Six Viewpoints for Assessing Egalitarian Distribution Schemes

Is Being Out of Step Intrinsically Bad?

Consider the following case. Imagine that at some time in the future there is no contact between Europe and North America. The long-term happiness levels of the populations of these two continents are roughly on a par, but in a given ten-year period while the people in North America are very happy, those in Europe are only fairly happy; and for the next ten-year segment of time the situation is reversed: people in Europe are very happy while those in North America are only fairly happy. Their happiness levels are out of step, as it were. Dennis McKerlie’s theory would force him to hold that this would be bad, while I find it hard to believe that it would matter. If the happiness levels of the two continents could somehow be brought into step at a cost of some loss of happiness for both, it would surely be perverse to do so.

McKerlie would have to take this stand because he rejects the common view that if equality is intrinsically valuable, it is equality between people’s lives taken as a whole that matters (“complete lives egalitarianism”). He argues instead that equality between fairly long simultaneous segments of lives is also intrinsically valuable (“segmental egalitarianism”) [Although he also discusses a number of other alternatives, I limit myself to the one he defends himself, simultaneous segments egalitarianism. Dennis McKerlie, “Equality and Time,” in this issue].

Equality and Utility

The trouble with seeking segmental equality as an end in itself is that it is likely to get in the way of maximizing happiness. Although it is arguable that complete-lives egalitarianism in regard to goods would probably maximize happiness, it seems likely that segmental egalitarianism would not because it would leave less room for distributing goods in the most beneficial way [Brandt argues persuasively that equality of goods (insofar as it is compatible with adequate incentives) is likely to maximize happiness because of the principle of diminishing returns and because deviations from an equal distribution of goods tend to be less beneficial when they are appropriate than harmful when they are inappropriate (Richard Brandt, "A Theory of the Good and the Right" (Oxford: Clarendon Press, 1979), pp. 311-16). McKerlie could, of course, try to show that segmental equality of goods is more likely than complete lives equality to maximize welfare. That would only show that it is extrinsically valuable, but it would still be an interesting and important thesis]. For example, complete lives egalitarianism would allow a great deal to be spent on young people’s educations if it were beneficial to do so, even though it would mean spending more on them than on people in other age groups, while segmental egalitarianism would find this temporary imbalance objectionable.

Still, the crucial question is whether segmental inequality is intrinsically bad. If it is, a sacrifice of happiness could be justified. I shall argue that although segmental inequality is sometimes bad (e.g., as in cases where it causes envy) it is never bad intrinsically [Since McKerlie is only concerned with the intrinsic value of the various kinds of inequality, not with the additional value that can be ascribed to helping people who are needy or in distress, I will assume that the individuals in question are neither needy nor in distress].

Although I think I have already shown in the two continents case that it is not intrinsically bad, perhaps I have not been entirely fair to McKerlie. Perhaps it is being out of step in conjunction with something else that is intrinsically bad. Maybe what is required is that the populations in question should know about the inequality. Would segmental inequality matter if the case were changed so that, despite a lack of physical contact, the two populations communicated by radio and knew that they were out of step? It is hard to believe that it would.

Or maybe it is avoidable segmental inequality that is bad. Suppose that it were possible for the two populations to rearrange their lives to bring them into step but that this would entail a significant loss in happiness. If, as seems likely, they were not willing to accept this loss, that would not seem to be either irrational or immoral.

Or it might be that segmental inequality is only objectionable between individuals in the same society. Consider a case in which a group of workers must choose between frequent two-week vacations and rare ten- year vacations. Two-week vacations would not cause the inequality between long segments of a life that McKerlie has in mind, while ten-year vacations would. Suppose that for one reason or another long vacations would give the workers somewhat happier lives, and that no one would be distressed by the thought of their vacationing comrades. Surely under those circumstances one would expect the workers to choose long vacations, and their doing so would not seem to be morally wrong.

