University of Southern California Dissertations and Theses

Intrinsic and extrinsic values of distributional equality

(USC Thesis Other)

Intrinsic and extrinsic values of distributional equality

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Intrinsic and Extrinsic Values ofDistributional Equality A Diss ertation Presented to the FACULTY OF THE USC GRADUATE SCHOOL UNIVERSIT Y OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY Doctoral Committee: (PHILOSOPHY ) By Yingying Tang May 2014 Associate Prof essor John Hugo Dreher, Chair Prof essor Gary Lynn Watson Associate Prof essor Robert David English Dedication To my husband II Acknowledgements I want to express my great appreciation to my advisor, John Dreher, for his continuous support and patient guidance throughout the entire journey. I am also truly grateful to my committee member Gary Watson for his careful comments and illuminating suggestions. I owe my deepest gratitude to my husband, Lei Zhong. The dissertation would not have been possible without his love and care, warm encouragement, insightful suggestions, and tremendous support. I would like to extend my thanks to Sharon Lloyd, Gideon Yaf fe, Zlatan Damnjanovic, Stephen Finlay, James Higginbotham, Janet Levin, David Manley, Andrei Marmor, Mark Schroeder, Scott Soames, James Van Cleve, Kadri Vihvelin, Dallas Willard, Elizabeth Anderson, Philip J. Ivanhoe, Jiwei Ci, and Robert English for their valuable ad vices during my graduate studies. Lastly, I thank my grandmother, my parents and my sister for their understanding and support; I also thank my friends, Chih- Ying Lin, Hsin- Wen Lee, Tiffany Chang, Xiao Wei, Jing Qiu, Kai Ren, and Xu Han, for their help and encouragement. Ill Abstract Contemporary political philosophers have provided diff erent reasons for distributional equality. Luck egalitarians contend that many unequal distributions cannot be justified because they are too sensitive to fa ctors that are not a result of people's eff orts, but a result of people's luck. Suf ficientarians embrace the view that more equal distributions would make more people live a suf ficiently decent life. For prioritarians, a more equal distribution is justifiable because it would bring more benefits to the worse off. Although luck egalitarians, sufficientarians and prioritarians attempt to provide different reasons for just if ying more equal distributions, they share two common features. First, they tend to assume that a unif ying principle or reason would suffice to just if y reducing all kinds of inequalities. Second, they tend to assume that equality of any good has no intrinsic value; instead, they attempt to just ify more equal distributions by appealing to the extrinsic values of equality. In contrast to all of those philosophical positions, I propose a pluralist approach on the reasons for justif ying more equal distributions. I argue that desert-based reasons (which may involve fa ctors ofluck to a large extent), suffi ciency-b ased reasons, priority-based reasons, and more distinctively the intrinsic value of equality can all be used to justif y a more equal distribution, albeit in diff erent circums tances. IV Dedication Acknowledgements Abstract Chapter 1 Introduction Table of Contents Chapter 2 Luck, Desert and Distributional Equality Chapter 3 Sufficiency and Distributional Equality Chapter 4 Priority and Distributional Equality Chapter 5 The Intrinsic Value of Equality Conclusion Bibliography II Ill IV 1 18 52 76 103 131 134 v Chapter 1 Introduction Significant social and economic inequality has been a serious problem in many countries, such as China and the United States. More equal distributions of social and economic goods within a society appeal to both academics and the general public. But how to just ify a more equal distribution? What are the reasons for distributional equality? Is equality of any kind by itself valuable? Or is it only of instrumental value for some other values? 1 Contemporary political philosophers have provided justi fications for more equal distributions by appealing to diff erent theories, among which the hotly discussed ones are oftentimes designated by contemporary academic community through the following three terms, "luck egalitarianism'', "sufficientarianism" and "prioritarianism" . Unfortunately, none of the three terms has a definite designation. A term, say "luck egalitarianism'', is actually used to designate or pick out a whole set of theoretic positions, some of which are even incomp atible with others within the same set. From now on, I only use "luck egalitarianism", "sufficientarianism" and "prioritarianism" to pick out some insightful theoretic positions that mean to address the important issues. But it is worth noting that I am not saying that I use a certain term to pick out all important theories that belong to a certain set of theoretic positions. Luck egalitarians contend that many unequal distributions cannot be justified because they are too sensitive to fa ctors that are not a result of people's 1 Since what connts as a just distribution is context sensitive, in order to avoid nnnecessaxy controver sy, this dissertation only focuses on the distribution within one connt:ry. 1 eff orts, but a result of people's luck. Sufficientarians embrace the view that more equal distributions would make more people live a sufficiently decent life. For prioritarians, a more equal distribution is justifiable because it would bring more benefits to the worse off Although luck egalitarians, sufficientarians and prioritarians attempt to provide different reasons for just if ying more equal distributions, they share two common features. First, most of them assume that a unifying principle or reason would suffice to just if y reducing all kinds of inequalities. Some of them do not assume that, but attempt to propose a unif ying principle or reason that would suffice to just if y reducing all kinds of inequalities, or that could apply to the justi fication of a more equal distribution in all cases or circumstances. Second, most of them, actually almost all sufficientarians and prioritarians, share the view that equality of any good has no intrinsic value. 2 It is worth noting that I follow a standard usage of the term "intrinsic value", according to which xis intrinsically valuable if xis valuable for its own sake, or 2 It is worth noting the distinction between the fact that a person values equality and the fact that a person values getting closer to equality. It is easy to recognize the conceptual distinction between the two facts. However, people oftentimes mistakenly asswne that anyone who values equality must value getting closer to equality. In fact, a person could value equality without valuing getting closer to equality. For example, a person may value writing as good a novel as Anna Karenina, meanwhile she wouldn't necessarily prefer being somewhat less terrible a novel writer than she now is, if she is still not in the same league with Leo Tolstoy. However, although we cannot say that it is irrational to value equality without valuing getting closer to equality, we must recognize the fact that, usually if a person values equality for its own sake, then she must value getting closer to it as well. This is especially true for the cases concerning people's well being. For instance, if it is hard to achieve full equal liberties in a certain society, and there are only two available distributions of which we may choose one, one is more equal and the other is less equal (say the more equal society is like a laissez faire society and the less equal society is like society with slavery), a person who values equal liberties would value the more equal society instead of the less equal one. So, although in some circwnstance, a person who values equality may not value getting closer to equality, we should also note that, actually in many cases, if a person values equality, then she would value getting closer to equality as well, and she values getting closer equality because she values equality. 2 in other words, valuable as an end. For most luck egalitarians, sufficientarians and prioritarians, a more equal distribution is at most a by-product of promoting other values, or instrumentally valuable for some other ideals. In other words, they contend that equality only has extrinsic values. For instance, Harry Frank furt, one of the most fa mous suffi cientarians, passionately denies that equality of any sort has intrinsic value (or inherent value in Frankf urt's own terminology) s ; instead, he contends that a particular type of social arrangement may promote equality for the sake of other morally worthy goals (Frankfurt 19 87, 2000) 4 In contrast to all of those philosophical positions, I hold a pluralist approach on the reasons for justif ying more equal distributions. I argue that desert-based reasons that may involve fa ctors of luck to a large extent, suffi ciency-b ased reasons, priority-based reasons, and more distinctively the intrinsic value of equality can all be used to justif y a more equal distribution, albeit of different goods or in diff erent circumstances. 3 Frankfrnt says, "In rejecting equality as a moral ideal, I intend the scope of my rejection to be entirely nnlimited. It applies without restriction to all egalitarian doctrines. A variety of particular modes of equality have been especially recommended or insisted upon by one group of advocates or another. Some egalitarians argue that the most attentive consideration ought to be given to guaranteeing equality in levels of income or of wealth. According to others, the sort of equality that is most indispensable to a just society is equality of opporhmity. Still others liTge that the highest priority should be given to establishing equality of welfare or well-b eing, or to ensliTing the equal enj oyment of certain basic rights or imrnnnities; and so on. I deny the inherent moral imp ortance of equality of every type. My claim is simply that equality as such, in any of its diverse modes, is always morally irre levant." (Frankfrnt 2000, p. 90) 4 For instance, he says, "[U]nder certain conditions efforts to reduce inequality may stimulate productive economic activity; and at the times of efforts of that sort may also be helpful in avoiding costly social disorder. Sometimes, however, striving to reduce inequalities may be obligatory simply because it will ameliorate nnnecessary suffering. Quite apart from considerations of self-inter est, there may be a stringent moral imperative that we make conscientious efforts to approach more closely to a state of affairs in which the circwnstances of people are- at least with respect to certain aspects of their lives -more substantially equal." (Frankfurt 2000, pp. 88-89.) 3 This dissertation aims at achieving two goals: First, to clar ify what luck egalitarianism, suffi cientarianism and prioritarianism are and to capture the important concepts that the relevant authors have in mind, or want to express clearly. Second, I propose five reasons for just if ying more equal distributions, two desert-based reasons (including a contribution-based reason and a difficulty based reason), a suffi ciency-b ased reason, a priority-b ased reason, and the intrinsic value of equality reason. Notice that I'm not going to give a systematic theory on how to weigh some of those reasons against others -t hat could be a topic of other projects. In the rest of the introduction chapter, let me introduce the general background of each theory that is examined and clarified in this dissertation, briefly introduce how I characterize the important concept that each theory aims at capturing and the five reasons that I propose for justif ying more equal distributions. 1. Luck Egalitarianism The term "luck egalitarianism'' comes from Elizabeth Anderson's fa mous paper, "What is the Point of Equality?" In that paper, Anderson distinguishes what she calls "luck egalitarianism'' from other concepts of equality (Anderson 19 99, p. 289). Luck egalitarians are mainly concerned with injustice involved in the cases where the fa ctors that are a result of people's luck, rather than their eff orts, significantly influence people's life prospects. Ronald Dworkin (Dworkin 1981a , 19 8 1b, 19 85, 2000, 2002) and Gerald Cohen (Cohen 19 89, 19 97, 2006, 2008) 4 have been recognized as the two most important luck egalitarians. Other luck egalitarians may include Richard Arneson, Thomas Nagel, Eric Rakowski, John Roemer and probably Philippe Van Pa rij s (Anderson 19 99). Luck egalitarianism originated from some of John Rawls's thoughts. In A Theory of Justice, Rawls rejects the idea that whatever distribution of goods is brought about by the operation of a competitive market economy is just. His objection is that in such a system distributive outcomes are improperly influenced by "natural and social contingencies", fa ctors that are "arbitrary from a moral point of view" 5 According to Rawls, these matters are entirely beyond the individual 's power to control. To the extent that they determine an individ ual's lif e prospects, they do so by sheer luck, luck that fa lls on a person directly and independently of any choice she might make. Rawls aspires to give a conception of justice that "nullifies the accidents of natural endowment and the contingencies of social circumstance as counters in quest for political and economic advantage" (Rawls 1971, p. 15 ). As with Rawls, Dworkin argues that the free market system is not a fa ir procedure of distributing resources. He maintains that the problem of a free market is that people do not start their lives on equal terms. For instance, "some begin with marked advantages of fa mily wealth or formal and informal education. 5 Rawls identifies three kinds of contingencies that are arbitrary fr om a moral point of view (Rawls 2001): (a) their (citizens') social class of origin: the class into which they are born and develop before the age of reason (inheriting wealth and social connections from one's parent); (b) their native endowments (as opposed to their realized endowments) (a genetic disposition to be more or less talented in various ways); and their opporhmities to develop these endowments as affected by their social class of origin (being well or badly educated and nliThrred in early childho od); (c) their good or ill forhme, or good or bad luck, over the colU'se of life (how they are affected by illness and accident; and, say, by periods of invohmtary nnemployment and regional economic decline). 5 Others suffer because their race is despised." "[P]eople are not equal in raw skill or intelligence or other native capacities ... they dif fer greatly, through no choice of their own, in the various capacities that the market tends to reward." (Dworkin 19 85, p. 207) Therefore, free exchange of commodities cannot guarantee that the results are just. 6 Cohen is also deeply impressed by how unfair it would be to let people's distributive shares be determined by sheer luck. For instance, he says, "It seems to me merely perverse to deny the intelligibility of outrage at the unfairness of mere luck causing a huge social inequality. We can argue about the correct further construal of the content of that outrage, but that its proper obj ect is inequality caused by luck seems to me evident, and there is nothing def eatingly 'trivial' about saying so." (Cohen 2008) 6 It is well-known that Dworkin distinguishes circumstance fr om choice. People's choices, he says, are linked to their personality that is made up principally of their ambition (broadly nnderstood) and character. He says, "Someone's arnbitions include all his tastes, preferen ces, and convictions as well as his overall plan of life: his ambitions fWT iish his reasons or motives for making one choice rather than another. Someone' s character consists of those traits of personality that do not supply him with motives but that nevert heless affect his pliTsuit of his ambitions: these include his application, energy, industry, doggedness, and ability to work now for distant rewards, each of which might be, for anyone, a positive or negative quality." (Dworkin 2000, p. 322) Circwnstances consist of an individual's personal and imp ersonal resolU"ces. Personal resolU"ces are an individual's physical and mental capabilities, including the market value of their talents. Imp ersonal resolU"ces are socially transferable goods, such as an individual's wealth and property and her legal opporhmities to make use of that wealth and property. (Dworkin 2000 p. 322-323) On the one hand, Dworkin contends that "[People should not] have different amonnts of wealth just because they have different inherent capacities to produce what others want, or are differently favored by chance." (Dworkin 1985, p. 207) On the other hand, he maintains that we need to construct a market system that is compatible with the ideal of equality of circwnstance. Cohen passionately embraces Dworkin's distinction between choice and luck in the shaping of people's fates (or prospects). (Cohen 1989) Compared to Dworkin, Cohen contended more explicitly and more wholeheartedly that the distinction between circwnstance and choice is the most fnndamental distinction for an egalitarian. For instance, he thinks that egalitarians ought to object to all and only those inequalities that do not appropriately reflect choice. As he says, "Equalities are just if and only if they reflect patterns of choice, and of fail liTes to choose, on the part of the people among whom the inequality prevail s." (Cohen 2008, p. 403) 6 Although luck egalitarians have raised an intriguing and important issue, their theories haven't been form ulated accurately. This is why their writings are oftentimes open to various interpretations T In order to carry out a fruitful discussion on luck egalitarianism and capture the fundamental moral intuitions that concern luck egalitarians, we need to first give an accurate form ulation of it, which would capture the ess ential ideas of those "luck egalitarians". Through examining what most luck egalitarians presuppose, I characterize luck egalitarianism in terms of the following thesis. (LE) Whatever is not a result of people's effort, but is rather a result of luck, cannot serve as a desert-based reason for just if ying an unequal distribution. (LE) has the im plication of fa voring a more equal distribution in the real world. In an unregulated market system, pervasive significant inequalities between the better off(e.g. super models and sports stars) and the worse off(e.g. care-takers and dirty industry workers) are mostly not a result of people's actual eff orts, but a result of luck. The society would tend to be more nearly equal if the distributions are less sensitive to the influence ofluck. In chapter 2, I examine LE and argue that it is a fa lse view. I argue that the fa ct that a person makes more contributions than another person serves as a desert-based reason for just if ying an unequal distribution between them, even when the diff erence in contributions is a result ofluck. I also argue that the fa ct 7 Some authors take luck egalitarianism to be the thesis: people must be held responsible for their choices and be compensated for brute misforhme. Some authors take luck egalitarianism to be the following thesis: the effects of brute luck ought to be equalized. Some take it to be a view as follows: a distribution is nnjust if it is sensitive to (or depends upon) people's sheer luck. All of these formulations are extremely ambiguous and fail to do justice to those luck egalitarians' thoughts. 7 that it is more difficult to bring about one socially beneficial outcome than to bring about another socially beneficial outcome can serve as a desert-based reason for just if ying an unequal distribution among people, even though difficulty is oftentimes a matter ofluck to them. The large inequality in the real world between the better off (e.g. super models and sports stars) and the worse off(e.g. care-takers and dirty industry workers) ought to be reduced for the sake of justice. But the in justice involved in the large inequality between those people who earn more in the unregulated market system and those who earn less is not because, as luck egalitarians assume, the inequality is sensitive to people's luck. Instead, there are two other reasons why the large inequality is un just: 1) the distribution isn't sensitive to how the better-offs contributions compare to the worse-offs contributions; and 2) the distribution isn't sensitive to how the difficulty of the better-offs occupation compares to the difficulty of the worse-offs occupation. 2. Sufficientarianism Sufficientarianism is embraced by many fa mous philosophers including Elizabeth Anderson, Roger Crisp, Harry Frankfurt, Joseph Raz, Alexander Rosenberg, and Amartya Sen. Frankfurt is the first philosopher who gives a relatively clear-cut account of sufficientarianism. In his 19 87 paper, "Equality as a Moral Ideal", Frankf urt raised "the doctrine of suf ficiency" : "With respect to the distribution of economic assets, what is important from the point of view of morality is not that everyone should have the same but that each should have enough. If everyone 8 had enough, it would be of no moral consequence whether some had more than others." (Frankfurt 19 87, p. 22) Although Frankf urt's doctrine of sufficiency is only concerned with the distribution of economic assets, in fa ct he believes that some of what he says about economic equality and suf ficiency applies to other benefits, like welf are, opportunity, respect and the satisfac tion of needs. (Frankfurt 19 87, p. 22, footnote 3) Actually almost all other sufficientarians mentioned above would think that the doctrine of sufficiency applies to other kinds of good as well. Sufficientarians all share the view that, the moral badness involved in significant inequality is not concerned with the fa ct some are worse off than others, but concerned with the fa ct that the worse off are too badly off (according to a proper standard). In other words, what is of moral significance is not concerned with how one's well-being compares with others in any respect, but concerned with one's absolute well-being in some respect. As Frankfurt says, "What I believe they find intuitively to be morally objectionable, in the types of situations characteristically cited as instances of economic inequality, is not the fa ct that some of the individuals in those situations have less money than others but the fa ct that those with less have too little." (Frankfurt 19 87, p. 32) Although suffi cientarianism has been widely discussed by contemporary political philosophers since the time when Frankfurt raised the doctrine of suffi ciency, the accurate characterization of suffi cientarianism is still unsettled. Through examining what both the proponents and the opponents of 9 suffi cientarianism have presupposed, I characterize suffi cientarianism in terms of the following two theses. (S l) a primary reason for distribution is to make the citizens reach or get closer to the sufficiency threshold; (S2) If a person reaches the suf ficiency threshold, then there wouldn't be any further reasons for distribution for the sake of improving her well being. In chapter 3, I first examine five fa mous accounts of the sufficiency threshold: suf ficiency as satiety, sufficiency as the satisfaction of basic needs, Crisp's account of sufficiency that appeals to an impartial spectator's compassion, Frankf urt's account of sufficiency that appeals to the pursuit of one's lif e projects, and Anderson's account of sufficiency as a list of personal goods. I argue that all the five accounts of sufficiency fa il to satisf y both of the two theses (although they may satisf y one of the two theses). Although none of the five accounts is tenable, it doesn't follow that sufficientarianism per se is fa lse. I instead propose an account of what I call "comparative sufficiency", which understands suf ficiency as the level of social average, i.e. the average wealth per capita. In my view, having a person reach or get closer to the level of social average serves as a primary sufficiency-based reason for distributing some goods to her. 3. Prioritarianism Bef ore introducing prioritarianism, it is worth mentioning weighted utilitarianism that was embraced by many economists, such as Kenneth Arrow, 10 Alfred Marshall, A. C. Pigou and Paul Samuelson, since it is very possible that prioritarians receive their flashes of ideas from weighted utilitarians. Following classic utilitarians, weighted utilitarians assume that the principle of maximizing the sum of individual utilities is the unif ying principle of distributive justice. But diff erent from classic utilitarians, weighted utilitarians seek to justif y a more equal distribution of economic goods by appealing to the value of maximizing the sum of individual utilities. Bef ore illustrating the diff erence between classic utilitarianism and weighted utilitarianism, let me introduce the following symbol, <x�, X2, Xs, ... , Xn>. I use the symbol to indicate a distribution of a certain good among n persons. In this symbol, Xi indicates a certain level of the distributive shares (of a given good) en joyed by the ith person. For instance, 300 hundred dollars is distributed between Smith and Tom. One distribution might be < 100, 200>, where Smith receives 100 hundred dollars and Tom receives 200 hundred dollars. A more equal distribution might be <1 20, 18 0>, where Smith receives 12 0 dollars and Tom receives 18 0 dollars, or <1 50, 15 0> where each person receives 15 0 dollars. Now suppose that, Smith has 1, 000 dollars, and Tom has 10, 000 dollars. We are going to distribute 100 dollars to one of them. A classic utilitarian would contend that we could distribute the money to either of them. She would argue that, as fa r as justice is concerned, there is no diff erence between the two distributions, <10 00 + 100, 10 000> and <10 00, 10 000 + 100>, since the sum of individual utilities under each distribution is the same. The classic utilitarian 11 reasoning is as follows. Suppose that the initial utility level of Smith is m and the initial utility level of Tom is n. The additional utility that is brought about through having Smith receive 100 dollars is a, and the additional utility that is brought about through having Tom receive 100 dollars is a as well. So under either distribution, the sum of individual utilities would be (m + n + a). In contrast, a weighted utilitarian would contend that the money should be distributed to Smith. She would argue that the sum of individual utilities in the distribution where Smith receives the money is greater than the sum of individual utilities in the distribution where Tom receives the money. Diff erent from classic utilitarians, weighted utilitarians recognize the truth of the law of diminishing marginal utility of money, according to which, the utility of the same amount of money diminishes as the owner becomes richer; in other words, the individual utility that a given amount of money brings is inversely proportional to the well-being of the recipient. For instance, Arrow says: "In the utilitarian discussion of income distribution, equality of income is derived from the maximization conditions if it is further assumed that individuals have the same utility functions, each with diminishing marginal utility." (Arrow 1971, p. 409) Samuelson says: "If extra dollar brings less and less satisfaction to a man, and if the rich and poor are alike in their capacity to en joy satisfaction, a dollar taxed away from a millionaire and given to a median-income person is supposed to add more to total utility than it subtracts." (Samuelson 1973, p. 16 4, n. 1) Given the truth of the law of diminishing marginal utility of money, suppose that the additional individual utility that is created by having Smith who initially 12 has 1, 000 dollars receive 100 dollars is a, and the additional individual utility that is created by having Tom who initially has 10, 000 dollars receive 100 dollars is b, it follows that a > b. Given the assumption that the initial utility level of Smith is m and the initial utility level of Tom is n, the sum of the individual utilities in the distribution where Smith receives 100 dollars would be (m + n + a), and the sum of the individual utilities in the distribution where Tom receives 100 dollars would be (m + n + b). Since a> b, (m + n +a)> (m + n + b). So according to weighted utilitarianism, the more equal distribution of money, <10 00 + 100, 10 000>, is more just than the less distribution of money, <10 00, 10000 + 100>, whereas according to classic utilitarianism, the more equal distribution and the less equal distribution are equally just. Although compared to classic utilitarianism, weighted utilitarianism could support a more equal distribution of money, it still allows a smaller utility gain of the worse off to be unlimitedly (or unrestrictedly) traded off for the greater utility gain of the better off. To put it in another way, since weighted utilitarianism shares the ethos of classic utilitarianism -- only maximizing the sum of individual utilities matters for distributive justice, it may in principle tolerate extremely unappealing inequalities 8 8 Rawls gives an excellent cri ticism of all versions of utilitarianism. He says, "The striking featill'e of the utilitarian view of justice is that it does not matter, except indirectly; how this swn of satisfactions is distributed among individuals any more than it matters, except indirect ly, how one man distributes his satisf actions over time .... The most natural way, then, of arriving at utilitarianism (although not, of colU'se, the only way of doing so) is to adopt for society as a whole the principle of rational choice for one man." (Rawls 1971, pp. 26-7) 13 In recent years, the term "prioritarianism'' has been used to refer to Derek Parfit's Priority View (Parfit 1991, 19 97, Crisp 2003, Holtug 2006, Otsuka and Voorhoeve 2009)· the view that benefiting people matters more the worse off these people are. Parfit has insightf ully distinguished the priority view from utilitarianism He says, "For Utilitarians, the moral importance of each benefit depends only on how great this benefit would be. For Prioritarians, it also depends on how well off the person is to whom this benefit comes." (Parfit 1997, p, 213 ) Moreover, Parfit maintains that benefiting the worse off matters more not because they are worse off than others, but because they are worse off absolutely (or non-comparatively ). As we can see, Parfit's priority view actually includes two theses, a positive thesis and a negative thesis. According to the positive thesis, the fa ct that a person is worse off absolutely serves as a reason for assi gning priority to benefiting her over benefiting a better-off person. According to the negative thesis, the mere fa ct that a person A is worse off than another person B cannot serve as a reason for assigning priority to A over B. The discussion of this chapter centers on the positive thesis. The negative thesis, as I will argue in chapter 5, is fal se. In this disser tation, by "prioritarianism'', I only mean the positive thesis of the priority view. Although prioritarianism can be vindicated, it is open to various interpretations as it stands -i t is not at all clear in what sense the fa ct that a person is worse off absolutely serves as a reason for assigning priority to benefiting her over benefiting a better -off person. 14 In chapter 4, I examine two interpretations ofprioritarianism, which I call PRl and PR2. According to PRl, the priority view is grounded in the principle that the moral value of benefiting the worse off is greater than the moral value of benefiting the better off According to PR2, the priority view is grounded in the principle that a moral agent has a stronger moral reason to benefit the worse off than to benefit the better off I argue that whereas PRl is fa lse, PR2 is tenable. I aim to show that even if in some cases the moral value of benefiting the worse off is the same as the moral value of benefiting the better off, a moral agent may still have a stronger reason to benefit the worse off than to benefit the better off. Moreover, I attempt to argue that the priority of benefiting the worse off is grounded in a more fundamental fa ct that a person would have a stronger prudential reason to improve her worse off prospect than to improve her better off prospect. According to my understanding of prioritarianism, the prioritarian slogan "the worse off matters more" is better interpreted as the thesis that a moral agent has a stronger moral reason to benefit the worse off than to benefit the better off Even if in some cases the moral value of benefiting the worse off is the same as the moral value of benefiting the better off, a moral agent may still have a stronger reason to benefit the worse off than to benefit the better off. I thus contend that there is a priority-based reason for fa voring a more equal distribution between the worse off person and the better off person. Moreover, as I have argued, the priority-b ased reason for distribution is grounded in a more 15 fundamental fa ct that a person would have a stronger prudential reason to improve her worse off prospect than to improve her better off prospect. 4. The Intrinsic Value of Equality Many philosophers reject the intrinsic value of equality. Some claim that it is hard to see how equality achieved through leveling-down the well-being of the better off with respect to any good could be intrinsically better. Some people worry that "envy" is one crucial emotion that motivates people to endorse the intrinsic value of equality. For instance, Raz contends that to value equal shares for their own sake "is to be irrational, or envious, or in the grip of a strange aesthetic ideal." (Raz 1986, p. 225.) Some people worry that the theories supporting the intrinsic value of equality encourage people to aim at something unimportant to their lives. For instance, Fran kf urt worries that acknowledging the moral importance of economic equality will alienate people. He says, A concern for economic equality, construed as desirable in itself, tends to divert a person's attention away from endeavoring to discover - within his experience of himself and of his lif e - what he himself really cares about and what will actually satisf y him, although this is the most basic and decisive task upon which an intelligent selection of economic goals depends. Exaggerating the moral importance of economic equality is harmful, in other words, because it is alienating. (Frankfurt 1987 , p. 23) I argue that all of the worries are based on some serious misunderstanding of the intrinsic value of equality. In chapter 5, the final chapter of this disser tation, I first examine some fa mous objections to intrinsic egalitarianism and then provide my positive argument for intrinsic egalitarianism. I argue that the equality is intrinsically valuable ins ofar as it is intrinsically valuable for some 16 persons, i.e. through having intrinsic personal value. In arguing that the equality has intrinsic personal value, I first propose an epistemic principle concerning intrinsic personal value as follows: It is (epistemically ) reasonable to regard x as an intrinsic value for a person S if S rationally desires x (as part of her lif e) for its own sake. After proposing the epistemic principle, I give some thought experiments to reveal our intuition that a person can rationally desire equality of respect, equality of liberties and equality of lif e years. 