There are certain dry questions which, to judge from conference speeches and articles, lawyers seem to spend an inordinate amount of time pondering. One such question is whether commercial arbitration awards should be routinely published and what their precedent value should be. One source (considered further below) apparently suggests that 64% of “stakeholders” in arbitration favour publishing redacted awards or summaries, and that a substantial minority (19%) favour publishing complete awards.

What metric should we apply when seeking to decide whether a given reform is desirable? I would suggest that a reform is only desirable if it is very likely to prevent or narrow future disputes, or make arbitrations quicker, cheaper or more likely to arrive at a just result. It is hard to see how the systematic publication of commercial awards could do any of those things.


International commercial arbitration is arbitration of the kind routinely conducted under the ICC Rules, LCIA Rules and similar. Parties (usually businesses) from different countries argue about their private law rights and obligations, arising from the contracts which they have entered into with one another. The LCIA does not publish awards. The ICC publishes anonymised extracts from a small minority of awards, though it is down to editorial discretion what is published and what is not.

Historically English law allowed appeals to be made to the English courts on the merits of arbitration awards. In the Arbitration Act 1979 parliament gave parties the right to contract out of appeals on the merits save in certain limited categories of case. The Arbitration Act 1996 (section 69) abolished the special categories where contracting out was not permitted. This right to contract out of appeals on the merits also appears in the UNCITRAL Model Law, on which many other states have based their arbitration laws. Section 45 of the Arbitration Act 1996 gives courts the power to determine a preliminary point of law on the application of a party. This right may be excluded by agreement. The ICC and LCIA Rules exclude appeals on the merits, but do not exclude the section 45 right.

The argument for systematically publishing awards in treaty arbitrations

Treaty arbitrations concern disputes between states and foreign businesses, where the rights in issue arise not from private contracts but from treaties which the state parties have entered into. Awards are routinely published.

The public interest arguments for publishing treaty-type awards are clear. Citizens should know whether their governments have broken laws which those same governments signed up to and propagated. There may be substantial sums of public money in issue. If the state loses, its citizens will want to understand why, so as to be able properly to hold its government to account (assuming, of course, that the respondent state is a democracy).

A further argument for publication of such awards is that they have value as precedents. Publishing awards, so the argument goes, will help ensure consistent decision making. The treaties are, after all, public documents. They confer rights not just on the claimant in a particular arbitration, but on other, unrelated parties. And tribunals often purport to apply principles of transnational law which (if they exist) should be universal. The same issues may, therefore, come up in future arbitrations, and future tribunals faced with the same questions may, so the argument goes, find it helpful to know how their predecessors decided.

Whatever one thinks of that argument one would have thought it should have little traction when applied to commercial arbitration. More than ten years ago in Arbitral Precedent: Dream, Necessity or Excuse? Arbitration International, Vol. 23, No. 3 Kauffmann Kohler put it like this:

In commercial arbitration, there is no need for developing consistent rules through arbitral awards because the disputes are most often fact and contract-driven. The outcome revolves around a unique set of facts and upon the interpretation of a unique contract that was negotiated between private actors to fit their specific needs.

Yet the question of whether to publish commercial awards perennially resurfaces.

Is there really demand for more precedents?

Commercial arbitration is popular with businesses. A 2013 report of a survey and interviews of in-house counsel (PWC Corporate Choices in International Arbitration 2013) found that for transnational disputes, businesses generally prefer arbitration over litigation, with arbitration being much more commonly used than litigation in the energy and construction sectors, and somewhat less commonly used than litigation in the financial services sector.

On the question of cost and delay the survey is hard to interpret. The authors wrote that “for respondents who considered arbitration not to be well suited to their industry, costs and delay were cited as the main reasons”. At the same time, respondents (presumably those from industries like energy, shipping and construction which did use arbitration) reportedly complained about ‘judicialisation’ - increased formality of proceedings and similarity to litigation, with increased costs and delays. “In-house counsel” say the authors of this report “value the features of the arbitration process that distinguish it from litigation”.

Respondents were not asked whether awards should be published or a doctrine of precedent introduced. None seems to have raised this issue in interview. From this, it would seem that a desire for more precedents is not foremost in the minds of in-house counsel and captains of industry.

Some more recent data comes from a 2015 survey (2015 International Arbitration Survey: Improvements and Innovations in International Arbitration). Regrettably, this was not limited to industry but sought instead the opinions of “stakeholders at all levels in international arbitration”. Most of those who contributed were practitioners, arbitrators, academics, institutions, experts and ‘others’ (students?) with only 8% being in-house counsel. One might have thought the only real “stakeholders” in international arbitration are the businesses whose disputes it is supposed to resolve, and who ultimately pay for all the arbitration practitioners, arbitrators, institutions and experts, but the authors of this report obviously felt differently.

