Well-Being and Fairness in the Distribution of Scarce Health Resources

Transcription

1 Journal of Medicine and Philosophy ISSN: (Print) (Online) Journal homepage: Well-Being and Fairness in the Distribution of Scarce Health Resources RE'EM SEGEV To cite this article: RE'EM SEGEV (2005) Well-Being and Fairness in the Distribution of Scarce Health Resources, Journal of Medicine and Philosophy, 30:3, To link to this article: Published online: 19 Aug Submit your article to this journal Article views: 536 Citing articles: 1 View citing articles Full Terms & Conditions of access and use can be found at

2 Journal of Medicine and Philosophy, 30: , 2005 Copyright Taylor & Francis, Inc. ISSN: print DOI: / Well-Being and Fairness in the Distribution of Scarce Health Resources NJMP Journal of Medicine and Philosophy, Vol. 30, No. 03, April 2005, pp. 0 0 Well-Being R. Segev and Fairness RE EM SEGEV Tel-Aviv University and Haifa University, Israel Based on a general thesis regarding the proper resolution of interpersonal conflicts, this paper suggests a normative framework for the distribution of scarce health resources. The proposed thesis includes two basic ideas. First, individual well-being is the fundamental value. Second, interpersonal conflicts affecting well-being should be resolved in light of several conceptions of fairness, reflecting the independent value of persons and the moral significance of responsibility of individuals for the existence of interpersonal conflicts. These ideas are elaborated in several principles that are applied with respect to the distribution of scarce health resources. Keywords: distribution, equal chance, fairness, interpersonal conflicts, responsibility, scarce health resources, well-being I. INTRODUCTION A fundamental normative problem concerns the resolution of interpersonal conflicts between the interests of individuals. An important kind of interpersonal conflict, involving basic aspects of well-being, mainly life and bodily integrity, arises with respect to the distribution of scarce health resources including medical equipment, organs for transplant, and attention from health-care agents. How should such resources be distributed among the individuals who could benefit from them? In this article, I argue that the answer to this question should be derived from a general thesis regarding the resolution of interpersonal conflicts, based on considerations of wellbeing and fairness. 1 I further argue that the proposed thesis constitutes a Address correspondence to: Re em Segev, Ph.D., 38 Hermon St., P.O.B , Mevaseret-Zion, 90805, Israel. rmsegev@yahoo.com 231

3 232 R. Segev plausible alternative to familiar moral theories, as well as to common, and sometimes legally formulated, practices, particularly with regard to the distribution of scarce health resources. The resolution of interpersonal conflicts raises many complex normative questions, most of which are relevant, in particular, with respect to the distribution of scarce health resources. This article focuses on the issues that are most relevant for the presentation of the main aspects of the proposed theory. It thus avoids certain important issues regarding the resolution of interpersonal conflicts, which although relevant also in the context of allocating health resources, do not raise special queries in this context. Particularly, the following general limitations should be mentioned at the outset. First, I consider the question of how to allocate health resources given scarcity (of a certain health resource), while avoiding the question of whether, and to what extent, more health resources should be produced in order to diminish scarcity. These questions are distinct, although they are of course related in various respects, 2 especially since the problem of scarce health resources might, and probably should, be significantly alleviated, though presumably not completely avoided, if more general resources would be directed to health care, or to the prevention of health problems, at the expense of other, less important, goals. 3 Second, like most discussions of the subject, I explore the question of which of several persons to prefer in distributing health resources, while avoiding the question of whether it is justified to positively harm one person in order to benefit another (e.g., whether it is justified to take the organs of one person and transplant them in another, thereby killing the first and saving the life of the second). 4 The latter question raises large issues that are beyond the scope of this article, particularly the moral significance of luck ( moral luck ) and the validity of deontological and rights-based agentrelative constraints on positive, as opposed to negative, agency (including, for example, the distinctions between doing harm and allowing harm or between intending harm and foreseeing harm). 5 Finally, the article is limited to cases in which only one of several persons could be helped in order to avoid the general question regarding the moral significance of the number of persons involved in interpersonal conflicts. 6,7 The article thus focuses, in light of these limitations, on given conflicts between two persons, each of whom could benefit from an available, but scarce, health resource. This limited focus sidesteps many difficult aspects of health resource allocation. The aim of this article is accordingly not to provide a comprehensive theory that could entail specific and final conclusions with respect to all aspects of interpersonal conflicts that are pertinent to health care allocation. 8 Rather, the more modest aim of this article encompasses two related aspects. The first is to emphasize the relation between interpersonal conflicts in general and the allocation of scarce health resources in particular. And the second is to present a thesis that

