Introduction

A common feminist and multiculturalist critique of liberal egalitarian theories of justice is that their narrowly distributive focus means that they ignore so-called status inequality or disadvantages due to misrecognition (Anderson 1999; Fraser 1997, 2003; Galeotti 2002; Taylor 1994; Young 1990). According to Nancy Fraser, for instance, status inequality or misrecognition is a question of injustice: ‘it is unjust that some individuals and groups are denied the status of full partners in social interaction simply as a consequence of institutionalized patters of cultural value in whose construction they have not equally participated and which disparage their distinctive characteristics or the distinctive characteristics assigned to them’ (Fraser 2003, p. 29). Moreover, writers more sympathetic to liberal egalitarianism, such as Andrew Mason, Roland Pierik and Ingrid Robeyns, also think that liberal egalitarians should pay more attention to the effects of social structure and socialization (Mason 2000; Pierik and Robeyns 2007; Robeyns 2003). In this paper, I will be concerned with whether this critique is applicable to one particular liberal egalitarian theory: Ronald Dworkin’s ‘equality of resources.’ Since Nancy Fraser is one of the foremost proponents of seeing recognition as a requirement of justice, I will focus my attention on her theory of recognition.Footnote 1

I start by clarifying what I will call the problem of recognition derived from Fraser’s position. I then turn to a brief summary of equality of resources followed by my main argument. The argument starts by conceding that equality of resources must be complemented with an account of which baseline rights and liberties a government must provide in order to treat everybody with equal concern. My response to the problem of recognition develops one of the principles guiding this baseline of rights beyond Dworkin’s own scarce remarks, i.e. the principle of independence, in four steps: first I provide an account of prejudice as comprising both negative or hostile attitudes and beliefs based on stereotypes. Second, I give an account of what renders prejudice unwarranted, wrongful and disadvantageous from a resource-egalitarian perspective: I claim that prejudices are wrongful if they are incompatible with people’s dignity as human beings, and I claim that discrimination based on wrongful prejudice is disadvantageous to an individual if he would have preferred his circumstances in the absence of prejudice. Third, I explore the right to protection from prejudicial discrimination that the principle of independence requires according to my interpretation. Finally, I ask whether the principle of independence as I have sketched it is adequate for responding to the problem of recognition, and I concede that this might not be so. But I argue that other aspects of the theory might be elaborated in order to address residual issues of misrecognition. I thus hope to show how equality of resources can be made reasonably sensitive to issues of misrecognition, and it is unclear that we need more sensitivity than that.

The Problem of Recognition

Nancy Fraser has claimed that distributive theories of justice are one-sided and inadequate because they are blind to issues of misrecognition and status inequalities. Fraser acknowledges that ‘many distributive theorists appreciate the importance of status over and above material well-being and seek to accommodate it in their accounts’. However, Fraser thinks their way of dealing with status is ‘not wholly satisfactory,’ the problem being their belief that ‘a just distribution of resources and rights is sufficient to preclude misrecognition’ (Fraser 2003, p. 34). According to Fraser, recognition-related problems cannot be reduced to an unequal distribution of rights and resources, as liberals have traditionally maintained (cf. Barry 2001). This is shown if we consider ‘an African-American Wall Street banker who cannot get a taxi to pick him up’ (Fraser 2003, p. 34). This individual has more than an equal share of economic resources and equal rights but remains in a disadvantaged position due to the informal patterns of values in society designating him as inferior. To take another example, suppose gays are treated with contempt by their colleagues if they do not conceal their sexuality. In these cases, some people are systematically treated differently and disadvantageously compared to others simply due to the category to which they are assigned by societal norms and representations.

Sune Lægaard has helpfully summarized Fraser’s view as comprising social theoretical and moral components (Lægaard 2005, pp. 330–331). What matters for Fraser as regards social theory is social status, which she understands as ‘an order of intersubjective subordination derived from institutionalized patterns of cultural value that constitute some members of society as less than full partners in interaction’ (Fraser 2003, p. 49). The status order thus becomes a product of institutionalized patterns of cultural value which in turn should be understood as more or less stable systems of evaluative norms and practices. Fraser takes the existence of a status order as motivating recognition by which she means establishing through a change in institutionalized norms ‘the subordinated party as a full partner in social life, able to interact with others as a peer’ (Fraser 2003, p. 30). The moral component of Fraser’s theory is thus a conception of justice which requires ‘participatory parity’ or the opportunity to ‘interact with each other as peers.’