This case is similar to the case of alternating peasants and nobles (p. 479) on which McKerlie seems to rely to give the strongest intuitive support for segmental egalitarianism. I have not used his example because one’s objection to the arrangement might reflect one’s belief that it is bad for there to be nobles and peasants rather than opposition to segmental inequality. This would be more obvious if the place changing were between slaves and slave owners, but it is hardly less natural in the case of peasants and lords. In any case, there is no danger of it when it is simply a matter of the length of vacations.

But it may be that the problem is envy. Perhaps segmental inequality is only objectionable when it is combined with envy. This would represent a considerable retreat from McKerlie’s original view that segmental inequality is bad in itself independently of any feelings it may arouse. Still it is worth examining. Consider a variant on the preceding case: this time the workers know that sometimes those working will be envious of the long vacationers. Nevertheless, if after taking into account the unpleasantness that this envy will cause both workers and vacationers it is still believed that long vacations will lead to happier lives, it would still seem rational to choose them.

This case involving envy is the only one where segmental inequality is not innocuous. Even so, segmental inequality is bad here because it causes envy, rather than bad in itself since, as we have seen, if it were not for the envy, the very same arrangement would be innocuous [Even though segmental inequality is not intrinsically bad, if it were almost always bad on the whole because of bad consequences, a Rule Utilitarian might argue that a moral code should be simple enough for an average person to learn, so that for the sake of simplicity it should be treated as if it were always bad. However, I doubt that it is tied that firmly to bad consequences].

Mckerlie’s Quasi-Maximin Response

In responding (McKerlie, n. 15) to the possibility that (as in the cases I have considered) prudence might tell us to institute changing-places egalitarianism if for some reason it would be in everyone’s long-term self-interest, McKerlie suggests as one possibility that a segmental egalitarian might say that “inequalities are not objectionable if they are voluntarily chosen by the people who will be worse off given the inequality.” This resembles the maximin principle in that it assumes that those who will come out worst should be given the deciding vote, and that it would be rational for them to vote for whatever would be best for themselves even if it would not be in the general interest (an assumption I will question later). In any case, since changing places will be in their own long-term interest along with that of everyone else, it seems likely that they would favor it.

Perspectives For Evaluating Segmental Egalitarianism

The Viewpoint of Unanimous Participants

Until now, I have considered segmental egalitarianism from the perspective of all of the individuals who would be affected by a scheme; there are, of course, a number of other perspectives from which a distribution scheme can be assessed. In order to do justice to McKerlie’s theory, I will also consider it from what I take to be the main alternative viewpoints.

The Viewpoint of a Future Participant behind the Veil of Ignorance

John Rawls defends a maximin principle for taxation on the grounds that it would be chosen by a self-interested person behind the veil of ignorance, but as Brian Barry has shown, Rawls ascribes too much weight to the small chance of having a lot that would be worse than the lowest position on a maximin scheme [Brian Barry, "On Social Justices' OxfordReview, no. 5 (1967), pp. 29-39. If the motivation for choosing a maximin distribution is that it is needed to avoid the possibility of the worst sort of poverty, this possibility could be avoided by setting some absolute economic minimum as a safety net. If the possibility of extreme poverty is not eliminated, I would expect a variety of different responses rather than the single ultraconservative response anticipated by Rawls. In any case, if, like McKerlie, one is concerned with segmental equality per se so that even decades of unequal bliss would be objectionable, to focus on that issue one should set aside the possibility of the worst sort of poverty, and with that eliminated, it is unlikely that one would sacrifice happiness for segmental equality]. Even so, since an evaluation of a principle of distribution from behind the veil of ignorance would be disinterested, it is worth applying this test to segmental egalitarianism. But if one had to decide whether to be in a world that takes long vacations or a less happy world that takes short ones in terms of intelligent self-interest and from behind the veil of ignorance, it would seem that once again one would choose the world with long vacations.

The point of view in my cases is obviously internal; in Rawls’s case, it might seem at first external, in that it is not the viewpoint of either a participant or a group of participants. But his evaluator is not really external: he must take the point of view of all members of the system because he might turn out to be any one of them. In both cases the outlook is internal in that it is the interests of the participants that is taken into account.

