Why Luck Egalitarianism Fails in Condemning Oppression

Luck egalitarianism has been criticized for (1) condoning some cases of oppression and (2) condemning others for the wrong reason — namely, that the victims were not responsible for their oppression. Oppression is unjust, however, the criticism says, regardless of whether victims are responsible for it, simply because it is contrary to the equal moral standing of persons. I argue that four luck egalitarian responses to this critique are inadequate. Two address only the first part of the objection and do so in a way that risks making luck egalitarianism inconsistent. A third severely dilutes the luck egalitarian doctrine. A fourth manages to denounce some instances of oppression for the right reason, but at the same time permits other instances of oppression and condemns yet others for the wrong reason. CYNTHIA A. STARK is an associate professor at the University of Utah. Her published works are on pornography, gaslighting, the gendered division of labor, disability, impartiality, social contract theory, egalitarianism, Rawls, and self-respect. They can be found in The Journal of Philosophy , Noûs , Journal of Political Philosophy , Social Theory and Practice , The Monist , Hypatia, and various other journals.

1 I owe thanks to members of the audience at the 2017 Feminist Ethics and Social Theory conference and at the Society for Analytic Feminism session at the 2017 Pacific APA conference. I am especially indebted to Kathryn Joyce for her copious feedback on both the SAF conference version and the penultimate version of this paper. I am also grateful for the excellent feedback I received from two anonymous referees for this journal.
(2011), Shlomi Segall (2007Segall ( , 2010, and Kok-Chor Tan (2008Tan ( , 2012 to combat the oppression objection. Finally, I explain the flaws in each of these approaches: the first two establish that oppression is (merely) conditionally unjust, which, I argue, threatens to render the approaches inconsistent; the third severely dilutes the luck egalitarian doctrine; and the last condones certain obvious cases of oppression and, further, condemns some cases as (merely) conditionally unjust.

Two Types of Egalitarianism
Egalitarians theories of justice generally fall into two categories. The first contains what are called "relational" or "democratic equality" accounts. This category is fairly eclectic. It includes Rawlsianism and other views that stress the importance of egalitarian social relationships. These views tend to foreground such values as equal citizenship, nonhierarchical social arrangements, and reciprocity (Rawls 1971;Anderson 1999;Scheffler 2003;Schemmel 2012a). The second category contains luck egalitarian approaches (Dworkin 1981;Arneson 1989;Cohen 1989). According to these views, differences in people's distributive shares are just if they are traceable to choice and unjust if they are traceable to brute luck. Luck egalitarians distinguish between "option luck" and "brute luck." According to Dworkin, who introduced the distinction, "Option luck is a matter of how deliberate and calculated gambles turn out-whether someone gains or loses through accepting an isolated risk he or she should have anticipated and might have declined. Brute luck is a matter of how risks fall out that are not in that sense deliberate gambles" (Dworkin 1981, 293). 2 The inequalities that luck egalitarians judge unjust are those caused only by brute luck. Inequalities arising from option luck, like those arising from choice, are just. Henceforth, "luck" is to be understood as "brute luck" unless otherwise noted.
Luck egalitarians invoke the distinction between luck and choice because they believe that people should be held responsible for their choices by paying the costs of those choices and that they should not be held responsible for their luck. Therefore, it is unjust for the state to aid persons who are comparatively badly off due to their choices, because in doing so, the state forces some people to subsidize the choices of others in which case those people are forced to take responsibility for the choices of others. It is just, however, for the state to aid persons who are comparatively badly off due to their luck because, in this case, their deficits occur through no fault of their own.

The Oppression Objection
Parties to the debate about luck egalitarianism and oppression rely on Iris Marion Young's (1990) account of oppression. She argues that oppression is, by definition, a structural phenomenon and that it has five faces: exploitation, marginalization, cultural domination, powerlessness, and violence. Proponents of the oppression objection state that these phenomena are unjust on the ground that they violate the equal moral standing of persons. They claim that luck egalitarians fail to recognize that equal moral standing requires that citizens relate to one another as social equals, which precludes any sort of oppressive social relations (Anderson 1999;Scheffler 2003). 3 This failure is rooted in the luck egalitarian aim of eliminating inequalities produced by luck but not those produced by choice. This aim causes luck egalitarians to make two mistakes. First, they end up condoning oppression when it is created by choice. For instance, they condone gender hierarchy created by women's informed choices to do unpaid care work. Second, when they do end up condemning oppression, they do so on the ground that it is caused by luck. For example, they condemn gender hierarchy created by discrimination because being subject to discrimination on the basis of one's gender is a matter of luck.

