n her recent, provocative essay "What Is the Point of Equality?", Elizabeth Anderson argues against a common ideal of egalitarian justice that she calls "luck egalitarianism" and in favor of an approach she calls "democratic equality." According to the luck egalitarian, the aim of justice as equality is to eliminate so far as is possible the impact on people's lives of bad luck that falls on them through no fault or choice of their own. In the ideal luck egalitarian society, there are no inequalities in people's life prospects except those that arise through processes of voluntary choice or faulty conduct, for which the agents involved can reasonably be held responsible. Anderson asserts that the adherents of luck egalitarianism, which can be elaborated in many different ways, include John Roemer, Erik Rakowski, Thomas Nagel, Ronald Dworkin, Gerald Cohen, Richard Arneson, and (with a qualification) Philippe Van Parijs. In contrast, according to the democratic equality conception, justice as equality requires an end to oppressive social relationships. In the ideal society of democratic equality, the social conditions of everyone's freedom are secured, each stands to every other in a relationship of fundamental equality, including equal respect, and all have real freedom to participate in democratic self-government.
Anderson's criticisms of luck egalitarianism score good points against a variety of views, including views I have defended. In this comment I do not aim to defend luck egalitarianism across the board, but rather to identify one member of the luck egalitarian family that is not vulnerable to Anderson's criticisms, is plausible in its own right, and in particular emerges as superior to the "democratic equality" conception of egalitarian justice.
First, a preliminary clarification. As characterized above, luck egalitarianism is ambiguous, and may not involve any affirmation of egalitarianism at all. After all, (1) It is bad if some are worse off than others through no fault or choice of their own is consistent with (2) It is bad if some are as well off as others through no merit of their own. Moreover, (1) and (2) together might be interpreted as an assertion of the principle of moral meritocracy, that is, (3) It is desirable that each person should be exactly as well off or badly off as she deserves. The principle of moral meritocracy affirms that the degree of overall good fortune each person gets should correspond to her moral merits or deservingness. This principle holds that it is desirable that the equally meritorious enjoy equal good fortune, but the equality here is not regarded as intrinsically morally desirable, just a byproduct of what really matters, namely, that each person should get what she deserves. In order to ensure that there must be some egalitarianism in luck egalitarianism, for purposes of discussion I shall stipulate that to qualify as a luck egalitarian, one must hold either (a) other things being equal, it is morally undesirable if some persons are worse off than others or (b) other things being equal, it is morally more desirable that a benefit should be gained for a person, the worse off the person is prior to receipt of this benefit. I favor the b variant.
Anderson urges that the luck egalitarians wrongly focus their attention on the distribution of privately owned goods among individuals. Against this shopping mall egalitarianism, we should insist that the point of equality is creating and sustaining a community of equals. What matters is equality in certain human relationships. So says Anderson. The issue raised here is how the egalitarian should be measuring inequality among persons. If egalitarian justice requires those who are better off to help those who are worse off, we need some standard that tells who are better off and who are worse off. One proposed standard is that people are equally well off when the resources they command that can be used to further their goals are the same or equivalent; variants of the proposal suggest different yardsticks for measuring individuals' overall resource levels. In these views the resources I have will surely include many goods that are not candidates or suitable candidates for private ownership.
But even the broadest and most inclusive measure of a person's resources would be rating a person's condition by the tools and means and opportunities she has available. I would suppose that what we fundamentally should care about when we are evaluating the distributive fairness of society is the quality of life that individuals reach by these means, in other words their level of well-being, where this is identified neither with bare preference satisfaction nor enjoyment but with achievement of what is objectively worthwhile or choiceworthy in human life. This stipulation defeats the fetishism objection that the luck egalitarian must claim as of fundamental moral importance what is at most of instrumental significance. The egalitarian of well-being is concerned with the quality of human relationships that people sustain in a society, but these are evaluated by their impact on well-being. If we were to institute relationships of perfect equality according to some measure of relational equality, but people ended up living avoidably miserable and blighted lives, then we should institute some inequality in relationships, in order to improve the quality of people's lives and the fair distribution of this aggregate well-being. Whether a type of relationship should qualify as oppressive should depend on whether or not it has an oppressive impact on people's well-being.
