affiliated to the International Association of Constitutional Law
Some readers might be surprised if told that one of the most significant cases on discrimination law generally, and race discrimination in particular, is likely to be decided by the Supreme Court before long. The UKSC heard the appeal against the Court of Appeal’s ruling in Home Office v Essop (2015) in December 2016. It is still to deliver its judgment.
The surprise will be understandable, given the media’s silence on this case. The case is not, after all, about outright racist prejudice or hate, the sort of thing that invites moral disgust and media outrage. It is a subtler case, but with profound implications for racial minorities in the UK. Readers can look up doctrinal niceties in a note on this case [132 Law Quarterly Review (2016) 35]. In this post, I wish to discuss its broader policy implications.
These are the facts: the Home Office administered a generic (i.e. not tailored to the actual job) Core Skills Assessment (CSA) as a first step towards determining its staff’s eligibility for promotion. The CSA selection rate for non-white candidates was 40.3% of the rate for white candidates, and there was a 0.1% likelihood that this could happen by chance. The selection rate of over-35 year olds was 37.4% of the selection rate for under-35 year olds, and again there was only a 0.1% risk that this could happen by chance (at -). Given these findings, there must have been some non-random explanation for the disproportionate impact of the CSA on non-white and older staff. However, neither the Home Office nor the claimants knew what this explanation was. The existence of an explanation was almost (99.9%) certain, what was lacking was the knowledge of what that explanation was. If one is permitted some jargon, the problem was epistemological rather than ontological: or, if you prefer, the Court of Appeal (CA) was confronted with a ‘known-unknown’. Could the claimants establish prima facie indirect discrimination in this case without showing what the explanation was? Sir Colin Rimer, writing for the CA, held that they couldn’t.
If affirmed by the SC, this outcome could reverse a significant turn in Anglo-American discrimination law that started with the decision of the US Supreme Court in Griggs v Duke Power Company (1971). In that case, an employer’s requirement of a graduation degree for a job—whose satisfactory performance bore no relation to one’s being a graduate, but one that had the effect of excluding most black candidates (disparate impact)—was deemed unlawful under the Civil Rights Act 1964. The concept of disparate impact was imported into the UK Sex Discrimination Act 1975 and Race Relations Act 1976 as ‘indirect discrimination’ (Hepple, p 608-9).
Ever since Griggs, claimants have only borne the burden of establishing disparate impact caused by the policy, practice or criterion in question, never the burden of also explaining that the disadvantaged suffered was because of their race. Sure, courts have often taken judicial notice of plausible factors associated with race, such as poverty, inadequate access to education, living in an under-achieving neighbourhood, absence of role-models, lack of friends and relatives who can pass on the cultural capital necessary for success etc. But they don’t ask claimants to establish that they failed because of their race, or a factor associated with race, rather than for other reasons (such as laziness, incompetence, lack of intelligence, failure to turn up on time for the exam and so on). After the claimant has made a prima facie case by showing disparate impact, the burden shifts to the defendant, who can still defend the policy by showing that it is a proportionate means of achieving a legitimate aim. The CA, in Essop, changes this long-established doctrine by placing the burden not only of causation, but also of explanation, on the claimant. It does so by appealing to the ostensible conceptual impossibility of proving causation without explanation , even though this is very much possible: popping an aspirin causes my headache to go away, whether or not I can explain precisely why that happens; asbestos is a known carcinogen, even though scientists cannot quite explain how it causes mesothelioma.
If confirmed, this judgment could have very serious consequences for the future of indirect discrimination law, whose entire point is to unearth and redress discrimination that is typically harder to detect, let alone explain. In indirect discrimination litigation, given the aggregate nature of the claims, it is hard enough for information-poor claimants to establish causation. The burden of explanation will all but kill the chances of a successful litigation in most cases, except those where an explanation is obvious (women and inflexible working hours, Sikh men and a turban ban etc). This is especially pernicious in the context of race (and, no doubt, other grounds) where indirect discrimination mainly occurs due to existing socio-cultural, political, and material disadvantages faced by a group. These disadvantage-based factors tend to be multiple, diffused, and affect everything we do, but in subtle ways: they include the postcode we live in, the schools and universities that we (and our parents) go to (if we/they do), the monthly income of our families, the social networks we have, the material and cultural capital we inherit and so on. It is a combination of these and other elusive factors that often result in the disproportionate impact on protected groups. Even some single-factorial explanations, such as unconscious bias against the group, may be very difficult to prove in particular cases. The worst hit would be claimants disproportionately disadvantaged by the well-disguised prejudice of a shrewd employer: direct discrimination cannot be proved (because the prejudice is well-disguised and the impact disproportionate); neither can indirect discrimination (because the only explanation for the disproportionate impact is disguised).
The move beyond what is still commonly understood by many laypersons as discrimination—direct, prejudicial, motivated—to indirect, structural, unconscious impacts on disadvantaged groups marked a remarkable shift in the legal meaning of discrimination (Khaitan, A Theory of Discrimination Law, OUP 2015 pp 1-4). In particular, this was a shift from conceptualising discrimination as something that happens in the mind of the discriminator to thinking of the phenomenon as experienced by its victims. The law moved away from a strict corrective justice model of reacting to wrongdoing, towards a hybrid-model, where a corrective justice structure was retained to serve broader distributive justice ends (Khaitan and Steel forthcoming). This move towards a hybrid model for indirect discrimination was reinforced by including a broad justification defence, making the possibility of damages exceedingly rare, and allowing the regulator to bring a claim.
The CA’s judgment in Essop seeks to turn the clock back on this development. By requiring the claimants to provide an explanation for their failure, it seeks to reimagine indirect discrimination in the image of direct discrimination: discrimination as bad for bad reasons rather than for its bad effects. The law’s broader understanding of discrimination entails the acknowledgement of the lived reality of those it is meant to protect. Going back to the older, narrower, meaning would be a significant reversal of hard-won gains.
Tarunabh Khaitan is an Associate Professor in Law at Wadham College, University of Oxford.
(Suggested citation: T. Khaitan, ‘Giving up on (Indirect) Discrimination Law’, U.K. Const. L. Blog (13th Mar 2017) (available at https://ukconstitutionallaw.org/))