Strategies for Combatting the Oppression Objection
Though they rebut the oppression objection in different ways, all the theorists listed above deny the objection's claim that luck egalitarianism ignores the implications of universal moral equality. They contend that luck egalitarians simply see those implications differently. Where the relational egalitarian sees universal moral equality as grounding a demand for nonhierarchical and nonoppressive social arrangements, the luck egalitarian sees universal moral equality as grounding the idea that inequalities produced by choice are just, while inequalities produced by luck are unjust. 4 For luck egalitarians, this luck/choice principle, which in turn grounds more specific distributive principles, 5 respects universal moral equality by holding moral agents responsible for their choices but not their luck (Tan 2012). As we will see below, their explicit endorsement of universal moral equality presents difficulties for the defenses presented by Barry, Navin, and (to some extent) Tan. Barry (2006) argues that if luck egalitarians adopt objective well-being as the metric of justice-as the good that should be distributed in accordance with the luck/choice principle-they can condemn virtually all types of oppression. Barry's two-part argument is as follows. The first part runs as follows: oppression inevitably reduces people's objective well-being. People are oppressed on the basis of their group affinities. In most cases, people do not choose their group affinities. Therefore, in these cases, people's oppression-caused reduced well-being is unjust on luck egalitarian grounds. The second part runs as follows: included in a person's well-being is the right to choose certain things that are central to their identity-for example, a religion-without being oppressed as a result. Hence if a person chooses to affiliate with an oppressed group, we should regard them as choosing their affiliation but not their oppression. Thus, the oppression-caused diminished wellbeing they experience is luck egalitarianly unjust. Hence, oppression, whether caused by a chosen or an unchosen group affiliation, is always unjust because it is always unchosen.

Oppression as Unchosen
Why must luck egalitarians adopt objective well-being as the currency of justice if they want to denounce oppression? Consider a comparison with the resourcist view advanced by Dworkin (1981). This view denounces inequalities in resources provided they are caused by luck and not choice. Suppose a person chooses to inhabit an identity that causes him to be oppressed. Let us say that he converts to Judaism and is then subjected to anti-Semitic exclusion and hostility in his community. The resourcist cannot register this inequality because it is not an inequality in resources. Our convert does not have fewer resources due to his religion. He has less well-being. The idea is that when the metric of justice is wellbeing, the luck egalitarian can easily register the inequalities instituted by oppressive practices, such as anti-Semitism, because these inevitably impact people's well-being.
Barry aims to show that luck egalitarians are "fundamentally" concerned with inequalities caused by, for example, gender, race, ethnicity, religion, and sexual orientation, and that "addressing social oppression becomes a very important goal of luck egalitarianism" once well-being is adopted as the currency of justice (Barry 2006, 93). The problem with his approach is that it evades the heart of the oppression objection. According to that objection, luck egalitarianism is flawed because it permits, not only in fact but in principle, certain oppressive relationships (namely, those created by choice). It is, therefore, in principle, incompatible with the equal moral worth of persons. Barry's response is to argue that (when outfitted with the proper currency of justice) luck egalitarianism, in fact, must prohibit all oppressive relationships and is therefore, in fact, compatible with the equal moral worth of persons.
The point is that where the proponent of the oppression objection sees oppression as unconditionally unjust, the luck egalitarian sees it as conditionally unjust. The proponent of the oppression objection says that oppression in itself violates the moral equality of persons and so is in itself unjust. She opposes the fact that, for the luck egalitarian, oppression is unjust on the condition that is unchosen. Hence, Barry's argument that this condition is always met does not address the whole of the objection.
Barry appears to be aware of this problem, for he observes that advocates of the oppression objection see "equality as about treating people as moral equals, independent of questions of well-being" (Barry 2006, 96). For them, he notes, oppression would be wrong even if it had no effect on the well-being of those involved. He seems to think that this is a flaw in the oppression objection, although he does not argue for this idea. He concedes that oppression is wrong independent of people's subjective welfare and suggests it is a strength of his view that it is sensitive to the effects of oppression on people's objective welfare. So, in the end, Barry does not defeat the oppression objection; he simply takes a stand against oppression.

