Poorly drafted international arbitration clauses often challenge US courts to find pro-arbitration solutions that meet the needs of the parties and are practicable within the bounds of the New York Convention and the FAA. That struggle was on display again recently in a case decided by the US Third Circuit Court of Appeals (Control Screening LLC v. Technological Application & Production Co., 2012 WL 3037824 (3d Cir. July 26, 2012).  Here a “pathological” arbitration clause identified a non-existent European arbitration institution, and the Court’s solution was to require arbitration in New Jersey, home turf of one of the parties, even though the other party had already commenced arbitration in Europe. The mission of this post is to explore whether a more optimal solution was available under Chapter Two of the FAA implementing the New York Convention.

The contract in Control Screening was for sale and distribution of goods from a New Jersey seller to a Vietnamese buyer.  The arbitration clause provided for arbitration “at” an institution that did not exist, i.e.: “at the International Center for Arbitration of the European Countries for claim in the suing party’s country under the rules of the Center.” The Vietnamese distributor commenced arbitration in Belgium “under the Belgium Judicial Code” (according to the Third Circuit) and the US company responded by filing a petition to compel a New Jersey arbitration in the US District Court for New Jersey.  The US company also asked the District Court to appoint the arbitrator and to enjoin its Vietnamese counterpart from proceeding with the arbitration filed in Belgium.  The District Court in an unreported decision evidently found meaning only in the contract’s words “in the suing party’s country,” and ordered that arbitration proceed in New Jersey on the basis that the movant was the “suing party” — even though the US movant had not commenced an arbitration, but had only declared that it would do so and was evidently hoping the court would appoint an arbitrator who would then proceed ad hoc.   The Third Circuit affirmed the District Court’s order compelling arbitration in New Jersey, but without adopting the District Court’s interpretation the arbitration clause (under which it would have been equally plausible to compel arbitration in Vietnam). The Third Circuit’s affirmance was solely on the basis that Section 4 of the FAA permits a District Court to order arbitration only in its own district, and that the District Court’s power to compel arbitration includes and depends upon its ability to designate its own district as the place of arbitration in a Convention-FAA Chapter Two case, if the parties have not clearly provided for a place of arbitration in the contract.

As the arbitration clause in Control Screening did nevertheless indicate an intention of the parties to arbitrate in Europe or under the auspices of a European institution, it is useful to consider if another solution was possible. The Court treated the designation of a non-existent “International Center for Arbitration of the European Countries” as null and void, and as severable from the agreement to arbitrate. But Europe is a place, and an order compelling arbitration in Europe would have complied with Section 206, even if it might not have been specific enough to dispatch the parties to arbitration without further judicial relief.  The reference in the contract to a non-existent arbitral institution could have been severed, but without severance and nullification of the geographic designation. If such an order, compelling arbitration in Europe, proved ineffective to launch the arbitration because the parties thereafter could not agree on a particular forum, lex arbitri, or administering body, they could return to the US District Court under FAA Section 5 for appointment of an arbitrator.

This solution would have had the practical advantage of sending the parties to a neutral venue rather than the home turf of the US party. It would also have avoided what was an arguable mis-application of FAA Section 4. Section 4 permits a party “aggrieved by the alleged failure, neglect, or refusal of another to arbitrate” to petition a US district court for an order directing the recalcitrant party to proceed to arbitration. Section 4 has been held to apply in Convention cases, with the exception that its jurisdictional clause — a requirement that the court would have had subject matter jurisdiction of the merits, under the Judiciary Code, but for the arbitration clause — does not apply because Chapter 2 grants original subject matter jurisdiction in Convention cases.

In Control Screening, it is unclear from the Third Circuit decision whether the US party petitioned under Section 4 or Section 206, or both.  But it is even less clear that this could have been a proper Section 4 petition, because the Vietnamese party had already commenced arbitration in Belgium, while Section 4 allows a petition to compel arbitration only where the non-movant had “fail[ed], refus[ed], or “neglect[ed]” to proceed with arbitration under the agreement. The US movant’s  application sought not only to compel arbitration in New Jersey but also to enjoin the arbitration in Belgium. The gravamen of the application was not to compel arbitration, but to obtain a favorable forum and to enjoin arbitration in a perceived unfavorable forum.

Given the ambiguity of the arbitration clause, the Belgian arbitration initiative appears to have been a good faith effort by the Vietnamese party to proceed in accordance with the agreement.  There was some logic to attempting to proceed under the arbitration clause by going to the capital of the European Union. Certainly there was no failure or refusal to arbitrate per se, nor was there a refusal to arbitrate according to the method clearly mandated by the agreement (which some US courts have held to qualify as “failure” and “refusal”). The Third Circuit did not address whether the US party did resort to Section 4, or could validly have resorted to Section 4 in the circumstances. Rather, the Court proceeded to apply only Section 4’s provision that arbitration shall take place in the district where the Section 4 petition is filed. The Third Circuit apparently assumed it was proper to refer to Section 4 for a place of arbitration gap-filler, when relief to compel arbitration is sought in a Convention case under Section 206 but the agreement does not provide for a place of arbitration. But that does not seem to be a correct use of Section 4, as Section 206 forecloses the option of compelling arbitration at a place not provided for in the agreement.  As the Third Circuit rejected interpretation of the arbitration clause as providing for arbitration in New Jersey, it should have considered whether movant fully qualified for relief under Section 4. Instead the Third Circuit simply lifted the place of arbitration clause from Section 4, and applied it without regard to the balance of Section 4. [fn.  In the domestic context, Section 4’s requirement that arbitration shall be in the district where the motion to compel is filed does not operate to enable a choice of forum contrary to the agreement of the parties.  If a party filed a motion to compel in New York under an agreement providing for arbitration in Seattle, the non-movant could seek to have the New York court dismiss the proceeding for improper venue, such that it would be re-filed in Seattle and if it did not do so, the non-movant would be deemed to have consented to a change in the agreed place of arbitration.]

