the award of a NAFTA arbitral tribunal in favor of a Canadian investor, on the ground that the damages awarded did not involve the investor’s investment in Mexico, and therefore were outside the jurisdiction of the tribunal?

What is immediately apparent — but the Ontario Court of Appeal does not confront directly — is that the NAFTA arbitral tribunal has an enormous comparative advantage over a national court in regard to the skills and knowledge applicable to deciding this “jurisdiction” question. The question is not remotely comparable to what passes for an arbitral jurisdiction issue in commercial arbitration. It involves fundamental issues of investment law, as to which depth of knowledge is presumably a main criterion for selection of the arbitrators (both in the treaty partners’ fashioning of their rosters, and in the parties’ case-specific choices). The Ontario decision reflects that the Tribunal’s decision on the jurisdiction issue brought into play NAFTA’s text, drafting history, and subsequent practice and understanding among the treaty parties; the Vienna Convention on the Law of Treaties; NAFTA arbitral case law dealing with similar issues of the territoriality vel non of claimed investor losses; principles of international law; and fact-intensive consideration of how Cargill structured its operations to sell high fructose corn syrup to food and beverage producers in Mexico. As appears often to be the case in investment arbitration, what is called jurisdiction is actually the heart of the matter — “the merits” — because the main issue is not what happened, nor whether the investor was harmed, but whether there is redress under the investment treaty for the harm that occurred as a result of the measures taken by the host State.

Why, then, would the Ontario court opt for a standard of review (“correctness”) that treats the NAFTA tribunal’s decision of a jurisdiction issue with any less deference than it would treat that tribunal’s award “on the merits”? The Ontario court’s opinion devotes considerable space to quotation of the UK Supreme Court’s decision in Dallah v Ministry of  Religious Affairs of the Government of Pakistan, wherein that Court maintained that an international arbitral tribunal (at least when it is a private tribunal in a commercial case) is normally entitled to no deference in its own determination of its own jurisdiction to adjudicate.  But Dallah was a commercial arbitration case, and the jurisdiction issue was a relatively straightforward one of whether the Pakistani ministry as a non-signatory of the agreement was bound to arbitrate. That is the kind of contract-agency law issue on which parties normally would expect judicial control of arbitral jurisdiction.

The Ontario Court of Appeal holds that the proper standard of review of the tribunal’s determination is “correctness,” and not (as proposed by Cargill) “reasonableness.” But to this reader/writer — a Canadian in spirit but not in Bar admission — there is less than satisfying clarity on what “correctness” means. The Court says that a court hearing a motion to vacate under the Model Law must be careful to determine if it is reviewing “a true question of jurisdiction,” because “courts are expected to intervene only in rare circumstances where there is a true question of jurisdiction.” But that does not seem to meaningfully limit the kinds of complicated questions involved in jurisdiction decisions of investment tribunals.

The Court gives as a hypothetical example of a jurisdiction ruling in a NAFTA case that could be vacated as incorrect a purported award to a Brazilian citizen, one who is manifestly not eligible to arbitrate as she is not a citizen of any of the treaty parties. But whereas the tribunal’s decision that Cargill’s lost US sales to its Mexican affiliate were within the tribunal’s jurisdiction to award was upheld as “correct,” does this not perhaps imply that an opposite decision by the tribunal would have been incorrect? The Ontario court does not appear to intend that we should understand its decision in this way. It rejects “reasonableness” as a standard of review, but one gets the clear sense that if the tribunal had ruled that Cargill’s distribution arrangements did not suffice to bring its lost US sales to its Mexican affiliate within NAFTA’s coverage, that outcome also would not have been set aside.

Rather, what “correctness” appears to mean is that the award will only be set aside if the error is plain on the face of the award in relation to the plain meaning of the NAFTA. It is akin to the “clerical or computational error” correction rules of most international arbitral institutions and national statutes: If the award adds two and two and gets to five not four, that outcome is entitled to no deference and may be changed by a competent court.

The reason such a rule makes sense is that it makes vacatur for jurisdictional error depend on criteria that a national court is no less competent than a NAFTA tribunal to apply. And so we may, without taking too much liberty, conclude that the transnational lesson of the Cargill decision from Ontario is that jurisdiction determinations of investment treaty tribunals, when they are subject to vacatur under an arbitration statute based on the UNCITRAL Model Law, shall not be vacated if the jurisdiction decision was based on criteria within the core competence of an investment treaty tribunal. The Ontario court says that the question whether lost capacity in Cargill’s US plants constituted damages by reason of Mexico’s breaches of treaty obligations “is a quintessential question for the expertise of the tribunal, rather than an issue of jurisdiction.” But one could equally say that the question was whether the damage was sustained by an investor in relation to its investment and so it was indeed an issue of jurisdiction. The key element is that this issue was in the tribunal’s “wheelhouse”! It is unhelpful to tack on an ill-defined dichotomy between jurisdiction and the merits in the context of investment arbitration.

