(Draft 20.12.06)
Equality, Priority, and Distributive Justice
Robert Huseby
The Ethics Programme and
Department of Political Science
University of Oslo
()
Abstract
In this paper I discuss and criticizetelic egalitarianism and prioritarianism.First, I consider some of the weaknesses of egalitarianism, focusing on the leveling down objection. In particular, I concentrate on a recent attempt to save the equality view from the force of this objection. In addition I present a continuation of the leveling down objection, claiming that on some versions, telic egalitarianism is even committed to the view that it would in one respect be better if some people perished, if this would lead to a more equal distribution. This argument, however, leans on the premise that inequality is in it self bad. The paper also considers the merits of prioritarianism. This view is by many thought to be superior to egalitarianism. Even so, I argue that this principle faces problems of its own. The paper concludes with a brief defense of the principle of sufficiency.
Equality, Priority, and Distributive Justice
1. Introduction[1]
Quite a few political theorists are self-professed egalitarians. Being an egalitarian implies that one values equality between people. As has become clear over the last decades, however, equality is in no way a straightforward concept. For instance, there are many different things that that one might think ought to be equal. What this equalisandum is thought to be varies a great deal. Contenders include resources, welfare, opportunity for welfare, capabilities, and access to advantage. I will not deal explicitly with that debate. Instead I will, in line with much of the literature discussed in this paper simply assume that some notion of welfare is the most plausible candidate.[2] I will alsoabstract from the further question of whether distribution of welfare (equal or not) ought to be conditioned by desert, merit, or responsibility.[3]
It seems to me that distributional equality is less vital than many egalitarian writers make it out to be. In the following, I will join ranks with those who suspect the ideal of equality to rest on weak foundations. In the firstpart of this paper (Sections 2-4), I will present some considerations that count against equality. Although these considerations surely will fall short of being conclusive, I take them to carry enough weight to prompt the search for alternatives. In the remainder (Sections5-7), I will consider the principle of priority. Prioritarianism has become a viable distributive ideal in the last years. One of the reasons is that this principle avoids some of the most troublesome implications of egalitarianism. I will argue that priority is nevertheless also problematic. It should be mentioned that the rejection of equality and priority presented here forms the negative part of a defense of sufficiency. Although I will not say very much about sufficiency in this paper, it is this principle that to me seems the best ideal for distributive justice.[4] I provide a brief account of the principle in the last section.
2. Equality
In the following I will discuss equality as it has been developed and elaborated by (among others)Derek Parfit and Larry Temkin. On this view, equality is concerned with how individuals compare with each other in certain respects. According to Parfit, egalitarians accept the
(1) Principle of equality: It is in itself bad if some people are worse off than others.[5]
This principle defines telic egalitarianism. Telic egalitarianism holds that unequal outcomes are in one respect bad. If two or more persons or groups enjoy different levels of welfare, then this is bad in some respect even when no one has deliberatively, or even causally, brought the situation about, and even if no person suffers or experiences any sense of loss. This is opposed to deontic egalitarianism. The latter rejects the above principle, and value equality for some “other moral reason”.[6]According to deontic egalitarianism, an outcome is not good or bad in itself regardless of how it was brought about. The outcome is bad if it was produced in the wrong way, for instance through unequal treatment. I will be concerned only with telic egalitarianism.[7]
In addition, equality is often thought to be intrinsically valuable, in the sense that it is valuable beyond the extent to which it serves to promote other goals.[8]It is easy to see how equality may be non-instrumentally valuable in certain situations. For instance, it may serve to dampen the potential for conflict over resources between different groups, or it may serve to secure the foundations for distributional schemes withinpolitical collectives. Nevertheless, such positive side-effects are essentially contingent, and on the view I am discussing here, equality is valuable even in the absence of any fortunate consequences.
Equality is also said to be impersonally valuable. This implies that one outcome may be superior, in one respect, to another, even if no person in the first outcome is better off than any person in the latter. Equality can have value independently from the value it may have for any individual person.[9]
It is common among egalitarians, however, to accept that there are also other moral values, for instance utility, that are important, and that these sometimes conflict with the value of equality.[10] Theories that combine equality with other ideals may be labeled pluralist egalitarian theories.
In spite of the fact that equality has received vast attention from political theorists and philosophers, Parfit suggests that many egalitarian writings are implicitly more concerned with sufficiency or priority, than with equality proper.[11] Consider the following societies:[12]
Society 1: This society is egalitarian. Resources are inadequate, but evenly distributed. As a result, everyone is starving to death.
Society 2: This society is inegalitarian. Resources are ample. No one starves. Everyone, including the worst off individuals, have at least some leisure time, and some resources to spare so that they can engage in personal projects.