Oppression as Something for Which People Are Not Responsible
Navin (2011) responds to the oppression objection by arguing that people who become members of oppressed groups through their choices are nevertheless typically not responsible for their oppression. This is not because, as Barry claims, they have chosen the group but not the oppression. Navin believes, contra Barry, that it is reasonable to think that persons who choose to join an oppressed group have indeed chosen to be oppressed, provided they had, at the time, adequate options and were capable of anticipating the consequences of their actions. There are two reasons, he says, that people are generally not responsible for their oppression when it is caused by their choices. The first is that people who have not experienced oppression, even when they understand the likely outcomes of their choice to join an oppressed group, are not capable of anticipating the way in which those outcomes will restrict their later options. The second is that even if a person may be responsible for becoming oppressed, she may not be responsible for remaining oppressed because the person who made the initial choice may not be the same person (in the sense relevant for responsibility) who is now suffering due to the initial choice.
To illustrate the first type of case, Navin gives the example of a man who chooses not to study and fails out of university. He has adequate options and knows that the result will be lower-paying and less-inspiring work. So, he is, in essence, choosing the oppression experienced by members of the working class. Navin claims that this person, however, cannot be expected to anticipate how the workplace powerlessness and marginalization that accompany unskilled labor will limit his future opportunities in life. He cannot be expected to see, that is, how joining the working class will constrain his life prospects, and so he is not responsible for his oppression.
To illustrate the second type of case, Navin provides the example of a 60year-old woman who suffers from job-market marginalization because, at age 19, she chose to sell illegal drugs and was convicted of a felony. He stipulates that, unlike the unmotivated university student, this woman was capable in her youth of comprehending the lifelong constraints that would be imposed on her as a convicted felon. He concludes that, nevertheless, she is not currently responsible for the choice she made as a teenager because she now has different beliefs, desires, and character traits than did the 19-year-old version of herself. She is, therefore, not responsible for her current oppression. 6 Whatever its merits, Navin's response, like Barry's, does not defeat the oppression objection, which says that oppression is unjust categorically, simply on grounds of the moral equality of persons. Like Barry, Navin holds that oppression is unjust conditionally. It is unjust only when people are not responsible for it. As it turns out, Navin says, they are virtually never responsible for it: those born into it are obviously not responsible, and those who choose it are typically not responsible either, for the reasons he provides. Navin, like Barry, argues that the luck egalitarian condition on which the injustice of oppression depends virtually always holds and that therefore luck egalitarianism virtually always prohibits oppression. 7 But the concern of the oppression objection is not merely that luck egalitarianism cannot prohibit all instances of oppression but that it does not prohibit oppression for the right reason. 6 Sher (2014) discusses cases of this sort. 7 Navin's (2011) approach brings out a problem with the luck egalitarian's approach to oppression that I do not discuss: some luck egalitarians seem to be committed in at least some cases to compensating people for their oppression rather than eradicating the oppression itself. For instance, suppose a working-class person's nonresponsibility for their oppression gives them a claim upon the state for extra resources or opportunities. This does not necessarily end their oppression, nor does it eradicate working class oppression in general.
Barry and Navin might respond by claiming that my expectation of a definitive refutation of the oppression objection is unreasonable. In fact, it is question-begging, they might say, to expect them to show that oppression is wrong according to luck egalitarianism on the same ground that it is wrong according to relational egalitarianism. As long as they can show that luck egalitarianism virtually always prohibits oppression, they have established its egalitarian credentials. To say that luck egalitarianism is defective unless it condemns oppression for the same reason that relational egalitarians do, they might argue, would be akin to saying that utilitarianism is defective unless it condemns slavery for the same reason that deontologists do. But that is simply to say that utilitarians should be deontologists.
This response does not pass muster. Notice that the analogy with the dispute between utilitarians and deontologists is inapt. Those theories have very different accounts of value: the first endorses the intrinsic worth of happiness, whereas the second endorses the intrinsic worth of persons. Many of their differences are traceable to that fundamental difference. However, both relational egalitarians and the luck egalitarians under discussion explicitly found their views on the equal intrinsic worth of persons. Hence in order to reject the claim that what makes oppression unjust is that it violates the equal-worth ideal, these luck egalitarians cannot deny the ideal nor view it as defeasible. Instead they must establish that transgressing the ideal through political arrangements is not an injustice, or they must show that oppression does not, in itself, transgress the ideal.