77% of respondents said there should be no appeals on the merits for commercial arbitrations. 33% said confidentiality and privacy - the traditional justifications which are given for non-publications - were one of the “most valuable characteristics of international arbitration”. But, at the same time, 64% of respondents said awards in “international arbitration” should be published in a redacted form or as summaries and 19% said full awards in “international arbitration” should be published.

In this survey, the term “international arbitration” was used to mean both investment arbitration and commercial arbitration, so it is impossible to gauge how much support there was for the publication of commercial awards specifically. There is also no breakdown given of the answers according to the respondent’s background, so it is impossible to say to what extent the publication of awards is something which in-house counsel are clamouring for as opposed to (say) academics.

In a 2016 lecture, entitled “Developing commercial law through the courts: rebalancing the relationship between the courts and arbitration” Lord Thomas variously said that:

The effect of the diminishing number of appeals compounds the problem that arises from the diversion of more claims from the courts to arbitration. It reduces the potential for the courts to develop and explain the law.

Arbitration confidentiality perpetuates public ignorance of continuing hazards, systemic problems, or public needs.

Such lack of openness … reduces the degree of certainty in the law that comes through the provision of authoritative decisions of the court.

… across many sectors of law traditionally developed in London, particularly relating to the construction industry, engineering, shipping, insurance and commodities, there is a real concern which has been expressed to me at the lack of case law on standard form contracts and on changes in commercial practice.

One cannot help but wonder whether it was really contractors, shipowners, insurers, commodities brokers etc. who were expressing this concern to Lord Thomas or lawyers in independent practice, and academics, who are operating at a rather greater remove from the commercial coalface.

His Lordship referred to a few means by which this (perceived) deficit of precedent might be addressed: (i) revision of the criteria for appeals; (ii) use of section 45 of the Arbitration Act 1996; and (iii) “greater recourse to the court instead of arbitration”.

Tellingly, his Lordship’s focus was on ensuring that these issues come before the courts, and are decided by judges, not arbitrators. He perceived there to be a need for more judicial precedents. He does not mention systematically publishing commercial arbitration awards as a means of obtaining valuable precedents. Presumably because such awards would be considered to have zero precedent value - no court asked to consider a standard form is likely to consider itself bound to follow what arbitrators had decided about that standard form in some previous case.

A further piece of evidence on the question of whether there is really a demand for more precedents is an ICC Report from November 2016. The ICC published a survey of financial institutions which concluded (emphasis added):

Apart from in the fields of M&A, asset management and banking advisory services, where confidentiality is of utmost importance, some financial institutions consider the lack of precedent to be a disadvantage of arbitration. Establishing precedents with respect to internationally recognised standard templates such as Loan Market Association (LMA) facility agreements or the ISDA Master Agreement is critically important. If the parties agree, an award that would otherwise be confidential may be published (with the possibility of redaction to avoid disclosure of sensitive information). In the sub-industries where precedent-setting is important, an industry-wide standard could be agreed, mandating institutions to publish redacted awards with the parties’ agreement.

The point that is being made here is that it is not for arbitral institutions like the ICC and LCIA to take it upon themselves to systematically publish awards. Arbitration is ultimately a creature of contract and the fact is that, if businesses really want awards to be published, then that is something which it is already within their power to agree. If the industry bodies which draft standard form contracts were really concerned about this, they could just include a clause mandating that awards are to be published in the next iteration of their standard form.

The very fact that, today, a great many business people remain content to have their disputes resolved by way of confidential arbitration is symptomatic of the fact that the non-publication of awards is not considered a big problem. If a given industry was really concerned about this issue it could amend its standard forms to mandate publication or vote with its feet, shift all its disputes to the courts, and thereby, over the next few years, obtain lots of precedents on the standard forms. The fact that this has not happened suggests that the supposed problem of a lack of precedents on standard forms has been overstated.

What is precedent anyway?

English law is a system of rules which the courts of England must apply in order to determine the rights and obligations of the litigants in a case before it. These rules are to be found in statute, enacted by Parliament and in the ratio decidendi of cases decided by the courts.

MacCormick (Why Cases have Rationes and What These Are in Goldstein (ed) Precedent in Law (1987) offers the following definition:

A ratio decidendi is a ruling expressly or impliedly given by a judge which is sufficient to settle a point of law put in issue by the parties’ arguments in a case, being a point on which a ruling was necessary to his justification … of the decision in the case.

