Anti-discrimination law is often referred to as “equality law” (Holmes 2005: 185). This is not coincidental. Conceptually, discrimination is tied to inequality. It is impossible to discriminate against someone unless there is some dimension in which the discriminator treats the discriminatee worse than those against whom she does not discriminate (Alexander 1992: 191; Fiss 1976: 109; Gardner 1996: 355; Heinrich 2007: 109). To be sure, for any given conduct that qualifies as discrimination due to adverse treatment in one dimension, there may be other dimensions in which the discriminator treats the discriminatee equally with others. A charity organization that serves pork chops to Muslim and Christian homeless alike treats them unequally in the dimension “serving food not prohibited by the recipient’s religion.” Yet, it also treats homeless people equally in the dimension “type of food, gastronomically speaking, served to recipients.” The latter does not show that the charity does not discriminate. It simply reflects that any action admits of an endless number of true descriptions.
Morally, it is natural to assume that discrimination is wrong, at least in part, in virtue of the inequality it involves. Moreover, if discrimination is necessarily pro tanto wrong, it must be wrong in virtue of its essential properties and unequal treatment is one such property. Yet, clearly it is not wrong per se to treat some worse than others in some dimensions, e.g., a judge who releases the innocent and convicts the guilty does not act wrongly. Hence, friends of an equality-based account of the wrongness of discrimination must specify the specific kind of unequal treatment that wrongful discrimination involves.
Equality-based accounts of the wrongness of discrimination differ from non-equality-based ones. One such account says that discrimination is wrong, when it is, because of the harmful effects it typically involves, where harm is determined relative to how well off the discriminatee would have been in the absence of discrimination (and not relative to, say, how well off the discriminatee would have been had everyone been treated equally. The subset of harm-based accounts that identifies harm through such an equality-fixed baseline—i.e., someone is harmed if she is worse off than she would have been had she been treated equally, e.g., by not having been subjected to discrimination, nepotism, etc.—is a species of equality-based accounts.) Since it is not a necessary property of discrimination that it harms anyone, harm-based accounts imply the possibility of non-wrongful discrimination. Unless, of course, they build into the very definition of discrimination that it is wrongful, harmful differential treatment. I favor a harm-based account of the wrongness of discrimination (Lippert-Rasmussen 2006). This account has problems of its own, but these are not my concern here.
A given token of discrimination may be wrong for equality as well as non-equality-related reasons. My focus is exclusively on the former. There are two ways, at least, in which the wrongness of discrimination could be explained on egalitarian grounds. Derek Parfit distinguishes between telic and deontic egalitarianism. According to the former view, “it is in itself bad if some people are worse off” than others through no choice or fault of their own (Parfit 1998: 3). According to the latter, “it is not in itself bad if some people are worse off than others” (Parfit 1998: 6). While it is often unjust that some are worse off than others, deontic egalitarians do not object to the inequality itself: “What is unjust, and therefore bad, is not strictly the state of affairs, but the way in which it was produced” (Parfit 1998: 7). It is not necessary for the inequality-producing process being unjust that it led to outcome inequality (Lippert-Rasmussen 2007a). Hence, deontic egalitarians may find discriminatory acts unjust even if, ironically, they result in discriminators and discriminatees being equally well off. Telic egalitarians may have no complaints in such a case.
Parfit says little about which features of the production of an outcome make it unjust. I assume that deontic egalitarians think that it is the feature of some people treating others unequally in a way that is morally wrongful. This needs elaboration and I return to it shortly. Suffice here to say that subscribing to the formal principle of equality that people who are not relevantly different should be treated equally does not make one a deontic egalitarian.
Telic egalitarianism implies that discrimination is not morally bad or wrong as such. Undoubtedly, when institutionalized and widely practiced discrimination results in large groups of people ending up worse off than others through no choice and fault of their own, telic egalitarianism condemns this outcome. However, discrimination may not result in anyone’s ending up worse off through no choice or fault of their own (cf. Holmes 2005: 193). First, discrimination against a certain group may counterbalance nonchoice advantages that it enjoys such that discrimination restores equality, in which case telic egalitarianism deems discrimination instrumentally good in one respect, e.g., ethnic prejudice against a certain wealthy group of immigrants may counterbalance inequalities in educational opportunities resulting from this immigrant group’s superior financial situation.
Second, suppose people are only discriminated against on the basis of properties they have through their own choice (or fault); suppose that, due to pharmaceutical inventions one can choose one’s sexuality through drugs; and suppose that people who are discriminated against end up worse off than others. If discrimination on the basis of sexuality is wrong in such a case, telic egalitarianism may fail to explain it.
