Introduction1

The New York Convention is 40 years old, and it is time to discover its full potential. National courts should now be prepared to allow the arbitral process to become truly international, by enforcing awards rendered in other countries on the basis of no other criteria than those which are internationally accepted. To be more specific, I propose that the annulment of an award by the courts in the country where it was rendered should not be a bar to enforcement elsewhere unless the grounds of that annulment were ones that are internationally recognized.

It is moreover my contention that legal bases already exist for disregarding annulments that do not meet this requirement.

The New York Convention was intended to ensure that arbitral awards would be enforced around the world unless the party resisting enforcement proved a fundamental impropriety such as excess of jurisdiction, wrongful constitution of the Arbitral Tribunal, or denial of the opportunity to be heard. Nevertheless, Article V(1)(e) of the Convention allows courts to decline to enforce a foreign award if it has been set aside in its country of origin, without distinction as to the grounds upon which it was annulled. In his seminal work on the Convention published in 1981, Prof. Albert Jan van den Berg noted the danger that the Convention could be 'undermined' if 'all kinds of particularities'2 of the law of the seat of arbitration were allowed to frustrate enforcement.

The anathema of local 'particularities' might conveniently be referred to as 'Local Standard Annulments' or LSAs. An example would be the setting aside of an award on the ground that it was not signed by all arbitrators. Such a requirement (which, to take a concrete case, once existed in Austria but was abandoned in 1983) is clearly at odds with contemporary international standards. Moreover, it contradicts modern rules of arbitration, which ensure that a valid award may be rendered even if one arbitrator refuses to cooperate. A party agreeing to international arbitration legitimately expects that a good claim cannot be defeated by the cynical expedient of appointing an arbitrator-saboteur. The choice of a particular place of arbitration should not facilitate subterfuge by dint of an unusual or internationally unacceptable local rule. (An example of an unusual rule would be one which imposed judicial review of the merits of an award; an example of an internationally unacceptable rule would be one which required all arbitrators to be of a particular religious confession, or to be men.) To those who would say that ignorance of the law is no excuse, one may answer that it would be unfortunate and unnecessary to allow the system of international [Page15:] arbitration to be a minefield for the unwary. Moreover, the place of international arbitration is often chosen by an institution, thus making it irrelevant whether a party is ignorant or well-informed.

It is my conclusion that Article V(1)(e) is not a bar to disregarding LSAs, and that they should be disregarded. Even if Article V(1)(e) were such a bar, it may, as we shall see, be overcome by Article VII.

The discussion is structured as follows. First I describe cases in which LSAs have been disregarded by enforcement courts, focusing especially on the recent and much-discussed Hilmarton and Chromalloy cases. Then I consider, and seek to refute, various arguments that all annulments, even LSAs, must be given effect. Finally, I propose criteria for deciding in what circumstances LSAs should be disregarded.

I. Instances of disregard of LSAs

The matter is simplified in legal systems which apply more liberal criteria for enforceability than those defined in the New York Convention. Since Article VII of the Convention requires application of any national law (or other international treaty) under which the award would be enforceable in the absence of the Convention,3 those legal systems simply disregard Article V.

More than a decade ago, I took the position that in principle there was no legal obstacle to such a result. 4 Despite more than respectable opposition to this view, 5 I have not resiled from it, and consider myself vindicated by the subsequent emergence of a number of cases in which enforcement courts did exactly what I thought could be done. 6 The recent Chromalloy case provides a further useful illustration, involving enforcement in the United States and in France of an award rendered in Egypt even though it had been set aside by the Court of Appeal of Cairo.

The Chromalloy award was rendered in 1994 (for US $16.2 million plus interest) against the Arab Republic of Egypt ('ARE'), in favour of Chromalloy Aeroservices Inc., a U.S. company which had contracted to provide inspection and repair services for helicopters owned by the Egyptian Air Force. The Arbitral Tribunal held by a majority that the contract had been improperly terminated. Two months later, Chromalloy petitioned the U.S. District Court in Washington D.C. for confirmation of the award, whereupon the ARE brought an action before the Court of Appeal of Cairo seeking annulment of the award. [Page16:]

This event provided an occasion to observe that Egypt's 1994 Law on Arbitration in at least one fundamental respect does not follow the UNCITRAL Model Law. In Article 53(1), it adds two uniquely Egyptian grounds for the annulment of awards, namely:

'(d) if the arbitral award fails to apply the law agreed by the parties to the subject matter of the dispute;

[...]

(g) if nullity occurs in the arbitral award, or if the arbitral proceedings are tainted by nullity affecting the award.' 7

The Egyptian Court considered only paragraph (d), which it held sufficient to justify setting the award aside (by an order of 5 December 1995). To reach this result, the Court first determined that the contract properly fell under the category of administrative contracts, and that therefore the contractual reference to 'Egypt Law' (sic) should be understood as an acceptance of the exclusive application of Egyptian administrative law. The Arbitral Tribunal had instead invoked the Egyptian Civil Code (while noting however that neither party had even alleged that a different result would follow if administrative law were applied). It had thus failed, in the Egyptian Court's opinion, to apply the law agreed by the parties and therefore run a foul of Article 53(1)(d).

So the choice facing the District of Columbia federal Court was whether it should enforce the award, or consider itself precluded from doing so by the Egyptian decision. On 31 July 1996, 8 the Court (Judge June L. Green) granted enforcement.

Chromalloy achieved similar success in France. The arbitral award had been granted exequatur (declaration of enforceability) by the Tribunal de Grande Instance of Paris on 4 May 1995, i.e. several months before the annulment of the award by the Court of Appeal of Cairo. The ARE asked the Paris Court of Appeal to overturn the exequatur on the grounds, inter alia, that it violated the 1982 Franco-Egyptian Treaty of Judicial Cooperation. This Treaty provides that arbitral awards rendered in one of the signatory States are enforceable in the other in conformity with the New York Convention. The Paris Court reasoned that the Treaty's unreserved reference to the New York Convention embraced the latter's Article VII, and concluded that the French courts were therefore required to apply French law to enforce the award; French law is more favourable than the Convention inasmuch as it does not recognize the annulment of an award abroad as a ground to refuse to enforce it. The Court stated flatly that as a consequence the Convention was 'put to the side' (écartée). 9

Chromalloy received payment in full in the wake of these decisions, and there were thus no appeals. 10 From the perspective of U.S. law, the District Court's decision is therefore all we have on this subject. It was in fact the first occasion on which a reported attempt had been made to enforce an annulled foreign award before a U.S. court. Much time may pass before the issue is revisited. As for the French decision, it confirms the other cases cited in note 6. [Page17:]

II. The objections to disregarding LSAs

A. 'The New York Convention requires alignment with foreign annulments'

The five subparagraphs of Article V(1) of the New York Convention define cases where applications to enforce foreign awards may be denied. Some readers are led to the hasty conclusion that if any of the subparagraphs apply, the award in question must be rejected. This would then be the case if, as subparagraph (e) puts it, the award:

'has been set aside ... by a competent authority of the country in which, or under the law of which, that award was made.'

This conclusion is, however, incorrect, because Article V(1) allows but does not require rejection of foreign awards that fall under any of the five subparagraphs. The relevant language reads 'recognition and enforcement may be refused ... only if' one of the subparagraphs applies.

The conditional 'may' leaps out at any lawyer, since it necessarily contemplates 'or may not.' There is no other guidance in the text of the Convention. 11 One therefore may fairly conclude that enforcement notwithstanding annulment elsewhere is a matter left to the States parties to the Convention. If such a State incorporates the rules of the Convention into its own law without alteration or qualification, the same discretionary freedom is passed on to the judge. 12

Decisions like those rendered in the U.S. and France in Chromalloy do not violate the New York Convention. The fact is that courts cannot violate the Convention by enforcing a foreign award. To put it the other way, it is impossible to violate the Convention otherwise than by refusing enforcement in the absence of one of the exceptions defined in Article V. This requires some elaboration.

Contrary to what many people seem to assume -quite naturally but just as wrongly - the New York Convention was not intended to establish a regime for the international enforcement of awards. Ways of enforcing awards across borders existed before the Convention. Ways of doing so without applying the Convention not only survive, but have evolved to a significant degree since 1958. For example, in countries which have adopted the 1985 UNCITRAL Model Law, enforcement of foreign awards may be achieved on exactly the same bases as under the Convention and without any need to refer to it. 13 Under Article 1502 of the French Code of Civil Procedure, foreign awards are more liberally enforced than under the Convention -and again with no need to refer to it. The same is also true of the other 26 States which are parties to the European Convention of 1961, all of which are at the same time parties to the New York Convention. [Page18:]

The purpose of the New York Convention was simply to make it easier to enforce foreign awards by establishing a minimum standard for obligating enforcement, not to establish a comprehensive and unitary regime. 14 To the extent that past or future rules allowed enforcement without the Convention, the latter was not intended to intrude; Article VII makes this clear, providing that national rules or indeed other treaties shall be given preference if they are more favourable to enforcement.

In other words, the Convention was not intended to fix things that were not broken, nor to prevent individual legal systems from going further. The question therefore is not whether it is possible for signatories to the Convention to enforce foreign awards even when the Convention would not so require, but whether and when it is wise - or indeed obligatory on another basis - for them to do so.