4 Well-Being and Fairness 233 identifies the main normative factors and suggests a basic structure concerning their interaction, with regard to a basic and important aspect of the general issue of interpersonal conflicts, which, I believe, it is especially beneficial to consider in the context of health resources allocation. The analysis of various concepts used in the article, such as impartiality, wellbeing, equality, and responsibility, is accordingly limited to the rudimentary level required for this purpose. II. THE GENERAL THESIS The general thesis I suggest for the resolution of interpersonal conflicts reflects two basic ideas. First, the fundamental value and source for reasons for actions, in the context of interpersonal conflicts, is individual well-being. Second, the resolution of interpersonal conflicts in which there is an internal clash, within the first idea, between reasons for action to protect the well-being of different persons should be done in light of several conceptions of fairness, which reflect the intrinsic and independent value of each person and the rational power of persons. In other words, according to the proposed thesis, individual well-being is what we should fundamentally care about, whereas fairness is concerned with how we should care for this value when we are confronted by interpersonal conflicts. It is important to emphasize that, according to the proposed thesis, the concepts of individual well-being and fairness do not represent competing fundamental values: the idea that interpersonal conflicts should be resolved fairly completes the idea that individual well-being is the fundamental value, rather than constitutes a contrasting ideas regarding the fundamental value. This feature becomes clear once we notice that the concept of fairness becomes applicable only when there is an internal clash within the concept of individual well-being. (As I will explain below, there could be, however, a clash between different notions of fairness, particularly a conception of impartiality, which entails equality, and a conception of priority for the more important interest. This last conception is sometimes described as based on a consideration of individual well-being, but it is in fact based on a conception of fairness in resolving interpersonal conflicts.) These two basic ideas are elaborated in several principles: 1. The Impartiality Principle: reasons for actions are agent-neutral, rather than agent-relative, namely, apply to all agents equally; particularly, reasons for actions should be evaluated from an impartial perspective, rather from an agent-relative perspective that accords special weight to the personal aspects of agents life. 2. The Well-Being Principle (WBP): there is a reason to protect and enhance the well-being of persons.

5 234 R. Segev 3. The Equal Chance Principle (ECP): in resolving interpersonal conflicts of well-being, there is a reason to accord equal weight to the well-being of each person, by following two hierarchal sub-principles (a) there is a reason to distribute benefits or inevitable costs between all persons equally, so that each would get the maximum possible (roughly) equal benefit or bear the minimum possible (roughly) equal loss provided that the benefit or reduction of cost for each person is significant; or (b) when this is impossible, there is a reason to give each person the highest possible equal chance to be preferred. 4. The Importance Principle (IP): the strength of the reason provided by the WBP depends on the importance of the interest at stake and the conjectured probabilities 9 concerning the possible effects of the considered action or inaction on it (positively or negatively). Assuming equal probabilities, the more important is the interest, the stronger is the reason to protect it. In resolving interpersonal conflicts, there is, therefore, a reason to prefer the person who would otherwise suffer the most severe harm or the person who could be benefited most significantly. 5. The Substantial Difference Principle (SDP): the reason provided by the IP prevails over the reason provided by the ECP if, assuming that all relevant probabilities are equal, there is a substantial gap in the importance of the competing interests. 6. The Principle of Fairness: Responsibility (PFR): when an interpersonal conflict requires a choice between the well-being of individuals, there is a reason to prefer a person who is not responsible for the existence of the conflict to a person who is and a person who is less responsible to a person who is more responsible. The resolution of clashes between the reason provided by the PFR and the reasons provided by the previous principles depends on the relative force of the clashing reasons, namely, the importance of the competing interests, the relevant probabilities and the degree of responsibility of each person for the existence of the conflict. My general claim is that these principles provide a more plausible alternative to common moral theories: on the one hand, deontological theories and theories of rights and, on the other hand, common, aggregative forms of consequentialism, especially utilitarianism. I will not repeat here my attempt to defend these principles in their general form. Rather, my aim in this article is to reinforce this defense by demonstrating that the proposed principles offer a plausible framework with regard to important aspects of the distribution of scarce health resources. In this context, I believe that the proposed thesis provides a plausible alternative to a number of theories regarding the proper allocation of health resources, particularly various forms of aggregate consequentialism (Savulescu, 1998; 1999), most notably utilitarianism (McKie, Richardson,