McKerlie and the External Viewpoint

In contrast, it might be argued that an outsider looking in on a group that would unanimously favor long vacations because they were in everyone’s best interest might hold that as long as the differences in happiness were not great, having short vacations would be better. As McKerlie puts it, a segmental egalitarian “might claim that there is still a moral objection to the system even if it is unanimously chosen and is to everyone’s best interest” (McKerlie, n. 15). My initial reaction is that if I were a member of such a group, I would tell McKerlie’s external observer that it was none of his business—that the evaluations that count are those of members of the group, potential members, or those concerned solely with the welfare of members.

But perhaps the external viewpoint should not be dismissed so lightly. It is the viewpoint taken, for example, by conservationists concerned with the preservation of a species of animal. They regard it as extremely important that the rhinoceros, for instance, should be preserved, but I have yet to hear a conservationist wonder whether the rhinoceros is a happy animal—whether his seemingly splenetic life is actually a life worth living—which is precisely the question one would ask if one approached the matter from an internal rather than an external point of view. The viewpoint of a conservationist is also external if he is not particularly concerned with whether individual animals of a given species are killed as long as there is no threat to the species itself. By contrast, Peter Singer’s concern for the miserable lives of the billions of chickens raised on factory farms is internal [Peter Singer, Animal Liberation(New York: Random House, 1975), pp. 100- 101].

A historian who took an extreme version of an externalist outlook would treat the history of mankind in much the same way that a music historian would treat the works of a composer who wrote two or three masterpieces and a great many insignificant works. Just as he might ignore the inferior works, not counting them negatively but treating them as if they did not exist, so a historian writing from an external viewpoint (e.g., Carlyle) concentrates on major figures, treating the pleasure and pain of ordinary men as insignificant.

Among philosophers, Nietzsche is the paradigm representative of the external point of view. This is his outlook when he favors a society like that of ancient Athens where the leisure required for the development of a high level of culture would have been impossible with an egalitarian distribution and was in fact dependent on slavery. Even if one is strongly opposed to slavery, and even if mankind had been destroyed after the golden age of Athens so that its cultural heritage could not be counted as a mitigating factor, even then, saying that the world would have been worse if Athens had been populated by an egalitarian but uncivilized tribe would have considerable intuitive appeal [Nietzsche could have adopted an internal rather than an external viewpoint by arguing that if the slaves had been rational they would have accepted the sacrifice involved in their slavery as a necessary condition for the high culture of the elite. I shall argue that it may sometimes be rational to accept major sacrifices for the sake of others, but Nietzsche seems instead simply to have ignored, or rather decried, the viewpoint of those who were not in the elite].

The most effective way to attack Nietzsche or at least to force him to qualify his approval of Athens is to reject the external viewpoint. The distributive system of ancient Athens would then be assessed from the point of view of all its members, giving equal weight to all their interests, slaves along with the cultured elite.

It seems likely that the internal point of view is also the best outlook from which to defend egalitarianism. If this is so, it is ill-advised to try to defend an egalitarian scheme by adopting instead an external viewpoint, as McKerlie does in suggesting that even if everyone in a group would be better off with changing places egalitarianism, adopting it might still be wrong. Intuitions from an internal point of view may not yield segmental egalitarianism, but they are likely to come far closer to it than intuitions from an external viewpoint. And even if our intuitions from an external point of view did favor a distribution scheme such as McKerlie’s, one would still have the problem of justifying the external perspective to those individuals, such as my group of workers, who would be affected by the scheme. In contrast the relevance of the internal perspective is obvious.

The Viewpoint of an Egoistic Participant

A fourth alternative is to assess a distribution scheme from the point of view of an egoistic participant fully aware of his own position in society. This would not yield a single disinterested judgment. There would instead be a number of conflicting judgments, and the problem of adjudicating between them. However, it may well be that both Rawls’s maximin principle and McKerlie’s quasi-maximin principle can be defended more plausibly from this viewpoint than in terms of either the veil of ignorance or from the external point of view, and I suspect that their intuitive attractiveness is based more on some such considerations as the following than on Rawls’s veil of ignorance argument.