If we had to choose between these two societies, we would probably go for the latter,[13] even though the first is completely egalitarian. Taken alone, however, the principle of equality would prefer 1. The fact that strict egalitarianism would favor situations marked by equality over situations marked by inequality, even if all persons are better off in the latter situation, is widely referred to as the leveling down objection.[14]
The leveling down objection is not restricted to situations in which the worst off actually suffer. Although it may be thought to have particular salience in such cases, it is also problematic in other circumstances. Compare society 2 with:
Society 3: This society is egalitarian. Resources are adequate and evenly distributed. Everyone enjoys a minimally sufficient level of welfare, equivalent to the worst off in 2.
In this case as well, the leveling down objection holds that it would be absurd to consider 3 to be preferable to 2, since in 2 no one is worse off, but some are far better off.
The force of the leveling down objection is surely recognized by contemporary egalitarians. Parfit and Temkin, for instance, argue that the objection can be met if egalitarians allow for other values, such as utility or efficiency. On this view, society 2 is preferable to both 1 and 3 because the amount of utility is so much greater in 2 (or so we can assume) that it outweighs the greater equality in 1 and 3.[15] This admission is, besides being reasonable, thought to take the edge off the leveling-down objection.
Even though egalitarians would prefer society 2 to societies 1 and 3 all things considered, many would concede, as Temkin does, that there is one sense in which both 1 and 3 is better than 2. On such a view, 1 and 3 are better than 2 in respect of being equal. That is, 1 and 3 are better in virtue of their equality, although worse in virtue of their inferior utility. Initially, this idea is plausible enough. Most people would accept that one solution may be preferable, all things considered, even though an alternative solution may be superior in one particular respect.
Even so, one may doubt that equality considered by itself makes a situation superior in any respect to an alternative situation in which no one is actually worse off. According to Crisp, this very claim counts heavily against egalitarianism. The reason, he says, is that things that are “independent of the welfare of individuals” cannot be morally salient.[16] In leveling down situations, some are made worse off, while no one is made better off. Crisp denies that there can be any respect in which such a situation can be better.
Note that the problem posed by the leveling down objection is general. If leveling down is not in any respect good, then equality is not in any respect good (unless defined contingently). In order to establish that the mere relation between the welfare of different individuals or groups matters morally in itself, rather than the absolute level of welfare of either group, the egalitarian needs to claim that equality matters impersonally. If it wasn’t for this claim, egalitarians would have no way of ever preferring a situation of equality over any situation of inequality even when the total utility level is the same.
3. Can Leveling Down be Defended?
If equality must be impersonally valuable in order to be valuable at all, then it rests on dubious grounds. And this, then, is not dependent on the leveling down objection. The objection only serves to highlight the implausibility of the view. To see this consider the argument presented by Campbell Brown.[17] Although Brown rejects the leveling down objection, I think his grounds for doing so can just as well be utilized in rejecting egalitarianism.
Brown wants argue that the leveling down objection fails. To this end he starts with the following question: What is it for something to be better in a respect? He suggests a comparison with two cricket players, Pollock and Cairns, and says that although Pollock is a better bowler, Cairns is a better all-round player, all things considered. If we imagine that as far as these players’ hypothetical twins (Pollock* and Cairns*) are concerned, bowling prowess is the only relevant difference, (that is, difference in any “all-rounder-making respects”), between the two. He then stipulates that there are no more than two respects of this kind: “bowling prowess and batting prowess”.[18]
My proposal is that in order to see what the claim that Pollock is better than Cairns in respect of bowling prowess comes to, we need to entertain these three propositions: (a) Pollock* and Pollock do not differ in respect of bowling prowess; (b) Cairns* and Cairns do not differ in respect of bowling prowess; and (c) Pollock* and Cairns* do not differ in respect of batting prowess. Given (a), (b), and (c), Pollock is better than Cairns, in respect of bowling prowess, just in case Pollock* is better than Cairns*, all things considered. In this way, we can analyse an in-a-respect betterness claim by reducing it to a (counterfactual) all-things-considered betterness claim.[19]
When applying such reduction to the leveling down scenario, we must stipulate two hypothetical states A* and B* which do not differ in any respect other than the respect of equality. All other possibly relevant aspects, such as total welfare or utility are equal for A* and B*.[20]
Browns example has the following features: In situation A group x have a welfare level of 10, while group y has a welfare level of 2 (A10/2). In (leveled down) situation B, both groups have a welfare level of 1 (B1/1). Situation A* is equal to A (A*10/2), while in B* both groups have a welfare of 6 (B*6/6). Thus, A* and B* are equal in all relevant respects (in this case total utility) except the respect of equality. The purpose, of course is to underline that B is, in fact better in one respect than A, because, if all other relevant respects were neutralized then B* would be better than A*, all things considered. Therefore, it is not implausible to argue that B is in one respect better than A. In addition, Brown argues that denying that B is in any respect better than A, is equivalent to denying that B* is in any respect better than A*. This leads him to conclude that
if there is disagreement over A* and B*, then we need not bother with A and B; all the leveling down business becomes quite redundant. For, if it was false that B* was better than A*, all things considered, then that in itself would be perfectly sufficient grounds for rejecting egalitarianism; hence there would be no point in considering what the right thing to say about A and B was.[21]
There is nevertheless reason to doubt whether this reduction of sentences really tells us much more than we already knew. There is no question that more equality obtains in 1/1 than in 12/2. No one could deny that. Therefore, it seems also to be true that it is correct in a certain sense, to say that 1/1 is better than 12/2 with respect to equality. I do not think that proponents of the leveling down objection in general are unable to agree that some object x can be better than another object y in respect of r. Instead, what they would argue is that the leveling down objection brings out the weak foundations that equality rests upon.