6 Well-Being and Fairness 235 Singer, & Kushse, 1998; Stein, 2001; 2002a; 2002b); equal opportunities theories (Daniels, 1981; 2002; Harris, 1996; 1999; Cohen & Burg, 2003); theories that combine utilitarian with deontological and justice-based considerations (Veatch, 2000, pp , ) 10 or take account of diverse considerations, such as likelihood of success, life expectancy, familial role and past and future contribution to society (Rescher, 1969); hypothetical decision theories (Dworkin, 1993); and autonomy-based theories, which enable individuals to choose among various health programs, each of which determines its priorities and allocation scheme with respect to health resources (Emanuel, 1991, Ch. 6; Elhauge, 1994). A comprehensive analysis of the differences between the proposed thesis and these theories involves many controversies that are beyond the scope of this article. However, it is important to note briefly at least the main general differences between the proposed thesis and other notable moral theories (more specific differences are discussed throughout the article in the appropriate places). The proposed theory differs from standard deontological and rightsbased theories in two main respects. First, the suggested principles all reflect agent-neutral reasons for actions, whereas standard deontological and rights-based theories assert agent-relative constraints. Thus, according to the proposed theory, if there is a conclusive reason to follow one of its principles in one case, but doing so would lead another person to violate another principle that is conclusive in another case, the first agent should not give precedence to following the principle that applies to his or her case, but rather consider the force of the reasons conveyed by the relevant principles in each case and decide which it is more important to follow. For example, suppose that, in one case, an agent has a conclusive reason to follow the ECP but this reason is relatively weak (since the competing interests are trivial) and doing this would lead another person to violate the SDP in another case, in which this principle is especially strong (since the gap in the importance of the competing interests is huge). According to the proposed theory, the agent should not follow the ECP in his or her case so that the SDP would be followed in the other case. Second, the proposed thesis considers individual well-being as the fundamental value. According to the proposed thesis, this fundamental value should be construed and supplemented in interpersonal conflicts, in light of the notion of fairness, in order to adjudicate internal clashes between different aspects of this value. However, contrary to deontological and rightsbased constraints, the proposed theory does not substitute or limit the value of individual well-being in light of other, competing, fundamental values, such as duties or rights. 11 The proposed theory also differs from standard consequential theories, particularly utilitarianism, in three main respects. First, according to the proposed theory, the right action is the one that best reflects all pertinent reasons for action, whereas consequential theories see the right action as the

7 236 R. Segev one that leads to the best possible overall state of affairs. These views might be reconciled, either by understanding the notion of states of affairs in terms of reasons for actions or by rephrasing my conclusions in terms of reasons to promote states of affairs. However, I believe that the notion of reasons for actions is preferable since it emphasizes the fact that what fundamentally matters is the fate of individuals and there is no common denominator that enables tradeoffs between (interests of) individuals. Second, and more importantly, the proposed theory considers individual well-being as the fundamental, but not the only, value that generates reasons for action and relies, in particular, also on the value of fairness, which includes both distributive and backward-looking considerations, such as the ECP and the PFR, respectively. In this respect, the proposed theory is different from common forms of consequentialism, and especially utilitarianism, which are often described as monistic (reducible to one normative factor) and which do not attach independent value to considerations of fairness. 12 Finally, the proposed theory is not aggregative, at least not in the sweeping form of utilitarian aggregation, which lumps together all aspects of different individuals well-being within one aggregative balance and thus completely disregards the nature of the interests involved. Since this article focuses on conflicts involving only one person on each side, the pertinent question in this respect is whether the general balance of well-being is morally significant in itself. The proposed theory answers this question in the negative. Despite these differences from common forms of consequentialism, however, the proposed theory is nevertheless essentially consequential in one important respect, namely, in that it considers individual well-being to be the fundamental value and source of reasons for actions, which is supplemented, but not limited, by notions of fairness. This is not a trivial sense. Every moral theory must go beyond the idea that some value or good matters in order to resolve internal clashes within this idea, such as those that emerge in interpersonal conflicts, including those concerning the allocation of limited resources. The utilitarian solution is the maximization of the aggregate balance of the relevant good (well-being). The proposed principles reflect an alternative, non-aggregative, solution, based on the concept of fairness. The proposed thesis is also different from common theories of justice in the way it combines considerations of individual well-being and fairness. First, the proposed theory assigns a central role for considerations of individual well-being: it considers individual well-being as the fundamental value in the sense explained above. This view is at odds, for example, with Thomas Scanlon s claim that individual well-being is not a central value (Scanlon, 1998, Ch. 3). Second, the proposed theory considers the notion of equality (equal shares or chances) as entailed by an impartial concern for the well-being of each person rather than by the view that equality is morally important in

