II. Two Versions of Luck Egalitarianism

Suppose, first, that we start with the notion of luck. When we say that some aspect of a person's situation is a matter of luck, what we generally mean is that it is not (and was not previously) within his control. Moreover, when we say that an aspect of a person's situation is or was within his control, we mean, at a minimum, that he made some choice that he knew would bring it about (or would significantly raise the likelihood of its coming about). Thus, whenever any aspect of a person's situation is not an expected consequence of any of his choices, it thereby qualifies as a matter of luck.

Because each choice is made by a single individual, whereas each inequality is a relation between the situations of different individuals, it is not immediately clear how we should interpret the claim that a given inequality is a matter of luck. It would not be plausible to say that an inequality is due to luck whenever the difference between the parties' situations was not a foreseen consequence of either party's choices; for, given that each party can typically affect only what he gets but not what the other party gets, this interpretation would imply that virtually all inequalities are due to luck.Footnote ² Thus, to give the luck-centric version of luck egalitarianism a fighting chance, we must interpret its conception of luck more narrowly. Instead of taking it to assert that an inequality is a matter of luck, and so is unjust, whenever the inequality itself was not a foreseen consequence of either party's choices, we must take it to assert only that an inequality is a matter of luck, and so is unjust, whenever each party's having what he currently does, as opposed to what the other has, was not a foreseen consequence of his previous choices. If we interpret luck egalitarianism in this way, it will imply that no inequality can be just unless having more than he now does was once within the less-advantaged party's control—unless, in particular, he was at some point aware of having some option which, had he taken it, would have resulted in his now having as much as his more-advantaged counterpart.Footnote ³

Bearing this interpretation in mind, let us now approach luck egalitarianism from the other direction. Instead of organizing our account of its central distinction around the notion of luck, let us try to organize it around the notion of responsibility. Because responsibility is a complex notion, any such approach is bound to raise many issues that lie beyond our scope. However, for present purposes, what matters is only that agents are often thought to be responsible for various outcomes of their acts that they did not foresee. The basic premise of negligence law is precisely that agents can be legally responsible for harms they were unaware of causing—for the effects of their inattention, lapses of memory or judgment, failures to recognize what is going on, and so on. Moreover, when we turn from legal to moral responsibility, we find that we often hold people morally responsible for the effects of these same failures.Footnote ⁴ Thus, if the luck egalitarian begins with our ordinary notion of responsibility and allows it to shape his understanding of luck, then he will not count every cognitive lapse that is beyond an agent's control as a matter of luck. As a result, his view will imply that an inequality can be just even though the less-advantaged party was never aware of having an option which, had he taken it, would have resulted in his now having as much as his more-advantaged counterpart.

We can now see why the two starting-points yield different versions of luck egalitarianism. Although the two versions agree about cases in which the less-advantaged agent is not responsible for his disadvantage and has not exercised control over it, and about cases in which the less-advantaged agent both is responsible and has exercised control, they disagree about cases in which the less advantaged agent did not foresee the effects of his choices, and so lacked control over the ensuing disadvantage, but in which he should have foreseen these effects and therefore is responsible for them. When one person has less than another because he has done something whose bad effects he should have foreseen but did not, the version of luck egalitarianism that is couched in terms of control will imply that the resulting inequality is unjust, while the version that is couched in terms of responsibility will allow that it may well be just.Footnote ⁵

III. Why the Distinction Matters

If cases of this sort were rare, then the fact that the different versions of luck egalitarianism disagree about them would not be terribly significant. However, far from being rare, such cases are very common. Although, of course, people do foolish things for many reasons—they may be weak-willed, may irrationally discount the future, may attach inordinate value to spontaneity, may be seduced by the pleasures of the moment, and so on—one very common enabling condition for each type of irrationality is precisely the agent's failure to bring into clear view the consequences of what he is doing. Although it may serve certain theoretical purposes to view each agent as fully aware of how each available act will affect him, or, at worst, as fully aware of the nature and likelihood of each possible outcome of each available act, the foolish acts that we encounter in the real world are very often performed by agents whose appreciation of their options is incomplete or distorted or blurred. Although muddled choices are especially characteristic of those who live disordered lives, they also account for much of the trouble that the rest of us make for ourselves.