17 Chapter 2 Luck, Desert and Distributional Equality Luck egalitarianism is an influential position in contemporary political philosophy, being proposed by several prestigious philosophers including John Rawls, Ronald Dworkin, G. A. Cohen, and Richard Arneson. In essence, luck egalitarianism is the view that whatever is not a result of people's eff ort, but is rather a result ofluck, cannot serve as a desert-based reason for just if ying an unequal distribution. In Section 1 of this chapter, I illustrate the central thesis of luck egalitarianism by appeal to a series of examples. In the rest of this chapter, I put forward several objections to luck egalitarianism. In Sections 2-4, I argue that the fa ct that a person makes greater contributions than another person serves as a desert-based reason for just if ying an unequal distribution between them, even when the diff erence in contributions is a result ofluck. Then in Sections 5, I attempt to explain how the difficulty level of a job matters to distributive justice. I argue that the fa ct that it is more difficult to bring about one socially beneficial outcome than another socially beneficial outcome can serve as a desert-based reason for just if ying an unequal distribution among people, even though difficulty is oftentimes a matter ofluck to them. It is important to note that both the contribution-based reason and the difficulty based reason are different from an effort-based reason for distributions. In Section 6, I intend to indicate why effort cannot serve as a desert-based reason in the context of distributive justice. 18 1 Imagine that all citizens within a society have gained an extensive scheme of equal basic liberties, and equality of opportunities to offices and positions has been achieved.9 Nevertheless, large inequalities in people's income and wealth remain. In the society, the distribution of economic goods is completely determined by a libertarian market system, i.e. an unregulated free market system. Suppose that this year one million people have paid 10 dollars to watch a supermodel's performa nce. One person, who is suff ering from a disability, has paid 20,000 dollars to a caretaker. As a result, this year, the supermodel earns 10,000,000 dollars, and the caretaker earns 20,000 dollars. In the unregulated free market system, the supermodel earns much more than the caretaker because the supermodel brings some pleasure to million people and the caretaker only takes care of one person. Luck egalitarians would disagree with some libertarians about whether the fa ct that the supermodel brings some pleasure to million people and the caretaker takes care of only one person serves as a desert-based reason for the inequality. A desert-based reason for distributing some goods to an individual is a reason solely concerned with activities performed by the individual . Desert-based reasons are to be contrasted with many other kinds of reasons for distributions, like needs-based reasons that are concerned with absolute needs (for example, healthy food, fresh air and 9 So OliT imaginary society satisfies both Rawls's first principle of justice, that each person has an equal right to the most extensive scheme of equal basic liberties compatible with a similar scheme of liberties for all, and the first part of his second principle of justice, that social and economic inequalities are justified only if they are attached to offices and positions open to all nnder conditions of fair equality of opporhmity. (Rawls 1982, p. 161) 19 sanitary water) and ef ficiency-catering reasons that are concerned to encourage increased contributions to the whole society 10 A libertarian, who thinks that the outcomes of the unregulated free market system are just, would claim that the fa ct that the supermodel brings pleasure to a million people and the caretaker takes care of only one person serves as a desert-based reason that is sufficient to just ify the inequality between the supermodel and the caretaker. A luck egalitarian would deny that unregulated free market outcomes just ify the inequality between them. This is because the fa ct is not a result of people's eff ort, but a result of luck - it happens that million people are willing to gain some pleasure from watching the supermodel's performance and only one person needs to be taken care of Luck egalitarians believe that the injustice involved in the inequality is concerned with how fa ctors of luck aff ect people's fates. It is important to note that there are many fa ctors of luck. In this disser tation, for the sake of clarity, I will leave some complexities aside and rather focus only on three types of luck, which I call "Constitutive Luck", "Circumstantial Luck" and "Opportunistic Luck". 10 Many people would make the judgment that the supermodel doesn't deserve so much more than the caretaker. But notice that the sentence "the supermodel doesn't deserve so much more than the caretaker" could be interpreted in two ways. According to the first interpretation, the sentence refers to the fact that the inequality between the supermodel and the caretaker is nnjust. According to the second interpretation, the sentence refers to the fact that the reasons for the inequality , no matter what they are, aren' t desert-based. Luck egalitarians would strong ly agree that the supermodel doesn't deserve so much more than the caretaker. Moreover, by this claim, a luck egalitarian would not mean that the inequality between the supermodel and the caretaker is nnjust, but merely that the reasons for justifYing the inequality , no matter what they are, aren't desert-based. Even so, it is very likely that a luck egalitarian would contend that the inequality between the supermodel and the caretaker is in fact nnjust, and would certainly contend that the reasons for justifYing the nnequal distribution do not outweigh competing reasons for a more equal distribution. 20 a) Constitutive luck is the fa ctor of luck concerned with a person's talents or natural endowments in making a certain product or providing a certain service. b) Circumstantial luck is the fa ctor of brute luck concerned with the external circumstance in the process of production. c) Opportunistic luck is the fa ctor of luck involved in the process of social exchange. Let me use some specific examples to illustrate what I mean by each type of luck. Imagine that there are two bread bakers in the same city, Smith and Tom, who run separate shops, where each bakes and sells bread. Suppose that, Smith bakes and sells 1,000 loaves ofbread per day, but Tom bakes and sells only 10 loaves of bread per day. Suppose that the price of one loaf of bread is one dollar. As a result, Smith's gross sale per day is 1, 000 dollars, while Tom's gross sales per day is 10 dollars. Suppose that Smith's net profit (gross sales minus rent and the cost ofbaking bread) is 905 dollars and Tom's net profit is 5 dollars. Now compare the following three scenarios. Notice that in all three of the cases, I assume that Smith shares a similar social background with Tom. For instance, they came from the same kind of fa mily and enjoyed equal opportunities for, say, education, employment, and rights. (1) [Effort] Smith and Tom are equally talented, and equally lucky in other ways. However, Smith is much more diligent than Tom, who is just as lazy as a couch potato. In this scenario, Smith sells more bread than Tom, and hence makes more profit than Tom, because he bakes more bread than Tom; the 21 diff erence between their productivity is a result of the diff erence between their dforts. (2) [Constitutive luck] Smith and Tom are equally hard-working - they both are very diligent. However, as far as bread baking is concerned, Smith has much more natural talent than Tom. As with (1), Smith sells more bread than Tom does, and hence makes more profit than Tom, because he bakes more bread than Tom. But in this scenario, the diff erence between their productivity is a result of the diff erence between their natural endowments, which are a matter of luck. (3) [Circumstantial luck] Smith and Tom are equally talented and industrious, but the temperature and humidity at Smith's bread shop just happens to be-more conducive to bread-baking than at Tom's So as with (1) and (2), Smith sells more bread and makes more profit than Tom because he is more productive than Tom. But in this scenario, the diff erence between their productivity is a result of brute luck involved in the process of bread baking. (4) [Opportunistic luck] Smith and Tom are equally talented and industrious. Moreover, the circumstance ofbread-baking in Smith's bread shop is the exactly the same as the circumstance ofbread-baking in Tom's bread shop. But it ha ppens that more customers buy bread from Smith's bread shop and fewer customers buy bread from Tom's bread shop. This scenario diff ers from all the others previously considered in that Smith sells more bread and makes more profit than Tom not because he is more productive than Tom, but because the demand for bread from his bread shop is greater than the demand for bread from 22 Tom's bread shop. In this example we are supposing that the diff erence between the two levels of demand is a result of brute luck, Now consider the question, "In any scenario, does the fa ct that Smith bakes and sells more bread than Tom serve as a desert-based reason for just if ying the outcome where Smith makes more profit than Tom?" A luck egalitarian might agree that, in (1 ), the fa ct that Smith bakes and sells more bread than Tom serves as a desert-based reason for justif ying the unequal distribution. However, a luck egalitarian would deny that in (2), (3) or (4), the fa ct that Smith bakes and sells more bread serves as a desert-based reason for just if ying the unequal distribution. Contrary to scenario (1 ), where Smith bakes and sells more bread than Tom because Smith worked harder, in (2), (3) or (4), Smith bakes and sells more bread than Tom due to a matter ofluck. After introducing the four examples and the three kinds of luck, I suggest form ulating luck egalitarianism as below. (LE) Whatever is not a result of people's eff ort, but is rather a result of luck, cannot serve as a desert-based reason for just if ying an unequal distribution. (LE) has the im plication of fa voring a more equal distribution in the real world. In our society, pervasive significant inequalities between the better off (e.g. super models and sports stars) and the worse off(e.g. care-takers and dirty industry workers) are mostly not a result of people's actual eff orts, but a result of luck. The society would tend to be more nearly equal if the distributions were less sensitive to the influence ofluck. 23 (LE) shouldn't be confused with another claim, according to which whatever isn't a result of people's free choice cannot serve as a desert-based reason for just if ying an unequal distribution between those people. Something might be a result of people's eff orts without being a result of their free choice. Luck egalitarians might admit, although this is not an implication of (LE), that in such a case there could be a desert-based reason for distribution. Consider the following two examples. As with the four scenarios discussed above, Smith bakes and sells 1,000 loaves ofbread per day, but Tom bakes and sells only 10 loaves of bread per day. As a result, Smith's net profit (gross sales minus rent and the cost of baking bread) is 905 dollars and Tom's net profit is 5 dollars. ( 5) [External freedom] Imagine that in the bread-baking process, a crazed person points a gun at Smith's head and forces him to work hard at bread baking. In contrast, Tom is as free as a normal person typically is. Suppose that Smith doesn't want to work hard at all: if he were as free as Tom, he would make only 10 loaves of bread per day. But Smith has no other choice. If he ref uses to follow the 'crazy guy's' demand, he believes that he will die. As a result, Smith bakes and sells 1, 000 loaves of bread per day under the coercion. (6) [Internal freedom] Suppose that Smith is an unwilling workaholic, whereas Tom is a normal person free of any inner compulsion. Smith's desire to work hard always moves him to act. However, Smith actually doesn't want the desire to move him to act. He very much wants to be a person who en joys a relaxing lif estyle. His mental state is similar to the mental state of an unwilling drug addict, who doesn't want the desire of taking drugs to move him to act, but 24 at the same time the desire actually moves him to act. 11 Smith, the unwilling workaholic, is not as free as Tom. But given that Smith is a workaholic, he bakes 1, 000 loaves of bread per day - if he were as free as Tom is, he would have baked only 10 loaves of bread per day. All other things being equal, does the fa ct that Smith bakes and sells more bread than Tom in a particular way described in (5) or (6) serve as a desert-based reason for a distribution where Smith is better off than Tom? Although in each case, the fa ct that Smith bakes and sells more bread isn't a result of his free choice, LE doesn't prevent that fa ct from serving as a desert- based reason for an outcome where Smith is better off than Tom because the diff erential outcome is still a result of the diff erence between the two persons' eff orts. 11 My nnderstanding of how an nnwilling workaholic is wllree is illwninated by Har:ry Frankfrnt' s acconnt on how an nnwilling addict is wllree. (Frankfrnt 1971) Frankfurt distinguishes two kinds of desires: first-order desires and second-order desires. A desire is first order if and only if its object is not itself a desire. For example, if you want to play poker, then you have the first-order desire for playing poker, and playing poker is that desire's object. A desire is second-order if and only if its object includes a first-order desire. For example, say that you wish you wanted to have dinner with yoliT fr iend tonight (but in fact you want to prepare for yoliT midterm exam). In that case, you want to want to have dinner with yoliT fr iend. Wanting to have dinner with yoliT fr iend is the object of yoliT desire. So the desire is second-order. Frankfurt then identifies a particular kind of first-order desires: a person's wills. A desire is the agent' s will if and only if (1) it is first-order; and (2) it actually moves her to perform some action in fliTtherance of it-in other words, it is "effective". For exan1ple, you want to prepare for the exan1, but you also want to have dinner with yoliT fr iend. But the desire to prepare for the exan1 moves you. So the desire to prepare for the exam, not the desire to have dinner with yo liT friend, is yoliT will. A desire connts as yo liT will as long as it moves you, even if in the end you fail to do what you want. Suppose that you want to prepare for the exam, but the electricity is suddenly shut down. In this case, yo liT desire for preparing for the exan1 still moves you, though you fail to satisfY the desire. After that, FrankfliTt distinguishes between two kinds of second-order desires: mere second-order desires and second-order volitions. A second-order desire is a second-order volition if and only if it is a desire that a particular first-order desire should be the agent' s will, or, equivalently, that a particular first-order desire should actually move the agent to act. In contrast, a mere second-order desire is a desire merely to have a particular first-order desire. With respect to taking drugs, an nnwilling addict does not have fr eedom of will becaus e although her desire for taking drugs is her will, this will doesn't accord with her second-order volition; in other words, the unwilling addict doesn't want the desire of taking drugs to move her to act, but at the san1e time the desire of taking drugs actually moves her. 25 In this chapter, I will put forward two arguments against (LE): one is based on contribution, and the other is based on dijji" cult:y. In Sections 2-4, I will clarif y what I mean by "contribution" and argue that contribution can serve as a reason for distribution, regardless of whether contribution is a result of luck. I agree with some luck egalitarians that in both (5) and (6), there might be a reason for just if ying an unequal distribution between Smith and Tom. However, the reason is not that Smith puts more eff ort into the process of bread baking than Tom does, but that Smith makes a "greater contribution" than Tom: while Smith contributes 1, 000 loaves of bread per day, Tom contributes only 10 loaves of bread per day. Moreover, in Sections 5-6, I will attempt to explain how the difficulty level of a job (rather than individual 's actual eff ort) matters to distributive justice. I will argue that the fa ct that it is more difficult to bring about a socially beneficial outcome S, than to bring about another socially beneficial outcome S2 serves as a desert-based reason for justif ying an unequal distribution among people, even though dif ficulty is oftentimes irrelevant to people's actual eff ort, but a matter ofluck to them. 2 In the following Sections, I argue for the thesis below: (C) The fa ct that a person A makes more "contributions" than another person B serves as a desert-based reason for justif ying an unequal distribution between A and B, even when the diff erence in contributions is a result of luck. To better define the notion of "contribution", let me first introduce the notion of "an agent's bringing about an outcome". For any agent A, and any outcome or 26 state of affairs S, A brings about S if a certain amount of effort exerted by A helps to cause the occurrence of S. 1 2 Then we can accordingly define two ideas concerning contribution, (i) the idea that an agent makes some contribution to the society and (ii) the idea that an agent makes a greater contribution than another agent does. An agent A makes a contribution to the society if and only if A brings about an outcome S that has n "amount" of social good (where n > 0). Suppose that there are two persons, Al and A2. If Al brings about an outcome Sl that has m "amount" of social goods, A2 brings about an outcome S2 that has n "amount" of social goods, then Al makes greater contributions to the society than A2 does if and only if m > n. I want to give some examples in which an agent A brings about an outcome S, although some fa ctors that are a matter of luck to A play a causal role in the occurrence of S. The first example: Consider two possible worlds, world, and world 2 . In world�, Smith is a very kind person; but in world 2 , he is cold-hearted and only has minimum moral sense. Suppose that Smith in world, and Smith in world2 are exactly the same in all other aspects. In each world, Smith happened to see a drowning baby nearby. While Smith saved the baby in world�, he didn't do that in world 2 . The only fa ctor that results in the diff erence is that Smith in world, is more sym pathetic toward others than Smith in world2. Due to the diff erence in symp athy, Smith in world,, unlike Smith in world2, 12 In a case where the agent brings about the outcome, the exercise of her ph ysical and mental abilities must play a significant causal role. Imagine that you intended to build a house. Some angel who likes you somehow knew yoliT intention. The other night, when you were still asleep, she used her superpower to construct a beautiful house for you. You can't seriously tell yoliT fr iend that it is you who built the house, since yoliT agency doesn't play any significant casual role in the construction of the house. 27 succeeded in overcoming his other desires (such as the desire to keep his clothing clean and dry) for saving the drowning baby. In world�, the causal connection between Smith's action and the outcome is to a large extent influenced by Smith's constitutive luck, i.e. his natural moral endowment. But we still think that it is Smith who saved the baby in world, -i n other words, Smith brought about the outcome in question. The second example: Suppose that Smith in world, and Smith in world2 are both nice persons. As with the first example, there was a drowning baby nearby. Both Smith in world, and Smith in world2 intended to save the drowning baby. However, the current of river in world, is less rapid than the current of river in world2. As a result, whereas Smith successf ully saved the baby in world,, he didn't make it in world2. In this case, some fa ctor involving circumstantial luck results in the diff erence in outcome. But we should still say that it is Smith who saved the baby in world,. The third example: Marie happened to hook a drowning person when she was fishing. Imagine that water plants covered the drowning person so that Marie didn't know that she had hooked a person; she thought instead that it was a big fish. In this example, Marie didn't intend to save the person at all; the causal relation between Marie's action and the outcome is to a large extent a matter ofluck. Nevertheless, it is still reasonable to believe that Marie saved the drowning person. In the above three examples, it is fa ir to say that the agent brings about the outcome, although luck plays some causal role in the occurrence of the 28 outcome. Then we can easily find cases in which an agent makes some contributions to the society as a matter of luck, as long as the outcome brought out by the agent has a certain "amount" of social good. Accordingly, we can easily find cases in which an agent Al makes contributions of greater value to the society than another agent A2 does as a matter of luck, as long as the outcome Sl brought out by Al has a greater "amount" of social good than the outcome S2 brought out by A2. 3 3.1 In what follows, I want to argue for the thesis that the fa ct that some persons make more contributions than other persons serves as a desert-based reason for just if ying an unequal distribution between them. I will attempt to show that this thesis is rooted in a fundamental moral principle, the 'Principle of Reciprocity" . This principle can be put in a more precise way as follows: the fa ct that an agent A brings about an outcome S that is beneficial to B with an amount of utility, serves as a desert-based reason forB to benefit A in return with a proportional amount of goods, when a fitting circumstance obtains. We can better appreciate the idea of reciprocity by appeal to some specific examples. Recall the scenario where Marie happened to hook a drowning person when she was fishing. In that example, Marie didn't intend to save the person at all; the causal relation between Marie's action and the outcome is to a 29 large extent a matter ofluck. Nevertheless, it is still reasonable to believe that Marie saved the drowning person. Suppose that Benjamin is the drowning person. The fa ct that Marie saved Benjamin seems to provide Benja min with a reason to be grateful to Marie. To say that a person x has a reason to be gratef ul to another person y is to say that x has a reason to do something good in return toy. Certainly, the fa ct that x has a reason to do something good "to" y doesn't imply that x should actually do something good "to" y. But it does imply that when a fitting circumstance obtains, x should help y. Imagine that in another occasion, Marie lost her job, and meanwhile she was suff ering from a serious disease. Suppose that her disease is curable; but she doesn't have enough money to pay for the medical treatment. If Benjamin knows about this and if he is rich, he seems to have a strong reason to help Marie. 3.2 Some people might deny that the fa ct that Marie saved Benjamin's life provides Benjamin with a reason to help Marie. But I believe that those people probably confuse reciprocal desert with moral desert. Although the mere fa ct that Marie saved Benjamin doesn't imply that Marie is morally praiseworthy, the fa ct itself does give Benjamin a reason to benefit Marie in return. A person could be a fitting sub ject of reciprocal desert purely in virtue of her action, but she couldn't be a fitting subj ect of moral desert purely in virtue of her doing. Whether a person is a fitting sub ject of moral desert depends upon her intention involved in her doing or attempting to do something. 30 Let's consider more examples to illustrate the diff erence between reciprocal desert and moral desert. Suppose that I van and Tom each saved a drowning person Benjamin at diff erent times. But there is a diff erence: whereas Ivan saved Benjamin purely for the sake oflifesaving, Tom saved Benjamin only with the intention of winning a bonus. In this case, while Ivan is morally praiseworthy, Tom is probably not. However, it seems that Benjamin also has a reason to help Tom in return in a fitting circumstance, regardless of Tom's intention to save him. It might be true that Benjamin has a stronger reason to help Ivan than to help Tom, when they are in need of help. But this is not because the fa ct that Tom saved Benja min doesn't serve as a reason for Benjamin to help Tom, but because Benjamin has an additional reason to help I van due to his morally praiseworthy act. Consider again a previous example. Smith in world 1 and Smith in world2 are both nice persons. They both attempted to save a drowning baby with the same level of eff ort or attempt; they both intended to save the baby simply for the sake of saving a lif e. But the current of river in world 1 is less rapid than the current of river in world2. As a result, Smith successf ully saved the baby in world 1 , but he fa iled to do it in world2. As far as moral desert is concerned, Smith in world 1 seems to be as praiseworthy as Smith in world2. However, there is a diff erence regarding reciprocal desert: whereas the baby's parents should benefit Smith in world 1 , the parents are not obliged to benefit Smith in world2. 1 3 1 3 Although whether the agent is morally praise worthy or blameworthy doesn't depend upon factors regarding circumstantial luck, constitutive luck may be relevant to moral desert. Consider the relation between a particular moral endowment, the feeling of empathy, and moral blameworthi ness. Let me give some examples. One sunny Saturday aftern oon, Smith went out :n Moreover, whether a person deserves moral praise for performing an act depends upon whether the action is a free one or not. An agent is not morally praiseworthy for her action if the action is not a free action. Suppose that a crazy guy points a gun at Smith, coercing him to save Benjamin. If Smith were free, then he wouldn't try his best to save Benjamin -p robably he would exert some eff orts directed at saving Benja min, but he would give up if it is not very easy for him to save Benjamin. Compare this scenario with another scenario where the lif esaving performed by Smith is based on his free choice. In both scenarios, the fa ct that Smith saved Benjamin serves as a reason for Benjamin to be gratef ul to Smith to a certain extent (although Benjamin may has a stronger reason to be grateful to Smith in the second scenario). But Smith should not receive moral praise in the scenario where his action is not a free action. Now imagine that Smith is addicted to life-saving unwillingly. That is, although upon reflection Smith doesn't want the lif e-saving desire to influence his action, this first-order desire still moves her to act. Suppose that Smith saved a drowning person, Benjamin. In this case, we have the intuition that the fa ct for a walk. Along the road to the park, he saw a drowning baby in a puddle. He didn' t save the baby although he was sliTely physically capable of doing it. At that time, his desire of enj oying the walking on the sunny aftern oon was overriding. Is Smith morally blamewort hy for not saving the baby's life? Suppose Smith is a psychopath who completely lacks the feeling of empathy. He lacks the ability of taking others' pain into consideration. The psychopath Smith may not be the object of moral blameworthines s. Or, suppose that Smith's capability of empathy is slightly better than a psychopath. Although his capability of empathy is still lower than most people in the society, his capability is high enough to be normal -he can form right moral intentions through taking others' pain into consideration. The slightly better Smith would be blamewort hy for not saving the baby's life. But it is worth noting that as long as moral endowments are above a certain level, how blameworthy or praiseworthy an agent is doesn't vary with the level of her moral endowmen ts. Now suppose Smith's capability of empathy is high enough and hence he is a very kind person in most cases. Comp ared to the slightly better Smith, the nice Smith with high endowment of empathy wouldn't deserve much more moral blameworthiness than the former. Actually they may just deserve the same degree of moral blameworthi ness. 32 that Smith saved Benjamin's lif e provides Benjamin with a reason to benefit Smith to a certain degree in a fitting circumstance, although Smith is not as praiseworthy as he would be if his lif e-saving act reflected his "deep" self , free of all addictions. As those examples have shown, the most distinctive feature of the idea of reciprocal desert is that an agent could be a fitting subj ect of reciprocal desert purely in virtue of her doing. A benefit recipient B ought to recognize the mere fa ct that A brings about a state of aff airs that is beneficial to B as a reason for benefiting A in return. The idea of reciprocal desert is not only diff erent from the idea of moral desert, but also different from some other ideas of desert. For instance, the idea of reciprocal desert should be distinguished from the idea of skill-based desert. Consider the following example given by David Miller. A very poor archer manages to persuade the local archery club to let him take part in its annual tournament. By sheer good luck, he shot three arrows into the goal, something that he could not repeat in a million attempts. This poor archer, according to Miller, could not on this basis deserve the trophy that is presented to her (Miller 1999, p. 144). This kind of desert in athletic competition is supposed to be based on an athlete's skill or ability. The success of achieving a goal or bringing about an outcome by itself is insuffi cient for the athlete to deserve a relevant reward. In order to deserve such a reward, her success in bringing about an outcome has to manif est her skill or ability in a certain respect. 33 3.3 Notice that a reciprocity-b ased reason is not a purely altruistic reason, and it is not a purely egoistic reason either. As Rawls says, "reciprocity is a moral idea situated between impartiality, which is altruistic, on the one side and mutual advantage on the other." (Rawls 2003, p. 77) The diff erence among the three kinds of reasons might be illustrated as follows. For example, B is a complete stranger to A, and A saved B's lif e when she was drowning. The lif e-saving act performed by A might be based on a purely altruistic reason. On another occasion, A treated B well because she wanted B to help her in the future. Here the reason for A's action is egoistic. A reason based on reciprocity is diff erent from either kind of reason. We oftentimes tip for good services, even if we would never see the waitress again in our life. This is a nice case of an action based on reciprocity. As Allan Gibbard says in his article "Constructing Justice", Ifl return fa vor for fa vor, I may be doing so in pursuit of my own advantage, as a means to keep the fa vors rolling. My motivation might, though, be more intrinsically reciprocal: I might be decent to him because he has been decent to me. I might pref er treating another well who has treated me well, even if he has no further power to aff ect me. We tip for good service in strange restaurants. (Gibbard 1991, p. 266) It is important to note that Raw Is uses the term "reciprocity" to express a concept that is broader than the concept Gibbard and I intend to express by the term "reciprocity" . In Rawls's terminology, the principle of reciprocity is a general ethical principle in human life, which says that we ought to respond to others' practices (attitudes or actions) in kind. 1 4 In his terminology, to give a 14 Rawls also observes the psychological fact that hwnan beings do have the tendency to answer in kind and the tendency already plays an extremely crucial role in hwnan life. He says, "Without it OliT nature would be very different and fr uitful and social cooperation fr agile if not person a moral praise he deserves could be an instance of reciprocity. But in my terminology (as well as Gibbard's), to say that A has a reciprocity-based reason to treat B is to say that A has a reason to benifit Bin return. Nevertheless, what Rawls says about the broader concept of reciprocity also applies to the narrower concept of reciprocity - that is, a reciprocity-b ased reason is neither a purely altruistic reason, nor a purely egoistic reason. 3.4 I have thus fa r argued that the fa ct that an agent A brings about an outcome S that is beneficial to another agent B serves as a desert-based reason for B to benefit A in return in a fitting circumstance. And it is reasonable to assume that other things being equal, the greater the benefit A brought to B is, the greater the benefit B should return to A is - in other words, B should benefit A in return with a proportional amount of goods. Compare the case where Marie happened to save Benjamin when she was fishing with the following case: Marie happened to hold a door open for Benjamin who was a little behind her. In both cases, Benjamin ought to recognize the fa ct that Marie benefited him as a reason for benefiting Marie in return in a fitting circumstance. But it seems that whereas he ought to benefit Marie in the lif esaving case, he should benefit her less in return in the door- opening case. For instance, although it is reasonable to say that the fa ct that Marie saved Benjamin's lif e provides him with a reason for lending some money impossible .... If we answered love with hate, or carne to dislike those who acted fairly toward us, or were aver se to activities that furthered OliT good, a comrnnnity would soon dissolve. Beings with a different psychology either have never existed or must soon have disappeared in the colU'se of evolution. A capacity for a sense of justice built up by responses in kind would appear to be a condition of h wnan sociability." (Rawls 1971, p. 495) 35 (or even giving some money) to Marie when she has financial problem, it isn't reasonable to say that the fa ct that Marie held the door open for Benjamin provides him with a reason for lending some money to her. In the af orementioned examples, the benefit recipient is an individual agent. But it seems to me that the principle of reciprocity can also apply to cases in which the benefit recipient is the society as a whole. By replacing a benefit recipient B with the society, we then have the following thesis: the fa ct that an agent A brings about an outcome S that is beneficial to the society with an "amount of social goods" serves as a desert-based reason for distributing a proportional amount of goods to A. It may be worth mentioning a diff erence between the case where an individual benefit recipient ought to remunerate a benefit giver in return and the case where the society as the benefit recipient ought to distribute some goods to a benefit giver. In the first case, it is the particular benefit recipient (rather than other individ uals) who has a desert-based reason to remunerate the benefit giver. In the second case, the benefit giver might benefit the society through benefiting some individual s. But it doesn't follow that only those benefited individuals have a desert-based reason to remunerate the benefit giver. Other citizens seem to have a desert-based reason to remunerate the benefit giver as well. For example, a care-taker has taken very good care of a person with severe disability. Suppose that the person who has been taken care of is very poor. She cannot aff ord to remunerate the care-taker. Does that mean that the care-taker should not receive 36 remuneration or receive lower remuneration? No. It seems to me that other citizens in the society ought to remunerate the care-taker decently. 