What then is the appropriate reaction of a court asked to enforce an award which has been set aside by a foreign court in the country where it was rendered?

This question should lead us to recall a core objective of the New York Convention: to free the international arbitral process from domination by the law of the place of arbitration.

• from Geneva to New York

The previous international framework was established by two instruments: the 1923 Geneva Protocol on Arbitration Clauses, Article 2 of which provides that:

'The arbitral procedure ... shall be governed by the will of the parties and by the law of the country in whose territory the arbitration takes place' (emphasis added),

and the 1927 Geneva Convention on the Execution of Foreign Arbitral Awards, which applied to awards based on agreements covered by the 1923 Protocol, and which required for the execution of foreign awards that they be 'final' in their country of origin. This criterion was defined in Article 1(d) as denoting an award which is neither open to any form of recourse nor the subject of pending proceedings contesting its validity. In practice, the only feasible way to demonstrate that an award was final was to obtain exequatur in its country of origin before seeking exequatur elsewhere -creating de facto what came to be called the 'double exequatur' requirement.

The combined effect of the Geneva Protocol and Convention was to make the enforcement mechanism exceedingly arduous, and the choice of an unfamiliar venue for arbitration fraught with danger. The existence of some unanticipated form of recourse in a neutral venue having no connection with the parties, or the unwitting failure to observe peculiar local forms, could destroy the award, or tie it up in years of litigation before enforcement in the practically relevant forum (i.e. the one where the losing party had assets) could even begin. This state of affairs engendered acute frustration in cases where the award was perfectly valid under the law of the enforcement forum. In addition, whenever the arbitral venue was in one of the litigants' home jurisdictions, the Geneva system put the process at the ultimate mercy of the procedural requirements (or even partiality) of that party's courts - the very thing international arbitration is intended to avoid.

The initiative that led to the New York Convention was a 'Draft Convention on the Enforcement of International Arbitral Awards' presented in 1953 to the U.N. Economic and Social Council by the International Chamber of Commerce (which of course is in U.N. parlance an NGO). This draft sought to address the inadequacies of the Geneva Protocol and Convention. As the new Convention took form, the internationalists' camp was naturally enough championed by Frédéric Eisemann, Secretary General of the ICC Court of Arbitration from 1947 to 1973. Mr Eisemann and his allies wanted a treaty that catered for 'international' awards; what they got was a convention on 'foreign' awards. Nevertheless, they succeeded in overturning the two principal obstacles created by the Geneva Protocol and Convention. First and foremost, Articles III and IV of the New York [Page19:] Convention compel enforcement of awards without requiring proof of 'finality' in their country of origin. Secondly, Article V(1)(d) allows the arbitral procedure to be defined by the parties alone; national law comes into play only with respect to procedural issues not covered by their agreement.

These were great achievements, and have had momentous and salutary practical effects; Lord Mustill once wrote that the New York Convention could perhaps 'lay claim to be the most effective instance of international legislation in the entire history of commercial law.' 15 But the Convention's failure to define 'international' awards, and to create an enforcement regime for them, was a real disappointment. This failure is manifested precisely in Article V(1)(e), which Mr Eisemann's successor described in 1978 as:

'a hitherto rock-solid rampart against the true internationalisation of arbitration, because in the award's country of origin all means of recourse and all grounds of nullity applicable to purely domestic awards may be used to oppose recognition abroad...' 16

Thus the New York Convention breathes hot and cold; the 'double exequatur' of the Geneva Convention is no longer required, but annulment in the country of origin may nevertheless be given effect (Article V(1)(e)) - unless the law of the enforcement forum provides otherwise (Article VII). 17

Mr Eisemann's hopes for a greater measure of internationalisation were shared by Prof. Lazare Kopelmanas, who was technical adviser to the U.N. Economic Commission for Europe, and in that capacity influenced the drafting of the 1961 European Convention on International Commercial Arbitration. Under Article IX(1) of this treaty, the fact that an award has been set aside in its country of origin may justify a refusal of enforcement elsewhere only if the grounds for the annulment are among those recognized by the Convention itself. The latter simply reiterates the grounds defined in Article V(1) of the New York Convention - except of course paragraph (e).

• the New York Convention de lege lata

If the drafters of the New York Convention thus balked at the last obstacle, where did they leave us?

The text of the Convention contains a set of absolute obligations on the part of contracting States. Their courts 'shall' defer to arbitration agreements (Article II). They 'shall' recognize and enforce foreign awards presented in accordance with the formalities defined in Article IV without imposing more onerous conditions than on domestic awards (Article III). The Convention 'shall not' prevent enforcement under treaties or laws which are more favourable than the Convention (Article VII). The Convention also contains discretionary provisions: Article V, under which enforcement 'may' be denied under the five familiar paragraphs; and Article VI, under which enforcement 'may' be adjourned pending the outcome of a challenge against the award in its country of origin. 18[Page20:]

This does not seem to require exalted legal analysis. By virtue of the combined effects of Articles V and VII, a country bound by the Convention may require its judges to refuse enforcement under the circumstances defined in Article V, or to the contrary require them to allow enforcement notwithstanding Article V unless stricter criteria for non-enforcement are met, or yet again leave them discretion (more or less explained) to decline enforcement under the circumstances defined in Article V. The French courts in Chromalloy and the other cases cited in note 6 found themselves operating under the second of these possibilities; they were bound to enforce awards that passed muster under French law irrespective of annulment in their country of origin. The same is true for any other country where the law has a narrower definition of defective awards than that of Article V of the Convention, such as the U.S. (if the Chromalloy Court was right19) where an award cannot be neutralized by reference to the objection which underlay the annulment by the Cairo Court of Appeal.

For the same reason, in cases where the European Convention applies any contracting State would be equally bound to enforce awards annulled in their country of origin, unless the annulment had been based on a defect recognized in that Convention.

B. 'An award annulled has no existence'

Mr Schwartz is clearly in this line of thought when he chides the Chromalloy judgment for not having 'addressed explicitly the question of the award's continuing legal existence,' since 'the essential question is whether there still exists an award to which Article VII can be applied.' 20

But this criticism misses the point, I believe, in assuming that the existence of an award is an exclusive matter for the courts of the country where it was rendered. This assumption strikes me as unwarranted. As long as the courts of country X have properly established their jurisdiction in a given case, they may conclude - and have indeed done so since for generations, as any glimpse of a textbook on conflict of laws will confirm - that a contract or a marriage or an adoption is valid and produces effects in country X, even if the relevant act had its origin in country Y and has been annulled there. The same is true for arbitral awards. 21

As we reach the end of the century, there ought not be much wind left in the sails of die-hard territorialists, who cling to the abstract notion that nothing of legal significance can happen anywhere if it is not approved or tolerated by the local sovereign. They consider it heretical for an award to be enforced if it has been set aside in the country where it was rendered. This abstract vision simply does not correspond to reality. Japanese and American businessmen may meet in an Indian airport hotel and sign a contract relating to a European venture which will be given legal effect in Europe without anyone pausing to consider whether the contract is 'heretical' because it has not complied with Indian formalities. If an Indian court intervened to declare the contract null and void it would have no existence in India, but this would not - just because India was the contract's place of origin - prevent another court from taking another view. Similarly, an arbitrator is not an emanation of a sovereign, and when he resolves an international dispute his award may be given effect without necessary reference to the acceptance or tolerance of the legal system of the place where he rendered his award.

Dr Francis Mann considered that no matter how the seat of arbitration had been selected, as long as it was consistent with the arbitration agreement the parties had made a 'direct or indirect, express or implied' choice of the law of that seat as lex arbitri. He took this to be equivalent to a choice between two available court jurisdictions: 'Just as the submission to the sovereign of the forum results in the application of his lex, so the determination of the arbitrator's seat establishes the identity of the lex arbitri.' 22 Dr Mann did not, however, [Page21:] demonstrate the foundation of his syllogism. The problem of course lies in the second of the following givens:

1) submission to the court results in the application of that country's procedural law;

2) a choice of arbitral venue is like submission to a court;

3) therefore a choice of arbitral venue means that the validity of the award will be governed by that country's law.

There is a world of difference between submission to the jurisdiction of a court and the choice of a venue for arbitral proceedings which by their nature are susceptible of being conducted anywhere. The international arbitrator cannot be deemed a manifestation of the power of the State. His mission, conferred by the parties' consent, is one of a private nature, and as Prof. Pierre Lalive has put it:

'it would be a rather artificial interpretation to deem his power to be derived, and very indirectly at that, from a tolerance of the State of the place of arbitration.' 23

This may be the place to say that one of the reasons Dr Mann is remembered so fondly is that the seriousness of most of his scholarship contrasted so charmingly with his penchant for occasional unsupported pronouncements which appeared on analysis to be but whims or idées fixes, irresistibly recalling a line from Le Mariage de Figaro: 'Prouver que j'ai raison serait accorder que je puis avoir tort.' 24

C. 'Inconsistent results must be avoided'

A more pertinent argument, expressed notably by Prof. van den Berg, 25 is based on a prudential perception of the orderly administration of justice. One cannot seriously promote the international arbitral process without considering the need for a safety valve, i.e. a way to neutralize improper awards. The victim of such an award - resulting say from an excess of authority, a violation of the right to be heard, or indeed corruption - ought to be able, under this analysis, to eradicate the evil at its roots, namely in the country where the award was rendered. Unless other legal systems respect that outcome, the consequence might be inconsistent decisions and vast confusion. The Hilmarton case has become a popular paradigm: an ICC award in Switzerland rejected a contractual claim on the basis that the contract contravened Algerian anti-corruption laws; the award was set aside in Switzerland on the (since-abolished) ground that it was arbitrary; the award was nevertheless granted recognition in France where the defect of 'arbitrariness' is not known as a reason to reject awards; subsequently a second award was rendered in Switzerland which rejected the corruption defence and ordered the Defendant to pay a consulting fee under the contract; whereupon both the Swiss judgment (that set aside the first award) and the second award itself were successfully presented for recognition by the French courts - without any retraction of the recognition of the first award. Unquestionably this example will inspire authors of law school examination questions.

