Thomas Schultz is a Reader in Commercial Law in the Dickson Poon School of Law at King’s College London and a Swiss National Science Foundation Research Professor in the International Law Department at the Graduate Institute of International and Development Studies, Geneva. He is also the Editor-in-Chief of the Journal of International Dispute Settlement (Oxford University Press).
Professor Puig’s article ‘Social Capital in the Arbitration Market’ is a quite wonderful contribution to a number of things: our understanding of the dynamics of investment arbitration, the literature on arbitrator appointments, the methodological diversification of studies in international law, and certainly a few more. And it brings us rather convincing evidence, in a field where claims and representations (not to speak of copious discussions of what other people happen to have said) are more readily found than data and studies to substantiate claims. It is, in other words, intellectually edifying. The experimental design is well done, the plan well executed and the findings credible. In this, it is intellectually responsive to developments in the social sciences and the humanities. We don’t even need a mood-elevating metaphor to set great store by this sort of works, and this work in particular. (Incidentally, the study is also a formidable ‘who’s who in investment arbitration’, which undoubtedly will make for welcome entertainment in certain circles.)
A few small methodological points would probably deserve more discussion. (I said ‘would deserve’, not ‘would have deserved’: the article is long enough as it is and this is a law journal after all.) For instance, the author says that ‘The network analysis advanced in this article relies upon displayed preferences by the appointing entity (litigation parties, arbitrators, and the institution) to provide a larger picture of the network’s aggregate topology.’ But how do we know the preferences of the appointing entity? Right, by looking at appointments. But do effective appointments really tell us what the preferences are? What if individuals, who are the preferred choices of the appointing entity, refuse an appointment, and the appointing entity has to turn to their second or third choice? Never happens. Well… Actually, could such situations be statistically relevant?
Another methodological point: Figure 8 is puzzling. Not puzzling as in ‘probably wrong’. Puzzling as in ‘how come’? Here’s the author’s accompanying notes: “Figure 8 shows how, despite the fact that most ICSID cases were registered in the last 10 years, most ‘power-brokers’ or those arbitrators at the top of the profession entered the network in or prior to 2004.” In other words, the mid-2000s is the moment when you see the network effects. Why? Why did the network stabilise at that point in time? The network seems to have acquired self-organisational elements at that point in time, but, again, why then? Any hypothesis? Just happenstance? Just puzzling.
Beyond methodological considerations, we may also wonder–and perhaps the author wants to elaborate on this–why, in fact, it is a bad thing that a small number of arbitrators decide a great number of cases. Here is the situation, in the author’s words: ‘10 per cent of the total pool accounts for half of the appointments’. Now, why is an oligopolistic situation undesirable in this situation? Granted, ‘the richer get richer’, as the author puts it. But if we can resist our mimetic desires for a moment, isn’t the world large enough for all of us who are interested in investment arbitration to do our own things? Granted number two, the cause of feminism is poorly served by investment arbitration. This is an important cause, which should be fought here as it it should be fought everywhere. But would it change investment arbitration?
Would arbitrators work better, or more efficiently, if there weren’t an oligopoly? Are we troubled by the fact that a small number of judges decide all cases at the highest level in a country? Is diversification really the problem? Or is the problem rather that we don’t really like what they have done–whoever they are? That we find the functions that investment arbitration actually fulfils to be disappointing? If this is so, whose fault is it? Where did the system go wrong? If we changed the people who make the decisions, would the decisions really change all that much? Or are the salient incentives and constraints that the current socio-legal system places on investment arbitrators the main determinants of their decisions (along with the law, yes, I’m not completely iconoclastic)? Conversely, if we like the output of the system of investment arbitration, why change its machinery, why change who sits behind the wheel?
To more prospective musings: what do Sergio’s points suggest about the future of arbitrator appointments in investment arbitration? Let us take the idea that one of the key determinants of arbitrator appointment is predictability of arbitrator behaviour. The author explains this as ‘the evil that we know is best’ and the importance, in appointments, of ‘risk aversion, and a desire for predictable outcomes on the part of those making the appointments’.
The argument, if we push it to an extreme to bring out its edge, means that what is important when choosing an arbitrator is not whether they are good or bad lawyers, whether they are efficient organisers or not, whether they are from a commercial law background or an international law background, whether they are disciplined and rigorous or not, how well they know investment law, how much experience they have in arbitration. What is important, here, is how much of a predictable machine they are: the ideal arbitrator would be a public and understandable algorithm that produces outcomes based on the rules of mathematics and logics. For a party, if the case fits the algorithm, the choice of arbitrator is easy. The closer an individual comes to this algorithm, the more valuable he becomes as an arbitrator candidate. Now, if this is true, what determines the proximity of an individual to this ideal? What makes an arbitrator’s behaviour predictable?
An individual who has decided a hundred cases offers more opportunities to ‘read’ his decision-making behaviour than an individual who has decided only one, or even no case. As the author says, ‘arbitrators who have been appointed more frequently are more likely to attract further appointments’. But more opportunities to read a decision-maker do not necessarily translate into more predictability. (Notice in passing how, arguably, some of the ‘best’ judges of many supreme courts are those whose decisions are the most unpredictable.) A judge, or an arbitrator, who has decided only a small number of cases but who systematically, invariably decides on the basis of a rather unsophisticated reading of the law and a strong identifiable inclination towards X or Y may be far closer to the ideal of the perspicuous algorithm. Does this mean that there is a market for hard-liners? That progressively the pool of arbitrators will expand and that the strongest new entrants will be those who take a hardline view on X or Y? If your case essentially revolves around X, you would look for a hardliner on X; if the case is about Y, you will want a hardliner on Y; and so on. The trick for new entrants, then, would be to make good guesses about how many occurrences of X and Y there will be in the coming years. And to communicate massively about their inclinations towards X or Y.
Finally, a particularly cynical reading of this article might be this: a small number of individuals are particularly influential in investment arbitration; they particularly strongly orient the development of the law produced by investment arbitration. These are really powerful individuals.
As Spiderman would say, with great power comes great responsibility. This is their gift. This is their curse. Are these the people responsible for the great backlash against investment arbitration? Ouch, I wouldn’t want to be one of them…