Review of A Theory of Discrimination Law by Tarunabh Khaitan
Comments made at a book launch
Time to read
1. It is not often one comes across a book that really sharpens the focus on a familiar subject. I have been teaching the law relating to discrimination all my professional life. After such a long engagement with the subject, I was not expecting to discover anything new in Dr Khaitan’s book, but my low expectations were confounded. This serious monograph has really got me thinking again very hard about the subject. Standing back from lots of fascinating detail in the meticulously argued position, I was particularly impressed, though also deeply troubled, by two big claims in the book.
2. First, Dr Khaitan joins what may still be a minority group in claiming that, despite the name of the UK statute being the Equality Act 2010, the general aim of discrimination laws is not primarily concerned with equality but rather with the well-being achieved through living an autonomous life. The general aim, he claims, is to reduce (and perhaps eventually remove) any significant gap in advantages to enjoy the essential goods for human flourishing, which are in this context liberty, access to valuable opportunities, and self-esteem, between a protected group defined by reference to an immutable or a valuable ground (such as respectively sex or religion) and a cognate group such a persons of the opposite sex or a different religion. According to Dr Khaitan, the reduction of inequalities between individuals is not the point of anti-discrimination law; and the reduction of inequalities between groups such as men and women is not the purpose of the law but rather a side-effect of attempting to create valuable opportunities and freedom for women as a relatively disadvantaged group. In making this claim, Dr Khaitain is explicitly rejecting theories of discrimination law that derive its underlying philosophy from ideas of either formal or substantive equality; and I think he is also rejecting theories that are sometimes described as ‘prioritarian’, which say in one way or another (like my own theory of social inclusion) that what the law is and should be doing is addressing the worst forms of disadvantage and social exclusion. So the first striking claim is that the general aim of the law of discrimination is not equality, not social inclusion, those two being the normal justifications put forward for these laws by politicians, but rather the pursuit of a soft perfectionist notion of freedom or autonomy, which says that discrimination law helps to achieve liberty, access to valuable opportunities, and self-esteem.
3. In a second big claim, Dr Khaitan asserts that, in pursuit of this general aim of promoting autonomy, the law intervenes to prohibit certain actions by certain duty-holders such as employers because those actions will (either intentionally or accidentally) worsen the gap between the advantages enjoyed by two cognate groups such as men and women in an unjustified way. This formulation tends to eliminate the contrast normally drawn between direct and indirect discrimination, because the difference is reduced to the issue of the probability of the action actually worsening the gap between the advantages of the two groups. In the case of sex discrimination, direct discrimination against women is bound to exacerbate the gap in advantages (if only in a small way as for instance depriving a woman of a job to which she was better qualified than a man); as for indirect discrimination, the application of a facially neutral rule may or may not increase the disadvantage to women or other disadvantaged groups, and so a finding of discrimination will become a matter of degree both in respect of the proportions of the groups affected and possibly the degree of harm suffered. This emphasis on the effects of actions also avoids any assessment of fault, blameworthiness, or motive in reaching a determination of whether or not there has been discrimination. This view is in line with the much criticised decision in James v Eastleigh BC. It rejects the (historically accurate) view that indirect discrimination exists merely as a supplementary technique to prevent employers from avoiding laws against direct discrimination by the device of introducing facially neutral rules that in practice achieved the same effect, and instead sees its disparate impact test as the key ingredient of all legal prohibitions on discrimination. Dr Khaitan’s formulation of a unified effects-based test of discrimination is also notable for the way in which a defence of justification is generally available for both indirect and direct discrimination, which is not generally thought to be the case in law. In sum, the prohibition on indirect discrimination, with its emphasis on adverse impact on disadvantaged groups, is, for Dr Khaitan, the paradigm of discrimination law, and direct discrimination is merely distinguishable because the adverse effect on a particular group seems certain.
