A movie director casts only male actors for the role of Romeo in his upcoming adaptation of Romeo and Juliet. A restaurant owner does not employ female chefs because if they become pregnant, he will have to hire a replacement. One exclusion is unproblematic, the other is wrongful discrimination. What explains the difference? In Faces of Inequality, Sophia Moreau sets out to answer when and why discrimination is wrong. Her main audience are moral and political philosophers, as well as lawyers. But the book is also of interest to economists and economically interested philosophers. If we set out to understand the extent of discrimination in labour markets, for example, we would need to explain why only the restaurant owner’s preference is discriminatory. And given that racial and gender inequalities, among others, are shaped substantially by discrimination, discrimination is central to egalitarianism more broadly.
Sophia Moreau, for some time a leading philosopher on discrimination, has written an excellent book on the topic. Moreau disagrees with an unstated assumption in the debate on discrimination that there has to be one explanation when and why discrimination is wrongful. In Faces of Inequality, Moreau proposes three answers unified in a pluralist theory of discrimination. The first answer, laid out in Chapter 2, is that discrimination is wrong when and because it unfairly subordinates members of a group. Moreau gives a rich and sophisticated account of unfair subordination. Subordination involves that members of subordinated groups have lesser power and de facto authority and are ascribed traits that attract lesser consideration by others. These facts are in turn rationalized and justified by stereotypes and by what Moreau calls ‘structural accommodations’, broadly policies that appear neutral but privilege the interests of dominant groups as the default.
The second answer, found in Chapter 3, is that discrimination denies individuals a specific kind of freedom. This deliberative freedom is as much a freedom of thought as it is a freedom of action. It is the freedom, in certain contexts, not to treat some fact about oneself, for example one’s sex, religion or race, as a cost. The third answer, developed in Chapter 4, is that discrimination is wrong when it denies individuals access to a basic good. A basic good in Moreau’s sense is a good that is necessary for individuals to have in order to be and be seen as an equal in society. While some basic goods are material goods (Moreau mentions the clean water crisis among Canada’s First Nations), the participation in shared institutions, such as the ability of homosexuals to take part in the institution of marriage, also counts as a basic good in this sense.
The first half of the book, after an introductory chapter, is devoted to exploring these three reasons why discrimination is wrong. Moreau does not claim that these are exhaustive but rather that they capture most cases of discrimination. In the second half of the book, Moreau highlights three key innovations of her theory. First, there is the pluralist nature of her theory of discrimination (Chapter 5). According to Moreau, what unifies the different reasons why discrimination is wrong is that they all are interpretations of the abstract idea of failing to treat people as equals. Second, Moreau advances an understanding of discrimination which treats indirect discrimination as much as a form of discrimination as direct discrimination (Chapter 6). Third, Moreau’s account of discrimination applies to the same extent to governments as well to private individuals (Chapter 7). While Moreau discusses these features in their own chapters, to get a sense of Moreau’s thinking on these matters one needs to read the whole book.
There is much to be admired in Moreau’s book which I believe will rightfully be a key text on the matter of discrimination. But in this review, I will challenge two key features of Moreau’s overarching theory. First, I disagree that the idea of failing to treat people as equals is enough to provide a unified theory of discrimination. Second, I challenge the idea that we should treat indirect discrimination on a par with direct discrimination.
1. Pluralism about discrimination
Pluralism about discrimination is appealing for a simple reason. No single theory of discrimination seems to fit well the varied nature of discrimination. Discrimination law has grown over the last half century or so and included more and more conduct within its ambit. Given that Moreau’s starting point is that discrimination law provides us with a good picture of discrimination, this means that the chances of there being a single reason why discrimination is wrong are dim (25–26). Why not instead appeal to a multitude of reasons that together explain when discrimination is wrong?
We are then still faced with the task of explaining why all these distinct reasons together form a theory of discrimination. Moreau’s answer is that the different accounts are unified because they all provide us with different interpretations of the general idea of discrimination, namely acting in ways that fail to treat some people as equals.
But it appears to me that the idea of failing to treat people as equals is broader than discrimination. If so, we need a more specific answer to what unifies the different accounts of discriminations. For example, a government that gives important political power to a royal family fails to treat all citizens as equal partners in government. Yet it would be a stretch to call this discrimination against commoners. The broad idea of failing to treat individuals as equals should make room both for a theory of political equality and for a theory of discrimination. It plausibly should also make room for a theory of economic equality. Moreau may want to dispute this. In discussing economic inequality (7–8), she seems to understand the concern about economic inequality as a concern about the pattern of a distribution, not about the ideal of treating people as equals. But it would be a mistake to entirely separate distributive and relational concerns. At points in the book, Moreau uses distributive language for her own account. She writes that judgements about social subordination rely on the ‘relative amounts of power, authority, deference, and structure accommodations enjoyed by different social groups’ (141). The focus on ‘amounts’ of power and so on suggests that distributive judgements are relevant to the relational idea of social subordination. The relational ideal is fundamental, but it makes distributions of some goods relevant. A similar structure can underlie a concern with economic inequality. We can understand Ronald Dworkin’s (Reference Dworkin2000) equality of resources in this way. Dworkin highlights that equality of resources is embedded in a broader ideal of treating people as equals. It makes the abstract idea more precise and tells us why we should be concerned with specific unequal distributions.
Discrimination then cannot be equivalent to failing to treat people as equals. We need to distinguish discrimination from forms of political and economic inequality that also fail to treat people as equals. Coincidentally, Moreau raises a second unifying feature of her different accounts of wrongful discrimination. What sets wrongful discrimination apart from other forms of failing to treat people as equals may be that they are all forms of discriminating between people on the basis of certain traits (157–158). (In the same way that we act indiscriminately if we fail to distinguish between different people.) But if this is all that distinguishes discrimination from (political or economic) inequality, then this answer sits awkwardly with other elements of Moreau’s theory. First, Moreau argues convincingly against any emphasis on differentiation. The key to a theory of wrongful discrimination is not, for Moreau, how people are classified (3–7). She writes ‘our worries seem best conceptualized, not just as concerns with inappropriate differentiation, but as concerns with disadvantage, and with disadvantage of a sort that are inappropriate because they fail to treat some people as the equals of others’ (6). Those concerned with economic inequality can say precisely this, they are concerned with the disadvantage of those relatively poor, a disadvantage that fails to treat them as equals. Second, Moreau is refreshingly open to which traits count as relevant for discrimination. She does not fix which kinds of traits are relevant by some prior definitional act. Traits matter because they are made relevant by an account of failing to treat people as equals. The idea here is correct. The different traits that are picked out by anti-discrimination law are unified because they are morally similar, not because of a definitional fiat. But then it would be implausible to appeal to a definitional fiat to explain how wrongful discrimination is unified.
A more plausible avenue would be to provide a thicker understanding of ‘failing to treat people as equals’. For pluralism to succeed, and I think Moreau’s discussion shows great promise, this idea would have to be thick enough to distinguish itself from other egalitarian concerns yet not so thick as to end up giving another a monist answer, for example by collapsing into an account of social subordination.
2. Indirect discrimination and causal contributions
When it comes to direct and indirect discrimination, Moreau sees little that morally distinguishes them. Direct discrimination refers to disparate treatment on the basis of a relevant trait, e.g. sex. Indirect discrimination refers to actions which use an otherwise unproblematic trait, e.g. height, which correlates with a relevant trait, e.g. sex, such that discriminating on the basis of height leads to disparate impacts between the sexes. While Moreau acknowledges that the two categories are useful to be kept apart for analytical purposes (185–191), both are wrongful and equally wrongful for the same reason. I disagree with Moreau; indirect discrimination is often best understood not as the personal wrong of discrimination but rather as a form of social injustice.
Moreau, however, describes the view that indirect discrimination is different from direct discrimination as an ‘odd place to start’ (20). For Moreau a theory of discrimination should start with the law of discrimination because ‘our ideas of discrimination seem to [her] to owe so much to our legal frameworks’ (14). Since the laws of most countries, except notably the USA, treat indirect discrimination as an instance of discrimination, this is Moreau’s starting point, too. Only if our attempt at theorizing discrimination leads to a different conclusion should we treat indirect discrimination as different. Accordingly, Moreau discusses in Chapters 2–4 how the different reasons why discrimination is wrongful apply to indirect discrimination, too.
I doubt that Moreau is correct that ‘our’ ideas of discrimination are strongly influenced by discrimination law. Her inclusive view widely expands what counts as discrimination. In a discussion of the gender pay gap (16–17) Moreau suggests that the practices that cause the gender pay gap are almost all forms of discrimination. This would include the fact that professions of typical ‘women’s jobs’ are less well paid than ‘men’s jobs’, or that women are more likely to work part-time. However, economists and other social scientists frequently use discrimination as a separate explanatory variable. Differences in job prestige or part-time status are separate explanations not attributable to discrimination. Economists evidently do not share the expansive view common among lawyers.
But Moreau provides additional arguments as to why we should adopt the unified view. Throughout the book, when discussing how her accounts apply to both direct and indirect discrimination, Moreau relies on contrasting cases of direct and indirect discrimination that appear very similar. In some cases, they show too much, however. Moreau compares a ban on turbans on football (’soccer’) fields which according to her is direct discrimination against Sikhs with a requirement to wear the specific ‘Stetson hat’ for police ceremonies which according to her is indirect discrimination against Sikhs (105). In this case it is indeed easy to see why there should be no relevant difference between the two policies. A requirement to wear the Stetson hat just implies a requirement not to wear a turban. It is mysterious why one policy is indirect and the other direct discrimination.
Regardless of whether the Stetson hat policy is indirect or direct discrimination, it appears very different from other cases of indirect discrimination. This suggests to me that key cases of indirect discrimination are importantly different from direct discrimination. For example, admission to socially prestigious engineering education in India is determined by means of a formal exam of mathematics and science. Using this exam means that, if it was not for affirmative action, hardly any Dalits or members of other oppressed castes would be admitted. This case appears to fit Moreau’s definition of wrongful indirect discrimination. The disproportionate outcomes in these competitive exams perpetuate stereotypes about members of oppressed castes as uneducated, untalented or lazy. By contrast, upper castes are portrayed as meritorious and naturally gifted. These stereotypes help justify differences in power and de facto authority by naturalizing essentially social disparities. Such entrance tests can also infringe on one’s deliberative freedom when year after year Dalits see how only very few fellow Dalits perform well in the tests. This makes caste an issue in the context of education. Moreau makes similar observations about employment tests used in North America.
When Moreau discusses these examples, she adds the detail that the reason why the tests fail members of disadvantaged groups is that they test cultural cues, rather than relevant ability (66, 106–107, 202–203). The Indian entrance examinations, however, test a plausible conception of ability by sticking to mathematics and the natural sciences. For Moreau’s theory this should not matter. Stereotypes are strengthened and caste becomes an issue regardless. Thus, we should say that these entrance examinations fail to treat Dalits as equals. However, Moreau is open to the idea that some forms of discrimination are justified despite treating people as inferior, second-class citizens (169–174). Unfortunately, she does not give a comprehensive theory of justification. She does, however, refer to the idea, found in some jurisdictions, that such indirectly discriminatory policies are justified only if there is no alternative and suitable test available that could increase the proportion of underrepresented groups (203). Since there is a plethora of ways to test academic potential, it seems likely that the entrance examinations would be wrongful discrimination.
This policy is, however, very different from direct discrimination such as banning Sikhs to play football. I agree with Moreau that there are good reasons why the law should require individual actors not to use such entrance examinations. However, I disagree with Moreau’s further claim that they are also a form of discrimination and not only an injustice. For Moreau, this is because both acts of indirect and direct discrimination often exhibit a form of negligence (199–208). The agent is failing to consider the interests of the discriminatee with appropriate weight. But this need not be the case here. The university may use a simple test of ability because it is unaware of all possible tests and their effect on minority representation, rather than out of negligent insensitivity to the interests of Dalits. Moreau thinks such cases are rare. We can be reasonably expected to be aware of the history of exclusion of subordinated groups and should therefore consider the impact on these groups (207). But precisely because we are aware of the history and presence of exclusion of Dalits and the lesser opportunities and educational capital enjoyed by Dalits, we should expect tests that track academic ability to be passed by fewer Dalits. What is relevant for the charge of indirect discrimination is the highly complex judgement of the exact proportion yielded by this test as opposed to others.
Even if indirect discrimination is often blameless, Moreau insists that when such acts contribute to unfair subordination or infringe deliberative freedom, they are discriminatory. The contribution clause is key here. At one point Moreau argues that the duty not to discriminate appears in many cases like an imperfect collective duty. She adds that her understanding of an imperfect duty does not imply that it is up to the agent to decide how to fulfil the duty. Rather, the agents’ circumstances and other demands determine the content of the duty (230–232). This idea leaves room for a possibility that Moreau does not discuss. It might be that many forms of prima facie indirect discrimination are not in fact discriminatory because the individual agent’s contribution is too small. One employer or one university using aptitude tests which, while tracking ability fail to do so in a way that improves the prospects of subordinated groups, contributes only little to the perpetuation of stereotypes and denial of deliberative freedom. Only if such tests are common across the job market or education sector, do these effects indeed happen.
This is not to suggest that private individuals should not be required to bear the burdens of eliminating these inequalities. Moreau argues convincingly that they should. After all, if they don’t shoulder the burden it will fall on already disadvantaged groups (197–198). But this argument is best read, in my view, as an argument that policy and law should distribute burdens to private agents, not that individuals commit the moral wrong of discriminating if they use such tests. The problems Moreau highlights which arise out of indirect discrimination are better analysed as a form of social injustice, not individual wrongs.
Conclusion
Despite my reservations, Moreau’s book is an outstanding contribution to the discussion of discrimination and beyond. Chapter 2 can be read as a terrific outline of relational egalitarianism that does more than many existing formulations to specify what it means to relate to others as an equal. There are interesting connections between Moreau’s idea of a basic good, in Chapter 4, and ideas of poverty and sufficiency. All in all, this is an excellent contribution that deserves thoughtful attention.
Acknowledgements
I thank Sophia Moreau, Thulasi K. Raj and Johanna Thoma for comments and discussion.
Bastian Steuwer is a post-doctoral associate at the Center for Population-Level Bioethics. His current research focuses on equality in the distribution of health care, risks and discrimination.