4 Given the idea of contribution defined in section 2, and given the Principle of Reciprocity, it seems to follow that the fa ct that an agent Al makes more contributions to the society than another agent A2 does serves as a desert-based reason for distributing more goods to Al than to A2, even when the diff erence in contributions is a result of luck. For instance, in cases (2), (3), and (4), Smith makes more contributions than Tom does, although the diff erence is not a result of their efforts, but a result of luck, constitutive luck in (2), circumstantial luck in (3), and opportunistic luck in (4). But the fa ct that Smith makes more contributions than Tom does serves as a desert-based reason for distributing more goods to Smith. I agree that sometimes there is no desert-based reason for distributing goods to a person if the beneficial outcome is not brought out by the person, but is completely a matter ofluck to her. For example, suppose that Smith wants to bake a lot of bread. A naughty angel who likes Smith plans to help him. The angel then produces 1,000 loaves of bread per day in Smith's kitchen. The exercise of Smith's physical or mental abilities doesn't play any significant causal role in the process of bread baking. I agree with luck egalitarians that we don't have a desert-based reason for just if ying an unequal distribution between Smith and Tom. However, the reason why there isn't such a desert-based reason is not 37 that the outcome is a result ofluck, but that Smith doesn't make more contributions than Tom in this case. Then consider cases (5) and (6) again. In (5), whereas Tom is as free as a normal person is, Smith is forced to work hard at bread baking. Smith doesn't want to work hard at all; if he were as free as Tom is, he would only bake 10 loaves of bread per day as Tom did. In (6), Smith is an unwilling workaholic. His desire to work hard always moves him to act. But he doesn't want the desire to move him to act. He very much wants to be a person who enjoys a relaxing lif estyle. Given that Smith is a workaholic, he bakes 1, 000 loaves of bread per day - ifhe were as free as Tom is, he would have baked only 10 loaves of bread per day. I agree with some luck egalitarians that in both ( 5) and ( 6), there might be a reason for an unequal distribution between Smith and Tom. However, the reason is not that Smith woks harder in the process of bread baking than Tom, but that Smith makes more contributions than Tom does. Of course, even though the fa ct that Smith makes more contributions than Tom does serves as a desert-based reason for distributing more goods to Smith, it might still be true that the fa ct that Smith works harder in the process of bread baking than Tom serves as another desert-based reason for distributing more goods to Smith. That is, recognizing the contribution-based reason doesn't by itself nullify the putative eff ort-based reason. In order to reject the putative effort-based reason, some independent arguments must be provided. I will discuss this in Section 5. 38 Someone might say that in the following two examples, the diff erence between two persons' contrib utions does NOT seem to serve as a desert-based reason for just if ying the inequality between their distributive shares. (7) [External freedom*] Smith is as free as a "normal" person. He bakes 1, 000 loaves ofbread per day. In contrast, Tom is not free in the bread-baking process: a crazy guy pointed a gun at Tom's head and forced Tom to work less hard at bread baking. Consequently, Tom only bakes 10 loaves ofbread per day. But ifTom were as free as Smith, he would bake as much bread as Smith. (8) [Internal freedom*] Smith is as free as a "normal" person. He bakes 1, 000 loaves ofbread per day. Tom is unwillingly addicted to a relaxing lif estyle. His desire of having fun always moves him to act. As a result, Tom only bakes 10 loaves ofbread per day. However, he doesn't want the desire to move him to act; he very much wants to be a person who devotes himself to hard work. If he were as free as Smith is, he would bake as much bread as Smith does. In either (7) or (8), is there a contrib ution-based reason for distributing more goods to Smith? A positive answer to this question may make some people feel uneasy. However, the uneasy feeling doesn't indicate that the diff erence in contribution is not a reason for just if ying the inequality between Smith and Tom, but rather that there might be some other reasons for limiting the inequality for instance, the reason that we should compensate Tom for the burdens that he bears. :39 5 5.1 In the previous sections, I proposed a contribution-based ob jection to (LE). In the current section, I want to put forward my second objection to (LE), a difficulty -based objection. This objection presupposes two theses: (D 1) Whether it is more dif ficult to bring about a socially beneficial outcome S, than to bring about another socially beneficial outcome S2 is sometimes irrelevant to people's eff orts, but rather a matter of luck to them. (D2) The fa ct that it is more difficult to bring about a socially beneficial outcome S, than to bring about another socially beneficial outcome S 2 serves as a desert-based reason for just if ying an unequal distribution among people. Ifthe two theses are correct, then it seems to follow that difficulty-related fa cts can provide desert-based reasons for distribution among people, even when difficulty is a matter ofluck to them. 5.2 To illustrate my diffi culty-b ased objection, let's consider an example. Imagine that Smith is a bread-baker, and Tom is a shoemaker. Smith bakes and sells 1, 000 loaves of bread; Tom makes and sells 2 pairs of shoes. Suppose that Smith and Tom are equally hard-working, and enj oy equal constitutive luck and circumstantial luck. Besides, let's assume that 1, 000 loaves ofbread and 2 pairs of shoes are equally beneficial. Compare scenario (a) and scenario (b). In (a), suppose that it is more dif ficult to bake 1, 000 loaves of bread than to make 2 pairs of shoes in the sense that under normal circumstance, a person with average skill would exert more eff orts in baking 1, 000 loaves of bread than in making 2 pairs of shoes. In scenario (b), suppose that it is easier to bake 1, 000 40 loaves of bread than to make 2 pairs of shoes - probably thanks to the invention of some efficient ovens, a person with average skill would exert less eff ort in baking 1,000 loaves of bread than in making 2 pairs of shoes under normal circumstance. It is not hard to see that (D 1) is quite plausible. Cases (a) and (b) are two particular examples of instantiating the truth of (D1 ). In case (a), the fa ct that to bake 1,000 loaves of bread is more difficult than to make 2 pairs of shoes, is a matter ofluck to both Smith and Tom. Similarly, in case (b), the fa ct that to bake 1, 000 loaves of bread is less difficult than to make 2 pairs of shoes, is a matter of luck to Smith and Tom. Smith and Tom are equally hard working, but the difficulty in baking 10 00 loaves of bread (or the dif ficulty in making 2 pairs of shoes) is barely relevant to people's eff orts, but rather a matter ofluck. It is intuitively true that the fa ct about the difficulty of a job is different from the fa ct about how much effort a particular benefit giver has to exert, or how great a burden a particular benefit-giver has to bear, in order to fulfill the job requirements. For instance, a talented mathematician might exert little ef fort in solving a very difficult mathematical problem; a highly patient caretaker who takes care of a person with disability might handle her caretaking job well, which is very burdensome according to an objective standard, with relatively little eff ort. On the other hand, a person who lacks mathematical talent might find it hard to solve a very simple mathematical problem; an impatient caretaker might feel very upset in engaging in a caretaking task that is quite easy for more patient people. 41 5.3 A luck egalitarian would deny that in either (a) or (b), there is a desert-based reason for just if ying an unequal distribution between Smith and Tom. By contrast, according to (D2), in case (a), the fa ct that it is more difficult to bake 1, 000 loaves of bread than to make 2 pairs of shoes serves as a desert-based reason for just if ying a distribution where Smith is better off than Tom; in case (b), the fa ct that it is less dif ficult to bake 1,000 loaves of bread than to make 2 pairs of shoes serves as a desert-based reason for just if ying a distribution where Smith is worse off than Tom. The truth of (D2) is rooted in a fundamental moral principle regarding the difficulty level ofbenefiting: for a benefit giver A and a benefit recipient B, how much benefit B ought to return to A depends in a proportional way upon how difficult it is for a "normal" human being to bring about the beneficial outcome B received. We can appreciate the truth of this principle by considering some specific cases in individual ethical lif e. Let me give one example. One day, Linda and Marie each do a fa vor to you. Linda takes care of your three-month baby for 6 hours; and Marie hangs out with you in a national park for the same length of time. Suppose that the utility you receive from having someone look after your baby is the same as the utility you receive from hanging out with someone in a national park. But you still have a reason for expressing more gratitude to Linda than to Marie - suppose that it is more dif ficult to look after a baby for 6 hours than to hang out in a national park for 6 hours. In our daily lif e and social practice, we can find many such examples. For a benefit giver A and a 42 benefit recipient B, B ought to regard the fa ct about how difficult it is for a human being to bring about the beneficial state of affairs that is brought about by A as a basis for her reactive attitude to A. Notice that even if it is more dif ficult to look after a baby than to hang out in a national park, it doesn't follow that Linda actually expends greater effort in looking after your baby than Marie does in hanging out with you in the national park. Perhaps Linda enjoys looking after a baby very much; or she is good at multitasking - she could work on her own business, and at the same time watch the baby very carefully. Nevertheless, even if Linda and Marie expend an equal amount of effort in benefiting you, you still ought to express more gratitude to Linda than to Marie because it is objectively more dif ficult for a human being to take care of a baby than to have fun in a national park. In the above example, the benefit recipient is an individual agent. But it seems to me that the principle regarding the dif ficulty level of benefiting can also apply to cases in which the benefit recipient is the society as a whole. By replacing a benefit recipient B with the society, we can easily get the thesis (D2): the fa ct that it is more difficult to bring about a socially beneficial outcome S, than to bring about another socially beneficial outcome S 2 serves as a desert based reason for just if ying an unequal distribution among people. Given the truth of(Dl) and (D2), it seems to follow that diffic ulty-related fa cts can provide desert-based reasons for distribution among people, even though dif ficulty is oftentimes irrelevant to people's actual eff ort, but a matter of luck to them. We then have another objection to (LE). 43 6 So fa r I've presented two argument s against LE, and have proposed two desert based reasons for distributions, the contribution-based reason and the difficulty based reason. Each reason contrasts with an effort-based reason. Moreover, as I've proposed in section 4, whereas some luck egalitarians contend that there is an effort-based reason for just if ying a distribution, I deny that individual s' actual eff orts can serve as a basis for distributive justice. I admit that, in many circumstances, a benefit recipient ought to recognize the amount of effort that a person spends in benefiting her as a basis for her reactive attitude toward the person. For example, both your friend Smith and your friend Tom send you a birthday gift. While Smith gives you an expensive watch, Tom only gives you an ordinary T-shirt. Assume that Smith is much richer than Tom, and buying the watch doesn't cost him much money, energy or time. Tom, however, makes the T -shirt by himself In this case, you wouldn't show more gratitude to Smith, although an expensive watch might be more beneficial to you than an ordinary T-shirt. On the contrary, you might feel more grateful to Tom, and your feeling is just ifia ble. The fa ct that Tom spends more effort in benefiting you by itself gives you a reason for being more grateful to him. Usually, we are more grateful to a person who makes more effort to benefit us. This is not only true with respect to how we ought to treat people who has intimate relationship with us, such as parents, friends and siblings, but also true with strangers who have no special relationship with us. A stranger 44 holds a door open for you when you are a little behind her. In one case, she makes a great eff ort to open the door for you since she is not very strong. In another case, she, as a very strong person, does the same fa vor to you without exerting any significant effort at all. Intuitively, the fa ct that she makes much more effort to open the door in the first case than in the second case serves as a reason for you to be more gratef ul to her in the first case than in the second one. However, in the context of distributive justice, our intuition concerning the role of individ uals' actual eff ort is quite diff erent. Consider the following example. Imagine that there are four bread bakers in the same street where you live, Al, A2, A3 and A4. Al is very productive, A2 and A3 are "ordinarily '' productive, and A4 is very clumsy. For instance, it takes Al one minute to produce one loaf of bread, A2 or A3 ten minutes, and A4 one hour. Suppose that the bread baked by the four bread bakers is equally tasty and healthy. Now you are considering how much money each bread baker ought to get from selling one loaf of bread. Does the fa ct that A4 expends more effort than the others and Al expends less effort than the others serve as a reason for just if ying a distribution where A4 receives more money than the others and Al receives less money than the others? We have a strong tendency to give a negative answer to this question. Indeed, the amount of effort that an individual actually spends in bringing about a beneficial outcome cannot serve as a basis for remunerating her. Remunerating people by appeal to the amount of their actual effort might lead to some counterintuitive results. As Miller says, all other things being equal, "We would have to pay clumsy people more than quick and skillful people for making 45 the same object, since the effort involved is greater. We would have to remunerate people who disliked their jobs more generously than people doing the same work who en joyed doing it." (Miller 1999, p. 18 3) So our intuition about whether individuals' actual effort could serve as a basis for distributive justice comes apart from our intuition about whether individ uals' actual effort could serve as a basis for a benefit recipient's reactive attitude toward her benefit giver. Here are two reasons that might be helpful to explain the disparity. First, in the circumstance of social cooperation, a benefit giver's knowledge that the benefit recipient would remunerate her according to her actual effort would weaken the benefit giver's motivation to become more skillf ul and work more efficiently. Of course, in many circumstances, we reward people for making great eff ort. As Miller says, "We congratulate people for making an effort - struggling to the top of Ben Nevis, for example - and we may give a prize to the hardest-working child in the class." (Miller 19 99, p. 183.) But in these cases, brave mountain climbers or diligent children would have a strong motivation to achieve their goals with as little eff ort as possible, even if they know that they would be rewarded for making great eff ort. Let me discuss the second reason why individual s' actual ef fort shouldn't serve as a basis for distributive justice. In individual ethical life, it is not the case that, in any circumstance, a benefit recipient ought to recognize the amount of eff ort that her benefit giver expends in benefiting her as a basis for her reactive attitude toward the benefit giver. The two cases discussed above, i.e. the birthday gift case and the door-opening case, are both concerned with a 46 particular kind of benefiting relation. That is, they both involve the fa ct that a benefit giver benefits a benefit recipient fo r the sake of the reci pient. We might call this particular kind of reciprocal relation "a caring relation". A caring relation could be either an intimate relation or not. Two strangers can stand in a caring relation with each other as well, as the door-opening case has shown. For a benefit giver A to stand in a caring relation with a benefit recipient B, A must not only benefit B, but also benefit B on the basis of the fa ct that she cares about B. But if a benefit giver benefits a benefit recipient without the attitude of caring, we wouldn't have a strong intuition that the benefit recipient ought to recognize the amount of effort exerted by the benefit giver as a basis for her reactive attitude toward the benefit giver. Recall the case where Marie happened to hold a door open for Benjamin. Suppose that she intended to hold the door open for her friend Linda who was a little behind Benjamin. Marie was even a little unwilling to hold the door open for Benjamin. In this circumstance, in regard to whatever reactive attitude Benjamin ought to have toward Marie, it doesn't matter how much eff ort Marie actually expends in holding the door open. It is worth noticing that in the context of distributive justice, the relation between a benefit giver and a benefit recipient is not a caring relation. For example, most of us benefit from people who work in Google or Microsoft. But surely they do not benefit us "f or us;" they do not benefit us on the basis of caring about us. 47 7 Let's return to the libertarian society introduced at the beginning of this chapter. In that society, a super model earns ten million dollars a year, but a caretaker earns two thousand dollars a year. This kind of inequality is very common in many societies in the real world. The large inequality between the better off (e.g. super models and sports stars) and the worse off (e.g. care-takers and dirty industry workers) ought to be reduced for the sake of justice. But the injustice involved in the large inequality between those people who earn more in the libertarian system and those who earn less is not because, as luck egalitarians assume, the inequality is sensitive to people's luck. Ins tead, there are two other reasons why the large inequality is unjust: 1) the distribution isn't sensitive to how the better-offs contributions compare to the worse-offs contributions; and 2) the distribution isn't sensitive to how the difficulty of the better-offs occupation compares to the difficulty of the worse-offs occupation. [Contribution] In the supermodel vs. caretaker case, while the supermodel makes one million people feel pleasant for an hour through her beauty, the caretaker takes good care of a person with disability in the process of overcoming many difficulties. I really doubt that the outcome where one million people feel pleasant for an hour is more beneficial than the outcome where a single person with disability is taken care of by others. We need to reflect upon the important issue regarding how the better-offs contributions compare to the worse-offs contributions in a real society. In order to address this issue, we need to have a theory of value. Surely how the value of one outcome x compares to the 48 value of another outcome y cannot be determined by how the market value of x compares to the market value ofy. This is not only because an individual consumer might irrationally pay for some product or service that is of no value or very low value, but also because, more importantly, the market system by itself cannot guarantee that a product's market value truly reflects its objective value. For example, as some scholars have argued, for any three goods, x, y and z, and any three persons Al, A2 and As, even if any individual 's personal preference of ranking the three goods is rational, it might be case that the collective preference of ranking the three goods is irrational. Suppose that x indicates being taken care of by a caretaker for two hours, y indicates listening to a lecture on poetry for two hours, and z indicates watching a superm odel's show for two hours. Al's pref erential ranking of the three goods is: x > y > z; A2's ranking is: y > z > x; AS's ranking is: z > x > y. Suppose that any person's ranking is rational, given some contingent fa cts about the person's personal conditions and circumstances. However, assuming the collective preference is determined by the maj ority rule, the collective pref erence seems to be irrational in the following sense. First, we have x > y, since according to both Al's ranking and AS's ranking, x > y; second, we have y > z, since according to both Al's ranking and A2's ranking, y > z. Then we get the conclusion that x > z, given the transitivity principle of rational pref erence (i.e. the principle that if x > y and y > z, then it must be the case that x > z ). On the other hand, according to both A2's ranking and AS's ranking, x < z. This contradicts the transitivity principle. Th us, the collective preference seems irrational. Some people might 49 take this kind of case as an argument for the hy pothesis that there isn't a standard of determining how one good's social value compares to another good's social value. But this understanding of this kind of case is fal se. The disparity between an individ ual's rational pref erence and the irrational collective preference doesn't imply that there isn't an objective standard of determining how the social value of one good compares to the social value of another good. Instead, it only implies that the maj ority rule that is used in any libertarian market system cannot help to determine the ranking of the social values of diff erent goods. [Difficulty] It goes without saying that the caretaker is engaging in a more difficult activity than the supermodel. We may need a general theory to determine how the dif ficulty of one occupation compares to the difficulty of another occupation, in order to justif y unequal distributions that appeal to difficulty -based reasons. But in this pro ject, I won't give a substantive theory of social values or a substantive theory of dif ficulty -t his could be a topic of another pro ject. Here it is reasonable to assume that some jobs (such as being a caretaker) are more difficult than others (such as being a model). Similarly, it is unjust that the workers who engage in dirty industries that are damaging to their health actually receive extremely low salaries. There might be at least two fa ctors that lead to the un just result. First, the workers have no skills or opportunit ies for better professions, so they have to engage in the dirty industries. Second, there are many such workers; so the boss can always find a replacement if one worker quits. Theref ore, the workers have very 50 low bargaining power. However, this logic of free market system is seriously unjust. It doesn't solely matter whether a person has no skill for better professions or whether there are many such persons. As long as engaging in a certain practice places heavy burden on a human being, the society ought to find a way to redistribute the social wealth in order to compensate the person who actually engages in dif ficult and burdensome practice. There are at least two possible ways of doing this. First, the society might find a way to force the company to raise the salaries of those workers. Second, the government might compensate the workers directly (through raising the tax rate and then redistributing the wealth to those workers). Figuring out which is the more ef ficient way of reaching the same goal is a matter for empirical investigation. At the end of this chapter, I want to note in passing that I am not saying that the inequality between, for example, the supermodel and the caretaker ought to be completely eliminated. Maybe there are some reasons to just if y an unequal distribution between a supermodel and a caretaker. But the inequality should not be so large 1 5 15 The fact that the inequality between the supermodel and the caretaker ought to be reduced doesn't imply that an individual person ought not to spend money watching the superm odel's show. But the individuals ought to vote for a policy that aims to reduce the inequality. 51 Chapter 3 Sufficiency and Distributional Equality Sufficientarianism has become more and more influential in recent decades. Sufficientarianism can be form ulated as the con junction of two theses: 1) a primary reason for any distribution is to make the citizens reach or get closer to the so-called "sufficiency threshold"; and 2) if a person reaches the sufficiency threshold, then there are no further reasons to alter a distribution for the sake of improving her well-being. Although a notion of the sufficiency threshold is central to any kind of sufficientarianism, different particular versions of suffi cientarianism hold diff erent views about the sufficiency threshold. In Sections 1-5, I first examine five fa mous accounts of the sufficiency threshold: suf ficiency as satiety, sufficiency as the satisfaction of basic needs, Crisp's account of sufficiency that appeals to an impartial spectator's compassion, Frankf urt's account of sufficiency that appeals to the pursuit of one's lif e pro ject, and Anderson's account of sufficiency as a list of personal goods. I argue that all the five accounts of sufficiency fa il to satisf y both of the two theses (although they may satisf y one of the two theses). Although none of the five accounts is tenable, it doesn't follow that sufficientarianism per se is fa lse. In Section 6, I instead propose an account of what I call "comparative sufficiency" , which understands suf ficiency as the level of social average, i.e. the average wealth per capita. In my view, having a person reach or get closer to the level of social average serves as a primary sufficiency-b ased reason for distributing some goods to her. 52 1 There are diff erent versions of suffi cientarianism. What those theories have in common is that they are committed to two principles as follows: (S l) a primary reason for distribution is to make the citizens reach or get closer to the sufficiency threshold; (S2) If a person reaches the suf ficiency threshold, then there wouldn't be any further reasons for distribution for the sake of improving her well being. Although diff erent versions of sufficientarianism all adopt some notion of suffi ciency, they disagree on what particular kind of suf ficiency can satisf y both (S l) and (S2). To avoid unneces sary and irrelevant complexities, the discussion in this chapter mainly focuses on the sufficiency threshold in regard to economic goods. Moreover, for the sake of convenience, the discussion is only concerned with the distributive circumstance where there is no diff erence in people's contributions or the difficulty level of people's beneficial practice. The first possible way of defining the sufficiency threshold that seems to be attractive is to define it as "the line of satiety". A person reaches the line of satiety in respect of a good if and only if having more of the good would not constitute any improvement for her well-being. Take the consumption of ice cream as an example. You find it very pleasant by having your first bite of ice cream, and the same is true with your second bite, third bite, and so on. But you could reach a point where consuming more ice cream would not improve your well-being. Perhaps your stomach is simply full, and you will not gain any more 53 pleasure. You even could reach a point where you will feel sick of having more ice cream. Probably you will feel that way when you already consume 500 grams of ice cream; then it can be said that you reach the line of satiety in respect of the consumption of ice cream. Similarly, a person reaches the line of satiety in respect of economic goods when having more wouldn't improve her well-being in any way. For instance, a person's economic goods satiate her when having more money does not make her happier, bring her more freedom, or increase her capabilities, but just causes her more difficulties in financial management. Whether a person reaches the line of satiety in respect of any good doesn't depend upon how she compares to others. For example, whether you reach the line of satiety in respect of the consumption of ice cream doesn't depend upon how the amount of ice cream consumed by you compares to the amount of ice cream consumed by others. How much ice cream would satiate you depends only upon your own biological, physiological and psychological features. The same is true with economic goods 1 6 To define the sufficiency threshold as the line of satiety in respect of economic goods is to affirm the following claim: a person reaches the sufficiency threshold in respect of economic goods if and only if she reaches a point where having more economic goods would not constitute any improvement for her well-being. I argue that suf ficiency as satiety may satisf y (S2), but cannot satisf y (Sl ). 16 In some cases, a person couldn't reach the line of satiety wlless others had less. But the fact that others have less doesn't by itself make the person reach the line of satiety. Instead, she would reach the line of society because she would have more if others have less. 54 Consider (S2) first. When a person reaches the line of satiety, then it automatically follows that distributing more goods to her would be unnecessa ry for the sake of justice, since having more economic goods would not constitute any improvement in her well-being. However, sufficiency as satiety cannot satisf y (Sl ). Since the level of satiety in respect of economic goods is extremely high or not even reachable, reaching the level of satiety or getting closer to it cannot bear any special moral significance in the context of distributive justice. For a very limited list of goods, such as healthy food, sanitary water and fresh air, the line of satiety is not that high. But for economic goods (or most goods), the line of satiety could be extremely high, if not unapproachab le. Consider the lif e of a person who earns one billion dollars per year. Even this billionaire could rationally want more income to for her and her loved ones. For example, suppose that some extremely saf e underground mansions have been built in order to keep the residents from being hurt by any natural and social disasters. But the mansions are extremely expensive. Billionaires would probably consider buying one for themselves and their loved ones. Or imagine that, thanks to the development of technology, people are capable of delaying aging for 20 years. After taking some drugs, a woman who is supposed to be forty years old would turn out to be as young and healthy as a twenty-year girl. Moreover, taking the drugs would not produce any negative side eff ects. But the drugs are very expensive. Although for the time being no such medical resource is available, it is rational for a rich person to 55 consider, "What if in the near future, people can delay aging? I should save enough money for the medical resource." Now look at how things are in the real world. Take the US as an example. Maybe only a few people, such as Bill Gates and Warren Buff ett, reach the line of satiety, or get very close to it. Almost everyone else would be better off by having more economic goods. That is, the amount of the total wealth of the society is far less than the amount that could make most people reach the line of satiety. Given the limited amount of total wealth, the reasons that the government actually uses to just if y the distributions of economic goods are seldom concerned with having people reach the line of satiety. Moreover, given the limited amount of social wealth, whether some people are closer to the line of satiety than others is normally irrelevant to the reasons for distributing wealth among them. Supp ose that a person would reach the line of satiety if she has one billion dollars. We are going to distribute 5,000 dollars between two persons, one has 10 ,000 dollars and the other has 15 ,000 dollars. Some sufficientarians might suggest distributing the money to the worse off person by appealing to the fa ct that the better off person is closer to the line of satiety. However, the better off can jus tifiably complain, "Am I that much closer to the line of satiety?! Compared to the diff erence between the billionaires and me, the diff erence between me and those even worse off is almost negligible." Indeed, when both parties are very far from the line of satiety, it would be unreasonable to justif y assigning priority to the worse off by appealing to the fa ct that the better off is closer to the line. So, if the sufficiency threshold is 56 understood as the line of satiety, it would be odd to say that having a person reach or even get closer to the sufficiency threshold is a primary reason for distribution. Theref ore, as fa r as economic goods are concerned, suf ficientarians have to define the sufficiency threshold as some point below the line of satiety, so that when a person reaches the suf ficiency threshold, she could still be better off by having more economic goods. 