y argument in "A Question for Egalitarians" challenges egalitarians to show that the logic of their position does not commit them to absurd policies. Arneson and Freeman respond by arguing that the challenge rests on confusions. The remarks that follow are intended to show that the alleged confusions are theirs, not mine.

To begin with Arneson, his argument has three parts: a brief and accurate account of the challenge; the statement of four confusions that he thinks vitiate the challenge; and a forthright statement of what must be called his egalitarian faith.

The first confusion he claims to have found is between the inequality of groups and persons. He says that "what should move us are disparities between individuals, not disparities between groups," (Arneson, p. 1), and claims that my argument addresses disparities between groups. That this is not so is made clear by the following passage: "These policies [i.e. redistribution and compensation] will be just only if they benefit victims of injustice, and the victims cannot be identified simply as poor, minorities, or women because they, as individuals, may not have suffered any injustice. ... Overcoming injustice requires, therefore, a much more precise identification of the victims than merely membership in such amorphous groups as those of women, minorities, or the poor." (Kekes, "Question," p. 663).

The second supposed confusion is between deep and shallow inequalities. Deep inequalities are "caused by factors entirely beyond individual power to control," whereas shallow inequalities are "brought about by differences in the choices that individuals make." (Arneson, p. 2). Arneson writes: "Kekes ascribes to egalitarians the (as he sees it, absurd) view that the difference between deep and shallow inequalities matters not at all from the standpoint of assessing the justice of redistribution." (Arneson, p. 2). This is a misunderstanding. My argument is that, regardless of whether or not the inequalities are beyond the power of individuals to control, egalitarians face the following dilemma: "If the policies of redistribution and compensation take into account the degree to which people are responsible for being among the worst off, then the justification of the policies must go beyond what egalitarians have been willing to provide. For the justification must then involve consideration of merit, desert, effort, and so forth. To the extent to which this is done, the justification ceases to be egalitarian. If, on the other hand, the policies of redistribution and compensation do not take into account the degree of responsibility people have for being among the worst off, then there is no disanalogy between men, who are worst off in respect to life expectancy, and women, minorities, or the poor, who are worst off in other respects," (Kekes, "Question," p. 663-4), and then the absurd policies that I point at follow from egalitarianism.

A further claim of Arneson in this connection is that "at most Kekes would have given us 'A Question for Some Egalitarians', not 'A Question for Egalitarians'." (Arneson, p. 3). In the first paragraph of the article to which Arneson is responding, I identify the egalitarian view that I am attacking, namely, that most prominently represented by Rawls and Nagel, and then say: "In order to avoid pedantry and verbosity, this view will be referred to simply as 'egalitarian,' although egalitarianism has other versions as well." (Kekes, "Question," p. 658). My argument, then, has not claimed more than what Arneson's "at most" grants.

Arneson's third response concerns interpersonal comparisons and the need for an index of benefits. Arneson correctly notes that he and I agree that "an egalitarian theory of justice requires a standard of interpersonal comparison," that "one ... needs an index ... determining ... how to attach weight to disparate goods," and that "egalitarian theorists have tended to underestimate the significance of the index problem." He goes on to say that my argument expresses a "global skepticism that the index problem is solvable without citing any reasons for this opinion." (Arneson, p. 3). Once again, this is not so. The reasons are provided: "Suppose ... there is an authoritative list of primary goods ... [and] that only those goods are included in this list that are directly or indirectly open to social control." It still remains the case that "primary goods are weighted differently by different people, they are possessed in different degrees, and people differ in their capacity to enjoy what they have or compensate for what they lack. ... These impossible comparisons must be made possible. That has not been done, and it is most unlikely that it could be done, especially since the comparisons must support policies which are bound to disregard individual circumstances." (Kekes, "Question," p.665). The reason Arneson has missed is that an index would have to be both impersonally applicable and personally variable, and that seems to be impossible.

The fourth confusion that Arneson claims to find concerns the distinction between natural and social inequalities. He writes: "This distinction between natural and social inequalities matters for the appraisal of Kekes's polemic because his sample inequality between men's and women's life expectancy could turn out to be purely natural. ... Yet at least one of his target egalitarians, John Rawls, clearly limits the scope of his egalitarian concern to a type of socially created inequality." (Arneson, p. 4). On the preceding page, however, Arneson says: "an inequality of life expectancy attributable to these factors [i.e. to "lifestyle choices of men and women," (Arneson, p. 3)] would be a shallow not deep inequality ... [and] the moral imperative of egalitarianism urges the elimination of only deep inequalities." (Arneson, p. 3). Deep inequalities, it will be remembered, "are caused by factors entirely beyond individual power to control" (Arneson, p. 3), and shallow inequalities are, at least to some extent, within individual power to control. Yet natural inequalities that affect good lives tend to be deep, e.g. genetic inheritance, and many social inequalities tend to be shallow, e.g. wealth. Arneson, therefore, is plainly inconsistent. He cannot claim that the unequal life expectancy of men and women is both shallow and natural, and that, because it is natural, it is irrelevant from an egalitarian point of view. He falls into this inconsistency by using Rawls's distinctions between natural and social goods and between deep and shallow inequalities. It is worth considering whether Rawls is likewise inconsistent, but that is beyond the scope of this reply to Arneson.