The prudential argument based on the orderly administration of justice has surface appeal. Speakers at conferences often amusingly combine it with a hair-raising scenario of conflicting decisions à la Hilmarton which gets the audience quickly to agree that such a grotesque regime surely could never have been intended.

Yet I find the argument specious.

First, the fact is that we can, and do, live with occasional situations of inconsistent decisions. 26 Until the very far-off day of a World Commercial Court of compulsory jurisdiction, there will exist [Page22:] a multitude of circumstances where untidy outcomes are possible. To remain in the context of arbitration, imagine an award challenged by the loser on the grounds of ultra petita, but nevertheless enforced in the Netherlands where an excellent provision of the Code of Civil Procedure (Article 1076(2)) excludes such an objection if it was not raised before the Arbitral Tribunal. Imagine further that the law of the country where the award was rendered does not contemplate such an implicit waiver, thereby preserving the jurisdictional objection, and the award is set aside. A new, entirely proper yet entirely inconsistent award may eventuate, and may in due course be presented in the Netherlands where the first award has already been enforced.

Such things are conceivable, but they are once-in-a-blue-moon occurrences and should not be allowed to turn the world upside down.

Hilmarton is like a two-headed white rhinoceros which might give us a thrill in the cinema but does not really endanger our daily walk to work. Its oddity, I believe, reveals precisely why this concern is more theoretical than real. In most situations in contemporary practice, an award sufficiently defective to be set aside in country A will not pass muster under the autonomous enforcement criteria of country B, whereas the Swiss annulment criterion in Hilmarton ('arbitrariness') was an old and much-criticised concept which would not have been applicable if that case had arisen under current Swiss law. Likewise, in most situations a court granting annulment could also require restitution of any sums paid out in consequence of the award, 27 whereas in Hilmarton the first application for recognition in France was purely tactical, since the award had rejected the claim and there was therefore no enforcement in France, nor any reason to ask for restitution in Switzerland. 28 Last but not least, need we recall how rarely awards are in fact set aside, but how often losing parties mount hopeless tactical challenges? Why should the predicament of a hypothetical party which might face the task of recovering amounts unduly paid cause us greater concern than the frustration of countless real parties whose recovery under perfectly valid awards is delayed and even destroyed by spurious challenges?

The second counter-argument seems conclusive. In order to be intellectually honest, anyone who maintains that inconsistent results are intolerable - no matter how rare in practice or how inconsequential in concrete effect - would have to accept that enforcement should never be granted until any possibility of challenge to the award in its country of origin has been disposed of. To the purist, there cannot be a difference [Page23:] between 'annulled' and 'might be annulled.' 29 In other words, 'may' should be understood as 'shall' in Article V of the New York Convention, and 'binding' in Article V(1)(e) should be understood as 'final,' with the effect that no enforcement is possible until the applicant proves that the award can no longer be set aside in its country of origin. This would in effect take us back to the Geneva Convention of 1927.

That is not all. Rigorous opposition to the possibility of inconsistent results would also require reengineering of Article VI under which enforcement of a foreign award may be 'adjourned' pending the outcome of an annulment application in its country of origin. Article VI is more important in practice than Article V(1)(e), because there are many annulment applications, but very few instances where they succeed. Such considerations should not deter the purist, who as a matter of logic ought to insist that adjournment of enforcement not be left to the vagaries of 'may, if it considers it proper,' but once again be subjected to a mandatory 'shall' in order to ensure that no award will ever be enforced until any possible challenges in the country of origin have been disposed of.

To be faithful to his thought that inconsistent judgments should be avoided, Mr van den Berg ought to insist that adjournments should always be granted. If his prudential argument were truly of great moment to him, he should campaign vigorously against any enforcement until the courts of the country of origin have pronounced themselves on the validity of the award. But to the contrary, Mr van den Berg sets a high standard for parties wishing to impede immediate enforcement, by inviting enforcement courts to make their own evaluation of the grounds presented to the foreign court in support of the application to set aside, and to grant enforcement unless it is shown that 'the award is tainted by a defect which is likely to cause its setting aside in the country of origin.' 30 In 1987, another commentator reviewed five national decisions ruling on applications for adjournment pending challenges elsewhere, and found that in three cases they were denied. 31 Since then, several other courts have also applied the high standard advocated by Mr van den Berg, particularly in England. 32 Such decisions would be unacceptable if the avoidance of inconsistent decisions were given paramount importance.

It is thus incongruous that Mr van den Berg at the same time maintains that adjournments should be granted parsimoniously and that under the New York Convention the setting aside of an award 'has extraterritorial effect as it precludes enforcement in the other Contracting States by virtue of ground e of Art. V(1).' 33 Surpassing strange it would be if the leading exponent of the New York Convention thus turned out to be half-stuck in 1927, in the mindset of the Geneva Convention .... Quandoque bonus dormitat Homerus. 34[Page24:]

D. 'Disregarding annulments insults the courts of the place of arbitration'

Chromalloy invited the U.S. judge to entertain criticism of the Cairo judgment, on the grounds that an evaluation of its bona fides was relevant to the judge's exercise of discretion under Article V of the New York Convention.

If that is the case, how should that discretion be exercised? Should it depend on the apparent bona fides of the annulment? If so, does the concept of bona fides allow the enforcement judge to apply subjective perceptions of the integrity of foreign legal systems, and thus give effect to annulments pronounced by 'good' foreign courts but disregard them when they emanate from 'bad' ones?

By avoiding Article V as a foundation for her decision, the U.S. judge steered clear of such vexatious questions altogether.

By relying on Article VII, Judge Green was able to state in a footnote35 that she assumed the Cairo decision to be 'proper under applicable Egyptian law' but nonetheless go on to enforce the award on the footing that U.S. law does not recognize the alleged infirmity which impressed her Egyptian colleagues.

There should however be no assumption that the courts of the place of arbitration are affronted if foreign judges disregard their opinions about the validity of awards. Prudent judges at the seat of arbitration may be diffident with respect to the erga omnes effect of their authority to set aside. 36 This makes particularly good sense in the case of international arbitration: if neither of the parties nor their dispute has a connection with the jurisdiction of the arbitral venue - which in many cases is precisely the reason for the choice of that venue - the judges there do not have any intrinsic interest in the controversy, or understanding of the culture and expectations of the parties, and thus may treat the matter superficially; if one of the parties is local, there is of course the uneasy paradox that the result of a neutral arbitration may be controlled by one party's home courts. 37

Enforcement notwithstanding an LSA implies no affront to the annulling court; individual systems should be entitled to enact local rules tailored to local perceptions without being encumbered by the thought that they are legislating for the whole world. Here as in many other areas of law, local rules are not automatically applicable to international situations. At the same time, the enforcement forum has the greatest interest in the use of State power with respect to goods situated in its territory, and it would be absurd to speak of insult if one remarks that the actions of its judges, in the absence of an unequivocal treaty obligation, cannot be subservient to what has been decided by foreign judges.

E. 'Parties who have chosen a venue where awards are susceptible to LSAs may have a contractual expectation that such annulments not be disregarded'

It may be - so this argument goes - that parties choose a venue in a country where LSAs are available precisely because they wish to put special limits on the arbitrators. 38 Thus they may welcome the idea that the local courts have the power to annul awards if they disagree with the arbitrators on the merits, even though the availability of such review goes squarely against modern international concepts. If a losing party is able to have the award overturned on this basis, it would thus be wrong for other jurisdictions to deprive that party of the benefits of the annulment it consciously sought to preserve as a remedy.

My sympathy for the expectations (real or invented post facto) of parties who might say that they chose a place of arbitration because its laws provided many ways of correcting wrongful awards is not sufficient to justify architectonic adjustments on [Page25:] their behalf. To do so would operate against the objective of keeping as much as possible of the parties' disputes out of the courts, and I believe that the law and practice of international trade has by now developed a wide consensus around this objective. Of course this affirmation - while it suggests that an alleged intent to opt for broad review of awards should not readily be assumed or inferred - does not meet the objection; are parties not entitled by agreement to establish a regime divergent from the consensus? To this I might answer yes in principle, but my understanding of practical reality leads me to counter with a Benthamite argument. For every case where the parties intentionally seek a maximum-review regime (and I recall few instances of clients articulating such an objective), there will be a hundred cases where an LSA would come as a bitter surprise. I would be astonished if either party in the Hilmarton case had the faintest idea that it was choosing a venue which (still) recognized an unpredictable 'arbitrariness' ground for overturning awards. I would be equally astonished if either party in the Chromalloy case had imagined that an award could be set aside if the arbitrators were to refer to the wrong branch of Egyptian law. Indeed, even if that had been the Chromalloy parties' expectation, the circumstances suggest that the error of Egyptian law was committed not by the Arbitral Tribunal, but by the Egyptian court in applying the new Egyptian arbitration law.

