International Quarterly — Issue 37

Appeals on a point of law: striking the right balance between finality of arbitral awards and judicial supervision

By Giuseppe Franco, Associate, and Amelia Adams, Paralegal

Implemented over 25 years ago, the Arbitration Act 1996 (the “Act”) is undergoing the review initiated by the Ministry of Justice in 2021. The Law Commission published their final report in September 20231 and, as of December 2023, the Arbitration Bill is under scrutiny of the House of Lords.2 The Bar Council has welcomed the review, emphasising the importance of upholding London’s status as a leading hub for international arbitration.3 Among other topics, the Law Commission took a stand on section 69 of the Act, which covers the appeal of arbitral awards on points of law. 

Section 69 according to the Law Commission

Section 69 of the Act allows arbitral parties to appeal an award on a question of law. To bring the appeal, the agreement of all parties involved in the arbitration is required. Alternatively, a party may seek to obtain permission from the court. This section operates on an opt-out basis, allowing parties to agree to exclude the right to appeal. Parties often waive such right by agreeing on arbitral rules that include a provision to that extent, such as the ICC Rules (Article 35.6) or the LCIA Rules (Article 26.8).

The Law Commission considered amending section 69 reasoning on two opposing goals. On the one hand, it considered that there is a need to enhance the finality of arbitral awards by limiting the possibility of appealing. On the other hand, the Law Commission observed that certain mechanisms of oversight are necessary to correct at least obvious errors of law.

In the end, however, the Law Commission did not recommend any reform of section 69, upholding it as “a defensible compromise between promoting the finality of arbitral awards … and correcting blatant errors of law”.4

Italy’s position towards appeals on questions of law

Akin to the Act, Italian law does not distinguish between domestic and international arbitration, applying the same provisions to all proceedings seated in Italy. Italian arbitration law is enshrined in Articles 806 to 840 of the Code of Civil Procedure (the “CCP”). In particular, Article 829(3) of the CCP deals with appeals for errors in law relating to the merits of the dispute, establishing that an appeal is admissible “if so expressly provided by the parties”. Parties can agree to opt into the appellate mechanism, both in the arbitration clause or in a subsequent, ad hoc agreement. 

It is interesting to note that, until 2006, Article 829(3) stated the opposite, namely that parties were always allowed to challenge the award for errors in law unless they had expressly agreed to exclude such recourse. In other words, with the 2006 reform, Italian arbitration law moved from an opt-out to an opt-in regime, aiming to consolidate the stability and finality of arbitral awards.

An opt-in mechanism for appeals on errors of law is also adopted in other jurisdictions,5 although in most instances, laws simply do not provide parties with appellate mechanisms and only allow for setting-aside proceedings.6


In line with the Law Commission’s recommendation, the Arbitration Bill does not contain any provisions amending section 69 of the Act. However, some consultees have argued that section 69 could benefit from moving to an opt-in mechanism. The main argument for this change affirms that an opt-out regime may be viewed as inconsistent with party autonomy to select arbitration as the only forum to resolve disputes.7 Here, it seems possible to counterargue that the parties’ choice to exclude court litigation would be safeguarded under both regimes, the only difference consisting in the conduct required of the parties to express such choice: an affirmative action to exclude under the opt-out regime; an omission under the opt-in regime. Indeed, in contract drafting, the parties’ deliberate omission to address certain issues can also be seen as a form of risk allocation and decision making.

This argument seems, however, to neglect the reality of drafting of dispute resolution clauses, which have been notoriously nicknamed “midnight clauses”. In what many litigators see as a reckless practice, parties often leave the review of dispute resolution clauses to the last moment (“at midnight”), especially when agreements are long and complex. In light of this risky context, one may understand why it could be beneficial to move to an opt-in regime in order to safeguard the parties’ choice to arbitrate. Indeed, even if parties were to overlook the drafting of the arbitration clause, their intention to resolve disputes by arbitration would need to be nonetheless upheld. Arguably, this can be better achieved through a system that, as a default rule, forbids appeals on questions of law.

A similar reasoning can apply to the argument of stability of arbitral awards, namely that access to appeal should be limited in order to enhance the finality of arbitration. If the goal of stability prevails over that of judicial supervision, it might be more sensible to prohibit appeals on questions of law unless the parties agree otherwise. Limiting the access to appeal has been seen, however, as a restraint on the development of commercial law.8 Reducing the number of applications for appeal would, indeed, mean fewer occasions for courts to develop and shape the law. Yet, this line of argument seems to imply that arbitral parties should owe some kind of higher duty to the advancement of the common law. The question might then be whether it is fair for party autonomy to be sacrificed in the name of higher, public policy reasons. One may suggest that businesspeople should not be held to such an obligation, especially if one considers the great sums that they spend to conduct arbitration.

Finally, the opt-out mechanism appears to suffer from the vexatious practice of bringing (meritless) appeals with the mere goal of exerting financial pressure on the counterparty. Between 2018 and 2021, of the 126 applications filed under section 69, only six (5%) were successful,9 indicating what can be seen as a party’s readiness to exploit the system. If the system were to be reversed to an opt-in mechanism, parties would arguably have fewer occasions to do so.


It feels that both the opt-in and the opt-out regime have their compelling reasons to be. From our perspective, appropriate consideration should be given to the (mal)practice of late drafting of dispute resolution provisions. Bearing this in mind, lawmakers ought to determine their priorities. An opt-in mechanism will probably be a better choice for lawmakers that are determined to consolidate party autonomy and finality. Conversely, an opt-out mechanism will better suit those systems that see judicial supervision as a tool to shape and control the development of the law. 

In case an opt-in approach is favoured, the question then arises as to whether it is still appropriate for courts to retain the authority to grant leave for appeal. While we recognise that this may be beneficial in order to oust unmeritorious appeals, we also consider that this might contradict the rationale of the opt-in regime, which is to promote party autonomy. Indeed, if parties have expressly agreed to opt into the possibility to appeal on a point of law, no further restriction should theoretically be imposed on their right to bring such challenge. As a matter of fact, this is the Italian position under Article 829 of the ICCP, which, indeed, does not require any permission to appeal.

Previous article

Subscribe to our newsletters

If you would like to receive a digital version of our newsletters please complete the subscription form.