The Supreme Court has arrived at what almost all arbitration practitioners and clients will view as the right result in the strange episode of Jivraj v Hashwani. The Supreme Court has unanimously allowed the appeal on the basis that an arbitrator is not an employee of the parties for the purposes of the Employment Equality (Religion or Belief) Regulations 2003 – a conclusion which, on its face, seems so unremarkable as to almost invite the question of what all the fuss is about.

Given that the Court of Appeal had reached the opposite view, however, this was no foregone conclusion. The practical significance of this development should not be underestimated. The Supreme Court has delivered an important and very welcome outcome.

The facts do not need to be repeated at length, not least because by now they may be very familiar to anyone reading this blog. The parties had entered into a joint venture agreement which contained an arbitration clause. The arbitration clause provided for specific and unusual appointment criteria: “All arbitrators shall be respected members of the Ismaili community and holders of high office within the community.

Was this a valid arbitration agreement? Mr Hashwani said not, arguing that it was void because it purported to discriminate on the grounds of religious belief. He said that the clause could not stand because the UK Equality (Religion and Belief) Regulations 2003 (giving effect to Council Directive 2007/78 EC) (the “Regulations”) prohibited employment discrimination on this basis.

At first instance, David Steel J rejected that case, essentially on the grounds that arbitrators could not properly be construed as being employees of the parties appearing before them. It is not easy to argue with that. However, the Court of Appeal then caused considerable dismay by holding that the arbitration agreement was discriminatory under the Regulations. It got worse: the Court of Appeal found itself unable to sever the discriminatory criterion from the rest of the clause, with the outcome that the arbitration agreement was void.

In order to reach that decision, the Court of Appeal of course had to find that an arbitrator is an employee for the purposes of the Equality Regulations – on its face, a surprising conclusion, and one based upon the premise that an arbitrator is somehow directly employed by the parties under “a contract personally to do any work.”

By upholding Mr Jivraj’s appeal, the Supreme Court favoured the approach of the first instance judge, David Steel J, over the line taken by the Court of Appeal. The Regulations did not apply, because in English law an arbitrator is not an employed person under “a contract personally to do any work.”

An arbitrator is, instead, an “independent provider of services”, a “quasi-judicial adjudicator” [whose] “functions and duties require him to rise above the partisan interests of the parties and not to act in, or so as to further, the particular interests of either party…he must determine how to resolve their competing interests. He is in no sense in a position of subordination to the parties; rather the contrary.” No-one who has served as an arbitrator, appeared before one or submitted a dispute to arbitration could argue credibly with any of this. Lord Clarke’s summation indicates, in diplomatic and veiled terms, just how badly the Court of Appeal missed the mark: “it is in my opinion plain that the arbitrators’ role is not one of employment under a contract personally to do work.”

There was a second issue, which was, if the Regulations did indeed apply, whether or not an arbitrator ‘employee’ could be employed by reference to discriminatory religious criteria because belonging to that faith was a “genuine occupational requirement for the job”. Unsurprisingly given the anti-discriminatory purpose of the Regulations, it is not easy to satisfy this test. It is strictly applied. On these facts, however, the majority of the Supreme Court found that the test would have been satisfied – so that even if the Regulations applied, and an arbitrator was an employee, then the arbitration clause would have been upheld in any event because it would have been a “genuine occupational requirement” that the arbitrator be a respected member of the Ismaili community. Lord Mance did not express a final view on this point.

The consequences of the appeal being upheld are much less draconian than those that would have followed had it been dismissed. They can probably be shortly stated: there will be less scope to query London as a good choice as an arbitral seat; London law firm model forms will change back to their pre-Jivraj incarnation; and, if they wish to resolve their dispute through arbitration, Messrs Jivraj and Hashwani will have to identify arbitrator nominees from the Ismaili community (which, given that the case has been described by counsel for Hashwani as a ‘hot potato’ within that community, may be no easy task).

This first point raises a question of anecdotal evidence. Throughout the life of the Jivraj saga, colleagues advocating the merits of arbitral seats competing with London for business have often pointed to the Jivraj factor as a reason not to touch London with the proverbial barge pole. But was the significance of the factor always overstated? How often, in reality, was the Jivraj factor determinative in causing parties to avoid London as an arbitral seat?

As for the judgment itself, it will not be easy to find an arbitration practitioner or client who believes that the Supreme Court got it wrong on either point, but any contrary views would be very welcome.


To make sure you do not miss out on regular updates from the Kluwer Arbitration Blog, please subscribe here. To submit a proposal for a blog post, please consult our Editorial Guidelines.

Kluwer Arbitration
This page as PDF


  1. Craig, have a look at the discussion at the LinkedIn group International Arbitration and the cases and directives referred to there.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.