Although these preliminary objections are decisive in my own thinking, I would moreover venture that mythical parties who might seek maximum review do have a solution. Those who approach international arbitration with such ambivalence could solve their problem by wording their arbitration agreement to exclude enforcement in the event of any type of annulment, i.e. even if it is only an LSA - or, why not, going all the way back to the 1920s, and precluding enforcement until the award has been positively and finally endorsed by the courts of the place of arbitration. 39

F. 'To enforce annulled awards violates an implicit allocation of national court authority'

Based on an intricate policy analysis, this theoretical objection has been articulated in considerable detail by Prof. Michael Reisman, in a monograph entitled Systems of Control in International Adjudication and Arbitration. 40 He argues that even if its text may be read to allow enforcement of annulled awards (as demonstrated under (A) above), the New York Convention contains an implicit bargain between contracting States to the effect that courts at the arbitral venue promise to control awards in exchange for the promise of other courts to respect the outcome of that control. To understand Mr Reisman's thesis, it is necessary to consider his short book as a whole. Its genesis was a series of three lectures in 1989, which in turn had their foundation in a series of six articles. The lectures dealt with very different subjects: the International Court of Justice, the International Centre for the Settlement of Investment Disputes (ICSID), and the New York Convention. To give the lectures (and thus the monograph) a leitmotiv, Mr Reisman posited that each of these systems was in a state of acute crisis. Thus, the chapter on the ICJ is subtitled 'The Atrophy of Informal Control Mechanisms;' that on ICSID, 'The Breakdown of International Institutional Control Mechanisms;' and the third, 'The Breakdown of the Control System of the New York Convention.'

But the leitmotiv dissolves as soon as one tries to perceive it. Most strikingly, the chapter on the 'breakdown' of the ICSID mechanism was based on a lecture given and an article published41 just after two much criticised ICSID annulment decisions, but just before matters were by and large set right by a third decision. 42 Mr Reisman added passages to acknowledge this important development, referring to 'a sweeping repudiation' of the initial annulments43 and concluding in a [Page26:] postscript that 'the prognosis is good' 44 - but without changing the title of the chapter or amending the ominous tone of the original essay. 45

In a word, the leitmotiv no longer fits the ICSID example.

As to the ICJ, the author's erudition is not to be questioned. (In fact five of the six articles that provided the basis of the book deal with the ICJ and issues of excès de pouvoir.) Nor do I seek to assess the accuracy of his critique. I wish only to show that this debate is irrelevant to our subject.

Mr Reisman believes that the ICJ has drifted away from the principle of restrictive jurisdiction. He sees this as a dangerous development given the fact that the Court decides its own jurisdiction without any satisfactory further control. Although under the U.N. Charter the Security Council 'may, if it deems necessary' act to 'give effect' to a judgment of the ICJ, which clearly implies that it may give none at all in case of an excès de pouvoir by the Court, this is not an acceptable safety valve. (It is political; it depends on who might be prepared to cast a veto; it provides no remedy of annulment and thus no apparent way to undo an illegitimate res judicata.) The author then describes three judgments which he believes reflect a breakdown of control in the ICJ's jurisprudence, and concludes that this has led to a lack of confidence in it.

Must it be pointed out that the thousands of tribunals which carry out the daily work of international commercial arbitration operate under constraints immeasurably more diverse and numerous than those of the ICJ? In the vast realm of arbitration that falls under the ambit of the New York Convention, the potential sources of judicial intervention to assist or neutralize the arbitral process are practically infinite - and moreover difficult to predict a priori with respect to each relationship subject to the relevant arbitration agreement.

What then is the 'crisis' in the control mechanism of the New York Convention?

Although Mr Reisman concedes that the text of the Convention allows an enforcing jurisdiction to disregard an annulment decision in the country where an award was rendered, he sets out to prove that a:

'coherent control theory and a proper interpretation of the texts require the jurisdictions implicated in the arbitration to perform, as the case may be, a primary or secondary control function under Article V.' 46

He then proceeds to define these concepts of 'primary' and 'secondary,' which are pure products of his own reflections. Without this distinction, he says, all States parties to the Convention would be jurisdictionally equal. Without the notion of a secondary jurisdiction, an 'unscrupulous loser' could run to any favourable court, obtain an annulment of the award, and claim that the annulment was effective worldwide. Without the notion of a primary jurisdiction, the winner of an illegitimate award could engage in endless worldwide harassment of the victim, who would have nowhere to go to seek a once-and-for-all nullification. Therefore, argues Mr Reisman, 'the scheme of Article V of the Convention represents a compact between two distinct groups;' 47 the countries where an arbitration takes place agree that their courts will act as primary jurisdictions to exercise ultimate control of the award, while other countries agree to enforce it, safe in the knowledge that it is subject to control in its country of origin, and with the proviso that they will accept the universal effectiveness of annulment in that country.

Having posited that this is the 'compact' of the Convention, Mr Reisman then explains that the 'crisis' he perceives is due to the violation of this compact by the legislatures of Belgium and Switzerland, and the danger that other countries may be tempted to follow their unacceptable example.

The Belgian violation was to amend Article 1717 of the Civil Code so as to exclude the jurisdiction of Belgian courts to entertain applications for the annulment of awards rendered in Belgium but implicating no Belgian nationals or residents. 48 The [Page27:] Swiss violation was to allow parties to arbitrations in that country, provided that neither is a Swiss national or resident, to agree that there may be no challenge to the award before the Swiss courts. 49

Mr Reisman considers that each of these two systems has for 'self-interested motives' refused to 'perform the functions assigned to it in the New York Convention's scheme.' 50 He asserts that they are acting as 'free riders ... taking advantage of the more stringent jurisdictional practices of other jurisdictions to attract a quantity of arbitral business that would otherwise not have selected their venue.' 51 In his view, a contracting State whose national has been deprived of protection against an unfavourable award might well seek reparation of that injury against Belgium or Switzerland as a treaty violator, i.e. on the level of public international law. He concludes his essay with the recommendation that the other States parties to the Convention 'should protest to those governments over their policy of defection from obligations they have assumed under the Convention.' 52

This appears (pace Reisman) to be a singularly unpromising project for creating diplomatic incidents, let alone seeking international adjudication. To begin with, it is impossible to demonstrate a violation of the explicit text; Mr Reisman simply cannot point to any definition of what he believes to be 'obligations ... assumed under the Convention.' Any complaint would have to be based on his notions of the 'assigned functions,' 'coherent theory,' and implicit 'compact' of the Convention, which not only lack textual foundation, but are contradicted by the demonstrable purpose of the Convention: to facilitate the enforcement of awards. But most of all, there is no true victim. No other country is about to attack Belgium or Switzerland for adopting a hands-off approach; enforcement jurisdictions are in a position to ensure that a defective award will not be effective on their territory. Indeed, they invariably insist on carrying out such a control. In fact Mr Reisman has turned the real world upside down; when complaints arise against the attitude of the courts of the place of arbitration, it is because of their excessive control - as in England under the pre-1979 obligation to 'state a case' to the courts, or as per the Egyptian decision in Chromalloy or the Swiss decision in Hilmarton - and not the reverse.

As for the notion that Belgium and Switzerland are gaining an unfair competitive advantage, the fact is that most parties prudently prefer a reasonable and predictable level of control. That explains why there has been no phenomenal rush to arbitrate in Belgium, and why I have yet to see an arbitration clause (although they surely exist) that uses the facility offered by Article 192(2) of the Swiss law.

The Convention itself does not define the categories 'primary' and 'secondary' as conceived by Mr Reisman, and I find them unwarranted. If anything, the primary jurisdiction should be the one where the economic or other consequences of an award are sought.

Mr Reisman's vision of things would mean that an annulment at the place of arbitration would be binding everywhere - no matter how parochial, or indeed odious, the grounds of the annulment (e.g. blatant xenophobia or religious discrimination). Even as a theoretical matter, this vision is unsatisfactory in that it evades the issue of asymmetry; Mr Reisman does not say that the upholding of an award by the courts of the country where it was rendered should oblige other jurisdictions to enforce it. 53[Page28:]

Given my great respect for the eminent author of the locus classicus, Nullity and Revision (1971), I have taken pains to discover the origins of what seems to be a misguided view of the Convention. It appears that the entire construct of Mr Reisman's third essay is built on the quicksand of an erroneous first premise. At the beginning of his discussion, he writes in describing the regime of Article V that:

'The carefully prescribed and limited grounds for reviewing an award set out in the Convention are the same for primary jurisdictions ... and secondary jurisdictions.' 54

This is of course a fundamental error. It is a misapprehension of the text of the Convention, which fails to define any - let alone 'carefully prescribed and limited' - grounds upon which the courts of the country of origin may annul an award. Moreover, the error also concerns the purpose of the Convention, which as explained above was not to impose a worldwide law of arbitration, but to enhance the enforcement of awards rendered abroad.

