THE ARB-MED-ARB PROTOCOL

• Author: Cheryl NG LLM Candidate (Columbia Law School).
• Date: 01 December 2020
• Published date: 01 December 2020
• Citation: (2020) 32 SAcLJ 124

Promising in Concept, Problematic in Design

Since its introduction on 5 November 2014, the Arb-Med-Arb Protocol has been enthusiastically promoted by the Singapore International Arbitration Centre, Singapore International Mediation Centre and various members of the Singapore government. Arbitration practitioners in Singapore appear generally to recognise its promise and utility, but also be cautious about how it would play out in practice. Two articles have recently been published in the Singapore Law Gazette, addressing what their authors perceive to be some “kinks” in the design of the protocol. This article aims to take a deeper look at the design of the protocol and to evaluate and suggest possible fixes to its potential problems.

1 At the Ninth Regional Arbitral Institutes Forum in Kuala Lumpur held on 9 May 2015, the former Attorney-General for Singapore, V K Rajah SC, delivered a speech in which he predicted that “the future [of international arbitration] belongs to hybrid dispute resolution mechanisms which marry both adversarial and consensual forums”.1 These hybrid mechanisms are more commonly known as “integrated” or “multi-tiered” dispute resolution systems,2 in which parties contractually agree to resolve their disputes by mediation or negotiation (consensual means) before proceeding to arbitration (adversarial means).3 Parties may also adopt a “reverse approach”, starting with arbitration and proceeding to mediation.4

2 One of the more prominent additions to the variety of hybrid mechanisms in recent years is an innovation known as the Singapore International Arbitration Centre (“SIAC”)-Singapore International Mediation Centre (“SIMC”) Arb-Med-Arb Protocol (the “AMA Protocol”). By way of a quick introduction,5 this is a three-stage process involving: (a) the initiation of arbitration proceedings under the SIAC; (b) the stay of the arbitration and submission of the case to mediation under the SIMC; and (c) the referral of the case back to arbitration for (i) the recording of a consent award, if the dispute was wholly or partially resolved in the mediation; and/or (ii) the full and final resolution of the dispute by arbitration, if the dispute was only partially or not at all resolved.6

3 The AMA Protocol was introduced by the SIAC and the SIMC on 5 November 2014.7 Since its introduction, it has been enthusiastically promoted by the SIAC, SIMC8 and various members of the Singapore government.9 Arbitration practitioners in Singapore and the region have generally welcomed the introduction of the AMA Protocol, recognising its utility in promoting the early resolution of disputes through mediation10 and ensuring the enforceability of mediated settlement

agreements (“MSAs”).11 That general optimism has been tempered by a sense of cautiousness, as some practitioners seem to prefer to see how the resolution of disputes in accordance with the AMA Protocol plays out in practice before recommending its adoption.12

4 Although the AMA Protocol has attracted a fair amount of attention, it does not appear to have been the subject of extensive study. While there exists some literature on the AMA Protocol, much of what has been written is in the nature of client bulletins aimed at informing commercial parties about the existence and features of the protocol and educating these parties about the general advantages and disadvantages of adopting the protocol as their dispute resolution process of choice.13 The limited number of academic papers on the AMA Protocol similarly focus on the features and general advantages and disadvantages of the

protocol.14 Many of these bulletins and papers highlight the relative pros and cons of the protocol over other hybrid mechanisms available to parties.

5 Given the existing literature, this article will not focus on the strengths and weaknesses of the AMA Protocol as a hybrid mechanism that parties can choose. Instead, it will evaluate the AMA Protocol on its own terms and with reference to its own aims. The AMA Protocol might, in concept, be a promising alternative to existing hybrid mechanisms and a more appropriate method of dispute resolution for parties for a variety of reasons specific to the relevant parties.15 However, if something about its design means that it is not working as intended or if there are hidden costs to its adoption, then these should be fixed. Two recent articles in the Singapore Law Gazette already address the issue of “kinks” in the AMA Protocol, but they are brief and possibly flawed.16 This article aims to take a critical look at the two Law Gazette articles and to build on the ideas in these articles.

