I. Two Theories of Equality
Equality is a popular but mysterious political ideal. People can become equal (or at least more equal) in one way with the consequence that they become unequal (or more unequal) in others. If people have equal income, for example, they will almost certainly differ in the amount of satisfaction they find in their lives. It does not follow, of course, that equality is worthless as an ideal. But it is necessary to state, more exactly than is commonly done, what form of equality is finally important.
This is not a linguistic or even conceptual question. It does not call for a definition of the word “equal” or an analysis of how that word is used in ordinary language. It requires that we distinguish various conceptions of equality, in order to decide which of these conceptions (or which combination) states an attractive political ideal, if any does. That exercise may be described, somewhat differently, using a distinction I have drawn in other contexts. There is a difference between treating people equally, with respect to one or another commodity or opportunity, and treating them as equals. Someone who argues that people should be more equal in income claims that a community that achieves equality of income is one that really treats people as equals. Someone who urges that people should instead be equally happy offers a different and competing theory about what society deserves that title. The question is then: which of the many different theories of that sort is the best theory?
In this chapter and the next I discuss one aspect of that question, which might be called the problem of distributional equality. Suppose some community must choose between alternative schemes for distributing money
and other resources to individuals. Which of the possible schemes treats people as equals? This is only one aspect of the more general problem of equality, because it sets aside a variety of issues that might be called, by way of contrast, issues about political equality. Distributional equality, as I describe it, is not concerned with the distribution of political power, for example, or with individual rights other than rights to some amount or share of resources. It is obvious, I think, that these questions I throw together under the label of political equality are not so independent from issues of distributional equality as the distinction might suggest. Someone who can play no role in determining, for example, whether an environment he cherishes should be preserved from pollution is poorer than someone who can play an important role in that decision. But it nevertheless seems likely that a full theory of equality, embracing a range of issues including political and distributional equality, is best approached by accepting initial, even though somewhat arbitrary, distinctions among these issues.
I shall consider two general theories of distributional equality. The first (which I shall call equality of welfare) holds that a distributional scheme treats people as equals when it distributes or transfers resources among them until no further transfer would leave them more equal in welfare. The second (equality of resources) holds that it treats them as equals when it distributes or transfers so that no further transfer would leave their shares of the total resources more equal. Each of these two theories, as I have just stated them, is very abstract, because, as we shall see, there are many different interpretations of what welfare is, and also different theories about what would count as equality of resources. Nevertheless, even in this abstract form, it should be plain that the two theories will offer different advice in many concrete cases.
Suppose, for example, that a man of some wealth has several children, one of whom is blind, another a playboy with expensive tastes, a third a prospective politician with expensive ambitions, another a poet with humble needs, another a sculptor who works in expensive material, and so forth. How shall he draw his will? If he takes equality of welfare as his goal, then he will take these differences among his children into account, so that he will not leave them equal shares. Of course he will have to decide on some interpretation of welfare and whether, for example, expensive tastes should figure in his calculations in the same way as handicaps or expensive ambitions. But if, on the contrary, he takes equality of resources as his goal, then, assuming his children have roughly equal wealth already, he may well decide that his goal
requires an equal division of his wealth. In any case the questions he will put to himself will then be very different.
It is true that the distinction between the two abstract theories will be less clear-cut in an ordinary political context, particularly when officials have very little information about the actual tastes and ambitions of particular citizens. If a welfare-egalitarian knows nothing of this sort about a large group of citizens, he may sensibly decide that his best strategy for securing equality of welfare would be to establish equality of income. But the theoretical difference between the two abstract theories of equality nevertheless remains important in politics, for a variety of reasons. Officials often do have sufficient general information about the distribution of tastes and handicaps to justify general adjustments to equality of resources (for example by special tax allowances) if their goal is equality of welfare. Even when they do not, some economic structures they might devise would be antecedently better calculated to reduce inequality of welfare under conditions of uncertainty, and others to reduce inequality of resources. But the main importance of the issue I now raise is theoretical. Egalitarians must decide whether the equality they seek is equality of resources or welfare, or some combination or something very different, in order plausibly to argue that equality is worth having at all.
I do not mean that only pure egalitarians need take any interest in this question. For even those who do not think that equality is the whole story in political morality usually concede that it is part of the story, so that it is at least a point in favor of some political arrangement, even if not decisive or even central, that it reduces inequality. People who assign equality even this modest weight must nevertheless identify what counts as equality. I must emphasize, however, that the two abstract conceptions of equality I shall consider do not exhaust the possible theories of equality, even in combination. There are other important theories that can be captured only artificially by either of these. Several philosophers, for example, hold meritocratic theories of distributional equality, some of which appeal to what is often called equality of opportunity. This claim takes different forms; but one prominent form holds that people are denied equality when their superior position in either welfare or resources is counted against them in the competition for university places or jobs, for example.