The last point concerns Arneson's declaration of his aggressive egalitarian faith. He says: "according to egalitarian justice as I conceive of it, I am obligated to transfer resources ... provided that the ratio of the cost that I incur from this transfer to the benefit that you gain by it is sufficiently favorable ... [and that] it would be appropriate for a third party to force me to comply." (Arneson, p. 4.). This astonishing faith is aggressive because it regards it as morally obligatory to forcefully deprive people of their legitimately acquired possessions in order to give them to those who are thought by the faithful to need them more. And it is a faith because its passionate affirmation is not affected by the reasons that numerous others, as well as myself, have given against it, nor does Arneson offer any reason in its support.

It is more difficult to respond to Freeman's comments because they lack the focus and clarity of Arneson's. In part, Freeman treats his readers to an exposition of Rawls's position. His assumption seems to be that the repetition of the position on which my arguments aim to cast doubt is sufficient to lay the doubts to rest. But it is not. Another part of Freeman's comments consists in attributing to me criticisms of egalitarianism which I have not expressed and do not hold, and then defending egalitarianism against them by claiming that I have misunderstood it.

My argument is that "it is a basic egalitarian belief that serious unjustified inequalities are morally objectionable and that a measure of a just society is the extent to which it eliminates or at least reduces them. Inequalities are serious if they affect primary goods, which are necessary for living a good life. ... All serious inequalities are unjustified unless they benefit everyone in one's society, especially those who are worst off. ... Overcoming unjustified inequalities requires the redistribution of primary goods." (Kekes, "Question," p. 658). I then point out that the life expectancy of men and women is unequal because men tend to have one-tenth shorter life than women. Since life expectancy is a primary good, and since the unequal life expectancy of men and women is both serious and unjustified, egalitarians must find it morally objectionable and must adopt policies to eliminate or reduce this inequality. These policies, however, are absurd. The question arises, therefore, whether other policies favored by egalitarians to overcome other inequalities are not similarly absurd.

Freeman's comment is: "Following Kekes's account, egalitarians should also be committed to making attractive people uglier, lean people fatter, and agile people clumsier." And then he goes on: "No egalitarian holds this position." (Freeman, p. 1). What Freeman says, however, does not follow from my account because being attractive, lean, and agile are not primary goods, and consequently inequalities in their respects are not serious. Nor do I anywhere attribute these ridiculous views to egalitarians. In the rest of Section II, taking up about one-third of his comments, Freeman repeats Rawls's familiar ideas to show that Rawls does not hold the view that Freeman mistakenly thinks I attribute to him. All of this is an ill-tempered non sequitor.

In Section III, Freeman writes: "Kekes contends liberal egalitarians are unclear whether primary goods are to be distributed specifically or in combination. This too is mistaken." (Freeman, p. 3). He then goes on to provide principles to show that what I claim to be unclear is in fact clear. But since these principles are Rawls's principles of justice, since the literature devoted to their interpretation, criticism, and amendment is vast, and since the difference principle relies on the "index" that even Arneson recognizes as problematic, it will perhaps be conceded that the unclarity I claim to hold has not been dissipated.

In Sections IV and V, Freeman discusses the difficulties I point at in the redistribution of primary goods. He deals with these difficulties by denying that there are any: "if we remain with Rawls's list of primary social goods, there is no major problem of practical implementation." (Freeman, p. 4). This will certainly be news to Rawls and Nagel who felt the need to write Political Liberalism and Equality and Impartiality in part to contend - unsuccessfully, in my opinion - with the major problems of practical implementation. Apart from Rawls and Nagel, however, the strategy of dealing with difficulties by denying that they exist leaves something to be desired.