If those at the bottom understand that they will be better off with a maximin distribution than with any alternative and are therefore satisfied from a completely egoistic point of view, it is tempting to assume that those who are better off will also be satisfied because they will, after all, have more than those at the bottom. However, this inference is highly questionable. Even if the upper classes are willing to ascribe more value to the distribution of goods to the lowest class than to themselves, they will not be willing to ascribe infinitely more value to it. And that is what the maximin principle requires: that any increase in goods to the higher classes, no matter how great, cannot justify any decrease in goods to the lowest class, no matter how small.

A maximin distribution might be held to be justifiable nevertheless because it would be irrational for an egoistic member of the lowest class to accept anything less, so that the alternative to a maximin distribution is alienation or even revolution. But it would also be irrational for the members of, say, the lower middle class to be satisfied, in their case even if they were not completely egoistic, with a scheme that attributed no intrinsic value at all to goods for themselves. It would seem then that if the lowest class must be satisfied from a completely egoistic point of view so that nothing less than a maximin distribution will satisfy them, there would be alienation or revolution with a maximin distribution, and alienation or revolution without it.

In actual practice, of course, there are times when the lower middle class would not revolt against a maximin scheme, and times when the lowest class would be delighted to have a utilitarian distribution rather than what might be called a “maximax” distribution. 

There may in fact be societies now where the higher classes would settle for a maximin distribution and the lowest class would need to be persuaded that it was in their own interest to allow just enough inequality to provide adequate incentives. It does not follow that a maximin distribution would be just, but it does follow that in certain contemporary contexts (e.g., Russia and China) it might well be expedient [In the Western democracies, on the other hand, it seems unlikely that the lower middle class would accept anything more egalitarian than a compromise between utilitarianism and the maximin principle].

The Viewpoint of an Ideal Observer as the Participants’ Representative

A distribution scheme could also be evaluated from the outlook of an ideal observer who, unlike McKerlie’s observer, is solely concerned with the welfare of the participants. Since he would not be a participant himself, he would be disinterested, and since benevolence would be part of his ideal nature, he would be concerned with the interests of all the participants so that his viewpoint would be internal. In fact, he could be considered their representative.

In order to know how he would respond to relatively fine distinctions in evaluating distributive schemes such as that between complete-lives egalitarianism and segmental egalitarianism, or between either of these and utilitarianism, it is necessary to have a fairly precise notion of his benevolence. But if ‘benevolent’ is construed in its ordinary sense, its meaning is not clear enough to determine which sort of distribution a benevolent observer would favor. One could, of course, build it into the notion of his benevolence that he would favor a utilitarian distribution, making it an analytic truth, but the attraction of the ideal observer test is that he is supposed to be benevolent in the ordinary sense. Nor would making it analytic help utilitarianism: it cannot be defended by a mere tautology; the same would be true, of course, if it were made analytic that the ideal observer would favor segmental egalitarianism.

The Viewpoint of a Slightly Generous Participant (Is It Always Irrational to Accept Less than an Equal Share?)

McKerlie considers (again in his n. 15) the possibility of an individual who “would prefer to have less for the sake of others” and asks whether we should instead “require her to accept an equal share of resources or happiness.” He seems to believe that it would be extremely unusual for an individual to choose to have less than an equal share for the sake of others and that even if she did it would not be rational for her to do so, so that the possibility has no bearing on the nature of distributive justice.

But people are often willing, for example, to undergo great hardships and even to risk their lives in what they regard as just wars where victory would lead to the general good. And although it is probably irrational for a normal person to make a sacrifice for others if they will not gain more than he will lose, a sacrifice may be perfectly rational if their gain will be many times his loss, even if it means that he will have less than an equal share of happiness or goods. I mean rational in the sense of being something that a person could choose to do with his eyes open, something that he could choose to do without being handicapped by ignorance, factual errors, or illogical thinking, and with a vivid idea of a representative sampling of the expected consequences of his action both for himself and others.