In the US, perhaps happily, such a “core competence” principle would not be needed, because we have the First Options-derived rule that when parties agree to arbitrate according to rules that permit arbitrators to decide upon their own jurisdiction, then judicial review in the context of a vacatur motion of the award on jurisdiction is as limited as review of any other issue decided in an award. The UNCITRAL arbitration rules of course give arbitrators such power and US courts (notably the Second Circuit in the Chevron-Ecuador saga) have treated BIT-based arbitration under those rules as an agreement to “arbitrate arbitrability.”  The ICSID Additional Facility Arbitration Rules confer similar power. So any investment treaty arbitration award on jurisdiction that one could reasonably imagine being the subject of a motion to vacate in a US court would fall within the limited scope of review rule of First Options and its progeny.  Nevertheless, it is helpful to US practitioners, as well as those abroad, to take away from the Cargill case the useful lesson that jurisdiction determinations of investment treaty tribunals that involve issues within the core competency of the tribunal shall be vacated only where it can be said that the error is one that the reviewing court has equivalent competency to detect and correct.

Today Arbitration Commentaries briefly notes a new decision from a respected federal district judge in Houston, Texas, holding that a sale of goods contract that was between two US companies, but which provided for discharge of the shipped goods in a foreign port to be designated by the buyer created a sufficient international nexus to make the New York Convention applicable to the contract’s arbitration clause. (Tricon Energy, Ltd. v. Vinmar International, Ltd., 2011 WL 4424802 (S.D. Tex. Sept. 21, 2011).

The issue is litigated from time to time, especially when one party sees a tactical advantage to holding award confirmation or vacatur proceedings in a state court, and federal jurisdiction based on diversity of citizenship would not exist.

Curiously the Convention itself does not directly address when an award between domiciliaries of the same Convention State is governed by the Convention. This is expressly left to domestic law by Article I(1): “[The Convention] shall also apply to arbitral awards not considered as domestic awards in the State where their recognition and enforcement are sought.” Chapter Two of the Federal Arbitration Act (FAA) in Section 202 provides that an agreement or award entirely between US citizens “shall  be deemed not to fall under the Convention unless that relationship involves property located abroad, envisages performance or enforcement abroad, or has some other reasonable relation with one or more foreign states.”

The Court in Tricon, calling the Section 202 standard “expansive,” had no difficulty in concluding that where the entire contracted product volume was to be delivered abroad, the award was a Convention award.

But suppose the contract had provided for a series of deliveries, most of which were to US ports? Suppose further, or alternatively, that the dispute arose over the deliveries scheduled for US rather than foreign ports? Perhaps the most critical words in Section 202 are “reasonable relation with one or more foreign states.” It is doubtful that Congress intended for the Convention’s coverage to extend to an entirely domestic dispute merely because a portion of the contract’s scope involves performance or property abroad. Nor is there any reason of arbitration policy why coverage should be so extended. If a contract involves a series of instances of performance, some of which have will occur abroad, it makes sense to look at the performance giving rise to the dispute as the relevant contract for purposes of the Convention’s coverage. (See,  e.g., Amato v. KPMG LLP, 433 F. Supp.2d 460 (M.D. Pa. 2006),  in which US investors contracts with a US affilate of Deutsche Bank to provide a. Certain iinvestment strategy, but execution of the strategy on each transaction at issue in the case involved significant elements of foreign performance including the purchase of foreign securities and a swap transaction with a foreign taxpayer).

The arbitration rules of America’s Financial Institutions Regulatory Authority (FINRA) present a special, and sui generis, but economically significant, species of arbitrability controversies. FINRA, as successor to the National Association of Securities Dealers (NASD) is the self-regulatory organization of the securities industry. Established pursuant to the federal securities laws, FINRA rules provide that member firms agree to resolve certain disputes by arbitration under FINRA’s arbitration rules.

Last week the US Second Circuit Court of Appeals ruled against the US arm of the Swiss investment banking giant UBS on a FINRA arbitrability issue, made complicated by the fact that the claims related to UBS’s role as a market maker in the market for Auction Rate Securities (ARS). (UBS Financial Servs. v. West Virginia University Hospitals, Inc., 2011 WL  4389991  (2d  Cir. Sept. 22, 2011)). The ARS market largely collapsed in 2008, forcing issuers of ARS debt obligations to pay penalty interest rates and to incur other unepxected costs. In this case, UBS had acted as underwriter in the initial placement of municipal bond issues structured as ARS for a West Virginia state university hospital, and, after the initial issuance, conducted the “Dutch auctions” in which the ARS were traded and their interest rates were reset according to market conditions.

FINRA’s arbitration rules provide that they may be invoked either where there is an agreement to arbitrate under those rules, or where the claimant is a customer of a FINRA Member Firm and the claim concerns business transactions of the Member Firm. Here UBS argued, inter alia, that even if Claimant was a “customer” for UBS’s  services as manager of the Dutch auctions (which itself was not conceded), the claims did not relate to the auctions but to the alleged nondisclosure of UBS’s role in the actions to induce the customer to issue ARS and engage UBS as underwriter for the initial issuance.