The premise has evolved in the Second Circuit case law, and elsewhere, is that an agreement to arbitrate, to be enforceable under the Convention and FAA Chapter Two, must provide for arbitration in the territory of a Convention Contracting State.  Whereas the Second Circuit has affirmed District Court orders compelling arbitration in New York under FAA Chapter Two where the agreement was silent as to the place of arbitration, it appears that such an agreement is deemed to “provide for” arbitration in New York, and thus in the territory of a Member State, if a motion to compel arbitration is made in a New York federal court. The other dimension of the supposed requirement of a Convention State place of arbitration is that the U.S. in ratifying the Convention agreed to apply the Convention on the basis of reciprocity only to the enforcement of awards made in the territory of another Contracting State, a reservation authorized by Art. I (3). But even though neither Art. I (3) nor the US accession declaration mentions enforcement of agreements to arbitrate, as opposed to awards, the notion that arbitration will not be compelled in the territory of a non-Contracting State has endured, based on judicial construction of the reciprocity declaration as evidence of broader Congressional intent. (See, e.g., DaPuzzo v. Globalvest Mgmt. Co., 263 F. Supp.2d 714, 726 (S.D.N.Y. 2003), citing National Iranian Oil Co. v. Ashland Oil, Inc., 817 F.2d 326, 331 (5th Cir.), cert. denied, 484 U.S. 943 (1987)).

But this case law does not foreclose the solution suggested here, that the court, instead of turning to Section 4 and ordering arbitration in its own district, may compel arbitration initially without reference to a place of arbitration, while reserving the ability to take measures calculated to ensure that the place of arbitration will be in a Convention State. The court’s order might further provide that if the parties have not agreed a place of arbitration or method for arbitrator selection within 21 days, either party may apply to the Court for appointment of an arbitrator under FAA Section 5. And the ensuing appointment order could provide that order the order shall be vacated in case the arbitrator selects a place of arbitration in the territory of a State that is not a signatory of the Convention.

The Third Circuit in Control Screening did not analyze whether the reference to a mis-identified European arbitral institution “provided for” a European place of arbitration (at least for purposes of case administration) even though no such place was identified specifically.  The few courts to have addressed this question have also not considered this. The Seventh Circuit in Jain v. de Méré, 51 F.3d 686 (7th Cir. 1995), for example, assumed without discussion that “provided for” in Section 206 meant “identified.” The arbitration clause in that case required that disputes be presented to “an arbitrary commission applying French laws.”  Clearly no arbitral situs was “identified.” And so the Seventh Circuit held that it was proper to resort to Section 4 and to compel arbitration in the District where the motion to compel arbitration had been filed, i.e. in Chicago (the Northern District of Illinois). But could the court have interpreted the clause to “provide for” arbitration in France? There being no other clues to the parties’ intent, the Court could reasonably have concluded that by providing for “French laws,” the parties intention liberally construed to give effect to the words they selected, was to provide for French substantive and procedural law, and in turn, logically, for a French situs. Would that not have been a solution more attuned to promoting an efficient arbitral process than forcing the parties to arbitrate in Chicago or at least with Chicago as the formal seat? Even if the parties in the Chicago-seated arbitration chose a tribunal of French arbitrators and agreed to a physical venue in Paris, there would likely be costs arising from having French arbitrators apply US lex arbitri, and the US District Court in Chicago would become the forum for any proceedings to vacate an award made under French law.

It is also curious that courts have not paused to consider whether Section 206 must necessarily be interpreted to condition the court’s power to compel arbitration under that Section on the parties’  having provided for a place of arbitration. That interpretation seems to be at odds with the Convention itself, which in Article II(3) requires the courts of Contracting States to refer the parties to arbitration upon finding merely that there is a written agreement to arbitrate.  The Convention does not require that a place of arbitration be identified in order to have an enforceable agreement to arbitrate. It seems plausible in view of the Convention to interpret the Section 206 language “at any place therein provided for…” simply to confirm (i) that arbitration shall not be compelled at a place that is not a place provided for in the agreement, and (ii) in case of the designation in the contract of a foreign place of arbitration, the US court does not lack power to enforce fully the agreement of the parties.

And if this interpretation of Section 206 were accepted, the gap-filling rationale for reading Section 4 as a source of power to compel arbitration in the district where the motion to compel is filed would not exist. And, moroever, taking this approach would permit the Courts to recognize a conflict between Section 206 and Section 4 that has been overlooked. Compelling arbitration in a US venue not provided for in the agreement is at odds with the command of Section 206 to compel arbitration in a particular place only if it is a place provided for in the agreement.

The concern will be raised that an order to compel arbitration which does not name a place of arbitration and where the agreement does not specify rules of arbitration is incomplete and inefficacious. But the problem does not lie with the gap in providing for a place.  The problem is the lack of a method to appoint arbitrators, which will normally be addressed in the law of the place.

The solution in the FAA is Section 5, which permits the court to appoint an arbitrator where (inter alia) the agreement provides no appointment method. In the Jain v. de Méré scenario, had the Court ordered arbitration without naming a place, and the parties could not find a solution, upon application the Court could have appointed a French arbitrator and left the matter of place of arbitration to be decided by the arbitrator. She would do so after hearing the parties on that question. This solution seems considerably more attuned to the principle of party autonomy than the imposition of a US place of arbitration.