But what is ‘participatory parity’ more precisely? Regarding the status-order, this standard is realized when ‘institutionalized patterns of cultural value express equal respect for all participants and ensure equal opportunity for achieving social esteem’ (Fraser 2003, p. 36). However, these are all very abstract concepts compatible with different conceptions. So how should they be understood? Fraser seems to leave this question open to be settled by actual democratic deliberation (Fraser 2003, pp. 42–45, pp. 70–72). I find this unsatisfactory, however, because I believe that political theorists can be expected to do more than merely point out abstract standards of justice; they can be expected to develop more concrete conceptions of these standards that can be evaluated in light of their implications for concrete cases and are ultimately candidates for practical guidance. Claiming this does not commit me to an anti-democratic posture in any way, as Fraser seems to think (ibid.). ‘Who should decide?’, ‘How should they decide?’ and ‘What should they decide?’ are in principle all separate questions, although the correct answers might be connected. At any rate, I will be concerned with the ‘what’ question, which is entirely compatible with my answers merely representing a contribution to a democratic dialogue, albeit expressing claims that I think others should accept.

By way of summary, I take Fraser’s view of recognition to entail a critique of equality of resources that comprises the following claims: (1) equality of resources is based on an (at best) incomplete social theory in ignoring the status or recognition order of society; and (2) equality of resources is consequently an (at best) incomplete theory of justice, because it overlooks impediments to people’s opportunities to participate as peers in social interaction, impediments that are unjust. I will refer to (1) and (2) as the problem of recognition. The remainder of this paper will discuss whether these claims are justified. I will attempt to show against (1) that Dworkin’s theory contains resources for acknowledging at least some recognition-type inequalities; and against (2) that the treatment of these inequalities by equality of resources is, if not all that is needed by way of justice, then at least an account of these problems that deserves serious consideration.

Equality of Resources

Equality of resources starts from the abstract egalitarian principle (cf. Dworkin 2002, p. 121) that government should treat all citizens with equal concern and asks what the implications of this requirement are for the distribution of impersonal resources. Impersonal resources are ‘those resources that can be reassigned from one person to another—his wealth and the other property he commands, and the opportunities provided him, under the reigning legal system, to use that property’ (Dworkin 2000, p. 323). The guiding principle behind the answer developed by the theory is that the distribution of impersonal resources should be both endowment-insensitive and ambition-sensitive (Dworkin 2000, p. 89). In other words, circumstances should be equalized in an ambition-sensitive manner (Williams 2004, pp. 143–145). But how should these claims be understood?

According to Dworkin, a person’s circumstances or endowments depend on her ambitions or conception of the good, i.e. her beliefs and judgments concerning what a successful life amounts to. Given a person’s ambitions, her circumstances are the features of her overall situation, including features of her body and mind, which provide means or impediments for realizing her goals (Dworkin 2000, p. 82). This means that we cannot specify whether a particular feature of a person, e.g. her height, is part of her circumstances without knowing her overall ambitions. Accordingly, equalizing people’s circumstances in any mechanical way is impossible. Rather, situating people equally requires accounting for her ambitions in the correct manner. To do so, we must ask whether the so-called envy test is satisfied, i.e. whether somebody prefers another person’s resource bundle. If so, the envy test is not satisfied. People have equal resources when nobody prefers others’ resource bundles.

The envy test renders the distribution of resources ambition-sensitive, since a person’s resource bundle reflects his ambitions. But there is also another way by which the distribution of resources should be ambition-sensitive and which is not captured by the envy test; this is due to the problem of dividing resources in a non-arbitrary manner. Every division of resources might favour the preferences of some while disadvantaging others. To see why, suppose that one form of resource, namely land, is initially divided into stadium-sized lots. The envy test might still be satisfied if nobody prefers another person’s bundle, but this division of resources would be less than fully ambition-sensitive, since it would be insensitive to the ambitions of those preferring a smaller tract of land. The division of resources should also be sensitive to differences in people’s ambitions. We therefore need a division rule giving everybody an equal impact on how resources are divided. The solution to this problem is to make use of an auction in which everybody has equal abstract purchasing power, securing an equal impact for everyone; where resources are auctioned in their most abstract form; and where people should be allowed to subdivide the auctioned lots until everybody is satisfied. People would then be able to tailor their resources to their ambitions to a much greater degree than if these subdivisions were not possible (Dworkin 2000, pp. 149–152).

The correct version of the envy test, then, is one applied in an auction in which people have equal initial abstract purchasing power and the opportunity to suggest subdivisions of auctioned lots until everybody is satisfied. This therefore results in a distribution which is optimally sensitive to people’s actual ambitions and preferences. Any other division rule would be arbitrary from a moral perspective, since it would favour the ambitions and preferences of some, effectively making them count for more than the ambitions and preferences of others. This second feature of an equal distribution counteracts what might be termed arbitrary divisions of resources.