8 Well-Being and Fairness 237 itself. In my opinion, the relative position of individuals, including in terms of equality, is not important in itself. 13 Equality is sometimes the right solution since, and when, the relevant reasons apply to all relevant individuals in the same way. (Of course, there might occasionally be various instrumental reasons for equality, in certain contexts, but these are not unique to egalitarian theories.) Third, the proposed thesis takes into account two additional considerations, which might justify a rejection of the starting point of equality the relative importance of the interests at stake and the responsibility of persons for the existence of interpersonal conflicts and suggests the SDP and the PFR with regard to their interaction with the notion of equality adopted in the ECP. This framework is different, for example, from Rawls principles of justice. For instance, the SDP differs from Rawls Difference Principle, which justifies deviations from (economic) equality (only) when it is for the benefit of the worst off. 14 Unlike the Difference Principle, the SDP takes account of all the persons who would be affected by the considered act (not just the least advantaged) and justifies deviation from equality only when there is a substantial gap in the importance of the competing interests (rather than any advantage for the worst off). Finally, the theory I suggest is not contractual, contrary to influential moral theories in contemporary thought (mainly Rawls, 1971; and Scanlon, 1998). (Of course, contractual theories can provide a framework for various considerations, depending on how they are elaborated.) I reach my conclusions based on what I consider the best reflection of valid reasons for action, rather than on contractual ideas such as Rawls original position (Rawls, 1971, Ch. 3) or Scanlon s justifiability to others test (Scanlon, 1998, pp. 4 5 and Ch. 5) (although I believe that the possibility of rationally justifying normative conclusions to the individuals who are affected by them is an important indication for the validity of these conclusions, I do not consider this test to be a substantive contractual standard). 15 Some of the general views reflected in the proposed principles are widely accepted, particularly in the context of allocating health resources. First, the Impartiality Principle, although generally controversial, is widely accepted with respect to the distribution of scarce health resources, since typically those who make the relevant decisions have a special obligation, derived from their role, to be impartial, especially with respect to the allocation of public resources and when they do not have a personal stake in the matter. To be sure, it is often argued that physicians have an agent-relative duty to prefer their patients in allocating attention and tangible resources. I do not think that this view is generally convincing (although there might be practical considerations that support a recommendation that health care professional should concentrate on patients they are familiar with when other things are equal). But, even if it were, it is implausible when formulating a theory of public resource allocation, as is often the case.

9 238 R. Segev Second, although the WBP raises various queries and disputes concerning its content, scope and place within a moral theory, there is a wide agreement concerning its hard core, namely, the idea that individual wellbeing is a value and that the effects of actions on central and basic aspects of the well-being of individuals, such as life and bodily integrity, should be considered. 16 This idea is therefore especially valid and accepted in the context of the allocation of health resources in which major aspects of wellbeing are at stake. My main concern in this article is with the interaction between the WBP and other values and corresponding reasons for actions, particularly those based on the notions of fairness outlined above, with respect to the allocation of health resources. This is the subject of the remainder of the article. III. THE EQUAL CHANCE PRINCIPLE The ECP addresses clashes of reasons generated by the WBP, in light of what I consider a conception of fairness. This conception reflects several related truths, which are especially conspicuous when important interests are at stake, as is often the case when distributing scarce health resources. First, each person has an independent value: the fate of each person as an individual matters. Second, there is no common denominator in light of which interests of different persons could be traded off without a loss. 17 Third, for each of the persons involved in an interpersonal conflict, its resolution is a question of all or nothing as far as the interest at stake is concerned and literally when it is life. This last general feature of interpersonal conflicts requires further clarification, particularly with respect to health resources allocation, since, at first sight, it might not seem accurate in this context, for two reasons. First, a person who does not get a certain health resource (since it is given to another) could often get another later. For example, a patient who was denied an organ for transplant might get the next available organ; and a patient who was denied the immediate attention of a health care professional might get the attention a few minutes later. Second, it is often possible to split a resource and give each patient some of it. For example, when two patients suffer severe pain and there is not enough morphine to fully relieve the pain of both (assuming that if all or most of the morphine is given to one it would fully relieve his or her pain but the other would continue to suffer severe pain), it might be possible to give each patient enough of the morphine to relieve most of the pain of each. However, these correct observations do not undermine the general claim that the resolution of each conflict is, for each of the patients involved, a question of all or nothing with respect to the pertinent interest.