Absent a conflicting statute, every court is also bound to follow the ratio decidendi of any case decided by a court above it in the hierarchy. Appellate courts, other than the Supreme Court, are bound by their own previous ratio decidendi.

This strict rule forms part of the “doctrine of precedent”, but that term also refers to a more nuanced practice, whereby courts look to previous decisions even in the absence of a binding ratio, and treat them as ‘persuasive’ or ‘suggestive’ as to how a court should be decided say where:

(a) The ratio decidendi of a previous decision does not settle the point of law in issue in the instant case, but the reasoning is in some sense analogous or suggestive of some wider principle which in turn suggests how some other point of law which is in issue in the instant case should be decided.

(b) A previous decision stated or suggested some wider rule or principle, but the decision in the earlier case would have been the same whether the wider rule or principle were correct or not (“obiter dicta”).

In deciding what weight is to be given to such precedents, a court might take into account the seniority of the court, the age of the case, whether the court had the benefit of all the arguments which are being presented in the present case, and whether other precedents were cited, how similar or how different the previous case was and so on. A court might even look to foreign judgments, or to academic commentary on judgments.

Precedent: costs and benefits

The doctrine of precedent, and argument from analogy is so ingrained in lawyers from the common law tradition that we tend to see the world through this prism, and rarely reflect upon how anomalous it is to reason in this way.

In other fields - science, engineering, business - and in our everyday lives we do not regard the fact that we decided one way in the past as raising some presumption that we should decide the same way in the future. If we consider our past decisions at all, it is only to ask: what worked? What didn’t? What could we do differently next time? We care only about making the correct decisions, not consistent ones. Consistency is a virtue only insofar as one is consistently right.

When it comes to making decisions about other people’s rights and obligations though, being right comes a distant second to being consistent and predictable. It is this which leads to judgments like Cavendish Square Holding BV v El Makdessi and ParkingEye Ltd v Beavis [2015] UKSC 67 where the Supreme Court was asked to consider simply abolishing the common law rule against penalties. Lord Neuberger and Lord Sumption noted that: “As early as 1801 … Lord Eldon confessed himself, not for the first time, “much embarrassed in ascertaining the principle on which [the rule was] founded”. Eighty years later … Sir George Jessel MR, not a judge noted for confessing ignorance, observed that “The ground of that doctrine I do not know”. In 1966 Diplock LJ, not a judge given to recognising defeat, declared that he could “make no attempt, where so many others have failed, to rationalise this common law rule”. Their Lordships said: “We rather doubt that the courts would have invented the rule today if their predecessors had not done so three centuries ago. But this is not the way in which English law develops, and we do not consider that judicial abolition would be a proper course for this court to take”. And so their Lordships continued to apply the rule.

The price of precedent, then, is that bad decisions may become ingrained, but this is supposedly outweighed by benefits of precedent. The theory is that consistency and predictability trumps the need to do justice in the instant case because our economies so depend upon consistency and predictability. Ultimately, the building you’re sitting in, the computer you’re reading this on (if anyone has made it this far), the pencils on your desk, and the electricity you used to boil the kettle for your cup of tea are only there because thousands of businesses were prepared to make contracts and to act on the assumption that those contracts were binding on them and on their counterparties and had a particular meaning and conduct their affairs accordingly.

Without wishing to be too dramatic, though, the fact is that all this has been possible without any systematic publication of commercial arbitral awards. Parties to arbitration agreements have been able to opt out of appeals on the merits for nearly 30 years, and arbitral awards have never been systematically published. And yet, in that 30 year period, industry bodies have continued to publish many successive iterations of their standard forms. Businesses have continued to make contracts on those standard forms (or more commonly, if we are completely honest, on bastardised versions of those standard forms). Things have continued to be manufactured, bought and sold. The world still turns.

It is noticeable that, over (say) the past 100 years ever more cases have come before the English courts. This is probably due to some combination of increased population, and increasingly complex society and businesses, globalisation, free trade, increased social mobility and entrepreneurship, increased regulation and a richer society with more resources in circulation and so more to argue over. Whatever the reasons, the fact is that there are far more case reports available than ever before and the number increases every day, these being easily searchable and retrievable through comprehensive online databases.

This presents something of a paradox. If precedents serve to fill in the gaps in the law and resolve uncertainties one might have expected that with an ever increasing pool of precedents there would be little left to argue about, and the law would get clearer and clearer. And yet legal arguments continue unchecked. Could it be that more precedents mean more to argue about, not less?