Mr Reisman's mistake is thus to have started with the postulate that the New York Convention itself prohibited LSAs. If that was so, this very article would be pointless. The objectives defined in my first paragraph would long since have been attained. Someone in the last 40 years would surely have noticed it.

III. When to disregard LSAs

A court which considers that its national arbitration law justifies the enforcement of an arbitration award must, by virtue of Article VII of the New York Convention, enforce the award without heed to its annulment elsewhere.

This is a good thing. In today's world injustice will result if too much credence is given to the law of the place of arbitration, for two reasons.

First, in some places this appears to incite courts to assist the losing party's attempt to overcome the arbitrators' decision on the merits. Courts in India and Pakistan have taken such regrettable attitudes into an entirely new dimension by purporting to exercise ultimate dominion over awards rendered outside their own territory. 55 They have done so on the ground that the awards in question were rendered 'under' their law, which, if this interpretation were shared by enforcement jurisdictions, would mean that an annulment would create the prospect of a worldwide obstacle to enforcement under Article V(1)(e) of the New York Convention. 56 (The Pakistani position is all the more unacceptable as Pakistan has not implemented the New York Convention; heads I win, tails you lose.) Mischief should not so easily be allowed.

Second, unusual or unacceptable annulments are destructive of legitimate and increasingly settled expectations.

It is nevertheless wise to eschew invidious comparisons. Therefore it surely must be preferable, whenever possible, to use the Article VII approach ('our law requires us to enforce this award even though your courts annulled it') rather than the Article V(1)(e) approach ('we exercise our discretion to enforce an award which your courts annulled even though we have often denied enforcement of awards annulled in other countries, because we are not convinced by what your judge did'). 57

An attractive way to work toward the internationalisation of the arbitral process would be for other national legal systems to follow the French example - as perhaps Belgium and the U.S. already have (if one can rely on two cases) - and to establish criteria for enforcement that do not take account of annulment in the country of origin. 58[Page29:]

This is not, however, the current legislative trend, which is inspired by the UNCITRAL Model Law. Like the New York Convention itself, the Model Law is a great achievement in many other ways but on this particular score perpetuates exactly59 the same half-hearted thinking, because Article 36(1)(a) simply repeats all five paragraphs of the Convention's Article V(1), and thereby proposes a national law on the enforcement of awards which contains the same discretionary hesitation with respect to paragraph (e) of the latter. How this will work out in individual countries remains to be seen, but uniformity is certainly not assured. Allow me to propose the following.

Since neither abstract territorialism nor the fear of inconsistent results poses insurmountable barriers to the enforcement of awards set aside in their country of origin, I suggest that the proper enforcement criterion should be whether the foreign decision was a 'local standard annulment' (LSA) as opposed to an 'international standard annulment' (ISA). 60 The rich experience of international trade law since 1958 has taught us what an ISA is: a decision consistent with the substantive provisions61 of the first four paragraphs of Article V(1) of the New York Convention and Article 36(1)(a) of the UNCITRAL Model Law. 62 Everything else would be an LSA, and entitled only to local effect.

When preparing for the Model Law in 1979, the UNCITRAL Secretariat prepared a 'Study on the New York Convention' which noted that the Convention:

'lends force to reasons which may be rather unexpected due to the disparity of national laws or which may be so much geared to particular local circumstances that their forced recognition in the country of enforcement would seem inappropriate.' 63

In less bureaucratic language, the Secretariat doubted that LSAs should be allowed to impede enforcement around the world.

Accordingly, a proposal was made to adopt the approach of Article IX of the European Convention, with the effect that the annulment of an award would impede its enforcement only if the annulment had been based on a ground of invalidity acknowledged in the Model Law itself. In 1983, however, 'the prevailing view' within the Working Group decided against this approach on the ground that: [Page30:]

'the restriction therein was not generally acceptable and, thus, too ambitious and its application could lead to difficulties.' 64

Today, our understanding of what is 'generally acceptable' and 'too ambitious' has evolved significantly, and I believe that we may now finally deal with the pernicious effects of local peculiarities.

Enforcement courts should consider the grounds upon which awards are set aside, and exercise their discretion to disregard LSAs. Moreover, an LSA does not become an ISA only because the annulling court used the words that correspond to international norms; there should be no reward for the hypocritical recitation of mantras. The enforcement judge, who is after all making a decision which will have practical consequences on resources located in his jurisdiction, has no less 'sovereign' authority to assess whether standards of enforcement have been met than a court in the country where the award was rendered. In sum, he or she may find the annulment decision more or less persuasive, but never automatically controlling.

At the same time, it should be understood that even an ISA need not exclude enforcement. The laws of a number of countries have gone beyond Article V of the New York Convention, which today in some respects appears somewhat old-fashioned and restrictive. 65 Given Article VII of the Convention, those countries are perfectly entitled to enforce awards on a more favourable basis.

The French solution may go too far, inasmuch as no foreign annulment is given effect. Even in the case of an ISA, the party resisting enforcement would have to make a new demonstration to the French court of the ground that justified the annulment (which by definition would also suffice to justify a refusal of exequatur in France). Evidence and argument that satisfied the foreign judge may not satisfy the French court; the latter may, moreover, accept and be convinced by new arguments and evidence in support of the award.

A more moderate treatment of ISAs would be to grant them a presumption of validity, reversing the general rule of the New York Convention and putting the resisting party to the burden of proving that the Arbitral Tribunal did not lack jurisdiction, was not wrongfully constituted, did not prevent a party from being heard, and so forth - i.e. that the foreign annulment was unjustified. 66

The fact that awards may thus suffer occasionally inconsistent fates is not, for reasons discussed above, a fundamental objection. Inconsistent decisions already do occur from time to time; they are the product of a world in which each country's legal system considers itself sovereign. No one has suggested that the fact that the courts of the place of arbitration have rejected an annulment application should require all other countries to enforce the award. 67 If it is not shocking that the enforcement court makes its own determination in deciding not to enforce, why should the reverse situation be of any greater concern? Indeed, the approach I advocate does no more than to accord court decisions from the country of origin equal authority, whether they uphold or annul the award. [Page31:]

The decision to enforce may either be predicated on the view that Article V(1) of the Convention is discretionary, or, pursuant to Article VII, be based on a more favourable law. (The latter would require rejection of Article 36(1)(a)(v) of the Model Law.) This is a secondary issue; national solutions may differ technically but yet reach the same result of disregarding LSAs.

Under my proposed approach, applications to adjourn enforcement actions should not be granted unless the enforcement court considers that an ISA is likely to materialize. 68 In no case should it grant an adjournment if the resisting party cannot show that it has invoked an internationally recognized standard as a ground for annulment and that the circumstances lend credence to its objection.

Conclusion

Participants in international transactions should play according to the same rules. The approach to the New York Convention advocated here will create incentives for national courts to conform to internationally accepted standards. As they come to see the futility of LSAs in the international context, national lawmakers and judges will limit their censure to arbitral misconduct of a kind which is recognized everywhere as justifying non-recognition. The unfortunate (though not intolerable) spectre of inconsistent results will materialize only as a consequence of court rulings in the most obtuse or chauvinistic jurisdictions. The cornerstone of this approach is the modern consensus as to the usefulness of a reliable international arbitral process. This consensus is reflected in the widely accepted UNCITRAL Model Law, as well as in innumerable court decisions, treaties, high-level political declarations, and influential scholarly writings. It should allow us now to achieve, by using the discretionary features built into the New York Convention, something that was premature half a century ago.

1
This article is adapted from the City University of Hong Kong's Eighth Annual Goff Lecture, given by the author at the Hong Kong International Arbitration Centre on 12 November 1997, in which he developed and updated a theme originally explored in 'Rediscovering the New York Convention: Further Reflections on Chromalloy,' Mealey's Int. Arb. Rep., April 1997, at 20, and 'The Case for Disregarding LSAs (Local Standard Annulments) under the New York Convention,' 7 Am. Rev. Int. Arb. 99 (1996). It is responsive to the following matters not dealt with in the two prior articles: the French Cour de cassation's final decision in the Hilmarton case, infra note 28, and the appearance of the following articles: Ph. Fouchard, 'La portée internationale de l'annulation de la sentence arbitrale dans son pays d'origine,' Rev. arb. 1997.329; E. Schwartz, 'A Commentary on Chromalloy: Hilmarton à l'américaine,' J. Int. Arb., June 1997, at 125; G. Sampliner, 'Enforcement of Nullified Foreign Arbitral Awards - Chromalloy Revisited,' J. Int. Arb., September 1997, at 141; G. Delaume, 'Enforcement Against a Foreign State of an Arbitral Award Annulled in the Foreign State,' Revue du droit des affaires internationales 1997.253; H. Gharavi, 'A Nightmare Called Hilmarton,' Mealey's Int. Arb. Rep., September 1997, at 20; and H. Gharavi, 'Enforcing Set Aside Awards: France's Controversial Steps Beyond the New York Convention,' 6 J. Transn. Law & Pol. 93 (1996).