10 Well-Being and Fairness 239 To be sure, this interest might be trivial. For example, when two persons suffer mild pain and there is only one doctor or nurse who can administer a painkiller, the one who would get the painkiller second would suffer only a few minutes of additional pain. Nevertheless, this observation only shows that, in this example, the interest at stake (i.e., avoiding a few minutes of mild pain) was trivial. It does not undermine the claim that, with respect to the relevant interest (in the above example, the avoidance of a few minutes of mild pain), the conflict indeed was a question of all or nothing. This feature is also present in the previous example in which the quantity of morphine is not enough to fully relieve the pain of either of the two patients who suffer, but is sufficient to relieve most of the pain of both by giving half of the morphine to each. In this example, we have to assume that if all or most of the morphine is given to one patient it would fully relieve his or her pain, but then the other patient would continue to suffer a severe pain. (If this assumption does not hold, namely, if giving more to one would not fully relieve her pain, then there is no interpersonal conflict.) However, when this assumption holds, it is a question of all or nothing with respect to the pain of each patient that it is possible to relieve, but which is not relieved (since the morphine, or part of it, is given to the other patient). The fact that it is possible to relieve most of the pain of both patients only means that the conflict is more limited, and perhaps even trivial (if the remaining pain is indeed very mild), but not that the conflict does not exist. After all, it is, by hypothesis, possible to relieve the remainder of the pain of each and either this is done with respect to each (by depriving the other of the morphine completely) or not. This general feature becomes significant, as noted above, when the interest at stake is important. For example, when two persons who are in an advanced stage of a major organ failure compete for a certain organ, not getting the organ might mean a loss of a substantial chance of getting an organ in time and, consequently, a loss of a substantial chance of living an additional significant period of time or of significantly improving quality of life. To be sure, in such cases there is typically still some chance of getting another organ in time. However, this does not undermine the claim that the chance was diminished; namely, that, with respect to this (potentially very important) chance, the resolution of the conflict was indeed a matter of all or nothing. These observations entail, in my opinion, a basic standard of according equal weight to the well-being of every person who might benefit from an available health resource. This basic standard should, I suggest, be elaborated in light of two hierarchical sub-principles. The best option is to distribute resources so that each person who might benefit from them would get the maximum possible (roughly) equal benefit or bear the minimum possible (roughly) equal loss provided that the benefit conferred or burden avoided, with respect to each person, is significant. For example, if each of

11 240 R. Segev several persons would benefit from the use of one dialysis machine, in proportion to the time it is available to each (until a certain limit), then the use of the machine should be divided so that each would get an equal but significant benefit from it. The next sub-principle applies when it is impossible to distribute health resources in a way that would benefit each person to a significant and roughly equal degree. In this case, I suggest that each person should be accorded the highest possible equal chance to get the relevant resource, for example, by tossing a coin. The concept of equal benefits, and consequently of equal chance to benefit, raises the question of how to elaborate the notion of equality with respect to its subject, namely, what should be equalized. One interpretation is forward-looking and refers to the magnitude of the effects of actions on the well-being of persons, namely, to the size of the potential contribution or loss for each individual s well-being. According to this interpretation, a scarce resource should be distributed in a way that it would contribute to each person s well-being equally. Another interpretation is backwardlooking and refers to the overall state of well-being of individuals taking into account all factors that affect the extent to which a person s life is good or bad, whether presently or throughout a person s life. According to this interpretation, a scarce resource should be distributed in a way that would bring all persons to the same (or to the closest possible) overall state of well-being. In this respect, this interpretation favors persons who are (relatively) worse off to persons who are (relatively) better off (not because the relative positions of persons itself is morally significant, but rather because we have a stronger reason to prefer a person who is less well off). 18 These interpretations reflect distinct factors and are therefore independent, although they are often related, since the overall state of a person might affect the magnitude of the effect of a certain measure on this person. The difference between the two interpretations might lead to different conclusions when the conflict is between persons whose overall state of wellbeing is unequal. For example, suppose that two persons, who differ in their overall state of well-being the overall state of one is 70 units and that of the other is 60 units could both benefit from a certain divisible resource equally (that is, the contribution of each unit of the resource to each person s well-being would be identical). According to the first interpretation, we should give each person half of the resource (a contribution of 10 units), thereby increasing the overall state of well-being of the first to 80 units and that of the second to 70 units. According to the second interpretation, we should give one-quarter of the resource (a contribution of 5 units) to the first person and three-quarters (a contribution of 15 units) to the second, so that the overall state of well-being of both would be equal (75). Another possibility is to take into account both the size of the contribution and the overall state of well-being, giving more or less weight to each factor.

