II. The Need for and the Dim Prospects for Pluralist Luck-Egalitarian Ranking

In its most simple monist form, luck egalitarianism holds that there is only one factor on the basis of which alternative distributions of welfare are to be ranked against one another; that factor is the extent of arbitrary inequality found within those distributions. The less arbitrary inequality there is within a distribution, the more highly ranked it is. I bypass here important questions for the luck egalitarian about who, if anyone, is capable of determining what portion of observed or projected inequalities of welfare among individuals is or would be arbitrary, that is, is or would be due to brute bad or good luck. In order to bypass these questions, all inequalities embedded in the various tables presented and discussed below are to be understood as arbitrary inequalities.

It is clear how to gauge degrees of (arbitrary) inequality if the distributions under consideration include only two parties. The degree of inequality is the difference in lifetime welfare between those two parties. However, the determination of which alternative distribution is more unequal becomes less obvious when more parties are introduced. Consider in Table 1 these two distributions of units of lifetime welfare.

Table 1.

Degrees of Inequality: X versus X*

In distribution X, two people who are equal to one another are each 8 units less well off than each of three others. In distribution X*, one person is 8 units worse off than each of the other four. Which distribution has more inequality? Also, consider in Table 2 these two alternative distributions.

Table 2.

Degrees of Inequality: Y versus Y*

In distribution Y, more people are in the gap between the worst-off and the best-off individual, but the gap between the worst-off and the best-off is smaller in distribution Y*. Which of these distributions has more inequality? These complications will plague any real-world attempt to identify what social and political policies and what personal conduct a luck egalitarian should endorse. However, I shall not pursue these complications in this essay.

Monist luck egalitarianism has radically counterintuitive implications. A broad implication is that all unequal distributions must be ranked lower than all equal distributions. Distributions <4, 5>, <6, 7>, and <15, 1,000> will all rank lower than <4, 4> because badness attaches to each of the first three while badness does not attach to <4, 4>. For monist luck egalitarianism, that is the end of the story. Similarly, badness attaches to <3, 10>, <3, 150>, and <3, 3,000>, so each of these must rank lower than <4, 4>. If equality is the only good-making feature of distributions, there is no basis for ranking <1,000, 1,000> over <4, 4>. It is purely ad hoc for a monist luck egalitarian to say that vastly greater aggregate welfare within one equal distribution, for example, <1,000, 1,000>, breaks its egalitarian tie with another equal distribution, for example, <4, 4>. Moreover, if aggregate welfare may be invoked to break this tie on behalf of <1,000, 1,000>, may not this consideration also be invoked to explain why <1,000, 999> is more valuable, all things considered, than <4, 4>? If so, some nonegalitarian measure of the value or disvalue of distributions has fully been brought into play.

In fact, serious luck egalitarians are pluralists, insisting that equality or inequality is only one dimension along which one distribution of welfare can be better or worse than another. For Segall, inequality is “one respect” in which a distribution of lifetime welfare can have disvalue (or value).Footnote ¹⁰ Larry Temkin says “the fact that ideals like equality, utility, and freedom sometimes have implausible, or even terrible implications, does not show that these ideals do not matter. It merely shows that each ideal, alone, is not all that matters.”Footnote ¹¹ He goes on to say that different conceptions of equality matter, thus all-things-considered egalitarian judgments must duly take account of both these different conceptions of equality and different ideals (only one of which is equality).

The first and most obvious further factor that comes into play for pluralist luck egalitarianism is aggregate welfare. In ranking, say, <3, 150> against <4, 4>, greater aggregate welfare in the former counts in its favor. Another obvious example that fits comfortably within the distributional perspective is Pareto superiority. One distribution, for example, X*, is Pareto superior to another, for example, X, if some individual is better off in it and no individual is worse off in it than they would be within the alternative distribution. For pluralist egalitarianism, the goodness that arises from the Pareto superiority of distribution <6, 7> relative to another distribution may outweigh the badness that arises from its inequality. One can also envision factors that outweigh counteregalitarian factors, for example, a bad-making feature of an unequal, albeit Pareto-superior distribution may be that the party who gains more (relative to an alternative equal distribution) gains at least N times more than the one who gains less.

For the egalitarian, the crucial implication of pluralism of weighty dimensions is that judgments that, all things considered, some unequal distributions, such as <3, 15> or <6, 7>, are more valuable than some equal distributions, such as <4, 4>, do not disconfirm the egalitarian claim that—within any distribution—inequality is a bad-making feature. Such all-things-considered judgments merely reflect the outweighing of this bad-making feature by one or more good-making features. The more the pluralist egalitarian emphasizes the role of distinct nonegalitarian features that confer goodness or badness on distributions of welfare, the less reason there is to label the pluralist doctrine “egalitarian.” The pluralist doctrine will be no more egalitarian than it is aggregative or Paretian, unless we have a credible account of how these different factors should be weighed in all-things-considered judgments that assign special significance to equality or inequality.