Some segments of the international arbitration community (particularly those spending their summer holidays in the Blogosphere), are abuzz with speculation that the US Second Circuit Court of Appeals may reconsider its jurisprudence concerning the arbitrability of arbitrability in a case called Thai-Lao Lignite (Thailand) Co. Ltd. v. Gov’t of the Lao People’s Democratic Republic, 2012 WL 2866275 (2d Cir. July 13, 2012) (summary order affirming district court order granting confirmation petition and denying motion to vacate award). A petition for rehearing en banc has been filed by the appellant Government of Laos, challenging the Court’s position that the parties’ agreement to arbitrate under UNCITRAL Rules was “clear and unmistakable evidence” that they had agreed to arbitrate arbitrability and had thereby foreclose de novo judicial consideration of objections to the arbitral tribunal’s jurisdiction. [Foreign readers take note: U.S. federal appellate decisions are made by panels of three judges. An en banc petition asks the full roster of appellate judges in that Circuit to participate in reconsideration.]

The lead-off premise of the en banc petition is that the District Court ignored established Second Circuit law that the district court, when asked to vacate a Convention award’s jurisdiction rulings, must review the arbitrator’s jurisdiction rulings “de novo,” i.e. as a new matter without treating the tribunal’s decisions with same deference accorded its decisions on the merits. In support of that position, they cite a decision written by Circuit Judge Lynch when he was on the district court, and the fact that the judgment resulting from that decision, confirming an award and denying a cross-motion to vacate, was affirmed by the Second Circuit. (Telenor Mobile Communications AS v. Storm LLC, 524 F. Supp.2d 332 (S.D.N.Y. 2007), aff’d, 584 F.3d 396 (2d Cir. 2009)).

The Telenor case involved a dispute over arbitrability of a shareholder agreement between Norwegian and Ukrainian shareholders of a Ukraine telecom. In the New York arbitration under the 1976 UNCITRAL Rules, the tribunal rejected the Ukraine shareholder’s jurisdiction objection, which was based on the alleged lack of authority of the representative who signed the agreement containing the arbitration clause. In the district court confirmation/vacatur proceeding, Judge Lynch rejected the appellant’s argument that, by virtue of the agreement to arbitrate under the 1976 UNCITRAL Rules which empowered arbitrators to rule on objections to jurisdiction, there was “clear and unmistakable evidence” of an agreement to arbitrate arbitrability. Judge Lynch acknowledged that prior decisions gave such presumption-rebutting effect to the competence-competence rules of the ICC and ICDR. But he considered the 1976 UNCITRAL Rule to be narrower because it authorized the tribunal not to “rule on jurisdiction” but only to rule on “objections to jurisdiction.” Nothing in Judge Lynch’s decision suggested that, if the presumption had been rebutted, in an award confirmation/vacatur proceeding the court would nevertheless review the arbitral decision on jurisdiction de novo. Judge Lynch proceeded with de novo examination of the arbitrability issue, and determined that there was an arbitrable dispute.

The Second Circuit in Telenor discussed the “review of arbitrability questions.” The Court wrote that the presumption that arbitrability is “to be decided by the courts, not the arbitrators themselves,” is rebutted “only by ‘clear and unmistakable evidence…that the parties intended that the question of arbitrability shall be decided by the arbitrator.'” But the Second Circuit did not address whether Judge Lynch was correct in finding that the parties’ agreement to arbitrate under the 1976 UNCITRAL Rules was not such “clear and unmistakable evidence.” The affirmance of the District Court’s decision, insofar as it pertained to the arbitral determination of arbitrability, was that appellant, the Ukraine company on the losing side, had failed to present evidence to the arbitral tribunal that would have justified a de novo trial on arbitrability in the district court. Had there been such evidence, then the Second Circuit would have had to decide whether Judge Lynch was correct that the presumption of judicial hegemony over arbitrability had not been overcome. But that was not the case. The Second Circuit appears to have assumed, without deciding, that de novo review of the arbitrability issue by the district court was proper.

So, insofar as the Thai-Lao Lignite en banc petition cites Judge Lynch’s Telenor opinion in the district court as Second Circuit law, because the district court’s judgment was affirmed, there is an evident flaw, because the affirmance in Telenor appears to have been “on other grounds” — i.e. not on the ground that district courts must (always or at least in arbitrations under the 1976 UNCITRAL Rules ) review arbitral arbitrability rulings de novo.

The next premise of the en banc petition is that, before July 2012, all the federal appellate decisions favoring deferential review of arbitral arbitrability rulings pursuant to arbitral compétence-compétence rules arose in cases where the district court had made a decision about the arbitrability of arbitrability before the arbitrators decided the question. The Second Circuit cases said to fall in that category  — Chevron and Contec, discussed below — Laos argues, are not controlling when the initial federal judicial decision on whether arbitrability was arbitrable occurs at the award confirmation/vacatur stage.

Of primary significance is the Second Circuit’s decision in Republic of Ecuador v. Chevron Corp., 638 F.3d 384 (2d Cir. 2011). The case reached the Second Circuit, exceptionally, while the BIT-based arbitration between Chevron and Ecuador was ongoing and there had been no arbitral decision on jurisdiction. This was so because in the district court the Republic of Ecuador joined with the so-called Lago-Agrio plaintiffs (individual citizens claiming environmental damages) in asking for an anti-arbitration injunction to stop Chevron from pursuing the BIT case. Immediate appealability was based on the denial of injunctive relief.