To bring the issue into sharper focus, it will be helpful to have some specific cases before us. Here, therefore, are a few examples of what I have in mind.

1. 1. Tattoo. Whisper, at age nineteen, has drifted into the Goth lifestyle. Like her friends, she favors studded leather jackets, spiked dog collars, and elaborate tattoos. The one she chooses is an attractive vine: it sprouts from her navel, winds sinuously up her body, extends its leaves around her neck, and spreads up the left side of her face and across her forehead. Because the tattoo is so visible and so permanent, her subsequent employment choices are limited to minimum-wage jobs at coffeehouses and low-end bars.

2. 2. No Scholar. Petey is an above-average student, but he is no scholar. Thus, no one is too surprised when he takes a leave from college to work as a roadie for a band. Petey recognizes the importance of returning to school and has every intention of doing so, but he misses the first year's reenrollment deadline and then stays away for a second year to be with his girlfriend. By the third year, his parents have become ill and need his care. Without ever deciding to, Petey has left school for good.

3. 3. Fondle. Yielding to an impulse, Dr. Rosen, an oral surgeon, fondles an anesthetized patient whom he wrongly assumes to be unconscious. Horrified, the patient files a complaint, and Dr. Rosen's license is suspended for several years. When he regains it, his practice is gone. To make ends meet, Dr. Rosen is forced to become a staff dentist at Dancing Bear State Prison in California's remote central valley, in the shadow of the forbidding Diablo Hills.

4. 4. Hothead. Salamandra is a competent employee, but she is a hothead. She frequently thinks others are treating her with disrespect, and when she feels slighted, she finds it unbearable to remain silent. Because her coworkers and supervisors find her outbursts disruptive, she rarely keeps a job for more than a few months. Over time, word of her temper gets around and she becomes unemployable.

5. 5. Balloon Payment. Porter wants to buy a house and has fallen in love with a split-level priced beyond his range. To make the purchase possible, the mortgage company offers a larger-than-usual loan with a low monthly rate and a balloon payment. Expecting to refinance before making the final payment, Porter accepts the offer. When the real estate market declines, Porter can neither refinance nor meet the payment, and thus loses his home.

6. 6. Little One. Ricia, a single woman, is pregnant and has decided to keep the baby. Although she has no real idea of who will watch the child when she is working, she hopes she can somehow muddle through. However, when the baby is born, she is unable to find a suitable caregiver, and thus is forced to leave her job. As the years pass, her skills atrophy, and she spends the rest of her life on public assistance.

7. 7. Mule. Whisper, whom we reencounter at age twenty-seven, has drifted southward, first to El Paso and then across the border to Juarez. There a recent acquaintance offers her five hundred dollars to transport some cocaine to the United States. Attracted by the opportunity to repay some debts, Whisper agrees, but she is apprehended when her tattoo attracts the attention of immigration officials. When she emerges from prison, she is saddled with a felony conviction that further degrades her already bleak employment prospects.

Because these cases are not described in much detail, it is possible to envision each agent as taking note of, but then deciding to ignore, the possible bad consequences of his or her act. However, it is also possible, and I think more realistic, to envision the agents as simply not thinking the issues through, and thus as not really registering just how badly wrong things might go. Although there obviously are degrees of clarity, there must be some threshold below which an agent is simply too unclear about the consequences of his act to be able to choose it in light of those consequences. It is true that anyone who lives on this planet must, in some sense, know that drug smugglers risk being caught and that real estate prices fluctuate. However, if an agent would grossly underestimate the risk if queried, and if his knowledge is buried so far below the surface of his consciousness that it plays no role in his deliberations and exerts no causal influence over his decision, then that knowledge is just too unconnected to his will to yield any meaningful form of control. In what follows, I will simply stipulate that Whisper, Dr. Rosen, and the others satisfy this description. If we understand them in this way, they nicely illustrate the manner in which a disadvantage can be a direct causal consequence of an agent's choices without being any part of what he has chosen.