Consider next what Freeman says about my alleged failure to distinguish between natural and social goods: "I focus on Rawls's account, since Kekes relies on the difference principle for his caricature. Rawls is clear that only social goods are the subject of distributive justice. Life expectancy is a natural, not a social good (a fact Kekes recognizes near the end of his discussion)." (Freeman, p. 2). In the article that Freeman undertook to comment on, I explicitly consider and respond to this point: "Another egalitarian attempt to avoid the force of this objection may be to claim that the primary goods ... in respect to which inequality is morally objectionable are social rather than natural goods. ... Life expectancy may then be said to be a natural good, not a social one." (Kekes, "Question," p. 666). I then go on: "This attempt, however, is unconvincing for several reasons. First, while the distribution of natural goods is not open to direct social control, it is open to indirect social control. Life expectancy cannot be directly controlled by public policies, but it can be controlled indirectly by policies which redistribute jobs, health care, vacations, and so on. Second, even if the indirect social control of the distribution of natural goods is difficult, it is not difficult to find policies that would compensate for the unequal distribution of natural goods. People with a shorter life expectancy can be provided with more primary social goods than people with longer life expectancy. Third, egalitarians are quite clear that they do aim to exercise social control over natural goods, and this attempt to avoid the objection is contrary to their clearly stated position. Rawls, for instance, says that the right conception of justice "nullifies the accidents of natural endowment and the contingencies of social circumstances" and that "we wish to set up the social system so that no one gains or loses from his arbitrary place in the distribution of natural assets ... without giving or receiving compensating advantages in return," (Rawls, A Theory of Justice, pp. 15, 102). (Kekes, "Question," pp. 666-667). Once again, therefore, Freeman falsely attributes to me a "caricature" that I explicitly disavow, and fails even to notice, let alone respond to, the reasons I give against this attempt to defend to egalitarianism.

A similarly egregious maneuver is to accuse me of misquotation: "Kekes also mistakenly contends liberal egalitarians aim to 'compensate' the naturally disadvantaged when equalizing natural goods is not feasible. He quotes ... Rawls on the principle of redress, which says that "undeserved inequalities ... are to be somehow compensated for.' But then Rawls goes on to deny the position Kekes attributes to him: 'The difference principle is not of course the principle of redress.' ... Distribution of Rawls's primary social goods has nothing to do with compensation." (Freeman, p. 3). Now on the same pages in A Theory of Justice, (pp. 100-101) from which both Freeman and I quote, Rawls says that "although the difference principle is not the same as that of redress, it does achieve some of the intent of the latter principle," and that "by selecting the so-called starting places one follows out the idea of mitigating the effects of natural accidents and social circumstances." Freeman is right, of course, that the difference principle is not the same as the principle of redress, but I have not claimed that they are the same. What I claimed was that Rawls is committed to the principle of redress and that has absurd consequences. Instead of responding to this criticism, Freeman obfuscates it by falsely accusing it of resting on a misquotation.

In Section VI, Freeman almost faces one fundamental issue that divides egalitarians and their critics: desert. The egalitarian view, in Freeman's words, requires "that social institutions ... be designed for everyone's benefit, to distribute primary social goods so that each person has fully adequate means to shape a meaningful life, given their [sic] capacities, and to exercise the powers of equal citizens." (Freeman, p. 6). These fine, if grammatically shaky, words, however, will inspire only those whose critical faculties have been lulled into acquiescence by the endlessly repeated egalitarian rhetoric. Why should social institutions be designed to treat equally criminals and their victims, terrorists and their hostages, the hard-working blue collar family and those who treat the welfare system as a milch cow, people who live reasonably and those who ignore the facts, take stupid risks, and abuse their talents? Why assume that people who are among the worst off in a society are in that position because of social injustice? And what is the justification of that remarkable slide of Freeman from the universally inclusive "each person" to the exclusionary "equal citizens"? Why is egalitarianism confined to fellow citizens? Freeman does not say.

That people should get what they deserve and should not get what they do not deserve is surely one of our most fundamental moral convictions. But it is just that conviction that egalitarians must violate because what people deserve depends of their merits, and merits are unequal. Most egalitarians simply ignore the question of what people deserve. In this respect, as in others, Rawls is more forthright than the rest. He writes: "There is a tendency for common sense to suppose that income and wealth, and the good things in life generally, should be distributed according to moral desert. ... Now justice as fairness [Rawls's theory] rejects this conception." (Rawls, A Theory of Justice, p. 310).

What Freeman, and also Arneson, ought to have asked is: how could the egalitarianism they defend be morally acceptable if it permits taking from people what they deserve to have in order to give it to others who do not deserve it? The aim of "A Question for Egalitarians" was to point out the absurdity of this version of egalitarianism and to challenge egalitarians to face it. It is a pity that Arneson and Freeman have not done so.

It will perhaps have struck readers that my responses to Arneson and Freeman are largely quotations from the article on which they are commenting. The reason for this is that their attribution of confusions to me rests on missing the passages I quote. Arneson and Freeman are competent philosophers, so it cries out for an explanation how they could have missed these passages. The explanation I tentatively propose in closing is that contemporary political philosophy in America largely consists of egalitarians arguing with each other about how far left of center it is legitimate to go. Given this orientation, challenges to egalitarianism appear as unwelcome interruptions of friendly discussions among political allies. Arneson responds to my challenges patiently, Freeman does not. But they, and others, believe that egalitarianism is fundamentally correct, and this leads them to take challenges to it as signs of confusion. Arneson and Freeman have read my argument in order to find these confusions. This is perhaps why they miss the passages that raise fundamental questions about egalitarianism. These questions deserve answers, even if they disrupt the amiable consensus among like-minded, but wrong-headed, political thinkers.