Ecuador’s position was that Chevron ceased to be eligible to invoke arbitration under the US-Ecuador BIT when it agreed to litigate in Ecuador with the Lago-Agrio plaintiffs as a condition for forum non conveniens dismissal of the plaintiffs’ claims in a US court. The Chevron Second Circuit panel — with Judge Lynch, now elevated to the appellate bench, as author of the opinion — stated the law in terms nearly identical to the Telenor opinion, i.e. that arbitrability issues “should be decided by the courts unless ‘there is clear and unmistakable evidence from the arbitration agreement …that the parties intended that [they] be decided by the arbitrator.'”

Judge Lynch’s Chevron opinion holds that Ecuador by agreeing in the BIT to resolve investment disputes under the UNCITRAL Rules (1976 version), “consented to sending challenges to the ‘validity’ of the arbitration agreement to the arbitral panel.” The opinion continues, on this point: “Because Ecuador’s waiver and estoppel claims go to the validity of the arbitration agreement, Article 21 of the UNCITRAL Rules requires that they be decided by the arbitral panel in the first instance.”

The words “in the first instance” are the fuel nourishing the Laotian Government’s en banc engine. The phrase should be understood, Laos argues, to mean that the compétence-compétence rule incorporated in an arbitration agreement gives the arbitrator the right without judicial obstruction to decide jurisdiction and move forward with the merits if appropriate, but leaves the final say on arbitrability to the courts in post-award de novo review.

But the interpretive clues within the Chevron opinion do not support this.

The first such clue is the Court’s citation, after its introduction of the phrase “in the first instance,” to a particular passage in a 2002 Second Circuit case, Bell v. Cendant Corp., 293 F.3d 563, 566. The cited passage in Bell quotes the Supreme Court in First Options, stating that “the parties to an arbitration agreement ‘may provide that the arbitrator, not the court, shall determine whether an issue is arbitrable.” Thus, Bell read First Options to mean that when the presumption of judicial hegemony is rebutted, the arbitrator decides arbitrability as an arbitrable issue like any other, meaning with the same limited scope for post-award review. Nothing in Bell or the quoted language of First Options suggests that this is only a “first bite at the apple” rule.

The second clue in Chevron about the import of the phrase “in the first instance” lies in the Court’s discussion of its opinion in Contec Corp. v. Remote Solution Corp., 398 F.3d 205 (2d Cir. 2005).  The Chevron court stated: “We concluded [in Contec] that [AAA Commercial] Rule 7’s language ’empower[ed] an arbitrator to decide issues of arbitrbility…. [and] served[d] as clear and unmistakable evidence of the parties’ intent to delegate such issues to an arbitrator.'”

Thus, Chevron reads Contec as holding that arbitral compétence-compétence rules incorporated in the arbitration agreement result in a “delegat[ion]” of the arbitrability issue to the arbitrator, which seems the antithesis of a principle that the arbitrator is only getting a “first bite at the [arbitrability] apple.” And neither Chevron nor Contec contains any suggestion that “delegation” occurs, and de novo review is foreclosed, only when the district court court opines on the arbitrability of arbitrability before the tribunal opines on arbitrability.  That happenstance of the procedural posture of Contec and Chevron does not appear to have affected the outcome. And Judge Lynch’s opinion in Telenor was not based on any such distinction, but only on a distinction between the UNCITRAL and AAA/ICC compétence-compétence rules which, in Chevron, he implicitly acknowledged to have been invalid.

Further, removing any doubt Telenor might have left about a distinction based on language variation between the AAA Rule and the 1976 version of UNCITRAL Art. 21, Judge Lynch in Chevron follows the Contec discussion immediately by stating that “Ecuador is now in the same situation” as the appellant in Contec.

With these elements in mind, one can readily understand why the district court in Thai-Lao Lignite responded to Laos’s reliance on Telenor by stating, in a footnote, that Telenor was “abrogated” by Chevron. That does appear to be the case, and if it is not, the perhaps Judge Lynch will lead the charge in rounding up votes on the Circuit to agree to review the Thai-Lao Lignite summary order en banc.

The en banc petition in Thai-Lao Lignite depiction of the state of the law in the Second Circuit may not be particularly convincing. But does this en banc petition have a broader agenda: to entice the Second Circuit to re-think its arbitrability-of-arbitrability jurisprudence? The Restatement Third of the Law of International Commercial Arbitration rejects the position of the case law (at least the majority of it) that treats adoption of arbitral compétence-compétence rules by the parties as clear and unmistakable evidence of an agreement to arbitrate arbitrability such that judicial review, where the US was the seat of arbitration, is essentially confined to the narrow FAA statutory grounds for vacatur (and more particularly to a generally- insurmountable “manifest disregard” standard for whether the arbitrator exceeded her powers). The position of the Restatement is that arbitral compétence-compétence rules do not, in general, speak to the scope of judicial review of arbitral jurisdiction rulings, and therefore the incorporation of such rules in an arbitration agreement, without more, should not be seen as clear and unmistakable evidence of the parties’ intention to have arbitrators decide arbitrability in a final and unreviewable fashion.

Surely the Restatement drafters have a point. When parties agree to “resolve” disputes under ICDR or ICC rules, their thoughts are often not focused on such fine points as jurisdiction. And those Restatement drafters — more voracious readers than this Commentator of foreign law sources on compétence-compétence (and not the least, recently, the UK Supreme Court decision in Dallah), will surely point out that scarcely any foreign jurisdictions with well-developed arbitration jurisprudence permit consensual foreclosure of meaningful judicial review of arbitral arbitrability rulings.

