Health, Luck, and Justice (HLJ) is an ambitious monograph that articulates a sophisticated luck-egalitarian theory of distributive justice and applies it to questions concerning the distribution of healthcare and health within and across nations. At the same time, it criticizes the two main egalitarian alternatives, Daniels's fair equality of opportunity account, and Anderson's sketchier democratic equality account. In laying out a luck egalitarian approach so lucidly, Shlomi Segall shows us not only the advantages of the luck egalitarianism that he argues for, but also its serious drawbacks both as a general view of distributive justice and specifically as a guide to justice with respect to healthcare and health.

After a first chapter that lays out a version of luck egalitarianism and argues for its virtues as opposed to alternative versions of egalitarianism, HLJ is divided into three parts. This review will focus on the first two parts, which deal respectively with healthcare and health. In the third part, Segall argues for a cosmopolitan view of health in which the claims of poor health either on healthcare resources or on non-medical means of redressing inequalities are independent of one's national affiliation. He also rejects ‘devolution’ – that is permitting sub-national groupings control over their own per-capita share of the health budget.

In the introductory chapter of Health, Luck, and Justice, Segall argues that ‘luck egalitarians would typically say that the reversal of bad luck is the most radical way in which social injustice can be addressed, and therefore any theory that does not fully neutralize bad luck falls short of meeting the requirements of social justice’ (HLJ: 11). ‘Luck’ here is ‘brute luck’ – those things for which individuals are not responsible or over which they have no control. How to distinguish brute luck from so-called ‘option luck’ has been a controversial matter. Segall offers a novel construal (though he mentions that his view is related to views developed in Sandbu (Reference Sandbu2004) and Stemplowska (Reference Stemplowska2009)).

His view of luck egalitarianism is accordingly, ‘It is unjust for individuals to be worse off than others due to outcomes that it would have been unreasonable to expect them to avoid’ (HLJ: 13). As Segall notes, what is at issue is ‘not whether the individual has acted in a reasonable way, but rather whether it is unreasonable for society to expect the individual to avoid a certain course of action’ (HLJ: 20). Segall spells out what it is and is not reasonable to expect agents to avoid mainly via his applications of luck egalitarianism to distributional questions concerning healthcare. So, for example, he sees no distributive injustice if reckless drivers are worse off due to car accidents. It is, in contrast, unjust when careful drivers are worse off as the result of car accidents. Other cases are more difficult. For example, whether it is reasonable to expect someone who contracts lung cancer as a result of working in an asbestos factory to have avoided the outcome depends presumably on what the agent knew and what alternatives were available. Segall's way of distinguishing between brute and option luck has the effect of reclassifying as brute luck a great deal of what Dworkin, for example, would have regarded as option luck (Reference Dworkin1981). For example, we cannot reasonably expect those who take significant but not foolhardy risks in starting new businesses and who wind up bankrupt in consequence not to have taken those risks. In my view, this reclassification of option luck as brute luck (which Segall does not discuss) is a welcome consequence of his revised version of luck egalitarianism.

As Elizabeth Anderson has pointed out with particular emphasis, luck egalitarianism apparently implies that there is nothing unjust about leaving the reckless driver bleeding by the road or refusing to help a needy lung cancer patient who has been a heavy smoker (Reference Anderson1999). This implication seems discordant with the impulses that drive egalitarians, and it is directly antithetical to Segall's view that justice requires universal healthcare. Accordingly in Part I, Segall attempts to show that ‘luck egalitarian justice can escape the abandonment objection and justify universal and unconditional healthcare’ (HLJ: 29).

The first step in Segall's response is to assert that luck egalitarians need not find anything unjust about equalities for which individuals are not responsible. In Segall's view, inequalities resulting from luck are unjust, while equalities that are due to luck are not. So even though luck egalitarianism by itself does not require treating the reckless driver, it permits him to be treated. From a luck egalitarian perspective, there is no injustice in doing so. That means that one can supplement luck egalitarianism with other principles, including other principles of justice, which do require one to treat the reckless driver. The other principle Segall espouses requires that society, to the extent to which it is able, meet everyone's basic needs. Since there is nothing anti-egalitarian about this principle, and satisfying it does not conflict with the requirements of luck egalitarianism, Segall regards the combined theory as a version of luck egalitarianism that meets the abandonment objection.