Consider the standard eyeball-equalization objection to egalitarianism. Jones has one functioning eyeball and Smith has two. According to the egalitarian, a certain degree of badness attaches to this distribution of eyeballs between Jones and Smith. Suppose that the only alternative to this situation is that both Jones and Smith have one functioning eyeball. Equality in eyeballs can be achieved only by “leveling down.”Footnote ¹² The anti-egalitarian asserts that the egalitarian is committed to the counterintuitive judgment that the <one eyeball, one eyeball> distribution is better (more valuable, less disvaluable) than the distribution in which Smith retains that second, offending eyeball. The pluralist egalitarian responds with a standard counterweighing response:

Hence, the egalitarian concludes, pluralist egalitarianism does not endorse the judgment that <one eyeball, one eyeball> has more value (or less disvalue) than <one eyeball, two eyeball>.

The problem with this sort of response is that it is too readily available since there are no limits set to what countervailing factors the egalitarian can invoke. Isn’t it counterintuitive to rank this more equal outcome higher than that more unequal one? No problem. Whenever this is the situation, one invokes a countervailing factor and declares that one’s pluralist egalitarianism itself accounts for the intuition that the unequal distribution ranks higher than the equal distribution. One could always mention that there are countervailing factors in favor of the unequal distribution and then presume that these factors are present to an extent that outweighs the badness of the inequality. In this way, pluralist egalitarianism threatens to be nonfalsifiable.

Consider Segall’s discussion of a similar case in which two individuals need a kidney to survive, but only one kidney is available in some organ bank. Segall follows John Broome in holding that the fairest thing to do “would be to give the kidney to neither of the patients,” so that “letting the kidney go to waste” is “the egalitarian course of action.”Footnote ¹³ How, then, can it be that, according to Segall, letting the kidney go to waste “correctly strikes us as sheer lunacy”?Footnote ¹⁴ This must be because the correct all-things-considered judgment is that countervailing considerations outweigh the disvalue of an unequal distribution of kidneys. Yet we are told nothing about why this would be the correct judgment. Despite the “correctly,” Segall hedges his bets by immediately saying that it “would probably be wrong all things considered”Footnote ¹⁵ to throw away the kidney. Why would it probably be wrong?

To block the threat of his doctrine becoming nonfalsifiable, the pluralist egalitarian needs an independently plausible account of how much badness inequalities of different degrees confer upon a distribution, how much goodness greater aggregate welfare of different degrees confer upon a distribution, and how much other good-making features (such as Pareto superiority) contribute to the goodness of unequal distributions.Footnote ¹⁶ The answer to the question “Why does the value engendered by the good-making features of this unequal distribution outweigh the disvalue engendered by the distribution’s inequality?” cannot be “They must generate enough value because sensitive people like myself intuit that this inequality is better than the alternative equal distribution.” On the one hand, for the pluralist egalitarian who is eager to employ the all-things-considered strategy to escape from counterintuitive monistic judgments, the more factors there are to consider, the better. On the other hand, the more things there are to consider, the more daunting is the task of explicating and vindicating a calculus that duly weighs all those factors that are supposed to be considered.

Recall Temkin’s claim that there are numerous distinct understandings of equality that have to be taken into consideration in any all-things-considered judgment. A parallel point might be made concerning welfare itself. There may well be different dimensions to human welfare, the values of which are not commensurable or not readily commensurated. Consider, for example, the values of aesthetic appreciation, efficacy within practical action, and community with others. It may be that either there is no truth about how much of one of those dimensions of human well-being is equal in value to how much of another of these dimensions or there is such a truth but it is not known to us. Thus, if the numbers within a given matrix report an equal number of units of welfare for two individuals, but those units represent different dimensions of human welfare, we cannot conclude that there is equality between the welfare of those two individuals. If the numbers reported are unequal, we cannot conclude that there is inequality between the welfare of those two individuals. If there are different dimensions to welfare, it will be yet more difficult to provide an account of how all the relevant factors are to be weighed against one another in the course of ranking alternative distributions of welfare.

Might egalitarians take some other route to escape counterintuitive judgments that would discredit egalitarianism? One possibility is that although an available equal distribution may or may seem to be more valuable than an unequal status quo, the conduct needed to convert the unequal distribution into the equal one may be morally impermissible. For instance, in the leveling-down case of eyeball equalization, if Smith does not consent to the removal of one of his eyeballs, the conduct needed to equalize eyeballs will violate Smith’s moral claim against severe interference in her life.Footnote ¹⁷ Of course, once the door is opened to there being such constraints on the means by which valuable outcomes may be promoted, most of the action in determining how individuals and institutions ought treat people may become a matter of identifying and gauging the stringency of those constraints rather than in refining and applying a calculus that will give due weight to all bad-making and good-making features. It will especially be necessary for the action to shift to identifying and gauging the stringency of those constraints, if no plausible account is forthcoming about how all those value-making and disvalue-making factors are to be weighed against one another.Footnote ¹⁸

Might the awfulness of the means involved in bringing about the (otherwise) best distribution be incorporated within a more comprehensive consequentialist calculation about which alternative distribution is best? Segall cites a passage in which Bertil Tungodden makes the common point that, for the pluralist egalitarian, the badness that attaches to an existing unequal distribution is only one factor in assessing that distribution in relation to an alternative equal distribution. Tungodden goes on to say: “Of course, if this badness can only be removed by torturing someone, then it is obvious that the new [that is, more equal] situation is worse all things considered.”Footnote ¹⁹ Here, the torture that would be needed to get a more equal distribution weighs negatively within the determination of which distribution is more valuable (or less disvaluable), all things considered.