Nevertheless the claims of both equality of welfare and equality of resources are both familiar and apparent, and it is these that I shall consider. In this chapter I examine, and on the whole reject, various versions of the
former claim. In Chapter 2
I shall develop and endorse a particular version of the latter. I might perhaps add two more caveats. It is widely believed that certain people (for example criminals) do not deserve distributional equality. I do not consider that question, though I do raise some questions about merit or desert in considering what distributional equality is. John Rawls (among others) has questioned whether distributional equality might not require deviations from an equal base when this is in the interests of the then worst-off group, so that, for example, equality of welfare is best served when the worst-off have less welfare than others but more than they would otherwise have. I discuss this claim in the next chapter, with respect to equality of resources, but not in this one, where I propose that equality in welfare is not a desirable political goal even when inequality in welfare would not improve the position of the worst-off.
II. A First Look
There is an immediate appeal in the idea that insofar as equality is important, it must ultimately be equality of welfare that counts. For the concept of welfare was invented or at least adopted by economists precisely to describe what is fundamental in life rather than what is merely instrumental. It was adopted, in fact, to provide a metric for assigning a proper value to resources: resources are valuable so far as they produce welfare. If we decide on equality, but then define equality in terms of resources unconnected with the welfare they bring, then we seem to be mistaking means for ends and indulging a fetishistic fascination for what we ought to treat only as instrumental. If we want genuinely to treat people as equals (or so it may seem), then we must contrive to make their lives equally desirable to them or give them the means to do so, not simply to make the figures in their bank accounts the same.
This immediate attraction of equality of welfare is supported by one aspect of the domestic example I described. When the question arises how wealth should be distributed among children, for example, those who are seriously physically or mentally handicapped do seem to have, in all fairness, a claim to more than others. The ideal of equality of welfare may seem a plausible explanation of why this is so. Because they are handicapped, the blind need more resources to achieve equal welfare. But the same domestic example also provides at least an initially troublesome problem for that ideal. For most people would resist the conclusion that those who have
expensive tastes are, for that reason, entitled to a larger share than others. Someone with champagne tastes (as we might describe his condition) also needs more resources to achieve welfare equal to those who prefer beer. But it does not seem fair that he should have more resources on that account. The case of the prospective politician, who needs a great deal of money to achieve his ambitions to do good, or the ambitious sculptor, who needs more expensive materials than the poet, perhaps falls in between. Their case for a larger share of their parent’s resources seems stronger than the case of the child with expensive tastes, but weaker than the case of the child who is blind.
The question therefore arises whether the ideal of equality in welfare can be accepted in part, as an ideal that has a place, but not the only place, in a general theory of equality. The theory as a whole might then provide that the handicapped must have more resources, because their welfare will otherwise be lower than it could be, but not the man of champagne tastes. There are a number of ways in which that compromise within the idea of equality might be constructed. We might, for example, accept that in principle social resources should be distributed so that people are as equal in welfare as possible, but provide, by way of exception, that no account should be taken of differences in welfare traceable to certain sources, such as differences in tastes for drink. That gives equality of welfare the dominant place, but it prunes the ideal of certain distinct and unappealing consequences. Or we might, at the other extreme, accept only that differences in welfare from certain specified sources, such as handicaps, should be minimized. On this account equality of welfare would play only a part—perhaps a very minor part—in any general theory of equality, whose main political force must then come from a very different direction.
I shall postpone, until later in this chapter, the question of how far such compromises or combinations or qualifications are in fact available and attractive, and also postpone until then consideration of the particular problems of expensive tastes and handicaps. But I want to single out and set aside in advance one form of objection to the feasibility of compromises of equality of welfare. It might be objected, against any such compromise, that the concept of welfare is insufficiently clear to permit the necessary distinctions. We cannot tell (it might be said) how much any welfare differences between two people who have equal wealth are in fact traceable to differences in the cost of their tastes or in the adequacy of their physical or mental powers, for example. So any theory that embraces equality of welfare must pay attention
to people’s welfare as a whole rather than welfare derived or lost through any particular source. Obviously, there is much in this sort of objection, though how strong the objection is must depend on the form of compromise proposed. I want, however, to set aside all objections about the feasibility of distinguishing welfare sources.
I also want to set aside the more general objection, that the concept of welfare is itself, even apart from distinctions as to source, too vague or impractical to provide the basis for any theory of equality. I said earlier that there are many different interpretations or conceptions of welfare, and that a theory of equality of welfare that uses one of these will have very different consequences, and will require a very different theoretical support, from a theory that uses another. Some philosophers think of welfare as a matter of pleasure or enjoyment or some other conscious state, for example, while others think of it as success in achieving one’s plans. We shall later have to identify the leading conceptions of welfare and look at the different conceptions of equality of welfare they supply. But we may notice, in advance, that each of the familiar conceptions of welfare raises obvious conceptual and practical problems about testing and comparing the welfare levels of different people. Each of them has the consequence that comparisons of welfare will often be indeterminate: it will often be the case that of two people neither will have less welfare, but that their welfare will not be equal. It does not follow, however, that the ideal of equality of welfare, on any interpretation, is either incoherent or useless. For that ideal states the political principle that, so far as is possible, no one should have less welfare than anyone else. If that principle is sound, then the ideal of equality of welfare may sensibly leave open the practical problem of how decisions should be made when the comparison of welfare makes sense but its result is unclear. It may also sensibly concede that there will be several cases in which the comparison is even theoretically pointless. Provided these cases are not too numerous, the ideal remains both practically and theoretically important.