12 Well-Being and Fairness 241 It is difficult to decide which interpretation should be adopted. The attraction of the second interpretation is clear: ultimately, the important factor seems to be the overall state of well-being. However, in resolving specific conflicts, the overall state of well-being might seem too detached, since it is comprised of factors that would often have nothing to do with the conflict at hand, such as the happiness or satisfaction of persons with respect to their lives in general. Eventually, however, it seems that this tendency to relate certain factors and to isolate others could not be justified. 19 Therefore, it seems that we should adopt the second interpretation, based on the overall state of persons. (It should be remembered, however, that persons might be responsible for various aspects of their overall level of well-being and then this consideration should also be taken into account.) 20 Several related points should be emphasized with respect to (mainly the second part of) the ECP. First, this principle reflects an agent-neutral, rather than agent-relative, reason for action: it is important that an equal share or chance would be given, regardless of the identity of the agent. Second, this principle reflects an intrinsically important reason for action, based on the concept of fairness: according each person an equal chance is a way of conveying impartial concern for all. There might occasionally be instrumental reasons for such a principle, but these are not the reasons the ECP is based upon. Third, the ECP is not (only) a method of overcoming (factual or normative) ignorance or fallibility or of reducing corruption or prejudices. We should follow it even assuming perfect (factual and normative) knowledge and impartiality. Fourth, the ECP, even when implemented through a random procedure or lottery, is not a way of evading responsibility or transferring responsibility to God or fate contrary to the assertions of some commentators (Häyry & Häyry, 1990, p. 10; Rescher, 1969, pp ; Elhauge, 1994, p. 1501). Rather, giving each person an equal chance is the right (fair) thing to do. Fifth, the (equal) chance to be preferred is, at least, of instrumental value for each person. In particular, it is important for each person, at least, in light of the possibility that it would lead to preferring this person, whether or not the chance of being preferred is also of intrinsic value (valuable in itself) regardless of the final resolution of the conflict. Finally, the ECP does not dictate the way in which an equal chance is accorded. The ECP should thus be distinguished from possible ways of implementing it, such as a lottery. (Generally, the ECP should be implemented in the best way of according each person an equal chance, which might be affected by the circumstances of each case.) On first sight, the ECP might be embraced by many. However, once its implications are fully understood, it turns out to be very controversial. In fact, the ECP is sometimes rejected and, more often, ignored or downplayed. It is rejected by the view that considers the answer to the question whom to prefer in conflicts in which all relevant considerations are equal or

13 242 R. Segev balanced a matter of indifference (Broome, 1984, pp. 40, 55; Harris, 1985, pp ; 1996, p. 278; Statman, 1995, Ch. 1). A more benign view of the ECP concedes that if all relevant factors are equal, then something like the ECP might be reasonable, but does not recognize its significance. This view gives the ECP only the force of a tiebreaker and rejects it whenever there is any special reason, no matter how trivial, to prefer one person to another (Broome, 1984, pp. 40, 50 54; Harris, 1996, p. 278). The most notable position that rejects the ECP, like considerations of fairness in distribution in general, is utilitarianism. According to utilitarianism, when the choice in each person would affect the aggregate balance of the good (namely, well-being) equally, it does not matter which alternative (person) is chosen. It is thus a mistake to write that, in such cases, utilitarianism would direct us to toss a coin (Ubel, Arnold, & Caplan, 1998, p. 265). Utilitarianism does not attach an intrinsic importance to the question which individual is benefited, and particularly to giving persons equal chances. Therefore, when all factors are equal (including the cost of saving or benefiting each person), utilitarianism would endorse the cheapest method of choice, presumably saving or giving a benefit to the first person that comes to mind, without a single further thought on the matter. On the other hand, when there is a difference between the options with respect to their expected effects on the aggregate balance of well-being, utilitarianism would straightforwardly prefer the person who would benefit most from the resource, even if she is expected to benefit only slightly more than another (Stein, 2002a). This view ignores the considerations underlying the ECP. This view that considers the ECP only as a tiebreaker seems to miss an important point: if, as I suggest, the ECP indeed reflects an important reason for action, then it should not be outweighed by any contrary reason but only by contrary reasons the weight of which are beyond a certain threshold. Clashes between the ECP and other considerations should be resolved in light of their relative force. The force of the ECP depends on the importance of the interests at stake: there is a stronger reason to give an equal share or chance in a conflict of lives than in a conflict between pieces of property (each belonging to a different person). Within the framework I suggest, the ECP might clash with the IP or the PFR, namely, with the reason to prefer a more important interest to a lesser one and with the reason to prefer a person who is not responsible for the existence of an interpersonal conflict to a person who is. These potentially clashing principles might outweigh the ECP when they are strong enough. As a single principle, the ECP leads to implausible implications (Rescher, 1969, p. 186, n. 22; Calabresi & Bobbitt, 1978, pp ): by giving the same chance to receive a scarce resource to a person who would benefit from it trivially as to a person who would benefit from it greatly (Stein, 2002a, p. 227; McKie et al., 1998, pp ); and by giving the same chance to a person who is not responsible for his predicament as to a person who is. But, it is important to