The fact that our innate nature is such that most people if they were rational would be at least a bit generous rather than totally egoistic may seem to have little significance for questions of distributive justice. But it plays a crucial role in showing that it is rational for us to be disinterested in our evaluation of national policies—that it might be rational for a doctor, for instance, to vote for nationalized medicine even though it would not be in his own best interests, because in such cases one’s loss would almost always be minute in comparison with the overall gains to the nation as a whole.

But how can one determine how much sacrifice would be rational if, for example, the suffering of a million people was at stake? Obviously, one cannot have a vivid idea of the suffering of that many people: one would simply go mad. And a mind that could vividly imagine the suffering of a million people would be so different from one’s own mind that it is doubtful that it would serve as a useful analogue. Further, it is not enough to say that one could at least have a vivid idea of a sampling of them. For what would be the difference between one’s vivid idea of a sampling of a thousand people suffering and one’s vivid idea of a sampling of a million people suffering? And suppose that one were moved more by the thought of a million people suffering than by the thought of a million and one people suffering. One would still know that it would be worse for a million and one to suffer than a million [In spite of what Taurek says to the contrary.) Compare Derek Parfit, “Enumerate Ethics,” Philosophy and Public Affairs 7 (1978): 285—301]. (If you have doubts, remember that it is as bad for a million people to suffer as for a million people to suffer, and then imagine that you are number one million one.)

It is possible, however, for me to do something like this. I can begin by forming a vivid idea of the suffering of a single person to get a rough notion of how much pain (or other equivalent loss) I would be willing to undergo to prevent that person’s suffering. I might notice too that although I do not initially tend to care as much about strangers as about people I know, this changes as soon as I have a vivid idea of them and of the kind of lives they lead; and if I begin to wonder why I should make sacrifices for others even though a large proportion of them would not do the same for me, I can remember that I usually feel more kindly even toward thieves and murderers when I acquire a more complete and vivid knowledge of their backgrounds and lives.

So far it has been largely a matter of determining how forming vivid ideas of various states of affairs would affect the degree of my generosity. But then I would do something quite different. Instead of trying to form a vivid idea of the suffering of a million people (or even of a fair sampling), I would take it as self-evident that it would be a million times as bad for a million people to suffer as for one person to suffer.

It would not follow that I would be willing to undergo a million times as much pain to prevent the suffering of a million people as I would to prevent the suffering of one person: the ratio of gain for others in proportion to loss for myself would have to increase as the amount of my loss increased, and for most of us there may well be some upper limit to how much loss we would be willing to accept.

This is not a serious problem when it is a matter of voting on national policies because the sacrifices required are seldom great, and even though most of us would be willing to sacrifice less in proportion to the gain for others as the amount of our sacrifice increased, a very small fraction of an extremely large gain in utility would still be more than enough to justify considerable sacrifice.

I have appealed here to two distinct kinds of rationality: what I call subjective rationality (the kind that Hume would have granted, though he might have underestimated its importance), and what I call objective rationality. If something is objectively rational, everyone should regard it as rational: if you do not, you are simply wrong; while the very same thing can be subjectively rational for one person but not for another.

It would be objectively irrational, for instance, to punish someone for an un-caused action; retributive punishment is objectively irrational whether the universe is deterministic or not, and it is also objectively irrational to say that numbers do not count. On the other hand, it may be subjectively rational for someone to endure one hour of pain in order to prevent each of three strangers from enduring an hour of pain, but not subjectively rational for someone else who is innately less generous. Rational considerations and what Brandt calls cognitive psychotherapy may show you that it would be subjectively rational for you to be more generous than you are, but there are innate limits to benevolence which vary from person to person, so that even if two individuals were influenced by rational considerations as much as they were capable of being, still one of them might be innately more generous than the other.

