The title of this President’s Message borrows liberally from a rather silly 1970s comedy, described in its Wikipedia entry as “a British farce written by Alistair Foot and Anthony Marriott, which premiered in London's West End on 3 June 1971. It was unanimously panned by critics, but played to full houses until 1987 at three different theatres (the Strand, the Garrick and the Duchess), totalling 6,761 performances”. The success was such that it was even made into a film starring Ronnie Corbett.
The reader may be puzzled by this introduction and query whether a comedy of dubious taste could have any relevance for arbitration. The answer is that it has none. What is of interest is the contrast between, on the one hand, the “unanimous” panning to which the play was treated by critics and, on the other, the immense enthusiasm and popularity it enjoyed among the greater public.
This brings us to the actual topic of this President’s Message. Our topic is another example of “expert wisdom” being drowned out by “popular perception”: the tenacious myth according to which arbitrators tend to “split the baby”. This subject has been referred to en passant in several recent President’s Messages and the time has come to address it squarely.
There exists a wealth of publications reporting on the commonly-held perception that arbitrators “split the baby” or even asserting that “baby-splitting” is a reality in arbitration, even in international arbitration. What is particularly troubling is that the belief appears to be held most strongly among corporate counsel; perhaps predictably, litigation lawyers come a close second, with judges taking third place. All of them opine that the tendency of arbitrators to render compromise decisions instead of “hard decisions” actually discourages resort to arbitration. This belief is very widely-held, despite not only the lack of any evidence that would comfort it, but even in the face of abundant evidence establishing the exact opposite.
For example, a report by the RAND Institute for Civil Justice on “Business-to-Business Arbitration in the United States – Perceptions of Corporate Counsel” published in 2011 shows that no less than 71% of survey respondents either agreed (43%) or strongly agreed (28%) that “arbitrators are less likely than a judge or jury to decide strongly in favor of one side or the other”.
This RAND report related exclusively to domestic arbitration in the United States; international arbitration was specifically excluded. Some might argue that this renders it of little or no relevance in relation to international arbitration. However, that argument would be valid only if the perceptions of corporate counsel or others distinguished between domestic and international arbitration; there is no evidence that any such mental distinction is made.
Confronting myth and facts: no “baby-splitting” on the merits
As will generally be the case of “myths” – otherwise we would not be calling them “myths” – the belief that arbitrators “split the baby” does not withstand a reality check.
For example the RAND report cited above included the following information, which shows an “apparent disconnect” between perception and reality:
“The prevalence of this perception prompted the AAA to analyze 111 of its awards in 2009. It found that
– 7 percent of decisions awarded approximately half (41 to
60 percent) of what was claimed
– 41 percent awarded more than 80 percent of the amount claimed
– 9 percent denied claims completely.”
These figures hardly support the notion that compromise rulings are the norm: in fact, with only 7% of decisions awarding between 41% and 60% of the amount claimed, they make absolute nonsense of the “baby-splitting” myth. At the time they were nothing new and have since been confirmed many times over. A recent study by Mr Carter Greenbaum (to which the author has referred in previous President’s Messages) should definitively “put the baby to rest”, even if it also concerns mainly (but not exclusively) domestic commercial arbitration in the United States.
Mr Greenbaum analyses about 400 cases administered by the Judicial Arbitration and Mediation Services (JAMS) group between August 2008 and August 2012. He compares the “claim amount” and the “award amount”. He defines the “claim amount” as the total of compensatory damages claimed, excluding punitive damages, attorneys’ fees, costs, interest as well as emotional damages or other forms of speculative damages. The “award amount” is the monetary relief actually awarded, exclusive of costs, punitive damages, interest, etc. Mr Greenbaum defines “claim splitting” rather broadly: for the purpose of his analysis “claim splitting” occurs where arbitrators award either less than 80% or more than 20% of the claim amount. 
One finding practically summarises it all: only 14.4% of all awards (including “slip and fall” tort cases) qualify as “claim splitting” as defined by Mr Greenbaum. When one excludes the “slip and fall” tort cases, the percentage of “claim splitting” (as defined above) drops from 14.4% to only 12.5%, leaving a whopping 87.5% of cases in which the amount awarded was either 80% or more, or less than 20%, of the claim amount.
That said, studies that take into account only the fate of monetary claims can give an incomplete picture. One must take things a step further, for other considerations can, rightly or wrongly, give an impression of “baby splitting”.
Consider these two examples, based on actual cases.