I've got some doubts. If one believes that ‘any theory that does not fully neutralize bad luck falls short of meeting the requirements of social justice’ (HLJ: 11), then it seems that one should object to a situation in which hard-working Ann is, as the result of bad brute luck, no better off than lazy Andy. In effect, Segall is arguing that luck egalitarianism is exhausted by what Fleurbaey calls ‘the compensation principle’, which says that differences in outcomes should not depend on anything other than factors for which an individual is responsible, and that luck egalitarians should not be committed to what Fleurbaey calls ‘the liberal reward principle’, which says that outcomes should depend on factors for which an individual is responsible. Speaking of ‘neutralizing brute luck’ suggests both of these. Furthermore, in a large population, if hard-working Ann is no better off than lazy Andy, then it is bound to be the case either that Ann is worse off than some equally hard-working Amy, or Andy is better off than some equally lazy Arnold.

Segall denies that what motivates a concern with inequalities due to brute luck also motivates a concern about equalities due to brute luck. What tempts us to think otherwise, he argues, is that we confuse luck egalitarianism with the view that individuals should get what they deserve. This complaint appears to be true of Larry Temkin, who denies that equality is, prima facie, good when there are differences in what people deserve (Hausman and Waldren forthcoming). Temkin writes, ‘In fact, I think that deserved inequalities . . . are not bad at all. The reason for this is simple. Undeserved inequality is unfair, but deserved inequality is not’ (Reference Temkin2003, p. 767; see also Temkin Reference Temkin1993: 138–140 and Kagan Reference Kagan, Pojman and McLeod1999). In Segall's view, once one carefully distinguishes the specifically egalitarian concerns that drive luck egalitarianism from concerns about desert, one can see that equalities are never of concern from the perspective of distributive justice, only inequalities.

What are the specifically egalitarian concerns that motivate a luck egalitarian? What is bad about inequality? One answer, which is not Segall's, is that inequalities often have considerable instrumental importance. ‘Such instrumental value (of equality) includes avoiding unacceptable forms of power and control, preventing stigma through differences in status, preventing harm to individuals’ self-esteem, removing impediments to community and fraternity, and facilitating and maintaining democratic institutions’ (HLJ: 113–114). Segall here cites work of Scanlon (Reference Scanlon2003) and Miller (Reference Miller1982). In Chapter 11 of Hausman and McPherson (Reference Hausman and McPherson2006) and in Hausman and Waldren (forthcoming), I argue (following Miller and Scanlon) that these ends that equality promotes are themselves egalitarian concerns (see Hausman and McPherson Reference Hausman and McPherson2006: chapter 11; Hausman and Waldren (forthcoming); O'Neill (Reference O'Neill2008) as well as Miller and Scanlon). But if these further consequences constitute what is wrong with inequalities, then equalities that are due to brute luck should also be condemned, because these can also harm self-esteem and create impediments to community and fraternity.

What seems to me to drive Segall's luck egalitarianism is an objection to arbitrary differences. When a tornado destroys a family's house, if the result is an inequality in wellbeing, then that result is, in Segall's view unjust. Why? It is awful that their house has been destroyed, but what is unjust about it? There are two possible answers here. The first is that it is unjust, because nature's action was unjust. Nature did not behave impartially. The resulting inequality has no justification. Such a view strikes me as implausible. Can one hold nature to standards of justice? Neither the actions of nature nor their results are justified or unjustified. Indeed, it seems to me presumptuous to suppose that it is up to us to pass judgement.

Moreover, the whole idea of subjecting nature to standards of justice is unnecessary to defend the luck egalitarian's view that the unfortunate family whose only home is demolished should be compensated. Rather than maintaining that they should be compensated because the tornado caused an unjust distribution, one can say that the failure of the society to remedy the inequality would create an injustice. Without holding that natural inequalities are just or unjust, luck egalitarians can maintain that they should not be tolerated by a just society. When the tornado collapses the house, nothing unjust occurs. But if neighbours or government agencies provide no assistance, then there is an injustice.