I assume the first claim is plainly indefensible. But the second is also unassailable. This is because social hierarchies have historically been justified by the idea of differences in the intrinsic worth of persons. Furthermore, even if these hierarchies are no longer so justified explicitly, they are maintained by structures that emerged under ideologies of natural group hierarchy and which permit some people to treat others as if they have less intrinsic worth. The luck egalitarians under discussion, moreover, cannot deny the existence of such structures, because they accept Young's (1990) account of oppression as a structural phenomenon. 8 It follows that they must concede that oppression as such violates universal moral equality.
In order to make their case, then, Barry and Navin must, first, establish the theoretical advantage of founding the injustice of oppression upon a principle derived from universal moral equality-namely, the luck/choice principle-rather than upon the ideal itself, even though oppression clearly violates the ideal directly. Second, they must demonstrate that their view is not internally inconsistent. Given that they ground their theory in universal moral equality, they are committed, for the reasons I stated above, to the categorical injustice of oppression. Yet, in grounding the injustice of oppression in the luck/choice principle, they are committed to the conditional injustice of oppression. It follows that Barry and Navin appear to endorse simultaneously the idea that oppression is unjust regardless of whether we are responsible for it and the idea that it is unjust only when we are not responsible for it.

Value-Pluralist Luck Egalitarianism
Segall's (2010) approach is distinct from that of Barry and Navin. He maintains that luck egalitarians should be value pluralists (Temkin 2003;Lippert-Rasmussen 2015). On his view, distributive justice is one aspect of justice, and justice is one aspect of morality. Distributive justice (or "fairness," in Segall's terms) is the comparative aspect of justice: it concerns people's shares of goods relative to one another. Because the luck/choice principle tells us which differences in shares among persons are just, it is strictly an ideal of fairness. The design of social institutions, Segall believes, should be responsive to considerations of morality, broadly speaking, and not merely to considerations of justice, much less merely to considerations of fairness. Relevant values outside of justice include such things as utility, self-respect, autonomy, compassion, and so on. It follows that the state may be allowed-or indeed, required-to assist those who are worse off than others due to their own choices because doing so may be demanded by some other moral value. A virtue of this approach, Segall explains, is that it enables luck egalitarianism to avoid the "abandonment criticism," which accuses luck egalitarians of leaving to terrible fates those whose choices have bad outcomes. 9 According to this criticism, luck egalitarianism demands, for example, withholding health care from lung cancer patients who chose to smoke or perhaps from those who are injured while skiing. 10 Segall's value pluralism relies on a specific reading of the luck/choice principle, in particular of the choice half of that principle (the "choice principle"). The luck egalitarian views I discussed earlier interpret the choice principle as stating that it is unjust for the state to ameliorate inequalities produced by choice; justice requires that those inequalities be sustained. Call this the "strict view." Segall holds what I will call the "lax view," which says that justice allows but does not require that inequalities produced by choice be sustained. Hence it is not contrary to justice to mitigate choice-generated inequalities.
The lax version of the choice principle is designed to provide flexibility when making value trade-offs, which are central to the value-pluralist approach. Segall identifies two types of trade-offs that might be required. The first arises when the choice principle is "indeterminate." There may be times, Segall says, when the principle does not issue a verdict as to whether or not an inequality produced by choice should be ameliorated. So, we must appeal to some other value to determine the moral acceptability of that inequality. The second type of trade-off arises when the choice principle recommends sustaining an equality produced by choice. In these cases, another value recommending that the inequality be eradicated might outweigh the verdict of the choice principle. In these types of cases, fairness requires sustaining an inequality, but fairness, it turns out, must be sacrificed to some other value. According to Segall, only the "less demanding" first type of tradeoff is required to avoid the abandonment objection because, in cases where that objection might be leveled, the choice principle is in fact indeterminate.
