[This is a simpler explanation of the argument I advanced in "Non-Egalitarianism" (Philosophical Studies 114 (2003): 147-171). --mh]
The form of egalitarianism I am concerned with holds that equality in the distribution of welfare across persons is intrinsically good. In other words, it is good for people to be equally well-off, and bad for some to be better off than others, apart from consideration of any further consequences of such equality or inequality.(1) Non-egalitarianism is the view that equality in the distribution of welfare across persons is intrinsically neutral--that is, apart from considerations of beneficial or harmful consequences, it does not matter whether people are equally well off. In the following two sections, I defend non-egalitarianism.
There are a number of reasons for thinking that equality in the distribution of welfare among persons ("interpersonal equality") is instrumentally good or instrumentally bad. For example, perhaps when people are unequal, this causes resentment and dissatisfaction on the part of those who have less, bringing their level of well-being even lower than it already was. Perhaps this even causes political instability. These would be reasons for thinking that equality is instrumentally good. On the other hand, perhaps inequality spurs people to work harder, resulting in an overall greater average level of well-being. This would be a reason for thinking that inequality is instrumentally good. These are not the sort of reasons I consider here. Here, I address only the question of whether equality has intrinsic value--that is, if we set aside the further consequences that equality may have, is it still better in itself for people to be equal?
There are many who believe equality has intrinsic value. Their reasoning is that, independent of whether inequality makes people unhappy, spurs people to work harder, and so on, it is unfair that some people fare better than others. Thus, they believe that, all other things being equal, it is better to have equality.
To test whether equality has intrinsic value, I ask you to imagine three very simple possible worlds. In each world, there are only two people, Antoinette and Bubba, who each live for 100 years.
Figure 1 depicts these three worlds.
According to egalitarianism, assuming all other things are equal (that is, there are no evaluatively relevant differences among these worlds that have not been stated in my descriptions), World 1 is clearly better than World 3. I'd like to convince you instead that World 1 and World 3 are equally good. My basic argument is this:
1. Worlds 1 and 2 are equally good.
2. Worlds 2 and 3 are equally good.
3. Therefore, worlds 1 and 3 are equally good.
To simplify the presentation, I introduce the following variables:
|V1 =||The value of World 1.|
|V2 =||The value of World 2.||V3 =||The value of World 3.|
|V2a =||The value of the first half of World 2.||V3a =||The value of the first half of World 3.|
|V2b =||The value of the second half of World 2.||V3b =||The value of the second half of World 3.|
Using these symbols, my argument goes as follows:
|1.||V1 = V2.||Premise.|
|2.||V2 = V3.||From a-e below.|
|a.||V2a = V2b.||Premise.|
|b.||V3a = V3b.||Premise.|
|c.||V2a = V3a.||Premise.|
|d.||V2 = V2a + V2b.||Premise.|
|e.||V3 = V3a + V3b.||Premise.|
|f.||V2 = V3.||From a-e.|
|3.||V1 = V3.||From 1, 2.|
The conclusion obviously follows from the premises and is obviously incompatible with egalitarianism. So now we need to examine the premises--why do I say (1), (2a), (2b), (2c), (2d) and (2e) are all true?
Why do I say that worlds 1 and 2 are equally good? Well, World 2 has the same average level of well-being (75) as World 1, they have the same number of people existing for the same amount of time, they are both perfectly equal, and we have stipulated that there are no other evaluatively relevant factors in the two worlds.
World 1 is obviously a world of equality. What about World 2? Well, in World 2, Antoinette is just as well off overall as Bubba is--both of them get 100 for half of their lives, and 50 for the other half. Both of them have an average welfare level of 75. If you are going to be dropped into a world like World 2, you have no rational reason to prefer being in Antoinette's position over being in Bubba's, or vice versa. In terms of the basic argument for egalitarianism, we can ask: to whom might World 2 be unfair? One might be tempted to answer: "For the first fifty years, life is unfair to Bubba, and for the last 50 years, it is unfair to Antoinette." But since both Antoinette and Bubba get the same amount of goods overall, it is hard to see how the world as a whole is unfair to either.
Some people might feel that Antoinette has things better in World 2, because she doesn't have to wait to get her high-quality segment of life; she thereby avoids the feeling of impatience resulting from having to wait for 50 years. On the other hand, some might feel that Bubba actually has it better, because he gets to enjoy the pleasure of anticipation for 50 years, whereas Antoinette, upon reaching age 50, knows that everything is downhill from there. In my example, however, neither of these things is the case: imagine that in fact, Bubba does not experience any extra unhappiness resulting from impatience, nor any extra pleasure resulting from anticipation. Both Antoinette and Bubba are indifferent as to when in their lives they get their goods. We also stipulate that neither of them feels any resentment or other form of unhappiness resulting from their unequal degrees of well-being. We make these stipulations so that we can focus on whether inequality is intrinsically bad, without being distracted by extraneous issues. Obviously, if world 2 contains less happiness overall, then it is worse than world 1, but examining that case would not be relevant to assessing egalitarianism; what is relevant to my argument is the case in which World 2 has the same amount of happiness overall as world 1, just differently distributed through time.
It seems that the two halves of World 2 are equally good, because they are qualitatively identical. The only difference is that Antoinette and Bubba have switched places; they have gone from Antoinette being the lucky one to Bubba being the lucky one. But neither Antoinette nor Bubba is more important than the other, so these two states of affairs are equally good.
It seems that the two halves of World 3 are equally good, because they are qualitatively identical, without even the difference of Antoinette and Bubba switching places. While there might be some reason to think World 3 is bad in one respect due to its inequality, there doesn't seem to be any reason for thinking the first 50 years of inequality are any better or worse than the later 50 years of inequality.
It seems that the first half of World 2 is as good as the first half of World 3, because they are qualitatively identical, without even the difference of Antoinette and Bubba swapping roles, and without even any temporal difference. Worlds 2 and 3 only diverge after year 50, after the first halves of the worlds have passed. What happens after year 50 should not affect the value of the first 50 years--one should not be able to retroactively improve or worsen the past.
It seems that the value of each world should be the sum of the values of its two halves. I call this principle "Temporal Additivity" (value adds over time). This principle is intuitive. However, an egalitarian might want to deny it, because (a) the egalitarian thinks inequality is intrinsically bad, and yet (b) inequality is obviously not additive over time (that is, you can't find out how much inequality the world contains by adding together the amounts of inequality that exist in each time period). In World 2, for example, we have a lot of inequality in the first fifty years, and a lot of inequality in the last fifty years; and yet the world as a whole exhibits equality, because the inequalities in the two halves of the world cancel out. So, because inequality is obviously non-additive, one might be tempted to conclude that value should be non-additive. One might be tempted to say, then, that V2 > V2a + V2b, because World 2 as a whole gets some extra value (above the value of its first segment and its second segment) due to its exhibiting equality, whereas neither the first nor the second half of the world exhibits this equality.
It is unclear where the burden of proof lies. The above argument against Temporal Additivity, obviously, assumes egalitarianism. If Temporal Additivity is intuitively plausible to begin with, then arguing against it by assuming egalitarianism--the very thesis that I am arguing against--might be said to beg the question. On the other hand, some might say that egalitarianism is so closely tied to non-additivity that my assuming Temporal Additivity begs the question against egalitarianism.
Fortunately, we need not let matters rest there. At least it is clear that if I can give an independent argument for Temporal Additivity, then the egalitarian cannot (without begging the question) continue to reject Temporal Additivity merely on the grounds of its incompatibility with egalitarianism. If the egalitarian were to do that, he would be adopting a dogmatic methodology in which no argument against his position can move him, since he will just use the assumed truth of his position to rebut any criticism.
There is an independent argument for Temporal Additivity: the denial of Temporal Additivity leads to a kind of paradox in decision theory. Imagine that God informs you that he is about to create a new world; call it World 4. This world will exist for 100 years, during which it will have no interaction with your world. God would like to give you a chance to play a role in the planning of World 4. He has already decided how the first half of this world is going to go, and He gives you a detailed description of it. Call the first half of this world, Part I. God then tells you that the second half of the world is going to go in one of two ways, which he also describes in detail--call these two possible ways Part IIa and Part IIb. Finally, suppose that you know that Part IIa is better than Part IIb, but, because value is non-additive, the combination of Part I with Part IIa is worse than the combination of Part I with Part IIb. In other words, if God creates Part IIa, then the second half of the world will be better, but the world as a whole will be worse. If Temporal Additivity is false, then all this is possible. Now suppose God asks you: "Which do you think I should create: Part I & Part IIa, or Part I & Part IIb?"
Suppose you are benevolent and just care about doing what is best--that is, you just want to make the choice that has the best consequences. So you say: "God, create World 4 with Part I & Part IIb, because that way, the world will be better overall." God says okay, then He disappears.
Fifty years later, God comes back and gives you an update: World 4 has been going for 50 years now, He tells you, and it is all going according to plan. Part I of the world has just finished, and Part II is just about to start. "By the way," He says, "I just wanted to check up on whether you stand by your earlier decision: do you still think I should allow Part IIb to come about, or would you like to change your mind and go with Part IIa instead?" You think about this latest question. You realize that since the first half of the world has already passed, there's nothing you can do to affect it now. All you can affect is how good the next 50 years will be for World 4. So, again being benevolent and wanting only to do what has the best consequences, you say: "You know, God, I have changed my mind: create Part IIa."
In both cases, you made the right decision: when God first approached you, you had to choose between (Part I + Part IIa) and (Part I + Part IIb). It was rational and correct for you to choose (Part I + Part IIb), because that was the better choice. But when God later approached you, you had to choose between Part IIa and Part IIb. It was rational and correct for you to choose Part IIa, because that was the better choice (figure 2).
This is paradoxical. A choice between (Part I + Part IIa) and (Part I + Part IIb) should be just equivalent to a choice between (Part IIa) and (Part IIb) given that Part I has occurred. If you make a rational and correct decision, and no new information appears--nothing happens that you didn't already know was going to happen--then it doesn't make sense that you should change your mind. (Imagine that someone asks you: "If it rains, do you want an umbrella?" and you say "Yes." Later, when it starts raining, he says, "So you'd like that umbrella now, right?" and you say, "No." It seems that there's some inconsistency here.) Yet this is the sort of thing that can happen if one denies Temporal Additivity. We can avoid the paradox by accepting that value adds over time.
Egalitarianism rests entirely on an appeal to intuition: the intuition that inequality is unfair and that this is bad. I defend the use of intuition in ethics. But intuitions should not be appealed to uncritically. Many intuitions are incorrect and distorted by emotional and other biases. Symptoms of the fact that an intuition is distorted include (a) that the intuition is correlated with political ideology, for example, that liberals are more likely to have the intuition than conservatives; (b) that the intuition is associated with particular emotions; and (c) that it leads to paradoxical results or conflicts with other intuitions that are less likely to be affected by bias. It is very likely that the intuition that equality is intrinsically valuable is a bias. It is, on the other hand, very unlikely that the intuition that value is additive, or that qualitatively identical events have the same value, is a bias. The latter principles have no noticeable emotions associated with them; are not important points of liberal, conservative, or any other ideology; and are in general more plausible candidates for being products of intellectual reflection. They are therefore the sort of principles that would be used to resolve disputes in ethics and to root out intuitive errors, if ever such disputes are to be resolved and such errors rooted out at all. We therefore should not respond to the argument of section 2 by rejecting one of the premises so that we can hold on to the intuition about equality. We should instead give up the value of equality in light of its incompatibility with abstract principles about value that are based on rational reflection.
1. Hereafter, I treat the intrinsic goodness of equality as equivalent to the intrinsic badness of inequality. More precisely, I am concerned (only) with whether equality is intrinsically better than inequality.