III. Conceptions of Equality of Welfare
There are several theories in the field about what welfare is, and therefore several conceptions of equality of welfare. I shall divide what I consider the most prominent and plausible such theories into two main groups, without, however, supposing that all the theories in the literature can fit comfortably into one or the other. The first group I shall call success theories of welfare.
These suppose that a person’s welfare is a matter of his success in fulfilling his preferences, goals, and ambitions, and so equality of success, as a conception of equality of welfare, recommends distribution and transfer of resources until no further transfer can decrease the extent to which people differ in such success. But since people have different sorts of preferences, different versions of equality of success are in principle available.
People have, first, what I shall call political preferences, though I use that term in a way that is both narrower and more extended than the way it is often used. I mean preferences about how the goods, resources, and opportunities of the community should be distributed to others. These preferences may be either formal political theories of the familiar sort, such as the theory that goods should be distributed in accordance with merit or desert, or more informal preferences that are not theories at all, such as the preference many people have that those they like or feel special sympathy for should have more than others. Second, people have what I shall call impersonal preferences, which are preferences about things other than their own or other people’s lives or situations. Some people care very much about the advance of scientific knowledge, for example, even though it will not be they (or any person they know) who will make the advance, while others care equally deeply about the conservation of certain kinds of beauty they will never see. Third, people have what I shall call personal preferences, by which I mean their preferences about their own experiences or situation. (I do not deny that these types of preferences might overlap, or that some preferences will resist classification into any of the three categories. Fortunately my arguments will not require the contrary assumption.)
The most unrestricted form of equality of success that I shall consider holds that redistribution should continue until, so far as this is possible, people are equal in the degree to which all their various preferences are fulfilled. I shall then consider the more restricted version that only nonpolitical preferences should be counted in this calculation, and then the still more restricted version that only personal preferences should count. More complex versions of equality of success, which combine the satisfaction of some but not all preferences from the different groups, are of course available, though I hope that the arguments I make will not require me to identify and consider such combinations.
The second class of theories of welfare I shall call conscious-state theories. Equality of welfare linked to that sort of theory holds that distribution should attempt to leave people as equal as possible in some aspect or quality
of their conscious life. Different conceptions of that ideal are constructed by choosing different accounts or descriptions of the state in question. Bentham and other early utilitarians took welfare to consist in pleasure and the avoidance of pain; equality of welfare, so conceived, would require distribution that tended to make people equal in their balance of pleasure over pain. But most utilitarians and other partisans of the conscious-state conception of welfare believe that “pleasure” and “pain” are much too narrow to represent the full range of conscious states that should be included. For example, “pleasure,” which suggests a specific kind of sensuous glow, poorly describes the experience produced by a harrowing piece of drama or poetry, an experience people nevertheless sometimes aim to have, and “pain” does not easily capture boredom or unease or depression.
I do not wish to discuss the issues this dispute raises. Instead I shall use the words “enjoyment” and “dissatisfaction” indiscriminately to name the full range of desirable and undesirable conscious states or emotions that any version of a conscious-state conception of equality of welfare might suppose to matter. This usage gives those words, of course, a broader sense than they have in ordinary language, but I intend that broad sense, provided only that they must name conscious states that people might aim to have or avoid for their own sakes, and states that are introspectively identifiable.
People often gain enjoyment or suffer dissatisfaction directly, from sensuous stimulation through sex or food or sun or cold or steel. But they also gain enjoyment or suffer dissatisfaction through the fulfillment or defeat of their preferences of different sorts. So there are unrestricted and restricted versions of the conscious-state conception of equality of welfare parallel to the versions I distinguished of conceptions of equality of success. One version aims to make people more equal in enjoyment without restriction as to source, another only in the enjoyment they take directly and from nonpolitical preferences, and another in the enjoyment they take directly and from personal preferences only. As in the case of equality of success, more discriminating versions that combine enjoyment from subdivisions of these different sorts of preferences are also available.
I shall also consider, though only very briefly, a third class of conceptions of equality of welfare, which I shall call objective conceptions. Many subdivisions and further classifications among these three classes of conceptions, beyond those I have just noticed, would have to be considered in any full account of possible theories of welfare, and there are theories of welfare not represented in this list at all. But these seem the most plausible candidates
for constructing theories of distribution. I shall just mention, however, two sorts of complexities that we should at least bear in mind. First, many of the...