IV. Responsible But Unlucky

Because such cases are very common, a good deal of the inequality that surrounds us is bound to be due to choices of whose consequences the less-advantaged parties were unaware. Thus, if the two versions of luck egalitarianism disagree about the justice of such inequalities, then the difference between the versions will be substantial. But must the two versions really disagree? To show that they need not, someone might try to reconcile them in either of two ways. On the one hand, he might try to assimilate the luck-centric version to the responsibility-centric version by arguing that the disadvantages agents unwittingly bring upon themselves are not really matters of luck. On the other hand, he might try to assimilate the responsibility-centric version to the luck-centric version by arguing that those disadvantages are not really ones for which the agents are responsible. Let us consider these possibilities in turn.

The claim that unwittingly incurred disadvantages are not really due to luck can itself be defended in more than one way. The most familiar strategy is to relocate the disadvantaged party's exercise of control to some earlier point in his history. If someone who unwittingly acts foolishly has previously decided not to bother acquiring information that he knew would prevent him from doing so—if he has knowingly performed what Holly Smith calls a “benighting act”Footnote ⁶—then even if he lacks control when he acts foolishly, he can plausibly be said to have exercised it when he made his earlier decision. Moreover, even if someone has performed no single benighting act, he may still have been aware that the cumulative effect of his choices was to prevent him from forming the sorts of intellectual habits which, if he now had them, would enable him to realize that he is acting foolishly. If someone's current failure to realize that he is acting foolishly can be traced to his previous decisions in either of these ways, then neither the resulting disadvantage nor the attendant inequality will be due to luck after all.

It is not hard to imagine Whisper and the others as having made just such decisions. We can, for example, envision Whisper as having smoked copious amounts of marijuana before entering the tattoo parlor, Petey as having avoided anyone who might try to convince him to return to school, and Dr. Rosen as having stubbornly refused the psychotherapy that his (now ex-) wife urged upon him. However, to establish that unwittingly incurred disadvantages are never or very rarely matters of luck, one would have to show not merely that those who incur them sometimes have exercised control at earlier points, but rather that they almost always have; and this last claim is far too strong to be credible. It is not at all unrealistic to suppose that Whisper has neither ingested any mind-altering substance nor done anything else that prevents her from recognizing the disastrous consequences of her choice of body art. Her lack of mental clarity may simply not be her doing. Also, of course, if Whisper can fail to realize that she is damaging her future prospects by getting a disfiguring tattoo, then she can also fail to realize that she is damaging her future prospects by smoking marijuana, and thus impairing her judgment, before setting off for the tattoo parlor.

Because there is often no previous moment at which someone who unwittingly acts foolishly has knowingly incurred the risk of doing so, we cannot close the gap between the two versions of luck egalitarianism by maintaining that even those self-inflicted disadvantages that appear to be due to luck are really due to earlier conscious choices. It may, however, still seem possible to close the gap by denying that the form of control that is opposed to luck requires conscious choice. To exercise a luck-canceling form of control over a given outcome, it might be suggested, an agent need not actually recognize his action as likely to bring about that outcome, but need only satisfy some weaker condition—for example, that the outcome in fact be caused by his actions, that the outcome be one whose likelihood he should have recognized, or that his failure to recognize its likelihood be due to his character or psychological makeup. Because Whisper, Petey, Dr. Rosen, and the others can all plausibly be said to satisfy each weaker condition, understanding control in any of these ways would allow us to say that their disadvantages were within their control in the luck-canceling sense. This would close the gap between the two versions of luck egalitarianism because it would mean that any resulting inequalities are just as acceptable to the theory's luck-centric version as to its responsibility-centric version.

I proposed an account of control of roughly this sort in my 2006 essay “Out of Control,” but I now think that account doesn't work.Footnote ⁷ One obvious problem is that it is, at best, linguistically awkward (and, at worst, incoherent) to speak of agents as exercising control over states of affairs of which they are unaware. However, in the current context, the more fundamental problem is that no such proposal seems consistent with the philosophical impulse that underlies the claim that luck and choice are opposed. When philosophers advance this claim—when they say, with Gerald Cohen, that “brute luck is an enemy of just equality, and, since effects of genuine choice contrast with brute luck, genuine choice excuses otherwise unacceptable inequalities”Footnote ⁸—their guiding thought is clearly that the justice of any given inequality depends on whether having what is in fact the lesser amount is part of what the less-advantaged party has chosen. However, no one can choose to have a given amount of something without choosing to perform an act, or to acquiesce in an omission, that he expects to result in his having that amount. Thus, the contrast these philosophers are drawing must be between brute luck and whatever effects of an agent's choices he foresaw when he made them. However, if so, then when luck egalitarians assert that inequalities are only just if they are or were within the control of the less-advantaged parties, the form of control they have in mind must indeed be conscious control.

