n "What's the Point of Equality?" (Ethics, January 1999) Elizabeth Anderson has written a rich and very thought provoking critique of some going accounts of equality and offered her own account as a superior alternative. There is much to think about in this fine essay but given limitations of space I have chosen just a few topics. Even these are covered in a necessarily cursory manner. In this piece I will discuss Anderson's two main arguments against what is called the luck egalitarian conception of equality. First, I will argue that luck egalitarians need not accept the characterization that Anderson attributes to them concerning the outcomes of voluntarily incurred risks. Second, I will argue that Anderson's description of the luck egalitarian approach to compensation for handicaps is unwarranted. Finally, I will return to the question of paternalism in the luck egalitarian approach to losses that result from voluntarily incurred risks. There are a number of important ideas in Anderson's paper that I will not be able to broach in this piece. I refer to her anti-welfarist approach to the question of equality of what and to her discussion of the problem of the relation between the recognitional components of justice and the distributive components.

Bad Luck and Respect for Persons
As Anderson describes it, luck egalitarianism is the principle that states that "the fundamental aim of equality is to compensate people for undeserved bad luck." (288. The numbers refer to the pages of Anderson's paper). In particular the idea is that the principle of equality implies that individuals' lives ought not to go worse than others merely as a result of bad luck. For example, egalitarian justice requires that individuals' lives ought not to go worse than others' lives merely on account of having been born in impoverished social circumstances, or with disabling handicaps, or with low levels of native talent. These cases are paradigm examples of brute bad luck, i.e. bad luck for which the individual has no responsibility whatsoever. A just society is one that attempts to eliminate some forms of bad luck, or tries to adjust social conditions to accommodate those with other forms of bad luck or may try to compensate those born with bad luck that cannot be alleviated efficiently in these other ways. These commitments on the part of a just society manifest the equal respect and concern for each person that is at the heart of egalitarian justice. The first main argument Anderson offers against luck egalitarianism is that its account of equality does not treat individuals with equal respect. The trouble arises when she considers what some luck egalitarians have said about bad option luck. Bad option luck is bad luck that occurs as a result of voluntarily incurred risks. Some luck egalitarians such as Eric Rakowski and Ronald Dworkin have argued that if a person takes a risk voluntarily which results in a bad outcome, then justice permits us to let the chips fall where they may. For example, if a motorist voluntarily does not buy auto insurance and gets into a terrible accident through her own fault, justice permits us to leave the motorist without helping her if she does not have the money to pay for medical care. This is a case of bad option luck since the motorist voluntarily took the risk of not having anyone come to her aid in the case of a disaster even though she had the resources and opportunities to buy insurance just like anyone else.

The main argument of the first section of the piece is that either luck egalitarians must allow the chips to fall where they may when individuals voluntarily incur risks or they are committed to treating them in a paternalistic way. Either way, she argues that one is failing to treat people with equal respect and concern. In the first horn of the dilemma, the person who loses from the risk taken is treated in a way that is incompatible with equal respect and concern because that person is abandoned to bad luck. Or the person is treated worse than other otherwise similarly placed individuals. The two problems can also be understood as the threshold problem and the discrimination problem. Anytime one allows a person to fall beneath a threshold of proper care one is treating them with less than equal concern and respect.

Notes on Option Luck
I found myself in agreement with Anderson's point that we ought not merely to let the chips fall where they may when individuals voluntarily take risks. At the same time I am not convinced the luck egalitarian need or ought to accept the distributive results of option luck, since it is not clear that these can be sufficiently morally distinguished from the results of brute luck. Suppose two people voluntarily take the same serious risks against a background of equality and one of them wins and the other loses. What is it that distinguishes these two people? What is it that accounts for the difference in their subsequent resource endowments? They made the same choices against the same background (let us suppose). They acted on the basis of the same preferences and with equal resources. The only thing that distinguishes them is luck. If we accept that luck egalitarianism has the implications that Dworkin, Rakowski and Anderson think it has, then on this account differences in luck alone are grounds for the differences in outcome. This would seem like a rather odd claim for someone who believes that differences in how well people do should not be based on morally arbitrary differences between the people. In addition it seems an odd claim for to make Dworkin since he asserts that differences in distribution ought to be sensitive to differences in ambition (or preference) and not to differences in circumstances.