Perhaps the solution in US law could be a presumption that parties reserve the right of de novo judicial review of all arbitral jurisdiction rulings, and this presumption should require separate rebuttal by clear and unmistakable evidence. But the consequence of such a principle would be waves of litigation over the presumption-rebutting significance of arbitration clause language (“resolve”, “settle” , “finally resolve”, “shall be final and unreviewable” ? And what if the clause specifically refers “the scope of the agreement to arbitrate” to arbitration, but lacks “finality” language?). The alternative might be for US courts to adopt a rule that prevents parties from contracting to affect the scope of judicial review of arbitral jurisdiction rulings in cases falling under the New York Convention — in essence declaring that whether there exists an agreement to arbitrate underlying a Convention award presented for confirmation is a question committed to the courts by the Convention, and therefore by FAA Chapter Two, and so the parties may not by contract constrict the judicial role in deciding whether the tribunal had jurisdiction to issue all or any portion of its award.

The Thai-Lao-Lignite en banc petition may well have dim prospects. But readers should stay tuned for a certiorari petition in this case, which may tantalize the Supreme Court with an opportunity to write an important new chapter of the U.S. law on compétence-compétence.

In various posts on this site over the three-plus years of its existence, Arbitration Commentaries has reported on court decisions that recognized the power of an arbitrator to decide “arbitrability” questions when the arbitration agreement calls for arbitration under rules that confer power on the arbitral tribunal to decide upon objections to its jurisdiction.
Last week the US Fifth Circuit Court of Appeals joined forces with at least the First, Second, Eighth, Eleventh and Federal Circuits, in holding that when parties agree to arbitrate under such rules, there is “clear and unmistakable evidence” that they have agreed to arbitrate arbitrability. Petrofac, Inc. v. DynMcDermott Petroleum Operations Co., 2012 WL 2892401 (5th Cir. July 17, 2012).
But even as the head-count of federal appeals courts embracing this principle was increasing by one, there were two district court decisions earlier this month — one in the Second Circuit, where the law is supposedly settled — that declined to apply it. This inconsistency invites a question: whether the principle of “agreement to arbitrate arbitrability” is adequately defined in the case law, so that judges will recognize the situations in which it should be applied.
It may or may not be analytically significant that each of the two recent district court cases in which the principle was not applied involved (i) a motion to compel arbitration in an action or proceeding before the court, (ii) no arbitration yet pending, and (ii) a request to compel arbitration of fewer than all issues presented in the case.
The first case was a high-stakes patent license/infringement dispute before a Virginia federal district court, between German and US companies. (Bayer Cropscience AG v. Dow Agrosciences LLC, 2012 WL 2878495 (E.D. Va. July 13, 2012)). The license agreement provided for arbitration under the ICC Rules. The parties agreed that issues of whether the license had been breached and lawfully terminated were arbitrable, but disagreed over arbitrability of patent infringement issues. Taking note of the “persuasive authority” of several federal circuits, and that there was a recent district court decision in the Virginia district adopting the principle, but also taking into account that the Fourth Circuit had not addressed the issue so that it “remained unsettled,” and that the general rule in the Fourth Circuit has been that questions of arbitrability are for the courts, the court held that “it may determine this threshold issue of arbitrability.” The Court proceeded to hold that the claims of patent infringement, in addition to the claims of license breach and termination, were arbitrable.
The second district court case, a Magistrate Judge’s Report in a New York federal court, is more unusual — for a number of reasons including the participation of this writer as counsel for certain parties. (Cardell Financial Corp. v. Suchodolski Associates, Inc., No. 09 Civ. 6148, slip op., ECF Docket Entry #102 (S.D.N.Y. July 17, 2012). The parties had engaged in multiple international arbitrations and collateral judicial proceedings in New York and Brazil since 2003. Eventually there was a final arbitration award that included an anti-suit injunction that forbade the award-loser to start any proceedings relating to the contracts at issue until the damages portion of the award had been satisfied. The District Court confirmed the award and the Second Circuit affirmed the confirmation, giving the arbitral anti-suit injunction the status of a final judgment of the district court. The award-loser then commenced a new lawsuit in Brazil, and the winner brought a contempt proceeding in the Southern District of New York, saying the injunction portion of the judgment had been violated. The loser/putative contemnor argued that the new Brazil suit was not related to the contract, and that whether this was correct required interpretation of the contract, and therefore required an arbitration. Further, whereas the arbitration clause provided not only for arbitration under AAA commercial arbitration rules but also expressly made arbitrable disputes over the scope of the arbitration clause, the putative contemnor cross-moved to compel arbitration of the arbitrability issue as well as the merits of whether the new Brazil lawsuit was related to the contract.
The Magistrate Judge’s report recommends denial of the motion to compel arbitration, not citing case law but holding in essence that this contempt proceeding presents a dispute solely concerning the language of the Court’s judgment, not the underlying contract, even though the scope of the judgment’s injunctive prohibition depends on whether the newly-commenced proceeding allegedly violating the injunction is “related to” the contract.
I will not argue here the merits of a case that is sub judice for one of my own clients. But I will note that in both recent instances where the courts declined to apply the “agreement to arbitrate arbitrability” principle, they were being asked to compel arbitration of the arbitrability issue, whereas in the recent Fifth Circuit case the issue was whether the arbitrator’s jurisdiction decision, presented for review with a final award, was to be reviewed de novo or with the same limited scope as any other arbitrable issue. The procedural posture of a completed arbitration understandably presented a less inviting situation for judicial control.
Does it, or should it, make a difference whether the question arises in a pre- or post-arbitration context? The same question is presented in either context: what did the parties agree to submit to arbitration? Perhaps what we see is that courts are more readily convinced of the parties’ intent to submit an arbitral jurisdiction issue to arbitration when the parties have already arbitrated and the loser is seeking expanded judicial review, or at least when there is pending a “live” arbitration in which some issues clearly must be arbitrated. In contrast, when one party is seeking to litigate the entire dispute, and perhaps no arbitration has yet begun, the agreement to arbitrate under rules that empower an arbitrator to resolve jurisdiction issues may look like a less convincing basis to draw any definite conclusions about the intentions of the parties to arbitrate the commercially obscure issue of arbitrability.
But under the case law of the circuits that have adopted the “arbitrate arbitrability” principle, these distinctions should not matter. The principle vindicates the parties’ ability to resolve disputes within a semi-autonomous legal system that largely stands alongside the judiciary rather than being a subordinate process subject to judicial control. Whether the arbitrability issue arises pre-, post-, or mid-arbitration, courts should respect clearly expressed intentions to commit arbitrability decisions to arbitral determination.