I bring up this passage for two reasons. First, it highlights how difficult it would be to incorporate the badness of torture—and also, one would think, the badness of other violations of bodily integrity, deception, and intimidation—into an account of how to weigh all the factors thought to be involved in the ranking of distributions. Second, without any hint of such an account, Tungodden declares that the badness of the torture of one individual overturns the ranking that would otherwise obtain in the new (egalitarian) distribution over the existing unequal distribution. This sort of casual invocation of countervailing factors threatens to descend into the doctrine that equal distributions are best—except when they aren’t. To avoid that doctrine, the pluralist egalitarian needs an account of what degree of torture (or deception or intimidation) of one person outweighs, and thereby overturns, rankings that otherwise would be correct.

Setting aside the fact that any proposed weighing of bad-making features of torture should take into account the degree of torture at issue—say, multiple crushed bones versus one fingernail pulled out—consider the rankings within these two pairs of distributions in Tables 3 and 4.

Table 3.

Two No-Torture Distributions

Table 4.

Distributions with and without Torture

Luck egalitarians like Tungodden rank (ii) over (i) and rank (i) over (iii). Given these two sets of rankings, the following proposition must be true: Adding torture to a world with distribution (ii) would more diminish value (or more increase disvalue) in that world than converting distribution (ii) into distribution (i) would decrease value (or increase disvalue) in that world.Footnote ²⁰ It would thus seem that anyone offering these two sets of rankings needs a credible account of the relative weight or force of all the good-making and bad-making factors that are taken to be in play to support this proposition. Such accounts do not seem to be on offer. It is difficult to imagine how one would even begin to assemble such a theory.Footnote ²¹

There is a further reason why the pluralist egalitarian needs a theory that indicates with some precision how much the degree to which each of the numerous bad-making and good-making factors contributes to the badness or goodness of the distributions being ranked. A pluralist egalitarian will sometimes favor an unequal distribution over an equal one, even though some other egalitarian-minded evaluators will favor the equal distribution. That same pluralist egalitarian will sometimes favor an equal distribution over an unequal alternative, even though some nonegalitarian-minded evaluators will favor that unequal distribution. In both of these situations, in order to get beyond dueling intuitions, that pluralist egalitarian needs to be able to show how her assessment flows from an account of the extent to which magnitudes of the bad-making and good-making factors affect the all-things-considered badness or goodness of the distributions in question. If and only if that account has some independent plausibility does it add weight to the pluralist egalitarian’s intuition concerning which of the contending distributions is to be preferred. Unfortunately for pluralist egalitarianism, there is little reason to believe that such an account can be provided.

III. Two Responses to the NAA Objection

I now consider two responses to my NAA objection. The first is that the pluralist egalitarian can helpfully appeal to the notion of “reflective equilibrium.”Footnote ²² “Reflective equilibrium” refers to a process in normative theorizing to which an investigator brings both a preliminary theory about how judgments within some normative domain are to be reached and a set of pre-theoretical intuitions about particular cases within that domain; the investigator works to refine the preliminary theory in light of the pre-theoretical intuitions as those intuitions themselves are refined (or reinforced or discredited) by the preliminary theory (or its refinements). If the normative domain concerns the proper ranking of alternative distributions of welfare, the investigator brings together a preliminary theory about how alternative distributions of welfare should be ranked and a set of pre-theoretical intuitions about how specific particular distributions should be ranked against one another. That investigator then seeks to resolve tensions between the preliminary doctrine and pre-theoretical data by rejecting intuitions that are discredited as discordant with the preliminary theory or by modifying the theory when it is discordant with intuitions that seem too secure to dismiss. The investigator repeats this process of reciprocal refinement until a stable equilibrium between the surviving theory and the surviving intuitions obtains. An example of a preliminary theory about the proper ranking of alternative distributions of welfare would be the contractarian idea that the proper ranking procedure for alternative distributions would be the procedure that would be agreed to by rational and fair-minded individuals.

However, the sort of pluralist egalitarians I have been discussing cannot helpfully appeal to reflective equilibrium. For pluralist egalitarianism does not offer any preliminary theory about how alternative distributions of welfare should be ranked to then bring into contact with pre-theoretical intuitions about particular rankings, so an equilibrium-engendering process cannot begin. There is no independently credible theoretical starting point that tentatively gauges (and is gauged by) an array of preliminary pre-theoretical judgments. Rather, each egalitarian pluralist will seek to bring his or her inventory of pre-theoretical intuitions into a consistent whole through placing greater weight on some and less on others. Significant differences between the initial sets of intuitions of different pluralist egalitarians will survive in the form of significant differences in their more polished judgments. Neither contending pluralist egalitarian will be in a position to say that his or her ranking reflects a credible reflective equilibrium and not merely a consolidation of pre-theoretical intuitions about the weight of various distinct considerations.

The second response to my NAA objection begins by noting that many plausible normative perspectives exist that affirm a pluralism of fundamental moral considerations and explicitly deny that there is some determinate decision procedure for aggregating the force of these distinct considerations or weighing the opposing force of them against one another. According to this response, my call for an adequate account of how these diverse moral factors are to be combined so as to yield determinate all-things-considered judgments leads to the dismissal of all such pluralist doctrines. This allegedly shows that my call for such a decision procedure is too demanding.Footnote ²³

My reply involves contrasting reasonable choice on the part of an individual concerning which alternative goal he or she will pursue and reasonable choice on the part of political leaders (or some societal mechanism) concerning which alternative societal goal—more specifically, which distribution of welfare—all members of the relevant society are to be required to serve. Let us suppose that individual choice and social and political choice among alternative outcomes are to be governed by the same set of distinct values, say, avoidance of arbitrary inequality, enhancement of the welfare of individuals without diminishing the welfare of anyone, and the enhancement of aggregate welfare.Footnote ²⁴ Suppose that Mary is deciding to which of the following four distributions of welfare across other members of society in Table 5 she will devote herself, her time, her energy, and her resources.

Table 5.

Distributions Available for Personal and for Societal Choice

If Mary chooses D₃ and she is challenged to justify her choice, I think she could reasonably say that aggregate welfare is one of the crucial values that (we are supposing) should guide such choices; aggregate welfare is significantly more realized under D₃ than under any of the other available distributions, so she has good reason to opt for D₃. She could also say that, had she chosen D₁, she would have had a good reason for doing so—and similarly for D_2, but not for D₄. In order to justify the choice that she makes, Mary does not have to show that the distribution to which she has reasonably chosen to devote herself, her time, her energy, and her resources is the most highly ranked distribution, all things considered. Faced with the same choice of options, John can justify his decision to promote any of the first three distributions by pointing to a good reason he has to favor that distribution without needing to show that his favored distribution is the mostly highly ranked one, all things considered. Note that whatever distributions Mary and John respectively and justifiably choose to promote, they do not come into conflict with one another. For each is simply making a decision about the end to which she or he will be devoting herself or himself. Neither will have any basis for complaint about the other’s choice, unless there is a decision procedure revealing that one of those distributions is the best one.

Suppose, instead, that some Great Leader or Democratic Majority decrees that everyone at least to a certain extent should promote and should be required to promote one of the first three distributions. Then, that Great Leader or supporters of that Democratic Majority must be able to show individuals who favor their own devotion to another of the available distributions that the distribution favored by the Great Leader or the Democratic Majority is, all things considered, the most valuable (or least disvaluable). That demonstration will require what is not on offer, namely, a credible account of how the normative force of each of the bad-making or good-making factors are to be aggregated or weighed against the force of the others.

The ultimate lesson here is that robust pluralism should not inspire the project of overcoming its robustness by reducing each of the affirmed values to entries in a calculus that informs everyone about what combination of those values all must act to advance. Rather, the diversity of ends affirmed by robust pluralism is to be served by the recognition of individuals’ separate spheres of freedom within which individuals can exercise discretionary choice about what combination of values they will respectively serve. Since the scope of what is determined by such choices is restricted to the decision-maker’s own sphere of freedom, such a liberal pluralism does not need—and happily avoids—an account of how the normative force of all the bad-making or good-making factors are to be aggregated or weighed against the force of one another.Footnote ²⁵ The NAA objection only targets versions of pluralism that aspire to politically or socially authoritative judgments about what outcome all may be required to serve and yet fail to supply an account of how those judgments are to be established.

IV. The CE Objection, Familiar Egalitarianism, and Asymmetrical Egalitarianism

As I mentioned in Section I above, it seems that the way to rebut the CE objection is to reformulate luck egalitarianism so that badness in arbitrary inequalities is grounded in the worse-off party’s arbitrary disadvantages rather than equally in such disadvantages and the arbitrary advantages of the better-off party. In this and the next section, I focus on Segall’s attempt to rebut or sidestep the CE objection by advocating an asymmetrical luck egalitarianism that ties the badness of arbitrary inequalities in welfare to arbitrary disadvantage rather than to arbitrary disadvantage or advantage. I begin with a somewhat fuller statement of the CE objection, describe Segall’s initial version of asymmetrical luck egalitarianism, and explain why it fails to rebut or sidestep the CE objection.

Suppose that, on a scale of 0 to 10, both Jones and Smith are on course for an unappealing lifetime welfare of 3. However, just for a lark, Smith luckily plucks from a tree and consumes a miraculous apple that provides him with eighty-five years of healthy life. As a result, Smith’s lifetime welfare prospects go from 3 to 8. (If he hadn’t plucked and consumed that apple, no one else would have.) Suppose instead that one unique drop of rain falls upon Smith, more fully immunizing him from some awful disease than does any vaccine available to Jones. If a gap opens up in these ways between the life prospects of Smith and Jones, complaining about Smith’s gain or the gap that necessarily accompanies that gain seems merely to be ill-willed, envious resentment of Smith’s good fortune. How could any decent person disvalue the gain to Smith or think there is anything regrettable about it or the gap that necessarily results from that gain, even if it is purely a matter of good luck? Why begrudge Smith his good fortune? Of course, if Jones is downcast because of Smith’s gain, Jones is worse off in a person-affecting way. She is worse off in the way that envy makes one worse off. However, that is Jones’s self-induced problem; it is not a problem that vindicates a complaint against Smith or Smith’s gain.