This variant of luck egalitarianism, which one can attribute to Segall, has the same implications in practice as the view of cosmic injustice defended by authors such as Larry Temkin, who writes, for example, ‘John's situation [he's been hit by a falling tree limb] strikes me as unfair. He has been the victim of terrible, cosmic, bad luck’ (Reference Temkin2003: 773). Where the two views differ is when there is nothing to be done. Luck egalitarians such as Temkin would say that if the tornado had killed the family, then there would be an even worse injustice, although there would be no way to neutralize the bad brute luck. According to the less extravagant alternative I am sketching here and attributing to Segall, the deaths of the family members would be tragic, but not in any way unjust. With respect to any issue where action is possible, the two views support the same conclusions.

If Segall accepts the view that attaches injustice exclusively to human actions or inactions and their outcomes, then he can provide a reasonable justification for his claim that inequalities that result from brute luck demand compensation, while equalities do not. Instead of regarding human society as a cosmic purifier charged with neutralizing brute luck, he can regard societies as charged with maintaining equality, except where inequality results from individual's choices. (But this interpretation might conflict with the cosmopolitanism that Segall defends in chapter 11.) Equalities do not need to be monitored or fixed, only inequalities. What is at issue is not neutralizing luck, but preserving equality, except when individual choices create inequality.

So, Segall argues, the luck egalitarian can respond to the abandonment objection by supplementing the luck egalitarian condemnation of inequalities due to brute luck with a compatible commitment to meeting everyone's basic needs. What then is his luck-egalitarian case for universal healthcare? It cannot plausibly rest on the claim that there should be no unchosen differences between individual lives. A world of such uniformity is thoroughly unappetizing and in any case impossible. The luck egalitarian seeks to neutralize the influence of luck on people's overall prospects or their overall welfare, not on their prospects for success in specific life plans. A luck egalitarian would for example have few concerns about a world in which Mary is a talented mathematician who cannot carry a tune while Martha is a superb violinist who has troubles with the multiplication table. Distributive justice does not require compensatory music lessons for Mary and math drill for Martha. The case for universal healthcare does not and cannot rest on the view that unchosen differences in specific outcomes are unjust. It must rest on the relevance of remediable health differences for some overall measure of welfare or prospects. Yet, as the controversies of the 1980s and 1990s showed, it is not easy to say exactly what overall measure luck egalitarians should seek to equalize. If they focus on welfare or on opportunity for welfare, then Tiny Tim, with his sunny disposition, may be better off without a wheelchair than many who are completely healthy (Cohen Reference Cohen1989: 918).

A luck egalitarian might be able to make the case that as a practical matter, health differences due to brute luck are almost always unjust. Health is so crucial to overall prospects that equalizing overall prospects may require roughly equalizing health. If universal healthcare is in turn crucial to equalizing health, then perhaps one has a case for universal healthcare. To make this case, one must say something about what would constitute equal health and one must make the case that healthcare is crucial to equalizing health. Both of these tasks are problematic. The notion of ‘equal health’ is obviously tricky. In one sense a robust 80-year-old and a frail 30-year-old may be equally healthy, but that is not the relevant sense. Nor should a luck egalitarian be content, as Segall is (HLJ: 97), with equal healthy life expectancy, because equal health expectancies do not neutralize luck. The relevant notion of equal health for a luck egalitarian is presumably equal lifetime health, or equal health in each life stage. Moreover, as Segall acknowledges (HLJ: 90), healthcare in fact makes a relatively minor contribution to overall health. It would appear that the luck egalitarian should be much more concerned with improving education, lessening crime, or strengthening families than with universal healthcare.

Why then favour ‘unconditional universal healthcare’? Segall's answer is that medical needs are basic needs, and his luck egalitarianism is supplemented with a commitment to meeting basic needs. This is a cogent position, but notice that the tail is wagging the dog: the luck egalitarianism in Segall's position is not doing any of the work. (For a truly luck-egalitarian argument for universal healthcare, see Legrand Reference Legrand1990: chapter 7.)

In justifying universal healthcare by the requirement that society meet people's basic needs, Segall also has a ready answer to Dworkin's case (Reference Dworkin2000: 61) of Roger, who needs an expensive wheelchair, but would rather have a Stradivarius that is no more expensive (HLJ: 84). Yet Segall is not content with this answer. To justify the in-kind care, Segall argues

I do not think that this passage is defensible. Luck egalitarians are concerned with eliminating unequal life prospects or unequal welfare for which individuals are not responsible. Nothing in luck egalitarianism implies that if the cause of the inequality is a health deficit, then the solution, where possible, should be medical care. Nothing within luck egalitarianism favours providing the wheelchair. Focusing on opportunity rather than outcomes is of no help. Both the wheelchair and the Stradivarius contribute to Roger's welfare only insofar as he makes use of them. They both improve his opportunity for welfare rather than directly improving his welfare. In any case, the whole discussion is unnecessary, since what is doing the work in Segall's account is the requirement that society meet people's basic needs; and Roger has no basic need for the Stradivarius.

When Segall turns from healthcare to health, he trades in luck egalitarianism for luck prioritarianism. He abandons luck egalitarianism, because he denies that levelling down to lessen differences in health promotes distributive justice. Segall does not maintain that levelling down never improves distributive justice, however much it may be undesirable on other moral grounds. But he finds levelling down with respect to health particularly unpalatable, because in the case of health, equality has essentially no instrumental value. Accordingly, he maintains that ‘it is permissible to abandon the quest for equality in health in favor of priority to the worse off’ (HLJ: 112). In chapter 9, Segall argues, in contrast, that we should be luck egalitarians rather that luck prioritarians with respect to the distribution of health enhancements, on the grounds that the distribution of enhancements, unlike the distribution of health is important instrumentally. The argument for a luck prioritarian approach to health inequalities strikes me as odd. First of all, it apparently places more weight on the instrumental value of equality than Segall in fact does. He is wise not to defend equality instrumentally, because if what makes one an egalitarian are the commitments to non-subjugation, mutual respect, and so forth to which equality contributes, it is hard to justify luck egalitarianism (Hausman and Waldren forthcoming). In addition, the vices of luck egalitarianism are no argument for the virtues of luck prioritarianism. What is the case for luck prioritarianism?

Segall formulates his luck prioritarianism with respect to health as follows:

Segall intends a strict lexical priority: ‘the luck prioritarian ideal tells us to compare levels of prudence first, and use the severity of the medical condition only as a tie-breaker between those who were equally prudent in looking after their health’ (HLJ: 119). In principle (though this may not be feasible in practice) we should determine each individual's health-prudence level and address the health shortfalls of the most prudent first. The less healthy among those who have been equally prudent should then get priority. One would have thought that the priorities should be reversed. Shouldn't the first concern of the prioritarian be to improve the health of the least healthy rather than to improve the health of the most prudent?

Consider the following schematic example. Suppose there are three individuals, Amy, Bill and Cathy. Their respective health levels (after their medical needs have been met as fully as possible) are .90, .85, and .60 on a scale of 0 for death and 1 for full health. Bill and Amy have been equally and highly prudent. Cathy's poor health is due in part to her imprudence. Other non-medical interventions to improve people's health, such as education, family therapy, job counselling and so forth are available. Suppose that for each $1000 one invests in these other medical interventions to improve Amy's health one gets a .02 improvement in health. For each $1000 invested in Bill's health one gets a .01 improvement in health, and for each $1000 invested in Cathy's health one gets a .04 improvement in health. Finally suppose one has $8000 to devote to improving people's health. Since Amy and Bill have been the most responsible and all the available funds can be spent on improving their health, Segall's principle recommends that we invest $7000 of the $8000 in improving Bill's health and $1000 in improving Amy's health, bringing them both up to .92. The result of applying Segall's principle is less equality than one would have achieved if one had aimed simply at maximizing health. To diminish the apparent implausibility of this result, Segall emphasizes that his principle of prioritizing the opportunity for health of the worse off is not meant to govern health policy all by itself (HLJ: 119). Other moral concerns, such as benevolence or efficiency also influence policy. The principle is meant only to spell out what distributive justice requires. Even with this caveat, one is left with doubts about both whether this principle captures what luck egalitarianism implies and, if it does, whether luck egalitarianism captures what egalitarians care about.

On Segall's view, the poor health of the less prudent generates claims for help only if resources are left over after doing everything possible for those who have taken good care of themselves. Since those who take care of themselves will on average be healthier than those who are less careful of their health, Segall's principle of ‘prioritizing the opportunity for health of the worse off’ will in fact prioritize improving the health of the healthier. This is an awkward bottom line for an egalitarian.

Though this review has canvassed some of the criticisms to which Segall's account appears to be subject, I hope it has also made clear how ambitious, rich, and challenging that account is. It is bound to provoke much more extensive discussion.