This argument, that FINRA arbitrability depends on a nexus between the Member-customer relationship and the transactions giving rise to the claim, found favor with the dissenting judge on the panel, who considered that such a nexus requirement is implicit in the FINRA scope of arbitration rule, and that the nexus was missing here because the fraud allegations related to the original issuance and underwriting as to which there was no customer relationship (UBS having purchased the ARS outright as part of the underwriting).

The majority, however, while equivocal as to whether the text of the rule compelled the conclusion that such a nexus must exist, concluded that the initial underwriting and the subsequent auctions were so closely related that any nexus requirement that might exist was fully satisfied.

One notable aspect of the case is simply the fact that UBS as a Member Firm of FINRA concluded as a business matter that it preferred litigation to arbitration before a panel, selected by the industry self-regulatory authority, that would normally include one or more members with considerable securities industry experience at a Member Firm. This may well be a function of recent large FINRA arbitration awards against Member Firms,  most notably the $400+ million award to ST Microelectronics against Credit Suisse.

Also notable is that the Second Circuit analyzed the arbitrability issue without reference to the pro-arbitration bias of federal arbitration jurisprudence. The reasons for this are understandable and sound. The Member Firm of FINRA, like a Member State of ICSID, makes a conditional offer to arbitrate, which may be accepted if the putative claimant meets certain conditions. In the FINRA Rule, those conditions are, at least, “customer” status of the claimant, and that the claim relates to business transactions of the Member Firm. Those two criteria must be satisfied before the Court may find that any agreement to arbitrate exists. In this context, arbitrability is simply a matter of contract law without an overlay of federal pro-arbitration presumptions.

In an American variation on the Dallah v. Pakistan fact pattern (an award enforcement proceeding against a foreign state that did not sign the arbitration agreement), the US Third Circuit Court of Appeals has affirmed a District Court order vacating a sister court’s earlier confirmation of a reinsurance arbitration award rendered against a non-signatory Brazilian state entity. (Aurum Asset Managers, LLC v. Bradesco Companhia de Seguros, 2011 WL 3562897, 3d Cir. Aug. 15, 2011)).

But unlike in Dallah, the non-signatory question faced by the District Court in the confirmation proceedings in Aurum involved the court’s ability to exercise the subject-matter jurisdiction over a New York Convention award ostensibly conferred by Chapter Two of the Federal Arbitration Act (FAA). If the respondent state-owned Brazilian financial services company was not a party to the arbitration agreement, then the District Court, while having ostensible award-enforcement jurisdiction under Chapter Two of the FAA, either had no such jurisdiction because jurisdiction was a function not only of the FAA but also the Foreign Sovereign Immunities Act (FSIA), or at least the court could not exercise the jurisdiction conferred by the FAA because the Brazilian state entity enjoyed sovereign immunity.

On a technical level, Aurum involved the legal standard for a District Court to grant relief setting aside its own prior judgment confirming an arbitration award, where the foreign state entity had preserved its objections to the jurisdiction of the arbitral tribunal and the District Court by declining to appear in either proceeding. The Third Circuit rejected claimant’s position that the original District Court order confirming the award should stand unless there was a “clear usurpation of power,” reasoning that such a narrow standard for vacating the award-confirming judgment was unjust to a foreign state entity that had neither waived sovereign immunity nor had a full and fair opportunity to litigate that issue, but instead had exercised its right, conferred by U.S. law, to resist jurisdiction by (i) not appearing in the proceedings,  (ii) permitting judgment to be entered in absentia, and (iii) later attacking the judgment in a collateral proceeding. In such a case, the Third Circuit held in Aurum, the District Court in the collateral proceeding properly reconsidered de novo whether the District Court in the original award confirmation proceeding had subject-matter jurisdiction.

Curiously missing from the Third Circuit’s decision is any mention of the New York Convention or FAA Chapter Two.  Viewed exclusively from a Convention/FAA perspective, the Court had subject-matter jurisdiction, as there was an arbitration award arising out of a trans-national commercial relationship (FAA Section 202), and therefore there was a proceeding to confirm that award was “deemed to arise under the laws and treaties of the United States” over which the District Courts “shall have original jurisdiction” (FAA Section 203).

However, Chapter Two of the FAA does not take explicit account of the circumstance that a party to the arbitration agreement or the award is a foreign state. Section 1330 of the Federal Judiciary Code (28 USC), the jurisdiction-conferring provision of the FSIA, provides that District Courts “shall have original jurisdiction” of all nonjury civil actions against foreign states where — and only where — “the foreign state does not enjoy immunity under the FSIA.”

