This blog is adapted from a keynote speech delivered at the Helsinki International Arbitration Day on 28 May 2015.

Most arbitration lawyers and users of arbitration services would probably agree that efficiency is a key issue in international arbitration, or that efficiency of arbitration is a key issue. However, these two ways of stating the issue tend to beg it: key to efficiency lies, precisely, in the distinction between efficiency in arbitration and efficiency of arbitration – in this hardly noticeable distinction between “in” and “of.” Efficiency in arbitration is about ensuring efficiency after the dispute has arisen. Efficiency of arbitration is about ensuring efficiency before the dispute has arisen.

When we look at efficiency in arbitration, we look at arbitration from the inside, as arbitration counsel and arbitrators. When we look at efficiency of arbitration, we look at arbitration from the outside, from the user’s perspective. The key is to combine the two perspectives.

As Yves Derains recently noted, in international arbitration due process always trumps efficiency. This is not to suggest that efficiency is not important in international arbitration; it certainly is, but it is always secondary to due process. Once an arbitration proceeding is underway, the task of the arbitrator is to ensure compliance with due process, and if there is a conflict between due process and efficiency, due process will always prevail.

Most arbitration lawyers would probably agree with this, as do in fact most arbitration laws – failure to respect due process is a basis for setting aside an arbitral award in most jurisdictions, but failure to respect cost efficiency is usually not.

Efficiency of arbitration is another matter. How do we ensure or at least enhance efficiency of arbitration before any dispute has arisen?

The answer lies not so much in the proper drafting of arbitration rules, although this could also be helpful. The answer lies in the proper drafting of the arbitration agreement. But before looking at this particular issue, let us reflect on what is meant by efficiency.

The type of efficiency that is relevant in international arbitration can be measured in two different ways – in terms of time, or in terms of money. We are all familiar with using time as a measure of efficiency, and we often say that an arbitration proceeding has been efficient if it lasts, say, a year rather than five years.

But a shorter arbitration is not necessarily a more efficient arbitration. You measure efficiency – in economics, engineering and other disciplines – by comparing two different things; efficiency is a ratio between the two, measured by the same unit. If you measure efficiency of arbitration in terms of time, one way to do this is by comparing the likely duration of an arbitration proceeding with of the duration of the commercial transaction out of which the dispute arises. Measured in these terms, an arbitration proceeding is the more efficient the greater the difference between the duration of the arbitration and the duration of the transaction. If it takes a year to arbitrate a dispute arising out of a long term – say, a ten-year – transaction, the ratio would be 1 to 10, or in other words, the transaction cost would be 10% in terms of time, so the arbitration option would be reasonably efficient.

However, if it were the other way round, efficiency would be negative; we could not say that a ten-year arbitration would be a time-efficient way of settling a dispute arising out of a one-year transaction.

This, incidentally, is one of the reasons why it usually does not make sense to arbitrate consumer or other one-off transactions – they are usually very short-term, so arbitrating them is not efficient in terms of time.

However, arbitration may be an efficient method even for short-term transactions if efficiency is not measured in terms of time but in terms of money. If you buy a lottery ticket, and win a 100 million euros, and the lottery company refuses to pay you, it would be efficient to try to arbitrate the dispute in terms of money, even if it would not be so in terms of time.

But how does one measure efficiency of arbitration in terms of money?

Again, you compare. One way to do this is by comparing the likely cost of arbitration proceedings with the value of the transaction from which the dispute arises. If the value of the transaction is say 100 million euros, in principle it would be cost efficient to agree to an arbitration clause since the cost of arbitration for any one party rarely exceeds a few million.

Conceptualizing the cost of arbitration as a transaction cost is of course a common way of measuring efficiency, and this is how users of arbitration tend to approach the issue: they measure the efficiency of arbitration by looking at whether it would be cost-efficient, in terms of money, to arbitrate disputes under the transaction they are about to enter into. The problem the users are facing when trying to achieve greater efficiency, and it is a very important problem, is that the value of the transaction is not always the same, and indeed is usually not the same, as the value of the dispute.

What is in dispute is not usually the value of the entire transaction – the entire 100 million in the example just mentioned – but something less. This something less is of course the amount claimed. This is how the value of the dispute is in practice quantified; you look at the amount claimed and not the value of the entire transaction.

Therefore, in order to determine whether arbitration of a particular dispute is a cost-efficient option, in terms of money, one must compare the cost of arbitration against the value of the dispute, and not the value of the whole transaction. When one makes this analysis, it may turn out that a standard arbitration proceeding is not a cost-efficient option because even if the dispute may arise from a 100 million transaction, the value of the dispute may be much less, say 3 million, and if the cost of arbitration is also 3 million, arbitration would not be a particularly efficient way of settling the dispute; if you win, you get the 3 million, but you would spend the same amount to get that 3 million. Most of us would agree that this is not particularly efficient; in fact, it would be a zero-sum game.

Is there any way to deal with this dilemma?

We can draft smarter arbitration agreements – agreements that take into account the uncertainty about the value of the dispute – the amount claimed. This can be done by developing structured arbitration agreements, that is, arbitration agreements that tailor the arbitration process to match the value of the dispute.

Standard arbitration clauses are the great missed opportunity to enhance the cost-efficiency of international arbitration. We should replace standard arbitration clauses – one-size-fits-all clauses, or clauses that fit all sizes of disputes – with real arbitration agreements that are tailored to the size of the claim and therefore provide a better fit.

For instance, if the value of the dispute is, say, 10 million euros or more, the parties may agree to have a full-fledged arbitration, with all the bells and whistles of an international arbitration proceeding – three arbitrators, two rounds of pleadings, document production, an oral hearing, post-hearing submissions. But if the value of the dispute is substantially less than 10 million, say 1 – 5 million, the parties may wish to consider whether to have a sole arbitrator, or agree in advance to have only one round of written pleadings, and no document production for instance. And if the value is even less, say below 1 million, the parties may wish to further simplify the process – for instance by arbitrating the case on the basis of documents only, without any hearing.

It is important to agree on the structure of the arbitration in advance because of the problem of moral hazard – a problem well known in economics, and in any attempt to increase the efficiency of regulation. When the parties negotiate a commercial transaction, and the related arbitration agreement, they don’t know yet which of them will be the claimant and which one will be the respondent – they will only know this once the dispute has arisen. Both parties are therefore required to think objectively, in their shared, common interest. They can think in terms of what is objectively economically rational and cost-efficient – like tailoring the arbitration process to the value of the dispute. But once the dispute has arisen, moral hazard always tends to raise its ugly head – it tends to turn into a legal lizard; a beast that risks poisoning the entire business relationship.

After the dispute has arisen, the parties are no longer in a symmetric situation in terms of information – one of them knows that it is the claimant and has something to gain, and the other knows that it is the respondent and has only something to lose – it only has the downside. Each party has its own interest to promote and protect, and it is unlikely that they will be able to agree, at that late stage, on such fundamental issues of principle such as a re-structuring of the standard arbitration process to reflect the value of the dispute.

The respondent in particular has no interest or incentive to shortcut the process – this would only tend to lessen the due process requirements that it needs to be able to effectively defend itself against the claims.

This is not a moral judgment; it is simply a description of the economic rationality that prevails in the asymmetric situation created by the dispute. Once the dispute has arisen, it is always already too late to regulate – the Pandora’s box, the arbitration case, is already open.

The way to achieve greater efficiency in international arbitration is much closer cooperation between corporate counsel and arbitration counsel when drafting arbitration agreements. We cannot consider ourselves as two different species of lawyers that only come together when it is already too late, after the dispute has arisen.

In conclusion, I would therefore like to make a concrete proposal. I would like to call for the creation of a key club of corporate counsel and arbitration lawyers, a gathering of lawyers that takes on the task of re-drafting of standard arbitration clauses – their conversion into proper arbitration agreements, structured products that take into account the very different types of disputes, in terms of both time and money, that may arise in the context of international commercial transactions. Such a gathering need not take place in any physical sense; a virtual meeting of minds would suffice.


To make sure you do not miss out on regular updates from the Kluwer Arbitration Blog, please subscribe here.



  1. Veijo, is there a reason to attack the problem in the arbitration agreement, a text that is almost always the result of negotiation among non-specialists in arbitration, instead of in the arbitration rules, which are always the product of experts?

    1. Michael, the parties are always much closer to the transaction than those drafting the rules and are therefore in a better position to tailor to process (whcih is not to say the arbitration rules cannot be improved), I guess one of the main points here was that arbitration specialists should be brought in when drafting arbitration agreements. You could then try to agree on a real arbitration agreement, instead of merely incorporating a standard clause.

      1. Veijo, while it may come as a surprise to many, there are occasional days when we in-house lawyers do not sit around eating bon bons…

        As I have argued previously here on Kluwer, however, the term “drafting” inaccurately describes how many dispute clauses come into existence.

        Rather than being artfully crafted, they are often a product of negotiation in which a party is unlikely to get 100% of what it wanted, especially when they are trying to expend negotiation leverage to obtain better terms on things like price and delivery.

        1. Michael — I certainly did not mean to suggest anything else. The idea rather was that arbitration clauses could (and in my modest view should) be negotiated not only in terms of the applicable law, rules and seat, but also in terms of the structure of the arbitration proceedings, based on the parties’ own assessment of the transaction costs they are prepared to bear, in view of the value ranges of potential disputes. The parties — with the support of their corporate and arbitration counsel of course — are certainly in a better position than the drafters of arbitration rules to assess the relationship between the two. This does not imply that either party will be in a position to impose its views on the other, but if the negotiations on the structure of possible arbitration proceedings are conducted at the time when the transaction itself is also being negotiated, the chances of reaching an economically rational (ie cost-efficient) agreement are much higher than after the dispute has arisen, when the moral hazard has already materialized and the parties are no longer in a symmetric situation.

  2. Veijo, thanks for an insightful post. What about the
    efficiency at the post-arbitration stage? Would you consider a possibility of enhancing structured arbitration clauses by
    provisions determining the deserved standard of review a viable option for the future?

    1. A very good question, Agnieszka. The parties could of course agree to waive at least some of the grounds for resisting enforcement under the New York Convention, but it is unclear how efficient this would be, and in any event this may not be advisable as almost all of the grounds relate to events taking place during the arbitration proceedings (and constitute elementary protections you may not wish to waive even if possible). But one might be able to consider ways of facilitating the enforceability of the award in circumstances where the award is not challenged within the statutory time limit.

Leave a Reply

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