2
A.J. van den Berg, The New York Arbitration Convention of 1958 (1981), at 355.

3
Article VII(1) of the Convention states: 'The provisions of the present Convention shall not affect the validity of multilateral or bilateral agreements concerning the recognition and enforcement of arbitral awards entered into by the Contracting States nor 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 the treaties of the country where such award is sought to be relied upon.'

4
See, e.g., the following articles: 'Arbitre et juge en Suède : Exposé général et réflexions sur la délocalisation des sentences arbitrales,' Rev. arb. 1980.441; 'France and the Arbitral Process in 1980,' Svensk och internationell skiljedom (arbitration yearbook of the Stockholm Chamber of Commerce) 1981.40; 'Arbitration Unbound,' ICLQ 1981.40; 'Procedural Law of the Forum,' Proceedings of the Chartered Institute of Arbitrators, Hong Kong 1981.35; 'Delocalisation of International Commercial Arbitration: When and Where It Matters,' ICLQ 1983.53; 'Arbitration Unbound (...) (...) in Belgium,' 2 Arb. Int. 68 (1986); 'The Extent of Independence of International Arbitration from the Law of the Situs,' in Contemporary Problems in International Arbitration, Julian Lew, ed., 1986.41.

5
See, e.g., F.A. Mann, 'Lex Facit Arbitrum,' in International Arbitration: Liber Amicorum for Martin Domke (1967), reprinted in 2 Arb. Int. 241 (1986); W.W. Park, 'The Lex Loci Arbitri and International Commercial Arbitration,' ICLQ 1983.25.

6
In 1983, I conceded that there was no 'satisfactory all-fours precedent' of an award annulled in its country of origin being enforced elsewhere. I wrote that this was hardly surprising given that 'in an overwhelming majority of cases,' a defect sufficient to nullify an award in its country of origin will also prevent its enforcement elsewhere, 'Delocalisation,' op. cit. note 4, at 60. Nevertheless, it did not take long for exceptional cases to emerge. The first arose in the aftermath of the unlovely SEEE v. Yugoslavia decision rendered in 1956 by a two-member tribunal sitting in Lausanne, which was declared by a Swiss court in 1957 not to be an arbitral award under the definition of local law but was nevertheless held enforceable in France (in 1984!); Court of Appeal of Rouen, 13 November 1984, Rev. arb. 1985.115, English translation in 24 Int. Legal Materials 345 (1985); see J. Paulsson, 'The Extent of Independence,' op. cit. note 4, at 142 et seq. (The Dutch Supreme Court in 1975 had gone the other way, holding that the Swiss court decision should preclude enforcement under Article V(1)(e) of the New York Convention, ibid. at 144-5.) Since then and up to Chromalloy, we have seen: Pabalk Ticaret Sirketi v. Norsolor S.A., French Cour de cassation, 9 October 1984, 112 J. dr. int. 679 (1984), English translation in 24 Int. Legal Materials 360 (1985) (award rendered in Austria in favour of Turkish party partially set aside in Austria on grounds of arbitral excess of power in basing award on equitable rather than legal principles; refusal by a French Court of Appeal to enforce award reversed by French Supreme Court on grounds that under Article VII of the (...)(...) New York Convention enforcement forum has a duty to determine whether its own law would allow enforcement irrespective of Article V of the Convention); SONATRACH v. Ford, Bacon and Davis Inc., Brussels Court of first instance, 6 December 1988, unreported, excerpts in English translation in 15 Yearbook Comm. Arb. 370 (1990) (ICC award rendered in Algiers in favour of U.S. party declared enforceable in Belgium notwithstanding its having been 'overruled' (infirmée) by the Court of Appeal of Algiers; New York Convention held inapplicable because Algeria acceded to it after award had been rendered, and no grounds for refusal of enforcement under Belgian law had been shown); Ministry of Public Works of Tunisia v. Société Bec Frères, Court of Appeal of Paris, 24 February 1994, Rev. arb. 1995.275; excerpts in English in 22 Yearbook Comm. Arb. 682 (1997) (award enforced in France notwithstanding a Tunisian court judgment declaring the arbitration clause to be null and void); Hilmarton Ltd. v. Omnium de traitement et de valorisation (OTV), French Cour de cassation, 23 March 1984, Rev. arb. 1994.327, excerpts in English in 20 Yearbook Comm. Arb. 663 (1995) (award which had dismissed claim on grounds that underlying contract was unenforceable, because contrary to Algerian anti-corruption law, recognized in France notwithstanding its annulment in Switzerland on the ground-subsequently abolished-of 'arbitrariness').

7
These passages are taken from the English translation which appears in the dissenting opinion of the Egyptian arbitrator, Dr El-Sharkawi, Mealey's Int. Arb. Rep., August 1996, at C52. No particular elucidation is provided by the French translation published in Rev. arb. 1994.763; Messrs B. Fillion-Dufouleur and Ph. Leboulanger, in their commentary on 'Le nouveau droit égyptien de l'arbitrage,' Rev. arb. 1994.665, agree that the circular and obscure text of paragraph (g) is likely to lead to applications for annulment.

8
Chromalloy Aeroservices v. Arab Republic of Egypt, 939 F. Supp. 907 (D.D.C. 1996).

9
Decision of 14 January 1997, Rev. arb. 1997.395; Mealey's Int. Arb. Rep., April 1997, B-1 (original text in French, with English translation); excerpts in English in 22 Yearbook Comm. Arb. 691 (1997).

10
Sampliner, op. cit. note 1, J. Int. Arb., September 1997, at 145.

11
The French text of the Convention, equally official, when translated more literally into English reads: recognition and enforcement of the award shall not be refused ... unless etc. Although this version does not provide explicitly and in such cases shall be refused, some commentators (such as Prof. Fouchard, op. cit. note 1, Rev. arb. 1997.329, at 344) believe that an implication to that effect is unambiguous in French. If that were so, Article VII of the Convention would be crucial in a Francophone context. I do not, however, agree that the French text excludes discretion under Article V. Moreover, the three other texts of the Convention (Chinese, Spanish, and Russian) are congruent with the English, and principles applicable to plurilingual treaties militate against une exception française; see J. Paulsson, 'May or Must under the New York Convention: An Exercise in Syntax and Linguistics,' 14 Arb. Int. (Issue No. 2, 1998). Mr Sampliner's useful review of the travaux préparatoires demonstrates that they are inconclusive, and I agree with him that if anything they suggest that discretionary enforcement is permissible; see Sampliner, op. cit. note 1, J. Int. Arb., September 1997, at 149. Given the Convention's pro-enforcement bias (per Article VII) a purposive interpretation would also militate against reading Article V(1) as commanding refusal in the cases there cited. As Prof. Pierre Lalive (Geneva) has impressed upon me in correspondence on this subject, it would violate the fundamental purposes of the Convention to interpret an isolated provision so as to make enforcement more, rather than less, difficult than in the absence of the Convention.

12
Accord, the Chromalloy judgment, 939 F. Supp. at 909; China Agribusiness Development Corp. v. Balli Trading, High Court of England, judgment of 20 January 1997, per Mr. Justice Longmore (unpublished); Delaume, op. cit. note 1, at 254; Schwartz, op. cit. note 1, at 133; H. Alvarez, 'The Enforcement of Arbitral Awards,' Commercial Arbitration - 1996 (Continuing Legal Education, June 1996), at 1.1.19.

13
This does not mean that the enforcement provisions of the UNCITRAL Model Law necessarily cover all cases falling under the ambit of the Convention. For example, countries which have signed the Convention without making a reservation to the effect that they wish to limit its effect to awards rendered in commercial matters have an enforcement obligation which applies to a wider range of cases than that dealt with by the Model Law, since the latter is limited to commercial arbitration. There may also be cases where an award is 'foreign' and therefore enforceable under the Convention even though it is not 'international' in the sense required to bring it within the ambit of the Model Law.

14
Contra Schwartz, op. cit. note 1, at 128, referring to 'the lofty goal' of the New York Convention as being 'the harmonious and uniform international treatment of arbitration awards.' In answer to Mr Schwartz, I would point out that if this had been the modest (not 'lofty') goal, there was no need to eliminate the double exequatur requirement under the 1927 Geneva Convention discussed infra, and thus no need at all for the new Convention. And quite obviously, if Mr Schwartz were correct on this point, the New York Convention would not include Article VII.

15
Arbitration: history and background,' 6 J. Int. Arb. 43, at 49 (1989). The President of the International Court of Justice, Judge Stephen Schwebel, put it this way in 'A Celebration of the United Nations New York Convention,' 12 Arb. Int. 83, at 84 (1996): 'The New York Convention is one of the many, and one of the most important, of the attainments of the United Nations in promoting a more effective and universal rule of law.'

