Essop and Naeem: understanding indirect discrimination
In the joined cases of Essop and Naeem ( UKSC 27) the Supreme Court has taken on a daunting task: the simplification of indirect discrimination law. This is not a case note in the usual sense. We have not set out the facts, the law and then a statement of what is novel. At the hearing we tried to give the Supreme Court a new vocabulary to use as a tool for its analysis. The aim of this note is to explain that language as simply as we can. If we succeed, what we have to say will seem obvious. Those reading Lady Hale’s judgment (with which all of their Lordships agreed) will have had that experience. She has distilled, from an area of law that was submerging into great complexity, a handful of principles that dispel confusion and whch make intractable issues straightforward.
Start with direct discrimination
Almost everyone has an immediate intuitive understanding of direct discrimination. That is not to say that there are no difficult cases, but the core concept is easily grasped. Imagine an employer with an express policy of refusing to employ women. In a case of that sort the discrimination is obvious. To use the language of Equality Act 2010, s. 13, the employer treats women less favourably because of their sex.
The reason for the simplicity of direct discrimination is that it usually needs no context for the discriminatory impact of the criterion to be apparent. The criterion is inherently discriminatory.
What makes indirect discrimination different? Introducing the “context factor”.
What if the employer had a different criterion; a minimum height requirement? Unlike the policy in the earlier example, that criterion is not inherently discriminatory – it focuses on height not sex. If men and women are, on average, the same height, the apparently neutral criterion will be neutral in application. But men and women are not the same height on average. Women are on average shorter. Therefore, the criterion will exclude more women than men. In order for the discriminatory impact of the criterion to be apparent one needs context. The lower average height of women is what we call a “context factor”.
Indirect discrimination occurs when the employer’s criterion combines with one or more context factors to produce a disparity in outcome between people with a particular protected characteristic and those that do not have it. Put pseudo-mathematically one might say:
PCP + Context Factor = indirect discrimination.
Context factors can come in many different forms. The height example is a genetic difference. Length of service criteria are well-understood to discriminate against women. The relevant context factor is the social expectation that women will be principal carers for children which leads to a greater likelihood of career interruption.
These examples highlight two important things about context factors. First, they do not need to be in the control of the employer. It is not, as it were, the employer’s fault that women tend to be shorter or that a social expectation exists about childcare responsibilities. When the employer’s criterion combines with the context factor the result is an uneven playing field. The employer must then justify the use of the criterion. That is not to say that a context factor may not be within the employer’s control. It may, in fact, be another PCP, as it was in Homer (2012] UKSC 15;  ICR 704). In that case the PCP was the employer’s requirement that employees should have a university degree if they were to be promoted. The context factor was the employer’s retirement age. When the PCP combined with the context factor it had a discriminatory impact: employees who were close to their retirement age were unable to acquire the necessary qualification in sufficient time for it to be of any use to them.
The second important thing about context factors is that they are always “but for” causes of the discriminatory impact. A height requirement will not be discriminatory if men and women are the same average height. A length of service criterion would not be discriminatory if men and women were equally likely to interrupt their careers to care for their children. Without the context factor, there is no discriminatory impact, thus the context factor is always a cause. Lady Hale acknowledges this causal contribution by calling the context factor the “reason for the disadvantage”.
It is equally true, however, that the PCP is always a cause. If there is no height requirement, it does not matter if there is a difference in the average height of men and women. Similarly, if you avoid using a length of service criterion you avoid the discrimination that might otherwise occur. So both the PCP and the context factor are but for causes of the discriminatory impact. They both contribute to uneven slope of the playing field.
