I was back in Kuala Lum­pur for Asia ADR Week 2024 and an Oxford-style debate on the top­ic of ‘The Extent and Lim­it­a­tions of Arbit­rat­ors’ Dis­cre­tion­ary Powers’.

I love con­fer­ences in Kuala Lum­pur and am happy to speak at them, but what is it with me and debates? Any­way, this one was mod­er­ated by Adri­an See. My oppon­ent, Teh Wai Fung, argued that ‘arbit­rat­ors should take an act­ive role in pro­ced­ur­al mat­ters and devel­op mech­an­isms to safe­guard the arbit­ral pro­cess’. I argued that ‘arbit­ra­tion is mainly a party-driv­en task and that arbit­rat­ors may rely on the cooper­a­tion of the parties to improve the over­all effect­ive­ness of arbit­ra­tion’. Here’s my state­ment and my rebuttal.

The Statement

I will argue why I believe that arbit­ra­tion is mainly a party-driv­en task and that arbit­rat­ors may rely on the cooper­a­tion of the parties to improve the over­all effect­ive­ness of arbit­ra­tion. I can­not do so in fun­da­ment­al dis­agree­ment with my learned friend, because I think he is very much right in his view. He’s not largely wrong, only partly. So my argu­ment is a simple yes, but.

Yes, …

Yes, arbit­rat­ors should take an act­ive role in pro­ced­ur­al mat­ters and devel­op mech­an­isms to safe­guard the arbit­ral pro­cess. Simply because those involved in arbit­ra­tion have allowed it to degen­er­ate in recent years and dec­ades. Parties have been com­plain­ing about this for as long as law­yers and arbit­rat­ors have been prom­ising improve­ments. But not much has changed. We often talk about inter­na­tion­al best prac­tices. This is not one of them. It is an inter­na­tion­al bad practice.

What am I talk­ing about?

Many inter­na­tion­al com­mer­cial arbit­ra­tions are too form­al and com­plic­ated, mak­ing a flex­ible and stream­lined pro­cess lengthy and expens­ive. In short, arbit­ra­tion is often inefficient.

This prob­lem is less pro­nounced in spe­cial­ised areas like mari­time arbit­ra­tion, where people bet­ter under­stand and util­ise pro­ced­ur­al flex­ib­il­ity. But in gen­er­al com­mer­cial arbit­ra­tion, flex­ib­il­ity has become a sig­ni­fic­ant dis­ad­vant­age. Unfor­tu­nately, this prob­lem often stems from how the parties, their law­yers and even the arbit­rat­ors handle the process.

The Theory

In arbit­ra­tion, the parties have almost com­plete con­trol over the pro­cess, thanks to the prin­ciple of private autonomy. This means they can tail­or the pro­cess to their spe­cif­ic needs, as long as the arbit­rat­ors fol­low due pro­cess. If the arbit­rat­ors do not fol­low due pro­cess, a court can later set aside their award. But ideally, this almost total con­trol of the pro­cess by the parties should mean a flex­ible and tailored pro­cess. Not too much, not too little. Just the right amount.

The Practice

In prac­tice, arbit­ra­tion often gets bogged down in excess­ive form­al­it­ies and motions, not for sense, but to com­plic­ate mat­ters for the oth­er side. These are ‘guer­rilla tac­tics’. In such a situ­ation, if an arbit­rat­or fears a court will set aside his award for not fol­low­ing due pro­cess, he often reacts with excess­ive cau­tion, grant­ing exten­sions of time and allow­ing fur­ther writ­ten sub­mis­sions when he shouldn’t. This is ‘due pro­cess para­noia’, and when guer­rilla tac­tics and due pro­cess para­noia meet, your inef­fi­cient arbit­ra­tion is perfect.

Against this back­ground, stake­hold­ers are rightly call­ing for arbit­rat­ors to engage in pro­ced­ur­al mat­ters and devel­op safe­guards for the arbit­ral pro­cess. After all, if arbit­ra­tion fails to be a good altern­at­ive dis­pute res­ol­u­tion, why should parties choose it over oth­er methods?

So yes, arbit­rat­ors should do that.

