+44 (0)20 7797 8600




Most of our readers are intelligent and well informed so the answer is probably "quite well”. After all, the concept has been superbly analysed by two former Presidents of the EAT in London Borough of Islington v Ladele [2009] IRLR 154 (Elias P) and Amnesty International v Ahmed [2009] ICR 1450 (Underhill P), and of course most authoritatively by the Supreme Court in the Jewish Free Schoolcase [2010] 2 AC 728.

So here is a pub quiz question to test the extent of that knowledge.  

A claimant ("C”) complains about a specific decision – say, to terminate a consultancy agreement, although it could be to dismiss, or not to promote, it does not matter. She asserts that the decision was taken on the ground of the protected characteristic age, but again it could be race or sex etc. The decision was taken by a sole decision maker ("X”), but his decision was based on the views of another employee ("Y”) who had compiled a report about C. There is prima facie evidence that X was indeed acting on the ground of C’s age, so the burden of proof shifts against the employer ("R”) against whom the complaint of direct discrimination is made.  Is it sufficient to discharge the burden of proof and defeat C’s complaint of direct age discrimination for R to show that X, the sole decision maker, was not acting on the grounds of age? Or is it also in principle necessary for R to show that Y, when compiling the report which significantly influenced X’s views, was also not acting on the grounds of age? And if it is in principle necessary to consider Y’s motivation as well, does it matter whether any complaint had in fact been made about those views by C? 

In fact those are multiple questions but no matter…

The scenario in which the act complained of is taken by a sole employee but is based on information provided by other employees is a familiar one, and the discrimination statutes have been around for 40 years or so, so you might think that the answers to those questions should be obvious.

Singh J provided answers in Reynolds v CILFIS (UK) Ltd[2014] ICR 907. He held that:

  • It was in principle necessary to consider the thought processes of Y, the report maker, as well as those of X, the ultimate decision maker. Otherwise "a real injustice might be done” to the C if the underlying report was based on reasons that were discriminatory and that report had a significant influence on the decision;
  • Support for that conclusion could be derived from the well- known speech of Lord Nicholls in Nagarajan v London Regional Transport [2000] 1 AC 501, in which he said at one point that "If racial grounds or protected acts had a significant influence on the outcome, discrimination is made out”.
  • Because Y’s mental processes were in principle relevant it was necessary to consider them whether or not any particular complaint had been made about Y by C, as the claim was brought against the employer (who was just as responsible for Y’s actions as it was for X’s actions) and the burden of proof had shifted against the employer. Once the burden of proof had shifted it was for the employer to show that the impugned decision was "in no sense whatsoever” on the prohibited ground: see Igen Ltd v Wong [2005] ICR 931, and that entailed proving that Y’s thought processes were not tainted by the prohibited ground.

These answers were clear…but also troubling. The implications for the conduct of discrimination claims were profound. Proving the actual reasons for a sole decision maker’s decision can be tricky enough. But at least the employer will know who it has to call to give the relevant evidence: the decision maker. On the Singh J analysis, however, a Respondent who called only that person (X), would be running a risk if the burden of proof was held to have shifted against it. Because although X could prove that in fact he was not acting on the prohibited ground, if the burden of proof shifted it would also be necessary for the employer to prove that the thought processes of Y, whose report influenced X’s decision, were also innocent. So it would be necessary to call Y to give evidence as well, even if no complaint were made about his thought processes. And if Y’s decision was itself based on information provided by other employees (Z and others) logically they too would have to be called. So even if a discrimination claim could not be simpler – a complaint about a single decision taken by a single person – it might be necessary to call multiple additional witnesses to prove that their prior decisions were in no sense whatsoever tainted by the protected characteristic. 

The Court of Appeal however has now overturned the decision of the EAT, in CLFIS (UK) Ltd v Dr Reynolds [2015] EWCA Civ 439. The sole substantive judgment is given by Underhill LJ and it is characteristically illuminating. He held that as a matter of principle it was necessary to consider only the thought processes of the sole decision maker. Although the problem of near endless regression – i.e. the problem identified above of having to call multiple witnesses to give evidence about prior acts - featured heavily in argument, Underhill LJ’s decision ultimately was based on a different consideration: the fact that it is possible to sue an individual as well as a corporate respondent for acts of discrimination. He said this (at para. 36): 

"…I believe that it is fundamental to the scheme of the legislation that liability can only attach to an employer where an individual employee or agent for whose act he is responsible has done an act which satisfies the definition of discrimination.  That means that the individual employee who did the act complained of must himself have been motivated by the protected characteristic.  I see no basis on which his act can be said to be discriminatory on the basis of someone else’s motivation.  If it were otherwise very unfair consequences would follow…. rendering [the employer] liable [for the act complained of] would make X liable too… It would be quite unjust for X to be liable to C where he personally was innocent of any discriminatory motivation.”

He also held that the burden of proof provisions did not have the effect of requiring the employer to prove that Y’s decision making was not tainted by the protected characteristic even if no complaint had been made about that decision. He held (at para. 51) 

The effect of [the burden of proof] provisions was not to place a blanket obligation on [the Respondent] to prove the absence of discrimination in every act of every employee that formed part of the chain of causation leading to the act complained of. On the contrary, the starting-point is that the Claimant was required to prove a prima facie case (in the sense explained in Madarassy) that the termination of her contract was discriminatory. Whether she reached that stage had to be decided by reference to the specific case which she advanced.  …. To put it another way, the burden of proof provisions apply for the resolution of the factual issues raised before the Tribunal: they cannot operate to extend those issues. 


The CA’s decision will obviously be welcomed by Respondents, but perhaps also by tribunals, who have the onerous task of trying to keep within reasonable bounds unwieldy discrimination complaints. The implication of the decision for Claimants is clear: it is necessary to identify with precision the specific act or acts complained of (whether it is the ultimate decision, the decision on which it was based or both), and to ensure that each specific act complained of is appropriately pleaded. Respondents now know that it will be necessary to call those responsible for the acts complained of to explain their reasoning, but that it will not be necessary to engage in a form of legal archaeology, picking over historic decisions that formed part of the chain of events leading to the ultimate decision, and wondering whether it is necessary to call those decision makers, even though no complaint has been made about their decisions. And individual respondents can breathe a sigh of relief that they cannot be held responsible for specific acts of discrimination on the basis of others’ impugned mental processes.

There it is. Quiz over… now for a drink!

Daniel Tatton Brown (instructed by Sarah Lamont and Jane Wallenstein at Bevan Brittan LLP) acted for the successful appellant in CLFIS (UK) Ltd v Dr Reynolds [2015] EWCA Civ 439. The judgment can be found here.

Posted: 01.05.2015 at 11:53
Tags:  Comments  Employment Law  Equality & Discrimination
Share this page
Print page

Cookies help us deliver our services. By continuing to browse this website, you agree to our use of cookies. OK