However, before pursuing luck egalitarianism’s capacity to escape the charge of countenancing begrudging envy, I consider briefly whether luck egalitarians can plausibly bite the envy bullet.Footnote ²⁶ Such an egalitarian would say:

There might be normative theories so well-grounded, robust, and illuminating in their implications that one should jettison intuitions that run counter to them, even if those intuitions have some appeal. However, I do not think that luck egalitarianism is such a well-grounded and robust theory. Rather, it is heavily dependent on the difficult-to-conjure intuition that badness attaches in a non-person-affecting way to all arbitrary inequalities in welfare. Against that intuition is the more common and easier to grasp intuition that there is something wrong with complaints against and grievances toward the pure good fortune of others. If one has to choose between biting the bullet of endorsing those complaints and grievances and biting the bullet of denying that non-person-affecting badness attaches to arbitrary inequalities, the latter is the less jaw-breaking bite to take.

Can the luck egalitarian who does not bite the bullet of endorsing egalitarian envy plausibly maintain that she does not begrudge Smith his good fortune? An egalitarian might assert that she only disvalues the welfare gap that Smith’s good fortune necessitates; moreover, she might assert that sometimes the nonegalitarian value associated with Smith’s gain will outweigh the egalitarian disvalue of the gap and, thus, the gap should be allowed to stand. However, the fact that sometimes, all things considered, the egalitarian will not favor the elimination of Smith’s gain does not show that the egalitarian does not always begrudge Smith that gain. For even when, all things considered, the gain is allowed to stand, egalitarianism insists that it be allowed grudgingly. For one should never forget that the allowed inequality is qua inequality intrinsically bad, as Segall indicates:

Yet, this insistence on the intrinsic badness of all (arbitrary) unequal distributions seems to underscore luck egalitarianism’s vulnerability to the CE objection.

Segall seeks to avoid this vulnerability by distinguishing between familiar and asymmetrical luck egalitarianism, contending that the latter position is not subject to this objection. The familiar view is well expressed by the claim that “[a]rbitrary inequality is (intrinsically) bad wherever and whenever it obtains.”Footnote ²⁸ We also have the familiar view before us when Segall envisions tracing the badness of some arbitrary inequalities to the arbitrary advantage of the better-off: “When advantaged holdings are arbitrary (say they are the product of brute luck) holders must provide a good reason (whether of justice or otherwise) why it is right for them to hold on to these privileged positions.”Footnote ²⁹

Segall illustrates the “very wide scope” of familiar luck egalitarian claims about the intrinsic badness of arbitrary inequalities with two dramatic examples. Insofar as it is arbitrary, the welfare gap between thirteenth-century Inca peasants and well-off members of twenty-first century prosperous societies is as intrinsically bad as would be the welfare gap between low-welfare Martian moral agents and their contemporaries among prosperous Earthling moral agents: “Any arbitrary inequality, no matter the space it traverses nor the centuries it spans, is bad.”Footnote ³⁰ The familiar egalitarian position is illustrated here because there is no suggestion that it matters whether the worse-off individuals are arbitrarily worse off because of their brute bad luck or because of the brute good luck of the better-off individuals.

In contrast to familiar egalitarianism, Segall’s initial asymmetrical egalitarianism explicitly ties the badness of (arbitrary) inequalities to the arbitrary disadvantage of the less well-off party: “What is bad about (unchosen) inequalities is precisely the fact that they leave individuals undeservedly disadvantaged compared to others.”Footnote ³¹ He also holds that

According to Segall’s initial asymmetrical egalitarianism, “[t]he badness of inequality … is rooted in the conjunction of arbitrary distributions and being worse off compared to others. It is being arbitrarily disadvantaged that lies at the foundations of telic luck egalitarianism.”Footnote ³³ Furthermore, he asserts: “It is the twin fact of being arbitrarily disadvantaged that is the source of badness according to this account. It is bad for one to be arbitrarily worse off compared to others, and consequently arbitrary inequalities are always bad.”Footnote ³⁴ Segall concludes that “[u]nlike the traditional, symmetrical view, the asymmetrical view ties the badness of inequality squarely and exclusively with the position of the worse off (‘it is bad for one to be worse off than another’), and not also with the better off one.”Footnote ³⁵

Asymmetrical egalitarianism is thus supposed to enable luck egalitarianism to rebut or sidestep the CE objection:

Since badness is supposed to attach to disadvantages and not to advantages, if Jones is worse off than Smith because of Jones’s brute bad luck, badness is supposed to attach to the inequality that obtains between Jones and Smith. Yet, if Smith is better off than Jones because of Smith’s brute good luck, badness is not supposed to attach to this inequality. On such a view, Jones’s brute bad luck would provide a reason to pump up Jones’s welfare, but Smith’s brute good luck would provide no reason to lower Smith’s welfare. Thus, the initial asymmetrical view seems not to assign disvalue to Smith’s lucky gain nor to begrudge Smith his gain through pure good fortune. This asymmetry seems to contribute somewhat to an account of how alternative distributions are to be ranked. An unequal distribution that contains an arbitrary disadvantage is to be ranked lower than a similarly unequal distribution that contains an arbitrary advantage.