2 Suppose that we are going to distribute some food between two persons. And further suppose that, bef ore the distribution, one person is hungry while the other is already full. To whom should we distribute the food? Most of us have the intuition that we should distribute the food to the hungry person. Even if we are choosing between providing ordinary food to the hungry person and bringing super delicious good to the person who is full, we would still have a stronger reason for the former action. What is of special moral importance here is to satisf y a person's basic needs (such as the freedom from hunger). Basic needs are those needs that are absolutely ess ential for a person's survival and reproduction, such as sanitary water, clean air and healthy food. We have to admit that, although diff erent persons have diff erent needs, they all share some basic needs. As Robert Huseby says, "Even though needs in general vary between persons, (nearly) everyone needs food, clothing, and shelter." These are basic human needs. Having these needs met is in most cases also a precondition for having one's further needs and pref erences met." (Huseby 2010, p. 180) 57 Moreover, whether a person's basic needs have been satisfied does not depend upon how she compares to others in any respect. As Andrei Marmor says, "These needs are determined by our biological makeup and environmental realities .... We can say that these biological needs are absolute, in that they are entirely determined by necessities of survival."(Ma rmor 2003, p. 131) Suf ficiency as the satisf action of basic needs can satisf y (S1 ). It is true that satisf ying a person's basic needs provides a fundamental and distinctive reason for distributing some goods to her. However, suf ficiency as the satisf action of basic needs cannot satisf y (S2). Suppose that everyone in the society has reached the level of basic needs satisf action. But significant inequality is still pervasive; some are as well off as those very rich people in the US, whereas some just meet the level of subsistence. Do we think that no further distribution is neces sary in the society? Probably not. So we can conclude that whereas it is unrealistic to attached special moral importance to reaching the line of satiety, the mere satisfac tion of basic needs is too low to be reasonably counted as the sufficiency threshold. The sufficientarian should define the threshold of sufficiency somewhere between the satisf action of basic needs and the satiety line. 3 Roger Crisp puts forward an account of the sufficiency threshold that lies between the satisf action of basic needs and the satiety line. He connects the idea of suf ficiency with the virtue of compassion, and endorses what he calls the "sufficiency principle": "compassion for any being B is appropriate up to the point 58 at which B has a level of welf are such that B can live a life which is sufficiently good." (Crisp 2003a, p. 762) According to Crisp, the feeling of compassion that helps to track sufficiency is not ordinary people's feeling of compassion, but what he calls an impartial spectator's feeling of compassion 1 7 Following Adam Smith, he takes compassion to be one of the original passions of human nature or a basic human emotion, like fear or anger. (Crisp 2008, p. 241) As Smith says (and Crisp quotes) in The Theory of Moral Sentiments: How soever selfish a man may be, there are evidently some principles in his nature, which interest him in the fortune of others, and render their happiness necessary to him, though he derives nothing from it except the pleasure of seeing it. Of this kind is pity or compassion, the emotion which we feel for the misery of others, when we either see it, or are made to conceive it in a very lively manner. (Smith 19 76, p. 8) In his writings, Crisp emphasizes the instinctive nature of compassion. For instance, he says, "In its most primitive form, as in the case of young babies who appear to be distressed by the sound of other babies' crying (so-called 'emotional contagion'), it is little more than instinctive distress caused by, ... , the suff ering of others." (Crisp 2008, p. 240) He suggests that we should not be misled by the fa ct that compassion is often takes on a subtle and sophisticated form or is felt for complex reasons into viewing it in an excessively narrow and cognitivist way. 17 It is worth noting that, although Crisp relates sufficiency to an impartial spectator' s comp assion, he never argues that the compassion can track all kinds of justice. For example, Crisp probably would agree that the comp assion cannot track justice in the following cases. (a) Other things being equal, it is nnjust that a person who makes more contributions to the society receives less than another person who makes fewer contributions, no matter whether an imp artial spectator would feel comp assion for the worse off in question. (b) It is nnjust that a person fails to get a job position because of her race. But the injustice doesn't have much to do with an imp artial spectator' s compassion. (c) It is justifiable to prnrish a person who shoots many innocent victims to death even though an impartial spectator would feel compassion for the crimin al. (d) Suppose that neither A's basic needs nor B's have been satisfied. A is badly off becaus e of bad luck, but B is badly off because of his own autonomous choice. Suppose that, given limited resolU"ces, only one person's basic needs can be met. It seems that A ought to be assigned priority, though an impartial spectator would have equal compassion for both persons. 59 (Crisp 2008, p. 24 1) Although Crisp doesn't believe that much analysis of the concept of compassion is necessary or even useful, he points out a crucial feature of compassion: it is a concern for another person's suff ering. (Crisp 2003b, p. 119 ) He thinks that the feeling of compassion is tied to the notion of a lack; where the individual in question has enough, compassion seems to give out. (Crisp 2003a, p. 761) Crisp doesn't offer extensive analysis on the concept of impartial spectator, either. Besides, in articulating a scheme of distributive justice by using the notion of an impartial spectator, Crisp is deeply influenced by Smith, just as he is in connecting justice with compassion. Smith says (and Crisp quotes), We can never survey our own sentiments and motives; we can never form any judgment concerning them; unless we remove ourselves, as it were, from our own natural station, and endeavor to view them as at a certain distance from us ... We endeavor to examine our own conduct as we imagine any other fa ir and impartial spectator would examine it. (Smith 19 76, p. 11 0.) According to Crisp's writings, the most important feature of an impartial spectator is that she is able to put herself in the shoes of all. "The ordinarily compassionate person, for example, is likely to be moved especially by suff ering that he or she witnesses or by the plight of those close to him or her. That cannot be true of the spectator." (Crisp 2003a, p. 757) Although Crisp doesn't give much analysis of compassion or of the impartial spectator, he indeed captures the essential feature of the two ideas. I agree with him that the most important feature of an impartial spectator is that she is able to put herself in the shoes of all and a crucial feature of compassion is that it is a concern for another's suff ering. However, to define the suf ficiency 60 threshold by appealing to an impartial spectator's compassion isn't a plausible approach, because (S2 ) couldn't be satisfied. Let's imagine the life of Mary, who lives in the downtown area of LA. Mary earns 2000 dollars per month. She can only aff ord to rent a studio, to buy basically healthy food and to buy some cheap clothes. She cannot afford to travel a lot or have fun with her friends very often. But she is both physically and mentally healthy. Would an impartial spectator feel compassion for Mary? Probably not. She may feel compassion for a homeless person or a person who lives in a poor rural place where sanitary water isn't accessible. Compassion, as a distinctive emotion, involves a significant degree of uneasiness or even sadness. An impartial spectator might have no such emotional attitude for Mary. Nevertheless, we may still think that Mary's lif e is not sufficiently good, and her well-being ought to be improved. In connecting the idea of sufficiency with an impartial spectator's compassion, Crisp contrasts rich people on Beverly Hills with very poor people whose basic needs are still not met. He might be right that, although an impartial spectator would have compassion for a very poor person whose basic needs are not met, she wouldn't feel compassion for those rich people on Beverly Hills, assuming they are healthy and happy. But it doesn't follow that the compassion can properly track suffi ciency, for an impartial spectator may not have compassion for a person who is better off than the very poor but worse off than the rich on Beverly Hills. 61 At some point, Crisp claims, "[O]n reflection, my own intuition is that, say, eighty years of high-quality life on this planet is enough, and plausibly more than enough, for any being." (Crisp 2003a, p. 762) Does he mean that a person reaches the sufficiency threshold on this planet only if she lives eighty years of high-quality life, or reaches some point below but close to that level? If he does mean something like that, then his theory strikes us as inconsistent. An impartial spectator wouldn't regard a person who lives seventy years of high- quality life, say a healthy, happy, flourishing and rewarding life, as a suffering one, assuming she is blessed with a peaceful death. The spectator may have no compassion for the person. 4 Another fa mous account of sufficiency that defines the sufficiency threshold in between the satisf action of basic needs and the satiety line is given by Frankfurt. He connects suf ficiency with a person's pursuit of her lif e project. In making this connection, Frankfurt first claims that to say a person has enough money means that it is reasonable for her to be content with having no more money than she has (Frankfurt 19 87, p. 37); 1 8 he then claims that it is reasonable for a person to 18 Here, I interpret Fra nkfliTt's writing with charity. Frankfrnt says, "To say a person has enough money means that he is content, or that it is reasonable for him to be content, with having no more money than he has. And to say this is, in turn, to say ... the person does not (or cannot reasonably) regard whatever (if anything) is nnsatisf}ring or distressing ab out his life as due to his having too little money ."(F rankfurt 1987, p. 37) Notice that fr om Frankfrnt' s writing, we can tell that he is not quite sill'e whether sufficiency is related to "actual contentment" or "reasonable contentmen t". Based on a charitable nnderstanding of his writing, we may consider him to relate sufficiency to reasonable contentment. Actual contentment cannot be a good sign for suffi ciency. Due to adaptive preferences, some people are actually content with their miserable lives. As Sen says, "A hopeless destitute with much poverty; or a downtrodden laborer living nnder exploitative economic arrangements, or a subjugated housewife in a society with entrenched gender 62 be content with what she has if and only if what she has is enough to enable her to carry out what she herself really cares about (in the sense of her life project) and whether what a person has is enough to carry out her project does not depend upon how she compares to others in any respect (Frankfurt 19 87, p. 23; Rosenberg, p. 66). Frankf urt is right in claiming that reasonable contentment might be helpful for tracking sufficiency. His description of the psychological state of being content with one's well-being is insightful, too. He says, "Saying that the person is content with what he has precludes his having an active interest in getting more. A contented person regards having more money as inessential to his being satisfied with his life." (Frankfurt 19 87, pp. 37-8 ) Moreover, he claims, and I agree, that it could be reasonable for a person to be content with her economic goods even if she would be better off by having more. In other words, it could be the case that a reasonably contented person still doesn't meet the line of satiety. 1 9 In fa ct, Frankfurt seems to contend that economic goods are insatiable. He says, From the supposition that a person tends to lose more and more interest in what he is consuming as his consumption of it increases, it plainly inequality , or a tyrannized citizen nnder brutal authoritarianism, may come to terms with her deprivati on. She may take whatever pleasill'e she can fr om small achievements, and adjust her desires to take note of feasibility (thereby helping the fulfillment of her ad justed desires). But her success in such adjustment would not make her deprivation go away. The metric of pleasill'e or desire may sometimes be quite inadequate in reflecting the extent of a person's substantive deprivati on." (Sen 2002, p. 82) 19 Once again, I interpret Frankfrnt's writing with charity. Frankfrnt says, "It is possible that someone who is content with the arnonnt of money he has might also be content with an even larger arnonnt of money." "From the supposition that a person is content with the amonnt of money he has it cannot even be inferred that he would not prefer to have more."(Fran kfliTt 1987, pp. 37-8) Frankfurt' s writing is open to various interpretations. One possible interpretation is: it is psychologically possible that a person who is actually content with her economic goods at the same time prefers to have more. Frankf liTt's view would be trivi alized through this interpretation. 63 cannot be inferred that he must also tend to lose interest in consumption itself or in the money that makes consumption possible. For there may always remain for him, no matter how tired he has become of what he has been doing, untried goods to be bought and fresh new pleasures to be enjoyed." (Frankfurt 19 87, p. 26) 2 0 Although Frankf urt is right in claiming that reasonable contentment might be helpful for tracking sufficiency, his substantive view about under what conditions it is reasonable for a person to be content with her well-being is fal se. It is quite doubtful that it is reasonable for a person to be content with what she has if what she has is enough to enable her to carry out her lif e project. Recall the case of Mary mentioned in the discussion of Crisp's account of suf ficiency. Suppose now that she is a non-established novelist. She aims at becoming a distinguished novelist and that is what she really cares about. Everyday she occupies herself with improving her writing skills, deepening her understanding of life, and sharpening her sense of aesthetics. Suppose that, by earning 2,000 dollars per month, Mary can aff ord to pursue her lif e project. She thus has no active interest in getting more money, and her attitude seems perfectly rational. Nevertheless, she could reasonably feel unsatisfied with her economic status. A person's life project is concerned with what she ought to do during her whole life; it is determined according to her interest, ability and individuality. A person could be reasonably unsatisfied with her well-being even when she already successfully carries out her life project. For example, although Vincent Van Gogh is a great artist, certainly he could reasonably be unsatisfied 20 It is worth noting that Frankfurt himself uses this passage to show that the marginal utility of money is not always diminishing. But I think it is better to regard the passage as showing the insatiability of money. 64 with his well-being in respect of his economic goods. The same is true with Ludwig Van Beethoven. Perhaps Frankf urt's supporters would defend his view by appealing to a case as follows. Back to Mary's case: Consider two possible scenarios. In the first scenario, poor Mary makes some new friends who are all very rich - say, the friends are as rich as the movie stars living in Beverly Hills. In the second scenario, Mary makes some new friends who are all very poor - even poorer than Mary. Consider this question, "Would it be reasonable for Mary to feel less content with her lif e in the first scenario than in the second scenario?" Frankf urt or his supporters would probably answer, "No. Mary would be unreasonable if her feeling about her lif e were sensitive to how she compares to others. If what she really cares about is to improve her writing skills and become a distinguished novelist, and what she has is enough to finally carry out what she really cares about, then she ought to be content with her life no matter how others in the society fa re." I agree that, in this particular case, Mary wouldn't be reasonable if how she feels about her lif e were aff ected by how she compares to her friends. However, this kind of case doesn't imply the conclusion that whether a person ought to feel content with her lif e is only concerned with what she really cares about (in the sense of a life project), since in general, how a person ought to feel about her lif e doesn't depend upon how she compares to others in this particular way. 65 5 So fa r I've examined four substantive accounts of sufficiency, and argued that none of them is plausible. Sufficiency as satiety is too demanding to satisf y (S 1 ); suf ficiency as the satisfaction of basic needs is too inadequate to satisf y (S2); neither defining the sufficiency threshold by appealing to an impartial spectator's compassion, as Crisp does, nor Frankfurt's particular version of sufficiency as reasonable contentment succeeds. In this section, I examine another suf ficientarian approach that specifies the sufficiency threshold in terms of a comprehensive list of goods. I take Anderson's characterization of a sufficiently decent lif e as an example. In her fa mous paper "What is the Point of Equality?" Anderson gives a paradigm list of capa bilities that should be equally distributed among citizens. As long as everyone has those capa bilities, no further redistribution would be necessary. According to her, people must be capable of functioning as equal citizens. To be capable of functioning as an equal citizen involves three aspects of individual functioning: as a human being, as a participant in a system of cooperative production, and as a citizen of a democratic state. To be capable of functioning as a human being requires eff ective access to the means of sustaining one's biological existence - food, shelter, clothing, medical care - and access to the basic conditions of human agency - knowledge of one's circumstances and options, the ability to deliberate about means and ends, the psychological conditions of autonomy, including the self -confidence to think and judge for ones elf, freedom of thought and movement. To be capable of functioning as an equal participant in a system of cooperative production requires eff ective access to the means of production, access to the education needed to develop one's talents, freedom of occupational choice, the right to make contracts and enter 66 into cooperative agreements with others, the right to receive fa ir value for one's labor, and recognition by others of one's productive contributions. To be capable of functioning as a citizen requires rights to political participation, such as freedom of speech and the franchise, and also eff ective access to the goods and relationships of civil society. This entails freedom of association, access to public spaces, such as roads, parks, and public accommo dations including public transportation, the postal service, and telecommunications. This also entails the social conditions of being accepted by others, such as the ability to appear in public without shame, and not being ascribed outcast status. (Anderson 19 99, p. 31 8) Roughly speaking, Anderson's paradigmatic list of the capabilities that should be equalized include the capabilities that people need to function in their personal and working lives and as citizens of a democratic society. Can we say that no further distribution for the sake of improving any individ ual's well-being would be neces sary as long as the equality of those capabilities is secure? To answer this question, consider this thought experiment. Imagine a good society where all of the capabilities that Anderson mentions are equally distributed among citizens. Let's call this society "Semi-utopia" . Consider this question: What if there is still an extremely large inequality in Semi-utopia? Does it matter? Or is it un just? Suppose that there are one thousand citizens in Semi-utopia, and there is a splendid space station. The citizens happen to discover the station. Through thorough research, they realize that a person may lead a fa ncy lif e on the station. For example, a person might arrange a party with dwellers from some other planets. Compare the following two cases. It is worth noting that in the cases the resource may include labor force, natural resources and the level of technology, 67 etc. To say that the resources are suffi cient for a society to do x is to say that the society has the labor force, natural resources and the level of technology to do x. (a) The resources of Semi-utopia can enable all the one thousand people to go to the space station once. But in reality, one hundred people can aff ord to go to the station ten times. Nobody else can aff ord to go there even one time. (b) The resources of Semi-utopia can enable all the one thousand people to go to the station once. In reality, everyone can aff ord to go there once. Which case is more just? My tentative answer is that, other things being equal, (b) is more just than (a). Indeed, Semi-utopia is already an extremely good society for us (here and now), but it seems that the inequality as such still matters in Semi-utopia. It is worth noting that maybe we (here and now) don't care about whether someone can aff ord to go to a space-station. But that's because we all know that the resources in this actual society are not sufficient to make even one percent of the whole population go to a space-station. Suppose that, in case (a), Mary is one of the nine hundred persons, who cannot aff ord to go to the station. Although Mary enjoys all of the capabilities listed by Anderson, there is still a reason for distribution for the sake of improving her well-being. So Anderson's account of suf ficiency as a comprehensive list of goods cannot satisf y (S2), if a person reaches the suf ficiency threshold, then there wouldn't be any further reasons for distributions designed to improve her well-being. In fa ct, by following the same logic, we can conceive an even higher standard of sufficiency, and argue that beyond the level of suffi ciency, further redistribution still seems to be necessary, for the sake of improving some 68 individ uals' well-being. Suppose that the resources of Semi-utopia can make everyone delay aging for 10 years. There are two available distributions. First, suppose that one hundred people can delay aging for 100 years, but nobody else can aff ord to delay aging at all; second, suppose that everyone can delay aging for 10 years. In this case, the government has a reason for choosing the second distribution for the sake of improving some individuals' well-being, i.e. delaying aging for 10 years. Nils Hotlug raises a similar objection to an account of sufficiency that appeals to a comprehensive list of goods. He says, Suppose that a given individual is at the threshold level ... Suppose also that, due to some (very fortunate) change in the world economy, everyone else becomes much better off than she is. So everyone but her now has a life of extreme luxury and happiness .... We would in fa ct attach greater value to benefits to her than to (f urther) benefits to others." (Holtug 2007, pp. 149 - 50) As with Semi-utopia cases, Holtug's case shows that it would be implausible to define the suf ficiency threshold independent of the total welf are of the whole society; a plausible idea of suf ficiency must be defined relative to the total welf are of a society. In fa ct, some sufficientarians have already attempted to combine suf ficiency with comparisons of this kind. For example, Robert Huseby, who also defines suf ficiency by appealing to reasonable contentment, says, "[T]here is reason to think that it is harder to be content if, for instance, many are much better off. This means that the higher the welf are level in society ... , the higher the threshold of sufficiency." (Huseby 2010, p. 182 ) 2 1 Although Huseby has 21 But Huseby' s writing seems to be misleading to some extent. The sentence "there is reason to think that it is harder to be content if, for instance, many are much better off " and the sentence "the higher the welfare level in society (or the higher the global welfare level), the higher the 69 noticed the fa ct that a plausible account of sufficiency must be defined relative to the total welf are of a society, he doesn't make it clear how to achieve this theoretic goal. In the next section, I propose a comparative account of suf ficiency - suf ficiency as the average wealth per capita, and argue that it can satisf y (S l) and (S2) at the same time. 6 6.1 Let's imagine, for the moment, that three people are going to share a cake. Suppose that they don't know any inf ormation about each other, except that they are all healthy and happy, and that, for each person, the more cake the better. Now they are considering how to distribute the cake fa irly. They would all agree that an equal distribution where everyone receives one third of the cake is fa ir. 22 Now back to the issue concerned with the distribution of economic goods among citizens within a society. Suppose that, (a) there is no information about how one's contributions compare to others'; (b) how the dif ficulty of the project that she is undertaking compares to the difficulty of the project that others are threshold of sufficiency" have very different meanings. Asswning a person's well-being is at a certain level, the fact that many are much better off doesn't imply the fact that the welfare level of the society is higher, since it could be case the many are much worse off at the same time. Besides, in some cases, although fewer people are better off than the person, but those people are much better off, the welfare level in the society might still be higher. 22 The moral reasoning of a person might be as follows. "I could either have less than one third of the cake, or more than one third, or exactly one third. How much I ought to receive depends upon what would happen if everyone receives the amount that I would receive. If everyone receives less than one third of the cake, then there would still be some arnonnt of cake left. That means I could have more. Since the more cake I have, the better it is for me, I ought not to receive less than one thir d. I ought not to receive more than one third, either, becaus e there wouldn't be enough cake to share if everyone receives more than one third. So I could reasonably feel content if and only if I receive exactly one third of the cake." Since everyone has the same line of reasoning, they would all agree that it would be fair if everyone receives exactly one third of the cake. 70 undertaking; and (c) economic goods of the same value would benefit each person to the same extent -p eople enj oy the same level of capacity to convert economic goods to welf are. Moreover, let me suppose that the society is a society of moderate scarcity, the resources of which are sufficient to satisf y everyone's basic needs. Then I propose that a person's economic goods reach the suf ficiency threshold if and only if she has the average wealth per capita, or to put it in another way, she reaches the level of social average. 2s 6.2 The comparative account of sufficiency can satisf y (S 1 ). As far as economic goods are concerned, making a person reach or get closer to the level of social average serves as a primary reason for distribution. As the cake-sharing case has shown, when no relevant information is available, having a person get the average amount per capita provides a reason for a certain distribution in regard to her. It is important to note that the reason is derived from a sense of fa irness rather than the intrinsic value of equality. To illustrate the diff erence, consider an even simpler case of cake sharing. Smith and Tom are going to share a cake. Suppose that they can either choose an unequal distribution, where Smith gets three quarters of the cake and Tom gets one quarter, or choose an equal distribution, where they each get half of the cake. Smith has two different reasons for choosing the equal distribution, although under the unequal distribution, he would have more cake. The first line of reasoning is as follows, "No information can be used to just ify treating Tom and me diff erently. So it might be unfair if 23 The level of social average could be defined in a more delicate way. But here I only want to show the point that the sufficiency threshold could be defined in a comparative way. 71 Tom receives only one quarter whereas I receive three quarters." The second line of reasoning is, "Tom would desire the equal distribution, and I should recognize his desire of equality as a reason for carrying out the equal distribution." The first line of reasoning is based on the value of fa irness, whereas the second line of reasoning is based on the intrinsic value of equality. Frankf urt also distinguishes the reason based on fa irness, or in Frankf urt's terminology, the standard of impartiality, from the reason based on the intrinsic value of equality. As he says, It does indeed make good sense to divide the cake equally, but the reason is not that equality en joys an inherent moral advantage over inequality. The reason the people ... should be treated the same lies in the stipulation that there is no relevant diff erence in what is known about them. They are not known to be relevantly alike, nor are they known to be relevantly diff erent. In other words, the relevant information that is available about each is exactly the same: namely, zero. That is why it would be improper to discriminate among them. The standard of equality does not prevail because it is reasonable or desirable for equality to be the defa ult. It prevails because the standard of impartiality requires that people are to be treated the same when, so fa r as we know, they are the same in every relevant way." (Frankfurt 2000, p. 98) Although Frankf urt is right that the reason based on fa irness or impartiality is diff erent from the reason based on the intrinsic value of equality, he is fa lse in claiming that equality of any sort has intrinsic value. In chapter 5, I give a detailed explanation on the intrinsic value of equality. But the general point here is that a comparative account of suf ficiency might have nothing to do with the intrinsic value of equality. From now on, by "the sufficiency-based reason", I only mean the reason that is concerned with having a person reach the level of social average and is derived from the value of fa irness. 72 The comparative account of sufficiency can satisf y (S2). If a person reaches the level of social average, then there wouldn't be further reasons for distribution for the sake of improving her well-being. Again, let's take the cake sharing case as an analogy. In that case, it is true that, if a person already receives one third of the cake, then it would be unnecessa ry to distribute more cake to her. Similarly, in the circumstance, where a given bundle of economic goods are distributed among a group of people, and no relevant information about those people is available except that everyone would be better off with having more economic goods, if a person already has the level of social average, then it would be unnecessary to distribute more economic goods to her, for the sake of improving her well-being. 6.3 In the above discussion, I assume that everyone has the same capacity to convert economic goods to welf are. But in a real world, people have diff erent capacities to convert income and wealth to welf are. As Sen says, "People have disparate physical characteristics with disability, illness, age or gender, and these make their needs diverse." (Sen 19 99, p. 70) An ill person may need more income to cure her illness; an older person needs more support; a pregnant woman needs more help; and so on. Then how could the sufficienc y-based reason apply to a real world where people have various basic needs? This question cannot be addressed only through philosophical analysis. But let me propose a general strategy of addressing it. Suppose that the total amount of social wealth is n, the population size is m, the welf are function of the ith person is U i (x). I propose 73 that a distribution <x 1 , X 2 , ... , Xm>, where Xi indicates the amount of economic goods received by the ith person, is fa ir if and only if it satisfies the following two equations: (1) U, (x 1 ) = U2 (x2) = ... = U (xm ); and (2 ) X 1 + X2 + ... + Xm = n. To illustrate the idea, consider the cake-sharing case again. The cake is divided into six equal slices. Three persons, Mary, Smith and Tom, are going to share the cake. For the sake of simp licity, let's assume that the utility function of Mary is U, (x) = ux, where xis the number of cake slices consumed by Mary and u is a positive constant, the utility function of Smith is U2 (x) = (u/2 )x, and the utility function of Tom is Us (x) = (u/3 )x. So Mary would gain u amount of utility through consuming one slice of cake, Smith would gain u/ 2 amount of utility through consuming one slice and Tom would gain u/3. How to fa irly distribute the cake? Here is the method. Suppose that the amount of cake that Mary ought to receive is x,, the amount that Smith ought to receive is x2, and the amount that Tom ought to receive is X s. The three numbers must satisf y the following two equations: (1) x, + X 2 + X s = 6; (2) ux, = (u/2 )x2 = (u/3 )xs. The distribution < 1, 2, 3>, where Mary receives one slice, Smith receives two slices and Tom receives three slices, satisfies the two equations. So the distribution < 1, 2, 3> is fa ir. Admittedly, a distribution scenario in the real world is far more complicated than the cake-sharing case. It is not easy to determine a person's utility function. Besides, given the utility functions, very probably no distribution could strictly satisf y both equations mentioned above. I tend to agree with Sen that probably a complete list of capa bilities that ought to be equally distributed among citizens cannot be determined by any given theory, 74 rather what capabilities ought to be included in the list varies with diff erent cultures and times and it is open to democratic discussion (Sen 2004). 75 Chapter 4 Priority and Distributional Equality As with luck egalitarianism and sufficientarianism, prioritarianism has been a hotly debated arena in recent years. But an accurate characterization of prioritarianism is unsettled. Broadly speaking, prioritarianism is understood as the following view: the fa ct that a person is worse off absolutely serves as a reason for assigning priority to benefiting her over benefiting a person who is better off. This characterization of prioritarianism as it stands is open to various interpretations. In the first two sections of this chapter, I examine two interpretations ofprioritarianism, which I call PRl and PR2. According to PRl, the priority view is grounded in the princi ple that the moral value of benefiting the worse off is greater than the moral value of benefiting the better off According to PR2, the priority view is grounded in the principle that a moral agent has a stronger moral reason to benefit the worse off than to benefit the better off. Then in Sections 3-4, I argue that whereas PRl is fa lse, PR2 is tenable. I aim to show that even if in some cases the moral value of benefiting the worse off is the same as the moral value of benefiting the better off, a moral agent may still have a stronger reason to benefit the worse off than to benefit the better off Moreover, I attempt to argue that the priority of benefiting the worse off is grounded in a more fundamental fa ct that a person would have a stronger prudential reason to improve her worse-off prospect than to improve her better off prospect. 