V. Unlucky But Responsible

Given all this, we cannot close the gap between the two versions of luck egalitarianism by expanding the scope of the disadvantages that we take to be, or to have been, within the relevant parties' control. But what, next, of the alternative strategy of contracting the scope of the disadvantages for which those parties are responsible? Here again, there are two possibilities to consider: the conclusion that the parties are not responsible for their disadvantages might be said to follow either from an adequate account of the conditions under which an agent is responsible, or from the fact that the psychological features that account for the parties' cognitive failures are external to their real identities.

Consider first the conditions for responsibility. It is widely agreed that to qualify as responsible for some feature or outcome of what he has done, an agent must satisfy both a volitional and an epistemic condition. He must, in some suitable sense, have been both free and adequately informed. Thus, one obvious way to narrow the scope of the disadvantages for which agents are responsible is to adopt a restrictive interpretation of responsibility's epistemic condition. There is, moreover, one familiar interpretation of that condition that is very restrictive indeed. Under this interpretation, the epistemic condition asserts that an agent is not responsible for a given feature or outcome of what he has done unless he realized that his act would have that feature or outcome when he performed it. Because this interpretation takes an agent's responsibility to extend no further than the features and outcomes of his actions that are illuminated by the searchlight of his consciousness, I have elsewhere referred to it as the searchlight view. Footnote ⁹

Whisper, Petey, Dr. Rosen, and the others were not aware of the disadvantages they were incurring when they performed their actions. Thus, according to the searchlight view, they are not responsible for those disadvantages. When this conclusion is combined with the responsibility-centric version of luck egalitarianism, it implies that it is unjust for Whisper or Petey or Dr. Rosen to have fewer advantages than others. Because this is also what the luck-centric version says, the strategy of backing the responsibility-centric version with the searchlight view will indeed eliminate the gap between the two versions.

But this strategy is only as convincing as the searchlight view itself, and despite that view's undeniable popularity, I think there is good reason not to accept it. One obvious problem is that it conflicts with so many of our intuitive judgments about who is responsible for what. As the cases of Whisper, Petey, Dr. Rosen, and the others themselves show, we often do view agents as responsible for disadvantages they unwittingly inflict upon themselves (and, we may add, for harms they unwittingly do to others).Footnote ¹⁰ If we accept the searchlight view, we must dismiss all such judgments as mistaken. Of course, by itself, this objection is hardly decisive, since despite the searchlight view's counterintuitive implications, we might be compelled to accept it if it could be given a sufficiently powerful independent defense. However, despite its familiarity, the searchlight view is very rarely defended, and the prospects for mounting a compelling defense do not seem bright.Footnote ¹¹ Perhaps for these reasons, few actual proponents of the responsibility-centric version of luck egalitarianism have, in fact, paired it with the searchlight view. Instead, most take the common-sense position that even if an agent is not aware that he is acting foolishly, he may still be responsible for his act's consequences if these consequences are ones he should have foreseen (or, equivalently, if they are ones a reasonable person in his situation would have foreseen).

As this last parenthetical clause reminds us, what we think a given person should have foreseen depends on where we draw the line between him and his situation. When someone with poor eyesight fails to notice something that is clearly visible, we do not say that a reasonable person in his situation would have noticed it, because we naturally think of the poor eyesight as itself part of the agent's situation. Thus, a second possible way of getting Whisper, Dr. Rosen, and the others off the hook is to maintain that their gormlessness, too, is part of the situation that determines what they should realize. If it is, then they will not be responsible for their self-imposed disadvantages because a reasonable person in their situation—that is, one who, like them, happened to be afflicted with gormlessness—would not have realized that he was incurring any disadvantage. This conclusion, when combined with the responsibility-centric version of luck egalitarianism, will again imply that it is unjust for Whisper or Petey or Dr. Rosen to have fewer advantages than others. Because that is also what the luck-centric version implies, the strategy of externalizing the determinants of a person's failure to realize that he is acting imprudently represents a second way of bringing the two versions together by contracting the person's sphere of responsibility.