Example 1: A contractor terminates a FIDIC contract for construction on the basis of the employer’s failure to provide evidence of its financial arrangement (Sub-Clauses 2.4 and 16.2(a) of the FIDIC Conditions of Contract for Construction; the “FIDIC Conditions of Contract”). The employer disputes the validity of the termination and in turn terminates the contract for abandonment of the works (Sub-Clause 15.2(b) of the FIDIC Conditions of Contract). The contractor brings arbitration and seeks a declaration that it validly terminated the contract, plus payment of works carried out, variations denied by the Engineer and other claims under Sub-Clause 16.4 of the FIDIC Conditions of Contract. The total sums claimed by the contractor are an eight-figure amount in Euros. The employer counterclaims in the arbitration for a declaration that the contract was validly terminated by it, plus payment of an eight-digit Euro amount for repair costs of defective works and delay penalties. The employer also reserves a very significant claim against the contractor for the additional costs that it will incur because it will need to have another contractor complete the works (Sub-Clause 15.4(c) of the FIDIC Conditions of Contract); this potential claim alone is estimated to be a multiple of the aggregated claims and counterclaims in the arbitration.
The arbitral tribunal finds in favour of the contractor on termination and grants the contractor about 75% of its “claim amount” (as defined by Mr Greenbaum). However, the arbitral tribunal finds in favour of the employer on the repair costs and delay penalties and also awards the employer about 75% of its “counterclaim amount”. The arbitral tribunal then offsets the amounts awarded to each Party, leaving a balance of about EUR 400,000 in favour of the employer.
Taking a purely monetary approach, this award would qualify as “claim splitting” as defined above. However, is this correct?
The central issue in this dispute was the validity of the termination. By finding that the contractor had validly terminated the contract, the arbitral tribunal shielded the contractor from the employer’s much higher potential claim for the costs of completing the work. It was also crucial for the contractor to avoid an award declaring the contract validly terminated by the employer because this circumstance would have had to be disclosed when bidding in future tenders. In other words, for the contractor, the outcome of the arbitration was an unqualified success, and well worth the EUR 400,000 that it had to pay to the employer.
Example 2: A general contractor terminates a subcontract on the basis of the subcontractor’s failure to commence and progress the works. The subcontractor disputes the termination and seeks damages in an amount of about USD 10 million. The general contractor counterclaims, seeking a declaration that it validly terminated the subcontract, as well as a payment of about USD 17 million, which includes a head of counterclaim for USD 16 million that is contingent upon a claim brought by the main contractor against the owner in separate proceedings.
The arbitral tribunal finds in favour of the general contractor on the termination and, for this reason, rejects the entirety of the subcontractor’s claims. With respect to the general contractor’s counterclaims, the arbitral tribunal dismisses without prejudice the USD 16 million contingent head of counterclaim, finding it to be premature. On the remainder of the counterclaims, the arbitral tribunal grants only about USD 100,000.
If one takes into account only the fate of the monetary claims, this award was not an instance of “claim splitting” (as defined by Mr Greenbaum), since the arbitral tribunal awarded 0% of the claim and only about 0.59% of the counterclaim. It also ruled in the main contractor’s favour on the termination.
However, for the main contractor, the finding that the subcontract was validly terminated did not have any value beyond nipping in the bud all of the subcontractor’s claims. Apart from that, the net effect of the award was almost a status quo, since both parties’ claims and counterclaims were dismissed either entirely or in their near-entirety. So, were the arbitrators “splitting the baby”? The answer (a piece of insider information) is no. The outcome was, very simply, the result of the Tribunal’s finding on termination, its finding on the premature nature of the bulk of the counterclaim and the weighing of the evidence for the rest of the counterclaim. But one may question whether this matters: is it possible that the parties perceived the arbitral tribunal to be engaging in “claim splitting”? The likely answer is yes, and if so, that is all that matters, since “perception is everything”.
To sum up at this stage, although empirical studies all conclude that “splitting the baby” is the exception, not the rule, this may not prevent a perception of Solomonic justice in individual cases. Hence, perhaps, the extraordinary tenacity of the arbitrators-split-the-baby myth.
What about “procedural baby-splitting”?
Another form of “baby-splitting” could be called “procedural baby-splitting”.
This is a more insidious creature. Here, arbitrators do not engage in “claim splitting”. What they do, or are perceived to be doing, is to issue procedural rulings that systematically appear to take the middle road.
Examples are legion and some usual suspects immediately come to mind:
- A party requests a three-week extension of time; the opposite party objects and says that one week is enough; the Tribunal gives two weeks;
- A party makes sixty requests for document production; the requested party states that it will comply voluntarily with twenty requests but objects to the remaining forty; the arbitral tribunal grants twenty out of the disputed requests;
- A party demands two days to cross-examine another party’s expert; the other party objects and submits that two hours are enough; the arbitral tribunal grants one day, with one hour “reserve time” the following day.
One particular twist here is that it is not only (or even mainly) people that are unfamiliar with arbitration who are complaining. On the contrary, “procedural baby-splitting” is a pet topic of many a war story exchanged among seasoned arbitration practitioners.