16
Yves Derains, 'Foreword,' Hommage à Frédéric Eisemann (liber amicorum) 5, at 13 (1978) (my translation).

17
Accord, Ch. Jarrosson, case commentary on Hilmarton, Rev. arb. 1994.328, at 332.

18
Finding that Article V does not command a refusal to enforce when there has been an annulment in the country of origin, the U.S. judge in Chromalloy focused on Article VII, which she found to contain an obligation to grant enforcement in the circumstances of the case before her. Thus Article V - and the Egyptian annulment - became irrelevancies: 'While Article V provides a discretionary standard, Article VII of the Convention requires that, 'The provisions of the present Convention shall not ... deprive any interested party of any right he shall have to avail himself of an arbitral award in the manner and to the extent allowed by the law ... of the country where such award is sought to be relied upon.' In other words, under the Convention, [Chromalloy] maintains all rights to the enforcement of this Arbitral Award that it would have had in the absence of the Convention.' 939 F. Supp. at 909-910. This reasoning led the judge to consider whether the award was enforceable as a matter of the U.S. Federal Arbitration Act, under which awards are 'presumed to be binding' and may be vacated (i.e. annulled) under only 'very limited circumstances.' The ARE's complaint against the award suggested only one of the grounds for annulment recognized by the U.S. Act, namely 'manifest disregard' of the law. In the case at hand, the arbitrators had responded to the ARE's argument that administrative law should apply by saying that they would reach the same result irrespective of whether they applied administrative or civil law. 'At worst, this decision constitutes a mistake of law, and thus is not subject to review by this Court,' ibid, at 911.

19
Mr Schwartz says no, op. cit. note 1, at 131-2; contra Sampliner, op. cit. note 1, J. Int. Arb., September 1997, at 1523. I express no view here, because my proposal does not depend on what may or may not be the correct current legal position in one particular country.

20
Op. cit. note 1, at 131. At the time the text of the New York Convention was emerging, Prof. Pieter Sanders wrote that once an award has been set aside by the courts of the country where it was rendered, other courts 'will ... refuse the enforcement as there does no longer exist an arbitral award,' in 'New York Convention on the Recognition and Enforcement of Foreign Arbitral Awards,' 6 Netherlands Int. L. Rev. 43, at 55 (1955). It is striking that Prof. Sanders, recognised as one of the fathers of the Convention, did not base his conclusion on anything in the text of Article V(1), but on an extrinsic (and I believe untenable) assumption. This confirms that Article V(1) does not preclude enforcement.

21
See note 26.

22
Op. cit. note 5, 2 Arb. Int. at 248 (1986).

23
P. Lalive, 'Les règles de conflit de lois appliquées au fond du litige par l'arbitre international siégeant en Suisse,' Rev. arb. 1976.155, at 159 (my translation).

24
Act I, scene I, Suzanne.

25
See, e.g., The New York Convention of 1958, at 355 (1981).

26
Accord, Ch. Jarrosson, Case Comment, Rev. arb. 1995.653. As Prof. Fouchard points out, op. cit. note 1, at 344, this issue is not limited to the fate of awards; contracts, marriages, or adoptions may also be annulled in one country but recognised in another. Harmonious results constitute a very important objective, but there are others as well which may have countervailing effects. I would add that inconsistent treatment of awards may occur even if an award is undisturbed in its country of origin; there have been a number of instances where enforcement has been allowed in a second country but denied in a third.

27
Accord, M. Tupman, 'Staying Enforcement of Arbitral Awards under the New York Convention,' 3 Arb. Int. 209, at 220 (1987). This was in fact exactly the position taken by the applicant in Nu-Swift v. White Knight et al., judgment of the Paris Court of Appeal, 21 January 1997, extracts in Rev. arb. 1997.429, full text and English translation in Mealey's Int. Arb. Rep., February 1997, at E-1, who had lost an ICC award in Paris and been compelled by the High Court in London to satisfy the award notwithstanding the pending annulment action in Paris. Accordingly the applicant asked the Paris court not only for annulment of the award, but for restitution of the sums paid in England. As happens in most cases everywhere, the award was upheld.

28
As for the fact that the Hilmarton saga littered the French countryside with exequaturs of irreconcilable Swiss awards and judgments, that is attributable to errors of French law committed by some of the French courts seized of the matter, and is of little interest to our analysis of the New York Convention. See Ch. Jarrosson, op. cit. note 26, at 651. The French confusion was effectively corrected by a final Cour de cassation decision of 10 June 1997, Rev. arb. 1997.376, excerpts in English in 22 Yearbook Comm. Arb. 696 (1997), which held that the recognition of the first award was res judicata, and that as a result the Swiss annulment decision and the subsequent second award could not be given effect in France. Mr Gharavi's comment on this dénouement is entitled 'A Nightmare Called Hilmarton,' op. cit. note 1, but contains no explanation why it should be considered as such. He writes, as though the use of ominous words might substitute for a reasoned critique: 'The practice of enforcing set-aside awards ... will ... continue to haunt the international arbitration community, which seems to underestimate this major 'dysfunction' which is progressively damaging the credibility of international arbitration,' at 22. These are portentous but hollow words. Nothing shocking happened in Hilmarton; the story started with a claimant losing his case in Geneva before an arbitrator, and it ended in France with a rejection of the same claimant's attempt to satisfy the same demand. (As for the possibility that the second award would be enforced elsewhere, I can only repeat that even if such an unlikely event were to occur, the international legal system can accommodate and survive such occasional inconsistencies.) What was 'dysfunctional' in Hilmarton was the Swiss court's decision to overrule the arbitrator on the merits, just as the Egyptian court's decision was the 'dysfunctional' element of Chromalloy. The cure in each case was to disregard the LSA on the strength of a more liberal regime in the enforcement forum.

29
Contra, Schwartz, op. cit. note 1 at 133, n. 31, pointing out that enforcement of an award that 'might be annulled' leaves open the possibility of restitution. In an international context, however, the remedy of restitution is often illusory: what is paid is gone. On the other hand, I accept Mr Schwartz' point when enforcement has been granted subject to the posting of security that would reestablish the status quo ante in the event of an annulment. Such conditional enforcement is not the subject of this study. At any rate, the condition of security would violate the New York Convention in the event the party seeking enforcement has an unfettered right thereto by virtue of Article VII. It is my proposal that such a condition should also be considered inappropriate when granting enforcement irrespective of Article V(1)(e) if the challenging party is seeking an LSA.

30
The New York Arbitration Convention of 1958, at 353-54 (1981). As to whether a legal provision of the country of origin to the effect that recourse against an award automatically suspends its enforcement may be considered a 'suspension' for the purposes of Article V(1)(e), Prof. van den Berg (in my view correctly) takes it for granted that this would clearly be wrong, 'Consolidated Commentary,' 21 Yearbook Comm. Arb. at 501 (1996).

31
Tupman, op. cit. note 27.

32
See e.g. White Knight et al., supra, note 27 (referring to applicant's request for reimbursement of sums paid under award by virtue of unpublished enforcement order of 14 July 1995 by English High Court); Soleh Boneh Int. Ltd. et al v. Republic of Uganda et al, 2 Lloyd's L. Rep. 208 (1993) (security rather than enforcement should be ordered if invalidity of award was 'seriously arguable'); Far Eastern Shipping Co. v. AKP Sovcomflot, 1 Lloyd's L. Rep. 520 (1995) (enforcement ordered notwithstanding continuing attempts to challenge maritime award in Russia); Ungerer v. Rapid Metaal B.V., judgment of 's Hertogenbosch Court of Appeal (Netherlands), 15 June 1994, excerpts in 20 Yearbook Comm. Arb. 750 (1995); Ukrvneshprom State Foreign Economic Enterprise v. Tradeway, 1996 U.S. Dist. LEXIS 2827 (SDNY, 12 March 1996) (award enforced against U.S. party notwithstanding pendency of its application to court in Ukraine to have award set aside); and Kersa Holding Co. v. Infancourtage et al., judgment of Luxembourg Court of Appeal, 24 November 1993, 21 Yearbook Comm. Arb. 617 (1996) (no adjournment of enforcement pending criminal action for fraud in the inducement brought in country where award rendered).

33
Consolidated Commentary,' 19 Yearbook Comm. Arb. 590 (1994). The same categorical (and in my view erroneous) statement is made by Mr Gharavi, op. cit. note 1, 6 J. Transn. Law & Pol. (1996), at 100: '... the setting aside of the award has ... extraterritorial effect insofar as Article V(1)(e) precludes enforcement of an award in other contracting states.'

34
I am of course playing with fire, and taking the risk that Mr van den Berg may find some way of throwing Pope's 'Essay on Criticism' back at me: 'Those oft are stratagems which errors seem, Nor is it Homer nods, but we that dream.'

35
939 F. Supp. 907, at 914, n. 6.

36
Belgium of course has ensured by law that its courts cannot entertain any action to set aside an award rendered in Belgium if the case involves no Belgian national or resident; see text at note 48.

37
As Professor Fouchard puts it, 'it is not legitimate to reserve to the judges of one state the power to destroy awards which displease them throughout the world,' op. cit. note 1, at 348 (my translation).

38
I am indebted to Peter Caldwell (Hong Kong) and Bo Nilsson (Stockholm) for putting this argument to me in separate communications; I make no claim that they will be satisfied by my answers. Mr Sampliner also gives weight to the desideratum of not frustrating the 'intent' of such parties, op. cit. note 1, J. Int. Arb., September 1997, at 162.

39
Some readers may find this discussion redolent of the Kyocera problem familiar to U.S. specialists. See, for the latest episode, Kyocera Corp. v. Prudential-Bache et al, judgment of 9 December 1997, U.S. 9th Cir., Mealey's Int. Arb. Rep., December 1997, at F-1. But in this connection no issue arises as to whether parties may contractually impose upon courts a regime for the review of awards different from the one defined by the law. Rather, the question is whether parties should be allowed to opt explicitly for special post-arbitral control of the award in accordance with relevant law (i.e. an LSA) and if so whether an LSA in such circumstances should be internationally accepted because it accords with the agreement to arbitrate.