14 Well-Being and Fairness 243 emphasize, the ECP might be decisive not only when other considerations are inapplicable namely, when the conflict is between roughly equal interests and none of the persons involved is (especially) responsible for the existence of the conflict but also when other considerations apply but are not strong enough to outweigh the ECP. When other considerations are not applicable or, more importantly, are outweighed by the ECP, the framework I propose differs significantly from prevalent suggestions and practices regarding the distribution of health care. Particularly, the proposed thesis applies the ECP with respect to every person that might benefit from the relevant resource. This framework excludes various considerations that might clash with the ECP, for example, citizenship or place of living (unless distance affects the magnitude of the expected contribution, in which case it should be considered within the IP). 21 The proposed conception also excludes another common consideration in the allocation of scarce health resources: waiting time (in waiting lists ) or the order of arrival ( first come, first served ). 22 Some believe that justice requires adhering to waiting time as a distributive criterion (Ubel, Arnold, & Caplan, 1998, p. 263; Veatch, 2000, pp ; Rhodes, 2002, p. 350). Particularly, it is suggested that waiting time is equivalent to a lottery that gives each person an equal chance (Calabresi & Bobbitt, 1978, p. 43; Kamm, 1993, p. 295; Harris, 1996, pp ; Veatch, 2000, pp. 17, 407; Elhauge, 1994, p. 1494). These views seem to me misguided. Waiting time and order of arrival are not, in themselves, morally significant factors. Waiting time is largely determined in light of whether the health need was created suddenly or gradually (during a long period of relatively slow deterioration). This in itself is not a morally significant consideration. Order of arrival, or waiting time in formal waiting lists, not only lack independent moral significance but, in addition, are not necessarily random: they are often affected by factors such as accessibility to medical care, which in turn is affected by socioeconomic factors (e.g., wealth) (Calabresi & Bobbitt, 1978, pp , , n. 93; Kamm, 1993, p. 295; Häyry & Häyry, 1990, p. 20; Koch, 2002, pp ). Waiting time might be occasionally relevant for contingent reasons (Kamm, 1993, pp ). For example, waiting time might sometimes be an approximation of just how poorly off persons actually are, either with respect to their overall state or with respect to the severity of the particular illness, because of the potential deterioration of the patient s condition that often occurs over time or because of the negative physical or mental effects such deterioration or prolonged waiting might engender. Similarly, waiting time on formal waiting lists might be an indication of the condition of patients if they were listed (in the waiting list) when their condition was equally serious and they have suffered more or less equally once they are listed. However, these contingent considerations are distinct from waiting time itself. I suggest that such considerations should be taken into account

15 244 R. Segev within the ECP and the IP (i.e., only when they indeed affect the overall state of the persons involved or the importance of the particular interest at stake). When, or to the extent that, waiting time does not affect these factors (e.g., when waiting longer does not lead to a worse condition), then, in my opinion, it should not be taken into account as an independent consideration. These remarks apply to waiting time or to order of arrival, as an independent factor, either as the only consideration or, as is typically the case in practice, as one among several other considerations (primarily considerations relating to the prospects of success, such as the extent of the match between an available organ and its possible recipients in organs allocation). However, adhering to waiting time or order of arrival as the only factor is a much less plausible option since it ignores several relevant considerations (Calabresi & Bobbitt, 1978, pp. 43, 187), particularly, the importance of the resource to the competing persons (Kamm, 1993, p. 295) and whether one or more of them is responsible for his predicament. IV. THE RELEVANT IMPORTANCE PRINCIPLE The IP determines the strength of the reason provided by the WBP in light of the importance of the interest, the probability that it will be affected (for better or for worse) if an action is not performed and the probability that an action will affect it. Assuming that all relevant probabilities are equal, the more important is the interest, the stronger is the reason to protect or enhance it. In distributing health resources, there is thus a reason to prefer the person who could be benefited most significantly by the resource (e.g., a person whose life could be saved to a person whose quality of life could be slightly improved). The interpretation of the IP raises a question parallel to the one discussed with respect to the ECP, namely, whether the importance of interests, as aspects of well-being, should be elaborated as referring to the size of contribution for each individual s well-being, to the overall state of wellbeing of each individual, so that a person who is relatively worse off would be preferred, or to both these factors. 23 Again, the difference between the two factors might lead to different conclusions when the conflict is between persons at unequal starting points in terms of their overall state of wellbeing. For example, suppose that two persons, who differ in their overall state of well-being, could benefit from a certain (indivisible) resource equally (i.e., the contribution of the resource to each person s well-being would be identical). According to the first interpretation, the IP does not apply in this case, since the size of the potential contribution to each is equal. On the other hand, according to the second interpretation, we should give the resource to the person whose overall state of well-being is lower.