Segall adds that “[e]ven though these [the disadvantage of one party and the advantage of the other] are two inevitable sides of the same coin, the badness of inequality is grounded in the grievance of the worse off, not in any potential ‘vice’ of the better off.”Footnote ³⁷ The last clause clearly indicates that Segall seeks to free luck egalitarianism from association with envy by formulating a version of egalitarianism that does not affirm “any potential ‘vice’ of the better off,” that is, that eschews any negative assessment of the better-off party’s lucky gain. Also notice, though, the crucial fact that comparative disadvantage and advantage are two sides of the same coin. Smith being arbitrarily advantaged necessitates Jones being arbitrarily disadvantaged. If Jones being arbitrarily disadvantaged necessitates the badness of the gap in welfare between Jones and Smith, so too does Smith being arbitrarily advantaged. Hence, there is no basis for asserting an asymmetry between the bad-making power of Jones being arbitrarily disadvantaged and the bad-making power of Smith being arbitrarily advantaged.

Nor is the initial asymmetrical view supported by Segall’s appeal to the complaint that, under the initial asymmetrical view, the worse-off has about gaps in welfare. According to Segall, this complaint “is always directed against someone else, namely, those individuals who are better off.”Footnote ³⁸ Indeed, he asserts:

Segall’s assertion of an asymmetry of valid complaints—that is, the worse-off always has a valid complaint against the better-off while the better-off never has a valid compliant against the worse-off—undermines the idea that asymmetry frees luck egalitarianism from fostering envy. This asymmetry of complaints straightforwardly expresses resentful envy.

The more sensible view for the luck egalitarian is that the disadvantaged party’s complaint is about the (purported) non-person-affecting badness of the gap in welfare between the worse-off and the better-off parties. Segall says that the asymmetrical view “locates the badness of inequality in the complaints that individuals have.”Footnote ⁴⁰ However, the (purported) badness is not in the complaint; if it is anywhere, it is in the inequality complained about. Yet, if there is such non-person-affecting badness in or attached to the inequality, shouldn’t the better-off party have the same complaint about it as the worse-off party, namely, that the gap adds disvalue to the world in a non-person-affecting way? Shouldn’t both parties have the same reason to disvalue and seek to diminish the gap, whether it opens up or increases due to one party’s brute bad luck or the other party’s brute good luck? Since the egalitarian ought to recognize a symmetry of complaints about non-person-affecting disvalue being in or added to the world, invoking complaints cannot support asymmetry. I conclude that Segall’s initial asymmetrical view does not enable egalitarianism to rebut or escape the CE objection.

V. Revised Asymmetrical Egalitarianism and the CE Objection

The core problem for the initial asymmetry view is that it attends only to comparative disadvantage and advantage. Since Jones is just as much comparatively disadvantaged by a non-comparative (“intrapersonal”) gain for Smith as by an equally extensive noncomparative (“intrapersonal”) loss for Jones, there seems to be no way to explain why badness attaches to inequalities that arise from Jones’s comparative disadvantage but does not attach to inequalities that arise from Smith’s comparative advantage. Thus, purely lucky gains for Smith must be assessed as negatively as purely unlucky losses for Jones.

Perhaps this difficulty can be overcome by offering a revised asymmetrical egalitarianism that has both an interpersonal and an intrapersonal dimension,Footnote ⁴¹ that is, that ties the badness of inequalities to one party being both comparatively worse off and non-comparatively worse off than she has been (or could have been). Segall proposes:

Presumably, it is bad for Jones to be worse off than she could have been, for example, to catch a cold when she could have not been infected, even if this does not make her worse off compared to others. It would be bad for her in a person-affecting way, even if everyone else has also caught that cold. The badness that Segall refers to in the revised view must not be the mundane badness that attaches to an individual being intrapersonally worse off. Rather, what Segall must mean is that, if and only if one becomes intrapersonally worse off (than one could have been) and this intrapersonal worsening makes one comparatively (interpersonally) worse off, badness of the non-person-affecting sort will attach to the resulting inequality (or increase in inequality). This conditioning of the non-person-affecting badness of an (arbitrary) inequality of welfare on the (arbitrary) intrapersonal loss of the worse-off party is at the core of each of Segall’s further statements of the revised view:

Segall asks us to imagine a “one-person world” inhabited by Robinson: “For egalitarian disvalue to obtain according to [revised egalitarianism], it is not enough that Robinson was better off yesterday, before he tripped and sprained his ankle… . Rather for an egalitarian disvalue to obtain, his sprained ankle must also make him worse off compared to others.”Footnote ⁴⁸ If Robinson suffers a welfare-reducing injury, that loss will engender a Robinson-affecting badness. However, it will not engender any badness of interpersonal inequality because there is no man Friday in comparison with whom Robinson becomes interpersonally worse off. In contrast, were Friday present—according to Segall, anywhere in space and time—and, say, at Robinson’s pre-injury welfare level, Robinson’s injury would engender an interpersonal inequality and non-person-affecting badness would attach to this engendered inequality.

