In the mysterious world of investment treaty arbitration, there has been much debate over several decades as to whether a given matter, said to condition the tribunal's right to hear the case at all, is truly one of jurisdiction, or is more properly classified as one of admissibility.
What is the difference? In essence, jurisdiction concerns whether a tribunal has the power to adjudicate the claim in question at all, whereas admissibility assumes that power and goes to whether the tribunal will choose to exercise it. As Butcher J put it in PAO Tatneft v Ukraine  EWHC 1797 (Comm) (at ):
"Issues of jurisdiction go to the existence or otherwise of a tribunal's power to judge the merits of a dispute; issues of admissibility go to whether the tribunal will exercise that power in relation to the claims submitted to it."
Perhaps even pithier is the explanation offered by the Singapore Court of Appeal last year in BBA v BAZ  SGCA 53 (at ):
"Jurisdiction is commonly defined to refer to 'the power of the tribunal to hear a case', whereas admissibility refers to 'whether it is appropriate for the tribunal to hear it'."
The stakes are high: if the decision of an arbitral tribunal goes beyond its proper jurisdictional limits, it may be shot down by a higher controlling authority (namely the courts of the seat). On the other hand, mistakenly classifying issues of admissibility as jurisdictional inevitably leads to unwarranted challenges and frustrates the parties' expectation that their dispute will be resolved by their chosen neutral tribunal.
In an 'English' context, the distinction lies between those decisions that are truly jurisdictional, which are subject to review under section 67 of the Arbitration Act 1996, and those that are classified as issues of admissibility, on which there is no appeal or review, absent some form of procedural irregularity or bad faith, for example.
Until relatively recently, the question has rarely been debated in what we may describe as purely commercial cases, i.e. where no state was involved. Indeed, almost all of the cases where it has been discussed (and in England all the reported decisions) have involved a sovereign state. Further afield, the writings on the issue are multitudinous. Standing out in the field must be Professor Jan Paulsson's 2005 contribution Jurisdiction and Admissibility published in an ICC series entitled Global Reflections on International Law, Commerce and Dispute Resolution, freely available online.
The High Court decision last week in Republic of Sierra Leone v S L Mining  EWHC 286 (Comm) is another example of a dispute involving a state, but it certainly has ramifications for commercial arbitration generally.
The key issue in Sierra Leone concerned the dispute resolution clause in a mining licence agreement. Its wording resembled that found in many modern day contracts, providing for the parties to spend a given period attempting to settle the dispute before commencing arbitration. In investor-state disputes, such a clause is commonly included in the underlying bilateral investment treaty. Readers may recall a similar provision that appeared in the distribution agreement in Emirates Trading Agency LLC v Prime Mineral Exports  EWHC 2104 (Comm), another Commercial Court decision from 2014.
In Emirates Trading, the dispute resolution clause contained an obligation to negotiate in good faith for not less than 28 days before commencing arbitration proceedings. The (relatively experienced) tribunal treated the clause as a jurisdictional precondition to starting the arbitration (but found that it had been complied with). Emirates Trading appealed under section 67. Teare J, while agreeing that the clause had been complied with, ruled that compliance with such a precondition was a jurisdictional issue. Much of the judgment was taken up with older English authorities, and a then recent Australian decision (United Group Rail Services v Rail Corporation New South Wales (2009) 127 Con LR 202), concerning the enforceability of an obligation to negotiate in good faith (albeit in United Group Rail the issue arose in rather a different context and was nothing to do with arbitrability).
By that time, of course, English law had recognised the enforceability of an obligation to mediate, but only by imposing a stay on proceedings (as in Cable & Wireless v IBM  EWHC 2059 (Comm)) or by penalising a party for unreasonably refusing to mediate (as in Halsey v Milton Keynes General NHS Trust  1 WLR 3002).
The Emirates Trading case went a step further, however, and introduced the concept of an obligation to negotiate in good faith as a breachable contractual duty. However, the decision did not fully grapple with the ramifications: what were the consequences of failing to comply with the pre-arbitral requirement to negotiate in good faith? And in this context, what was meant by "in good faith"? What if the negotiations only lasted two days, and not 28 days? Or one party alleged bad faith? If the negotiations are privileged, how would bad faith be proven? What exactly was required by the duty to negotiate? What would be the consequence of a breach? How could loss be proved? Teare J posited the notion of a claim based on loss of chance, but how would that stand up to scrutiny? All valid questions, though perhaps the biggest criticism of the decision in Emirates Trading was its failure to treat the compliance with the requirement to negotiate as purely an issue of admissibility rather than jurisdiction, and therefore not susceptible to a challenge at all.
In Sierra Leone, the clause contained a similar requirement for the parties to "in good faith endeavour to reach an amicable settlement of all differences of opinion or disputes which may arise between them", and that in the event of a failure to do so within three months of a written notice of dispute and a request for such "amicable settlement", either party could march off to the ICC and start arbitrating in London. Unfortunately, the licensee decided not to wait the full three months, and issued its request for arbitration after about six weeks. The state complained to the tribunal that it had no jurisdiction as a result. The Sierra Leone tribunal took a different tack to the Emirates Trading troika; it regarded the compliance as purely an issue of admissibility, and ruled in favour of the licensee. The state launched its challenge and, praying in aid Emirates Trading, involved section 67 of the Act.
In an admirably clear and concise judgment, Sir Michael Burton considered carefully three recent English cases in which a distinction had been made between jurisdiction and admissibility, two of whom (including Tatneft v Ukraine, above) had been the subjects of judgments by Butcher J, albeit both involving states. He also noted criticism of the decision in Emirates Trading by Merkin and Flannery on the Arbitration Act 1996 (as well as in an earlier Arbitration International article by the same authors), particularly its failure to consider whether compliance with pre-arbitral procedures was really a matter of substantive jurisdiction, rather than admissibility.
In noting that firstly, the decision in Emirates Trading had not considered the distinction between admissibility and jurisdiction (it not having been argued in that case) and secondly, that on the facts, the decision on the effect of the clause was strictly obiter, Sir Michael Burton firmly concluded that compliance with pre-arbitral procedures was indeed an issue of admissibility. Section 67 was thus not engaged. In this, as he noted, he was singing from the same hymn sheet as that of the Singapore Court of Appeal in the BBA decision, from which we took the pithy quotation above.
In conclusion, we can say that the decision in Sierra Leone is to be welcomed, not least for it finally adding to the chorus of disapproval of the dictum in Emirates Trading, but for clearly and comprehensively allocating compliance with pre-arbitral procedures to the tray marked "admissibility", not the one marked "jurisdiction".