40
Duke University Press (1992).

41
'The Breakdown of the Control Mechanism in ICSID Arbitration,' 4 Duke L. J. 739 (1990).

42
Maritime International Nominees Establishment (MINE) v. Guinea, ad hoc decision of 22 December 1989, 5 ICSID Rev. -For. Investment L. J. 95 (1990), discussed in J. Paulsson, 'ICSID's Achievements and Prospects,' 6 ICSID Rev. - For. Investment L. J. 380 (1991).

43
At page 79.

44
At page 106.

45
Thus, at page 87: 'A viable and useful dispute resolving instrument is now in peril because its control system has spun out of control.'

46
At page 116 (emphasis in the original).

47
At page 119.

48
See 'Arbitration Unbound in Belgium,' op. cit. note 4.

49
See Article 192(2), Swiss Private International Law Act.

50
At page 133. Mr Reisman is not shy about laying the blame for what he views as these legislative aberrations where he thinks it lies: the self-interest of local lawyers stimulated by the 'attractive fees for that part of the bar' which is active in international arbitration. But this makes no sense. The elimination of any possibility of challenges means that no local lawyers are needed in connection with arbitrations in Belgium and Switzerland that fall within the scope of the criticised laws; there is no local rule to be complied with, nor any possibility for them to be engaged to apply for a local annulment. It would make more sense to accuse the local bar of foolishness than of greed. The villains must be elsewhere. But surely this beer is too small to have mobilized the Belgian and Swiss hotel lobbies?

51
At page 131.

52
At page 140. It is unclear how the present law in France has escaped Mr Reisman's censure. After criticising the French Norsolor decision, cited in note 6, supra, as 'inconsistent with the obligations of a secondary jurisdiction,' at page 122, he goes on to approve Article 1502 of the French Code of Civil Procedure ('a legislative instruction to national courts to perform control functions with regard to international commercial arbitration within the meaning of Article V ...,' at page 125) as though he has not noticed that Article 1502 does not list annulment in the country of origin among the exclusively enumerated grounds for refusing enforcement. If Switzerland and Belgium should be attacked for failing to control awards, surely France should be just as aggressively pursued for failing to respect decisions of 'primary' jurisdictions which have performed their controlling function and set an award aside.

53
Cf. note 67.

54
At page 114.

55
See M. Kerr, 'Concord and Conflict in International Arbitration,' Arb. Int. 1997.137; J. Paulsson, 'The New York Convention's Misadventures in India,' Mealey's Int. Arb. Rep., June 1992, at 18.

56
Happily the 1996 Indian Arbitration Act has repaired the textual defect which facilitated such overreaching, but one still finds Indian judges willing to issue injunctions against international arbitral institutions and arbitrators conducting proceedings outside India.

57
The interests of international respect, morality, and intellectual discipline are better served by foreigners saying 'if your courts misapply the law we will avoid them' than by their saying 'your courts can do what they like; we will disregard them if they step out of bounds.' As Judge Schwebel has put it: 'If the courts of a State bend the Convention to suit domestic parties, penalties of the market will substitute for those of the law,' op. cit. note 15, at 86.

58
The ultimate irony here might be that the equivalent of NYC Article VII was already present in Article 5 of the Geneva Convention. In other words, the Chromalloy (...) (...) case would have been decided the same way in both France and the U.S. if the New York Convention had not existed, and only the Geneva Convention had applied. This should not however cause anyone to leap to the conclusion that Mr Eisemann and his supporters might better have forgotten about the New York Convention and instead sought to convince individual countries to adopt laws like the 1981 French decree by which Article 1502 was inserted into the Code of Civil Procedure.

59
Albeit applicable to a slightly smaller category of awards, because the Model Law applies only in defined 'commercial' and 'international' situations.

60
One commentator who disapproved the U.S. Chromalloy decision nevertheless conceded that Article V(1)(e) is 'outdated' and 'incomplete' in failing to 'harmonise the grounds for setting aside awards,' and urges the view that reform must be effected on 'an international level,' H. Gharavi, 'Chromalloy: Another View,' Mealey's Int. Arb. Rep., January 1997, 21, at 24. Mr Gharavi thus calls for a new treaty to effect the desired reform. This seems a remote possibility; an evolution in the law and jurisprudence of enforcement fora in the manner described in this article seems more plausible. Moreover, the wisdom of such an approach is questionable. Presumably the search for uniformity would not only define internationally acceptable grounds for annulment in the country of origin, but would also impose a duty on enforcement courts to accept such annulments. In other words, Article VII would have to disappear. This would have several disadvantages. Enforcement courts, which after all have the most natural claim to authority over use of State power to enforce legal rights with respect to property located within their own national territory, would be subjugated to the will of courts of the place of arbitration, which, as experience has unfortunately shown, are sometimes guilty of egregious chauvinism. Moreover, this approach would also bar substantive evolution of the criteria by which awards are evaluated. For example, it would be most unfortunate if the courts of the Netherlands were prevented by such a uniform regime from applying the salutary(...)(...) Article 1076 of that country's Code of Civil Procedure, which deems objections to awards to have been waived if they could have been raised in the arbitration, but were not.

61
To keep LSAs from returning through the back door, one must disregard the incidental renvois to the law of the venue in paragraphs (a) and (d).

62
Mr Sampliner objects, op. cit. note 1, at 162, to the proposal of disregarding LSAs on the grounds that Article V(1)(e) would be redundant 'if the only grounds upon which nullification decisions could be honoured were the grounds already set forth in Articles V(1)(a-d).' This critique misses the point that the New York Convention establishes a minimum standard, not a uniform regime for the international enforcement of awards. Actus interpretandus est potius ut valeat quam ut pereat (the presumption against superfluous terms) has no place here as a rule of interpretation. A treaty providing that children may not work more than 10 hours per week is not violated by a signatory State which precludes child labour altogether. There are in fact a number of elements within Articles V(1)(a-d) which are progressively being rendered irrelevant by modern legislation; see Fouchard, op. cit. note 1, at 347.

63
Doc. A/CN.9/168, 20 April 1979, para. 43, quoted in H. Holtzmann and J. Neuhaus, A Guide to the UNCITRAL Model Law on International Commercial Arbitration: Legislative History and Commentary 1069 (1989).

64
Third Working Group Report, Doc. A/CN.9/233, 28 March 1983, para. 171, quoted in Holtzmann and Neuhaus, op. cit. note 63, at 1086-7.

65
Fouchard, op. cit. note 1, at 347.

66
This interesting suggestion has been made to me by Prof. Pierre Mayer (Paris), who notes that such a reversal of the burden of proof would have the same practical result as the mechanism of pouvoir de révision by which French courts once evaluated foreign judgments. Prof. Mayer also proposes, with respect to ISAs, a test akin to that known in French administrative law as erreur manifeste d'appréciation. These and other ways of assessing ISAs merit further study.

67
But see B. Leurent, 'Reflections on the International Effectiveness of Arbitration Awards,' 12 Arb. Int. 269, at 2845 (1996) suggesting that such a result should be sought by means of a treaty among countries whose legal systems are 'comparable' and characterized by a reciprocal 'high level of confidence.' He proposes as a 'natural circle' the signatories to the Brussels, San Sebastian, and Lugano Conventions on court judgments. Mr Leurent's suggestion is interesting and perhaps salutary, but since it would at best be limited to countries having such confidence in each others' legal systems that they are willing to conclude multilateral judgments conventions, it would, at least for the foreseeable future, exclude important countries like China, India, Russia - and the United States. In this connection, one might reflect on the final argument raised by the ARE in Chromalloy, namely that the U.S. court should recognize the Cairo decision as res judicata. The U.S. does not have a treaty with Egypt concerning the enforcement of judgments (nor indeed with any other country). Nevertheless, U.S. enforcement of foreign judgments is possible under case law, going back to the hoary precedent of Hilton v. Guyot (1895). Such a possibility may however be entertained only if certain criteria are met, including that the original claim must not violate U.S. public policy. In this case, the Court found that the original claim in Egypt (i.e. the claim for annulment) violated the 'unmistakable' U.S. public policy in favour of final and binding arbitration of commercial disputes.

68
Mr Gharavi has proposed that '[i]nstead of blindly and automatically applying the more-favourable-right provision' of Article VII, the French courts should suspend enforcement in cases such as Hilmarton pending the outcome of the annulment application, op. cit. note 1, J. Transn. Law & Pol. 93 (1996) at 106. But Article VII, as the U.S. court in Chromalloy noted, see note 18, is not discretionary; courts must give parties the benefit of their more favourable right. Those who dislike Article VII should promote disentitlement to more favourable rights, not criticise courts for respecting the law 'blindly and automatically.' There is obviously no reason to adjourn enforcement of an award pending the outcome of an annulment application in cases where an annulment per se would not impede enforcement. Such is the case in France. Indeed, in Hilmarton the Swis (...)(...) annulment had already been decreed when the award was presented for recognition in France - a fact that Mr Gharavi's recommendation seems to overlook. Adjournment would have been nonsense; there was nothing to wait for.