Anderson holds that luck egalitarianism builds consideration for individual responsibility into the theory of justice in the wrong way, with disastrous results. The luck egalitarian identifies justice with minimizing and equalizing the effects of bad brute luck on people, luck that falls on people in ways that are beyond their power to control, but this involves a harsh toleration of misfortune that falls on people through their fault or choice. In a society that is just by luck egalitarian standards, some members of society must be allowed to fall into utter destitution that is deemed to arise through their fault or choice. The luck egalitarian imperative of making social decisions to help or decline to help needy individuals on the basis of the degree to which they have exercised or failed to exercise responsibility in socially approved ways is unfair to the needy who are labelled faulty and left to languish. But the social process of distinguishing responsible from irresponsible, deserving from nondeserving citizens is inherently disrespectful and unfair to all members of society. Those who receive aid are stigmatized as incompetent failures. Those who are deemed unworthy of aid are stigmatized as morally irresponsible and undeserving. In a luck egalitarian society all members will find their privacy violated by intrusive and offensive investigative procedures that aim to classify them according to the level of badness of their lives and the degree of irresponsibility of their life choices. These invasions of privacy signal that taking the imperative of justice to be undoing the effects of all brute bad luck inherently erases the line between what is the legitimate concern of society and what should be left to individual discretion.
A deeper objection to luck egalitarianism demands attention. According to Anderson, this theory fails to express equal concern and respect for all persons. The considerations that are the basis for adopting luck egalitarian principles essentially involve a contemptuous pity of the unfortunate on the part of the fortunate and in return envy of the haves gnawing at the have-nots. Neither the attitude of pity for those viewed as worse off nor envy of those deemed to be better off is compatible with a proper egalitarian regard for persons. Luck egalitarian principles embody the idea that what fundamentally matters morally is how well off one person is as compared to others. At the root of this conception is a morally incorrect perspective that leads to distorted notions of what we owe to one another.
This objection misfires if it is aimed at the prioritarian branch of the family of egalitarian principles. Prioritarianism holds that institutions and practices should be set and actions chosen to maximize moral value, with the stipulation that the moral value of obtaining a benefit (avoiding a loss) for a person is greater, the greater the well-being gain that the person would get from it (the smaller the loss in well-being), and greater, the lower the person's lifetime expectation of well-being prior to receipt of the benefit (loss). Prioritarianism is egalitarian in tilting in favor of those who are badly off. But priority is assigned to aiding an individual in virtue of how badly his life is going, as measured by an objective scale of well-being, not intrinsically by any comparison between his life and that of others. If the attitude that a theory expresses is given by the reasons that warrant its adoption, then I see no basis for associating with prioritarianism a psychology of pity and envy. If one wants to be fair and do what is just, and after full reflection one is convinced that prioritarianism is the correct theory of justice, then one's adoption of prioritarianism reflects one's belief that this doctrine picks out what justice requires coupled with one's desire to conform to the requirements of justice. One expresses due respect for persons and treats them respectfully by acting toward persons in accordance with the moral principles that are best supported by reasons. In this sense respect for persons looks to be an unobjectionable but purely formal idea, neither a clue to what principles are best supported by moral reasons nor a constraint on what principles might be chosen.
Prioritarianism as stated does not attribute moral value per se to channeling benefits toward the more deserving and responsible, though such considerations would no doubt play an instrumental role in a fully articulated prioritarian theory. I myself am inclined to think that if two persons voluntarily engage in high stakes gambling, from which the loser emerges with unfavorable future life prospects, it is intrinsically, not merely instrumentally more valuable to provide the means to a one-unit gain of well-being to someone who is just as badly off as the unlucky gambler but arrives at this condition through bad luck that is beyond his power to control than to the unlucky gambler. Hence it is better to amend prioritarianism to responsibility-catering prioritarianism. According to the latter doctrine, the moral value of altering a state of affairs in a way that makes someone better off or worse off depends, other things being equal, on the degree of responsibility the person bears for her present condition. It is morally more valuable to provide a gain in well-being of a given size for a person with a given well-being prospect if she is less rather than more responsible for her present condition (if it is bad). In a similar way, less moral disvalue is produced by bringing about a loss in well-being of a given size to a person with a given well-being prospect if the person is less rather than more responsible for her present condition (if it is good). To have a theory, rather than a quick sketch of a theory, one would have to provide an account of responsibility and attach weights to the three elements of well-being, priority for the badly off, and responsibility in responsibility-catering prioritarianism (RCP).