Nobody will prefer the resources of others if the envy test is satisfied, although there may be considerable differences between what they actually end up with. This is, namely, the result of differences in their preferences, aspirations and beliefs about what makes a life successful, together with the choices flowing from these differences. Conversely, the envy test is not satisfied if people have different resource bundles as a result of physical disabilities or different talents. Consequently, some people envy others’ bundles. Dworkin therefore proposes compensating for these differences in circumstances according to a complex hypothetical insurance scheme. Briefly, the idea is that inequality in terms of initial circumstances defined as the existence of envy can be countered by establishing an insurance scheme which allows people to insure against the impact of their circumstances on their opportunities for realizing their ambitions, taking into account their own views on risk and chance and the value of particular options as they see it.Footnote 2

In summary, equality of resources requires that people’s circumstances are equalized in an ambition-sensitive manner. This requires (1) satisfying the envy test, (2) countering arbitrary divisions of resource lots, and (3) redistributing according to a hypothetical insurance scheme. Before proceeding, however, we should ask whether the recognition problem really applies to equality of resources as outlined in this section. In other words, does equality of resources ignore status inequalities? If so, is this a problem? Equality of resources clearly does not have anything particular to say about the various evaluative norms embodied in society. The theory is concerned with how we should take account of differences in ambitions, preferences and circumstances narrowly in the distribution of impersonal resources, and although these differences obviously have a connection to evaluative norms, the significance of these norms is not explicitly thematized. The question is whether this is a problem. Consider the African-American Wall Street banker who cannot hail a taxi. He is clearly subject to an injustice, and any theory of justice that cannot account for this seems inadequate. So what does equality of resources imply for his case?

The answer seems to be that unless the banker prefers the circumstances of his white peers, including the colour of their skin, he has no claim of justice according to equality of resources. But it seems profoundly insulting to require of the banker that he should prefer to have another skin colour to have such a claim (cf. Robeyns 2003, p. 541). In treating the injustice incorporated in the example in this manner, equality of resources fails to acknowledge the real source of inequality, which is racist-evaluative norms, i.e. those that pertain to the recognition dimension of society. On the face of it, then, the problem of recognition seems to represent a profound and insurmountable challenge to equality of resources. As I will argue, however, this appearance is deceptive.

Disadvantaged by Prejudice

In Dworkin’s original formulation (Dworkin 2000, p. 65), the scope of his theory is restricted to how privately owned resources should be distributed, but he has subsequently devoted attention to the wider institutional background necessary for an equal distribution. In ‘The Place of Liberty,’ Dworkin argues that the auction and envy tests already presume a certain ‘liberty/constraint system;’ i.e. a baseline of rights determining what people are entitled to do with their resources. The liberty/constraint system is crucial to how the market will determine prices. How this liberty/constraint system is to be designed cannot itself be determined by private decisions on a market, since a market already presupposes a particular baseline (Dworkin 2000, pp. 143–144). As originally stated, equality of resources is thus incomplete; we cannot determine how the liberty/constraint system should be designed by consulting the envy test and the insurance device. We should attend to the abstract egalitarian principle that requires a political community to treat all of its members with equal concern and ask which baseline of rights follows from this principle instead (Dworkin 2000, p. 147). Dworkin’s answer is, roughly, a baseline satisfying a range of principles which (1) deal with the effect of prejudice, (2) correct for market imperfections, (3) secure the opportunities to develop authentic beliefs and preferences, and (4) generally protect people’s liberty.

My response to the problem of recognition appeals (primarily) to one of the background principles, which, according to Dworkin, must be satisfied in order for a political community to treat its citizens with equal concern: the principle of independence (PI). Footnote 3 According to PI, no one should be worse off as a consequence of others’ prejudices. As seen above, physical disabilities and lack of talent possibly constitute a resource deficit and therefore something which can trigger resource-egalitarian concern. Analogously, prejudice can also place some at a disadvantage compared to others. For instance, if racists are sufficiently numerous in a given society and choose to exclude other ethnic groups from working and trading in a particular area, the victims of this treatment are worse off than others. These actions are therefore ruled out by PI.

So according to Dworkin (2000, p. 162), prejudice constitutes a ‘structurally related problem’ to physical disabilities. As Browne and Stears (2005) rightly point out, however, it is not entirely clear how this analogy is supposed to work. More specifically, we require more detailed answers to the following questions: (1) what constitutes a prejudice? (2) When is a person disadvantaged by such prejudice in a resource-egalitarian sense? And (3) what does PI imply regarding the effects of prejudice? In answering these questions, I will argue that prejudice in fact poses a problem closer to that of the stadium example presented above than to the problem of physical disabilities; the difference between being subject to lack of talent or having a disability and being subject to the detrimental effects of prejudice is that the latter can object to the prejudicial beliefs that disadvantage him, while the former two cannot.