Was the Third Circuit correct to cite lack of subject-matter jurisdiction as the proper basis for the District Court’s order vacating the original confirmation order? Or did the Brazilian entity merely enjoy sovereign immunity from the District Court’s exercise of the jurisdiction conferred by the FAA in Chapter Two? The answer lies in the US Supreme Court’s interpretation of 28 USC §1330 in the Verlinden case in 1983. (Verlinden B.V. v. Central Bank of Nigeria, 461 U.S. 480). In Verlinden, the Supreme Court held that “[a]t the threshold of every action in a District Court against a foreign state… the court must satisfy itself that one of the [FSIA] exceptions [to sovereign immunity] applies,” and went further to say that despite legislative history calling sovereign immunity “an affirmative defense that must be specifically pleaded,” under the FSIA “subject matter jurisdiction turns on the existence of an exception to foreign sovereign immunity. 28 USC § 1330(a). Accordingly, even if the foreign state does not enter an appearance to assert an immunity defense, a District Court still must determine that immunity is unavailable under the Act.“

It seems useful, therefore, to take note that FAA Chapter Two, as a jurisdiction-conferring statute, contains this flaw. It is ineffective to confer on federal courts jurisdiction to confirm a Convention award against a foreign state unless there is subject-matter jurisdiction over the foreign state based on an exception to sovereign immunity under the FSIA. When there has been no waiver and the foreign state was not involved in commercial activity in the United States, the District Court must analyze whether the so-called “arbitration exception” to sovereign immunity applies, i.e. whether the award “is or may be governed by a treaty or other international agreement in force for the United States calling for the recognition and enforcement of arbitral awards.” (FSIA, Section 1605(a)(6)).  In that event, the arbitral tribunal’s determination that it had jurisdiction over the foreign state is entitled to no deference, as the District Court to have jurisdiction to confirm the award must determine de novo whether the foreign state was a party to the arbitration agreement underlying the award. Whereas the FSIA’s arbitration exception to immunity in Section 1605(a)(6) applies to awards resulting from arbitration agreements “made by the foreign state,” it is an open and difficult question whether arbitral jurisdiction over the foreign state based on principles extending the duty to arbitrate to non-signatories is sufficient meet the FSIA’s requirements for federal court jurisdiction in a proceeding to confirm an arbitral award.

The arbitration community should derive much satisfaction from reading a recent thorough and well-annotated application of the New York Convention and the Foreign Sovereign Immunities Act to confirm an arbitration award against a foreign state, authored by an experienced and well-respected U.S. District Court Judge in New York.  Thai-Lao (Thailand) Lignite Coal Ltd. v. Government of Lao People’s Democratic Republic, 2011 WL 3516154 (S.D.N.Y. August 3, 2011)).  I will find one portion of the decision against which to register a complaint: the possibility of applying the doctrine of forum non conveniens to deny consideration of a petition to confirm a Convention award. But this quarrel is not so much with the Court’s decision as it is with the Second Circuit precedent that the Court was constrained to respect. (Arbitration Commentaries has taken issue with the Second Circuit’s position before. See “Denial of Award Enforcement Under Article III ‘Rules of Procedure’: An Expanded Commentary on Zeevi Holdings v. Republic of Bulgaria,” April 26, 2011).

First, a brief summary of the Thai-Lao decision.   Two private companies in a joint mining venture, one Thai and the other Laotian, arbitrated claims against the Government of Laos under a contract that provided for arbitration under the UNCITRAL Rules in Kuala Lumpur, Malaysia. An all-American arbitral tribunal — two members from New York and the third from the Washington office of a New York firm — made its award at Kuala Lumpur (after merits hearings held there), under New York law as the contract provided, in favor of Claimants for a sum in excess of $40 million.  Claimants petitioned to confirm the award in the Southern District of New York, and the Court did indeed confirm the award after finding:

(1)             that it had personal jurisdiction over the Government of Laos, as Laos had broadly waived sovereign immunity from jurisdiction, attachment, and execution, in the contract; that the FSIA in any event provided an express exception to immunity for proceedings to enforce a Convention award; and that foreign states under Second Circuit law do not enjoy rights under the US Constitution’s due process clause to assert that requiring them to defend an action in a US court violates due process;

(2)             that while the doctrine of forum non conveniens is, under Second Circuit law, a New York Convention Article III “Rule of Procedure” of the forum  whose application could potentially result in the Court declining to entertain to a Convention award-confirmation summary proceeding, the principles governing application of that doctrine did not justify declining to the hear this case; and

(3)             the arbitral tribunal’s decision that the dispute was within the scope of the arbitration agreement made among the parties was an “arbitrability” determination entitled to judicial deference, because a contract calling for arbitration under the UNCITRAL Rules – rules that confer on arbitral tribunals competence to decide whether they have jurisdiction — provides “clear and unmistakable evidence” of an agreement to arbitrate arbitrability, an agreement that under US arbitration law reverses the presumption that arbitrability issues should be decided by courts, either in the first instance or with little deference to an prior view expressed by the arbitral tribunal.

All of these holdings are unquestionably right. But in the application of the doctrine of forum non conveniens, or I should rather say in the analysis leading to the non-application of that doctrine, the District Court assumed that the doctrine of forum non conveniens has the same content in a New York Convention award confirmation summary proceedings as it does when it is raised as an objection to the Court entertaining a case on the merits. Although that view may be implicit in the Second Circuit’s position that the doctrine does apply in Convention cases, there is no necessary reason why a prudential doctrine mixing considerations of comity, competence, and conservation of judicial resources, should be applied uniformly to ordinary litigations and proceedings with the singular characteristics of award-confirmations under the New York Convention.