16 Well-Being and Fairness 245 With respect to the ECP, I have suggested that the overall interpretation would be favored, since the aim in that context is equality and, in this regard, there is no reason to prefer one aspect of well-being to another. However, in formulating the IP, the question is different; namely, what reasons exist for preferring one person to another? In this respect, it seems that both factors should be taken into account: there is reason to prefer a person who can benefit more over a person who could benefit less and also a person who is relatively worse off over a person who is relatively better off. With respect to the distribution of scarce health resources, when both persons are in the same overall state of well-being, the relevant consideration for determining the force of the IP is the expected contribution of the resource to the quality and length of life (which together constitute the well-being) of individuals. This consideration combines two factors. The first factor is the (expected) importance of the interest at stake, as an aspect of individual well-being (i.e., its [expected] contribution to the quality and length of life of the patient). For example, there are significant differences, in this respect, between death, loss of a major limb, and suffering a minor inconvenience. The second factor is the (expected) contribution of the resource to the interest at stake and thus to the overall well-being of the individual. This factor includes the probability of success, both short-term (i.e., the chance of achieving the immediate aim of the procedure) and long-term (i.e., the chance of full or significant recovery). These two factors, which highlight different aspects of the general consideration of the expected contribution of the resource to the quality and length of life, must be considered together. For example, it is impossible to determine what is the interest at stake without knowing what it is possible to prevent with available resources. This analysis has an interesting implication. Many describe a clash between two considerations in the allocation of scarce health resources. The first is the urgency of the need ; namely, the prospects of each person without the resource 24 (e.g., the life expectancy without the resource when the danger is death). And the second is the magnitude of the expected benefit of the resource (e.g., the life expectancy with it). It is often claimed, for instance, that while the first consideration typically favors the oldest and the sickest, the latter typically favors the youngest and the healthiest (Caplan & Coelho, 1998, pp ). This claim is correct in pointing out that urgency and expected benefit are distinct factors, which might pull in different directions. The proposed analysis, however, suggests that these two factors are related: both are part of the more general consideration of the expected contribution of a resource to a person s quality and length of life. This is so since the contribution of a resource to a person namely, the additional benefit a resource would contribute to a person is affected by the person s prospects both with and without the resource. On the one hand, the more urgent is the need, the greater the expected contribution of the resource

17 246 R. Segev (e.g., other things being equal, a person who is expected to die in one year without a resource would benefit from it more than a person who is expected to die in five years without it). 25 On the other hand, the longer and better the life of a person is expected to be with a resource, the greater its expected contribution (e.g., other things being equal, a person who is expected to live with a resource for ten years would benefit from it more than a person who is expected to live for six years with it). When we consider the factors together, for instance by combining these two examples, the conclusion is that, other things being equal, the expected contribution to both persons is the same since the life of each could be prolonged by five years (from one year to six years or from five years to ten years). To be sure, my suggestion that urgency and expected benefit are part of the same general consideration of expected contribution does not mean that these factors are not distinct and potentially clashing. The point is only that these considerations are related in an important respect. The IP in itself is widely accepted, especially in the context of health care. The idea that the contribution of health resources to individual wellbeing, as reflected in the above factors, is a relevant consideration in the allocation of scarce health resources is commonplace (Veatch, 2000, p. 289). The crucial question is what should be the scope and weight of this consideration, relative to that of other, potentially competing, considerations. This is the subject of the following sections. Here it is important to briefly note several preliminary points concerning the implications of the IP itself and its interaction with other considerations and conceptions. First, the IP excludes considerations that are not related to the wellbeing of individuals. For example, the IP considers the cost of health resources, but only to the extent that it affects individual well-being. Of course, the cost of a resource could typically affect the availability of other health care resources that might contribute to the well-being of individuals. But this is not necessarily the case: it is possible that if a certain amount of money would not be used to produce a certain health resource, it would be used for some other purpose that does not contribute in any way to the well-being of persons. 26 Second, the IP is just one of the principles I suggest, in addition to the other two conceptions of fairness reflected in the ECP and the PFR. This distinguishes the proposed theory from the prevalent view that, in decisions regarding the allocation of scarce health resources, factors that are part of the IP (e.g., the size of the benefit and the prospects of success) are not only relevant, but should be given precedence, 27 often as the only (or, at least, the preliminary) excluding factors (Rescher, 1969, p. 177; Rhodes, 2002, p. 351). This, I believe, is unjustified: while the considerations reflected in the IP, such as the size of the benefit and the prospects of success, are valid considerations, they are not the only ones and should therefore not be considered a priori decisive.