Although I know of no luck egalitarian who has explicitly taken this line, it remains worth considering because a number of theorists have said things that point naturally to it. One such philosopher is Ronald Dworkin, who acknowledges that his approach to equality “produces a certain view of the distinction between a person and his circumstances, and assigns his tastes and ambitions to his person, and his physical and mental powers to his circumstances.”Footnote ¹² Another is Richard Arneson, who at one point writes that “having preferences that increase the difficulty for the individual of delaying gratification can be an affliction on all fours with suffering from asthma or chronic headaches,”Footnote ¹³ and who at another point asserts that we cannot reasonably expect as much from a person who “[finds] work extremely aversive” as we can from someone with a “zest for hard work.”Footnote ¹⁴ These passages are relevant because once we have taken a person's circumstances to include his mental powers in general, and his reluctance to delay gratification and aversion to work in particular, it is hard to see on what principled grounds we might refuse to take his circumstances to include the (other) causes of his failure to realize that he is acting imprudently.

But precisely because the siphoning away of any of a person's mental powers is bound to create a powerful suction, there is a real danger that this strategy will undermine itself. It will be self-undermining if it commits us to exporting so many of an agent's features that there is no self left to be responsible for anything; for, in that case, the same considerations that bring the luck-centric and responsibility-centric versions of luck egalitarianism together will also empty the responsibility-centric version of its content. Thus, the challenge for those who seek to externalize the features that account for an agent's failure to realize that he is acting imprudently is to make sense of the “him” to whom those features are said to be external.

Because this challenge again raises questions that lie beyond the scope of this essay, I can do little more than register my deep skepticism that the challenge can be met. Expressed in the broadest terms, the reason for my skepticism is the degree to which the sources of the relevant failures are, or are entwined with, the dispositions, attitudes, and states that make up a person's character. Although my previous description of Whisper, Petey, and the others as gormless was obviously a shorthand, it seems equally obvious that any adequate expansion of it will have to specify the parties' characteristic attitudes and tendencies of thought. When Ricia does not face up to her inability to combine motherhood and work, the explanation may lie in her hostility to her disapproving parents, her tendency to engage in magical thinking, or her incurable optimism. When Dr. Rosen does not realize that he is jeopardizing his future, the reason may be that he is too arrogant to take such threats seriously or that he has poor impulse control. When Whisper fails to register the stigma to which she is exposing herself, her failure may reflect her deep passivity or her low sense of self-worth. In each of these cases and in innumerable others, the causes of the agent's failure to realize that he is acting imprudently cannot plausibly be attributed to his situation, because its causes are far too deeply implicated in the tangle of dispositions, attitudes, and traits that together make him the person he is.

VI. Different Versions, Different Tasks

So far, we have seen that the two versions of luck egalitarianism yield different verdicts about the justice of many inequalities. On the one hand, the disadvantages incurred by Whisper, Petey, and Dr. Rosen are unforeseen and thus beyond their control, and are therefore the source of inequalities that the luck-centric version judges to be unjust. On the other hand, Whisper, Petey, and Dr. Rosen do seem responsible for the disadvantages that flow from their unwitting choices, and thus the resulting inequalities do not seem unjust according to the responsibility-centric version. But how, in that case, should we understand the many formulations of luck egalitarianism that blur the distinction between its two versions—assertions such as “it is bad if some people are worse off than others through no voluntary choice or fault of their own,”Footnote ¹⁵ and “the right cut is between responsibility and bad luck”?Footnote ¹⁶ When luck egalitarians say things like this, what are they saying about agents like Whisper and Petey and Dr. Rosen?