Do we need to be able to identify the context factor? (Essop)
This was the critical issue in Essop. In that case civil servants needed to pass a skills test if they wanted to obtain a promotion. BME employees did less well in the test. So did older employees. Statistics suggested that the was only a 0.1% likelihood that this on average poorer performance was unconnected to race. The same percentage likelihood applied to its being unconnected to age. What that meant was that there was a PCP – the test – and the playing field was demonstrably uneven and statistics suggested overwhelmingly that a context factor was in play but no-one could say what it was. The question for the Supreme Court was whose problem that was. The employer’s position was that the claimants had to identify the precise nature of the context factor. The Supreme Court has decided that they do not need to do so. If, as in Essop itself, the statistics point towards there being context factor that results in a disadvantage that is linked to the protected characteristic, that is all the Act requires. The employer must turn to justification. It cannot say: “it looks likely my PCP is indirectly discriminatory but unless you can explain exactly how it is having that effect, I should escape liability.” Indirect discrimination requires a claimant to establish not just the uneven playing field (i.e. the group disadvantage) but that they were individually disadvantaged. This led to much agonising in the Court of Appeal about “coat-tailing”. What if the claimant fails because he or she does not turn up for the test at all? If the requirement that they establish an individual disadvantage means only that they have to show that they failed the test (as colleagues who share their protected characteristic are, on average, more likely to) they win. That seems an unjust result. Lady Hale spends some time seeing off the difficulty. Again, it is suggested, the context factor analysis is helpful. In a case where an employee simply does not turn up, it is clear that his failure to attend is the whole of the causal context. The context factor that causes his disadvantage is not the same as the context factor that causes the group disadvantage. One can imagine a harder case. What if he is 10 minutes late for a two hour test? The tribunal is persuaded that his lateness was a cause of his failure but that it is satisfied on balance that the mystery context factor causing the group disadvantage also played a role? We think the tribunal should call on the employer to justify in those circumstances. Both context factors may be playing a role but each context factor only has to be a cause. It does not have to be the sole or principal cause. So the Tribunal should ask:
- Is there a PCP;
- Is there a context factor;
- Do they combine to create a group disadvantage;
- Is the same context factor a cause of the individual disadvantage?
If the answer to all four questions is yes, the employer should be made to justify the use of the PCP even if there were other context factors in play. If the tribunal is satisfied that some other factor entirely explains the individual disadvantage, the fourth question must be answered in the negative and the claim should fail.
Does the context factor need to be independently unlawful? (Naeem)
Mr Naeem is an imam who works for the Home Office as a prison chaplain. In his case the PCP was his incremental pay scale. Like all such incremental scales, the longer your service the more likely you were to have advanced higher up the scale. Although it was found as fact that a prison chaplain with 6 year’s experience had all the knowledge and expertise they were likely to need, it took a lot longer than that to progress to the top of the scale.
The relevant context factor was that the Home Office had not been willing until 2002 to recruit Muslim chaplains as employees. They were engaged instead on a sessional basis. The justification was that there were insufficient numbers of Muslim prisoners to justify employing imams.
The combined effect of the PCP and context factor was an uneven playing field: Muslim chaplains were more likely to be found towards the lower end of the scale. Naeem, then, is a case like Homer, where two PCPs (i.e. the incremental pay scale and the pre 2002 policy of non-recruitment) are combining to produce the discriminatory effect.
The Court of Appeal took the view that the “real” cause of the uneven playing field was the context factor and that the decision not directly to employ imams prior to 2002 was justified. But of course the context factor causes the disadvantage. If it did not have a causal impact, it would not be a context factor. But the question the Act requires is whether the PCP causes the disadvantage. It plainly does. If there were no incremental scale and chaplains were paid a flat rate, the playing field would be even. It does not matter if it is not the “main” cause, it simply has to be a cause. Nor does it matter if the context factor is “justified”. As observed above, context factors may well be things over which the employer has no control at all and which, on any view are lawful (e.g. the fact women are on average shorter than men). The proper question is whether the use of a PCP is justified in a context in which it produces a discriminatory impact. This was the analysis adopted by the Supreme Court and its ability to untangle the knot the created by the Court of Appeal’s approach makes it a tool for analysis that is likely to assist tribunals very considerably.