One can infer, from guidance published by the courts, that there is already a problem with too many precedents being cited by parties. The Commercial Court Guide says:

Authorities should only be cited when they contain some principle of law relevant to an issue arising on the application and where their substance is not to be found in the decision of a Court of higher authority.

PD 52c says that, for appeals to the Court of Appeal:

(4) The bundle should not–

(a) include authorities for propositions not in dispute; or

(b) contain more than 10 authorities unless the issues in the appeal justify more extensive citation.

What precedent value would arbitral awards even have?

Arbitral tribunals are not courts. No court is bound to follow an arbitral tribunal’s ratio decidendi. At most a party might refer to a tribunal’s award on some issue as a persuasive, but the court would not feel bound to follow it.

The same would be true of tribunals. Assume a contract requires an arbitral tribunal to apply English law. English law is statute and rationes decidendi from decisions of the English courts. Decisions of arbitral tribunals are not part of English law, and so arbitral tribunals who are to apply English law are not required to follow the decisions of previous arbitral tribunals. Such awards will be, at most, persuasive but never binding. Publishing an arbitral award is therefore never going to prevent any future arbitral tribunal deciding the same point differently, and it is never going to prevent different parties in a future dispute arguing about the same issues.

The basic principles of commercial and contract law were settled over 100 years ago in cases about horses and repairing water mill components. Commercial and contract law are not ‘feeling their way’ (unlike investment treaty law, which is relatively new). The result is that, in commercial arbitrations, it is relatively rare to have a serious disagreement about what the law is on (say) the interpretation of contracts, remoteness of damages, termination and all the other fundamental rules which come up over and over. The parties will often be (broadly) in agreement about what the legal principles are, at least where the lawyers are competent. The real battleground will be as to what the facts are, and how the legal principles, which are not themselves in issue, apply to those facts.

In practice, then, most commercial awards will be of little value as precedents, because so many of the issues which a tribunal looks at in any given award will be specific to that case and those facts and will never arise again.

Eventually, though, similar issues might come up in successive arbitrations (particularly where cases concern standard forms). But, since the second tribunal is not bound by the previous decisions, and might well decide differently the result is inconsistent precedents. This problem would be much more acute in commercial arbitration than in court, because in commercial arbitrations there are no appeals and there is no hierarchy of tribunals. So all these precedents have the same weight, and no award will ever provide a definitive answer. A bad decision will stand forever.

So what is the end result of systematically publishing awards? The answer is simply lots of awards. All of limited precedent value and with many being inconsistent to some degree. How many awards is hard to guess. But at least in some industries more disputes are resolved by arbitration than through the courts. So it could be that the number of awards published each year in those sectors would be similar to or exceed the number of judgments presently being reported.

For all these reasons, it seems unlikely that publishing arbitration awards will do much in terms of preventing future disputes, whether about standard forms or anything else.

Publishing awards: Cui bono?

If the systematic publication of awards is unlikely to prevent future disputes about the meaning of standard forms, then it is hard to see how systematic publication would deliver any substantial benefit for the businesses that arbitration is supposed to serve.

That is not to say, though, that no one will benefit from systematically publishing commercial arbitration awards. Someone will need to publish the awards, write headnotes and commentaries. Law firms, chambers and libraries will need to buy subscriptions to all this new material. Lawyers will have to spend time reading and arguing about all these extra cases. Arbitrators may even spend longer writing their awards, knowing they will be published. Ultimately, the additional costs of all this endeavor by other “arbitration stakeholders” will be borne by businesses which are involved in commercial disputes.

Of course, arbitrators, and lawyers who specialise in arbitration might welcome the publication of awards for other reasons too. The cases they work on will enjoy something of a raised profile. They will see themselves referred to in published awards and will be able to give more detail about their work in their promotional materials in the same way that their colleagues who specialise more in court litigation presently do. One reason why lawyers and academics might want to see the publication of commercial awards is also simply curiosity about the detail of these cases, where very substantial sums may be in issue.

I would even go so far as to say that it might sometimes be quite useful to see how an arbitrator has decided a particular point or interpreted a particular clause in a standard form, especially when deciding which arbitrators to appoint, or preparing a case for a given tribunal. But that it not the same as saying that published awards will have a strong precedent value, or be determinative of these issues to the point of preventing or narrowing future disputes.

The fact remains, though, that the systematic publication of commercial awards seems unlikely to prevent or narrow disputes or make arbitration quicker or cheaper and more likely to arrive at a just result. For that reason, such publication should be resisted.