Let us compare what I call the “disadvantage case” and the “advantage case.” In the disadvantage case, Jones is on course for a lifetime welfare of 8 and so is Smith. However, through brute bad luck Jones’s lifetime welfare level drops from 8 to 3:

$$ \mathrm{Disadvantage}\ \mathrm{Case}:<8,\hskip0.2em 8>\to <3,\hskip0.2em 8> $$

This brute bad luck makes Jones intrapersonally disadvantaged and, through this intrapersonal disadvantage, Jones becomes interpersonally worse off. According to revised egalitarianism, badness (egalitarian disvalue) attaches to the resulting interpersonal inequality. Contrast this with the advantage case in which Jones is on course for a lifetime welfare of 3 and so is Smith. However, through brute good luck Smith’s lifetime welfare level rises from 3 to 8:

$$ \mathrm{Advantage}\ \mathrm{Case}:<3,\hskip0.2em 3>\to <3,\hskip0.2em 8> $$

This brute good luck makes Smith intrapersonally advantaged and, through this intrapersonal advantage, Smith becomes interpersonally better off and Jones becomes interpersonally worse off. However, according to revised egalitarianism, badness of the non-person-affecting sort (egalitarian disvalue) does not attach to the resulting interpersonal inequality because it does not arise from anyone’s intrapersonal disadvantage.Footnote ⁴⁹ Thus, according to Segall’s revised egalitarianism, Smith’s good fortune does not infuse badness into the resulting (arbitrary) welfare inequality between Jones and Smith. Hence, it seems that this egalitarianism cannot properly be charged with countenancing churlish envy of Smith’s good fortune.

It is not clear exactly how to understand the interplay between the worse-off party’s intrapersonal loss and that party’s resulting interpersonal disadvantage so as to engender the badness that is said to attach to that inequality. Perhaps the interplay goes something like this: Jones’s intrapersonal loss creates a potentiality for non-person-affecting badness, but this potentiality is actualized, if and only if an interpersonal inequality obtains in which that badness can be realized. The potential badness hovers unrealized unless and until an interpersonal inequality arises in which that badness can be realized.

I shall pose three difficulties for the revised version of asymmetrical luck egalitarianism. The first and primary difficulty with this revised view is that no independent rationale is offered for its key feature, namely, that the intrapersonal disadvantage of the worse-off party is a crucial condition for the badness of the interpersonal inequality between that party and the better-off party that arises because of that intrapersonal disadvantage. Conditioning the non-person-affecting badness of inequalities on the worse-off party’s intrapersonal loss yields the conclusion that Segall wants. If the conditioning is justified, the luck egalitarian sidesteps the CE objection. However, what can be said for the postulation of this condition beyond its having the desired implication? Why (else) believe that the worse-off party’s mundane, person-affecting, intrapersonal loss explains the non-person-affecting badness that purportedly adheres to the interpersonal inequality that arises because of that intrapersonal loss?Footnote ⁵⁰

Neither Jones’s welfare loss nor the disvalue of that loss for Jones provides a plausible basis for ascribing non-person-affecting disvalue to an inequality of welfare that may open up because of Jones’s loss. To be clear, I am not denying the possibility of non-person affecting badness. Rather, I am questioning that a party’s intrapersonal disadvantage or the disvalue that attaches to it helps us to understand the (purported) egalitarian disvalue of the interpersonal inequality that comes into existence because of that intrapersonal loss. The interesting novel idea within luck egalitarianism is that there can be a different species of badness that attaches to interpersonal inequalities or disadvantages rather than to negative conditions within the lives of individuals. However, the more different in kind the badness of interpersonal inequality or disadvantage is supposed to be from the badness of intrapersonal disadvantage, the more difficult it is to see how the occurrence of the latter can explain the occurrence of the former.

A second difficulty for revised egalitarianism also arises in connection with explaining the badness of interpersonal inequalities on the basis of intrapersonal inequalities, that is, intrapersonal losses or gains. The difficulty here is that, if one party’s arbitrary intrapersonal loss has the exotic power of attaching non-person-affecting badness to the interpersonal inequality that arises because of that intrapersonal loss, why doesn’t the other party’s arbitrary intrapersonal gain have the exotic power of attaching non-person-affecting badness to the interpersonal inequality that arises because of that intrapersonal gain? Why should only bad intrapersonal changes in individual welfare trigger non-person-affecting badness in resulting interpersonal inequalities? Invoking the potency of Robinson’s arbitrary intrapersonal disadvantage to explain the non-person-affecting disvalue of the resulting arbitrary interpersonal inequality seems no more reasonable than invoking the potency of Friday’s arbitrary intrapersonal advantage to explain the non-person-affecting disvalue of the resulting interpersonal inequality. However, if Friday’s intrapersonal gain engenders as much egalitarian disvalue in the resulting inequality as Robinson’s intrapersonal loss does, we are back to a symmetrical egalitarianism that supports begrudging Friday his good fortune.Footnote ⁵¹

Anyone who compares the emergence of an inequality by way of one party gaining with the emergence of the same inequality arising by way of the other party losing is apt to look more favorably upon the first of these inequalities.Footnote ⁵² Anyone is apt to favor the advantage transition <3, 3> ➔ <3, 8> over the disadvantage transition <8, 8> ➔ <3, 8>. An advocate of revised luck egalitarianism might say that this preference reflects and, therefore, is evidence for the fact that badness attaches to inequalities that arise from intrapersonal disadvantage but not to inequalities that arise from intrapersonal advantage. Nevertheless, less speculative explanations are available for the advantage transition being favored over the disadvantage transition.

For nonegalitarians any combination of the considerations we have been considering as attaching value to a distribution—namely, the second party gains, someone gains and no one loses, and aggregate welfare increases—is a reason for favoring the advantage transition over the disadvantage case. For familiar pluralist egalitarians, in both cases the change in welfare level for one party introduces egalitarian disvalue. However, according to the familiar pluralist egalitarian, in the advantage transition, this egalitarian disvalue is outweighed by the combination of the value of the second party’s personal gain, the value of a gain to someone without a loss to anyone, and the value of an increase in aggregate welfare. In contrast, in the disadvantage transition, some combination of the disvalue of the first party’s personal loss, the disvalue of no one gaining, and the disvalue of diminishing aggregate welfare combines with the (supposed) egalitarian disvalue to bolster the overall disvalue of the transition. Neither the nonegalitarian explanation nor the familiar egalitarian explanation appeals to the idea that a transition from an egalitarian to a nonegalitarian distribution introduces badness, if and only if the interpersonal inequality arises from an intrapersonal loss.

The third difficulty for revised egalitarianism involves an implication of this doctrine that is apt to make uncomfortable any advocate of luck egalitarianism. A key feature of luck-egalitarian doctrines is that any departure from equality requires justification. There is a normative cost to any departure from equality and, hence, any such departure is unjustified, unless the departure also adds a large enough infusion of countervailing value by improving the world along some nonegalitarian dimensions. However, in postulating a necessary—indeed, a dominant—role for intrapersonal disadvantage in explaining the badness of interpersonal inequality, revised egalitarianism opens the door to there being many departures from equality that have no moral cost, namely, all those departures that do not arise because of one party’s intrapersonal loss. None of those departures will be unjustified and all of those departures that improve the world to any degree along some nonegalitarian dimension will be positively justified.

This point is especially clear, if we consider the good-making dimension of Pareto superiority. With respect to this good-making feature, both distributions B and B* are to be ranked higher than distribution A in Table 6.

Table 6.

Equal and Pareto-Superior Distributions

Within familiar forms of pluralist egalitarianism, taking note of the good-making feature of Pareto superiority is only one step within the all-things-considered ranking of distributions A and B or distributions A and B*. The next step involves weighing the disvalue of the inequality of distributions B or B* against the value of their Pareto superiority. As pluralist egalitarians are eager to say, even if one ends up ranking distribution B (or B*) over distribution A, all things considered, one should not surrender the crucial egalitarian claim that the inequality of distribution B (or B*) is in itself intrinsically bad.

However, this next step has no place within the ranking procedure called for by revised egalitarianism. For, according to revised egalitarianism, since Jones is not intrapersonally disadvantaged under any Pareto-superior inequality, such as under either distribution B or B*, no intrinsic badness attaches to the worse-off party’s interpersonal disadvantage. On the revised view, any value that attaches to a Pareto-superior distribution must outweigh the disvalue of its inequality because its inequality can have no disvalue no matter how great that inequality might be.

Project this result on to the ranking of alternative sets of basic institutions and norms that affect the distribution of welfare among members of large-scale, complex, ongoing societies. Many distinct sets of institutions and norms, the establishment and enforcement of which will engender distributions of welfare among the members of the relevant society, will be Pareto superior to institutionally enforced equality of welfare.Footnote ⁵³ A common view among egalitarian theorists is that a burden of justification must be satisfied by anyone who advocates the establishment and enforcement of any one of those mutually advantageous sets of basic institutions and norms. Some case has to be made that enough value is added by adopting the relevant Pareto-improving institutions to justify the move away from an initial state of equality. However, revised egalitarianism unintentionally eliminates this burden of justification. For, according to revised egalitarianism, as long as no one is intrapersonally disadvantaged, there can be no disvalue in the resulting distribution of welfare that nonegalitarian value must be great enough to overcome. On the basis of these three difficulties, I conclude that revised luck egalitarianism is neither a viable doctrine nor one that accords with the general thrust of luck egalitarianism. Thus, it is not a basis for luck egalitarianism to rebut or sidestep the CE objection.

However, here is a complication.Footnote ⁵⁴ Segall has a very broad understanding of a person being made intrapersonally worse off than she could have been. According to Segall, if Jones is not born wealthy and Smith is, then Jones is worse off than she could have been. Whenever there is a windfall for Smith that could have fallen upon Jones, Jones is intrapersonally worse off than she could have been and, due to that, is also interpersonally worse off: “Jones … has suffered an intrapersonal misfortune (the misfortune of not being born into wealth) which, in turn, made him worse off compared to [Smith or Rockefeller].”Footnote ⁵⁵ Segall’s view seems to be that Jones suffers intrapersonal misfortune, even if none of the windfall for Smith or Rockefeller can be redistributed to Jones. Given this understanding of a person being made intrapersonally worse off than she could have been, every case of a person being interpersonally worse off is a case of that person being intrapersonally worse off. Nothing more is required for Jones to be intrapersonally disadvantaged than for Jones to be interpersonally advantaged!Footnote ⁵⁶ If we attend to this view of being made intrapersonally worse off than one could be, even revised egalitarianism negatively assesses all brute good luck gains to Smith. In this light, contrary to my argument concerning the third difficulty for it, revised egalitarianism does not clash with familiar egalitarian intuitions. However, that is because it collapses back into familiar egalitarianism and, for that reason, cannot distance itself from countenancing churlish envy.Footnote ⁵⁷