It might be suggested that telic egalitarians should be concerned not with outcomes, but with opportunities. However, as the case of wrongful age discrimination reminds us, wrongful discrimination can coexist with equality of opportunities where this is cashed out in terms of the Rawlsian requirement that “any two persons with the same native talent and the same ambition should have the same prospects of success in the competition for positions of advantage that distribute primary social goods” (Arneson 1999: 73; cf. Rawls 1971: 66, 72). Equality of opportunity so understood fails to assess negatively wrongful discriminatory patterns of socialization that cause, say, gender differences in ambitions. Also, like with equality of outcome, wrongful discrimination against a certain group could in principle counterbalance its better opportunities in nondiscrimination dimensions.
Some have argued that telic egalitarians misconstrue the point of equality (Anderson 1999; Miller 1998; Scheffler 2003). It is not that everyone has equal amounts of some good but that individuals relate to one another as equals. This is incompatible with various forms of domination, servility and oppression, but it does not require distributive equality, i.e., in an egalitarian, democratic society not so well-off people need not see themselves as socially inferior to well-off people despite the latter people’s superior income and fortune. In response, Fleurbaey (2008: 245) has argued that on a suitably permissive understanding of the relevant good, this view is a species of telic egalitarianism, say, one that deems inequality of social standing to be bad. Whatever the merits of this response, it is important here that while, historically, discrimination has often generated social hierarchies, it may subsist together with their absence and yet seem morally objectionable. Suppose we live in a multiethnic society in which there are ten equally powerful ethnic groups. Members of each group discriminate against members of all other groups such that everyone is equally subjected to ethnic discrimination and no ethnic group dominates the others. Here discrimination does not translate into “highly visible and morally irrelevant differences into systematic social disadvantage” (Sunstein 1994: 2411), thereby violating any anti-caste considerations: no group is disadvantaged relative to any other and no citizen is a second-class citizen. If discrimination in this scenario is morally wrong, a concern for the absence of social hierarchy cannot explain it. So let us turn to the other of Parfit’s two main egalitarian categories, deontic egalitarians.
Deontic egalitarians hold that discrimination is unjust to the extent that it involves treating others unequally in a way that is morally wrongful. There are a number of ways in which discrimination may be thought to do so. First, some egalitarians have held a mental state-focused position according to which a discriminator acts unjustly if her differential treatment of others reflects what I will call an inegalitarian mental state, e.g., a belief to the effect that people do not have equal moral status. Second, others share with mental accounts the focus on the lesser moral status expressed through the discriminatory act but think that what matters morally is that the unequal treatment is objectively, i.e., independently of the discriminator’s mental states, demeaning and, thus, clashes with equality of moral status. Third, some believe that discrimination involves unfairness in that it treats the claims of different individuals unequally and that this is what makes discrimination wrong (cf. Broome 1994: 37; Halldenius 2005: 456). These accounts do not exhaust the logical space of possible deontic egalitarian accounts, but they all explain the wrongness of discrimination in a way that involves an essential reference to the ideal of equality.
I focus on the first two deontic accounts: in part because they have been most fully developed in the literature on discrimination; in part because they seem well placed to explain the wrongness of some of the most repugnant forms of discrimination. More specifically, I focus on Larry Alexander’s and Deborah Hellman’s deontic egalitarian accounts of the wrongness of certain forms of discrimination. My aim is to show why, despite their ingenuity and insightfulness, neither is satisfactory. More generally, I am skeptical of egalitarian mental state- and objective meaning-based accounts of wrongness. Some of my objections to Alexander’s and Hellman’s accounts apply to these accounts in general.
One caveat: whether Alexander’s or Hellman’s accounts are correct bears indirectly only on how discrimination should be regulated legally. First, there is no direct inference from the immorality of certain actions to what legal status they, morally speaking, ought to have and, second, while their accounts have implications for which kinds of differential treatment constitute wrongful discrimination, their primary focus lies on what makes differential treatment wrongful discrimination. In many contexts, the fact that an action is morally wrong is one, albeit by no means a decisive, reason for making it illegal. However, nothing below hinges on this assumption.
Discrimination Based on Beliefs About Inferior Moral Worth
Mental-state accounts contend that discrimination is wrong, when it is, because it reflects a certain morally objectionable mental state on behalf of the discriminator. The idea that mental states determine the wrongness of an action is well known outside the context of discrimination. For instance, the doctrine of double effect holds that actions involving some bad effect may be permissible when this effect is merely foreseen but not when it is intended. The doctrine is often invoked to explain why terrorism differs morally from conventional warfare involving a similar number of innocent casualties (McMahan 2009: 346). More generally, Anderson and Pildes have defended the view that “what makes an action morally right depends on whether it expresses the appropriate valuations of … persons” where “‘;Expression’ refers to the ways that an action or a statement (or any other vehicle of expression) manifests a state of mind” (Anderson and Pildes 2000: 1504, 1506).
