The English Arbitration Act 1996 (AA 1996) proceeds on the basis that it is for the tribunal, not the court, to rule on its own jurisdiction, at least in the first instance. The Departmental Advisory Committee (DAC) hoped that legislative endorsement of the principle of kompetenz-kompetenz would “avoid delays and difficulties when a question is raised as the jurisdiction of the tribunal”.
It is clear, however, that where an arbitrator has ruled on jurisdiction, that ruling may be challenged in court as of right under the mandatory provisions of section 67 of the AA 1996. Any such challenge takes the form of a complete rehearing, rather than a review of the tribunal’s decision. The arbitrator’s ruling enjoys no privileged status; the court will take its own, independent, view of the underlying facts and evidence. At highest, the tribunal’s ruling may be “of interest” to the court (The Kalisti). Further, either party is entitled to adduce evidence which was not placed before the arbitrator. Although this is a logical consequence of the arbitrator’s inability to make a finally binding ruling on jurisdiction (because to do so might involve him or her pulling him or herself up by his or her bootstraps), it does involve an element of waste and inefficiency, which is arguably inconsistent with the overall policy in favour of efficient and speedy arbitration.
A striking example of this was provided earlier this year in The Pounda (Jiangsu Shagang Group Co Ltd v Loki Owning Company Ltd). The underlying issue in that case was whether the claimant had guaranteed the obligations of a charterer to pay hire to the defendant vessel owners. That, in turn, depended on whether the claimant had authorised the charterers to enter the guarantee on its behalf (an issue described as one of pure fact), turning in large part upon whether or not a conversation had taken place between SWM, a representative of the claimant, and the charterers.
When arbitration was commenced against the claimant, it challenged the tribunal’s jurisdiction. The tribunal disposed of the jurisdictional issue as a preliminary point. At the hearing, evidence was heard with an interpreter via video link from China. The tribunal rejected the challenge, ruling that it had jurisdiction to hear the claims. In reaching its decision, it rejected the video link evidence given by SWM, and found that he had in fact authorised the guarantee. They also drew adverse inferences from gaps in the disclosure provided by the claimant. The tribunal then went on to award over US $70 million to the defendant vessel owners.
The claimant challenged the ruling pursuant to section 67. This time, SWM travelled to the UK to attend the hearing and give live evidence. Carr J held that, contrary to the tribunal’s finding, the owners had failed to establish that the claimant had authorised the charterers to offer the guarantee. Her assessment of SWM’s evidence was entirely different to the tribunal’s. Despite inconsistencies between his evidence at the two hearings, she found him to be a credible and reliable witness, and accepted his account. The ultimate consequence was the setting aside not only of the jurisdictional ruling but also the substantive award.
The case provides a striking example of the waste and inefficiency that can arise in jurisdictional challenges, particularly those which turn on issues of fact. Is there any way of avoiding this?
In previous cases, the courts have suggested that such waste can be avoided by proceeding directly to court pursuant to section 32 of the AA 1996. However, an application under section 32 may proceed only if the parties agree or if the tribunal so directs. Interestingly, the DAC commented that section 32 was intended only for “exceptional cases” and should not become the usual route for challenging jurisdiction. In practice, agreement between the parties is unlikely once an issue has arisen, and the tribunal may well be reluctant to relinquish control over the issue of its own jurisdiction. In the present case, Carr J’s judgment records that the claimant’s application to have the jurisdictional issue determined by the court under section 32 was resisted by the defendant owners and then rejected by the arbitrators, who proceeded to decide the point themselves. The courts have, furthermore, confirmed that appeal to the court’s inherent jurisdiction to grant declarations is unlikely to provide a workaround, given the direction in section 1(c) of the AA 1996 that the court “should not intervene” save as provided by the AA 1996.
That suggests that the only possible means of deterring wasteful section 67 rehearings may be to invoke the court’s purely procedural powers to limit evidence (on case management grounds) and to visit costs consequences where spurious challenges fail. Further (and unlike in The Pounda), it may be worth encouraging the tribunal to refrain from proceeding to a final award unless and until the court has finally ruled on the jurisdictional issues. However, these are partial solutions: parties to London arbitration must take on board that any jurisdictional challenge is likely to be heard not once but twice, and will need to factor that into their costs.