RCP even if fully articulated would be an abstract moral theory, a set of principles of justice, not a specification of just institutions or just practices. These latter would vary with circumstances, which determine what institutions and practices and actions would best achieve the RCP moral goals. On any remotely plausible theory of human well-being, even if in principle interpersonal cardinal well-being and responsibility judgments can be made, in practice individuals and institutions would not have access to such information, so in practice we would be designing the most relevant and appropriate proxies we can find for the values that really matter to us. Anderson's attractive ideals of democratic equality are pitched at a somewhat lower level of abstraction than RCP, and might for all I know be a reasonably good set of means for implementing prioritarian values under favorable modern circumstances. Indeed, some of her criticisms of luck egalitarianism might be interpreted as criticisms of inept strategies for implementing RCP values.
Real disagreement arises when allegedly wrongful policies that invade privacy, restrict people's liberty for their own good, and restrict rights to equal participation in democratic politics would improve the quality of people's lives and distribute these improvements fairly according to the weighted well-being standard. One aspect of the disagreement is that Anderson accords priority to freedom on her favored interpretation of it (see below). According to RCP, having real freedom to achieve basic human goods is valuable both instrumentally and for its own sake, insofar as having wide freedom is itself a constituent of a good human life. But freedoms according to RCP are important as constituents of well-being, and the ultimate moral standard is the extent of (appropriately weighted) well-being that we achieve. To enhance weighted well-being, this or that freedom must sometimes give way. Freedom, even freedom on its morally most adequate interpretation, is not an absolute moral value that trumps all others. For example, perhaps all paternalism inevitably carries some cost of insult and stigma imposed on those whose freedom is restricted for their own good. But when paternalistic policy satisfies RCP, the insult and stigma cost is outweighed by genuine well-being gains, and is not then inherently disrespectful.
Anderson's arguments against luck egalitarianism pave the way for the democratic equality conception that she proposes to replace it. The latter is complex; this discussion just highlights some main features. The democratic equality ideal requires that all members of society should have a fundamental equal status, constituted by the real freedom possessed by all over the entire course of their lives to function as humans, to participate in civil society, and to participate in democratic political decision making. In other words, all persons are equally guaranteed the capacity to achieve a threshold acceptable level in these three domains, the generic human, the sphere of association, and the political. The maintenance of these equal freedoms is to be guaranteed over the course of people's entire adult lives, come what may. This guarantee is asserted to contrast favorably with luck egalitarianism, which countenances allowing people to languish in bondage or squalor if they are deemed to have had a fair opportunity and squandered their opportunities through their own fault or choice. Democratic equality guarantees only freedom at an acceptable threshold level. Inequalities above the threshold are not deemed per se morally undesirable. This limited guarantee imposes on individuals the responsibility to order their lives as they choose above the threshold and eschews the politics of envy that Anderson associates with luck egalitarianism. Personal responsibility also receives its due in the democratic equality norm in another significant way: The guarantees that democratic equality enforces are guarantees of access to functionings (real freedom to achieve a set level of functioning), not a guarantee of any achieved level of functioning.
