There are some contracts that cannot be easily analysed under existing classifications of legal relationships. An example with which readers may be familiar is a software transfer agreement, which, in the words of Professors Green and Saidov, “…is not truly analogous to any conventional chattel with which the law is familiar”. Similarly, identifying the nature of the contract between an arbitrator and the parties is more elusive than one might think. One possibility – that it is agency – is fraught with obvious difficulties. Another – service – is attractive at first sight but is also open to objections. The issue is not merely academic, for modern legislation is often triggered by a classificatory question – for instance, it is necessary to examine whether a software transfer agreement is a sale or a licence or something else entirely, in order to ascertain whether a buyer is entitled to the heightened protection that the Sale of Goods Act confers, whether the transaction can be brought to sales tax, and so on. So it is with the arbitration contract, which the UK Supreme Court has discussed at length in its recent judgment in Jivraj v Hashwani.
In 1981, Mr Jivraj and Mr Hashwani entered into a Joint Venture Agreement, governed by English law, which provided that any dispute would be referred to three arbitrators, all of whom must be respected members of the Ismaili community in the UK. Disputes arose in the late 80s and finally Mr Hashwani appointed Sir Anthony Colman as an arbitrator. Sir Anthony is, of course, a highly respected judge who retired recently from the English High Court, but not a member of the Ismaili community. Mr Hashwani argued that he was nevertheless entitled to appoint him because the arbitration agreement was void insofar as it purported to discriminate on the basis of religion, and contrary to the Employment Equality (Religion or Belief) Regulations 2003 [“the Regulations”].
That issue is, of course, a question of public law. The reason it is important for our purposes is that the Regulations apply only to “employment under a contract of service or of apprenticeship or a contract personally to do any work…”, and it was therefore necessary to ascertain whether an arbitrator’s contract falls within one of these descriptions. At first instance, Steel J. held that an arbitrator is not “employed” within the meaning of the Regulations, but was reversed by the Court of Appeal, which held that an arbitrator is employed because his contract with the parties is a “contract personally to do any work”. The Supreme Court reversed the Court of Appeal, restored the judgment of Steel J. and its judgment contains a valuable discussion of the nature of the arbitrator’s contract.
Lord Clarke began by noting that the expression used in the Regulations was not simply “contract personally to do any work” but employment under such a contract, and said that “the role of an arbitrator is not naturally described as employment under a contract personally to do work.” Uncertainty in this area arose principally from apparently divergent approaches adopted by the English courts, as a matter of domestic law, and the European Court of Justice. The Court of Justice had held in three decisions, of which the most important is Allonby  ICR 1328, that the gravamen of the test is whether the person in question is an “independent provider of services … not in a relationship of subordination” with the person receiving the services. On the other hand, English domestic cases had repeatedly adopted what Lord Clarke called the “dominant purpose” test ie whether the dominant purpose of the contract is the execution of personal work or labour. It is not necessary for our purpose to explore whether this test is consistent with European law, and it suffices to say that Lord Clarke held that dominant purpose cannot at any rate be the sole test, because the European cases regard the relationship between the recipient and provider of the service as more important than the purpose of the contract.
Applying this test, it was fairly clear that an arbitrator does not fall within the Regulations – as Lord Clarke notes, even though an arbitrator pays VAT and is for that purpose a service provider, he does not “perform those services or earn his fees for and under the direction of the parties…” The understanding of an arbitrator’s role in general as well as in English law is that he acts independent of the control of the parties, short of both parties agreeing to terminate the arbitration. Lord Clarke was careful to distinguish this from the position of a judge because although judges too act independently in the adjudicatory process, “they are in other respects not free agents to work as and when they choose”, subject as they typically are to service rules of the State they serve. As Lord Clarke notes in paragraph 45, the analogy between a judge and an arbitrator is misleading, and the true analogy – for the purpose of this question – is not the relationship between the judge and the State, but the relationship between the judge and litigants.
Lord Mance gave a brief concurring judgment and referred to two sources that help considerably in identifying the proper analysis of an arbitrator’s contract. The first was a decision of a German court in 1904 that held that an arbitrator cannot be treated as equivalent to a “representative, employee or an entrepreneur… his office has an entirely special character”. In short, an arbitrator’s contract is a sui generis agreement. That is precisely the conclusion that Lord Mance’s second citation – Gary Born – advances, noting that while it is prima facie attractive to treat an arbitrator as a service provider, it ignores that the adjudicative aspect of his function crucially distinguishes him from a banker or a doctor.