If you practice arbitration law internationally from a base of operation outside the United States, you might consider that the arbitrability of cross-border disputes involving insurance is rather non-controversial. After all, arbitration lawyers wish to have large and growing practices, and insurance disputes help enormously. Simple. You would think.

But here in the United States, with our great legal tradition of making simple matters difficult, this subject is fertile ground for controversy, large legal fees, and lengthy opinions from appellate courts. To understand this quagmire, readers might study ESAB Group, Inc. v. Zurich Insurance PLC, 2012 WL 2697020 (4th Cir. July 9, 2012). Or they might read this post and have time remaining for the crossword puzzle.

It is another chapter in the perpetual power struggle between the federal Montagues and the state Capulets (whose descendants, when they came ashore in South Carolina, changed their respective family names to Hatfield and McCoy).

By tradition, the issuance of an insurance policy was, in America, viewed as a local matter, and was regulated by the governments of the several States. But the Supreme Court in the 1940s held that federal power to regulate interstate commerce allowed the US government to regulate insurance, and Congress swiftly stepped in to restore home field advantage to the states, with a law called the McCarran-Ferguson Act. McCarran-Ferguson said “no Act of Congress shall be construed” to interfere with state regulation of insurance, unless it does so explicitly.

Many states took advantage of this Congressional restoration of state hegemony to pass laws prohibiting arbitration of insurance disputes. The Federal Arbitration Act was considered to be an “Act of Congress,” and compelling arbitration of an insurance dispute under the FAA would entail construing the FAA to regulate insurance. Hence it came to be understood that state law could and did, via McCarran-Ferguson, “reverse pre-empt” the FAA, and require a judicial forum for insurance cases.

Then along came the New York Convention. Rather slowly, to be sure, as the US dallied from 1958 to 1970 before giving legislative blessing to a treaty thrashed out on the shores of the East River. And it was only a matter of time before a state law requiring a judicial forum for an insurance dispute was held to apply to a transnational dispute and thus came into a collision with a motion to compel arbitration based on the New York Convention and FAA Chapter Two.

Last week’s decision from the federal Fourth Circuit Court of Appeals arose from an insurance coverage dispute between a South Carolina manufacturer and the Swedish insurance firm that issued global coverage to the manufacturer’s Swedish parent company. The Court affirmed the ruling of the federal district court, which had accepted jurisdiction under the New York Convention and granted the motion to compel arbitration made by Zurich Insurance as the assignee of the Swedish insurer’s loss portfolio.

Credit the Fourth Circuit for shaping its analysis of this “reverse pre-emption” issue around a central, fundamental policy premise: that the United States seeks to speak with one (federal) voice in foreign affairs, including the regulation of foreign and international commerce.

Invokintedg that premise to guide its inquiry into Congress’s intent, the Court conclude that the “Act[s] of Congress” referenced in McCarran-Ferguson, i.e. those that shall not be construed to impair state regulation of insurance, could not have been meant include statutes like FAA Chapter Two that implement treaties. In other words, Congress did not intend for the New York Convention to take a back seat to state regulation of insurance. The result surely was greeted warmly at Zurich Insurance, who executives must have wondered what dysfuntionality of the US legal system could force them to litigate a high stakes coverage dispute in a US court, despite Zurich having inherited from its Swedish assignor a standard arbitration clause that originated between a Swedish insurer and a Swedish insured.

Aficionados of treaty-interpretation jurisprudence among US readers will take note that the Fourth Circuit panel, following in the path set by the Fifth Circuit in an en banc opinion three years ago (Safety Nat’l Casualty Corp. v. Certain Underwriters at Lloyd’s, 587 F.3d 714 (5th Cir. 2009) (en banc), cert. denied, 131 S.Ct. 65 (2010)), decided that it did not need to decide whether the New York Convention is a “self-executing treaty” under US law. The anti-arbitration argument made in each case was that (i) the Convention is not a complete self-contained direct mandate to courts of the Contracting States, (2) therefore FAA Chapter 2 was needed to make the Convention effective as US law, and (3) therefore a court asked to compel arbitration of an insurance dispute, in contradiction of a state law guranteeing a judicial forum for insurance disputes,  is invited to construe an “Act of Congress,” FAA Chapter Two, as interfering with state insurance regulation, in violation of McCarran- Ferguson.

The Fifth Circuit had rejected this argument by finding that, whether or not the Convention was a self-executing treaty, a motion to compel arbitration in a Convention case calls upon the Court to construe the Convention, not FAA Chapter Two.  The Fourth Circuit’s rejection is more broadly stated: FAA Chapter Two, as legislation implementing a treaty, is not among the “Act[s] of Congress” covered by McCarran-Ferguson, which concerned only Congressional legislation regulating domestic commerce. Perhaps the Fourth Circuit approach offers more predictable protection for international arbitration, as it avoids controversy that might arise in a particular case about the meaning of particular language in FAA Chapter Two.