… But

For all our love of effi­ciency, our love of private autonomy should be great­er. An arbit­rat­or is not a judge! In addi­tion to decid­ing dis­putes, a judge main­tains and devel­ops the law – espe­cially in pre­ced­ent-based jur­is­dic­tions – and pre­serves the unity of the law and pub­lic order. A judge must there­fore pro­tect not only the interests of the parties, but also those of the gen­er­al public.

The Core Concept of Arbitration

An arbit­rat­or need not do all this; indeed, an arbit­rat­or must not do all this. Instead, an arbit­rat­or must decide the dis­pute before him – and only that dis­pute – and why? Because the parties have brought it before him in their private capa­city. A court exists because the con­sti­tu­tion says so, but an arbit­ral tribunal exists because of an arbit­ra­tion agreement.

We there­fore accept as a core concept of arbit­ra­tion that an arbit­rat­or decides the dis­pute com­pletely detached from gen­er­al con­sid­er­a­tions, because his exist­ence as an arbit­rat­or is based on the will of private parties.

We must accept this concept in its entirety. Not only can the parties privately and autonom­ously agree to arbit­rate, they can also agree to arbit­rate inef­fi­ciently, and if they do, the arbit­rat­or must accept this (although per­haps he should ensure that he is paid prop­erly and on time). The arbit­rat­or must not put his own view of effi­ciency above that of the parties. If the parties agree that there is no need to safe­guard the arbit­ral pro­cess, the arbit­rat­or must not safe­guard it.

Private Autonomy and Inefficient Arbitration

Cer­tainly, the parties will not nor­mally agree to an inef­fi­cient pro­ced­ure (whatever that may be). If one party sab­ot­ages the agreed pro­ced­ure, it will be in breach of the arbit­ra­tion agree­ment. So how do you avoid this or any oth­er unwanted inef­fi­ciency? Well, first and fore­most, you don’t do it by throw­ing some­thing at the parties that doesn’t come from them. Arbit­ra­tion, as a private pro­cess, requires that you do it by involving the parties. In short: safe­guard­ing the pro­cess is what the parties say it is.

This means that if an arbit­ra­tion is inef­fi­cient accord­ing to the arbit­ra­tion agree­ment, arbit­rat­ors should primar­ily work with the parties to make improve­ments. Only under this premise and with­in this frame­work should arbit­rat­ors be pro­ced­ur­ally act­ive and devel­op safe­guards for the arbit­ral pro­cess. It seems to me that this is not always respected.

Is it bad for the parties to get involved before the arbit­rat­or act­ively safe­guards the pro­cess? No. Prop­erly applied, this approach allows soph­ist­ic­ated parties to use arbit­ra­tion without facing a nanny mech­an­ism that pro­tects interests determ­ined by the arbit­rat­or against their will. At the same time, it pro­tects inex­per­i­enced parties if their arbit­ra­tion agree­ment requires such protection.

This is an approach that respects arbitration’s basic ten­et: private autonomy. It can cre­ate new inter­na­tion­al best prac­tice where (and only where) needed, without risk­ing the intro­duc­tion of anoth­er bad practice.

The Rebuttal

My learned friend said that by agree­ing to arbit­ra­tion, the parties have del­eg­ated the power to man­age the pro­cess to the arbit­rat­or. He lit­er­ally said, ‘Parties may not know what’s good for them.’

Really? Who are we to decide what’s good for the parties?

Don’t get me wrong, I’m all for arbit­rat­ors tak­ing a very act­ive role in pro­ced­ur­al mat­ters and safe­guard­ing the pro­cess – in fact, as an arbit­rat­or myself, I live by that prin­ciple. But only when the parties them­selves can­not agree on the pro­ced­ure (which, as we know, is often the case). And nev­er, ever when the parties do agree on the pro­ced­ure. Party autonomy doesn’t end with the con­sti­tu­tion of the arbit­ral tribunal. This cor­rect pos­i­tion is sup­por­ted by the UNCITRAL Mod­el Law and the New York Con­ven­tion, which provide that an arbit­ral award may be set aside or its recog­ni­tion refused if it is proved that the arbit­ral pro­ced­ure was not in accord­ance with the agree­ment of the parties (Art­icles 34(2)(a)(iv), 36(1)(a)(iv) of the UNCITRAL Mod­el Law; Art­icle V(1)(d) of the New York Convention).

See also my Linked­In post and those of Asia Inter­na­tion­al Arbit­ra­tion Centre and Wai Fung on this event.