In today’s workplace it is not uncommon to find individuals engaged on terms not constituting a classic employer/employee relationship. Many people operate under the auspices of “brass-plate” companies or other arms-length arrangements entirely legitimately, even if the relationship they create might in the eyes of the uninformed outsider be seen as employment in all but name. As we know, this may not be the case.
The recent “gig economy” cases involving Uber, CitySprint and Addison Lee have brought these unorthodox arrangements to into the spotlight but these have all concerned rights attributable to “workers” under the definition in s.230(3) Employment Rights Act 1996. However, under that statute it is largely unknown for individuals to be on the receiving end of claims (a notable exception being the ability to name individual tortfeasors as Respondents to whistleblowing detriment claims under s.47B). The position under the Equality Act 2010 is markedly different and the conceptual distinctions between it and the Employment Rights Act can lead to individuals being named inappropriately or not named at all when it would be prudent to do so.
The starting point in any discussion of this nature is s.110 Equality Act 2010. This section, referring back to s.109’s liability for employers, sets out that an employee may be personally liable for discriminatory acts so long as they have been done in the course of employment. It is well known that the definition of “employee” for Equality Act purposes is much broader than under s.230 Employment Rights Act 1996 and (generally) encompasses any individual who would, under the latter, be either an “employee” or a “worker”. Section 110, however, also uses the concept of “agent” as being someone who might be liable for discrimination. This is a concept not found under the 1996 Act.
Problems have arisen where an individual is quite obviously not an employee, even in a broad sense, yet they are potentially responsible for acts of discrimination. If that individual is not an “agent” they cannot personally be liable, nor – arguably more importantly – can the Claimant’s employer be liable for what they do. There therefore exists a lacuna, and since third-party harassment (s.40(2) Equality Act 2010) was abolished in 2013 no matter how heinous the discriminatory act is said to be, the boundaries of liability stop at the agent’s doorway.
Historically there had been a tendency to view the concept of “agent” quite broadly. This approach was rejected by the Court of Appeal in Kemeh v Ministry of Defence  IRLR 377. Although Elias LJ accepted that it was not necessary for an agent to have the power to affect a principal’s relations with third parties, the absence of any statutory definition of “agent” in the Equality Act 2010 must follow the common law position. In summary, the correct post-Kemeh formulation of the definition of “agency” consisted of a fiduciary relationship which existed between two persons, one of whom expressly or impliedly consents that the other should act on his behalf… and the other of whom similarly consents so to act or so acts pursuant to the manifestation. Simply carrying out work on behalf of another person would not be enough. It was, as can be seen, a restrictive approach which widened the lacuna.
What was of significance in Kemeh was the emphasis on the fiduciary nature of agency, and it has fallen to Employment Tribunals to consider that specific element where it is otherwise agreed that an individual had the principal’s authority to do something. Would individual x, for example, have had a duty to act without conflict of interest as regards the putative principal? Would x have had a duty to act in the putative principal’s best interests? What was the extent of the autonomy granted to x? And perhaps most crucially of all, did even a quasi-trust and confidence relationship exist between them? These are difficult conceptual questions.
In Unite the Union v Nailard  EWCA Civ 1203 the Court of Appeal considered several issues but also the question of whether elected trade union officials could be categorised as “agents” for Equality Act purposes. On the facts, the Employment Tribunal found that they were and the EAT agreed.
Underhill LJ, giving judgment, considered the history of the question insofar as it involved fiduciary considerations but distilled the analysis in Kemeh down to a “starting point”, namely that “the principal will be liable wherever the agent discriminates in the course of carrying out the functions he is authorised to do” (paragraph 42). Following Nailard, the focus must initially be on the scope of x’s authority. As Underhill LJ recognised, this essentially equates the test with that of the “course of employment” test for liability familiar to us from s.109(1) and the infamous “works night out” lines of authority.
However, it is important to observe that Nailard does not overturn Kemeh, nor does it challenge Elias LJ’s reasoning on the key question of who is an agent. That question remains, and Underhill LJ was careful to distinguish between the case of a union official, where those individuals would by definition have the union’s authority to do certain things, and the case of the alleged tortfeasor in Kemeh, who merely carried out work on behalf of the Claimant’s employer through a hierarchy of contractors. In the latter case it would be necessary to identify whether there was an agency relationship at all, and to that end Kemeh’s rigid approach survives (paragraphs 42 and 43).
Therefore, those wishing to pursue or defend claims where the alleged discriminator is not an employee will out of necessity require a careful analysis of whether sufficient fiduciary qualities exist in the relationship between the putative agent and principal. In reality, do they have authority to do anything? This will often be a difficult judgment call for practitioners but a consideration of the scope of authority, coupled with due regard to the type of questions posed above should, I hope, prove useful.