Should we rejoice or commiserate over the decision during the past week, by a panel of the US Eleventh Circuit Court of Appeals, holding that a party to a domestic arbitration in Ecuador could obtain US discovery for the Ecuador case pursuant to Section 1782 of the US Judicial Code? (Consorcio Ecuatoriano de Telecomunicaciones S.A. v. JAS Forwarding (USA), Inc., 2012 WL 2369166 (11th Cir. Jun. 25, 2012)).  Surely litigation lawyers in Florida will be pleased, as the news will spread rapidly in Central and South America, and whereas US sources of evidence for Latin arbitrations are somewhat more likely to be found in Florida than elsewhere.  But some readers of the Section 1782 canon will wonder if the Eleventh Circuit panel majority has overstepped a boundary in concluding that its holding is a logical extension of the Supreme Court’s decision in the Intel Corp. v. Advanced Micro Devices, Inc., 542 U.S. 241 (2004).

The issue is whether an arbitral tribunal that is a creature of a private commercial contract is a “foreign or international tribunal” as that term is used in Section 1782. The US Supreme Court in the Intel case held that the European Commission and the EU courts with adjudicatory jurisdiction over claims involving violations of EU competition law and regulations are such tribunals,  and that evidence to be gathered in the US for presentation to the EU’s investigative Directorate

General — Competition (“DG-Competition) was “for use in” such a “tribunal” because the investigative record of the DG-Competition proceedings would be the basis for European Commission adjudication imposing or declining to imposed penalties, which would be subject to judicial review in the EU Court of First Instance and on further appeal to the European Court of Justice.

But in Intel, there was no doubt about the sovereign DNA of the DG-Competition, the European Commission, or the EU courts; they were creatures of the EU Treaty; and so the Supreme Court’s decision only had to focus on whether the connection between evidence presented to the DG-Competition and the ensuing adjudications was sufficiently clear that evidence gathered for presentation to DG-Competition was “for use in” an EU adjudication, even though the adjudicative proceedings was not yet pending.

The Supreme Court in Intel did not state this premise explicitly. One can perhaps read the Court’s dicta about “foreign or international tribunal” to relate only to adjudicatory function, irrespective of the sovereign or private character of the tribunal. And it was this uncertainty about the Intel Court’s analysis that fueled debate in commentaries and conferences, and a few district courts, over whether a privately-constituted arbitral tribunal seated outside the US is a “foreign or international tribunal” under 1782 in light of Intel.

The phrase “foreign or international” by itself suggests that the tribunal must be a unit of a sovereign State or States. If the intention of Congress had been only to identify tribunals situated outside the US, sovereign or private, the phrase “foreign or international” is redundant and imprecise. But we also know that Section 1782 in its pre-1964 iteration was available to get evidence for use in “foreign or international judicial proceedings.”  And the language change made in 1964 via the drafting of the late Professor Hans Smit, to substitute “tribunals” for “judicial proceedings,” appeared to have as its purpose to take cognizance of the fact that foreign States, like the US, had a variety of adjudicatory bodies that might be left out if the catch phrase remained “judicial.” But the contemporaneous record of the 1964 amendment — six years prior to US accession to the New York Convention — furnishes no indication that Congress intended to obliterate distinctions between sovereign and private arbitral adjudications.

The Eleventh Circuit records in a footnote that its decision is in conflict with pre-Intel decisions of the Second and Fifth Circuits, whose decisions placed private arbitral tribunals outside Section 1782 because the statute was only “intended to cover governmental or inter-governmental arbitral tribunals and conventional courts and other state-sponsored adjudicatory bodies.” But, says the Eleventh Circuit panel majority: that was then (pre-Intel) and this is now (post-Intel), and the Eleventh Circuit majority reads Intel as “set[t]ing forth a far broader and wholly functional definition of the term ‘tribunal.'” (emphasis supplied).

It is at least arguable that the Supreme Court in Intel did not intend to cast aside the sovereign connection requirement.  First, the sovereign- private distinction was not before the Court. The EU’s DG-Competition and the EU’s competition law courts were clearly inter-governmental bodies. Second, Intel records that the 1964 amendment of Section 1782 flowed from the establishment by Congress in 1958 of a “Commission on International Rules of Judicial Procedure” whose mandate as set forth in the enactment was to “investigate and study existing practices of judicial assistance between the United States and foreign countries with a view to achieving improvements.” (emphasis supplied). Third, Intel notes that the Senate Report on the 1964 change “explains that Congress introduced the word ‘tribunal’ to ensure that ‘assistance is not confined to conventional courts,’ but extends also to ‘administrative and quasi-judicial proceedings.'” (emphasis supplied). Fourth, the Intel Court, citing the amicus brief filed by the European Commission, made a precise record of the DG-Competition’s status as a creature of inter-governmental sovereignty under the European Union treaty. The Court then undertook, in order “to place this case in context,” to trace the role of the DG-Competition in enforcing EU competition law and regulations. Fifth, the holding in Intel is that “[t]he statute authorizes…a federal district court to provide assistance to a complainant in a European Commission proceeding that leads to a dispositive ruling, i.e. a final administrative action both responsive to the complaint and reviewable in court.” (emphasis supplied). Sixth, the Intel Court notes that among the discretionary factors a district court may consider is “the receptivity of the foreign government or the court or agency abroad to U.S. federal-court judicial assistance.” (emphasis supplied). Seventh, in its analysis of whether a Section 1782 applicant must show that the matter would be discoverable in the foreign tribunal, the Court “question[ed] whether foreign governments would in fact be offended by a domestic prescription permitting, but not requiring, judicial assistance.”  (emphasis supplied). Eighth, the Intel decision includes no reference to the pre-Intel Second and Fifth Circuit decisions holding that a private arbitral tribunal is not within Section 1782. Ninth, the Intel court quoted Professor Smit’s International Litigation treatise, which included a reference to “arbitral tribunals” amidst a litany of clearly sovereign-related adjudicative bodies, and the Court quoted the Smit treatise not for the purpose of suggesting that 1782 reaches non-sovereign arbitral tribunals, but only to show that 1782 clearly covered the EU courts that would hear EU antitrust claims based on the investigative record made before the EU’s DG-Competition.