That said, is it really a problem? And if so, when and why?
At the start of the proceedings, in particular when the parties and the arbitrators are sitting down at their first case management conference to agree on the basic ground rules to govern the proceedings, at least some kind of give-and-take (also called sensible compromise) is not only unavoidable: it is indispensable. This goes to the very heart of arbitration. Anyone who thinks of this as “procedural baby-splitting” should think again.
Likewise, over the course of the proceedings, the arbitral tribunal will often need to strike a balance between competing interests, all of which may be perfectly legitimate. The arbitral tribunal must also keep the process on track and the parties on board. Taking things a step further, “procedural baby-splitting” can even help to protect an award if the arbitral tribunal has an idea where the cases is headed on the merits and does not want to give the likely losing party real or imagined grounds for procedural complaints. If some measure of “procedural baby-splitting” (by taking the middle road or by fudging some decisions) is the price to be paid for achieving these goals, this is still a good bargain.
Matters become more complicated where the arbitral tribunal is faced with a party that behaves like a bully or, more generally, that acts in disregard of the rules of the game and thereby threatens to derail the process. Arbitrators may be tempted to keep this type of party “on board” by accommodating some of its requests, even requests that the arbitral tribunal might not have otherwise considered had they come from a less obnoxious litigant. The problem with bullies is that accommodation tends quickly to turn into appeasement, with predictably dismal results. Firmness must be shown, and the sooner the better. This may sound obvious, and in theory it is. In practice, it is not: fortunate is the arbitrator who can tell an assertive party from a bully at first sight.
Even where one party is not bullying the arbitral tribunal and/or the other parties, another risk of “procedural baby-splitting” is that it can ultimately undermine the Tribunal’s authority. There are few things that are more disconcerting to users (in-house counsel and/or the business management) than an arbitral tribunal that does not uphold its own rules. For the author, this is a deadly sin that an arbitrator must avoid at all costs. When forging the rules of the game, compromise and understanding for the background and culture of the parties and their counsel is indispensable. When applying the rules, some flexibility can and should be shown where one party’s failure to comply with the rules does not result in any irremediable prejudice to another and where procedural fairness and balance require it. But when a party is allowed to get away with too many short-cuts, the integrity of the entire process is in danger.
Thus, whilst there are many studies showing that “claim-splitting” is rare when it comes to the merits, perhaps the time has come for in-depth empirical examination whether “procedural baby-splitting” is figment of the imagination or a real issue.
Let us close with two observations.
First, it is ironic that the terms “splitting the baby” should be associated with King Solomon’s famous judgment. As we all know (or should know), he did not split the baby but awarded 100% of the object in dispute to its true mother (admittedly, his methods of taking and weighing the evidence were somewhat novel). His judgment would not qualify as “claim-splitting” as defined above. Thus, even the terms “baby-splitting” are misguided. Had someone pointed this out earlier, arbitrators might have actually welcomed a comparison with such illustrious precedent.
Second, turning back to the 2011 RAND report, one remark leaps from the page:
“We found that respondents who used arbitration clauses most frequently tended to disagree with the view that arbitrators split decisions (…).”
These two observations bring us to an obvious and upbeat conclusion: whilst perception may be everything, even at the expense of fact, there is hope that experience and sharing of hard evidence will triumph over even deeply ingrained myth. It falls upon us to pursue this noble aim.
 Even if innocuous by today’s standards, the title of this work of art could cause embarrassment, which explains the rather cryptic reference. Readers can find useful information by googling the search terms “no please we’re British”.
 A cursory internet search is enough to establish how widespread this perception remains.
 Douglas Shontz, Fred Kipperman, Vanessa Soma, Business-to-Business Arbitration in the United States – Perceptions of Corporate Counsel, RAND Corporation, 2011, p. 12.
 Shontz, Kipperman, Soma, op. cit., p. 3, footnote 13.
 Shontz, Kipperman, Soma, op. cit., p. 11.
 Carter Greenbaum, Putting the Baby to Rest; Dispelling a Common Arbitration Myth, 26 American Review of International Arbitration  101-130.
 Greenbaum, op. cit., pp. 115-121.
 Greenbaum, op. cit., p. 121. The analysis drills down deeper and reviews mean amounts; the conclusions are the same in substance.
 Another deserving topic would be an examination of “claim-splitting” when it comes to costs. Given the broad discretion that arbitrators enjoy in this context, the potential for real or perceived “baby-splitting” is particularly high. The subject is too vast to be addressed in this President’s Message.
 See Kings 1, chapter 3, verse 28; few arbitrators would object to receiving praise of this sort.
 Shontz, Kipperman, Soma, op. cit., p. 11. The language that has been shamefully relegated into the ellipse reads, “although our sample size was not large enough to fully characterize this distinction”. The author makes this disclosure out of intellectual honesty.