What constitutes a prejudice? Dworkin (2000, p. 162) assimilates prejudice to negative attitudes of dislike and a wish to avoid, which suggests something akin to what Richard Arneson (2006) calls ‘unwarranted animus’ and ‘unwarranted prejudice’. At any rate, that is how I will understand the term ‘prejudice’ here. Animus towards an individual is a negative attitude towards him, while Arneson understands unwarranted prejudice against the individual as

beliefs about the person’s characteristics that are either inferred from one’s beliefs about persons of that type or directly caused by one’s reaction to the type, these beliefs being formed in some culpably defective way (Arneson 2006, p. 788).

If this interpretation is accepted as an answer to (1), then PI refers to two sorts of problematic attitudes: first, it refers to unwarranted negative attitudes such as dislike, hostility and contempt concerning particular types of individuals; second, it refers to beliefs concerning particular types of persons formed in an unwarrantedly faulty or defective way, or beliefs inferred from other beliefs so formed, i.e. unwarranted prejudice in the narrow sense. However, prejudice in the narrow sense warrants further discussion. Obviously, belief formation processes might be tainted by all sorts of epistemic errors, but we should take note of three important classes of epistemic failure that seem especially important. These types of failure are basing one’s beliefs concerning a particular person on statistical, simply false and/or undignified normative stereotypes (Appiah 2005, pp. 194–195). I base my belief about X on a statistical stereotype if I ascribe

a property to an individual on the basis of the belief that it is characteristic of some social group to which she belongs, where there is indeed a statistical correlation between that property and being a member of that group, but where, in fact, she does not have that property (Appiah 2005, p. 194).

Simply false stereotypes, on the other hand, simply represent false beliefs about a group. The mistake of basing one’s beliefs on simply false stereotypes can be exemplified by ‘the case of the strong woman who presents herself for a job as a firefighter and is told that she will not be considered because “women are not strong enough to be firemen”’(Appiah 2005, pp. 194–195). An example of the second kind of failure, i.e. basing one’s beliefs on statistical stereotypes, would be ‘ethnic stereotypes, which lead people, say, not to do business with members of a group because they are purportedly “shifty and dishonest” when, in fact, they are not, or not especially so’ (Appiah 2005, p. 195).

A third kind of stereotype is normative stereotypes:

In such cases, a stereotype is not a view about how members of the group behave simpliciter: it is a view about how they behave grounded in a social consensus about how they ought to behave in order to conform appropriately to the norms associated with membership in their group (Appiah 2005, p. 195).

There is a question here whether all normative stereotypes should be counted as prejudicial for our purposes. I want to leave open the possibility that normative stereotypes can be warranted, and I have therefore only referred to undignified normative stereotypes as prejudicial above. Giving an account of what undignified means, however, can best be pursued in the course of answering (2): When is prejudicial discrimination disadvantageous in a resource-egalitarian sense?

I will divide my answer into two parts, namely (a) an account of why and when prejudicial treatment of an individual is wrongful to the victim from the perspective of equality of resources; and (b) an account of how such wrongful discrimination can be seen as unfairly disadvantageous from a resource-egalitarian perspective. As to (a) I deem prejudicial treatment wrongful to the victim if it is contrary to their dignity as a human being; concerning (b) I say that a person is unfairly disadvantaged by a particular kind of wrongful treatment if he would have preferred the opportunities that would have been available to him in the counterfactual situation where no prejudiced treatment were present. Take the African-American Wall Street banker who is unable to get a taxi to pick him up, while his white peers have no such difficulty. I will suggest that he is unfairly disadvantaged from a resource-egalitarian perspective, then, because the prejudicial treatment violates his dignity, and he would have been better able to fulfil his preferences if the prejudicial discrimination had been absent.

To motivate this account, recall that we are concerned with which background of rights and opportunities will treat everybody with equal concern. To answer this question and its implications for prejudiced discrimination in particular, I now propose interpreting the abstract egalitarian principle in a particular way, namely as requiring recognition of two principles of human dignity (cf. Dworkin 2006, pp. 9–21). The justification is that a government cannot be said to treat every person with equal concern if it allows serious insults to their dignity. Obviously, it is worse if the government itself engages in undignified treatment, and it is primarily responsible for not doing so; but it should also be held accountable if it merely allows people to treat each other in undignified ways.

So what are the principles of human dignity I have referred to? First, there is the principle of intrinsic value. This principle states that every human life has a special kind of objective importance; it has value as potentiality, meaning that once it has begun, it matters how well it goes (Dworkin 2006, p. 9). Moreover, since there is no good reason to think that any particular life is objectively more important than anybody else’s, the principle requires us to assign equal objective importance to each human being’s life (Dworkin 2006, p. 16).