This democratic equality ideal is intended to be a sketch of a theory that needs further refinement, so criticism may be premature. But as presented so far, the implications of democratic equality are implausible where they disagree with those of RCP. Democratic equality holds that once someone is above the basic capability threshold, justice is unconcerned with whether or not his life goes better or worse. Why not? Suppose that society faces an issue, say a choice of tax policy, where the interests of those who are far above the basic capability threshold (and thus on the average high in well-being) are starkly opposed to the interests of those who are just above the threshold (and thus on average significantly lower in well-being). Unfortunately someone's ox must be gored. Whose? RCP says that on the facts as described, other things equal we should favor the worse off in order to fulfill the requirements of justice. Democratic equality says that the issue is a "don't care" from the standpoint of justice. I disagree.
The force of this criticism could be blunted to some extent if the threshold of basic capability is set at a very high level. But only to an extent. Moreover, this move brings another difficulty into view. Democratic equality extends an unconditional guarantee that each member of society shall have access to the basic functioning level. But this priority ranking is too stringent. When misfortune strikes, it is a regrettable fact that some people cannot be sustained at the threshold level no matter what resources are poured into the coffers earmarked for their aid. In other cases, sustaining an individual at the threshold level is possible only at too great a cost. Morally sensitive cost and benefit calculation must be carried out to determine whether maintaining an individual at the guaranteed level (or at some specified distance from the level) is morally worthwhile all things considered, but democratic equality is inhospitable to the needed tradeoffs. The higher the threshold level of basic capability is set, the more glaring this problem becomes.
Democratic equality eschews moralizing judgments about the quality of individual lives and hectoring assessments of the degree to which individuals have behaved responsibly. Is this avoidance an advantage? To focus on the relevant issues, ignore questions concerning the availability of the information needed for making these judgments and assessments and concerning the moral cost of discovering this information if it is available. RCP affirms that what is morally right to do depends on this information. But in circumstances in which the information is unavailable or costly to obtain, RCP affirms whatever norms and policies will most efficiently advance the RCP goals. Consider then simple examples in which the relevant information is readily available. Suppose that a national park service rescue team can choose between one of three lifesaving missions. Each involves significant risk of severe harm to rescue workers, but promises a significant net saving of lives. Suppose these risks and benefits are the same for each of the three rival missions. The park rescue team must choose either to assist (a) a party of stranded schoolchildren caught in an unanticipated blizzard while on a school outing, (b) a party of experienced climbers who carefully chose to pursue a difficult route under hazardous conditions which then suddenly turned desperate, or (c) a party of tourists who ignored warning signs and the stern advice of park rangers to venture on a foolhardy hike across a treacherous steep slope, rendered more treacherous by their mid-hike alcohol consumption. One might suppose that the rescue team's policy should be set in part by consideration of its incentive effects on the behavior of future park visitors, but suppose the park is about to be shut down and there are no such incentive effects to consider. I take it to be a datum in this case that the fully voluntary choice of the climbers to shoulder the risk they take and the grossly reckless conduct of the hikers reduce their moral claims to be aided by comparison with the claim of the stranded school children. This is the basic idea of the responsibility-catering element in responsibility-catering prioritarianism. A proposed theory of justice that excludes it excludes a factor that is intrinsically morally important.
RCP also incorporates the belief that what we should do depends both on how much good we can do for people and how badly off they will be absent our intervention. In the example, policy should respond to the consideration that it is objectively worse to have one's life cut short as a child, other things being equal, than to have one's life abruptly ended after one has lived longer and had ample opportunity to sample the goods basic to a normal human life. No doubt measurements of well-being and well-being prospects are beset by conceptual difficulties such that commensurability is only partial even in principle, quite aside from the practical difficulties of acquiring the information that in theory is needed for assessment. But I have never seen a good argument for maintaining an asymmetry between the good and the right in this regard: If one supposes no rational agreement on the good is possible, one's skepticism will by parity of reasoning lead one to conclude that no rational agreement on the right is possible, and one should abandon moral theory as a lost cause.
If one rejects wholesale moral skepticism, then liberal egalitarian teleology remains a viable contender for our reflective allegiance, and responsibility-catering prioritarianism stands as one interpretation of its fundamental principle. The point of equality I would say is to improve people's life prospects, tilting in favor of those who are worse off, and in favor of those who have done as well as could reasonably be expected with the cards that fate has dealt them.