The 2005 Choice-of-Court Agreements Convention (“Convention”) has been widely promoted by the Hague Conference on Private International Law (“Hague Conference”) and others.  This post continues the discussion in two prior posts (Part I and Part II) in this series which argued that it was inappropriate to transpose the New York Convention’s basis structure and terms to the very different setting of national courts.  This post argues that the Convention also significantly dilutes essential protections that the New York Convention provides for both party autonomy and procedural fairness. In doing so, the Convention significantly exacerbates the risks that arise from the prevalence of judicial corruption and the lack of judicial independence and commercial expertise that exist in many legal systems. The Convention also departs radically from all previous proposals of global judgment recognition conventions, including the 1971 Hague Convention on the Recognition and Enforcement of Foreign Judgments and the 2019 Hague Convention on the Recognition and Enforcement of Foreign Judgments. These and other instruments included mechanisms for dealing with the realities of judicial corruption and the lack of judicial independence (permitting states to provide for the mutual recognition of judgments on a bilateral, rather than global, basis). Strikingly, the Convention abandoned these mechanisms.

Most states have been reluctant to ratify the Convention. As discussed in this post, that reluctance is well-considered. Given the Convention’s serious defects, states should not ratify the Convention and, if they have done so, they should exercise their right to withdraw from it promptly.


The Convention’s Dilution of Safeguards for Party Autonomy

There is no dispute that the principles of party autonomy and consent are fundamental to contemporary private international law regimes and, in particular, to matters of international dispute resolution. Indeed, proponents of the Convention emphasize that it is intended to “protect party autonomy”1)Andrea Schulz, The Hague Convention of 30 June 2005 on Choice of Court Agreements, 2 J. Priv. Int’l L. 267 (2006). and “remov[e] obstacles to productive commercial relations, which are best served by party autonomy.”2)Hans van Loon, The 2005 Hague Convention on Choice of Court Agreements – An Introduction, 18 Annals Fac. L.U. Zenica 12 (2016).

Party autonomy does not, however, mean giving effect to every alleged international arbitration clause or forum selection agreement.  Rather, respect for party autonomy means giving effect to those dispute resolution agreements that commercial parties have in fact validly concluded. As a consequence, the provisions of both the Convention and the New York Convention that govern the treatment of challenges to the existence, validity or scope of dispute resolution agreements – and hence the parties’ consent to a particular forum for adjudication – are of central importance. The Convention does not parallel the New York Convention’s treatment of party autonomy in this respect. Under the New York Convention, the existence, validity or scope of an arbitration agreement can generally be challenged at three stages: (a) in challenges to the validity of the arbitration agreement, in both the arbitral proceeding and litigation in the arbitral seat (and often elsewhere); (b) in challenges to an arbitral award in annulment proceedings in national courts which supervise the arbitral process in the seat of the arbitration; and (c) in challenges to the recognition of the award in proceedings in foreign courts outside the arbitral seat.  The results of any one of these challenges in a particular national court system (or the arbitral proceedings) will ordinarily not have preclusive effect in other jurisdictions.3)Gary Born, International Commercial Arbitration 3797-3808, 3995-4000 (3d ed. 2021).  As a consequence, parties will not be required to arbitrate, nor be bound by an award, unless several independent inquiries into the existence and scope of valid consent to arbitrate have been satisfied, including inquiries by both the arbitrators themselves and by national courts in the recognition forum.

Significantly, the Convention dispenses with inquiries into the existence of valid consent to a choice-of-court agreement that would parallel those of the New York Convention. The existence and validity of a choice-of-court agreement may be challenged under Articles 5 and 6 of the Convention – generally paralleling Article II of the New York Convention. However, if such a challenge is made, and rejected by the putatively chosen legal system, then no further avenues for inquiry into the existence or validity of the agreement are possible in other forums. If the court putatively chosen by a choice-of-court agreement has decided that the agreement exists and is valid under the chosen court’s law, then Article 9(a) provides that the requested court must accept this decision: a judgment may be denied recognition if “the [choice-of-court] agreement was null and void under the law of the State of the chosen court, unless the chosen court has determined that the agreement is valid.”

Relatedly, Article 8(2) of the Convention also provides that “The court addressed [in recognition proceedings] shall be bound by the findings of fact on which the court of origin based its jurisdiction, unless the judgment was given by default.”  Thus, even if the chosen court has not decided on the existence and validity of the choice-of-court agreement, Article 8(2) makes its factual determinations binding in subsequent recognition proceedings.