Second, there is the principle of personal responsibility. This principle says that each person has a special responsibility for realizing the success of his own life, including a responsibility for exercising judgment about what success is for him (Dworkin 2006, p. 10). The decisions regarding what is worthwhile and important for my life are decisions I must make myself:

We may not subordinate ourselves to the will of other human beings in making those decisions; we must not accept the right of anyone else to force us to conform to a view of success that but for that coercion we would not choose (Dworkin 2006, p. 17).

Furthermore, we have no reason to believe that only some kinds of people have this kind of special responsibility for the success of their own lives, i.e. as long as everybody has the capacities for basic rationality (Dworkin 2006, p. 18).

Now, my claim is that interpreting the abstract egalitarian principle in the light of these principles of human dignity allows us to explain when and why discrimination on the basis of prejudice and animus is wrongful. To see this, take first discrimination based on animus or dislike. Hostility and dislike are obviously sometimes warranted attitudes, e.g. if a person tries to kidnap one’s children. But if the negative attitudes are simply a matter of contempt, as in the case of the African-American banker, it seems clear that treatment based on these attitudes is wrongful discrimination. The banker could rightly object to this treatment, and he could do so with reference to the principle of intrinsic value. The dislike in question expresses the judgment that the African-American banker’s life is less important than the lives of others.Footnote 4

Next, consider discrimination based on statistical or simply false stereotypes. Both are by definition inaccurate, and I will suggest a similar judgment concerning their use. When a strong woman, for instance, is told that she cannot be considered for a job as a fire-fighter because ‘women are not strong enough to be fire-fighters’ she can complain of unequal treatment: if the employer had cared to check, he would have discovered the falsity of his views. Using such a statistical measure as a means of selecting among applicants disregards their equal claim to be assessed in a reasonably individual manner. Measures for selecting amongst candidates must themselves be compatible with the equal importance of them all.

Finally, consider discrimination based on undignified normative stereotypes. In exemplifying, I will concentrate on violations of the principle of personal responsibility. The principle of personal responsibility forbids subordination, i.e. it

forbids my accepting that other people have the right to dictate what I am to think about what makes a good life or to forbid me to act as I wish because they think my ethical values are wrong. It therefore forbids me to accept any manipulation of my culture that is both collective and deliberate – that deploys the collective power and treasure of the community as a whole and that aims to affect the ethical choices and values of its members. That is subordination. (Dworkin 2000, p. 76).

Deploying the powers of the state to enforce normative stereotypes expressing particular ethical values, i.e. personal judgments concerning what a successful life amounts to, would accordingly amount to subordination. However, I suggest that other, more informal, kinds of discrimination based on normative stereotypes similarly seem to subordinate the victims.

For an example, consider a case of non-governmental, normatively stereotyped discrimination, which intuitively seems problematic: Deborah is part of a group characterized by very patriarchal stereotypes of gender roles according to which women are primarily responsible for raising and caring for children, while she herself would rather pursue a professional career. Moreover, women who do not care for their children personally are sanctioned by other members of their culture, who treat such women in various obnoxious ways, e.g. by refusing to employ them and by shunning them in general. The kind of discrimination Deborah is subject to is based on normative stereotypes, i.e. beliefs about what women can or cannot do derived from social norms. In my view, Deborah is treated wrongfully on the grounds that her personal responsibility for her own life is not respected. The normative stereotypes coupled with harsh sanctions are collectively adopted by the majority of her cultural group in order to coerce Deborah and other women to conform; the stereotypes involve particular views concerning what a successful life must contain which are incompatible with particular types of persons making their own autonomous decisions concerning how to live their lives, and which are coercively imposed on them in order to force them to conform to a view of success that they would not have chosen but for that coercion (cf. Dworkin 2006, p. 17). Discrimination based on these stereotypes, therefore, is also a form of subordination, since it is a form of ‘manipulation’ of Deborah’s ‘culture that is both collective and deliberate’ and ‘that aims to affect the ethical choices and values of its members’ (Dworkin 2006, p. 76).

So my claim is that discrimination based on normative stereotypes can be wrongful from a resource-egalitarian perspective, e.g. when the stereotypical discrimination is contrary to the principle of personal responsibility because it amounts to collectively manipulating the culture in order to enforce conformity with the stereotypes. Normatively stereotyped discrimination of this kind entails subordination of particular persons and is incompatible with their equal dignity. The upshot is that both discriminatory political action and more informal discrimination based on normative stereotypes can violate the principle of personal responsibility and are thus incompatible with the equal dignity of all persons. These forms of discrimination are therefore wrongful.

If we now relate these points to the second condition of unfair prejudicial discrimination mentioned above, we can say of Deborah that (a) she is subjected to wrongful treatment when facing discrimination based on undignified normative stereotypes regarding the proper roles of women effected through the threat of social sanctions; and (b) she would have been better off seen from her own perspective if the discriminatory treatment had not been present, since she would then have been able to fulfil more of her ambitions. It is important to stress here that equality of resources provides an account of the inequality of Deborah and other victims of prejudiced discrimination that does not rely on a welfare metric. According to this account, we should ask not whether the banker has less welfare (or opportunity for welfare) than his white peers, but whether he would have had better opportunities seen from his own perspective given a different set of options, not tainted by prejudice. The relevant analogy here is the case in ‘Equality of Resources,’ where the divider of resources transforms all resources into old claret and plover’s eggs (Dworkin 2000, p. 67). In that case, the character who hates wine and plover’s eggs can complain that he has not been treated equally in the selection of the set of resources available. It is not that he prefers the bundle of any other person within the actual set. His complaint is that the selection of the set unfairly disfavours his preferences. Similarly for Deborah – she can complain that a similar set of options not influenced by prejudice would have been less unfavourable to her preferences. The comparison is to herself in a counterfactual situation. And since the prejudiced discrimination in question is incompatible with her dignity as an autonomous and responsible human being, it is problematic from a resource-egalitarian perspective.Footnote 5

These admittedly sketchy remarks concerning wrongful and disadvantageous, prejudiced discrimination leaves us with our third question, namely (3) what should we take PI to require regarding wrongful prejudice? I propose interpreting PI as a right to protection from the unfair disadvantageous effects of wrongful prejudice that the government has an obligation to uphold. Such a right, however, must be interpreted, and it has different consequences in different contexts. Since it is a right that individuals have against the government, it is clear that the implications of the right depend on what the government can reasonably do in a given context, but also on differences in character between different kinds of social relationships. I will distinguish between three different spheres of influence. First, we have a sphere consisting of the official laws, policies and institutions of the state itself. The PI applies directly in this sphere, meaning that laws, policies and institutions must not be based on prejudice. Next, we have the semi-public sphere of civil society, including the labour market. This sphere is indirectly within the scope of government influence. Finally, we have the private sphere, e.g. of family life. Here again, the government has indirect influence. In civil society and the private sphere, where government influence is only indirect, the implications of PI are more difficult to interpret than in the sphere directly amenable to government action. I will offer a few comments regarding how PI should be understood in the latter spheres.

Intuitively, there is a difference between how much the government should justifiably intervene in civil society and the private sphere. How can this difference be explained? I will suggest that it has something to do with whether a given practice is constitutively dependent on individuals having the right intentions or whether the practice is not so dependent. Consider friendship as an example of the first. Friends liking and respecting one another seems to be constitutive of friendship. People might wrongfully discriminate when choosing their friends, but it is not wrongful in a way that brings it within the reach of government intervention, because intervention would be self-defeating. The government cannot legitimately force a friendship between two people if they do not like each other. Note, however, that although government cannot directly intervene to prevent private discrimination by force, it might have a duty under PI to work to change people’s prejudicial preferences in other, ‘softer’ ways, e.g. via education and campaigns.

Consider next a company interacting on the market. It does not seem to be constitutive of such companies that the individuals employed in the company particularly like each other or do not harbour defectively formed beliefs about each other, as long as they suppress these attitudes to a necessary degree in their interaction. It does seem justified for government, then, to attempt to influence commercial interaction in a less discriminatory direction through the imposition of legal sanctions. In some cases, it might also require companies to engage in affirmative action. Private associations are found between those two cases. What government can legitimately do to influence such associations depends upon whether their purposes are mainly instrumental to other ends, as in the commercial case, or whether they are more like friends in being intrinsically dependent on mutual relations involving positive attitudes.

By way of summary, I have argued that the principle of independence should be interpreted as requiring a right to protection from unfair, disadvantageous and wrongful prejudicial discrimination. Wrongful prejudice can either consist of unwarranted animosity against particular types of persons or as unwarranted prejudice in a narrow sense as beliefs formed in an unwarrantedly defective way. In both cases, the beliefs or attitudes are wrongful as a result of being contrary to people’s dignity. A person is unfairly disadvantaged by prejudice when people harbouring wrongful prejudice against him treat him differently than others because of their prejudice, and he would prefer the counterfactual circumstances where he would not be treated in a prejudiced manner. The problem with prejudicial discrimination is, then, that it wrongfully structures the victim’s opportunities in a disadvantageous manner. The right to protection against the effects of prejudice claimed by PI must be interpreted differently in different circumstances, however, depending on the possible effects of government intervention. At a minimum, government would be required to enact non-discriminatory laws and policies; to legally sanction discrimination in parts of civil society such as the labour market; and to campaign for belief change in a less prejudiced direction.Footnote 6

Is Overcoming Prejudice Enough?

We now have a detailed picture of equality of resources and the answers this theory entails for prejudicial discrimination. I now turn to a discussion of whether these answers represent adequate responses to the problem of recognition. More precisely, I consider whether there are justice-relevant dimensions of the status order that are left out by the principle of independence, and I concede that this in fact seems to be so. However, other aspects of equality of resources might be elaborated in order to handle the residual problems of recognition left unaddressed by the principle of independence.

Recall that the problem of recognition was that equality of resources overlooks the status or recognition order of society; and that it consequently ignores unjust inequalities produced by this order. The status order consists of the institutionalized patterns of evaluative norms that assign differences in importance, worth and prestige to different types of actions and individuals. The status order subordinates particular categories of persons to others. So is it true that equality of resources, with its principle of independence, ignores this status order? PI states a right to protection from the unfair effects of wrongful, prejudiced discrimination, and wrongful prejudice in turn should be understood as referring to both unwarranted negative attitudes and various unwarranted stereotypes. These effects surely include the impact of prejudicial, institutionalized norms. PI thus clearly addresses some of the issues Fraser assigns to the status order of society, e.g. the case of the African-American banker.

The question then is whether parts of the status order cannot be reduced to prejudiced norms. That clearly seems to be so. For instance, there might be a norm in society that the profession of a medical doctor is more important and prestigious than the profession of a taxi driver. This norm does not seem to be based on prejudice, although it is hardly self-evidently justified. So clearly there are evaluative norms and ‘schemas of interpretation’ which affect the status order of society but which are not singled out as problematic by PI. However, the real question is whether this presents a problem for the theory, i.e. whether such cases constitute injustices which show the need for an alternative theory of justice, e.g. based on participatory parity. The latter would surely require a more detailed elaboration of this alternative standard. Be that as it may, if we abstract from the plausible claim that the case of the doctor and taxi driver might embody injustices related to the distribution of resources, this example does not seem to involve any serious injustice. Thus the bare fact that some social roles are evaluated as more important and prestigious than others does not seem to support the claim that this is unjust and that equality of resources is inadequate in failing to address this injustice, though perhaps other facts will.

Consider, for instance, the cases of

US court rulings holding that employers’ failure to provide pregnancy leave does not constitute sex discrimination because it does not deny women a benefit provided to men; firefighter job application procedures that test climbing speed on ladders designed for persons whose height falls in the normal range for men, thus disadvantaging many women; and regulations mandating uniform headgear for Canadian mounted police, effectively closing that occupation to observant Sikhs (Fraser 2008, p. 86).

Or consider, finally, a case where a person cannot get a particular job because the employer dislikes his appearance.Footnote 7 She only wants to employ (what she takes to be) physically attractive people in her business.

These cases all seem to be unjust, and according to Fraser the injustices consist in denials of participatory parity. But what does equality of resources have to say about them? They do not seem to directly violate PI or the resource egalitarian tenets more generally. In particular, resource egalitarians would not say that the very fact that a rule has disparate impact on different people makes that rule discriminatory or unjust per se.Footnote 8 Furthermore, the cases under examination do not seem to be grounded in beliefs or preferences incompatible with the two principles of human dignity, i.e. they are not unfairly discriminatory.

But whether these cases actually do represent injustices from a resource egalitarian perspective is actually a more complex question. There are, namely, further principles guiding the baseline of rights which might be relevant to the cases in question. Consider the stadium example mentioned above. In that case, people preferring smaller tracts of land are unfairly treated because the property regime arbitrarily disfavours their preferences. This arbitrariness is wrong from a resource egalitarian perspective because it violates the principle of abstraction (Dworkin 2000, pp. 150–151). Rules and procedures are arbitrary in the sense relevant to this principle if they do not allow maximum flexibility for each person’s ambitions. The principle as applied to the market requires that resources are available in their most abstract form, allowing people ‘the greatest possible flexibility in fine-tuning bids to plans and preferences’ because this is more ambition-sensitive (Dworkin 2000, p. 151); and I suggest that we generalize this requirement to all rules and evaluate them in light of how much flexibility they allow people in satisfying their different preferences and ambitions. Accordingly, I will understand the principle of abstraction to require government to maximize ambition-sensitivity in what it prescribes and proscribes.

Returning now to the cases posed by Fraser, the question is whether these are compatible with the principle of abstraction, or whether they arbitrarily restrict the opportunities available in an ambition-insensitive manner. That is a complex question, but I will offer a few brief comments. First, it seems that the Sikh example violates the principle of abstraction since the rule in question makes some ambitions impossible to satisfy whereas it is not required to realize the particular values that a rule for uniforms is intended to serve; another rule allowing turbans could realize these goals equally well.

Second, it seems that the fire-fighter case might similarly violate the principle of abstraction, but whether this is actually so depends on whether the procedure for testing applicants could be made more sensitive to women’s ambitions without sacrificing accuracy in identifying relevant qualifications. Clearly, if another procedure has the consequence that fewer people will be saved from fires, then that rule is less ambition-sensitive in some respects—less sensitive to the ambitions of the victims of fires, for instance. We must therefore consider the overall effects of the procedure on people’s abilities to satisfy their preferences and pursue their goals. The correct judgment concerning the fire-fighter case thus depends on further contextual facts.

Concerning the third case, I would suggest the following analysis. From a resource egalitarian perspective in ideal circumstances, the decision to provide pregnancy leave should be left to the employer. But in those circumstances, people who contemplated having children would probably demand to have the right to paid leave included in their employment contract. In these circumstances, they would, for instance, have access to reasonable unemployment benefits that would improve their bargaining position. But in non-ideal circumstances where the distribution of economic resources is very unequal, the decision to provide pregnancy leave might very well be taken away from the employers in order to reproduce circumstances that are as similar to the ideal situation as possible. I tentatively conclude, then, that all three cases that Fraser mentions can be handled by other dimensions of equality of resources. But what about the case of ugliness?

I assume here for the sake of the argument that this case does not in fact violate PI or the principle of abstraction. But if that is so, the reason might be that there is nothing actually wrong with such aesthetic norms. It is at least not obvious that such aesthetic norms are unjust. Moreover, it seems to me that many that are deemed ugly by the norms of beauty in question would themselves subscribe to these norms, and it seems patronizing to treat this as simply false consciousness. As to the inequality that these people face, equality of resources certainly has an answer: since their ugliness is clearly a disability in their own minds, they are entitled to compensation if they would have insured against it. Depending on the scale of the problem, this kind of reasoning might lead to publicly funded cosmetic surgery for particular kinds of ‘ugliness-related’ problems. So in the ideal resource egalitarian world, employers would perhaps be allowed to discriminate on the basis of aesthetic norms (as long as this in fact would be both compatible with people’s dignity and in accordance with the principle of abstraction); however, in the non-ideal world where there is no compensation to ugly people preferring other people’s circumstances, it seems plausible to say that the aesthetic discrimination should be prohibited by analogy with the maternity leave case discussed above to bring the situation closer to the ideal world.

Summarizing, I have argued, first, that PI goes some way toward addressing the problem of recognition in acknowledging many forms of inequalities caused by institutionalized evaluative norms. Second, I have conceded that some purportedly unjust examples of misrecognition are not addressed by PI. But I have argued, elaborating on another principle applying to the baseline of rights, that resource egalitarians might in fact object to these cases as entailing forms of arbitrary treatment not reducible to prejudiced discrimination—that is, insofar as these cases really manifest injustices, and my brief discussion suggests that this is actually a more complex question than Fraser seems to acknowledge. I conclude then that although overcoming prejudice might not be enough, equality of resources as a whole seems capable of responding to the problem of recognition.

Conclusion

The examples discussed above possibly illustrate how overcoming prejudice is insufficient to realize justice in a society marked by misrecognition. But they do not show that we need something beyond equality of resources. The principle of independence is part of a complex theory that deals with different kinds of unjust inequalities in different ways. Other parts of the theory can, for instance, handle the problem of ugliness, and it is at least not clearly wrong to treat this (and the other) problem(s) as equality of resources does. However, I have not argued in any comprehensive way for equality of resources in this paper, and I will conclude by outlining what I take the argument to have accomplished.

First, I have clarified what I take to be an important problem for liberal egalitarian theories in general and equality of resources in particular, namely the problem of recognition; second, I have shown how one might develop Dworkin’s brief remarks about prejudice and discrimination in more detail in what I think is a plausible way that coheres well with the basic principles of his theory; third, I have argued that the refined account of equality of resources does not seem to fall prey to the problem of recognition—rather equality of resources arguably goes a long way toward explaining when and why particular recognition issues are problematic.

Hence, I hope to have plausibly sketched how one coherently can come to hold both the view that equality of resources attractively articulates important dimensions of a theory of justice and the view that evaluative norms of status and prestige must figure in any such adequate theory of justice in non-ideal circumstances. Whether the resource-egalitarian answers proposed to the problem of recognition are ultimately convincing is a question warranting further consideration.