4. One of the attractions of avoiding a general justifying aim formulated in terms of equality is that it can more straightforwardly justify various kinds of affirmative action programmes that favour disadvantaged groups. Affirmative action or positive action is a bit different from the general law of discrimination because it does not prohibit particular actions. Instead, it may either provide an excuse for direct discrimination against a privileged group such as men for the sake of helping a disadvantaged group such as women, or it may suggest that a duty to act in favour of a particular group should be placed upon controllers of opportunities such as employers. Dr Khaitan observes (p. 164) that such legal provisions for affirmative action encounter hostility because they permit direct discrimination, ie discrimination against men solely for reasons of their sex, in order to help women, and therefore seems to contradict the aim of discrimination laws if it is formulated in terms of equality. Viewed from the perspective of their effects on group disadvantage, affirmative action measures seem, however, to fit exactly into the model of discrimination law proposed by Dr Khaitan that intervenes to reduce comparative group disadvantage. Incidentally, he notes in his discussion how affirmative action that targets people within the group who are particularly disadvantaged would be desirable because that would make the case for imposing a duty stronger. I note his embracing of a more prioritarian approach at this point, and wonder whether he has done enough to separate his position from it. He is tempted down that path, I suspect, because he is worried about flouting too ostentatiously the expressive dimension of discrimination law and the very far from empty idea of equality. Even so, it is to my mind one of the major attractions of Dr Khaitan’s approach that affirmative action falls within the purpose of discrimination laws rather than being a contradiction of its purpose
5. A wider possible attraction of this theory is that it permits justifications of direct discrimination, not only in cases of affirmative action, but also apparently in a wider context. Although direct discrimination is bound to adversely affect the disadvantaged group, Dr Khaitan leaves open the door for justifications (as he is forced to do when he amalgamates direct and indirect discrimination). Of course, in the law of direct discrimination there are some defences or justifications already – such as some kind of genuine occupational qualification for a job, or the possibility of justification in relation to disability where there is no possibility of accommodation. What additional justifications, if any, for direct discrimination does Dr Khaitan envisage? He introduces (at p.181) the novel idea of ‘permitted discrimination’ with the illustration that in choosing my friends I can reject people on any ground whatsoever. Later he explains that I am not a duty bearer in any case when it comes to choosing my friends, because the law does not impose anti-discrimination duties on private individuals when choosing their friends. But he asserts that permitted discrimination is not the same as justified discrimination. Justified direct discrimination will arise, Tarun concludes, ‘in those rare cases where there is no blame whatsoever’ (p. 185). On reading that statement, however, I began to wonder if this argument smuggles intention or motive back into direct discrimination by the back door? Having insisted that James v Eastleigh was right in its emphasis on the effects of the rule about concessions in the swimming pool and that the benign intention of the council was irrelevant to an effects based test, why does the author now bring back in the defence of the ‘blameless’ conduct of the Council? He sees the problem I have identified (namely that he is dropping the effects test and reintroducing the lay approach of moral blameworthiness in the test of justification) and so he suggests in response to this kind of objection that the Council in James v Eastleigh acted recklessly and so were blameworthy after all (p. 185). He suggests that the Council would not have been guilty of direct discrimination if the concessions had been limited to those actually living on pensions (entirely?). Is that a convincing distinction regarding moral blameworthiness between choosing those entitled to pensions (fault) and those living on pensions (blameless), and, in any case, it is impracticable task to distinguish these categories at the ticket booth. But my general point is that the introduction of a justification for direct discrimination based upon ‘morally blameless’ conduct will run the risk of undermining his proposed effects-based test of discrimination and restoring some lay view that discrimination implies fault.
6. Although that concern about justifications for direct discrimination is a small point, it is in fact part of a much broader question about the architecture of his theory.
In Chapter 1, Dr Khaitain adopts the HLA Hart’s structure of separating the general justifying aim (GJA) of a particular branch of the law from the more detailed distributional rules that impose rights and obligations on citizens within that general framework set by the aim. I have long thought that such a structure could be applicable to all legal subjects, including crime, tort, contract, and discrimination law and would produce fruitful insights. The question that has always troubled me about this structure, however, is to identify or understand the conceptual linkage (to try to use a neutral word) between the general justifying aim and the more detailed distributional rules. Hart stresses how the link is not a tight one, but the question remains what kind of link or constraint there exists between the aim and the detailed rules of application. Dr Khaitan argues that the general justifying aim for discrimination law is (as we have noted) to reduce relative disadvantages of groups with respect to the conditions for autonomy. With respect to the detailed distributional rules, he adopts (as we have seen) an effects based test of discrimination: in other words, if an action harms a member of a group because of membership of that group, there is discrimination, though it may be justified. For Dr Khaitan, the main attraction of an effects-based test for the distributional rules of discrimination is surely because it gels well with the general justifying aim: if the aim is to remedy group disadvantage, then a test for discrimination that requires an instance of the exacerbation of group disadvantage looks entirely appropriate. Indeed, the Hartian structure that separates out the questions begins to appear completely unnecessary owing to the close fit between GJA and distributional rules. Dr Khaitan’s theory of discrimination law begins to look like what Hart would disparagingly call a univocal theory. Dr Khaitan is aware of this issue and has to wrestle at great length with the question of how far he is going to deviate from his chosen tram lines. Consider, for instance, his discussion of why discrimination law permits claims by men (as in James v Eastleigh itself). Assuming that men are a favoured group, claims by men are not within the GJA as defined in this book and nor do they pass his effects test of discrimination. So does he advocate abolition of such claims? No, that would be a step too far, since then Dr Khaitan’s theory would really not fit a great deal of the law, and he is supposed to be writing a theory of the law as it is. He introduces two complex ideas to justify this extension of discrimination law to men (and other favoured groups where there is symmetrical protection). The first is what he calls ‘collateral discrimination’ (p.145), which simply permits some symmetrical claims, and the second is the ‘expressive clause’ which explains that such collateral claims are permitted when not allowing the dominant group to make a claim (in this case men), would damage the interests of the disadvantaged group (women). Although these arguments are clever and intriguing, though perhaps indeterminate, Dr Khaitan dodges what may be the simple and correct explanation of why men are permitted to bring claims for sex discrimination, which is simply that the legislator passed a rule demanding formal equality of treatment between men and women as a symbolic gesture of formal equality. Why does he not like that straightforward explanation? My surmise is that the formal equality rule does not fit his conception of the general justifying aim of discrimination law, so he does not want to embrace it as an explanation. If that surmise is correct, then what we see in his book is what I would regard as dangerous univocal tendencies. In other words, having adopted in Chapter 1 Hart’s insights regarding the difference between questions about goals and questions about distributive mechanisms, the author then proceeds to ignore those insights and tries to sell us another univocal theory like all his competitors. It is no doubt a great theory, full of brilliant insights and original ideas, but, like all univocal theories of legal doctrines or bodies of law, it must be wrong because the life of the law is not logic but experience, and experience is very messy.