Let us consider a few ways some egalitarians might defend the idea that option luck may legitimately have differential effects on people's lives. The first simply asserts that regardless of the impact on subsequent outcomes, one must live with the outcomes of one's choices because choice itself has the moral power to make the outcomes just at least as long as the conditions under which the choice was made are just. The mere fact of having chosen a course of action that produces an outcome makes that outcome just. This general claim does not seem to me to be compatible with an egalitarian position. If the conditions under which the choice is made are not egalitarian, then the choice cannot confer justice on the outcome. For example, if you and I choose the same profession and it turns out (previously unbeknownst to us) that you have a talent for it while I do not, the luck egalitarian will not endorse the unequal outcome merely because both courses of action were chosen. This is because an underlying inequality, that has not itself been chosen, has determined the outcome. If one denies this claim, then it is hard to see how one can maintain an egalitarian position. For the most part, individuals are quite uncertain of what talents they have in the marketplace. The great majority of choices of productive activity are made in the light of this uncertainty. If such choices can be thought of as voluntary assumptions of risk, then the so-called egalitarian will be committed to endorsing nearly all the inequalities that arise from the differential talents of the participants. The view in question nearly reduces to an equality of opportunity principle. If one rejects the implied claim, one must pull back from the idea that what one chooses under risk must legitimate the outcome.

Moreover, In these other cases, an underlying inequality that has not been chosen, determines the outcome.

This last argument suggests a dialectical argument against allowing option luck to determine distribution at least in general. The above case is not relevantly different from cases of differential option luck determining differences of distribution. In the above case, the source of the uncertainty is ignorance of talent and not ignorance of external facts (strictly speaking, economic conceptions of talent define it on the basis of external facts as well, namely supply and demand for abilities). If the inequality generated in this way cannot be legitimated through choice, how can inequalities generated through more garden-variety forms of risk taking be made just by choice? It is hard to see the principle that would help distinguish a paradigm case of the illegitimate impact of brute luck from clear cases of option luck. In fact one might ask why at least most cases of option luck aren't just intentional actions done in ignorance of brute luck effects. Suppose I decide to invest in a method of production and that method turns out to be, contrary to my conscientiously arrived at beliefs, not very reliable while you have invested in a highly reliable process without being any more conscientious. What has happened here is that I have had bad brute luck for me and you have had good luck. Only in a rather minimal sense in do you or I have control here. To ground differences in subsequent distribution on these differences in luck would appear to be unjust.

The second argument (offered by Dworkin) for allowing option luck to determine distribution is that if one requires that winners to distribute their winnings to losers then one has in effect deprived both of the opportunity to live their lives as they wish. This is because one has in effect made the risk taking impossible. Hence, both the winners and the losers have grounds for complaint. The suggestion is that in some way those who wish to take risks are not being accorded the same respect as those who do not wish to take risks. All are forced to live their lives according to the desires of the risk averse. Hence, there is a kind of inequality. The interests of some are being favored over those of others, at least if their preferences are reasonable indicators of their interests.

This second argument is not a bad one except that it does not support the general principle that states that one ought to let the chips fall where they may when people voluntarily incur risks. Risks can be assumed for a variety of reasons. Some assume risks because they like living dangerously. Indeed, there is a bit of that in all of us. We enjoy the excitement and stimulation of actions undertaken under uncertainty and risk. Some like this kind of thing a lot more than others do. Some enjoy gambling; others enjoy sky diving and bungy jumping. And there are many other kinds of risks people enjoy more or less for there own sakes. Dworkin's argument works for these cases reasonably well. If two people take the same risks at a gambling table and both know that if they lose, their losses will be compensated, they will indeed have been deprived of the pleasures of risk taking. As a consequence they will not be able to experience the kinds of pleasures that are important to them. It also makes sense to say of these people that if they lose then at least they have gained something in return, namely the risky pleasures.

Two questions remain about this kind of risk taking. What does one do with the winners? They receive both the benefit of the risk taking as well as the beneficial outcome. The second question is whether this supports a principle of letting the chips fall where they may in this kind of case. It is not clear that it does. We do not need to feel concerned if someone loses something as a result of taking a pleasurable risk, for that person has gained something as well. And presumably it was worth it. But presumably a person who gets into a motorcycle accident and is paralyzed from the neck down has not merely paid his due for the pleasure of taking the risk. It is hard to see how letting the chips fall where they may is just if there is a way of rectifying it.

In any case, the kind of risk taking mentioned above is not the usual case. Normally people take risks in the sense that there is a degree of inevitable ignorance regarding the outcomes of their actions. They do not court risk for its own sake; they accept it as an inevitable worry associated with their actions. They would just as soon not have to experience the risk at all. In general, if they lose, they usually think that they made a mistake. They would in no way be unhappy if the chance for making a mistake were diminished. And they would not lose anything if the consequences of risk taking were mitigated. In this kind of case, if one were to redistribute from winners to losers, the losers would not be deprived of something they wanted. And so the argument Dworkin suggests, even if it does work in part in cases of risk seeking activities, does not work in the standard cases of voluntary risk taking. A third argument for allowing the chips to fall where they may in cases of option luck has to do with the efficiency of allowing this in a society that encourages at least some markets. Surely a society that always redistributed from winners to losers would have a dampening effect on people's willingness to take the risks in the first place. To the extent that technological innovation occurs as a result of risk taking, constant redistribution of the proceeds of option luck would seriously slow down the productivity of some sectors of the society. Hence there is some reason for allowing winners to keep at least some of the proceeds of their good luck. In addition, it is important that people be able to learn from their mistakes as well as from their good fortune. This argument, however, does not show that one ought to let the chips fall where they may merely because individuals voluntarily incur risks. First it merely says that there ought to be limits on redistribution for the sake of the common good. It is no different from the idea that we might in some cases want to reward high talent in order to ensure higher productivity. Indeed, if my reasoning above is cogent, this shouldn't be a surprise. Second it does not say that the chips should fall where they may, it places limits on the kinds of differences that are allowed to arise from risk taking.

Hence, it is not clear to me that the dilemma that Anderson associates with luck egalitarianism is a real one. Luck egalitarians ought not to accept the principle that the chips fall where they may in cases of option luck.

Inferiority and Lack of Talent
Anderson critiques the luck egalitarian idea that a person is to be compensated for bad brute luck. The luck egalitarian idea is that if a person is born with an inferior resource endowment, then that person ought to be compensated since it is not by choice or preference that they have this inferior resource endowment. If one is born with a handicap or with very little talent, then one ought to receive some sort of compensation in the form of rectifying the handicap or of some other kind of resource.

Anderson critiques this on the ground that a luck egalitarian policy along these lines expresses the idea that the person being compensated is inferior and she says this is incompatible with equality. She says on this account: "People lay claim to the resources of egalitarian redistribution in virtue of their inferiority to others." (306) If you are to be compensated for having less talent, that implies that you are inferior in this respect and that you need to be compensated for that inferiority by those who have the superior talent. Hence, there is an implicit failure in the luck egalitarian scheme to recognize the equal worth of all individuals. Here I think that Anderson has miscast the egalitarian position. There are a number of ways in which the significance of claims about handicaps and inferior talents can be understood and Anderson seems to miss the distinctions here. One, one can say that the person is of inferior moral worth and so must be compensated and two, one can say that the person is of the same worth but has less with which to pursue a decent life. Presumably egalitarians have the latter claim in mind. This latter claim, however, is in no way incompatible with the idea of equality of respect and concern.

Anderson may have in mind that those who are handicapped are thought of not as having inferior worth but that they are thought of as having inherently a lesser capacity to have a decent life. (305) For example, she thinks that egalitarians are committed to the claim that being deaf is in and of itself worse than being able to hear. She seems to think that egalitarians must be committed to this claim in order to justify redistributive measures to the deaf such as establishing special means of communication in public facilities for the deaf. And this commitment to the inherent inferiority of the life of a deaf person would be insulting to them. Anderson then observes that deaf people deny that their lives are or need be inherently inferior in comparison to that of the hearing and seem to think that this idea is insulting to them (304). Certainly many people think this way about the lives of deaf people. But it is in no way required by egalitarian views that they think that the lives of the deaf are inherently inferior. All they need to do is suppose that in a society that is strongly coordinated on communication by sound, deaf people are at a disadvantage in the process of communication within that society. It is useful for societies to be coordinated in this kind of way. One can argue that the best way to benefit from coordination on the means of communication and to give deaf people an equal resource base with which to communicate, is to supply them with what is necessary to communicate with others. If one takes no stand on whether it is inherently better to be deaf or not one can still see the sense in which deaf people are worse off than others in a society like our own. It is here that the egalitarian can make a distinctive contribution to understanding the demands of justice regarding the deaf. Now we can understand the egalitarian impulse as saying that a particular highly efficient method of coordination has left a small minority with lesser resources and that this ought to be rectified. There is no implication here about the inferior moral worth of the deaf or their inherently inferior capacity to live a decent life. No humiliation is necessary. But this should give us the key to understanding the importance of redistribution to those who have less talent than others have at least in the sense that their talents are less productive in a market.

Moreover, it does seem that one can say that some people may conceivably be worse off than others because of the condition they are in. Individuals who have chronic pain are surely worse off than others are, even if they do have all the capacities and resources that others have. We can think that it is appropriate in these special cases to attempt either to lessen the pain or to compensate the persons with provision of some other good. None of these claims or policies in any way suggests that the person is morally inferior. And even though there is a sense in which we are saying that initially, the person's life is inherently worse than others' lives, there does not seem to be any insult in this observation.

Insurance, Paternalism and Justice
We discussed above the idea that in cases of voluntarily incurred risks, we do not accept the principle of letting the chips fall where they may. This can be seen in cases where two people undertake the same risks and experience the opposite outcomes. But we must ask what is to be done when individuals take different risks and some win while others lose; or some win and lose while others do not take any risk. Insurance is the most obvious case of this. Those who insure themselves are not willing to take certain risks and so they pay into a fund that will pay some of the costs if things go badly. They forgo certain benefits in order to lessen the chance of certain kinds of harms.

Anderson seems to think that the luck egalitarian is committed to allowing the chips fall where they may for those who fail to insure themselves. On her account, whether one has insured oneself or not ought not to make any difference at least for the more serious losses incurred. One has a duty to help people in serious need. It seems to me that we can see from the above arguments that the luck egalitarian is not committed to letting the chips fall where they may. On the other hand, I am not convinced by Anderson's approach either. There are two considerations that lead me to question her conclusion here. First, the person who has insured herself has made a genuine sacrifice for the sake of her well being in cases of bad luck. The person who has not insured himself has not made this sacrifice and has opted to use his resources in some other way. Both of these decisions can be perfectly prudent ones. But one has made a sacrifice of resources for the sake of greater security while the other has not. It is hard to see how this cannot make a difference for an egalitarian. The relation is similar to that between the person who gives up something in order to buy a desirable object and the person who receives the object without having to exchange anything. The second seems to be receiving more than the first. How can this be just?

A second consideration comes from the following example. Suppose that two people have fallen ill and the first person has insured herself while the second has not. Suppose further that the health of both persons is deteriorating rapidly and the longer one waits before treating the illness, the worse off each will be. Finally suppose that an ambulance can bring only one person to the hospital at a time for treatment. This implies that the first person will receive treatment sufficiently soon to avoid any future difficulties while the second person is in danger of dying and in any case is likely to experience permanent damage as a result of having to wait for the second ambulance run.

If both persons were insured then clearly we would have to flip a coin to determine who gets treatment first. And since we have rejected the principle that the chips should be allowed to fall where they may, the same would hold if both were uninsured. But it seems clear to me that it would be unjust to flip a coin in the case we are looking at. And it seems clearer that it would be unjust to take the uninsured person first.

Still, since we have rejected the principle that allows the chips to fall where they may, it is hard to see how the egalitarian is going to be able to allow the severely injured but uninsured person to go without help. This simply follows from the principle of equality and the rejection of the idea that luck alone can determine differences in how well people's lives go. There will be an egalitarian duty to help here and it will be a matter of justice. How extensive the duty is unclear. Hence, it is not clear that Anderson's view is all that different from the luck egalitarian view on this.

Perhaps it would be desirable to require that people take basic insurance for these kinds of basic harms. Is this paternalistic? It seems not, for two reasons. First, it can be perfectly prudent for a person not to take out insurance even on the possibility of basic harms, because of the low likelihood of harm and the great benefit of using the resources for other purposes. It would be foolish to think that one was always requiring people to insure themselves for their own good. Only a self-deceived paternalism could be in play here. Second, justice requires that we not allow the chips to fall where they may. And since some may wish to free ride on the dutiful actions of others when they get into trouble, we might prefer to make the insurance a collective good so that all are covered and all must pay in. In a sense, the insurance payment is part of doing one's own part in maintaining justice.

Here we can see that the luck egalitarian need not accept the characterization of paternalism when it requires that everyone have basic insurance. In a way, the luck egalitarian view seems to anticipate the views that Anderson expresses on this subject.