What Brown’s argument does bring out with clarity is that the problem, if any, revealed by the leveling down objection goes all the way down. It is a general problem. Arguing that there is no respect in which 1/1 is better than 12/2 is equivalent to saying that there is no respect in which 6/6 is better than 2/12. But this is true, of course, only so long as the respect in question remains the respect of equality. One might perfectly well agree that 6/6 is preferable to 12/2, but this need not entail that the reason is that 6/6 is more equal. One might think, for instance that in 6/6 the half who increased their share (from 2) benefits more than the half who got their share reduced (from 12) suffered. But this is a contingent matter. Here equality is instrumentally valuable in a way that it would not be in 12/2 versus 1/1.
What Brown underlines, in essence, is that 1/1 is more equal in distribution than 12/2. He does not show that there is a respect in which 1/1 is better than 12/2. In order to show this, it seems, he must show that equality (in distribution) is intrinsically valuable. If there is more x in A than in B, then this does not imply that A is better than B. And this further, and vital, argument is still missing. To compare: If God one day removed all the colors in the world except green, there would be no denying that more green would obtain in the world. The world would, in one sense, be better in respect of greenness. One might of course try to argue that Green World is better than Ordinary World, but that would require a justification for the goodness of green. The crucial question of course, is whether greenness is in fact good. The same goes for equality. If equality is intrinsically and impersonally valuable, 1/1 is in one respect better than 2/12 (and 49/50 for that matter). Throughout his analysis, Brown actually presupposes that equality is good. If he is correct, then he is correct all the way down. If he is not, on the other hand, he is wrong all the way down. If equality is not intrinsically valuable, then 6/6 is no better than 2/12 for reasons of equality. Brown seems to deny this. He assumes that to the extent that we would prefer 6/6 to 2/12, we are more or less committed to the view that there is some respect in which 1/1 is better than 2/12. But this need not be the case. We might have instrumental reasons to prefer an equal situation E to an unequal situation U that has nothing to do with equality as such, and that does not commit us to the view that any situations in which equality obtains is in one respect better than any situation in which equality does not obtain. If it is the case, for instance, that the level of sufficiency is located above 2, but below 6, the principle of sufficiency would clearly say that 6/6 is better than 2/12.
4. The Wiping Out Objection
The real question remains of course. Is equality in distribution important at all? Compare the welfare of two societies. These societies occupy Parfit’s “divided world”. This means that they are separated and now nothing about each others’ existence.
(a) Half at 100, Half at 200,
and
(b) Both at 150,
In order to focus on the right issue here, remember that the total amount of utility in 4 and 6 are roughly the same, so that no society is superior to the other on that score. The sole difference between the two is that 6 is egalitarian and 4 is not. In both societies all persons have a sufficient amount of resources to lead a decent and worthwhile life. (Assume that the level of sufficiency is well below 100.) No one’s existence is at stake, and no important needs are unmet. Is 6 morally preferable to 4? Many egalitarians would claim that this is the case. To them, equality is intrinsically valuable.
Those who hold 6 to be preferable to 4 must answer why, when everyone has enough of what is important, it would be better in itself that everyone had equally much of it. Clearly, no reference to needs or basic resources is appropriate in order to answer this question. No one lacks anything that is vital to lead a decent life. It is the fact that inequality obtains that must be explained to be bad. Temkin argues that concern for equality is best seen as a part of a concern for “fairness that focuses on how people fare relative to others”.[22] It is thus the relation between A and B that makes a situation bad with respect to equality.