6 Part I17 of this article attempts to set out an understanding of the following aspects of the AMA Protocol: its (a) nature; (b) existing design; and (c) aims. This discussion is intended to fill the gaps in the existing literature and to provide an understanding that will underpin the analysis in Part II18 of this article. Part II identifies the shortcomings of the AMA Protocol in relation to two issues that might be expected to arise in an arbitration — (a) jurisdictional challenges; and (b) interim measures —

and evaluates the proposed amendments to deal with these issues; and recommends a possible fix for these issues.

7 The increased interest in hybrid mechanisms in general, and the blend of optimism and cautiousness concerning the AMA Protocol in particular, make the protocol a subject worthy of further study. If any design flaws in the AMA Protocol can be identified and fixed, then transnational commercial dispute resolution would be all the richer for having an additional tool in their toolkit of dispute resolution mechanisms that is optimised for effective resolution. If, on the other hand, these design flaws are allowed to lie unchanged, then the consequences of such design flaws might go beyond the particular set of commercial parties first caught unawares by the problem. Instead, these design flaws might engender disillusionment on the part of the watchful community of arbitration practitioners.

I. The AMA Protocol

8 A study of the design flaws of the AMA Protocol should begin, first and foremost, with an understanding of (a) the nature of the protocol; (b) the existing design of the protocol, which provides a baseline for the evaluation of the protocol; and (c) the aims of the protocol, which provide the yardstick by which the existing protocol and any proposed amendments can be evaluated.

A. Nature of the AMA Protocol

9 What is the nature of the AMA Protocol? As its name suggests, the AMA Protocol is neither a dispute resolution clause nor a body of procedural rules, at least as those terms are conventionally understood. Instead, it is a “protocol” — a seemingly sui generis innovation that appears to combine elements of both dispute resolution clauses and procedural rules.

10 The AMA Protocol is not quite a dispute resolution clause, because it only takes effect if there is already a pre-existing agreement between parties to submit their dispute for resolution under the AMA Protocol.19 This agreement can take the form of the SIAC's and SIMC's

Singapore Arb-Med-Arb clause20 (“the Singapore AMA Clause”) or some other similar clause (that is, an “AMA Clause”) or a specific agreement by parties to submit their dispute for resolution under the protocol (“Submission Agreement”). However, it has elements of a dispute resolution clause, in that it provides that parties “agree that any dispute settled in the course of the mediation at the [SIMC] shall fall within the scope of their arbitration agreement”.21

11 The AMA Protocol is also unlike conventional bodies of procedural rules, in that it is not designed to be self-contained and/or standalone. While it makes some prescriptions as to the procedural steps by which a dispute between parties is intended to be resolved, it also provides that the rules applicable to the arbitration (“the Arbitration Rules”)22 and the SIMC Mediation Rules will govern some aspects of the dispute resolution process, like the constitution of the arbitral tribunal and the commencement of the mediation. In this regard, the AMA Protocol provides that the Arbitration Rules must be either: (a) the arbitration rules of the SIAC,23 as may be revised from time to time (“the SIAC Rules”) or (b) the UNCITRAL arbitration rules,24 as may be revised from time to time (“the UNCITRAL Rules”), where parties have agreed that the SIAC shall administer the arbitration.

12 The sui generis nature of the AMA Protocol and the gaps in the existing literature leave unanswered many questions about the relationships between: (a) the protocol and the parties' agreement on the matter of dispute resolution (as captured in the AMA Clause or Submission Agreement); and (b) the protocol and applicable procedural rules (that is, the Arbitration Rules and the SIMC Mediation Rules).

13 If the AMA Protocol is considered to be part of, or more akin to, the procedural rules chosen by the parties, then an arbitrator and mediator might have the power and discretion under the Arbitration Rules or SIMC Mediation Rules to deviate from or add to the order of proceedings contemplated in the protocol by, for example, deciding to

hear jurisdictional challenges or applications for interim relief before staying the arbitration. This may mean that the provisions of the AMA Protocol might be more flexibly applied depending on the specific circumstances of the dispute.

14 If, on the other hand, the AMA Protocol is considered to be incorporated into the parties' agreement to arbitrate, then the provisions of the protocol (whatever their effects are interpreted to be) might override an arbitrator's general discretion to “conduct the arbitration in such manner as it considers appropriate”25 or “direct the order of proceedings”.26 This is because Art V(1)(d) of the Convention on the Recognition and Enforcement of Foreign Arbitral...