There is a case to be made, therefore, that Intel was not pronouncing a “strictly functional” approach to what is a “foreign or international tribunal,” but instead that, having no need to deal with the private versus sovereign dichotomy, the Court adopted a functional approach to the question whether the “for use in” requirement of Section 1782 was met when (i) the only forum to which the US-gathered evidence could be directly submitted, the DG-Competition, was investigative not adjudicatory, but (ii) the applicant for the 1782 discovery might eventually use the DG-Competition record to pursue its claims before the EU Court of First Instance and then on appeal to the European Court of Justice.

A final word about the decision from the Eleventh Circuit. Only two of the three members of the panel accepted the position that a private arbitral tribunal is covered by Section 1782. And one of those, the non-author of the opinion, was a district court judge in Florida sitting by designation. The other regular Eleventh Circuit appellate judge on the panel concurred specially, to state that he would have affirmed the decision of the district court to allow 1782 discovery only because the arbitral claimant in Ecuador was in addition contemplating civil and private criminal suits in Ecuador courts, courts that are indisputably “foreign tribunals” within the ambit of Section 1782. Such division of opinion within the appellate panel might lead to en banc review by the full Eleventh Circuit, and arbitration law watchers will be watching for that in the coming weeks.

It is occasionally the pleasant duty of Arbitration Commentaries to inform its readers that American trial judges do understand the New York Convention and Chapter 2 of the Federal Arbitration Act (“Convention Act”), and that they often apply the Convention and the Convention Act sensibly to advance international arbitration and the predictability and stability of American law that supports it.

This post is such an occasion. Within the past two weeks:

(1) A federal district judge in New York properly rejected the attempt of a party to an ongoing international maritime arbitration to get judicial relief from the arbitral tribunal’s procedural ruling staying the proceedings until the Claimant complied with a partial final award directing Claimant to post security for Respondent’s counterclaims. (SH Tankers Ltd. v. Koch Shipping, Inc., 2012 WL 2357314 (S.D.N.Y. Jun. 19, 2012); and

(2) A federal district judge in Dallas held that an action to confirm an award  in an arbitration in Texas between brothers who were Indian citizens residing in Texas, concerning their respective interests in a California corporation headquartered in Texas, was properly removed to the federal court from a Texas state court because the brothers’ foreign citizenship, not their Texas domicile, determined that the award in their arbitration was an award not considered as domestic in the United States, and therefore was an award to which the Convention applies. (Nanda v. Nanda, 2012 WL 2122181 (N.D. Tex. Jun. 12, 2012)).

The New York case is a tale of an unhappy arbitral litigant seeking an interlocutory judicial bailout. Claimant owned an oil tanker that it chartered to Respondent for shipment of $170 million worth of oil from Iraq to the U.S. En route, the tanker was hijacked by Somali pirates and held hostage for seven months until Claimant’s war-risk insurer paid a $9 million ransom. Claimant commenced arbitration to recover the ransom from its customer, the Respondent charterer, and the Respondent asserted counterclaims. Further, concerned that  Claimant could not, and that Claimant’s insurer would not, satisfy an award on the counterclaims, Respondent obtained a partial final award for security of $13 million. When that award was not complied with, Respondent, preferring not to expend resources to prove a claim it would be unable to collect, then obtained a procedural order from the Panel staying all proceedings pending compliance.

Claimant then opened its litigation front, asking the federal district court to compel Respondent to arbitrate its counterclaims, i.e. to proceed without Claimant first posting security.   Claimant thus presented the highly unusual question of whether a conditional refusal to proceed within an arbitration, based on a Tribunal’s order permitting such conditional refusal to proceed ,  is a “failure, refusal, or neglect” to proceed with arbitration covered by FAA Section 4. But the Court held that a refusal to proceed, based on the Panel’s granting that party a stay, was not a refusal to arbitrate within the meaning of the FAA, and presented no occasion for the Court to assist the aggrieved party by entering an order to compel arbitration. Claimant also asked the Court to vacate the Panel’s stay order, but the Court held that the stay order was not an award, even if its practical effect was to freeze the arbitration indefinitely, because the Panel was not functus officio as to that issue and retained power to lift the stay.

The Texas case concerned the “removal” provision, Section 205 of the FAA.  (Foreign readers will recall that “removal” is an idiosyncrasy of American law, concerning the allocation between US federal courts and courts of the individual states of cases over which there is subject matter jurisdiction in both). Following a final award made in Texas in arbitration between brothers who were each a citizen of India but a resident of Texas, the winner sought confirmation in a court of the State of Texas and the loser sought to remove the case to the US District Court in Dallas. The winner argued that case law had construed the Convention as establishing independent criteria of when an award made in the State (i.e. a nation that is a Convention Contracting State) where enforcement is sought falls under the Convention. But the Court rejected this view, holding that the language of Convention Article I (1) (Convention “shall also apply to arbitral awards not considered as domestic in the State where their recognition and enforcement is sought”) left it entirely up to the State (i.e. nation) in which enforcement is sought to define criteria of non-domesticity. The Court held that the FAA had clearly defined what the US regards as non-domestic, in Section 202, and accordingly an award between two foreign citizens residing in the US was a non-domestic award, and fell under the Convention, and the action for confirmation in a court of the State of Texas was therefore properly removed to US District Court.