The consequences of these provisions of the Convention are very significant.  Their effect is to give the national legal system putatively chosen in a choice-of-court agreement the sole authority to decide on the existence and validity of that agreement, without the possibility of review in recognition proceedings. That is a striking contrast to the New York Convention, where recognition courts are granted the authority by Article V(1)(a) to deny recognition based upon the absence of a valid arbitration agreement – notwithstanding an arbitral tribunal’s ruling that such an agreement existed and notwithstanding an annulment court’s decision to the same effect. Given the central importance of consent and party autonomy to both arbitration agreements and choice-of-court agreements, the Convention’s elimination of Article V(1)(a)’s safeguard is highly problematic: it creates a very real risk of parties being forced to litigate in, and bound to judgments by, courts to whose authority they never consented.

The Convention’s treatment of consent is also subject to additional serious criticisms. Under Article 9, there is no provision for denying recognition based upon the chosen court’s excess of authority, including by deciding disputes that are not within the scope of the parties’ choice-of-court agreement. In particular, there is no analog to Article V(1)(c) of the New York Convention, and parallel annulment provisions of national arbitration legislation, which permits a recognition court to deny recognition where the arbitral tribunal made an award that deals with issues not contemplated under the terms of the submission or decisions on matters beyond the scope of the submission to arbitration (ultra petita). However, the Convention eliminates Article V(1)(c)’s protections – leaving the chosen court, of any Contracting State, as the sole judge of the scope of its own jurisdiction. Again, this does not protect, but undermines, party autonomy.

Contrary to the assurances of its proponents, the Convention does not protect party autonomy; instead, it eliminates essential mechanisms for ensuring that the parties’ autonomy is validly exercised and genuinely respected. This is a particularly serious issue in jurisdictions with legal systems of doubtful integrity, independence and competence (in contrast to arbitral tribunals, where the opposite is true, and, in any event, where proceedings are supervised by the courts of the arbitral seat and recognition forums).


The Convention’s Dilution of Safeguards for Procedural Fairness

No less important than respect for party autonomy in international adjudication are requirements of procedural fairness, the basic elements of which include that the failure of a court or arbitral tribunal to respect these principles constitutes a denial of justice and deprives its rulings of both validity and legitimacy.

Again, contrary to its proponents’ commentary, the Convention does not parallel the New York Convention’s treatment of procedural fairness.  Under the New York Convention, an award may be denied recognition if “the party against whom the award is invoked was not given proper notice of the appointment of the arbitrator or of the arbitration proceedings or was otherwise unable to present his case.”  Moreover, many aspects of the procedures in international arbitration are a product of the parties’ consensual arrangements, with the New York Convention again providing for non-recognition of awards if these procedural agreements are not complied with. These protections complement the parties’ rights to “equality of treatment” and a “full opportunity to be heard” under virtually all national arbitration legislation and the availability of annulment of awards for violations of these guarantees of procedural unfairness.4)Born, at 3822-3, 3829.  These protections also parallel the protections that are available in most jurisdictions against foreign judgments rendered in procedurally unfair proceedings.

Together with the parties’ role in the selection of the arbitral tribunal, Articles V(1)(b) and V(1)(d) of the New York Convention and analogous provisions in annulment proceedings provide effective protections for the parties’ due process rights in international arbitration. At both the annulment and recognition phases, the procedural decisions of the arbitrators are subject to scrutiny by national courts – in order to ensure that the proceedings were conducted fairly.

Importantly, the Convention does not replicate these safeguards for procedural fairness. Article 9(d) of the Convention permits non-recognition of a judgment where it was “obtained by fraud in connection with a matter of procedure,” defined as “deliberate dishonesty or deliberate wrongdoing,”5)Trevor Hartley & Masato Dogauchi, Hague Conference on Private International Law Convention of 30 June 2005 on Choice of Court Agreements Explanatory Report 831 (2005).  Although important, this provision is directed only to deliberately fraudulent conduct – not to other denials of procedural fairness, including through incompetent, negligent, inadvertent or biased decision-making by a national court: the provision does not replicate the New York Convention’s protections for due process rights.

In addition, Article 9(e) of the Choice-of-Court allows a court to deny recognition of a judgment if “recognition and enforcement would be manifestly incompatible with the public policy of the requested State, including situations where the specific proceedings leading to the judgment were incompatible with fundamental principles of procedural fairness of that State.”  Article 9(e) provides more extensive protections than Article 9(d), but it too does not provide the safeguards that exist under the New York Convention.

First, Article 9(e) treats procedural unfairness as a subcategory of the public policy of the requested state, prescribing an elevated and two-pronged standard of proof – that recognition of a judgment be “manifestly incompatible” with a state’s public policy – and requiring that the “specific proceedings leading to the judgment” have been “incompatible with fundamental principles of procedural fairness.” By treating procedural unfairness solely as a sub-set of public policy, Article 9(e) dilutes the specific procedural protections that are provided under Articles V(1)(b) and V(1)(d) of the New York Convention.

Second, Article 9(e) of the Convention also limits non-recognition to cases where “the specific proceedings leading to the judgment” were procedurally unfair. By so doing, the Convention forbids inquiry into the fairness and independence of the legal system of the Contracting State whose courts rendered the judgment in question.  In the words of one commentator:

“[Article 9(e)’s] words were chosen with care.  Review may be had in the court addressed of something which may have occurred in the particular case leading to the particular judgment for which recognition and enforcement is sought.  Article 9(e) is not an invitation to a broad scale attack on the nature, character, or alleged conduct of the foreign judicial or legal system as a whole.”6)Ronald A. Brand & Paul M. Herrup, The 2005 Hague Convention on Choice of Court Agreements: Commentary and Documents 118 (3d ed. 2008) (emphasis added).

This approach is seriously flawed in an instrument aspiring to be a global convention: Article 9(e) mandatorily requires recognition of judgments rendered by judicial systems that lack basic guarantees of independence, integrity and competence – which is a characterization that, as discussed above, describes a substantial number of states. Efforts to detect, and prove, judicial corruption are notoriously challenging and seldom successful. Demonstrating corruption in an individual foreign judicial proceeding is even more challenging, because of the difficulties in obtaining evidence, language and other obstacles (i.e., cost, risks of official interference and the like). Likewise, proving governmental interference in individual proceedings is extremely difficult. As a consequence, the Convention’s provisions regarding procedural fairness are virtually certain to prove inadequate as safeguards against the types of misconduct that are endemic in far too many jurisdictions.

The Convention’s treatment of procedural fairness also significantly dilutes the protections that are available under national law in many jurisdictions. Under both common law and civil law standards in most states, courts may deny recognition of awards rendered by legal systems that lack independence or impartiality. The vital importance of these procedural protections is underscored by the U.S. Supreme Court’s classic treatment of common law standards in Hilton v. Guyot:

“Where there has been opportunity for a full and fair trial abroad before a court of competent jurisdiction, conducting the trial upon regular proceedings, after due citation or voluntary appearance of the defendant, and under a system of jurisprudence likely to secure an impartial administration of justice between the citizens of its own country and those of other countries, and there is nothing to show either prejudice in the court, or in the system of laws under which it was sitting, or fraud in procuring the judgment, or any other special reason why the comity of this nation should not allow it full effect, the merits of the case should not, in an action brought in this country upon the judgment, be tried afresh.”7)Hilton v. Guyot, 159 U.S. 113, 202-03 (1895) (emphasis added).

Guarantees of procedural fairness in other jurisdictions are worded differently but are very similar in substance.

These safeguards, paralleling those of Article V(1)(b) of the New York Convention, are not incidental or merely “nice to have”: they are essential attributes of any ruling that is to be given binding effect in a developed legal system. Despite that, the Convention very significantly dilutes the procedural protections of both existing private international law rules in most jurisdictions and the New York Convention. Even if the New York Convention model were considered appropriate for litigation, however, it is extremely difficult to accept the proposition that its procedural protections should be materially diluted for foreign judicial proceedings.



The Convention aspires to be a worldwide charter governing international forum selection agreements and national court judgments and is promoted as a significant milestone in the development of international civil procedure.  Despite these ambitions, there are fundamental defects in the Convention’s structure and terms, making it unsuitable for ratification by jurisdictions committed to the rule of law.

The Convention purports to transplant principles from the New York Convention to cross-border choice-of-court agreements, notwithstanding decisive differences between the international arbitral process and proceedings in national courts.  Moreover, the Convention dilutes critical safeguards that the New York Convention guarantees for both the parties’ autonomy and the procedural integrity of the adjudicative process.  In doing so, the Convention again suffers from serious flaws which makes it unsuitable for adoption on a global scale.

On other occasions, the Hague Conference has acknowledged that judiciaries in many countries lack the integrity, independence, and competence to justify recognition of their judgments, even where those judgments were made pursuant to a legitimate jurisdictional base.  The 1971 Hague Convention on the Recognition and Enforcement of Foreign Judgments included, in Article 21, a provision that the Convention would apply only where two Contracting States had agreed to its application on a bilateral basis.  Similarly, the 2019 Hague Convention on the Recognition and Enforcement of Foreign Judgments included, in Article 29, a provision allowing states to opt-out of the Convention’s application as to any other Contracting State.

In both cases, these provisions applied even where jurisdiction over the judgment-debtor was undisputed, including where it was established by consent.  The reason for these provisions was pervasive doubts about the integrity, independence, and competence of courts in many countries – which led to insistence on provisions allowing Contracting States to opt out of application of the Convention as to such states.  The same conclusions apply equally under the Convention: notwithstanding a valid choice-of-court agreement, there is no justification for recognizing judgments from courts whose integrity and independence are suspect.

Given the grave defects in the Convention, states that have not already adopted the Convention – like the United States, India, Brazil and China – should not ratify it.  States that have already done so, like Singapore and Mexico, should reconsider and, utilizing Article 33, should withdraw from it.  Doing so is necessary to preserve the rule of law and the security of international trade and investment.  Failure to do so will inevitably result in gross procedural and other unfairness to innocent parties and will deter, rather than promote, cross-border trade and investment.


To make sure you do not miss out on regular updates from the Kluwer Arbitration Blog, please subscribe here. To submit a proposal for a blog post, please consult our Editorial Guidelines.

Profile Navigator and Relationship Indicator
Access 17,000+ data-driven profiles of arbitrators, expert witnesses, and counsels, derived from Kluwer Arbitration's comprehensive collection of international cases and awards and appointment data of leading arbitral institutions, to uncover potential conflicts of interest.

Learn how Kluwer Arbitration can support you.

Kluwer Arbitration


1 Andrea Schulz, The Hague Convention of 30 June 2005 on Choice of Court Agreements, 2 J. Priv. Int’l L. 267 (2006).
2 Hans van Loon, The 2005 Hague Convention on Choice of Court Agreements – An Introduction, 18 Annals Fac. L.U. Zenica 12 (2016).
3 Gary Born, International Commercial Arbitration 3797-3808, 3995-4000 (3d ed. 2021).
4 Born, at 3822-3, 3829.
5 Trevor Hartley & Masato Dogauchi, Hague Conference on Private International Law Convention of 30 June 2005 on Choice of Court Agreements Explanatory Report 831 (2005).
6 Ronald A. Brand & Paul M. Herrup, The 2005 Hague Convention on Choice of Court Agreements: Commentary and Documents 118 (3d ed. 2008) (emphasis added).
7 Hilton v. Guyot, 159 U.S. 113, 202-03 (1895) (emphasis added).
This page as PDF


  1. Mr Born’s contribution is one of the poorest I have ever seen from an international lawyer. He writes that the Hague Convention “seeks to transpose the New York Convention’s legal regime” (Part I). This is of course wrong. The Convention’s drafters were simply inspired by the New York Convention and wanted to find a legal regime also efficient in respect of choice of court agreements. It is said that the Convention “ignores the realities of endemic courruption” (I). Equally wrong: there are many safeguards in Article 9 and in particular the public police protection. (I). It is completely wrong to assert that the New York Convention prevents anything like corruption occurring in arbitration proceedings (II). And even if this would be the case, there is no reason not to undertake an effort to ensure the enforcement of foreign judgments based on choice of court agreements considered as valid in a Hague Conference Member State. Precisely, choice of court allows the parties to avoid jurisdictions where corruption may occur. In many respects, the enforcement of arbitral awards under the New York Convention faces the same difficulties, which Mr Born tends to ignore. It is said that “the Convention would replace existing private international rules”: equally wrong, the Convention does not promote an exclusively applicable regime. The validity of any choice of court agreement may be verifies in the enforcement State, the same way as the validity of an arbitration agreement may be checked under the New York Convention. For Mr Born, the Convention dilutes the protections that are available under national law (III). He ignores the large spectrum of protections under Article 9 and the fact that the New York Convention does not do it better. In sum: All of Mr Born’s assertions are FAKE NEWS.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.