76 1 The notion of "prioritarianism'' was originally presented by Derek Parfit, who maintains that benefiting the worse off matters more, and that benefiting the worse off matters more not because they are worse off than others, but because they are worse off absolutely (or non-comparatively ). As we can see, Parfit's priority view actually includes two theses, a positive thesis and a negative thesis. According to the positive thesis, the fa ct that a person is worse off absolutely serves as a reason for assigning priority to benefiting her over benefiting a better-off person. According to the negative thesis, the mere fa ct that a person A is worse off than another person B cannot serve as a reason for assigning priority to A over B. The discussion of this chapter centers on the positive thesis. The negative thesis, as I will argue in chapter 5, is fa lse. From now on, by "prioritarianism'' , I only mean the positive thesis of the priority view. Although prioritarianism can be vindicated, it is open to various interpretations as it stands - it is not at all clear in what sense the fa ct that a person is worse off absolutely serves as a reason for assigning priority to benefiting her over benefiting a better -off person. Parfit explains the non-comparative ness of prioritarianism by appealing to the correlation between two types of cases, which we might call "two-person cases" and "one-person cases". Suppose that you could, in a two-person case, either benefit a person who is badly off, or give the same benefit (or a slightly greater benefit ) to another person who is well off. You probably would pref er benefiting the person who is badly off to benefiting the person who is well off 77 Now imagine that you could, in a one-person case, either benefit the person if he is badly off, or give the person the same benefit (or a slightly greater benefit) if she is well off In this one-person scenario, you probably prefer benefiting the person if she is badly off to benefiting her if she is well off. So there seems to be some correlation between assi gning priority to the worse-off person in the two person case and assigning priority to the person's worse-off prospect in the one person case (Parfit 2012). Let's consider another example that illustrates the putative correlation between a one-person case and a two-person case. The basic setup of the example is given by Michael Ostuka and Alex Voorhoeve in their influential paper, "What it really matters when some are worse off than others". Imagine that a young and perf ectly healthy adult, say, Marie, will soon develop one of the following impairme nts and has a 50 percent chance of developing each: slight impairment -a condition that renders it difficult for one to walk more than 2 km, and very severe impairment -a condition that leaves one bedridden, save for the fa ct that one will be able to sit in a chair and be moved around in a wheelchair for part of the day if assisted by others. Suppose that you could either choose a treatment for the very severe impairment or a treatment for the slight impairment. If you treat the very severe impairment, you could move Marie's condition up from the very severe impairment to severe impairment -a condition in which she is no longer bedridden, rather, she is able to sit up on her own for the entire day but requires the assis tance of others to move about. Or if you treat the slight impairment, you would completely eliminate the mild 78 disability. However, in order for any treatment to be eff ective, it must be taken before it is known which impairment she will suffer. Each treatment, if eff ective, provides the same increase in utility over no treatment. The question is, "Which treatment would you give to Marie supposing you are a morally motivated stranger who could come to provide her with one or the other treatment?" (Ostuka and Voorhoeve 2009, pp. 171-2.) Here is the corresponding two-person case. Imagine that a young and perfectly healthy adult, Smith, will soon develop the very severe impairment; another young and perfectly healthy adult, Tom, will soon develop the slight impairment. Suppose that you can either choose the treatment for Smith's very severe impairment, or the treatment for Tom's slight impairment. But you cannot choose both treatments. The function of each treatment in the two- patient case is exactly the same as the function of each treatment in the one patient case. Moreover, as with the one-patient case, each treatment, if eff ective, provides the same increase in utility over no treatment. The question is, "As a morally motivated stranger, which treatment would you choose?" It appears that in both the one-patient case and the two-patient case, a moral agent ought to choose the treatment for the very severe im pairment. So assigning priority to the worse- off person in a two-person case seems to be correlated with assigning priority to the worse-off prospect of the person in a corres ponding one-person case. It is worth noticing that in the one-person case, benefiting the worse-off prospect matters more because the prospect is worse off absolutely (i.e. non-comparatively ) - there is only one person in question after all. 79 Parfit thus concludes that, in the two-person case, benefiting the worse-off person matters more because she is worse off absolutely (Parfit 1991, 19 97, 2012 ). 2 However, Parfit's explanation of the non-comparativeness ofprioritarianism is not at all clear. Philosophers have attempted to develop prioritarianism in diff erent ways. In this chapter, I examine two important interpretations of the view, which I call "PR1" and "PR2". According to PR1, the moral value of benefiting the worse off is greater than the moral value of benefiting the better off (let's call it the "eval uative interpretation" ). According to PR2, the priority view is grounded in the principle that a moral agent has a stronger moral reason to benefit the worse off than to benefit the better off (call it the "deliberative interpretation" ). Whereas some philosophers interpret prioritarianism as PR1 (Broome 1991; Hirose 2009; McCarthy 2006, 2008; Rabinowicz 2002 ), other philosophers interpret prioritarianism as PR2 (Crisp 201 1; Ostuka and Voorhoeve 2009 ). To better understand the distinction between PR1 and PR2, let's consider some specific examples. Deliberative One-person Cases One example of the kind of cases is the one-patient case with uncertainties introduced in Section 1. Let me briefly restate the case. Marie has a 50 percent chance of developing the very severe impairment, and a 50 percent chance of developing the slight impairm ent. Suppose that a moral agent could either 80 choose the treatment for the very severe impairment or the treatment for her slight impairme nt. In order for any treatment to be eff ective, it must be taken before it is known which impairment she will suffer. Besides, each treatment, if eff ective, provides the same increase in utility over no treatment. Which treatment ought a moral agent to choose? A deliberative one-person case could be a one-person case without uncertainties. Suppose that Marie will develop the very severe impairment at age 20, and without any treatment, the impairment will last 1 year. And suppose that she will develop the slight impairment at age 30, and without treatment, it will last 1 year. In other words, it is certain that, without any treatments, Marie would suf fer from the very severe impairment from 20 years old to 21 years old, and would suffer from the slight impairment from 30 years old to 31 years old. Suppose that a moral agent could either choose the treatment for the very severe impairment that would make Marie only suffer from the severe impairment from 20 to 21, or choose the treatment for the slight impairment that would make Marie perfectly healthy from 30 to 31 (but the moral agent cannot treat both ). The question is, "Which treatment ought a moral agent to choose?" Deliberative two-per son cases Let me briefly restate an example I gave in Section 1. A young and perfectly healthy adult, Smith, will soon develop the very severe impairment; another young and perfectly healthy adult, Tom, will soon develop the slight impairme nt. Suppose that a moral agent can either choose the treatment for Smith's very 81 severe impairment, or the treatment for Tom's slight impairment. But she cannot choose both treatments. Which treatment ought a moral agent to choose? A deliberative two-person case could be a case with uncertainties. Suppose that one of the two persons, Smith and Tom, will soon develop the very severe impairment and the other will soon develop the slight im pairment. The probability that Smith has the very severe impairment and Tom has the slight impairment is the same as the probability that Tom has the very severe impairment and Smith has the slight impairment. Before it is known what will happen, a moral agent ought to choose either the treatment for the very severe impairment or the treatment for the slight im pairment. The question is, "Which action ought a moral agent to choose?" Evaluative One-person cases Let me convert the deliberative one-patient cases to evaluative one-patient cases. At an earlier time t�, a moral agent gave Marie who was suffering from the very severe impairment at that time the treatment that can move Marie's condition from the very severe impairment up to the severe impairme nt. At a later time t2, the same moral agent gave Marie who was suffering from the slight impairment the treatment that can move Marie's condition from the slight impairment up to a healthy physical state. As with all cases mentioned before, let's assume that the utility gain from the very severe impairment to the severe impairment is the same as the utility gain from the slight impairment to the healthy physical state. The question is, "How does the moral value of the earlier treatment received by Marie compare to the moral value of the later treatment received by her?" 82 The evaluative one-patient case could be conceived in a diff erent way. Compare a pair of possible worlds. In world!, a moral agent gave Marie who was suff ering from the very severe impairment the treatment that can move Marie's condition from the very severe impairment up to the severe im pairment. In world 2 , the same moral agent gave Marie who was suff ering from the slight impairment the treatment that can move Marie's condition from the slight impairment up to a healthy physical state. The question is, "How does the moral value of the treatment carried out in world! compare to the moral value of the treatment carried out in world2?" Evaluative two-person cases The evaluative one-patient cases could be easily converted to evaluative two patient cases. On an occasion, a moral agent gave Smith who was suff ering from the very severe impairment the treatment that can move his condition from the very severe impairment up to the severe impairment. On another occasion, the same moral agent gave Tom who was suffering from the slight impairment the treatment that can move his condition from the slight impairment up to a healthy physical state. The question is, "How does the moral value of the treatment received by Smith compare to the moral value of the later treatment received by Tom?" 83 3 3.1 According to PRl, the worse off matters more because benefiting the worse off (either a worse-off person or the same person's worse-off prospect) has greater moral value than benefiting the better off. For instance, i n the evaluative one patient case, the treatment for the very severe im pairment received by Marie at time tl has greater moral value than the treatment for the slight impairment received by her at time t2. In the evaluative two-patient cases, the treatment for the Smith's very severe impairment has greater moral value than the treatment for Tom's slight im pairment. Then according to PRl, the fa ct that the treatment for the very severe impairment has greater moral value than the treatment for the slight impairment, serves as a reason for choosing the treatment for the very severe impairment rather than the treatment for the slight impairme nt. People who interpret prioritarianism as PRl seem to presuppose what I call "the principle of diminishing marginal moral value". On this view, the moral value of an action that brings a given amount of utility gain (or benefit ) decreases as the well-being of the recipient increases; in other words, the moral value of an action that brings a given amount of utility gain is inversely proportional to the well-being of the recipient. To put it in an even simpler way, the better off the recipient is, the lower the moral value of an action that brings a given amount of utility gain is. To explain the principle of diminishing marginal moral value, it might be helpful to contrast it with the law of diminishing marginal utility of money, 84 according to which, the utility of the same amount of money diminishes as the owner becomes richer. Consider the utility of having 100 dollars. Let "a" indicate the utility of having 100 more dollars when you have 1, 000 dollars, and "b" indicate the utility of having 100 more dollars when you already have 10, 000 dollars. According to the law of diminishing marginal utility of money, a > b. Now consider a certain amount of utility gain, say the utility gain received by a person who has the very severe impairment from moving up to the physical state of severe impairment, or the utility gain received by a person who has the slight impairment from moving up to a healthy physical state. Let "m'' indicate the moral value of the treatment for the very severe impairment, and "n" indicate the moral value of the slight impairment. According to the principle of diminishing marginal moral value, m > n. Admittedly, a person could affirm the view that, in some evaluative cases, benefiting the worse off has greater moral value than benefiting the better off, without affirming the principle of diminishing marginal moral value. Nevertheless, if a person holds the former view, she would tend to hold the latter view as well, because the latter view provides a unif ying basis for the former view. 3.2 In the rest of Section 3, I want to argue against PRl. I will attempt to show that, in an evaluative case, the moral value of an action that brings a given amount of utility gain does not decrease as the well-being of the recipient increases; instead, the moral value could be constant. Consider the following evaluative case. In world!, Marie is poor - she earns 2000 dollars per month; in world2, she earns 85 20, 000 dollars per month. At time t, in each world, she suff ers from the slight impairment. At a later time t*, in each world, she receives the treatment for the slight impairment and becomes healthy. In this case, Marie in world! is significantly worse off than Marie in world2, and it is reasonable to believe that the utility gain that the treatment for Marie's impairment brings to her in world! is the same as the utility gain that the treatment for her impairment brings to her in world2. The question is, "How does the moral value of the treatment for Marie's impairment in world! compare to the moral value of the treatment for her impairment in world2?" It seems that the two values are the same. Consider a corresponding evaluative two-person case. Smith and Tom both suf fer from the slight impairment. The diff erence between them is: Smith is poor - he earns 2000 dollars per month, and Tom is rich - he earns 20, 000 dollars per month. Now they both receive the treatment for the slight impairment. In this case, Smith is significantly worse off than Tom, and the utility gain of the treatment for Smith's impairment and the utility gain of the treatment for Tom's impairment are the same. The question is, "Is the moral value of the treatment for Smith's impairment greater than the moral value of the treatment for Tom's impairme nt?" It seems not; instead, the moral values of the two treatments are roughly the same. To test our intuition, consider an evaluative two-s ociety case. Compare the following two societies, "Misery" and "Flourishing". Suppose that Misery and Flourishing are the same in all other aspects except the following diff erence: 86 citizens in Misery all live under poor conditions -t hey are as well off as Marie in world, (Or Smith in the evaluative two-person case ); by contrast, citizens in Flourishing are all rich. At an earlier time t, all citizens in Misery suffer from the slight impairment and all citizens in Flourishing suf fer from the slight impairment as well. At a later time t*, all citizens in Misery and all citizens in Flourishing receive the treatment for the slight impairment and they all become healthy. Needless to say, Misery at time t* is better than Misery at time t, and Flourishing at time t* is better than Flourishing at time t. The question is, "Is later Flourishing better than earlier Flourishing to the same extent as later Misery is better than earlier Misery?" It seems we can hardly have a reason to say "No". Based on our intuitions on the above evaluative cases, it is saf e to conclude that the moral value of an action that brings a given amount of utility gain does not decrease as the well-being of the recipient increases; instead, the moral value would be constant. Our moral intuitions on the cases are consistent with the following moral presupposition: our actions, attitudes and feelings, as long as they significantly benefit people, can always have significant moral value. Therefore, PRl is problematic. Notice that the evaluative circumstances introduced in Section 2 are slightly different from the evaluative circumstances discussed in this section. The former cases are concerned with the contrast between the moral value of the treatment for the very severe impairment and the moral value of the treatment for the slight impairment; the latter cases are concerned with the contrast 87 between the moral value of the treatment for a poor person's slight impairment and the moral value of the treatment for a rich person's slight impairme nt. Whereas in the latter evaluative circumstances, our intuition that the two moral values are the same is clear, in the former evaluative circums tances, many people probably do not have a clear intuition that the two moral values are the same. Does it mean that, in regard to the former evaluative circumstances, the two moral values are not the same, and the moral value of the treatment for the very severe impairment is greater than the moral value of the slight impairment? No. Those people who don't have the intuition that the two moral values are the same, don't fully appreciate the fa ct that the utility gain from the treatment for the very severe impairment is the same as the utility gain from the treatment for the slight im pairment. They probably assume that the utility gain of the treatment for the very severe impairment is greater than the utility gain of the treatment for the slight impairment. But according to PR2, the priority of benefiting doesn't rely on how the moral value of benefiting the worse off compares to the moral value of benefiting the better off Instead, the worse off matters more because a moral agent would have a stronger moral reason to benefit the worse off (either a worse-off person or the same person's worse-off prospect) than to benefit the better off. For instance, in the deliberative one-patient cases, a moral agent ought to choose the treatment for the person's very severe impairment over the treatment for the person's slight impairment; and in the deliberative two-patient cases, a moral 88 agent ought to choose the treatment for the person who has very severe impairment over the treatment for the person who has slight impairment. 4 4.1 I will argue that whereas PRl is fa lse, PR2 is tenable. Someone might say that PR2 is actually equivalent to, or grounded in, PRl. On this view, a moral agent ought to choose the action that could benefit the worse-off person rather than the action that could benefit the better-off person, precisely because the moral value of the action that benefits the worse off is greater than the moral value of an action that benefits the better off But I want to point out that there is a disanalogy between the evaluative cases and the deliberative cases. To begin with, it may be worth noticing that the weight of the moral reason for a given action in a deliberative case doesn't equal to the weight of the moral value of a corresponding action in an evaluative case. For instance, the weight of the moral reason to treat the very severe impairment in a deliberative case doesn't equal to the weight of the moral value of the same treatment in an evaluative case; similarly, the weight of the moral reason to treat the light impairment in a deliberative case doesn't equal to the weight of the moral value of the same treatment in an evaluative case. Here is the argum ent. Compare the following pair of cases. In the deliberative case, a moral agent could choose one action from one million options. She could save one person, or save two persons, or save three person ... , or save 89 one million persons. In the evaluative case, the moral agent saves one person at time t 1 , saves two persons at time tz, saves three persons at times ts ... , and saves one million persons at time tn. In the deliberative case, the weight the moral agent should give for saving one person is very low. We might say that, in the deliberative case, compared to the moral weight of saving one million people, the moral weight of saving one person is very small and even negligible. However, we don't think that in the evaluative case, the moral value of saving one person is negligible; instead, we think that the moral agent's lif e-saving at time t, is praiseworthy to a very serious extent, and the moral agent ought to feel proud of herself Theref ore, the weight of the moral reason for a given action in a deliberative case doesn't equal to the weight of the moral value of a corresponding action in an evaluative case. The proponents ofPRl might respond that even if the weight of the moral reason for saving one person doesn't equal to the weight of the moral value of saving one person, it may still be the case that a moral agent has a stronger moral reason to save one billion people than to save one person because the moral value of saving one billion people is greater than the moral value of saving one person. In other words, there may still be a correspondence relationship between moral reason and moral value. I acknowledge that this is what is at issue. In the following passages, I shall argue that even if in some cases the moral value of benefiting the worse off is the same as the moral value of benefiting the better off, a moral agent may still have a stronger reason to benefit the worse off than to benefit the better off 90 4.2 Some might wonder, "How could it be possible that, in a deliberative case, the action that brings a given amount benefit to the worse off matters more than the action that brings the same amount of benefit to the better off, if in a corresponding evaluative case, the moral value of benefiting the worse off is the same as the moral value of benefiting the better off?" For example, why, in the deliberative one-patient case, does the treatment for Marie's very severe impairment matter more than the treatment for her slight impairment, if it is not because in a corresponding evaluative case, the treatment for the very severe impairment has greater moral value than the treatment for the slight impairment? Here we can just focus on one-person cases, in order to avoid unnecessary complexities concerning inter-personal contexts. Let me attempt to address this question. In a deliberative one-patient case, a moral agent has a stronger moral reason for carrying out the treatment for the very severe impairment rather than the treatment for the slight impairment, because the former treatment is better fo r the patient, or to put it another way, the patient would have a stronger prudential reason to receive the former treatment than to receive the latter treatment. Recall the deliberative one-patient cases. Now let's revise the cases to be what I call "deliberative one-patient cases of prudence". First revise the one patient case to account for uncertainties in one-person case of prudence. Suppose that Marie has a 50 percent chance of developing the very severe impairment and a 50 percent chance of developing the slight im pairment. She 91 could either choose the treatment for the very severe impairment or choose the treatment for the slight impairment. But she cannot choose both. The question is, "Is the treatment for the very severe impairment better for her than the other treatment?" Or, "Does Marie have a stronger prudential reason to receive the treatment for the very severe impairment than to receive the other treatment?" Now revise the one-patient case without uncertainties and take it to be a case of prudence. Suppose that Marie will develop the very severe impairment at 20, and without any treatment, the impairment will last 1 year; and she will develop the slight impairment at 30, and without treatment, it will last 1 year. Suppose that she could either choose the treatment for the very severe impairment or choose the treatment for the slight impairment. But she cannot choose both. The question is, "Is the treatment for the very severe impairment better for her than the other treatment?" Or, "Does Marie have a stronger prudential reason to receive the treatment for the very severe impairment than to receive the other treatment?" The basic setup of each deliberative one-patient case of prudence is exact! y the same with the basic setup of each deliberative one-patient case introduced earlier in section 2. The diff erence is that in a case of prudence, the question to consider is not about which treatment a moral agent ought to choose, but about which treatment is better for a patient. For the sake of convenience, from now on, I will use the term "deliberative one-person cases of morality'' to refer to the deliberative one-person cases introduced in section 2. In the deliberative one-patient cases of morality, we have the moral intuition that a moral agent ought to choose the treatment for the person's very 92 severe impairme nt. This is because we assume that, a moral agent ought to choose a treatment on behalf of the patient. A moral agent ought to choose the treatment that is better for the patient. Moreover, we assume that the treatment for the very severe impairment is better for the patient. Reviewing the one patient cases of prudence introduced above; we indeed have the intuition that Marie has a stronger prudential reason to receive the treatment for the very severe impairment than to receive the treatment for the slight im pairment. In a deliberative one-patient case, the treatment for the very severe impairment is better for the patient than the treatment for the slight impairme nt. By contrast, in an evaluative one-patient case, the treatment for the very severe impairment is as good as the treatment for the slight impairment for the patient. To understand this, let's contrast evaluative one-person cases of prudence with deliberative one-person case of prudence. Now revise the evaluative one-patient cases introduced in Section 2 to be evaluative one-patient cases of prudence. Suppose that at an earlier time t�, Marie received the treatment for the very severe impairment that she was suff ering at that time; at a later time t 2 , she received the treatment for the slight impairment that she was suff ering. The question is, "Is the treatment for the very severe impairment better for Marie than the treatment for the slight im pairment?" Or put the question in another way, "Should Marie rank the earlier treatment higher than the later treatment?" Look at the evaluative one-patient case of prudence from a diff erent perspective. Compare a pair of possible worlds. In world!, Marie received the treatment for the very severe impairment that she was suff ering. In world 2 , Marie received the 93 treatment for the slight impairment that she was suff ering. How does the prudential value of the treatment for the very severe impairment compare to the prudential value of the treatment for the slight impairment? I contend that, in the evaluative one-patient cases of prudence, the two treatments have the same prudential value, but in the deliberative one-patient case of prudence, the treatment for the very severe impairment has greater prudential value than the treatment for the slight impairment, that is, the treatment for the very severe impairment is better for Marie than the treatment for the slight impairment, or she has a stronger prudential reason to receive the treatment for the very sever impairme nt. The essential diff erence between the evaluative one-person cases of prudence and the deliberative one-person cases of prudence is that the latter involves the circumstance of deliberative choice, whereas the former doesn't. This diff erence plays an extremely important role in grounding the diff erence between the prudential valuation in the evaluative circumstances and the prudential reasoning in the deliberative circumstances. Notice that in both types of one-person cases, each action has two functions, bringing the recipient a certain amount of utility gain and eliminating a certain prospect. For example, in both types of one-patient cases of prudence, the treatment for the very severe impairment brings Marie a certain amount of utility gain, and at the same time eliminate the prospect of suff ering from the very severe impairment; the treatment for the slight impairment brings Marie 94 the same amount of utility gain, and at the same time eliminate the prospect of suff ering from the slight impairm ent. Given the two functions of each treatment, in the deliberative one-patient cases of prudence, in the process of deliberating on the alternative treatments, Marie not only needs to compare the benefit (or utility gain) that each treatment brings, but also needs to compare the weight of the elimination of the worse-off prospect with the weight of the elimination of the bette r-of f prospect. By contrast, in the evaluative one-patient cases of prudence, when reflecting upon how the prudential value of one treatment compares to the prudential value of the other treatment, Marie only needs to compare the benefit (or utility gain) that each treatment brin gs - she doesn't need to compare the weight of the elimination of the worse- off prospect with the weight of the elimination of the better-off prospect. This is because, in the deliberative circumstance, the elimination of the worse- off prospect and the elimination of the bette r-of f prospect are competitive (in the sense that the agent can only choose one option), whereas the competitiveness disappears in the evaluative circumstance. I also contend that, in the circumstance where the elimination of a worse off prospect and the elimination of a bette r-of f prospect are competitive, a rational person ought to assign priority to the worse-off prospect. So in the deliberative cases of prudence, Marie has a stronger prudential reason to receive the treatment for the very severe impairment than to receive the other treatment because she has a prudential reason to avoid the prospect of suff ering from the very severe impairment than to avoid the prospect of suff ering from the slight 95 impairment. By contrast, in the evaluative cases of prudence, the two treatments have equal prudential value, since they bring the same amount of utility gain for Marie; she doesn't need to compare the weight of avoiding the prospect of suff ering from the very severe impairment with the weight of avoiding the prospect of suff ering from the slight impairment at all. 4.3 My view on prioritarianism is based on a distinctive account of prudential rationale, which we might call the "risk aversion rationale". According to the rationale, in the circumstance where a person could either choose a lottery that brings her a certain amount of benefit when she is worse off, or choose a lottery that brings her the same amount of benefit or a slightly larger benefit when she is better off, she has a prima fa cie prudential reason for preferring eliminating the worse-off prospect to improving the bette r-of f prospect. The risk aversion rationale is different from the so-called "expected utility rationale" that is oftentimes assumed by contemporary economists and philosophers. Let me briefly introduce expected utility rationale. A person's expected utility under a lottery equals to the sum of the utility of each possible prospect multiplied by the probability of the prospect. According to expected utility rationale, a rational person has a stronger prudential reason for a lottery, L,, than another lottery, L2, if and only if her expected utility under L, is greater than her expected utility under L2. Recall the deliberative one-patient cases of prudence. For the sake of clarity, let's use a number to indicate the utility level of a certain prospect. 96 Suppose that the utility level of the very severe impairment is 1, the utility level of the severe impairment is 3, the utility level of the slight impairment is 4, and the utility level of a healthy physical state is 6. Given the assumptions, Marie's expected utility under the lottery of receiving the treatment for the very severe impairment would be: 3*50% + 4*50%. Her expected utility under the lottery of receiving the treatment for the slight impairment would be: 1 *50% + 6* 50%. Since the two values are equal, according to the expected utility rationale, if Marie is rational, she would feel indifferent between the two treatments. However, the expected utility rationale cannot always serve as the correct rationale for rational choice. As the deliberative one-patient cases of prudence have shown, although Marie's expected utility under the lottery of receiving the treatment for the very severe impairment is the same as her expected utility under the lottery of receiving the treatment for the slight impairment, we still have the intuition that the treatment for the very severe impairment is better for Marie, or that she has a stronger prudential reason to receive the treatment for the very severe impairment. Indeed, imagine that you were Marie. Imagine that you have a 50 percent chance of developing the very severe impairment and a 50 percent chance of developing the slight impairme nt. There is a treatment that is available for each of these conditions, but, in order for it to be eff ective, it must be taken before it is known which impairment you will suffer. Would you want to have the medical treatment for the very severe impairment or the treatment for the slight impairment? I think most of us would want the treatment for the very severe impairm ent. 97 4.4 In many circumstances, we have a stronger prudential reason for a lottery L, than another lottery L 2 , even if L 2 would bring us greater expected utility. Here it is worth mentioning Thomas Nagel's fa mous two-child case 2 4 and its corresponding one-child cases. In the two-child case, Nagel imagines that he has two children; one of them is healthy and happy, and the other is suff ering from some painful handicap . Suppose his fa mily could either move to a city where the suff ering child would receive special medical treatment, or move to a suburb where the happy child would flourish. Nagel also supposes "the gain to the first child of moving to the suburb is substantially greater than the gain to the second child of moving to the city." Given all of the assumptions, Nagel thinks that, all things considered, it is better to move to the city, because "it is more urgent to benefit the second child, even though the benefit we can give him is less than the benefit we can give to the first child."