Segall's approach can arguably be adapted to address the oppression objection. 11 Indeed, the allegation that luck egalitarianism cannot oppose oppression is plausibly understood as a type of abandonment objection (or at least, like the abandonment objection, as a type of harsh-treatment objection). Luck egalitarianism, one might say, abandons those whose choices regarding group membership cause them to suffer marginalization, exploitation, powerlessness, cultural imperialism, or violence. It abandons, for instance, the privileged person whose choices lead him to drug addiction and homelessness and therefore to severe marginalization. In response, a proponent of Segall's view can say that, for example, a moral duty of respect for persons requires that the inequality in this type of case be eradicated. Value pluralism, then, provides luck egalitarians the resources they need to counter the oppression objection: while the luck egalitarian principle itself does not condemn oppression, another principle does.
There are four problems with Segall's view, which together strongly suggest that it is not viable. The first is that the indeterminacy of the choice principle makes it problematically weak. The second is that the demand to revert from the choice principle to an alternative value to counter the abandonment objection is invoked in an ad hoc manner. The third is that Segall assumes without warrant that the alternative value we must revert to will inevitably recommend assisting, and not abandoning, those with severe disadvantages resulting from their choices. The fourth problem is that if the choice half of the luck/choice principle is indeterminate, it stands to reason that the luck half is also indeterminate. So, to the extent that indeterminacy weakens the choice principle, it also weakens the luck principle.
Consider the first problem. According to Segall, for the (lax) choice principle to be indeterminate (when it is indeterminate) is for it to be silent "for one reason or another" on the issue of whether or not a choice-generated inequality should be offset (Segall 2010, 26). It is indeterminate, in other words, when it implies, for some reason, that rectifying such an inequality is neither fair nor unfair. The lax choice principle, though, is not only sometimes indeterminate. It is always indeterminate because what makes the lax principle lax is that it is indeterminate. The lax version of the choice principle asserts that not offsetting inequalities produced by choice is not unfair. It thereby refrains from stating whether or not offsetting a particular inequality produced by choice is fair. Therefore, it is, as such, indeterminate with respect to whether an inequality produced by choice should be offset-it is always silent on this issue. 12 It follows that the second type of trade-off identified by Segall is impossible: if the lax principle as such never delivers a verdict on the matter of whether to ameliorate an inequality produced by choice, then it never delivers a verdict requiring that such inequalities be sustained. Hence we are never required to sacrifice fairness for the sake of another value. Indeed, we have no option but to appeal to another value to determine whether or not to rectify any given choicegenerated inequality.
This might seem like a strength of the lax choice principle, for the more demanding trade-offs are more contested, and they force us to sacrifice something of great value-namely, fairness. Nevertheless, the lax choice principle is feeble. It gives us little guidance regarding inequalities traceable to choice, despite the fact that this is its purpose. It tells us merely that those inequalities may be sustained, but it does not tell us in any case whether they should be. On Segall's approach, then, all cases of inequalities produced by choice must be subjected to some principle other than the choice principle in order to determine whether they should be sustained. Hence, the luck/choice principle itself gives us limited guidance on issues of distributive justice on Segall's interpretation of it. Now consider the second issue-the problem of ad hocism. Let us assume that I am wrong and that the lax choice principle is only sometimes indeterminate, as Segall claims. This would mean that, in some instances, it demands maintaining a choice-produced inequality, and in other instances, it is silent. If this is the case, we need to know how to tell in which instances the choice principle is silent and hence amenable to a less-demanding trade-off. This is where the threat of ad hocism emerges. Segall's outline of value pluralism suggests that it is precisely when its verdict is counterintuitive that the choice principle is indeterminate: "The abandonment objection" he says, "points to the sort of case, it seems to me, where luck egalitarian distributive justice is indeterminate" (Segall 2010, 65). As it stands, this is an ad hoc rescue of the luck egalitarian ideal.
To be fair, Segall may believe that there are independent criteria of indeterminacy and that it so happens that those criteria imply that virtually all those who would be abandoned by the choice principle should be assisted. For example, he might hold that the choice principle is indeterminate whenever it is hard to establish whether or not a person's disadvantages are the result of brute luck or option luck. 13 But Segall should identify these independent criteria, or at least suggest how they might be constructed, so that the critic can see how the appeal to value pluralism legitimately shores up the luck egalitarian theory. Moreover, constructing these criteria is no small undertaking, for it is difficult to see how the criteria will avoid merely catering to the harsh-treatment objections that arguably prompted the appeal to value pluralism in the first place.
The third problem with Segall's view is that it fails to acknowledge that value pluralism cuts both ways. In fact, Segall seems to believe that the values we appeal to when the luck/choice principle is inert will inevitably demand assisting those who suffer excessive disadvantages produced by their choices. So, oppressed people will be rescued for sure. It is conceivable, though, that the value that steps in to fill the void created by the indeterminacy of the choice principle will seal the fate of the person whose choices turned out badly. Perhaps a strongly libertarian ideal of personal responsibility is applicable in the case of a person who is convicted of a felony as a young adult. This ideal would likely demand not compensating that person for the inequalities she suffers once released from prison. We cannot know how often value pluralism yields this type of result until we know more about the values at play and about the norms for weighing those values against one another. Segall overlooks this double-edged feature of his view because he invokes value pluralism specifically to avoid harsh-treatment criticisms; his aim is to show how luck egalitarians can support policies such as full medical treatment for lung cancer patients who chose to smoke.
The fourth problem with Segall's view concerns the provisional nature of the verdicts of the luck/choice principle under a regime of value pluralism. Absent an argument to the contrary, consistency would seem to demand that if a value can outweigh the verdict of the choice half of the luck/choice principle when that principle is determinate (assuming that it is sometimes determinate), then a value can outweigh the verdict of the luck half (which is always determinate on Segall's view). That is, if value pluralism allows fairness to be overruled in some cases where inequalities have been produced by choice, it must allow fairness to be overruled in some cases where inequalities have been produced by luck. Hence, value pluralism implies that perhaps some inequalities caused by luck should be sustained. Allowing that other values can trump the verdicts of both halves of the luck/choice principle significantly weakens the principle in the sense that its prescriptions are always and inevitably provisional.
The provisional nature of the luck/choice principle under value pluralism is a problem because it makes the structure of the view hard to distinguish from some relational egalitarian positions, such as Rawls's. Luck egalitarians claim that Rawls also deploys the moral distinction between luck and choice (Tan 2012;Navin 2011). Specifically, in justifying the difference principle, 14 Rawls (1971, 73) maintains that inequalities produced by one's initial social position or by one's place in the distribution of natural talents are "arbitrary from a moral point of view" and therefore unjust. The difference between their view and Rawls's, luck egalitarians say, is that in Rawls's theory, the luck/choice distinction is not foundational, while it is in luck egalitarianism. However, when luck egalitarianism is embedded in the doctrine of value pluralism, the luck/choice distinction's foundational status is compromised. This is because it is possible that the luck and the choice principles might be outweighed simultaneously, resulting in a verdict that is completely untethered from the luck/choice principle. For example, the value of utility might demand that the last available ICU bed be given to the patient whose illness was caused by choice and not to the patient whose illness was caused by luck.

Luck and Choice in Various Domains
Like Segall, Tan (2008, 2012 believes that the moral terrain has a "plural character," and like Segall, he believes that the luck/choice principle is therefore limited in its applicability. 15 So, like Segall, he believes that luck egalitarians can, in many cases, condemn oppression regardless of whether it is caused by luck or choice. Two limits on the luck/choice principle are relevant to the issue of oppression. 16 First, the luck/choice principle, on Tan's account, is an ideal of distributive justice only. It is a comparative principle that regulates the distribution of economic goods such as resources and opportunities. It does not apply within the domain of political justice, which determines citizens' basic civil and political rights and liberties. It follows that, when oppression manifests itself as civil or political inequality, as it frequently does, it is not governed by the luck/choice principle.
Consider the recently revoked prohibition on same-sex marriage in the United States. The luck choice/principle would not apply here, on Tan's view, because the deficit experienced by gay people was a deficit in civil rights. So, whether preferring a same-sex marriage partner is a matter of luck or choice is irrelevant to the justice of eliminating this deficit: a principle of equal civil rights, which is grounded in universal moral equality, demands ending the state's oppression of gay Americans by legalizing same-sex marriage.
The second limit is this: the luck/choice principle, for Tan, applies only above a certain threshold of basic need. If that threshold is not met, the relevant principles are duties of humanity, which ignore the distinction between luck and choice and are grounded in the moral equality of persons. The plight of persons who are economically very badly off in the absolute sense, then, is not determined by a luck egalitarian distributive principle. 17 So, when people are economically very badly off due to oppression, they are owed aid regardless of whether or not their situation arose from their choices. The starving tenant farmer, then, is owed humanitarian assistance regardless of whether his crop failure is traceable to luck or choice.
On Tan's view, the luck egalitarian as such is not committed to a particular account of political justice or humanitarian aid. Therefore, he can rely on strongly egalitarian principles of political justice and/or stringent duties of humanity, which would enable him to condemn a variety of types of oppression on the ground that they violate the equal standing of persons. However, the luck/choice principle is not 16 A third limit Tan (2008Tan ( , 2012 places upon the luck/choice principle is that it applies only to the design of institutions. Institutions, on his view, must refrain from converting matters of luck into distributive inequality. Therefore, the luck/choice principle does not apply to people's personal pursuits. For example, it does not apply to inequalities produced by people giving one another gifts. Moreover, it does not apply to nature. For example, it does not apply to inequalities caused by people's differing degrees of attractiveness. Because oppression is a structural phenomenon, which virtually always involves social institutions, there is no oppression caused by personal pursuits alone or by nature. Therefore, Tan's distinction between the institutional and noninstitutional domains is not relevant to the issue of oppression. 17 This restriction is what allows Tan to avoid the abandonment objection. exempt, on Tan's view from evaluating inequalities associated with oppression. Remember, it applies whenever there are comparative deficits in economic goods that are not a matter of basic need, for that is when the question of fair distribution arises. The fact that it applies, and cannot be defeated, whenever this question arises, is what makes Tan's view a distinctively luck egalitarian view, in spite of the constraints he imposes upon the applicability of the luck/choice principle.
It follows that the luck/choice principle applies to inequalities in resources or opportunities that are caused by race, ethnicity, sex, gender, religion, sexuality, or any other factor that qualifies people for oppression. So, it applies to a large swath of inequality. According to Tan, the luck/choice principle condones inequalities caused by these factors when the factors are chosen, and it condemns inequalities caused by these factors when the factors are unchosen. So, Tan's view implies that inequalities in wealth or opportunity caused by, for example, race, sex, or ethnicity are unjust; and inequalities in those goods caused by, for example, adopted religion, social class, or gender status are just. It follows that Tan's view does not succeed in avoiding the oppression objection, for it both permits some types of oppression and condemns some types conditionally; and in condemning those types conditionally, Tan's view is vulnerable to the inconsistency worry that applies to the views of Barry and Navin.

Conclusion
Above I have challenged three strategies used by luck egalitarians to rebut the oppression objection, which says that luck egalitarianism not only condones some cases of oppression but also wrongly locates the injustice of (unjust) oppression in the fact that victims are not responsible for it. The first strategy, employed by Barry, argues for a specific metric of justice that enables luck egalitarianism to classify oppression based on both chosen and unchosen group affiliations as a matter of luck. On this approach, virtually all cases of oppression are unjust. The second, employed by Navin, drives a wedge between choice and responsibility: people should not be held responsible for choices that lead to oppression. On this approach, too, virtually all cases of oppression are unjust.
I maintained that neither of these arguments addresses the entirety of the oppression objection. Though they may establish that oppression is virtually always unjust, because people are never responsible for their oppression, these arguments fail to address the objection's claim that oppression is unjust because it violates the ideal of equal moral worth. I argued, further, that because proponents of these arguments in fact endorse the ideal of equal worth (as the ground of the luck/choice principle), their theories may be inconsistent. Oppression as such is clearly incompatible with equal moral worth, I argued, and so it seems that luck egalitarians, as proponents of that ideal, are committed to the categorical injustice of oppression. Yet at the same time, they argue that oppression is unjust only when (and because) people are not responsible for it, which implies that oppression is conditionally unjust, even if it is always unjust. But oppression cannot be both unjust categorically, as a violation of universal moral equality, and unjust only on the condition that we are not responsible for it.
The third strategy, used by Segall and Tan, is to restrict the purview of the luck/choice principle. Segall argues that the luck/choice principle can be defeated in some cases by other principles. Tan argues that the luck/choice principle applies only within a particular domain. This restriction strategy is subject to the following dilemma. If the luck/choice principle is limited so as to condemn all cases of oppression as categorically unjust, as it arguably is on Segall's view, the resulting theory is not distinctively luck egalitarian. If the luck/choice principle is limited so as to preserve the theory's distinctively luck egalitarian character, as it is on Tan's view, it condones some cases of oppression, and some of those it condemns, it condemns conditionally rather than categorically. So, it does not defeat the oppression objection.