Petitioners in Thai-Lao – claimants and award-winners in the arbitration — stated that the reason for seeking confirmation of the award in a US court was that the Respondent, the Government of Laos, was believed to have assets in the US that could be applied to satisfy a judgment upon the award. As that is foremost among the reasons that the Convention and FAA Chapter 2 permit an award to be enforced in a US court without regard to the foreign nationality of all of the parties, it seems odd that any further justification should need to be offered to oppose a forum non conveniens motion lodged against a petition to confirm a Convention award.

The Court understood its mandate under Second Circuit law to be that it should apply essentially the same forum non conveniens criteria that have been developed by US courts, mainly in domestic cases, to decide whether a particular US court rather than some other US or foreign court should be the exclusive forum to hear the merits and make findings of fact and conclusions of law. But no such exclusive role is contemplated by Convention confirmation proceedings. The Convention contemplates that recognition and enforcement will be sought in multiple for a – a point noted by the Court. And the very limited and defined grounds on which the courts may refuse recognition and enforcement under the Convention should rarely entail the kind of elaborate evidentiary proceedings whose presumed burdens on courts and witnesses who must travel to the forum justify a resource-conserving doctrine like forum non-conveniens.

Would it not therefore be more appropriate, for so long as forum non conveniens remains relevant in Convention cases, for a court simply to address whether the proceeding has been brought mainly for an award enforcement purpose and not as a device to gain some other tactical advantage from having a proceeding in a US court? If the proceedings  has been brought for that purpose, then there should be a strong presumption in favor of the court entertaining the case.

And then what factors might serve to rebut this presumption? The standard forum non conveniens criteria seem ill-suited. “Forum shopping” into a jurisdiction whose courts apply the Convention in a pro-enforcement fashion is consistent with the Convention’s purposes. The “convenience of parties and witnesses” is not a good fit with a “summary proceeding” (a motion, on papers) that should rarely require either discovery or a live testimonial appearance by a witness. And the presence or absence of connective factors other than the losing party’s assets should not matter very much if there is a good faith belief that assets may be found in the jurisdiction of the court. It should not have been necessary for this Court to note, as it did, that two of the arbitrators were New Yorkers, that a procedural hearing was held in New York, that New York law governed, that the award discussed and cited New York law extensively, that the award was in English, and that the damages were stated in US dollars. Had these factors been absent, while assets of the loser were believed to be present, there should have been no less reason for the US court to hear the case.

Nor does it make very much sense to consider the existence of an “adequate alternative forum” where – unlike litigation on the merits – an inherent aspect of the legal regime of the New York Convention is that the same proceeding will be undertaken concurrently in several jurisdictions. Rather the forum non conveniens doctrine as applied in Convention cases (until the Second Circuit reconsiders or the US Supreme Court considers whether it should be applied) should address whether a particular foreign forum is specially suited, or, stated differently, whether the US court is specially unsuited, to address the legal, factual, or public policy  issues raised.

One can imagine scenarios in which a Convention defense to recognition and enforcement would depend upon a de novo application of foreign law – e.g., the Dallah v Pakistan scenario, where it might have been quite reasonable for the U.K. court to have refrained from deciding, in the first instance, whether Pakistan was a proper party to the arbitration under French arbitration law, thereby virtually requiring Dallah to seek enforcement of the award in France.  But if the issues presented in an award-confirmation case are not clearly more suitable — substantively and in efficiency terms — for exclusive resolution by a foreign court, forum non conveniens dismissal of a confirmation petition where assets to satisfy the award are believed to be present in the US should be possible only in exceptional circumstances.

Institutional rules governing international commercial arbitration permit (and thereby to some extent encourage) arbitrators to render partial final awards when appropriate. And those rules, institutional guidelines, and published commentaries suggest or recommend (i) the structuring of complex arbitral proceedings by issue, and (ii) the early determination of some key issues whose resolution might advance the prospects of settlement on remaining contentious issues. Parties might assume that if an arbitral tribunal issues its final determination of a particular issue in the form of an award, in a fashion that complies with applicable arbitral rules about the form and content of awards, and in terms that reflect the arbitrators’ understanding that they give up any power to adjudicate that issue further, then the award should have the same status as an award terminating the entire case, in regard to confirmation or vacatur by a federal district court under the Federal Arbitration Act (FAA). But US law on when partial awards may be presented for confirmation or vacatur – awards which are final in terms of the issues decided, but non-final in relation to the universe of issues submitted for determination — has not kept pace with arbitral practice.

If one takes at face value a decision last week from a federal district court in Connecticut (Pearl Seas Cruises, LLC v. Irving Shipbuilding, Inc., 2011 WL 3475469 (D. Conn. Aug. 9, 2011)), the law in the Second Circuit on FAA confirmation and vacatur of partial final awards is quite rigid. According to this court’s reading of Second Circuit law, an award that does not terminate the arbitration may be presented for confirmation or vacatur in only two narrow circumstances: (1) when the award finally decides a separate independent claim, and (2) where the parties jointly asked the tribunal in the course of the arbitration to bifurcate the issues of liability and damages on a claim and to enter a partial final award on liability.

If that is indeed all that Second Circuit law permits, a modernization of the law is in order, to bring judicial practice into harmony with arbitral practice. A reading of the relevant case law suggests that this modernizing process can be achieved as a logical progression from the existing principles.  It is not a problem of the law having taken an unsuitably narrow approach to partial final awards, but rather a case of the law having developed only to address the particular circumstances presented, without serious effort to anticipate the range of situations and issues on which arbitrators might render partial final awards.

A Preliminary Issue: Does the Stolt-Nielsen Case Imply That All or Most Partial Final Awards are FAA-Reviewable?

Some readers will surely wonder whether the law in this area was clarified by the Supreme Court’s decision last year in the Stolt-Nielsen case.  After all, the arbitral award reviewed by the District Court, the Second Circuit, and the Supreme Court was a “clause construction award,” a quintessentially “partial final” award that resolved only the issue of whether the arbitration clause in the parties’ contract permitted class proceedings in the arbitration. In Stolt-Nielsen, the AAA’s Rule 3 of its Supplementary Rules for Class Arbitrations provided in pertinent part that the arbitrator “shall stay all proceedings” for at least 30 days to permit any party to move a court of competent jurisdiction to confirm or vacate the Clause Construction Award. Thus the drafters of the AAA class arbitration rules evidently thought there was no non-finality obstacle to FAA review.

When the Stolt-Nielsen clause construction award reached the District Court, non-finality was not raised by any party or sua sponte by the Court. The Second Circuit was equally silent on the point. In the US Supreme Court, the majority and dissenting opinions sparred over whether the case was ripe for review in the Supreme Court. But the majority did not squarely address whether the FAA permitted judicial review in the District Court. Rather, the majority in a footnote addressed the constitutional doctrine of “ripeness,” and concluded that the hardship faced by the respondent, forced to endure “class determination proceedings,”  coupled with the fact that if respondent failed to proceed in the arbitration it would most likely face an FAA Section 4 petition to compel arbitration, satisfied the constitutional test for ripeness of a non-final adjudication. Justice Ginsburg’s dissent questioned “judicial intervention so early in the game,” (emphasis supplied) but had very little to say about what the FAA requires or permits, offering only a cursory review of the positions of federal courts of appeals on the question of confirmation or vacatur of awards resolving fewer than all submitted issues.

How then do we fit Stolt-Nielsen into the jurisprudence on reviewability of interim and partial awards? One’s view depends in part on whether the amenability of an award to FAA confirmation or vacatur is a question of “ripeness” or “justiciability” that a court may raise sua sponte, or whether the absence of an “award” as FAA common law defines that term means the petition may be dismissed for failure to state a claim. If the latter, then a party may waive the failure to state a claim defense. I believe the latter concept is more fitting. The absence of an FAA “award” bars a vacatur action under the FAA in the same way that the absence of a “security” bars an action under Section 10(b) of the Exchange Act — an essential legal element of the statutory cause of action is missing, and so there is a failure to state a claim under the statute creating a cause of action. That is a failure of legal substance.  Some courts use the term “non-justiciable” to refer to an arbitration award that is not final enough to be an FAA “award.” But “justiciability” as it relates to issues presented to courts for initial decision refers to a series of judicially-developed doctrines relating to the suitability of exercising judicial power – the case-or-controversy requirement, the political question doctrine, the Act of State doctrine. These are concerns the courts can and do raise sua sponte.  Whether an arbitral decision styled an “award” by the arbitral tribunal, and complying with the requirements for form and content of an award prescribed by the rules or laws under which the arbitration took place, is also an “award” to which the FAA applies is strictly a question of statutory construction. If the petitioner seeks confirmation or vacatur and claims the arbitral decision involved is an FAA award, and the respondent does not disagree, the non-finality issue is waived, and the court must address the merits provided it has jurisdiction. This, I submit, is the best way to explain FAA review in Stolt-Nielsen: the non-finality issue was waived in the District Court and the Second Circuit. And if that view is taken, Stolt-Nielsen offers very little assistance in a case where the alleged interlocutory character of an award is raised by a party as a statutory bar to confirmation or vacatur.

Indeed, as the federal appellate cases mentioned in the Stolt-Nielsen take various positions on when an arbitrator’s partial award is an FAA award, it is useful to consider on what basis the courts have decided that question, and whether their approaches are justified as an exercise in statutory construction. I will take the Second Circuit as a case study.

Second Circuit Law of FAA Review of Partial Final Awards

Taking as the first of the “modern” Second Circuit cases on the issue Michaels v. Mariforum Shipping S.A., 624 F.2d 411 (2d Cir. 1980), we find that the Court there held that an FAA motion to vacate an interim award on liability, where damages issues had been reserved, was “premature.” The Court in Michaels stated that under the FAA “a district court does not have the power to review an interlocutory ruling by an arbitration panel.” But the 1957 Second Circuit case cited in support of this proposition involved proposed review of evidence rulings in an ongoing arbitration, scarcely a solid basis to conclude that the statute bars a petition to vacate a final determination of liability issues. But the Michaels court also rested its decision on “policy considerations, no less than the language of the Act and precedent construing it.” The Court cited the potential inefficiency in the arbitral process that could result from piecemeal review. But whereas Michaels holds that such policy considerations are relevant to whether a given award is eligible for confirmation or vacatur, the decision implies that there could be some situations where pro-arbitration policy considerations would weigh in favor of FAA review of an interim or partial award.

A significant advance in the Second Circuit’s thinking about partial final awards took place in Trade & Transport, Inc. v. National Petroleum Charterers, Inc., 933 F.2d 191 (2d Cir. 1991). The question in Trade & Transport was not finality for purposes of confirmation or vacatur strictly speaking, but finality sufficient to avoid re-hearing of the issues decided in a partial final award when one member of the tribunal thereafter died and had to be replaced. The Court reasoned that whereas the parties had specifically stipulated that the tribunal should enter a partial final award finally deciding some but not all of the issues submitted to arbitration, the tribunal upon rendering such award was functus officio on the submitted issues, and therefore the proceedings on those issues did not need to be repeated before the re-constituted tribunal.

Several years later in Rocket Jewelry Box, Inc. v. Noble Gift Packaging, Inc., 157 F.3d 174 (2d Cir. 1998), the Court cited Trade & Transport for the proposition that “an arbitration award, to be final, must resolve all the issues submitted to arbitration, and … it must resolve them definitively enough so that the rights and obligations of the two parties, with respect to the issues submitted, do not stand in need of further adjudication.” (emphasis in original). Rocket Jewelry could possibly be read as a step backwards – because Trade & Transport clearly involved an interim award that left unresolved many issues the parties had “submitted” in the sense that they were presented in the arbitral pleadings and proceedings for eventual arbitral determination.  And whereas the argument made against finality of the award in Rocket Jewelry was that the parties had unresolved disputes that they had excluded entirely from arbitration and that had to be resolved in litigation, the Second Circuit had no occasion to delve more deeply into the question of when an award may be final even though issues remain for arbitral determination. But Rocket Jewelry did confirm Trade & Transport as good law, and so the term “submitted” as used by the Court is best read as a reference to the issues the parties have asked the tribunal to determine in an award, whether that requested award terminates the arbitration or leaves issues to be resolved in a further award.

The Second Circuit did not explain in Rocket Jewelry, and in case law since then has not explained, what elements of the procedure that led to the partial final award in Trade & Transport caused that award to be seen, through the prism of subsequent decisions, as an FAA-reviewable award. As a result, cases such as last week’s decision in the District Court in Connecticut cite Trade & Transport for the proposition that when the parties by stipulation have explicitly bifurcated liability and damages, the resulting award on liability is FAA-reviewable, as an “exception” to the “general rule” that FAA-reviewable award is one that resolves all issues. This regrettably elevates a description of what occurred in Trade & Transport into a purported rule of law. But neither the Trade & Transport decision nor the Rocket Jewelry decision suggests that only this precise procedural pattern will qualify for FAA review an award that finally resolves fewer than all issues. What if the “bifurcated” (or separated) issue were a jurisdiction objection, or liability and damages on one of several causes of action, or the cause of action against one respondent but not the others? What if the decision to issue a partial final award had been made by the tribunal, and had been announced in advance of the proceedings leading to it, over the objection of one party that urged the tribunal to decide all issues together in a single case-terminating award? Is that situation outside the “exception” of Trade & Transport because there was no stipulation of the parties?

On a careful reading of Trade & Transport, it is certainly clear that there is no subject matter limitation, i.e. a liability determination with damages issues reserved.  The rationale of the Trade & Transport decision was the parties’ agreement that the tribunal should issue a partial final award vested the tribunal with the authority to do so.  On this basis, district courts in the Second Circuit should have no hesitation to conclude that partial final awards addressing jurisdiction issues, or claims against fewer than all parties, or critical legal or factual threshold issues germane to a cause of action, may be the subjects of a partial final award that is immediately eligible for FAA confirmation or vacatur.

The question remains whether an interlocutory stipulation that there should be a partial final award is a necessary element. I suggest that this is not the case. The Second Circuit in Trade & Transport looked to the joint application for a partial award that was made in that case as the source of the tribunal’s authority to issue the partial final award. What was critical to the Court was that the partial final award was authorized by the parties, not that the authorization had to take the form of a joint request in the course of the proceedings.  If I am correct in that regard, then whenever the parties have agreed to arbitrate under rules or arbitration laws that authorize tribunals to issue partial final awards, the tribunal’s authority to issue the partial final award is established — at least provided the parties have been given an opportunity to be heard on whether there should be a partial final award and were on notice that a certain phase of the proceedings would culminate in a partial final award on particular issues or claims.

One often finds the statement in decisions concerning FAA reviewability of a partial award that it is the content of the arbitral decision, not the label “award” or even “partial final award,” that determines whether judicial action under the FAA may be sought. The matter is not quite so black-and-white, as the arbitrators’ intentions may well be relevant. The content-trumps-label principle is sound insofar as it relates to the functus officio doctrine: if the arbitrators have expressly left open the possibility to modify an interim ruling, then they cannot trigger FAA review by mis-identifying their decision a “partial final award.” Equally, a tribunal cannot trigger FAA review of a purely procedural decision by calling the ruling, for example, a “partial final award on admissibility of evidence.” (See, e.g., Accenture LLP v. Spreng, 2011 WL 2090825 (2d Cir. May 27, 2011), holding that an arbitral order denying leave to amend the request for arbitration to add a fraud claim on the eve of the merits hearing, which did not prevent claimant from making that claim in a separate case, was not an “award” eligible for vacatur under the FAA). But where an interim award resolves in what appears to be a final way an issue of jurisdiction or the merits, then the tribunal’s declaration that it is a “partial final award” is a useful clue to the tribunal’s intentions and also should invite the district court to satisfy itself that the tribunal had authority, by rule, by law, or by agreement of the parties, to issue a partial final award.

A separate strand of the Second Circuit’s award finality jurisprudence holds that “an award which finally and definitely disposes of a separate independent claim may be confirmed although it does not dispose of all the claims that were submitted to arbitration” A leading case is Metallgesellschaft A.G. v. M/V Capitan Constante, 790 F.2d 280 (2d Cir. 1986), where the Court affirmed confirmation of a partial final award for maritime freight charges that left still to be resolved certain issues of counterclaim and setoff.

The Second Circuit has had little occasion to shed more light on the “separate independent claim” formula.  In Zeiler v. Deitsch, 500 F.3d 157 (2d Cir. 2007), the Court affirmed confirmation of a series of orders that required one joint owner of property to render accountings to the other and to transfer certain documents. The Court observed that these “accounting orders” were final awards for FAA purposes in part because they “require[d] specific action and [did] not serve as a preparation or a basis for further decisions by the arbitrators.” In a footnote, the Court distinguished the Michaels case, and stated that “the accounting orders in the pending case are not segments of a future conclusive award, nor are they determinations required for furtherance of arbitration.”  But the Court went on to say that the “confirmable nature” of these accounting orders “stem[med] from the unique character of the arbitration,” whereby the parties had agreed to have the arbitrators “preside over the continuing process of sorting out the details of a commercial relationship.” Thus Zeiler may fairly be said to belong in the Trade & Transport strand of the jurisprudence which focuses not on the nature of what the tribunal decided in the partial final award but upon the authority conferred on the tribunal to decide in a final way at a non-final stage a subset of the issues submitted in the arbitration.

This state of the law in the Second Circuit leaves uncertain the position the Court would adopt concerning partial final awards that resolve questions of arbitral jurisdiction over a claim or a party.   An objection lodged against a claim based on alleged non-arbitrability is not a “separate independent claim” but rather it is a defense to a claim positing that the tribunal lacks power to resolve the claim.  If the “separate independent claim” requirement is literally applied, then a partial award on jurisdiction that does not qualify for FAA review based on the Trade & Transport vesting-of-authority principle would possibly fail to qualify for FAA confirmation or vacatur.   Perhaps the issue will be treated by courts in the Second Circuit as a question of first impression to be decided based upon essential principles of FAA arbitrability law: especially that parties should not be required to arbitrate if they have not agreed to do so, and if they have agreed to arbitrate they should only be required to arbitrate those disputes they have elected to so submit.

That principle suggests that partial final awards on jurisdiction should be FAA-eligible for confirmation or vacatur at the time they are made. That is not to say that in every such instance the district should immediately decide the petition for confirmation or vacatur. There could be prudential reasons to await further arbitral proceedings. But equally there could be compelling reasons to decide a jurisdiction issue before the tribunal proceeds much further, as for example when the arbitrators’ jurisdiction decision appears prima facie to be contrary to settled arbitrability law and the costs to the parties that would result if the correction is made only after the arbitration is concluded would be unacceptably large.  This approach would apply even to situations where the parties have agreed to “arbitrate arbitrability,” as the agreement of the parties to empower the arbitrators to resolve jurisdiction issues means only that the arbitrators’ jurisdiction decision is reviewed judicially as an award under the FAA and not de novo. The agreement to “arbitrate arbitrability” implies nothing about the timing of FAA review of the arbitrators’ arbitrability decision.  

*****

It is trite to say that arbitrations are growing more complex, with more parties, more claims and defenses, and more issues of liability and damages. For better or worse, the FAA scheme for confirmation and vacatur dates from an earlier era when “award” had a universally understood meaning as the single decision by which an arbitrator would decide the entire dispute. But the courts appear to have accepted that the FAA concept of an “award” must be fluid and should keep pace with the parties’ legitimate expectations about how arbitration will serve their needs. It is to be hoped that future court decisions will incrementally expand the universe of FAA-eligible partial final awards in accordance with this view.