Although most of us care about the welfare of those at the bottom, we care about others as well, and the maximin principle has been sufficiently discredited so that it cannot be said that this is irrational of us. It is not the case that goods or happiness for others are regarded as having value only insofar as they provide incentives for work that will be beneficial to the lowest class, and this is true whether the lowest class is picked in terms of complete lives or current segments of lives.

I now think that I have shown that even though the level of innate generosity most of us have is such that it is only rational for us to make sacrifices for others if those others will gain a number of times more than we would lose, it is virtually certain that that requirement will be fulfilled in voting for a beneficial national policy. Any detrimental effect that it may have on oneself will be minute in proportion to its effect on the nation as a whole.

This means that when you are voting on a distribution scheme for a large number of people, a disinterested vote is likely to be rational even if you will be adversely affected by it. It may be rational to vote for a distribution scheme that will be beneficial on the whole even though it will leave you less happy or less affluent than others. It follows from this (plus the fact that our disinterested concern is not limited to the lowest class) that if McKerlie’s hypothetical person voted against a segmental egalitarian policy even though she would benefit from it, choosing instead a policy that would come closer to maximizing the general good, it might well be rational for her to do this and irrational for us to try to stop her.

In summary, I have attacked McKerlie’s form of egalitarianism by giving a series of cases designed to show that being out of step is not undesirable per se, that the combination of segmental inequality with various other factors is not undesirable per se, and that when segmental inequality arouses envy, that should be taken as showing that changing places egalitarianism may have bad consequences, not that it is bad in itself.

I have also tried to take the discussion to a deeper level by considering the main points of view from which one could assess distribution systems. I have arrived at the following conclusions:
  1. If a group of individuals believes that on the whole they will have happier lives with complete lives egalitarianism than with segmental egalitarianism, it would be neither irrational nor wrong for them to choose the alternative that would give them happier lives.
  2. Although an ideal observer would be disinterested and benevolent in his judgments, if benevolence is taken in its ordinary sense, it is not clear what distributive system his benevolence would lead him to favor.
  3. It seems unlikely that judgments made from the external viewpoint would favor segmental egalitarianism. Even if they did, it is hard to see why participants should accept a distributive scheme chosen from an external viewpoint.
  4. A scheme that would satisfy a completely egoistic lowest class would not satisfy the rest of a society if they were egoistic too. In fact, it would not satisfy them unless they ascribed no intrinsic value to the welfare of anyone but members of the lowest class.
  5. A distributive system selected from behind the veil of ignorance would be disinterested and it is arguable that it would be fair, but to be a good citizen one needs an element of generosity as well as knowledge of what is fair. If you were a complete egoist and knew your position in society, it would only be rational for you to favor a fair distribution when it was in your interest to do so. If you were rich, you would vote for a maximax policy; if you were poor, for a maximin. Similarly, if you were a complete egoist you would not be motivated to accept a distribution scheme that was not in your interest, by learning that it would be favored by a benevolent observer.
  6. However, with the exception of psychopaths and near psychopaths, whenever very large numbers of people would be affected by a distribution scheme, all of us have more than enough innate generosity to justify a disinterested vote—a vote that would take full account of the interests of everyone involved. Obviously, even the limited degree of generosity required for such a vote is not always manifest, but it would be if we were subjectively and objectively rational, and what is relevant to distributive justice is the way we would vote if we were rational.
Perhaps the first of these tests, the one requiring unanimity, is best for making it clear that when segmental inequality is bad it is only bad extrinsically, as when it causes envy; and this final test may provide the best basis for arguing that even though some individuals in any large society would be bound to favor segmental egalitarianism, whether one is rich or poor it is rational to vote for the distribution scheme that would be best for one’s society, even if for some reason segmental egalitarianism would be best for oneself.

Six Viewpoints for Assessing Egalitarian Distribution Schemes
Author(s): R. I. Sikora
Source: Ethics, Vol. 99, No. 3 (Apr., 1989), pp. 492-502
Published by: The University of Chicago Press
Stable URL: