No I did not retire. And I didn’t have long Covid, or any Covid. I just had stuff to do. Mostly stuff I can’t tell you about. Confidentiality, tender egos, taciturn Tribunals, all that. It’s better to be quiet. But seriously, readers, if you really like this Blog and crave regular Blog posts with your morning coffee, just stop sending me arbitral appointments, or stop making motions in my pending cases, and then I will have enough time to do what I am meant to do, which is to entertain you. To make you smile. Which we all need to do. And yes, thanks for asking, I WILL do the 560-mile Empire State Ride again this year. (July 24-30, no motions please). Last year I got two new appointments in the week of the ride. So what’s not to like? You can donate to cancer research at Roswell Park Cancer Center in Buffalo (Go Bills!) by supporting my ride at www.empirestateride.com, or my 100-mile warm-up ride on June 25 at www.RideforRoswell.com. Now, back to the entertainment!

Some people just won’t admit that they read this Blog. The judges of the Ninth Circuit, for example.
When last read by many of you, in February 2021 (sorry for the long silence; I’ve had to work), your Commentator reported on a decision in the US Ninth Circuit Court of Appeals in which the panel majority, despite a persuasive dissent, held that the law governing the arbitrability of US federal trademark claims, when contested in a US court whose jurisdiction is based on those federal claims, and contested between a non-signatory and signatory, each of Indian nationality, of an arbitration agreement made in India, concerning a family-owned Mumbai-based incense-selling business, is … US federal common law. (See the Post on this site “Nonsignatories Unmasked,” dated February 11, 2021, discussing Setty v. Shrinivas Sugandhalaya LLP, 2020 WL 192820 (9th Cir. Jan. 20, 2021), since then officially reported at 986 F.3d 1139, and withdrawn, at 998 F.3d 897 (9th Cir. June 4, 2021)).
I sided with the dissenting member of the Ninth Circuit panel, who would have held that the law applicable to such an arbitrability issue is the ordinary contract law of the relevant state (whether a US state or a foreign state), as determined by applying the forum’s choice-of-law rules. And I predicted, in that February 2021 Commentary. that an application for review in the US Supreme Court might ensue.
That is not quite what has happened, but it appears that a sensible outcome for US international arbitration law has emerged, by another procedural route. The disappointed proponent of arbitration in India, in the Setty case, submitted a petition for rehearing en banc in the Ninth Circuit. (Non-US readers, that is a request for review by an enlarged appellate panel, and in the populous and raucous Ninth Circuit it meant that 11 appellate judges, from the Court’s full roster of 29, would sit on the re-hearing). But after the re-hearing briefs were in, and before the 11 made a ruling, a funny thing happened on the way to the Courthouse: the majority in original panel of three judges withdrew its original opinion, in an order dated June 4, 2021 that instructed that the withdrawn opinion “may not be cited by or to this court or any district court of the Ninth Circuit.” (No injunction there against bloggers on the US East Coast). “A new disposition will be filed in due course. … [T]he petition for rehearing en banc is DENIED as moot.” Hmm.
Granted that the Ninth Circuit spans the full length of the coffee-centric US West Coast, from Canada to Mexico, and that this case came up on appeal from the US District Court for the Headquarters of Starbucks, but still, a bit of TEA LEAF reading is in order: It looks as if one or both judges in the original Setty panel majority will now join the dissenter, and hold that arbitrability under a contract made in India, as between competing factions of a family in the incense business in Mumbai and Bangalore, is to be determined by a US district court under the contract law of India provided that is the contract law applicable to the arbitration agreement, as indicated by applying the usual choice-of-law rules of the forum. And it looks as if the panel will hold, or at least strongly indicate, that this approach prevails in any procedural context in which enforcement of an arbitration agreement to which the New York Convention applies is presented to a US district court – that is to say, not only in a Setty-like case, where the plaintiff has invoked federal court jurisdiction by suing under the federal trademark laws, but also (1) if an action were brought in federal court based on diversity of citizenship, asserting only violations of state law, (2) if the action were a free-standing petition to compel arbitration under the Federal Arbitration Act, and (3) in a proceeding to recognize and enforce, or to vacate, an arbitral award governed by the New York Convention.
We do care about this. Because here in the US our Supreme Court just one year ago settled, once and for all, that when international arbitration is sought to be compelled in a US District Court, by or against a non-signatory, principles of equitable estoppel available under the applicable state contract law are not made inoperative by the “agreement in writing” requirement of the New York Convention (GE Energy Power v. Outokumpu, 140 S.Ct. 1637). So now, in nonsignatory arbitrability cases that fall under the Convention, the question “What law is the applicable ‘state’ contract law?” comes to the forefront.
The Ninth Circuit panel majority in Setty initially thought that the answer was “federal common law” because the underlying claim was a federal statutory claim under US trademark law and that statutory claim was the plaintiff’s admission ticket to a US district court. But the precedent it relied on for that choice was a case decided in the context of a domestic federal securities law dispute, and in that case the choice of law question was a species of what American lawyers call an “Erie question” (never mind why): that is, a question whose answer is tied up with the tug-of-war for legal hegemony between the states of our Imperfect Union and E Pluribus Unum itself. In that older case (OK insomniacs, it’s Letizia v. Prudential Bache Securities, 802 F.2d 1185 (9th Cir. 1986)) it made sense that arbitrability of federal statutory claims should not potentially vary from one US district court to another based on variations in the common law of contracts from one state to another. In that context, a choice of “federal common law” was a specialized choice of law rule based on special considerations of federal supremacy and US internal legal consistency.
But such internal federalism considerations have little resonance when the choice is between the contract law of India and federal common law of contracts. The risk calculus is different. The court need not worry about whether American businesses might be subject to different arbitrability outcomes on claims governed by federal law, depending on the state in which a US district court was asked to compel arbitration. Instead, the concern should be that federal common law rules might be influenced by concerns that are not relevant to litigants domiciled outside the US – like protecting a judicial forum for litigation of federal statutory claims, a not-quite-moribund legacy of the bad old days before ~ 1985, when Mitsubishi was just Mitsubishi, another cool foreign car.
That sort of influence appears to have been present in the Ninth Circuit’s legacy federal common law rule, from the Letizia case, about equitable estoppel as a basis for a nonsignatory to require a signatory to arbitrate: the rule was that, for equitable estoppel to be a basis for compelling arbitration, the claim proposed to be arbitrated had to be closely “intertwined” with the contract containing the arbitration clause (a rule that seems to be quite the mash-up of contract law and substantive federal arbitration law, and scarcely a rule of generally applicable contract law – as to this tension, see the Arthur Andersen case discussed in the February 11 Post). Finding no such “intertwining” in the Setty case, the panel majority had originally rejected the motion to compel arbitration.
Happily it seems that the prevailing view among the participating judges in the Ninth Circuit (the back-channel en banc ‘hearing’ that evidently motivated the panel’s change of heart) is that a “traditional” choice of law rule – taking into account relevant contacts and interests of India and the US in applying their contract law to the question – was the way to go. That seems to be a ‘win’ for international arbitration in the US, whether or not it is ultimately determined that, under the law of India, these Indian parties with competing incense businesses in Mumbai and Bangalore, each a derivative of the family business whose partnership agreement called for arbitration, are required to arbitrate. However it is decided, at least they (and other foreign parties in disputes over their US business activity) can take comfort that a US court thought seriously about the geographic and juridical context in which the arbitration agreement was made, and in which the parties’ respective connections to that arbitration agreement were forged.
Stand by for a replacement judgment from the original 9th Circuit panel, which presumably will remand the case to the US District Court from whence it came, in Seattle (home of Starbucks), with instructions to re-determine the law applicable to the arbitrability issue – US or India – and to apply the law so determined.

This Commentary begins, innocently, with a report of a new decision in the US Ninth Circuit Court of Appeals, about choice of law in the arbitrability realm. I realize you are watching lot of daytime TV these days. The Edge of Night, etc. So I will try to make it interesting. Settle in, remove your mask, stay awhile.

This story actually begins with the Outokumpu case. You remember Outokumpu, from last June. It is famous for a few reasons. One is that Justice Thomas wrote the opinion for a unanimous US Supreme Court. How often has that happened in 30 years? It’s also famous, in our world, because the Court held that the New York Convention doesn’t prevent a Court from compelling arbitration for or against a non-signatory based on equitable estoppel. What’s sauce for the FAA Chapter 1 domestic goose, held the Court, is also sauce for the FAA Chapter 2 international gander. Birds of a feather. No conflict, no problem.

Well, maybe a technical problem. Because the 11th Circuit US Court of Appeals, from whence Outokumpu emerged, was not the only such US Court – the 9th Circuit was another — to have held that the New York Convention not only required the courts of Contracting States to enforce arbitration agreements in writing within the coverage of the Convention upon the application of one of the parties, but also that the Convention by implication prohibited such courts from enforcing such agreements upon the application of a non-party who invokes estoppel principles in support of enforcement. So, in the 9th Circuit case that is the text for today’s sermon, the Supreme Court granted certiorari, and sent the case back to the 9th Circuit to figure out what happens next in light of Outokumpu. The 9th Circuit heard new arguments, and then held that, in a federal district court, federal common law, not state law, controls the question of whether non-signatories can enforce arbitration agreements based on equitable estoppel when federal statutory claims are asserted in the district court in the underlying dispute. And so the 9th Circuit held that the District Court did not err, in the first go-round, in denying the motion to compel arbitration. One judge dissented, elaborately, on the question of what law governs the equitable estoppel/arbitrability issue. The position taken was that the choice of law for the non-signatory arbitrability issue ought probably to be the law of India, and that a remand to the District Court to re-decide the motion to compel arbitration after considering anew the choice-of-law issue should have been the outcome. So here we are, with material for a Commentary. Setty v. Shrinivas Sugandhalaya LLP, 2020 WL 192820 (9th Cir. Jan. 20, 2021).

Where I would like to take you in this Commentary eventually, dear Readers, is to the question of what a US-seated arbitrator should do if confronted with a jurisdiction objection raising the equitable estoppel issue, in an international case between parties of non-US nationalities. That is a particular challenge because in the USA, maybe more than in other beehives of international arbitration activity, because the US jurisprudence on arbitrability mostly concerns how judges should act or refrain from acting. Translating that into guidance for arbitrators, for whom US arbitration law is lex arbitri, can be a major challenge.

So let’s first sort out the different views of the Panel majority and the dissent in the Setty case. That should qualify you to sit on the District Court in some reliably sunny place, say Phoenix or LA or San Diego. Even Seattle, if you like coffee and rain. But since you are, or aspire eventually to be, an international arbitrator sitting in such places from time to time, let’s eventually try to sort out what a case like Setty means for you.

Setty is about a family business in India selling incense. Due to a family feud, one branch of the family ended up with its incense business in Mumbai, the other in Bangalore. The only question for the 9th Circuit, after the post-Outokumpu remand from the Supreme Court, was what law applies to the question of using equitable estoppel as a basis for a non-signatory to enforce an arbitration clause against a signatory, to require the signatory to arbitrate not litigate its claims against the non-signatory. Is it the law of India, or federal common law? The arbitration clause is found in a “partnership deed”, and the family members who run the Mumbai operation are not parties to the deed. They are parties to a federal district court lawsuit brought against them by the Bangalore business, claiming trademark infringement under US federal trademark laws. Even though the partnership deed evidently was made in India between partners who were of Indian nationality and provided for arbitration governed by the law of India, the Ninth Circuit swiftly dismissed the partnership deed as irrelevant to the choice of law issue, reasoning that, under 9th Circuit precedent, what it called a “threshold issue” of arbitrability arising under FAA Chapter 2 and the New York Convention is governed by federal common law. The Panel cited a 2016 9th Circuit case holding that when it is for a court to decide whether a binding agreement to arbitrate exists in an FAA Chapter Two case, i.e. whether an agreement to arbitrate falling under the New York Convention was formed, that issue is determined based on federal common law. So the first problem with the Setty Panel majority’s reasoning is that it relies on a precedent that involved whether any arbitration agreement ever was made, not whether a non-party could enforce such an agreement that indisputably was made.

As the next and essentially final step in its analysis, the Panel majority in Setty relied on a 1986 9th Circuit decision in which it was held – in a domestic arbitration context – that when a party to an arbitration agreement seeks to arbitrate US federal statutory claims (there, under the federal securities laws’ anti-fraud provisions) against non-signatories, federal common law rather than state law governs the arbitrability issue. Appellant in Setty had contended this conclusion was incorrect, invoking a US Supreme Court decision in 2009, Arthur Andersen LLP v. Carlisle, 556 U.S. 624, where the Court held in a domestic FAA case that the relevant state contract law determines whether a non-signatory of the arbitration agreement may require its enforcement. But per the 9th Circuit majority in Setty, prior 9th Circuit decisions had only applied Arthur Andersen as the choice of law rule in cases involving state substantive law claims that came before the federal district court based on diversity of citizenship jurisdiction, and (by implication) not in cases raising federal statutory substantive law questions and no in cases involving arbitration itself as a federal question under FAA Chapter Two. The affirmative limitation of Arthur Andersen to a non-federal claims and domestic arbitration agreements is, for the 9th Circuit, evidently a new development in Setty. And whereas the federal common law version of equitable estoppel for purposes of arbitrability requires that the claims be “intertwined” with the contract containing the arbitration agreement, and the District Court had already found that there was no such intertwining, the Panel re-affirmed its earlier judgment that sustained the District Court’s denial of the motion to compel arbitration.

So, you may ask, tell us more about Arthur Andersen. OK. Simplified, the Supreme Court in Arthur Andersen held that when a party to a federal lawsuit seeks a stay of the action under FAA Section 3, claiming the dispute is arbitrable, and the arbitrability issue is a non-signatories issue (i.e. who is bound by a valid arbitration agreement) that issue is determined according to the ordinary, generally applicable principles of state contract law that apply to contracts generally. The litigation in an Ohio federal court underlying Arthur Andersen did not involve federal statutory causes of action, and this was a domestic FAA case where diversity of citizenship provided the needed independent basis for federal subject-matter jurisdiction. But nothing in the Court’s opinion states expressly that a choice of federal arbitrability law would have been required had there been federal statutory claims. And nothing said expressly in Arthur Andersen limited its holding to FAA Section 3 motions (to stay litigation pending arbitration) affecting purely domestic arbitration agreements.

Not only your Commentator, but also the dissenting 9th Circuit panelist in Setty, thinks the Setty panel majority’s attempt to pigeon-hole Arthur Andersen to domestic arbitration of non-federal substantive claims was not justified. And not only your Commentator, but also the dissenting 9th Circuit panelist in Setty, believes the Supreme Court would transpose the notion of “state law” in Arthur Andersen in the context of a dispute between non-US parties to mean the law governing the enforceability of the arbitration agreement by or against a non-party as determined by the applicable choice-of-law principles in the forum where enforcement is sought. So, to conclude the first section of this Commentary, I would say to stay very tuned for a petition for writ of certiorari to the US Supreme Court in the Setty case. An elaboration of how the choice-of-law principle of Arthur Andersen applies in the setting of international arbitration and/or in the setting of underlying US federal statutory substantive law claims may attract the interest of the Court.

Ok so now let’s consider how all of this arcane US law on arbitrability affects you as an arbitrator if you must decide arbitrability. Suppose the facts of the Setty case (Mumbaians v. Bangaloreans, incensed), and that the partnership deed identifies no seat of arbitration but does provide for ICC arbitration. The parties make submissions to the ICC Court concerning the place of arbitration. The Claimant prefers Mumbai and the Respondents prefer Seattle (where the federal district court case leading to Setty was heard, after a transfer from Alabama). The ICC Court picks Seattle, and the Parties eventually, by agreement, pick YOU!

When you must rule on an objection to the jurisdiction of the Tribunal, with the objecting parties urging that the non-signatories cannot rely on equitable estopped principles to enforce the arbitration agreement, what choice of law rules shall you apply? The answer in an ICC arbitration may be less than clear. Article 21 of the ICC Rules states that with respect to the law applicable to the merits of the dispute, in the absence of agreement of the parties the Tribunal shall apply the laws or rules of law that it considers to be appropriate. But is the choice of law rule at issue applicable to the merits? It seems that the issue of jurisdiction is preliminary to the merits, and the ICC Rules contain no express reference to law applicable to the issue of arbitral jurisdiction. If your case were under the ICDR Rules, on the other hand, Article 31.1 provides, without limitation only to “the merits,” that the Tribunal in the absence of party agreement “shall apply such law(s) or rules of law as it determines to be appropriate.” Under the literal text of ICDR Art. 31.1, if you reject the Panel majority approach in Setty and instead apply a traditional contacts-and-connections based choice of law rule, your choice of law rule is protected by the agreement of the parties to arbitrate the jurisdiction issue under the ICDR Rules, and the Federal Arbitration Act protects the agreement, so the courts ought not second guess you.

With only slightly less confidence, I believe the same outcome is achieved under the ICC Rules. Article 42 of the ICC Rules provides that “[i]n all matters not expressly provided for in the Rules, the Court and the arbitral tribunal shall act in the spirit of the Rules….” And of course the empowerment of the arbitrator to decide upon her own jurisdiction is embedded in those Rules. Your ability to select a choice of law rule, in relation to the extension of the arbitration agreement to non-signatories, seems to be protected by the agreement to arbitrate jurisdiction under the ICC Rules, and in turn by the FAA whose core principle is to enforce the agreement. Add to this the observation – which I make from recollection and not from a systematic review of Supreme Court decisions got this Commentary, but I believe it to be correct – that the Supreme Court’s interpretations of the requirements of the FAA pertaining to arbitrability – apart from allocation of power between courts and arbitrators — have concerned how courts shall address arbitrability when it is required to be decided by a court. That is to say, the US jurisprudence on arbitrability is not mainly concerned, perhaps concerned almost not at all, with what rules of law the arbitrator decides to apply to arbitrability issues, unless the arbitration agreement imposes rules she must apply and it is claimed that she failed to apply them.

If this characterization of US arbitration law is accurate, then unlike many leading arbitral jurisdictions the United States really does not have an American arbitration law approach to the law applicable to issues of arbitrability that binds arbitrators. When you instinctively apply New York law because the contract provides for a New York seat and New York substantive law, be aware that you are not enforcing an implied choice of law on arbitrability. Instead you are adopting one of several perfectly supportable approaches to the choice of law issue. And so it is a useful bank of knowledge for the US-seated international arbitrator, or the arbitrator applying a contract that generally selects New York or other US state law, to have in mind generally how other major arbitral jurisdictions treat the question. Here are some basics.

1.The Law of England: The UK Supreme Court chimed in on October 9, 2020 with a decision called Enka v Chubb ([2020] UKSC 38). Per this decision, a general choice of law clause in the contract will also govern the arbitration agreement if the parties have not chosen other law applicable specifically to the arbitration agreement. If no choice of law is made, there will be a presumption that the law of the seat of arbitration governs the arbitration agreement, a presumption derived from the UK conflict rule that points to the law with which the arbitration agreement is most closely connected. Caveat: this was a case concerning the scope of arbitrable disputes and not about who other than a signatory may enforce or be bound by the clause.

2. The Law of Singapore: At least pending any new decision, where the influence of Enka v Chubb might be felt, Singapore law declines to treat the contract’s general choice of law clause as a choice made for the arbitration agreement where no specific choice for the latter has been expressed. It was so held by the Singapore Court of Appeal in BNA v. BNB et ano. ([2019] SGCA 84). The “proper law” of the underlying contract (what we US types call the substantive law applicable) is however presumptively the proper law of the arbitration agreement in the absence of an express choice for the latter. As to what circumstances might bring about rebuttal of this presumptive choice, first on this list is where the proper law of the contract would deny enforcement of the arbitration agreement despite other clear manifestations of the parties’ intent to arbitrate. Stay alert for new developments in Singapore: the presumption sustained in BNA was drawn from UK case law prior to Enka v Chubb, notably in an English Court of Appeal case called (for short) Sulamerica. Does the “Sulamerica Presumption” survive Enka? It’s Topic #1 at the outdoor bar at the Raffles Hotel. Listen closely to the chatter! And caveat reader once more: this is not a case about enforcement by or against non-signatories!

3. The Law of Switzerland: Treading carefully now in the potential quicksand of civil law tradition, a federal arbitration statute without a decisive single construction on the question by the nation’s highest court, and a vast and influential academic and professional literature, an educated guess at the Swiss law position is as follows: One key provision of the Swiss federal arbitration statute (referring to relevant portions of the Swiss Private International Law Act (PILA)) treats an arbitration agreement as valid if it is valid under one of three sources of law: “the law chosen by the parties, or the law governing the subject-matter of the dispute, in particular the main contract, or [under] Swiss law.” (PILA Art. 178(2)). In Europe they call this “in favorem validitatis”; which at your kitchen table means (roughly) “whatever works.” But here’s the thing. Clause “validity” is obviously just a small piece of larger arbitrability puzzle, so it is robustly debatable whether the Swiss position on clause validity applies to non-signatory issues (what in European lexicon pertains to the “extension of” the arbitration clause). But wait, the plot thickens : If you go down a ways in the PILA you get to Art. 187 (1) which is a “special conflict of laws provision just for international arbitration” (said an actual reliable source, Pierre Karrer, in 2014). It says “The Arbitral Tribunal shall decide the case according to the rules of law chosen by the parties or, in the absence thereof, according to the rules of law with which the case has the closest connection.” For some scholars (e.g. Prof. Karrer) this test inevitably leads to applying the law of the seat in the absence of a different express choice. But what if the seat in Switzerland is not chosen by agreement, but by the ICC Court, as a compromise choice wanted by neither party? Happily there is clarity in the Swiss law on extension of the arbitration clause to non-signatories if the conclusion on choice of law is that Swiss law applies: a concept effectively equivalent to common law equitable estoppel prevails – “behavior with respect to the conclusion or the performance of the principal agreement that either clearly demonstrates its (implied) intent to be bound by the arbitration clause, or that under the general principle of good faith may and must have been understood by the other party in such a way that the non-signatory intended to join the principal agreement, including the arbitration clause being part thereof.” (I quote Georg von Segesser paraphrasing the Swiss Supreme Court in a 2009 Kluwer blog post).

4. The Law of France: Choice of law is easy in France because no national law stands to be chosen. The arbitration clause is considered to be governed by general principles focused on the common intent of the parties that are not associated with any national legal system. Voila, another good reason to go to Paris, if ever again we can.

It is a distinctive aspect of American arbitration law there really is not an “American law approach” to be added to this list. Effectively, American arbitration law prescribes to some extent how judges decide issues of arbitrability, but does not prescribe rules about how arbitrators decide issues of arbitrability so long as they respect the agreement of the parties. So it is open to arbitrators sitting in the US to consider all of foregoing approaches to choice of law applicable to the arbitration agreement. And it is clearly open to arbitrators sitting in the US to treat US case law like the Setty case as a unique law-of-the-forum applicable to judicial proceedings and not even particularly instructive about how arbitrators sitting in the US should approach their arbitrability task.

With the northern border closed except to Essential Workers, which You, sorry to say, are NOT, these are challenging times to know what’s really going on Up There in the Great White North. OK, you can binge-watch Bubble Hockey from Toronto and Edmonton. In a timeout You might catch a stirring glimpse of Lake Ontario at sunset, or a mild-mannered Albertan Black Bear seeking entry to the ice arena without a mask. 🐻 😷. But what you really crave is news of what’s happening with NAFTA (USMCA) arbitration. Arbitration Commentaries is here to help.

The case that motivates this post was issued July 20. It’s called United Mexican States (yes, readers, that’s México) v. Burr, written by a Justice of the Ontario Superior Court in Toronto (…. well, by July 20 the Justice was probably having a crackling gin & tonic on the patio of his Georgian Bay cottage, but ICSID said the seat of the arbitration was Toronto). It’s a case about arbitral jurisdiction in a NAFTA arbitration and, specifically, whether NAFTA’s requirement that Investor Claimants send the Respondent State a Notice of Intent to Arbitrate at least 90 days before starting the arbitration, if it’s violated which it was by about half of the 30+ Investor Claimants, means there is no arbitral jurisdiction over the laggards’ claims. Even more specifically than that, one piece of this case, the one Your Commentator will take on here, is what the arbitrators are supposed to do when the non-party treaty countries – here Canada and the USA (as it then was) — chime in on the side of the Respondent State, to say that they think the Treaty means exactly what México says it means — i.e. no 90 Day Advance Notice = No State Party Consent to Arbitrate. Can they give the back of the hand, which the Majority seem to have done, to evidence of “subsequent practice” of the Treaty Parties, when the governing international law says this evidence SHALL be taken into account?

Now, I realize what some of you Americans are thinking: “We did this dance 💃 in BG Group v Argentina, no? This is just SO 2014…” But wait. Wait. Fingers off the delete key. Let’s check in on the different legal frameworks of American and Canadian arbitration law, and see how like cases get treated differently.

BG Group was indeed about how if at all a court reviews an investment treaty advance notice requirement. In that case, the UK-Argentina BIT required an investor first to file a lawsuit in a Host State court, and either lose or chill-out for nine months before starting arbitration. At the Washington DC seat of arbitration, where FAA jurisprudence governs vacatur and the Model Law is chat room fodder, the outcome was dictated by some well-developed legal presumptions in the FAA case law (what the cool 😎 people call “First Options”): that parties presumably want arbitrators not courts to mainly decide whether a procedural hurdle has been jumped, and presumably want courts not arbitrators to mainly decide whether they consented to arbitrate at all. What looks like a procedural hurdle – in a contract or a treaty — might potentially be an issue of consent to arbitrate, but only if the text convincingly says so. And the text of UK-Argentina BIT didn’t. Game to the arbitrators, siding with the Investors, in the BG Group case.

Ok now back to Canada. (I know people who know people… just follow me, via Buffalo, across Lake Erie. Backstroke will do). Here the Model Law is everywhere. You go in Tim Horton’s for morning coffee and you can get the Model Law for $1.99 if you buy an extra large Double-Double and small box of Timbits. You can even find it in the Courthouse, where, if you are México or are otherwise unhappy with a Toronto-seated (or even Moose Jaw) international arbitration award, the motion to vacate is based on Art. 34 of the Model Law which says they can only vacate on specified grounds which just happen to be – can you believe it? – the NY Convention Article V grounds to deny recognition of a foreign award. (Wow, conforming domestic law to international law – what a scandal! The USA used to allow a handful of lawyers in the American Law Institute to talk about this idea, but then this ALI cell was discovered to be a posse of radical-left anarchists and that was the end of that). Since Art. 34 leaves the Canadian court without much choice but to determine whether the Arbitral Tribunal had jurisdiction, the energy for deference to arbitrators on that issue has been channeled into whether what’s presented to the Court is a “true” question of jurisdiction. If it’s the real turtle soup, not merely the mock, then the standard of review is “correctness” – a polite Canadian term for de novo review. Add to the mix the fact that Canada unlike its friend to the south (you have to learn what “friend” means, to get a degree from a Canadian law school) is a Party – a full-fledged, acceded, ratified, card-carrying Member State – of the Vienna Convention on the Law of Treaties (VCLT). You don’t hear much about VCLT in American treaty arbitration case law, what there is of it. In New York, if you cite VCLT in a brief the law clerks mistake it for a lunch order.

So in this México v Burr case the VCLT was in play as governing law, not lunch, and its says in relevant part that in the interpretation of a treaty any “subsequent agreement” or “subsequent practice” of the Treaty parties “shall be taken into account.” Into the Mexico v. Burr arbitration come lawyers for Canada and lawyers for the US (under NAFTA’s Greek Chorus clause, Art. 1128) who say “yes just like México says, we always have and always will treat the failure to send that 90-Day Advance Notice of Intent as a failure to engage the Consent of the Respondent State, and that rises to the level of a subsequent agreement or at least a subsequent practice and either way México wins on jurisdiction and the laggard Investors are out.” But it seems that the Arbitral Tribunal majority more or less understood these pleas as opportunistic messages from litigants who might want to win similar cases — that is to say, Canada and the USA, as serial NAFTA arbitration Respondents with serial motivation to win, are motivated to argue for and to develop some jurisprudence that treats the non-observance of the Advance Notice requirement (NAFTA Art. 1119) as jurisdictional, but the arguments say more about the State-employed litigators’ motives in common to win their cases than they do about the States’ “agreement” or uniform “practice” in the interpretation of the NAFTA.

On a practical level, don’t you wonder what it is about NAFTA Art. 1119 that is necessary to “engage the consent” of the Respondent State? “You live where? On the Upper West Side? Nah, Canada doesn’t do UWS! You’re claiming what, an FET breach? Nah, we only arbitrate expropriation. You want how much? $300 million? Well, we‘ll think about it and let you know in 90 days.” Is this really about consent? Or is it so the lawyer can tell her boss that she has a big new case? Canada isn’t really going to withhold consent for UWS Claimants. Really?? In BG Group terms, the Treaty text evidence that this Art. 1119 Notice is a condition of the Host State keeping its standing offer to arbitrate upright looks pretty thin.

Can the Treaty Parties change things by lining up their briefs in the arbitration (one as Respondent, two as Greek Choraleers under NAFTA Art. 1128) to say “Ever since 1995, all three of us have consistently viewed that 90 day Advance Notice as a condition of consent. No Notice, No Offer to Arbitrate!”? Do those brief-y submissions make out a “subsequent practice in the application of the treaty which establishes the agreement of the Parties regarding its interpretation” that NAFTA arbitrators, bound to heed (not eat) the VCLT, “shall take into account” ? Hmm.

You will appreciate why this is a tough question for any judge who spends her waking hours parsing the opportunistic arguments of litigants. Because whether States bound by a Treaty have a “practice” tantamount to an “agreement” about how they give consent to Treaty-based arbitration is, naturally, about what these States have argued in other cases. It’s not about objective activity out there in the Real World, like how far into its coastal waters a fishing treaty group of States permits its treaty partners’ fishers to fish. Crunching the Ontario judge’s very detailed examination of NAFTA jurisprudence into a digestible summary for You, the Commentaries Consumer, the court holds that (1) “subsequent practice” in regard to NAFTA‘s arbitration commencement provisions may possibly be established by aligned litigation submissions made in one case by the Host State Respondent supported by the other States as Art. 1128 Choraleers, BUT (2) the Court looks at the content of those submissions with three factors in mind: precision, repetition, and collaboration.

These are my interpretive phrases not the Court’s. So what do I mean by “precision”: Well, Canada, if you (or México or USA) argued in earlier cases that investor compliance with the NAFTA case-commencement procedures generally is jurisdictional, that’s not as precise as having argued in prior cases the exact issue here: the alleged jurisdictional significance of the 90 Day Advance Notice under Art. 1119. Repetition?: Much easier. How many times has the precise same argument been presented? Showing three or four prior instances of a State having made precisely the same Art. 1119 Advance Notice argument may be better evidence of “practice” indicative of “agreement” than six or more prior instances of making the more general argument about NAFTA case-initiation procedures. Collaboration?: Also easier. In how many of those prior arbitrations were the three NAFTA States all making the identical argument, in harmony, and calling it a VCLT “practice” tantamount to an “agreement on interpretation of the Treaty.”? As a final step, the Court considered whether arbitral tribunals or courts in the prior cases cited by the States found the arguments persuasive.

Said the Court in its own words: “[T]his is the first case in which the NAFTA parties have offered unanimous submissions on the interdependency of Articles 1119 and 1121 (with the exception of Canada’s submissions in Mondev, where the tribunal did not give effect to those submissions). … While the NAFTA parties do have a long history of arguing for strict adherence to procedural requirements, their track record of success is inconsistent. … Having found that the NAFTA parties’ Article 1128 submissions do not constitute subsequent practice, I find that it was not jurisdictional error for the Majority to treat the submissions as they did.”

Ok why am I telling you all this, you ask, if NAFTA was replaced by the USMCA as of July 1, 2020 and arbitration under the revised treaty has been kicked to the curb? Well, not so fast. First, there are pre-USMCA cases in the pipeline. Second, NAFTA dispute resolution methods are intact until 7/1/2023 for investments made and continuously maintained before and up to 7/1/2020. Third, NAFTA’s arbitration procedures are carried forward for US-México disputes. Fourth, the carry-forward also extends to investors who made government contracts in five key economic sectors. Fifth, Canada and México have not given up on treaty arbitration; henceforth they will do it under the Trans-Pacific Partnership Treaty (CPTPP), and there is surely plenty of room for persuasive construction and application of the VCLT in those cases. Finally, as to the USA and as fans of the Brooklyn Dodgers used to say except in 1955: Wait ‘Til Next Year!

Some of you know that the editorial staff of Arbitration Commentaries has a soft spot in its heart for stories about changing the place of arbitration. And now, just in time for your beach reading (masked, distanced, and only if your beach is open), comes a new tale – or perhaps one should say a new installment of a melodrama that goes back two decades. It involves a certain Sovereign State in South America that, of late, struck out trying to swing a withdrawal of its own gold from the Bank of England. But the judicial activity that motivates this report comes from that rowdy collection of once-united states that, in Congress assembled, rather openly spurned any further nourishment of the Bank of England, already ~244 years ago. Thus the text for today’s sermon: Northrop Grumman Ship Systems, Inc. v. Ministry of Defense of the Republic of Venezuela, 2020 WL 1584378 (S.D. Miss. Mar. 31, 2020), appeal filed, US Court of Appeals for the Fifth Circuit, No. 20-63047, Apr. 29, 2020, Appellant’s brief filed, June 29, 2020.

Here’s how the story goes:

Back when Hugo Chavez was still a college instructor, Venezuela (“VZ”) had some broken boats. VZ signed up a Northrop Grumman (“NG”) affiliate to fix them. The contract called for arbitration in Caracas and for application of VZ substantive law and the VZ Code of Civil Procedure (of which the VZ Arbitration Act, channeling the UNCITRAL Model Law, was and is a chapter).

When NG first moved to compel arbitration, it resisted Caracas and pitched for Mississippi (evidently the Gulf-shore venue of the boat repair shop). The parties initially struck a deal, to arbitrate in México, but the Mexican arbitration fizzled and the parties made their way back to the federal district court in Mississippi on new cross-motions to compel arbitration. Now it is 2010, and Hugo Chavez is running VZ, and NG tells the court “this isn’t the Caracas we signed up for”, and the district court has to make a decision on cross-motions to compel arbitration, that of NG, still pitching for Mississippi or alternatively anywhere but Caracas, and VZ, still pitching for Caracas because that’s what the contract says.

The district court, citing US case law that permits courts to decline to enforce “forum selection clauses” that are “unreasonable,” grants each party’s motion to compel arbitration but only in part: neither party’s pitch regarding the place of arbitration is accepted, and the Court directs the parties to try to reach an agreement on that point: “The Court concludes that the parties should find a mutually agreeable alternative forum as the location for the arbitration proceedings.” The Court in its Order accepted NG’s factual submissions that there was a big risk of partisan judicial interference with the arbitral process in Caracas, through actions of the Venezuelan State in favor of the VZ-affiliated entity. Whether that 2010 order was correct is Issue #1, and really the only issue, in the current Fifth Circuit appeal. So let’s mark it! And come back to it! (Issue #2 is whether the District Court should have refused to confirm the final award under the Panama Convention, but essentially the asserted Convention ground for non-recognition is that the procedure wasn’t what the parties agreed to because they agreed to arbitrate in Caracas. See Issue #1).

After an abortive try by VZ to appeal that Order – rejected by the 5^th Circuit on jurisdiction grounds* — an actual arbitration broke out. NG chose an arbitrator. VZ chose an arbitrator. NG and VZ agreed that the ICC would act (only) as appointing authority, and the ICC appointed a President of the Tribunal. The parties agreed – sort of – that Washington DC would be the place of arbitration – and VZ did not disavow that agreement until the Tribunal was already duly constituted. But now came VZ before the Tribunal to say – as relevant here –  this can only be and is a Caracas-seated arbitration because that is what the contract says! And so, in the critical portion of the critical 2014 procedural order of the Arbitral Tribunal that is now ripe for consideration by US Fifth Circuit in the context of an appeal from an order of the district court granting recognition and execution of the Tribunal’s final award, the Tribunal decided:

1) that the venue provision of the contract, selecting Caracas as the place of arbitration, could not be enforced because the unenforceability of that venue provision had been determined by the district court in Mississippi, and was res judicata,

2) that this left unimpaired the parties’ agreement in the contract that any arbitration would be governed by the VZ Code of Civil Procedure,

3) that the Tribunal would apply the VZ Arbitration Act (as part of the VZ Code of Civil Procedure) as the “curial law” of the arbitration – explained very precisely by the Tribunal to mean the law governing the parties’ conduct before the Tribunal, the proceedings before the Tribunal, and the parties’ relationship with the Tribunal — but notably omitting from this descriptive recitation any reference to the judicial competence of the VZ courts in relation to the arbitration or its outcome;

4) whereas the VZ Arbitration Act leaves selection of the place of arbitration to be decided by the Tribunal if that question is not resolved by party agreement, and the parties’ agreement having been invalidated in a judicial proceeding whose outcome binds the parties, it was incumbent upon the Tribunal under the curial law to select a location, and

5) whereas the place of arbitration to be selected could not be in VZ (see #1), and as a matter of fairness ought not to be in the United States save by agreement, the Tribunal designated Rio de Janeiro as the location for the proceedings.

Fast forward to 2018. In the intervening years, the Tribunal hears the merits, deliberates, and enters a final award. VZ loses, for about $138 million USD. Ouch. So back we go to Mississippi for award enforcement (or not) – for which the district court retained jurisdiction.   The district court’s decision granting recognition and enforcement of that final award is, for our purposes, just a warm-up for the main event in the Fifth Circuit. The only arguable procedural irregularity in the Tribunal’s selection of Rio as the place of arbitration was the Tribunal’s putative lack of authority to make the decision. But that was the issue decided by the district court in its 2010 order, which was final, and the law of the case, for purposes of the district court’s proceedings. This matter was effectively just passing through the district court in 2019-20 on its way to the Fifth Circuit, where the 2010 Order could finally be the subject of an appeal in 2020.

***

What should the Fifth Circuit sensibly do in its post-arbitration assessment of the 10-year-old district court Order? Ironically, one conceivable outcome would be to hold that the Parties’ respective motions to compel arbitration should have been denied.  If the contractual choice of Caracas as the place of arbitration was not severable from, but was integral to, the agreement to arbitrate – and accepting the Court’s 2010 findings that arbitration by these parties of this case could not fairly take place in Caracas — then both motions to compel arbitration arguably ought to have been denied. The unavailability of fair process in Caracas would, on this view, mean that the arbitration agreement, not merely the forum selection clause, was frustrated, commercially impracticable and/or impossible under general contract principles (cf. the New York Convention, Article II (3): “null and void, inoperative or incapable of being performed” – language with no parallel phrasing in the Panama Convention). What then? Would both parties try their luck in Brazil, for recognition of the award and its annulment, respectively?

The 2010 district court Order did not directly address the severability question, and probably should have. But can we not accept that the unstated premise of that Order was that when a business from one State makes a commercial contract with an agency of a foreign State, from the business’s point of view arbitration, not arbitration in the foreign State forum, is the essence of the dispute resolution bargain. The main objective is to resolve disputes outside the foreign State’s courts, with the portability for enforcement offered by the Conventions. The location of the arbitration is a risk accepted as a price for these benefits. Looking at things through the eyes of the Sovereign State party, the view at least presumptively is much the same: it primary objective in agreeing to arbitrate is to eliminate the risk that it could be forced to litigate in a judicial forum of its foreign business partner that it fears may be hostile if not clearly biased, and that it views as unfamiliar and therefore tactically more to the advantage of its opponent. Winning, in the bargain, the right to arbitration on its own territory under its arbitration law and its substantive law, while certainly important advantages, cannot (at least presumptively) be said in the circumstances to be fundamental to the State’s willingness to arbitrate at all. Had there been a negotiating impasse with Northrop Grumman over the place of arbitration, would VZ more likely have accepted the risk of merits litigation in a US court, or the risk that, by leaving the place of arbitration to be agreed later or decided by the Tribunal, the Tribunal might ultimately select a place of arbitration outside of VZ?

The US case law concerning the severability of particular aspects of the agreement to arbitrate, such as the provider institution or the identity of the arbitrator, is mainly in a domestic context and therefore largely fails to capture the considerations that factor into the arbitration v. litigation choice in international business deals generally, and acutely in deals that involve foreign State-affiliated entities. Thus in a domestic context where the parties chose an arbitration provider as the “forum” and the provider no longer exists when a dispute arises, the courts say that they will look to the language of the arbitration agreement for objective indications whether the parties wanted to arbitrate or only wanted to arbitrate if they could do so in that forum. Such case law and such standards are unhelpful to decide cases like Northrop Grumman v. Venezuela. Let’s watch to see if the Fifth Circuit takes this occasion to develop some significant law on this point.

Venezuela in its brief to the Fifth Circuit, says in its lead-off argument that the 2010 district court Order wrongly applied a “[FAA] Chapter 1 standard” to the parties’ respective motions to compel arbitration, each of which arose under Chapter 3 and the Panama Convention. What this argument means is less than clear, but let’s try to unpack it in objective Convention/FAA terms. Article 1 of the Panama Convention states: “An agreement in which the parties undertake to submit to arbitral decision any differences that may arise or have arisen between them with respect to a commercial transaction shall be valid.” Segué to FAA Chapter 3, and we see that Section 303 tells courts they may compel arbitration at any place provided for in the agreement, and if no place is provided for then direct that arbitration shall be held “in accordance with Article 3 of the [Panama] Convention.” The referenced Article 3 says that in the absence of an express agreement the arbitration shall be conducted under the procedural rules of the Inter-American Commercial Arbitration Commission (“Yak-Yak” to IACAC buffs).  

Let’s agree that the 2010 district court Order did not describe this statutory and treaty framework. But it did apply traditional contract principles to determine – in relevant part, as to the choice of Caracas as the Seat – the contents and validity of the parties’ agreement: commercial impracticability, impossibility, and (at least by implication) frustration. What’s wrong with that? Nothing in the Panama Convention or FAA Chapter 3 limits the district court in its methods to find out what the parties’ agreement really is. Besides (you were about to say, if I didn’t) FAA Chapter 1 applies under Chapter 3, where there is no conflict, and FAA Section 2 (in Chapter 1) says arbitration agreements are valid and enforceable save upon such grounds as exist for the revocation of any contract. How about revocation for commercial impracticality, impossibility, or frustration? Where is the conflict between Chapter 1, Section 2 and Chapter 3 /Panama Convention? Maybe the district court in 2010 didn’t exactly check all these analytical boxes the way readers of Arbitration Commentaries would have, but still, it’s looking pretty affirmable. And nothing the district court did was contrary to compelling arbitration in accordance with the Panama Convention – directing the parties to try to agree on a place of arbitration was surely within what the district court could do under FAA Section 303(b) [An express agreement of the parties on a place of arbitration is surely “in accordance with Article 3 of the [Panama] Convention.” The parties eventually reported to the district court that they had agreed to Washington DC as the place of arbitration. When Venezuela later disavowed that agreement, the matter was already before the Tribunal, and both parties then submitted the question of place of arbitration to the Tribunal, as provided for in the governing Venezuela Arbitration Act.

The brief for Venezuela goes on to argue that the 2010 district court Order failed to analyze whether the choice of Caracas as the place of arbitration was severable from rather than integral to the agreement to arbitrate. That’s a fair and accurate critique – but what’s a Fifth Circuit to do now, after nearly a decade of arbitration that culminated in a final award? Isn’t this an occasion to invoke the principle, almost as hoary as your Commentator, that district court judgments can be affirmed on any ground (not waived) that the record will support? As discussed above, it is indeed time for US courts to develop some principles, attuned to international business transactions, to assess whether a choice of the seat of arbitration is integral to, or ancillary to, the agreement to arbitrate. But if the adopted guiding principle is, as it ought to be, that the parties presumptively chose international arbitration primarily (i) to avoid the risk of being forced to litigate in a hostile foreign judicial forum, and (ii) to secure the award enforcement advantages of the New York and Panama Conventions, then it should ordinarily be an uphill battle to show that the choice of one party’s home State as the place of arbitration was such an integral part of the agreement to arbitrate that the entire agreement must fall if the agreement on the seat is no longer valid.

***

The case under discussion presages, and might eventually involve, an issue of interpretation of Article (V)(1)(e). You remember Article (V)(1)(e), which says recognition and enforcement may be refused if the award has been annulled “by a competent authority of the country in which, or under the law of which, that award was made.” Must a court of “secondary jurisdiction,” invited to refuse recognition and enforcement under Article (V)(1)(e), accept that there is potentially concurrent jurisdiction to annul an award in the courts of two foreign States, the State where the award was made and another State under whose arbitration law the award was made?  This question has not been answered by a US court to this writer’s knowledge.

Mainstream international arbitration scholarship instructs us that “an award is ‘made’ at the place at which the arbitration is held, i.e., at the arbitral seat.” (credit F.A. Mann, quoted by other Mainstreamers…). Professor van den Berg tells us that “or under the law of which the award was made…” in Article V(1)(e) (of each Convention) refers to a situation where “on the basis of an agreement of the parties the award is governed by an arbitration law which is different from the arbitration law of country in which the award was made.” That seems to imply that the parties may agree to derogate from the normal “primary jurisdiction”-conferring consequence of naming a place of arbitration, and that the result of such an agreement remains that there is only one State whose courts are to be seen as competent to annul the award. And this view would be consistent with the legacy notion of a single “country of origin” – a notion inherited from the bad old days of double exequatur before 1958.

But the closest common analogy to the situation in Northrop Grumman v. Venezuela is not the side-deal-in-derogation example given by Prof. van den Berg, but one perhaps at least as common: where the parties have chosen procedural law that does not make a default designation of the place of arbitration, and the parties have made no separate agreement about the place of arbitration. If your arbitration agreement in an international case provides that the Ontario International Arbitration Act (adopting the Model Law) shall apply, but names no place of arbitration, then the place of arbitration “shall be determined by the Tribunal” (Model Law Art. 20) if the parties can’t agree. If the Tribunal picks Rio, well, maybe by the time you lose the arbitration and move for an annulment, there will be flights to Brazil, live Samba in the clubs until the wee hours, a Brazilian President recovered from the Virus, and a vaccine.

It seems right to view the Northrop Grumman case just as you would if the parties had agreed initially on the Venezuelan Arbitration Act as the lex arbitri (the “curial law”)  but without naming a place of arbitration. Whereas the parties’ initial selection of Caracas was invalidated in an Order that the Arbitral Tribunal treated as res judicata, and which in Mississippi is the law of the case, any potential annulment judgment from a Venezuelan court, made on the footing that the place of arbitration had to remain and did remain in Caracas, would invite a US court to rule – à la Pemex — that this is a rare instance when a foreign award annulment judgment should not be respected. If the Respondent returns to the district court in Jackson MI, maybe next year, brandishing an annulment judgment from a court in Caracas, not one in Rio, remember what you read here! 🔮

***

A final word about your Commentator’s soft spot for Seat stories. In 2000 as an advocate I had occasion to submit to the ICC Court that Belgrade 2001 was not the same place, in Seat terms, as the Belgrade 1990 that our client had signed up for – certainly not for an ICC arbitration against a Balkanized fragment (Serbia) of a failed State (Yugoslavia). The ICC Court delegated the Tribunal to do the fact-finding.  Post-hearing, and while the Tribunal was deliberating over the status of Belgrade, the facts on the ground changed materially: the Serbian autocrat Milosevic was ousted and there was a “normalization” of Belgrade at least in respects deemed by the Tribunal to be germane to its status as a juridical Seat. Belgrade still lacked good fondue venues, so the hearings remained in Geneva.

One member of the Northrop Grumman arbitral tribunal had close exposure to that episode. If my storytelling here is unfaithful to reality, I expect to hear about it and to pass along any needed corrections!

*Just a question for hard-core FAA buffs to mull, over Zoom cocktails, about the Fifth Circuit’s 2011 dismissal of the initial appeal: Was the 2011 appeal properly seen by the 5^th Circuit as arising from orders granting motions to compel arbitration, and therefore not jurisdictionally sound, under the FAA, until the arbitration is over? Or was it an appeal from an order denying enforcement of a “forum selection clause” – an ancillary agreement about the place of arbitration — that might have been appealable as an interlocutory order as a matter of discretion, or even as a mandatory injunction against a Caracas-seated arbitration?

Let’s talk today about enforcement in New York of arbitral awards annulled in Lagos, Nigeria. Big ones. USD $2 Billion, give or take. I know, you think you have heard and read enough about this topic. But really, where are you going to go today? Wash your hands and read on.

The scripture for today’s sermon is a judgment of the US District Court for the Southern District of New York, now in the throes of an appeal before the US Second Circuit Court of Appeals, in Esso Exploration & Production Nigeria Ltd. v. Nigerian National Petroleum Corp., 397 F. Supp.3d 323 (S.D.N.Y. Sept. 4, 2019).

ICYMI, the raw facts are these: In the early 1990s Esso, Texas folk now called ExxonMobil, corporate descendants of John D. Rockefeller’s Standard Oil (S.O., get it?) made an exploration and production deal with Nigeria’s state oil company. Being pragmatic (even if reluctant) Texas oilmen, they agreed to ICC arbitration in Lagos under Nigerian law. Esso spent a small fortune on the drilling, produced a lot of oil, but had the rug pulled out when Nigeria unilaterally changed the contractually-agreed allocations. (“My oil is my oil, your oil is my oil.”). Esso won an ICC arbitration for the aforesaid $2 billion give or take, and then was wiped out by a series of decisions in the Nigerian courts that effectively annulled the award. On the eve of time bar for U.S. enforcement, Esso brought the award (or its dessicated remnants) to the Southern District of New York, and in the above-cited decision, lost. And this was on the basis of the doctrine of “international comity,” as it has been filtered through the prevailing US jurisprudence on enforcement, or not, of New York Convention awards that have been set aside by a competent court at the seat of the arbitration.

I propose to do two, maybe three things, in this rant.  First, I will trace the principle of international comity as a line running through the US annulled award enforcement case law, mainly Second Circuit case law that is stare decisis, more or less, for the pending Esso-Nigeria appeal. Second, I will hark back to the 19^th Century notion of international comity in US law, and show that it has had a robust recent history elsewhere in the federal judicial system, notably in the Ninth and Eleventh Circuits, much in contrast to the constricted version that has taken root in the Second Circuit’s annulled arbitration award case law. Third and finally, I will suggest that perhaps this more robust outer-Circuit* version of what international comity means US law, foreign States that systematically nullify arbitration awards against their State instrumentalities of the State would face closer scrutiny in US courts of their alleged manipulations of their own judicial organs.

Onward.

We start, again, with Chromalloy (939 F. Supp. 907 (D.D.C. 1996)), because we must. Chromalloy has fallen out of favor in part for the US District Court’s emphasis on Art. VII of the New York Convention** in its framing of the rationale for enforcing, under the New York Convention and the Federal Arbitration Act, a Convention award against Egypt that a Cairo court had tossed into the Nile.  But about the role of international comity in the annulled award enforcement equation the District Court judge in Washington DC had this to say, mainly quoting a 1984 D.C. Circuit judgment in the Laker Airways case (731 F.2d 909): “‘No nation is under an unremitting obligation to enforce foreign interests which are fundamentally prejudicial to those of the domestic forum… [C]omity never obligates a national forum to ignore ‘the rights of its own citizens or of other persons who are under the protection of its laws.’”  (The italics mean I like it, and as we shall see shortly even the Ninth and Eleventh Circuits evidently like this interests-based articulation of the international comity question).

The next case in the US canon, Baker Marine from the Second Circuit in 1999 (191 F.3d 194), did not develop the comity theme; the panel found it sufficient to reject the appellant’s Chromalloy-inspired reliance on Art. VII of the Convention, and to take note of the fact that this had been a Nigeria-seated arbitration under Nigerian law between Nigerian entities (non-State) on both sides – never mind that, on both sides, the contestants were corporate affiliates of large US energy industry players who evidently had organized affiliates in Nigeria in order to do business there. The Court’s conclusion that appellant “ha[d] shown no adequate reason for refusing to recognize the judgments of the Nigerian court” (here the italics are for clarity only) left quite open the question of what might constitute an “adequate reason,” to call out Nigeria for international lawlessness, especially in a case that might involve (as Baker Marine did not but Esso-Nigeria of course does) the Nigerian State as a party-affiliate.

But the D.C. Circuit was back on the track of State entity defendants in the Termorio case (487 F.3d 928).  In this case, Colombia’s state electric utility acted in breach of a deal to buy power, arbitrated under the ICC arbitration clause in the power purchase contract, lost the case for $60-odd million,  and then insisted in a Colombian court on a defense rejected by the arbitrators: that in Colombia after all you really CAN’T arbitrate at the ICC; you have to do it at the Bogota Chamber of Commerce. Award annulled in Bogota.  Justice in American nevertheless? Negative. Recall this from the D.C. Circuit: “Because there is nothing in the record here indicating that the proceedings before the [Colombian court] were tainted or that the judgment of that court is other than authentic, the District Court was, as it held, obliged to respect it.”  487 F.23d at 930.  And to take matters to even greater depths of divergence from US notions of international comity (look back at that quote from Laker Airways, a few stanzas ago), the Court, seemingly channeling a law clerk’s fascination with the writings of a certain Dutch law professor, stated: “For us to endorse what appellants seek would seriously undermine a principal precept of the New York Convention:  an arbitration award does not exist to be enforced in other Contracting States if it has been lawfully ‘set aside’ by a competent authority in the State in which the award was made. This principle controls the disposition of this case.” Id. at 936.  

Mercifully, the Second Circuit has not gone this far, and this notion of a New York Convention near-mandate for recognition of the non-existence of an annulled award has not gained much traction.  But still Termorio helps explain why in the Second Circuit the annulled award enforcement question has become compressed into a “basic notions of morality and justice” pigeonhole: There is an express public policy exception to enforcement in the Convention (Art. V(2)(b)) said the DC Circuit, but not even a glancing reference to public policy in the section about annulled awards. And so, said the Court, even if you could say there is a public policy “gloss” on the annulled awards provision (Art. V(1)(e)) it can’t be any broader than the express public policy ground for non-recognition of an award, and therefore the settled test in that context — basic notions of what is decent and just, etc. — sets the limit of judicial discretion to decide that an foreign award annulment was unlawful and that the award still exists to be recognized and enforced in the USA.  Got it?  Mark this, because when we get back to Esso-Nigeria, we will see that this “basic notions” or “most basic notions” formula is, well, formulaic, and it acts as sort of a strait-jacket inhibiting what ought to be (says your Commentator) a more fulsome international comity analysis.  

But first a couple of stops on the Second Circuit Expressway.   Most importantly, Pemex (832 F.3d 92 (2d Cir. 2016), which looks a lot like Esso-Nigeria on the surface at least: foreign oil-producing State loses oil-related arbitration to what is effectively a US contractor, and then argues in the State’s courts that the arbitration was non-arbitrable; State Court sides with State Oil, and annuls the award. I will spare you the details of the Mexican legislative and judicial actions that formed the foundation for the Second Circuit’s adjudication. The comparison of those actions to the actions taken by the Nigerian State is the essential rhetorical framing of the Esso-Nigeria case, in the District Court’s decision and in the Second Circuit briefing so far. My job is different: first and foremost, to make you smile if not laugh, and to a lesser extent to advocate for courts to think about annulled award cases in a broader framework of international comity jurisprudence. So mainly I want to show you here (1) how the Second Circuit in Pemex seized the “most basic notions of morality and justice” strait-jacket from Termorio, and (ii) how a court somewhere might make its way from Pemex to a more fulsome comity analysis, taking guidance from Pemex and doing no disservice to it.

I start with the fact that the Pemex case really does frame the basic question as whether comity should be given to the foreign award annulment judgment: “[B]ecause the Southern District’s holding necessarily encompassed its decision to deny comity to a foreign judgment, the standard of review is modified: ‘[w]e review a district court’s decision to extend or deny comity to a foreign proceeding for abuse of discretion.’” (832 F.3d at 100).  Discretion to recognize and enforce an annulled award under the Convention, said the Second Circuit panel, “is constrained by the prudential concern of international comity, which remains vital notwithstanding that it is not expressly codified in the Panama [or New York] Convention. … Although courts in this country have long recognized the principles of international comity and have advocated them in order to promote cooperation and reciprocity with foreign lands, comity remains a rule of practice, convenience, and expediency, rather than law … When construing a statute, the doctrine of international comity is best understood as a guide where the issues to be resolved are entangled in international relations.” (832 F.3d at 106) (Internal citations and quotation marks omitted!) (Italics in lieu of an applause emoji).

From this point on, the Pemex reasoning turned down a narrower path, focused on foreign judgment res judicata principles, and in our reductionist world of bottom lines and catch phrases we end up, maybe for worse not better, with an understanding that Pemex stands for a narrow “public policy exception” to a default rule of non-recognition of annulled awards, an exception confined to judgments that are “repugnant to fundamental notions of what is decent and just.” (Id. at 106-07).  But we can find a foundation for a less cramped comity analysis in the text of Pemex:     

“A judgment that tends clearly to undermine the public interest, the public confidence in the administration of the law, or security for individual rights of personal liberty or of private property is against public policy….” (Internal citations and quotation marks omitted). Too bad this is, alas, dictum.

Nearly one year after Pemex, the Second Circuit issued its long-awaited judgment in Thai-Lao Lignite (Thailand) Co., Ltd. v. Government of the Lao People’s Democratic Republic, 864 F.3d 172 (2d Cir. 2017) – long-awaited in part because the panel decided to wait for the Pemex panel to issue its opinion and for Pemex’s application for US Supreme Court review to run its course (certiorari was denied), and then to permit the parties to supplement the briefing in the case to take Pemex into account. Pemex was not fully determinative of the outcome, because the decision turned in part on the fact that the foreign award in Thai-Lao Lignite case had previously been recognized and enforced in a New York federal district court before the issuance of the annulment judgment by the competent court at the arbitral seat in Malaysia. My takeaway from Thai-Lao Lignite, for purposes of this Commentary, is that the panel recognized the separate strands of Pemex that I have identified above – international comity and public policy – and at least initially considered that they might have separate but complementary roles in an annulled awards enforcement setting.  That is my take on the following language from the opinion: “[W]e ruled [in Pemex] that although Article V(1)(e)’s permissive language could be read to suggest that a district court has unfettered discretion as to whether to enforce [an annulled] award, the court’s exercise of that discretion should rather be treated as constrained by the prudential concern of international comity. … Pemex also carved out a public policy exception to the comity principle for occasions when enforcing an arbitral award annulled in the primary jurisdiction is needed to vindicate fundamental notions of what is decent and just in the United States.” 864 F.3d at 176 (boldface supplied for emphasis). From this one might conclude that the Second Circuit here read Pemex as endorsing a fulsome international comity analysis in annulled award enforcement cases, with the public policy “exception” as the exceptional circumstance where deference to the foreign annulment judgment almost inevitably must be rejected. But in the dispositive section of the judgment, the Thai-Lao Lignite Court took a more restrictive view, effectively treating the Pemex version of the public policy exception as exhaustive of the scope of US judicial discretion to enforce an annulled award: “In Pemex, we adopted [our] view of the scope of our courts’ discretion under Article V(1)(e) []: that refusal to recognize a foreign judgment nullifying an award is appropriate only to vindicate fundamental notions of what is decent and just in the United States.” Id. at 184 (Internal citations and quotation marks omitted).  

So now you know why the issue framed for decision by the Appellant Esso in Esso-Nigeria is whether the Nigerian judicial annulment of Esso’s $2 Billion award offends “fundamental notions of what is decent and just in the United States.”  But perhaps you wonder, as I do, why the Second Circuit and the DC Circuit have so consciously turned their backs on the possibility of a fulsome application of the US doctrine of international comity – indeed have not in the annulled award cases even recognized that there is such a more fulsome comity test available (as described in one contemporary version below). Instead the annulled award cases from New York and Washington pay lip service to comity doctrine and segué to a restrictive, deferential test of essential public policy. The reason for this appears to be the perception that the New York Convention assigns special status to the competent court at the seat of the arbitration, recognizing that court’s power to nullify the outcome of the arbitration on any basis short of transparently arbitrary confiscation from the winner. But isn’t it more reasonable to assume – there being no decisive travaux préparatoires on this point cited in the case law or known to this Commentator – that the drafters (huddled in First Avenue taverns during the 1958 Yankees-Braves World Series) had in mind that Contracting States would have arbitration laws that more or less congregate around an international norm, with the occasional benign local quirk like requiring wing arbitrators to be accomplished mogul skiers or permitting vacatur for “manifest disregard of the law”?  Isn’t that the strand of arbitration theory that inspired the evolution of the UNCITRAL Model Law (in the same First Avenue taverns)? And isn’t such a normative limitation on the power of a Convention Contracting State to slaughter its own awards to be fairly implied from that State’s acceptance of an international legal obligation to recognize and enforce foreign awards?

OK you say, show me this so-called fulsome international comity test. I will, and don’t lose patience just because it comes from California. Your pomegranates come from California.  

If you were a Second Circuit judge just poking around for the intellectual fun of it, and feeling un-cabined by the precedents of Baker Marine and Pemex and Thai-Lao Lignite, at least in the privacy of your own chambers and thoughts, you would discover that US comity jurisprudence comes in a variety of shapes and sizes, depending on the purpose for which a comity argument is raised. And to narrow the field a bit, you would discover a sub-category of comity doctrine that is referred to, thematically, as “Adjudicatory Comity.” (E.g., Mujica v. AirScan, Inc., 771 F.3d 580 (9^th Cir. 2014)). That label seems to apply to a number of adjudicatory situations, ranging from one where there is already a final judgment on the same or nearly the same claim from a foreign court, to a parallel proceedings scenario, to something akin to a forum non conveniens motion on a claim not brought but that could be brought in some form in a foreign court. In the Mujica case just cited, the underlying claims were those of victims of a bombing attack by the Colombian military against rebel insurgents in Colombia in 1998. And while claims arising from those events were pending in Colombia’s courts, the claims in the US courts were under our Alien Tort Statute against US corporations allegedly complicit in the bombing, and those precise claims could not be litigated in Colombia. In all events this was referred to as a “prospective” Adjudicatory Comity situation – looking to whether a foreign court would be treated as the exclusive forum for a merits adjudication to be obtained.

So let’s first have a look at the comity principles identified in that setting. And then let’s consider to what extent it might be appropriate to invoke those principles in what is clearly a “retrospective” context – the potential recognition and enforcement in the US of an already-annulled foreign arbitration award, most especially one annulled by the State court of a losing State party.
The Ninth Circuit in Mujica adopted in broad outlines a three-part standard (at least) that had been stated by the Eleventh Circuit, inviting courts to “evaluate several factors, including [1] the strength of the United States’ interest in using a foreign (or domestic) forum, [2] the strength of the foreign governments’ interests, and [3] the adequacy of the alternative forum.” (quoting from Ungaro-Benages v. Dresdner Bank AG, 379 F.3d 1227, 1238 (11^th Cir. 2004)).  Expanding on the US interests factor, the Ninth Circuit elaborated five factors: “(1) the location of the conduct in question, (2) the nationality of the parties, (3) the character of the conduct in question, (4) the foreign policy interests of the United States, and (5) any public policy interests.” (771 F.3d at 604). After a more detailed exposition of the considerations in applying each of these five factors, the Ninth Circuit observed that “[t]he proper analysis of foreign interests essentially mirrors the consideration of U.S. interests.” (Id. at 607). And in the third branch of the exposition of its Adjudicatory Comity analysis, the Court looked at how to assess the adequacy or inadequacy of the foreign forum. Here the Court stated: “[C]ourts consider decisions rendered by the alternative forum and ask ‘(1) whether the judgment was rendered via fraud; (2) whether the judgment was rendered by a competent court utilizing proceedings consistent with civilized jurisprudence; and (3) whether the foreign judgment is prejudicial [and] … repugnant to fundamental principles of what is decent and just.’” (Id. at 608, quoting from Belize Telecom, Ltd. v. Gov’t of Belize, 528 F.3d 1298, 1307 (11^th Cir. 2008)).

“Aha” you say – “finally something that looks familiar.” And indeed it is notable, and important to this discussion, that — courtesy of the Termorio and Pemex and Thai-Lao Lignite — the entire analysis in the Second Circuit of whether to grant or refuse enforcement of an annulled foreign award has been telescoped into what the Ninth and Eleventh Circuits treat as merely one factor in the third branch (“adequacy of the foreign forum”) of a fulsome examination of Adjudicatory Comity. But consider the shortcomings of this telescoping in a case like Esso-Nigeria. Arguably it results in little or no examination of the state of diplomatic relations between the United States and Nigeria. Arguably it results in marginal consideration of available evidence concerning judicial independence in Nigeria broadly speaking, as opposed to what can be brought to bear in the record concerning the specific judicial proceedings in which the annulment occurred. Arguably the narrow focus on what is “fundamentally decent and just” results in no examination beyond the case-specific evidence of Nigeria’s faithfulness or faithlessness as a Contracting State of the New York Convention, or Nigeria’s conduct in regard other bilateral or multilateral treaties to which the US and Nigeria are parties (and note here that two key features of the older notion of comity drawn from 19^th Century and earlier jurisprudence: reciprocity between nations, and the essentially permissive and prudential character of comity. See Hilton v. Guyot, 159 U.S. 113 (1895)). Arguably the Second Circuit’s narrow “fundamental notions” approach also results in insufficient attention to the duration and size of the financial commitment made by the US party to the arbitration, not to mention the actual circumstances of the US party’s acceptance of Nigeria as the arbitral seat and its acceptance of Nigerian law as the governing law. Arguably it results in insufficient attention to the position of the US party in the US and global economy and the economic and security interests of the US and our allies and strategic partners that might be implicated. 

As you read the District Court judgment in Esso-Nigeria and the Appellant’s Brief in the Second Circuit, take note of the extensive record that was made in the District Court concerning many of the factors identified in the preceding paragraph – but especially judicial independence in Nigeria — and observe how little bearing that evidence had on the District Court’s analysis.  That analysis not only focused on the Pemex “public policy exception” formula, but effectively proceeded to narrow that test even further by comparing one-by-one the egregiousness or not of Nigeria’s actions toward Esso to the steps taken by Mexico/Pemex against the US affiliated Pemex contractor.  Appellant in Esso-Nigeria essentially accepts the existing framework of Second Circuit law, and seeks reversal within it. But there are clear overtones in its brief appealing to a broader perspective on comity of the type described as Adjudicatory Comity in the Ninth Circuit.  We should expect to see the Esso-Nigeria case, in its appellate outcome, as provoking more discussion about a shift in that direction in the US jurisprudence concerning recognition of annulled foreign awards.

*Manhattanites thinks of Brooklyn, Queens and the Bronx as “Outer Boroughs.” Manhattan international arbitration lawyers can be charged with thinking of places like the Ninth and Eleventh Circuits as “Outer Circuits,” mainly because they are less often called upon, as compared to the Second and DC Circuits, to adjudicate important issues of US international arbitration law.

* In pertinent part Art. VII says that the Convention does not “deprive any interested party of any right he may have to avail himself of an arbitral award in the manner and to the extent allowed by the law or treaties of the country where such award is sought to be relied upon.” The position of the Court in Chromalloy was this meant that if an award would not be subject to vacatur under FAA Section 10, the domestic FAA, then it should be enforced despite the foreign seat court annulment.  That view has not gained traction in case law and has been criticized in literature.