Before I can answer this question, I will have to make a brief detour through the theoretical underpinnings of luck egalitarianism. Like a proponent of any other normative theory, a defender of luck egalitarianism confronts both an analytical and a justificatory task. In its analytical aspect, his task is to explain exactly what he takes justice to require; in its justificatory aspect, it is to explain why we should accept his account of justice. Because what sets luck egalitarianism apart from other forms of egalitarianism is precisely its willingness to accept some inequalities as just, a crucial part of the analytical task is to account for the justice of these inequalities within the egalitarian rubric. (Another part, which I mention only to set it aside, is to provide a well-motivated answer to the question, “Equality of what?”) By contrast, in its justificatory aspect, a crucial part of the luck egalitarian's task is to defend his account against various objections. Predictably, some of these, such as objections that appeal to property rights, maintain that the account goes too far in the direction of equality, while others, such as objections that appeal to the suffering of the disadvantaged, maintain that it does not go far enough. Of these two tasks, the analytical one is internal to the theory of egalitarianism, while the justificatory one is external to it.

The importance of the distinction between analysis and justification begins to emerge when we notice that the notions of choice and the absence of luck line up naturally with the luck egalitarian's analytical task, while the notion of responsibility lines up no less naturally with the justificatory one. In his magisterial early statement of luck egalitarianism, Ronald Dworkin made it very clear that his aim was fundamentally analytical. He said he wanted to articulate a version of egalitarianism that was “ambition sensitive”—elsewhere he has described it as “choice sensitive”Footnote ¹⁷—and his rationale for wanting to do this was precisely that under equality of resources,

By thus reconceptualizing equality as a relation between the advantages enjoyed by persons considered as temporally extended entities, and by tolerating inequalities that reflect people's choices about how to use their equal resources, Dworkin in effect offered an egalitarian response to Robert Nozick's purportedly anti-egalitarian observation that “liberty upsets patterns.”Footnote ¹⁹ Because it emerged in the context of an attempt to develop an “attractive theoretical development of equality of resources,”Footnote ²⁰ the role that Dworkin assigned to choice (and so, by contrast, to its opposite, brute luck) was clearly internal to his theory of equality. Moreover, although I shall not argue the point, I think the case for assigning this sort of role to choice is just as strong if we say that what egalitarianism seeks to equalize is not resources but opportunities for welfare or advantage.

Significantly, terms like “fault” and “desert” do not appear in Dworkin's original analytical discussion. However, and in stark contrast, such terms do appear in many discussions that take the crucial division to be between disadvantages for which agents are responsible and those for which they are not. When luck egalitarians speak of disadvantages that agents are responsible for incurring, they very often go over into speaking of these disadvantages as deserved. To cite just one example, Nicholas Barry has recently written that

This association of responsibility with desert is significant because attributions of desert have independent normative force. To say that someone deserves a disadvantage is to attribute some sort of value to his having it. Moreover, even without bringing in desert, many attach special moral significance to the advantages and disadvantages for which agents are responsible. In light of these facts, the introduction of responsibility in this context is best understood not as part of the luck egalitarian's specification of what equality requires, but rather as contributing to his justification of the proposed deviations from a more comprehensive equality.

Taken together, these considerations shed light on the underlying structure of luck egalitarianism. They suggest that its luck-centric version represents its core analytical idea, while its responsibility-centric version adds a justificatory element. More specifically, the responsibility-centric version is what we get when we combine the Dworkinian insight that supports the luck-centric version—the insight that to equalize the advantages that are available to people over their lifetimes, we must allow them to make choices that will lead to inequalities both in their welfare and in the resources they command—with a natural way of softening the impact of the fact that accepting such inequalities means allowing people to suffer the consequences of their foolish choices and unsuccessful gambles. Because the resulting inequalities can be glaring, and because they are easily portrayed as morally objectionable, it is not surprising that luck egalitarians have sought to block this objection by pointing out that the worse-off parties, who have chosen to incur or risk incurring their disadvantages, therefore deserve or are responsible for them. When we incorporate this observation into our description of the deviations from equality that are called for by a whole-life view of equality, we arrive at the responsibility-centric version of luck egalitarianism. And, hence, that version stands revealed as a hybrid which combines a partial account of what equality requires with a partial answer to the question of why we should embrace it.

VII. Resolving the Tension

If, when someone's opportunities match those of others, he consciously chooses to incur or risk a disadvantage, he becomes responsible both for that disadvantage and for any attendant inequality. Because such inequalities are precisely the ones that the luck-centric version of luck egalitarianism accepts as just, it follows that any inequality that is just according to the luck-centric version is also just according to the responsibility-centric version. However, there remains a tension between the two versions because an agent can also be responsible for a disadvantage that he has not consciously chosen to incur or risk. This is the tension that Tattoo, No Scholar, and my other examples (in Section III) were introduced to illustrate, and, to conclude, I will explain how the remarks of the previous section are relevant to this tension.

I have suggested that anyone who accepts the responsibility-centric version of luck egalitarianism must accept both (1) a principle of equality that endorses inequalities that reflect the choices of the less-advantaged parties, and (2) the view that there is value in people's getting what they deserve or living with outcomes for which they are responsible. Because these are distinct normative claims, anyone who accepts them both must be some sort of pluralist. However, within this general rubric, there are two sub-possibilities: one might hold either that the value of equality always dominates the value of responsibility or desert when these conflict, or that it does not. Bearing this distinction in mind, let us ask what luck egalitarians of each stripe would say about Whisper, Petey, Dr. Rosen, and the others.

Consider, first, the claim that desert and responsibility do have some independent weight, but that equality always dominates in cases of conflict. If a luck egalitarian holds this position, then he will happily take the value of an agent's getting what he deserves, or of his living with the outcomes for which he is responsible, to blunt the force of the objection that allowing agents to live with whatever disadvantages they chose to incur or risk is too harsh to be just; for this supplementary argument will be consistent with a principle of equality that classifies the attendant inequalities as just. By contrast, such a luck egalitarian will not take the value of desert or responsibility to justify any unchosen inequalities for which agents are responsible; for these inequalities will be classified as unjust by the principle of equality that he regards as dominant. Thus, to a luck egalitarian who believes that equality does dominate desert and responsibility, the inequalities that flow from the actions of Whisper, Petey, and the others will all be unjust.

But the other sort of pluralistic luck egalitarian, who thinks desert and responsibility have independent weight and that equality does not always dominate them in cases of conflict, will draw a very different conclusion. No less than the luck egalitarian who believes that equality does always dominate, he will happily invoke the value of an agent's getting what he deserves, or of his living with the outcomes for which he is responsible, to blunt the force of the objection that allowing him to live with the disadvantageous consequences of his previous choices is too harsh to be just. However, because this sort of luck egalitarian does not regard equality as the dominant factor in determining what justice requires, he will also hold that it can be just for someone to get what he deserves, or to live with disadvantages for which he is responsible, despite the fact that those disadvantages are unchosen and therefore are not endorsed by the principle of equality he favors. Hence, if a luck egalitarian believes that equality does not dominate desert or responsibility, then he will acknowledge that the inequalities that flow from the actions of Whisper, Petey, and the others are not necessarily unjust.

Of these two ways of combining the values of responsibility and desert with a theory of equality, which should we prefer? Should we take those values to be weighty enough to imply that it is not unjust for Whisper, Dr. Rosen, and Ricia to end up with less than others because of disadvantages that they neither foresaw nor chose but for which they are nevertheless responsible? Or should we say, instead, that because equality always takes priority over the values of responsibility and desert, the fact that these inequalities violate the favored principle of equality tells decisively in favor of mitigating them?

Because these questions again raise issues that lie far beyond the scope of the present essay, I will not try to answer them, but will simply end by registering my convictions on the matter. For what it's worth, I think that if we are going to be egalitarians at all, then a luck egalitarian is the best kind of egalitarian to be; and I think, further, that the best kind of luck egalitarianism is one that does not take equality to dominate all other values. This last conviction is based partly on my general resistance to absolute priority relations among normative considerations, but partly also on my intuitions about specific cases. I simply do not think it is always unjust for agents like Whisper and Dr. Rosen to have to live with consequences of their behavior that they did not choose but for which they are nevertheless responsible. There are, of course, many cases—perhaps some of the ones I have offered are among them—in which not mitigating even the richly deserved effects of people's folly is unacceptably harsh. However, to explain why mitigation is called for in such cases, we need not wheel in the heavy machinery of justice, but can instead appeal to simple humanity or charity.