There is a wide range of objectionable mental states which an account of the wrongness of discrimination may invoke. For instance, mental-state accounts may ground the wrongness of discrimination in noncognitive states such as hostility or in cognitive states such as the belief that members of a certain group are inferior. More generally, mental-state accounts vary in terms of which propositional attitudes—e.g., believing, expecting, hoping, doubting—and in terms of which contents—i.e., object, say, “I will win the lottery”—believed, expected, hoped for or doubted, etc.—they contend are the loci of the wrongness of discrimination. Also, they may vary in terms of how they flesh out the relation of reflection that obtains between the mental state and the relevant discriminatory action.
Not all mental-state accounts are concerned with equality. For instance, Richard Arneson proposes that: “Discrimination that is intrinsically morally wrong occurs when an agent treats a person identified as being of a certain type differently than she otherwise would have done because of unwarranted animus or prejudice against persons of that type” (Arneson 2006: 779). A person who treats Copts differently than Muslims because of unwarranted animus or prejudice against Copts responds differently to members of those two groups. But she might have no mental states to the effect that Copts have lesser worth than Muslims, e.g., because she is an unreflective person who does not bother to rationalize her aversions. Similarly, Matt Cavanagh (2002: 166) believes that discrimination which involves treating people “with unwarranted contempt” is wrong. Being contemptuous need not clash with equality of moral status. I can feel contempt for someone, say, whom I consider a superior person, on account of how she has squandered her talents. Similarly, I may think of an animal as having a lower moral status and yet harbor no contempt for it. With this broader picture as background, I turn to Alexander’s account.
In an article which deservedly has become a standard reference in the last two decades’ writings on discrimination, Alexander observes that “discrimination is not one thing, but many” (Alexander 1992: 153). Most forms of discrimination are wrong, when they are, for contingent consequentialist reasons. However, in his view one species of discrimination is intrinsically wrong for a different and Kantian reason, namely discrimination “premised on the belief that some types of people are morally worthier than others” (Alexander 1992: 161), where for some to have greater moral worth than others is for them to merit “greater moral concern” than others (Alexander 1992: 160). Presumably, by “intrinsically wrong” Alexander does not rule out the possibility that, all things considered, some such actions may be morally permissible, e.g., because they minimize the overall number of such intrinsically wrongful acts.
One attractive feature of this suggestion is that some paradigm forms of racist discrimination have involved this belief, e.g., the belief held by the Nazis that Jews are sub-human. It is unclear, however, that all paradigm forms of discrimination involve beliefs about unequal moral status. Paternalistic sexism holds that men and women have different functions, but in principle this is compatible with the view that they have equal moral worth. Similarly, discrimination against disabled persons often involves various forms of condescending treatment, but those who engage in it hardly believe that disabled people merit less moral concern than the able-bodied.
Alexander suggests three reasons why this kind of discrimination is intrinsically morally wrong. First, he submits that the underlying biases “are intrinsically morally wrong because they reflect incorrect moral judgments” (Alexander 1992: 161). However, it is not in general true that, because one acts on the basis of false beliefs, one’s action is intrinsically morally wrong. If I hold the false but justified belief that people like myself are less inclined to act altruistically, and treat people who are different from me better than myself for this reason, my act is ill informed but it is not intrinsically morally wrong.
Second, commenting on Nazi biases, Alexander writes: “Their biases were intrinsically morally wrong because Jews are clearly not of lesser moral worth than Aryans” (Alexander 1992: 158–59). This suggests an explanation that is more specific than the previous one. Discrimination is not intrinsically morally wrong simply because it involves a bias based on a false belief, but because it involves a bias that is based on a belief that is clearly (presumably, even to the discriminator) false. But again: acting on the basis of beliefs that are clearly false is not wrong in general. Suppose I believe that I deserve less than others. Suppose, moreover, that this belief is clearly false and yet I act on it. Setting aside controversial moral duties to self, here my act is based on a clearly false belief without being morally wrong. One might suspect that there is some aspect of the mind of the mental states of the Nazi bias holder that explains why she fails to grasp the easily accessible truth of equal moral worth, but then it is this mental state that renders her discriminatory act morally wrong, not her failing to believe a clearly true proposition.