(Nagel 1979, pp. 12 3-4) Prioritarians agree with Nagel's judgment on the two-child case (Parfit 19 95, 19 97, Crisp 2012). 2 5 Moreover, prioritarians contend that one reason for why benefiting the child with handicap matters more than benefiting the healthy child is that the child with handicap is worse off absolutely. As Parfit says, "it would be just as urgent to benefit the handicapped child even if he had no sibling who was better off" (Parfit 19 97) 24 It might be interesting to note that Parfit makes it clear that he comes to embrace the Priority View through his reflection upon Thomas Nagel's famous two-child case. (Parfit 1995, 1997) 25 For instance, Parfit makes it quite clear that, moving to the city is a better choice because "benefiting people matters more the worse off these people are". (Parfit 1997) 98 But prioritarians fa il to give a further explanation for how exactly the fa ct that the child with handicap is worse off absolutely serves as a reason for assigning priority to benefiting the child over benefiting the healthy child. I propose that, in the two-child case, one reason for why benefiting the child with handicap matters more than benefiting the healthy child is that, in a corresponding deliberative one-child case of morality, improving the child's worse-off prospect matters more than benefiting his better-off prospect. Let's revise the two-child case to be a deliberative one-child case of morality. Suppose that a child has a 50 percent chance of ending up with the handicap and a 50 percent chance of ending up able-bodied. Suppose that the parent must take the move to the city or the suburb before it is known whether the child will end up with the handicap . As with the two-child case, in the one-child case, I agree with prioritarians that the parent ought to move to the city, since they ought to give more weight to avoiding the possible outcome in which the child would be worse off Moreover, in a corresponding deliberative one-child case of morality, benefiting the only child's worse-off prospect matters more than benefiting his better-off prospect because moving to the city is better for the child. Consider the following deliberative one-child case of prudence. Suppose that you are the child who has a 50 percent chance of ending up with the handicap and a 50 percent chance of being perfectly healthy. Suppose your fa mily could either move to the city where you could receive special medical treatment should you end up with the handicap, or move to the suburbs where you would flourish should you 99 turn out to be perfectly healthy. The question is, "Would you want your parent to move to the city or the suburbs?" I think most of us would believe that it would be worse if we couldn't get proper medical treatment when we end up with the handicap . It's worth noting that, the fa ct that, in a deliberative one-person case, a person has a stronger prudential reason to improve her worse-off prospect, doesn't imply that a moral agent ought to, all things considered, choose the benefit to her worse- off prospect. Sometimes, although a person has a stronger prudential reason to improve her worse-off prospect, she might actually pref er the moral agent to improve her better-off prospect. In a case of this kind, the moral agent has a reason to respect the agent's autonomy. But when a moral agent doesn't know what the person pref ers her to choose, the agent ought to carry out the action that the person would have a stronger prudential reason to choose. As Parfit says, "when we have to make some decision on someone's behalf , and we don't know how this person would prefer us to act, we may believe that we ought to be cautious, or risk averse .... As Prioritarians claim, we ought to give more weight to avoiding possible outcomes in which this person would be worse off" (Parfit 2012, p. 423) 4.5 To sum up, according to my understanding of prioritarianism, the prioritarian slogan "the worse off matters more" is better interpreted as the thesis that a moral agent has a stronger moral reason to benefit the worse off than to benefit the better off Even if in some cases the moral value of benefiting the worse off is 100 the same as the moral value of benefiting the better off, a moral agent may still have a stronger reason to benefit the worse off than to benefit the better off I thus contend that there is a priority-based reason for fa voring a more equal distribution between the worse-off person and the better-off person. Moreover, as I have argued, the priority-b ased reason for distribution is grounded in a more fundamental fa ct that a person would have a stronger prudential reason to improve her worse-off prospect than to improve her better-off prospect. Some might doubt that the priority-b ased reason can play any distinctive role in justif ying a more equal distribution. It seems that a person has a stronger prudential reason to improve her worse-off prospect than to improve her better off prospect if and only if the worse-off prospect is below a certain threshold, say the threshold as a healthy physical state and the threshold as being as well off as a middle class person is. But if so, then isn't the priority-based reason grounded on a reason concerned with the satisfac tion of basic needs or reaching the suf ficiency threshold? I argue that the priority-based reason can play a distinctive role in justif ying a more equal distribution. I have two reasons. First, the fa ct that a person has a stronger prudential reason to improve her worse-off prospect than to improve her better -off prospect would by itself serve as a reason for justif ying the priority of benefiting a worse-off person over benefiting a better-off person, even though the fa ct is grounded on a more fundamental fa ct that the worse-off prospect is below a certain threshold. Second, for some prospect, even if it is above the suf ficiency threshold, a person still has a stronger prudential reason to improve her worse-off prospect 101 than to improve her better -off prospect. Now consider another one-child case. Suppose that a child has a 50 percent chance of ending up "normally '' healthy, and a 50 percent chance of ending up extraordinarily healthy. Suppose that if the child were normally healthy, living in the city would bring her n more units of utility than living in the suburb. Suppose further that if the child were extraordinarily healthy, living in the suburb would bring her n more units of utility than living in the city. The parent must take the child to move to the city or the suburbs before it is known how well off the child will end up. In contrast to the previous one-child cases, both prospects are above the sufficiency threshold. In this case, the child seems to have a stronger prudential reason to move to the city (due to the risk-aversion consideration). 102 Chapter 5 The Intrinsic Value of Equality In the recent literature of distributive justice, philosophers have debated whether distributional equality could be intrinsically valuable (Holtug and Lippert- Rasmussen 2007). I use the term "intrinsic egalitarianism'' to label the view that equal distribution of (at least) some goods has intrinsic value. 2 6 There are diff erent theoretical approaches, such as luck egalitarianism, sufficientarianism and prioritarianism, which might just if y an equal (or more equal rather than less 2 7) pattern of distribution, but those approaches would make the value of equal distribution non-intrinsic. What distinguishes intrinsic egalitarianism from those competing theories is its insistence on the intrinsic value of distributional equality. In the final chapter, I aim to defend the intrinsic value of distributional equality. This chapter is divided into three sections. In Section 1, I introduce the fa mous "leveling-down'' objection, which seems to raise a serious challenge to intrinsic egalitarianism. The leveling-down ob je ction is grounded in the person- aff ecting requirement of equality. In Section 2, I attempt to argue that equality can benefit persons in an intrinsic way. I first propose an epistemic principle concerning intrinsic personal value as follows: it is (epistemically) reasonable to 26 A list of distributional goods may include fr eedom and rights, opporhmities, happiness, capabilities, respect, income and wealth, etc. Here "distributional goods" should be nnderstood broadly, including not only direct metrics of distribution, but also distributive states that are achieved by appeal to the metrics of distribution. 27 The egalitarian can agree that there could be pro tanto reasons in favor of an unequal distribution of some goods, although there are pro tanto reasons in favor of an equal distribution. So, the egalitarian doesn't have to propose a strictly equal distribution of goods, but only a more equal rather than less pattern of distribution. 103 regard x as an intrinsic value for a person S if S rationally desires x (as part of her life) for its own sake. Then I introduce some thought experiments to reveal our intuition that a person can rationally desire equality of at least some goods (e.g. respect, liberties, and longevity) for its own sake. In Section 3, I consider several possible objections to my view and then make further responses. I attempt to show that those objections don't really threaten my main thesis. 1 1.1 Let me first define what is meant by "intrinsic value". In this dissertation, I follow a standard usage, according to which xis intrinsically valuable if xis valuable for its own sake, or in other words, valuable as an end. 2 8 Two remarks are worth mentioning. First, something can be intrinsically valuable even if its value derives from deeper values or reasons (Raz 2008). For example, suppose that beauty is an intrinsic value. But there could be deeper reasons and values (aesthetic or otherwise) that explain why beauty is valuable. We should distinguish intrinsic values from fundamental values, values that are not grounded by any other values. The mere fa ct that the value of something is based on other values is not a reason to deny that it could still be intrinsically valua ble. 28 Some philosophers may adopt a different terminol ogy. For example, Korsgaard uses the term 'intrinsic value' in the sense that xis an intrinsic value if the value doesn't depend upon anything external to x (Korsgaard 1983). She seems to equate the distinction between intrinsic and extrinsic values with the distinction between non-relational and relational values. For OliT pliTpose, we stick with the standard usage of 'intrinsic value' adopted by most egalitarians/non egalitarians (Parfit 1997; Raz 2008). 104 Second, something can be intrinsically valuable even if its value is contingent on certain conditions. The distinction between intrinsic and extrinsic values is not parallel with the distinction between unconditional and conditional values. For something to be an intrinsic value, it doesn't have to be valuable (much less intrinsically valuable) in every context - it only needs to be so in some circumstances. For instance, pleasure is a typical example of intrinsic value. But as Jonathan Dancy argues, pleasure might be not valuable in certain circumstances, such as the circumstances in which one feels happy about killing a baby (see Dancy 2004). For our purpose, we only need to argue that distributional equality is intrinsically valuable under certain conditions or in certain circumstances, but not that it is always intrinsically valua ble. Of course, although the opponents of intrinsic egalitarianism cast doubts upon the intrinsic value of distributional equality, they don't deny that distributional equality could be valuable in other ways. Distributional equality can be instrumentally valuable if equal distribution of goods causally brings about other valuable states of aff airs (e.g. peace, social harmony and non domination). Moreover, distributional equality can be valuable as a by-product of promoting other values. For example, suppose that it is valuable for a person to enj oy freedom. Then it seems to follow that it is also valuable for everyone to enj oy freedom. The equal distribution of freedom is valuable indeed, but, according to the non-egalitarian, valuable only as a side-eff ect of promoting freedom maximall:y or universall:y. 105 1.2 Then let's discuss the leveling-down objection. Suppose that a certain kind of good G is to be distributed between two persons A and B. Compare the following two distributions, S 1 ( < 10 0, 200>) and S2 ( < 100, 1 00> ). 200 180 160 140 120 100 B 80 60 40 20 S1 S2 For the sake of argument, suppose that all other things are equal. For example, A and B make equal contributions; the difficulty level of their beneficial practice is the same; the leveling-down situation wouldn't instrumentally benefit anyone in any respect. And it is reasonable to say that neither sufficiency-b ased reasons, nor priority-b ased reasons can be used to just if y the leveling-down distribution < 10 0, 1 00>. Then the only possible way to just ify the leveling-down distribution is to appeal to the intrinsic value of equality. But according to many philosophers, it is hard to see how a leveling-down situation with respect to any good could be intrinsically better. This is the basic idea of the leveling-down objection. There are diff erent versions of the leveling-down objection. Let's consider the first version (L1 ). 1. If an equal distribution is intrinsically valuable, then the leveling down situation S2 is overall better than the unequal situation S 1; 2. But S2 is not overall better than SJ; 106 3. Therefore, an equal distribution is not intrinsically valua ble. However, as many philosophers indicate, this version of the leveling-down objection is unsound, as premise 1 is clearly fa lse (Raz 2008). People who launch the above argument seem to assume (mistakenly) that if a person endorses the intrinsic value of equality, then she must regard it as the most important value or the overriding value. But one can maintain that distributional equality has intrinsic value but at the same time coherently agree that this value could be traded off for the sake of other values. The opponents of intrinsic egalitarianism may revise the first leveling- down objection. Let's consider a modified version (L2). 1. If an equal distribution is intrinsically valuable, then the leveling down situation S2 is better than the unequal situation S I in some respect; 2. But S2 is not better than S I in any respect; 3. Therefore, an equal distribution is not intrinsically valuab le. Peter Railton gives a vivid example to illustrate this leveling-down objection to egalitarianism: Suppose there is a genetic diff erence among actual humans that is normally expressed in diff erences in the resilience or intensity of the respect they have for others. People with the R allele, say, are harder to push into disrespect for others out of stress, disappointment, fear, etc. than those with the NR allele. Now this is also an inequality in respect. R individuals will, for example, feel and show respect for others in conditions where the NR individuals around them fa il to do so. Suppose that otherwise R individual s are no happier or more prosperous than NR individuals- perhaps those more easily pushed into disrespect also have more gratif ying experiences of restoring their respect for others, say. The inequality, then, is limited to respect. Suppose there is no treatment to replace the NR allele with R, but there is a button you can push to eliminate the R in future generations. Would there be anything to be said for that? A systematic inequality in respect would be eliminated. Of 107 course, there might be something to be said for looking for a way to make the R allele universal. But then the rationale could be increasing respect for others. (Railton 2008, p. 4) In this thought experiment, Railton compares two ways of achieving the equality of respect, the leveling-down situation through the elimination of the R allele and the leveling-up situation through making the R allele universal. He argues that, if we equalize the amount of respect by making the R allele universal, this leveling-up situation would augment overall intrinsic value. But it seems that this is not due to the intrinsic value of distributional equality, but because of an increase in a kind of intrinsic value (i.e. respect). Then, Railton asks, what intrinsic value can the leveling-down action possibly bring about? It seems a very hard task for the intrinsic egalitarian to show where the intrinsic value of distributional equality lies. According to many philosophers, equality has no intrinsic value precisely because equality cannot intrinsically aff ect a person's well-being (call this "the person-affecting challenge" ). The person-affecting challenge is what actually motivates the leveling-down objection to intrinsic egalitarianism. In the following two sub-sections, I will distinguish between two diff erent person- aff ecting challenges. 1.3 Consider the first person- aff ecting challenge (P 1 ): 1. A state of aff airs x cannot be better than another state of aff airs y in any respect ifthere is no one for whom xis better than yin any respect (the person-a ffecting requireme nt of value); 2. There is no one for whom S2 is better than S1 in any respect; 3. Th us, S2 isn't better than S1 in any respect; 108 4. If equality is intrinsically valua ble, then S2 would be better than S1 in at least one respect; 5. Therefore, equality isn't intrinsically valua ble. Although this argument gives some further explanation of why S2 isn't better than S 1 in any respect, it is unsound. Both premise 1 and premise 2 are problematic. Here I only focus on premise 1 (I will discuss premise 2 later). Temkin' s "Saints vs. Sinners" case is a very good case against the person- aff ecting requirement of value. Temkin invites us to imagine two alternative afterlives. "In the first, the saints get what they deserve, and so do the sinners, who fa re much worse. In the second, the saints get what they deserve, and the sinners fa re much better than the saints." (Temkin 2003b, p. 25) According to Temkin, there is one respect in which the second afterlif e is worse than the first, even if there is no one for whom the second afterlif e is worse than the first afterlif e. That is, the second afterl ife is worse than the first because proportional justice is not exemplified in the second afterlife. It seems to be appealing to say that proportional justice is one respect in which a situation can be better or worse than another without being better or worse for anyone in that situation. Someone may argue that proportional justice is justified because it can influence people's choices. For instance, a person might restrain himself from doing bad things after imagining that he will fa re much worse if he keeps doing those bad things. Th us, proportional justice would only be of instrumental value for some personal good. In response, let's assume that in a possible world, either people don't know that sinners would fa re much worse than saints or people's choices are not influenced by this knowledge. Even if so, it still seems true that the 109 second afterlif e where the sinners fa re much better than the saints is worse than the first afterlife. Proportional justice is an ideal that we use to evaluate an outcome (or state of aff airs). But it is very hard to include proportional justice in any theory of personal good. However, it is important to note that even though the person- aff ecting requirement of value is fa lse, it doesn't follow that equality can be valuable without being person-affecting. A state of aff airs could be an instantiation of many values. Suppose that a state of affairs S instantiates both proportional justice and equality. As we discussed, even if S is not better for any person, it can be better in virtue ojbeing an instantiation of proportional justice. However, it doesn't follow that S is better in virtue of equality (although S also instantiates equality). People may hold the view that the person-aff ecting requirement applies to equality, even if it doesn't work for other values (say, proportional justice). For example, Raz argues that if distributional equality cannot promote people's well-being, it would be dubious to regard it as a value. Raz puts it this way: Even ifth ere are values which cannot benefit people the view that equality is among them would surprise and disappoint most egalitarians, as the ethos of egalitarianism is deeply humanist, that is concerned with the fortunes of people. (Raz 2008, p. 10) Raz's view could be labeled as "the person-affecting requirement of equality", which says that equality is intrinsically valuable only if it intrinsically benefits persons. Now let's consider the second person-affecting challenge (P2), which assumes the person-affecting requirement of equality. This new challenge seems more plausible than (P 1 ): 11 0 1. Equality (of any sort) is intrinsically valuable only if it is intrinsically good for persons (the perso n-aff ecting requirement of equality); 2. Equality cannot be intrinsically good for persons; 3. Therefore, equality is not intrinsically valuab le. There are two strategies the intrinsic egalitarian may adopt for rejec ting this argument. She can either deny either premise 1 or premise 2. But I agree with Raz that the person-affecting requirement applies to the value of equality, even if it may not work for other values. It would be odd if the equal distribution of goods didn't matter to each person's well-being. So, I will assume the person- aff ecting requirement of equality without further discussion. Notice that premise 2 in this argument could be regarded as an implication of premise 2 in the first person-affecting challenge. The claim that there is no one for whom S2 is better than S1 in any respect would entail that there is no one for whom S2 is better than S1 in virtue of the equality involved in S2. Of course, a person who attempts to defend the view that there is no one for whom S2 is better than S1 in virtue of the equality involved in S2, doesn't need to hold the view that there is no one for whom S2 is better than S1 in any respect. In the next section, I will, by appeal to some thought experiments, argue against premise 2 -i n other words, I will attempt to establish that equality (of some goods) can intrinsically benefit people. 111 2 2.1 In order to examine whether distributional equality is intrinsically good for people, I suggest that we should first determine how to find out whether something has intrinsic value or not. To answer this question, we had better consider paradigms of intrinsic personal values, such as pleasure and being respected by fellow citizens. Why do philosophers almost universally agree that pleasure is intrinsically valuable for people? What is the evidence for this claim? A natural answer is that this is because people rationally desire pleasure for its own sake. 2 9 How do we know that it is good for a person to be respected by fellow citizens? One piece of evidence is that people rationally desire being respected by fellow citizens for its own sake. The thought of connecting values with desires has a long history in ethics. (Hume 1978; Mill 19 98 ) In contemporary metaethics, this tradition has been highly developed and sophisticatedly elaborated by various sorts of theories that aim to reduce evaluative properties to desire-based dispositional properties. (Lewis 19 89; Railton 19 86; Smith 2002) For example, Railton proposes a fa mous account of non-moral goodness that appeals to rational desires or wants, according to which xis non-morally good for A if and only if an idealized self of A, A+, would want A to want x were A+ to find himself in the actual condition and circumstances of A. (Railton 19 86 ) For example, eating food is good for an 29 Here I use the term 'desire' in a broad sense to refer to any motivational state that has a 'world-to mind' direction of fit. The object of a desire could be either a particular thing or a state of affairs. 112 apositia patient, as the idealized self of the apositia patient would want her to want to eat food. However, in this pro ject I am not proposing any dispositional account of value; for the time being I remain completely silent to the metaethical question of whether values are desire-based dispositional properties. All I wish to stress here is just that our knowled ge about what things have personal value is typ ically based on evidence regarding rational desirability - this phenomenon on value epistemology is exactly the starting point from which a dispositional theory of value is developed. I do think that it would be fruitful for our current ethical and political inquirie s to draw upon the resources of dispositional theories of value, which pay much attention to the close relationship between values and rational desires. Now I want to tentatively propose an epistemic principle concerning intrinsic personal value as follows: (Principle D) It is reasonable to regard x as an intrinsic value for a person Si f Sr ationally desires x (as part ofher life) for its own sake. It is worth noting two points. First, although it is very hard to give an accurate definition of 'rational', it would be helpful to list three typical conditions on the rationality of desires: (1) Freedom fr om errors. A desire is rational only if it is a result of correct (descriptive) beliefs and competent reasoning; (2) Freedom from ignorance. A rational desire must be based on full descriptive information about the relevant situation; and (3) Consistency. For a desire to be rational, it must be coherent with other desires had by the agent. For example, there shouldn't be 113 inconsistency between first-order and second-order desires, between basic and derivative desires, etc. Second, in order for something to be a value fo r someone, the person must rationally desire it (as part of her life), not just merely endorse or believe in its value. Imagine we somehow know that there is a beautiful forest in a remote planet but we can never observe the forest by any means. Suppose that while we believe that the forest is beautiful, we don't really desire it (although we may desire that there be such a forest in our city). Then the beautiful forest is not valuable for us, even if it may be valuable as such. 2.2 In what follows, I will argue that people can rationally desire equal distribution of certain goods for its own sake in certain circumstances. We can easily decide whether we rationally desire certain things (e.g. pleasure) for their own sake. But it is much more dif ficult to judge whether we have the same attitude toward distributional equality. So, I attempt to appeal to thought experiments to reveal our intuitions on distributional equality. Let me introduce three cases regarding distribution of respect, distribution of freedom and distribution of longevity. Respect Consider two planets, a and �- a-dwellers show a minimal level of respect to each other. They do respect others' dignity and rights, and regard some actions (e.g. killing, beating, manipulation, rape, and stealing) as morally im permissible. But besides that they are cold-hearted and disrespectful. They seldom have warm feelings to each other, and don't care whether their behaviors would hurt 11 4 others' feelings -t hey always criticize, mock, and despise their fellows. In contrast, �-dwellers are very warm-hearted. They always try to be as nice as possible to each other, even under depressing and stressful circumstances. In a word, we can say that �-dwellers enj oy a substantially higher level of reciprocal respect than a-dwellers. But there is one thing in common between a-dwellers and �-dwellers. Both a-dwellers and �-dwellers would treat any aliens coldly and disrespectfully (at the same level as a-dwellers treat their kind). That is, whereas on planet a there is an equal distribution of respect no matter whether aliens immigrate there, on planet � an alien would be treated less respectfully than the original dwellers are treated. Let's suppose that the diff erences between a and � would not lead to any further good or bad consequences. Moreover, let's also suppose that, for some reason, an alien on planet � wouldn't be aware of the fa ct that she is treated with less respect than others. Now imagine that the earth is no longer livable (perhap s due to global warming) and we need to immigrate to another planet. We can choose to immigrate to either a or �- Suppose that we know many things about the two planets, including the above-mentioned fac ts. Suppose also that we will show the same amount of respect to people in the two planets as they show to us no matter which planet we will live on. The question is: Which planet would we choose to live on? I have a strong intuition that we would desire to live on planet a. 115 Freedom Consider again two planets -f or the sake of convenience, I also call them a and � On planet a, everyone is entitled to the same scope of freedom, including political liberties, property rights, freedom of speech, etc. But for some reason, they lack the freedom of religion; they are forced to embrace one particular religion, R. On planet �' everyone possesses at least the scope of freedom a dwellers possess. But there is one diff erence: original �-dwellers enj oy one additional sort of freedom, that is, freedom of religion, whereas immigrants do not -l ike a-dwellers, they must also embrace religion R. In a word, while on planet a there is an equal distribution of freedom no matter whether aliens immigrate there, on planet � an alien would possess a narro wer scope of freedom than original dwellers have. Suppose further that the diff erences between a and � would not lead to any further good or bad consequences. As with the case about respect, let's also suppose that, for some reason, an alien on planet � wouldn't be aware of the fa ct that she en joys less freedom than others. Now we need to immigrate to another planet. We can choose to immigrate to either a or �- Suppose that we know many things about the two planets, including the above-mentioned fac ts. Suppose also that we don't want to embrace that particular religion. The question is: Which planet would we choose to live on? Again, I have a strong intuition that we would desire to live on planet a. 116 Longevity Now let's test our intuition on whether we could rationally desire equality of life years for its own sake. On planet a, everyone lives to 80 years old. On planet �' everyone lives to 12 0 years old. We need to immigrate to another planet. We can choose to immigrate to either a or �- Suppose that we would live to 80 years old on either planet. Which planet would we choose to live on? Again, I have a strong intuition that we would desire to live on planet a. There is no reason to deny that our desires in the three cases are rational. The desires are not based on any errors, incompetence or ignorance. We prefer to live on a planet (an egalitarian society) than � planet, even though, as we have assumed, all other things are equal. Since we know that this choice won't lead to other good consequences, it is reasonable to say that we intrinsically desire to live on a, not for other purposes. Then according to Principle D, it is reasonable to say that living on a planet is intrinsically valuable for us (given that our desires are rational). The general lesson is that what matters to our well-being are not only the absolute, non-comparative features of our life, but also how our lif e compares with the lives of others. To put in another way, the equality through leveling-down the well-being of the better off in some respect can intrinsically benefit people who are worse off in the original unequal position. Some might still doubt that a person can really desire equality of any of the goods for its own sake. They may wonder, if a person wouldn't know that she is treated unequally, and there wouldn't be any other bad consequence, why could she rationally care about how she compares to others? In response, I invite 117 everyone to consider a fa mous example given by Richard Hare. In one scenario, a person's wif e commits adultery and this person never comes to know of it. Besides, the adultery doesn't lead to any other bad consequence. In another scenario, everything is the same as the first scenario except that the wif e doesn't commit adultery. Very few people would doubt that the husband could rationally prefer the second scenario to the first (Hare 1971, p. 131). But if it is reasonable to believe that a person can rationally desire a fa ithful relationship for its own sake, then it should be also reasonable to believe that a person can rationally desire equality for its own sake. Now we are in a position to answer the question raised by the leveling down objector. Consider two possible situations. In one situation, an earthling lives on a (call it situation A), whereas in the other she lives on � (call it situation B). Situation A can be regarded as a leveling down result of situation B. In situation B, there is inequality regarding respect, freedom or longevity: an earthling enjoys fewer goods than other original �-dwellers. But in situation A, everyone en joys as much good as an earthling possesses in situation B. As we have argued, whereas living on a-planet could be intrinsically valuable for the earthling, living on �-planet wouldn't have this kind of value -t hat is to say, leveling down would bring the earthling a distinctive, equality-based intrinsic value. To generalize, leveling down can intrinsically benefit people that are worse off in the original unequal position. Since anyone's well-being counts, it is reasonable to conclude that equal distribution is intrinsically valuable at least in 118 a pro tanto sense - it is another issue whether or not this value outweighs losses caused by leveling-down. It is worth noticing that my thought experiments allow the chooser to know her immigration status. But according to many contractualists (such as John Rawls), if we want to figure out what kind of distribution people should choose for the whole society, the choosers should NOT know where they will end up in the society for which they choose. However, as I mentioned, the aim of my thought experiments is not to find what kind of distribution, overall speaking, we should choose, but just to discuss whether equal distribution can have intrinsic value. For my purposes, it is fine to assume that the person knows where she will end up. It is possible that �i s overall better than u, supposing that the greater goods override the value of equality (accordingly, a rational person without knowing her immigration status would choose � rather than a). The intrinsic egalitarian can agree with this, as she is only committed to the view that equal distribution has pro tanto value, but not the view that equality is the overriding value in every context. So fa r, I've used respect, freedom and longevity, as tenable examples for vindicating the intrinsic value of distributional equality. I believe that my discussion of those goods can be generalized to argue for the intrinsic value of equal distribution of some other goods. There is something I want to note. Raz seems to believe that if the equal distribution of some good G is intrinsically valua ble, then G itself must be also intrinsically valuable (Raz 2008). I am not sure that the intrinsic value of equal distribution of G necessarily requires that G 119 be intrinsically valua ble. At least, this doesn't seem to be a conceptual truth. Even though G is not intrinsically valuable, it is still possible that the equal distribution of G somehow has intrinsic value. Take moral value as an analogy. Although health care per se is just a personal good rather than a moral value, the equal distribution of health care may be morally worthy or even morally required. Similarly, although income and wealth, for example, have no intrinsic value, it is possible that the equal distribution of income and wealth is intrinsically valuable (Marmor 2003). This may or may not be true, but Raz hasn't provided any arguments to bolster his claim. None of this tends to undermine my thesis, which is the more modest claim that the equal distribution of some intrinsic goods, such as respect, freedom and longevity, is intrinsically valua ble. 2.3 Let me make some more remarks on the idea of equality of respect. It is worth noting that Stephen Darwall distinguishes between two concepts of respect, recognition respect and appraisal respect. A person could both be an object of recognition respect and an obj ect of appraisal respect. Recognition respect is the kind of attitude that a person x shows to another person y purely in virtue of the fa ct that y is a person. Appraisal respect is the kind of attitude that a person x shows to another person y in virtue of the excellence that y manif ests. As Darwall says, Persons can be the obj ect of recognition respect. Indeed, it is just this sort of respect which is said to be owed to all persons. To say that persons as such are entitled to respect is to say that they are entitled to 12 0 have other persons take seriously and weigh appropriately the fa ct that they are persons in deliberating about what to do ... There is another attitude which diff ers importantly from recognition respect but which we likewise ref er to by the term "respect". Unlike recognition respect, its exclusive objects are persons or features which are held to manif est their excellence as persons or as engaged in some specific pursuit. For example, one may have such respect for someone's integrity, for someone's good qualities on the whole, or for someone as a musician. Such respect, then consists in an attitude of positive appraisal of that person either as a person or as engaged in some particular pursuit. Accordingly the appropriate ground for such respect is that the person has manif ested characteristics which make him deserving of such positive appraisal. .. I shall call it appraisal respect. (Darwall 1977, pp. 38-39) The idea of respect that is discussed in this pro ject is not appraisal respect, but one kind of recognition respect. That is, the idea of respect is about the kind of attitude that a person x shows to another person y purely in virtue of the fa ct that y is a person -' 0 The attitudes or treatments that express our recognition respect to others could be very broad. A core type of recognition respect is respect for others' liberties and rights. However, recognition respect is not only about basic rights and liberties, but also about an important attitude that we care about in our daily life, the attitude of being nice. Let me take an example to illustrate this kind of attitude. I was having lunch in Caf e 84 the other day. After I chose the dishes that I like, I looked around trying to find a decent table. I saw a girl, who was sitting by a table alone. I asked her, "Do you mind ifl sit here?" She 30 The idea of person might be characteri zed by appealing to Rawls's idea of political conception of person. According to Rawls, a person has two moral powers. The first is the capacity for a sense of justice: it is the capacity to nnderstand, to apply, and to act fr om (not merely in accordance with) the principles of political justice. The second is the capacity for a conception of the good: it is the capacity to have, to revise, and rationally to pliTsue a conception of the good (Rawls 2001, p. 19). \Vhereas I adopt Rawls's characterization of the second moral power, I suggest revising the characterization of the first moral power. To say that x has the capacity for a sense of justice is to say that x has the capacity to nnderstand, to apply, and to act from the moral reasons for justice. 121 answered decisively, "Sure." After I put my dishes on the table, just before starting to enj oy my lunch, the girl said, "I am sorry. I need to use my computer. So I have to go over there, since there is an electronic plug." She was pointing at a table that is besides the wall of the restaurant, when she was talking to me. At first, I was a little surprised when she said that she was sorry; actually I was even a little puzzled about why she said that to me. Indeed, why does she bother to explain her intention of moving over to another table to me? Suddenly, I realized that she is purely nice, kind and respectful. She was af raid that I, a complete stranger to her, would feel uneasy or uncomfortable if she moved over to another table immediately after I sat by the table at which she was initially sitting. Darwall is quite right in characterizing the general feature of the content of recognition respect: "to say that persons as such are entitled to respect is to say that they are entitled to have other persons take seriously and weigh appropriately the fa ct that they are persons in deliberating about what to do." When the girl explained to me her reason for moving over to another table, she was taking me seriously purely in virtue of the fa ct that I am a person. This is a particular kind of recognition respect. Both the equality of rights and liberties and the equality of respect by others (in the sense of being nicely treated) are momentous to us. My thought experiments already reveal how deeply we care about the two kinds of equality. Now let me use another example to show how important the equality of being nicely treated really is. Imagine the following scenario. Smith works as a junior accountant in a prestigious accounting company. As every on weekday, he came 122 to work this morning. He first met Mary, a janitor, who was cleaning the floor. When Mary saw Smith, she warmly says, "Good morning, Smith. How are you?" Smith, however, passed by Mary without any response; he didn't even look at Mary. He was fully aware the existence of Mary, and heard her voice as well. But he had no motivation to say hello to her. Several seconds later, Smith ran into Tom, who is a senior accountant in the same company. Tom is young, social and smart. Immediately after he noticed Tom, Smith changed his fa cial expression from a cold and serious style to a warm and kind style. He says, "Morning, Tom. How is it going?" Let's assume that Smith and Tom are not friends; they are just colleagues. I believe that most of us would agree that Mary has a reason to feel upset if she knows that Smith treats her and Tom unequally in this particular way; it is rational for Mary to feel upset. This may not have anything to do with whether Mary's self -respect is aff ected or not. Mary could be a person with high level of self -respect, but still rationally feel upset about this. What Mary cares about is mainly concerned with the fa ct that Smith treats her and Tom unequally. If Smith were cold to everyone, Mary wouldn't feel so upset about being ignored by Smith. I believe that it is a fundamental psychological fa ct that we human beings care about being respected equally. The disappointing fa ct is that we, as human beings, oftentimes show unequal respect to others. We oftentimes fa il to recognize the fa ct that the two persons whom we respect unequally are both persons. We are oftentimes distracted by some unimportant diff erence between two persons, such as one 123 person is more or less fa mous than the other, one person has more or less political power than the other, one person is more or less wealthy than the other, etc. Although both the equality of rights and liberties and the equality of respect by others (in the sense of being nicely treated ) are significant, the equality of basic rights and liberties is prior to the equality of respect by others. For example, in normal circumstances, it is not rational for a person to trade off the equality of rights and liberties for having even a large amount of income and wealth, but it may be rational for a person to trade off the equality of respect by others for having a large amount of income and wealth. Notice that I am not saying that a person can never rationally desire to trade off the equality of rights and liberties for some amount of income and wealth. Moreover, I am not saying that it is always rational for a person to trade off the equality of respect by others for some amount of income and wealth. To illustrate the view that the equality of basic rights and liberties is prior to the equality of respect from others, consider the following example. Suppose that on planet u, we would receive equal respect, but would be very poor, say earn 20,000 dollars a year; on planet � ' we would receive unequal respect, but would be rich, say earn 1, 000,000 dollars a year. Suppose that all a dwellers would be as economically well off as we would be; and all �-dwellers would be as economically well off as we would be. Then it may be rational for us to choose to live on planet �- 12 4 Now suppose that on planet u, we would enj oy equal rights and liberties, but would be very poor, say earn 20,000 dollars a year; on planet �' we would wouldn't enj oy equal rights and liberties, but would be rich, say earn 1, 000,000 dollars a year. Suppose that all a-dwellers would be as economically well off as we would be; and all �-dwellers would be as economically well off as we would be. It seems irrational for us to choose to live on planet �- 3 In this section, I will consider several possible objections to my view and attempt to show that none of them really threaten my main thesis. (1) Someone might say that it is not always the case that we (intrinsically) desire distributional equality. In some circumstances, we may be indifferent to whether other people live better than us. Quite a few philosophers appeal to thought experiments regarding remote galaxies for rej ecting egalitarianism. For instance, Raz invites us to imagine the below scenario: Imagine that a state of ideal equality prevails on our planet. However, there is another planet, in a galaxy too remote for us to be able to interact with, where there are human beings living in conditions of ideal equality with each other, but whose conditions are not equal to those of people on planet earth, not equal in the respect in which equality is a good thing. The people in either planet do not know of the existence of the other, nor can they find out (given the laws of nature). Bertie is one of those people. He lives on earth. The question is: Would Bertie benefit if the conditions of the people on the other planet changed and became equal to the conditions of his lif e (and that of other people on earth)? (Raz 2008, p. 11) Let's borrow Raz's example back to our current cases. Suppose that in some possible world w,, people on that remote planet receive a greater amount of 125 respect or freedom than earthlings do. Now consider another possible world, w 2 , in which earthlings possess the same amount of respect or freedom as they do in possible world w�, but the amount of goods people on the remote planet receive is leveled down to the amount of goods earthlings possess. Again, suppose that the diff erence between the two possible worlds won't lead to any further good or bad consequences. Should we earthlings desire to live in w 2 rather than w,? No. It seems we are indiff erent to the two options. Thus, since the worse off don't always rationally desire equal distribution of goods, it seems to follow that equal distribution of goods is not an intrinsic value for the worse off. My reply would be very simple. The value of distributional equality depends upon the context in which distributional equality is imposed. To be more specific, the value of distributional equality depends upon whether equal distribution is established within the same society or not. Here I use the term 'society' in the sense that people are in the same society if they can have social, political, economic and sym bolic interactions with each other. In the contemporary era of globalization, the boundary of a society is beyond the boundary of a village, a city, or even a country; the scope of the society for a human being probably includes the most part (if not the whole) of the earth. As some philosophers argued, it may be just a normative fa ct that we should care about comparisons with other fellows in the same society, but not with people that are too remote to engage us (Marmor 2003). People who use examples regarding remote galaxies to refute egalitarianism seem to have a (f alse) assumption that if equal distribution is 126 intrinsically valuable, then it must be intrinsically valuable in every circumstance. As I have mentioned at the beginning, intrinsic values could be conditional values. In order to argue against non-egalitarians, I don't have to argue that equal distribution is always intrinsically valuable; rather, I only need to argue that equal distribution is intrinsically valuable in certain circumstances. Moreover, it may be worth mentioning that intrinsic egalitarianism does not imply the view that equality of any good is intrinsically valuab le. Temkin claims that there is some respect in which a world where only some are blind is worse than one where all are. (Temkin 19 93) I believe that this is a mistake. Suppose that on planet u, everyone is blind. On planet �' everyone is sighted. Smith would be blind on either planet. All other things being equal, could Smith rationally choose to live on planet u? We don't have strong intuition that he could rationally desire to live on planet u. (2) As psychologists and anthropologists may point out, humans generally don't want to live less well than their fellows - the emotion of envy seems to be deeply rooted in human psychology. The envy-rooted psychology can explain why we have the desire to live on a-planet rather than �-planet even though the absolute level of goods is the same; it can also explain why we don't care whether people on a remote planet live better than we do. But it is hard to say that an envy-driven desire is indicative of any value. Even if the desire for distributional equality satisfies all the requirements of rationality I listed, it doesn't neces sarily follow that distributional equality is (intrinsically ) valuable- 127 perhaps because rational desirability is not a sufficient condition for genuine values. I have two responses. First, I don't claim that there is a necessary connection between rational desirability and value. Instead I just want to say that a desire's being rational is reliable evidence that the object of the desire is valua ble. The non-egalitarian needs to provide independent argument s to show why the objects of so-called "envy-based" desires must be valueless- it is right to place the burden of proof on the shoulders of the non-egalitarian. But if the non-egalitarian merely labels any desire involving interpersonal comparison as an envy-b ased desire, I am happy to admit that the desire for distributional equality is "envy-based" according to her terminology. So what? The fa ct that an alien would feel envy in planet � may indicate that there is something wrong with the distribution of goods in �- Second, even if humans generally have envy-based desires the objects of which are valueless, it doesn't follow that our desire for equal distribution of respect or freedom, a desire after rational deliberation, must be the product of our actual envy-rooted psychology. If there are some cases in which we don't care, after rational deliberation, whether our fellows possess the same amount of goods with us, then it is dubious to appeal to envy in explaining why we have the desire for equal distribution of respect. Let's consider an example concerning distribution of pleasure. Imagine two planets. In one planet (call it planet x), people equally enj oy a reasonable amount of pleasure, say, 100 units, whereas in the other planet (call it planet y), 128 people equally enj oy a substantially higher amount of pleasure, 200 units. Now an earthling, Smith, is deciding which planet he should live on. Suppose that for unknown biological and physical reasons, Smith will have 100 units of pleasure no matter which planet he will live on. If he lives on planet x, there will be an equal distribution of pleasure. But if he lives on planet y, he will en joy a smaller amount of pleasure than others. Let's assume that all other things are equal. Which planet should Smith choose? It seems that we don't have a strong intuition that Smith should choose planet x over planet y- unlike the respect case and the freedom case, the equal distribution of pleasure doesn't seem to be intrinsically desira ble.Sl Since after rational reflection we don't always desire equal distribution of goods, it is unreasonable to explain away our (rational) desire for equal distribution of respect or freedom in terms of envy. (3) Someone may worry that my argument for the intrinsic value of distributional equality (in certain contexts) is paired with an argument for the intrinsic value of inequality in analogous contexts. Take the freedom of religion as an example. Consider two planets: C-p lanet in which everyone including aliens must practice a particular religion, and D-planet which is like C except that although every original dweller must practice this religion, aliens are free to choose whatever religion to practice. Which planet would you prefer to live in, as a resident alien? It seems we would pref er to live in planet D. 31 The discussion of why equal distribution of some goods (e.g. respect) rather than others (e.g. pleasliTe) is intrinsically valuable is beyond the scope of this paper. My tentative hy pothesis is that equal distribution of goods that involve interpersonal relationship such as respect, care, fr eedom, and opporhmities may have intrinsic value, whereas equal distribution of goods that don't involve such relationship like pleasill'e, health and capacities is less likely to have intrinsic value. 129 I agree that a resident alien would probably choose planet D. But I tend to believe that this is not because inequality is intrinsically valuable, but because the person would possess more goods (here, the freedom of religion) in D than she would have in C. To see whether people (rationally) desire inequality for its own sake, let's consider another scenario. Here are two planets, again: E-planet in which everyone (including aliens) enjoys the freedom of religion, and F-p lanet which is like E except that only aliens have freedom of religion. Which planet would you pref er to live in as a resident alien? It seems that we don't pref er planet F over planet E-w e would even feel bad about ourselves if we prefer to live in planet F than to live in planet E. Theref ore, whereas equality can be rationally desired in an intrinsic way, it is not rational to desire inequality for its own sake. ( 4) Some may worry that affirming the intrinsic value of equality would distract a person's attention from her personal project to something that she ought not to focus on. This worry is unnecessary. In fa ct, many things could be intrinsically valuable for us without being regarded as our personal goals. For example, being loved by others is intrinsically valuable for a person, but usually we won't take being loved to be among our goals. Since being loved, to a large extent, is a matter of luck, a person can only focus on what she can have control over, or what she can achieve through her own eff ort. Promoting equality is, to a large extent, something that the worse off cannot control. It relies on whether the better off would recognize it as a reason for just if ying a certain distribution. 130 Conclusion At the end of this dissertation I want to summarize the main points I have discussed. 1. In Chapter 2, I have argue against luck egalitarianism, which is the view that whatever is not a result of people's eff ort, but is rather a result ofluck, cannot serve as a desert-based reason for just if ying an unequal distribution. I have attempted to show that 1) the fa ct that a person makes more contributions than another person serves as a desert-based reason for justif ying an unequal distribution between them, even when the diff erence in contributions is not a result of people's eff ort, but a result of luck; and 2) the fa ct that it is more difficult to bring about a socially beneficial outcome than to bring about another socially beneficial outcome serves as a desert-based reason for just if ying an unequal distribution among people, even though difficulty is oftentimes a matter of luck to them. 2. Chapter 3 is on sufficientarianism, which is formulated as the con junction of two theses: 1) a primary reason for distribution is to make the citizens reach or get closer to the so-called "sufficiency threshold"; and 2) if a person reaches the sufficiency threshold, then there are no further reasons for distribution for the sake of improving her well-being. Although suffi cientarianism per se is true, the five fa mous accounts of the sufficiency threshold are all problematic: Suf ficiency as satiety, sufficiency as the satisfac tion of basic needs, Crisp's account of sufficiency that appeals to an 131 impartial spectator's compassion, Frankf urt's account of sufficiency that appeals to the pursuit of one's lif e projects, and Anderson's account of sufficiency as a list of personal goods. All the five accounts of sufficiency fa il to satisf y both of the two theses (although they may satisf y one of the two theses). I've instead proposed a comparative account of suffi ciency, which understands suf ficiency as the level of social average, i.e. the average wealth per capita. 3. Chapter 4 discusses prioritarianism, which is the view that the fa ct that a person is worse off absolutely serves as a reason for assi gning priority to benefiting her over benefiting a person who is better off. There are two important interpretations of prioritarianism, which I call PRl and PR2. According to PRl, the priority view is grounded in the principle that the moral value of benefiting the worse off is greater than the moral value of benefiting the better off. According to PR2, the priority view is grounded in the principle that a moral agent has a stronger moral reason to benefit the worse off than to benefit the better off I have argued that whereas PRl is fa lse, PR2 is tenable. I have shown that even if in some cases the moral value of benefiting the worse off is the same as the moral value of benefiting the better off, a moral agent may still have a stronger reason to benefit the worse off than to benefit the better off. Moreover, I have argued that the priority of benefiting the worse off is grounded in a more fundamental fa ct, to wit, that a person would have a stronger prudential reason to improve her worse-of f-prospect than to improve her better off -prospect. 132 4. In the final chapter, I have def ended the intrinsic value of distributional equality of some goods (e.g. respect, liberties, and longevity ). I first propose an epistemic principle concerning intrinsic personal value as follows: It is reasonable to regard x as an intrinsic value for a person S if S rationally desires x (as part of her lif e) for its own sake. Then I give some thought experiments to reveal our intuition that a person can rationally desire equality of at least some goods for its own sake. 133 Bibliography Alexander, John M. (2008) Ca pabilities and Social Justice, Ashgate Press. Andrew, Williams (2012) "The Priority View Bites the Dust?" Utilitas 24: 31 5- 331. Anderson, Elizabeth (1 999) "What is the Point of Equality?" Ethi cs 109: 287- 337. Anderson, Elizabeth (2009) "Def ending the Capabilities Approach to Justice," in Brighouse and Robeyns (eds.), Measuring Justice: Ca pabilities and Primary Goods, Cambridge University Press, 81 -10 0. Arneson, Richard (1 989) "Equality and Equal Opportunity for Welfare," Philoso phical Studies 56: 77-93. Arneson, Richard (2000) "Luck Egalitarianism and Prioritarianism," Et hics 11 0: 339-349. Arneson, Richard (2008) "Rawls, Responsibility, and Distributive Justice," in Fleurbaey, Salles and Weymark ( eds. ), Justice, Political Liberalism, and Utilitarianism, Cambridge University Press, 80-108. Arrow, Kenneth J. (1971) "A Utilitarian Approach to the Concept of Equality in Public Expenditures," The Quarterl:y Jo urnal of Economi cs 85: 409-4 15. Broome, John (1991) Wei ghing Goods, Blackwell Press. Brown, Campbell ( 2005) "Priority or Suf ficiency ... or Both?" Economi cs and Philosoph y 21: 19 9-220. Casal, Paula (2007) "Why Suf ficiency is not Enough," Et hics 11 7: 296-326. Cohen, Gerald A. (19 89) "On the Currency of Egalitarian Justice," Et hics 99: 906- 944. Cohen, Gerald A. (1 997) "Where the Action is: On the Site of Distributive Justice," Philosoph y and Public Affa irs 26: 1-30. Cohen, Gerald A. (2006) "Luck and Equality: A Reply to Hurley," Philosoph y and Phenomenological Research 72: 439-446. Cohen, Gerald A. (2008) Rescuing Justice and Equality, Harvard University Press. Crisp, Roger (2003a) "Equality, Priority, and Compassion," Ethi cs 11 3: 745-763. Crisp, Roger ( 2003b) "Egalitarianism and Compassion," Ethics 11 4: 119 -1 26. Crisp, Roger (2008) "Compassion and Beyond," Ethical Theory and Moral Practice 11 : 233-246. Darwall, Stephen (1 977) "Two Kinds of Respect," Ethi cs 88: 36-49. Darwall, Stephen (1 998) Philoso phical Et hics, Westview Press. Darwall, Stephen (2004) "Respect and the Second-Person Standpoint," Proceedings and Addresses of the American Philoso phical Association 78: 43-59. Darwall, Stephen (2006) The Second-Person Stand point- Morality, Respect and Accountability, Harvard University Press. Dworkin, Ronald (198 1a) "What is Equality? Part 1: Equality ofWelfare," Philosoph y and Public Af fa irs 10: 18 5-246. Dworkin, Ronald (198 1b) "What is Equality? Part 2: Equality of Resources," Philosoph y and Public Affa irs 10: 283-345. 134 Dworkin, Ronald (19 85) A Matter of Princi ple, Harvard University Press. Dworkin, Ronald (2000) Sovereign Virtue: the Theory and Practice if Equality, Harvard University Press. Dworkin, Ronald (2002) "Sovereign Virtue Revisited," Ethics 11 3: 106-143. Frank furt, Harry (1971) "Freedom of the Will and the Concept of a Person," Jo urnal of Philosoph y 68: 5-20. Frank furt, Harry (1 987) "Equality as a Moral Ideal," Ethics 98: 21- 43. Frank furt, Harry (1 997) "Equality and Respect," Social Research 64: 3-15. Frank furt, Harry (2000) "The Moral Irrelevance of Equality," Public Affa irs Quarterl:y 14: 87-103. Freeman, Samuel (2007a) Justice and the Social Contract, Oxf ord University Press. Freeman, Samuel ( 2007b) Rawls, Routledge Press. Gibbard, Allan (199 0) Wise Choices, Apt Feeli ngs, Harvard University Press. Gibbard, Allan (1991) "Constructing Justice," Philosoph y and Public Affa irs 20: 264-279. Hausman, Daniel M. and McPherson, Michael S. (2006) Economic Anal:ysi s, Moral Philosoph y, and Public Policy, Cambridge University Press. Harsanyi, John (19 75) "Can the Maximin Principle Serve as a Basis for Morality? A Critique of John Rawls's Theory," The Ameri can Political Science Review 69: 594-606. Hirose, Iwao (2009) "Reconsidering the Value of Equality," Australasian Journal of Philoso phy 87: 301-312. Holtug, Nils (2007) "Prioritarianism," in Holtug and Lippert-Rasmussen (eds.), Egalitarianism: Ne w Essa ys on the Nature and Value of Equality, Oxf ord University Press, 12 5-133. Huseby, Robert (2010) "Suff iciency: Restated and Defended," The Jo urnal of Political Philoso phy 18: 178-197. Kagan, Shelly (19 88) "The Additive Fallacy," Ethi cs 99: 5-3 1. Kagan, Shelly (19 98) "Rethinking Intrinsic Value," The Jo urnal of Et hics 2: 277- 297. Lewis, David (19 89) "Dispositional Theories of Value," Proceedings of the Aristotelian Society, Supplementary Vol. 63: 113 -13 7. Marmor, Andrei (2003) "The Intrinsic Value of Economic Equality," in Meyer, Paulson and Pogge (eds.), Rights, Culture and the Law: Themes fr om the Legal and Political Philosoph y if Josep h Raz, Oxf ord University Press, 12 7-141. McCarthy, David (2006) "Utilitarianism and Prioritarianism I," Economi cs and Philosoph y 22: 335-363. McCarthy, David (2008) "Utilitarianism and Prioritarianism II," Economi cs and Philosoph y 24: 1-33. Mill, John S. (1 998) Utilitarianism, Oxf ord University Press. Miller, David (1 997) "Equality and Justice," Ratio 10: 222-237. Miller, David (1 999) Princi ples of Social Justice, Harvard University Press. Miller, David (2004) "The Concepts of Desert," in Clayton and Williams (eds.), Social Justice, Blackwell Press, 18 6-2 01. 135 Nagel, Thomas (1 979) Mortal Questions, Cambridge University Press. Nagel, Thomas (1991) Equality and Partiality, Oxf ord University Press. Nozick, Robert (1 974) Anarch y, State, and Utopia, Basic Books Press. O'Neill, Martin (2008) "What Should Egalitarians Believe?" Phil osoph y and Public Af fa irs 36: 119 -156. O'Neill, Martin (2012) "Priority, Preference and Value," Utilitas 24: 332 -348. Ot suka, Michael (2002) "Luck, Insurance, and Equality," Ethi cs 11 3: 40-54. Ot suka, Michael and Voorhoeve, Alex (2009) "Why It Matters That Some Are Worse Off than Others: An Argument against the Priority View," Philosoph y and Public Affa irs 37: 171-19 9. Ot suka, Michael (2012) "Prioritarianism and the Separateness of Persons," Utilitas 24: 365-380. Parfit, Derek (1 982) "Future Generations: Further Problems," Phil osoph y and Public Affa irs 11: 113 -1 72. Parfit, Derek (1991) "Equality or Priority?" The Lindley Lecture, University of Kansas. Parfit, Derek (1 997) "Equality and Priority," Ratio 10: 202 -22 1. Parfit, Derek (2 011) On What Matters, Oxf ord University Press. Parfit, Derek (2012) "Another Def ense of the Priority View," Utilitas 24: 399- 440. Porter, Thomas (2012) "In Def ense of the Priority View," Utilitas 24: 349-364. Rabinowicz, Wlodek (2002) "Prioritarianism for Prospects," Utilitas 14: 2-2 1. Rawls, John (19 55) "Two Concepts of Rules," The Philoso phical Review 64: 3-32. Rawls, John (1971) A Theory of Justice, Harvard University Press. Rawls, John (1 974 -1 975) "The Independence of Moral Theory," Proceedings and Addresses of the American Philoso phical Association 48: 5-22. Rawls, John (1 980) "Kantian Constr uctivism in Moral Theory," The Journal if Philosoph y 77: 515 -572. Rawls, John (19 82) "Social Unity and Primary Goods," in Sen and Williams (eds.), Utilitarianism and Beyond , Cambridge University Press, 15 9-185. Rawls, John (19 93) Political Liberalism, Columbia University Press. Rawls, John (1 999a) The Law of Peo ples, Cambridge University Press. Rawls, John (1 999b) A Theory of Justice, revised edition, Harvard University Press. Rawls, John (1 999c) Collected Pa pers, Harvard University Press. Rawls, John (20 01) Justice as Fairness: A Restatement , Harvard University Press. Railton, Peter (1 986) "Moral Realism," The Philoso phical Review 95: 16 3-207. Railton, Peter (2008) "The Problem of Well-being: Respect, Equality, and the Self," Tenth Con ference of the In ternati onal Society fo r Utilitarian Studies, 1- 30. Raz, Joseph (19 86) The Morality of Freedom, Clarendon Press. Raz, Joseph (2008) "On the Value of Distributional Equality," Legal Research Pa pers Series 41: 1-19. Roemer, John (1 996) Theories of Distributive Justice. Harvard University Press. 136 Rosenberg, Alexander (19 95) "Equality, Sufficiency, and Opportunity in the Just Society," Social Philosoph y and Policy 12: 54-7 1. Samuelson, Paul (1 973) Economi cs, McGraw-Hill Press. Scanlon, Thomas (1 998) lVhat We Owe to Each Other, Harvard University Press. Scanlon, Thomas (2001) "Sen and Consequentialism," Economi cs and Philoso phy 17 : 39-50. Scanlon, Thomas (2003) "The Diversity of Objections to Inequality," The Dijji .culties of Tolerance, Cambridge University Press. Scanlon, Thomas (2009) "When Does Equality Matter?" Legal Theory Workshop, University of Calif ornia, Los Angeles. ScheiDer, Samuel (2003) "What is Egalitarianism?" Philosoph y and Public Affa irs 31: 5-39. Sen, Amartya (1 970) "The Impossibility of a Paretian Liberal," Journal of Political Economy 78: 15 2- 15 7. Sen, Amartya (19 73) "Behavior and the Concept of Pref erence," Economica 41: 241-259. Sen, Amartya (1 980) "Equality of What?" in McMurrin (ed.) Tanner Lectures on Human Values, Vol. 1, Cambridge University Press, 257-280. Sen, Amartya (19 95) In equality Reexamined, Harvard University Press. Sen, Amartya (1 999) Development as Freedom, Knopf Press. Sen, Amartya ( 2004) "Capabilities, Lists, and Public Reason: Continuing the Conversation," FeminiSt Economi cs 10: 77-80. Shroeder, Mark (2009) "Buck-Passers' Negative Thesis," Philoso phical Ex plorations 12: 341-347. Shields, Liam (2012) "The Prospects for Sufficientarianism," Utilitas 24: 101-1 17 . Smith, Michael (2002) "Exploring the Implications of the Dispositional Theory of Value," Nous 36: 329-347. Steiner, Hillel (1 997) "Choice and Circumstance," Ratio 10: 296-3 12 . Steiner, Hillel (2003) "Equality, Incommensurability, and Rights," in Meyer, Paulson and Pogge (eds.), Ri ghts, Culture and the Law: Themes fr om the Legal and Political Philosoph y of Jose ph Raz, Oxf ord University Press, 119 - 12 6. Temkin, Larry S. (1 993) In equality, Oxf ord University Press. Temkin, Larry S. (2003a) "Egalitarianism Def ended," Ethics 11 3: 764-782. Temkin, Larry S. ( 2003b) "Personal versus Impersonal Princi ples: Reconsidering the Slogan," Theoria 69: 21-31. Voorhoeve, Alex and Fleurbaey, Marc (2012) "Egalitarianism and the Separateness of Persons," Utilitas 24: 38 1-398. Watson, Gary (19 7 5) "Free Agency," The Jo urnal of Philosoph y 72: 205-220. Weirich, Paul (19 83) "Utility Tempered with Equality," Nous 17: 423-439. Williams, Bernard (1981) Moral Luck, Cambridge University Press. Williams, Andrew (2012) "The Priority View Bites the Dust?" Utilitas 24: 31 5- 331. 137

Abstract (if available)

Abstract Contemporary political philosophers have provided different reasons for distributional equality. Luck egalitarians contend that many unequal distributions cannot be justified because they are too sensitive to factors that are not a result of people’s efforts, but a result of people’s luck. Sufficientarians embrace the view that more equal distributions would make more people live a sufficiently decent life. For prioritarians, a more equal distribution is justifiable because it would bring more benefits to the worse off. ❧ Although luck egalitarians, sufficientarians and prioritarians attempt to provide different reasons for justifying more equal distributions, they share two common features. First, they tend to assume that a unifying principle or reason would suffice to justify reducing all kinds of inequalities. Second, they tend to assume that equality of any good has no intrinsic value

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Distributive equality and labor burdens

PDF

Defending Rawls's Law of Peoples against cosmopolitan critiques

PDF

Three essays on cooperation, social interactions, and religion

PDF

Reasoning with uncertainty and epistemic modals

PDF

Ideological shifts in portrayals of ethnic gangs and gangsters in New York novels and film adaptations

PDF

The boundaries of the intentional

PDF

Where number lies: plural marking, numerals, and the collective-distributive distinction

PDF

On the variety and inter-relation of ethical judgments

PDF

Queer enthusiasms: cross-gender awakening and the affective remnants of religious feeling

PDF

Essays on the empirics of risk and time preferences in Indonesia

PDF

Washington gold rush: the competition for congressional earmarks

PDF

QoS-aware algorithm design for distributed systems

PDF

Finite dimensional approximation and convergence in the estimation of the distribution of, and input to, random abstract parabolic systems with application to the deconvolution of blood/breath al...

PDF

Crime and punishment: examining proportionality issues within criminal law

PDF

When and how infants discriminate between declaratives and interrogatives

PDF

The economic and political impacts of U.S. federal carbon emissions trading policy across households, sectors and states

PDF

Feature selection in high-dimensional modeling with thresholded regression

PDF

On being bound: law, authority, and the politics of obligation

PDF

Mistaken defense and normative conventions

PDF

Imperial injuries: race, disease, and disability in North American narratives of resistance, 1908-2006

Asset Metadata

Creator Tang, Yingying (author)

Core Title Intrinsic and extrinsic values of distributional equality

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Philosophy

Publication Date 02/20/2014

Defense Date 12/16/2013

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag distributive justice,equality,intrinsic egalitarianism,luck egalitarianism,OAI-PMH Harvest,prioritarianism,sufficientarianism

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Dreher, John H. (committee chair), English, Robert D. (committee member), Watson, Gary L. (committee member)

Creator Email hannahtang@gmail.com,yingyint@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-365405

Unique identifier UC11296259

Identifier etd-TangYingyi-2258.pdf (filename),usctheses-c3-365405 (legacy record id)

Legacy Identifier etd-TangYingyi-2258.pdf

Dmrecord 365405

Document Type Dissertation

Format application/pdf (imt)

Rights Tang, Yingying

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA