Temkin holds that our notion of inequality is complex, multifaceted, and incomplete, and he provides a wealth of ingenious examples to support this view. This by itself—irrespective of the book’s many other virtues—would suffice to make it an essential part of the literature on inequality. He shows, for instance, that some intuitions support the view that complaints about inequality should be based on divergence from the individual who is best off, that others support the view that it should be based on divergence from an average individual, and that still others support the view that it should be based on divergences from everyone better off than the person with a complaint. He shows too that there is intuitive support for a variety of views as to which complaints should count. The maximin principle honors only those of the lowest class. The additive principle honors all complaints and uses their sum to determine the amount of inequality. The weighted additive principle also honors all complaints but gives extra weight to greater complaints so that a complaint that was twice as serious as another complaint might count four times as much.
But consider (P2) the principle that, ceteris paribus, it is wrong to change a situation for the sake of any increase in equality, however great, if it involves any loss, however small, for the disadvantaged. Surely P2 is highly plausible, but even though P2 isn’t directly concerned with the value of equality, and Temkin doesn’t hold that equality is all that matters, he should still reject it. For if he granted that an enormous gain in equality would not justify a minute loss for the disadvantaged, he could ascribe virtually no value to equality per se.
Both PI and P2 are relevant to the following case. Planet Alpha has a thousand blissful occupants; planet Beta, a thousand moderately happy occupants. The planets are thousands of light years apart so that Alphans can’t help the Betans. But there is also a fairly powerful god. Troubled by the inequality, he goes to Beta explaining that he can produce complete equality, but only by lowering their happiness level a bit and reducing the level on Alpha to that on Beta. He is not sure that the enormous loss for the Alphans would not outweigh the gain in equality, but he argues that if he can solve that problem the Betans should accept the loss—say a holiday a year—for the great gain in equality. Puzzled by their refusal, he calls it selfish and immoral. They protest vigorously on the basis of P2 and add that inequalities should be evaluated from the point of view of an insider rather than an outsider.
Temkin defends the external point of view, arguing that a world where happiness was linked to desert would be better than a somewhat happier world where it wasn’t (p. 260). But, as I’ve argued elsewhere, the kinds of freedom offered by soft determinists cannot justify the ascription of intrinsic value to distribution based on desert, and the Kantian kind of freedom is logically impossible since it is incompatible with both determinism and indeterminism (R. I. Sikora, “Utilitarianism and Applied Ethics,” in Applied Ethics: A Reader, ed. E. R. Winkler and J. R. Coombs [Oxford: Blackwell, 1993]: 87— 109).
Temkin also defends the external point of view by suggesting that a world with virtuous people might be better than a world where people were happier but less virtuous. But there are two problems. First the claim that virtue is intrinsically valuable rather than only valuable because virtuous behavior is conducive to happiness is usually regarded as contingent on our being free in the Kantian sense and, therefore, ultimately responsible for our behavior. Second, if, in spite of this, virtue is intrinsically valuable, it would seem to be good from one’s own point of view to be virtuous, even if this entails some loss in happiness, so that a world with less happiness but more virtue could be better from an internal point of view than a happier but less virtuous world.
As for the Betans’ appeal to P2, Temkin would object that they are motivated by self-interest rather than concern for the overall outcome. But they would respond that, even ignoring the loss to the Alphans, and despite the immense increase in equality, there cannot be an overall improvement if it entails any loss for the disadvantaged. Temkin would reply that he is only concerned here with improvements in equality, not overall improvements. But if he concedes that P2 provides a necessary condition for an overall improvement, it would be a barren victory since overall considerations trump all other considerations.
He also defends PI against Parfit’s view that it is all right to cause inequality by adding happy people at no cost to happier people who already exist by arguing that, since adding happy people is morally neutral, the egalitarian consideration can still prevail (pp. 252,253). But since adding wretched people is bad, if adding happy people were neutral, ceteris paribus, we should prevent the existence of any large group with even one wretched person. For even if all the rest would be happy, if adding them were neutral, it could not offset the badness of adding the wretched person.
In actual practice the denial of PI may have little effect as long as it is granted that increases in equality may be outweighed by decreases in utility. This raises the question of how much each factor should count when there is a conflict between them. Here as elsewhere, he seems to believe that there are objective moral truths, but he does not consider the formidable meta-ethical objections to any view that implies that there are objective moral qualities (e.g., see P. F. Strawson, “Ethical Intuitionism,” Philosophy 24 : 347-57). This is understandable—for example, most writers on applied ethics do the same—but this omission leaves an important gap in his arguments.
Even if there are no objective values, there may be areas where most of us either share much the same intrinsic values or would do so if we were not handicapped by rational errors, ignorance of concepts or facts, or the lack of vivid ideas of the values in question. Although this sort of agreement is contingent on what we are capable of caring about, such values are “quasi-objective” because of the possibility, at least ideally, of attaining widespread agreement on them by cognitive means. I suspect that a largely hedonic theory of values is quasi objective. This leaves room for the ascription of extra-hedonic value to such things as knowledge, mystical experiences, aesthetic experiences, and romantic love, but it seems unlikely that there would anything approaching enough agreement on the value of any of these for them to qualify as quasi objective, and I suspect that the same would hold true of the amount of intrinsic value that various people would ascribe to equality versus utility when there was a conflict between them.
Fortunately, the question of how much utility and equality should count in comparison with one another can usually be avoided. Richard Brandt has argued persuasively that—aside from the need for incentives and special needs of the handicapped—the equal distribution of wealth is likely to maximize utility, since deviations from an equal distribution are likely to lead to larger losses when they are harmful than to gains when they are beneficial (Richard Brandt, A Theory of the Good and the Right [Oxford: Clarendon Press, 1979], p. 315).
Although I have attacked Temkin’s central thesis, there is much that I agree with in his book, particularly his claim that it is inequalities between individuals that count rather than inequalities in the average level of various classes. Furthermore, he is right in observing that “since even mistaken views can be illuminating, there is reason to hope this book’s approach will shed light on the nature of inequality even for those rejecting its central tenets” (p. 283).