II. PROBABILISTIC PRINCIPLES

Suppose that, in Case One, we could either

or

Utilitarians would believe that we ought to do Y, since all that matters is the size of these possible benefits. Egalitarians would disagree, believing instead that we ought to do X, since it also matters whether some people are worse off than others. On this view, we have one reason to do Y, given by the greater size of the benefit to Ted, but we have a conflicting reason to do X, given by the fact that Tom is worse off. By benefiting Tom rather than Ted, we would reduce rather than increase the inequality between these people. This egalitarian reason to do X would be stronger than our reason to do Y, since the benefit to Tom would be only slightly smaller than the benefit to Ted, and Tom is much worse off than Ted.

Prioritarians would also believe that we ought to do X, but they would defend this belief in a different way. On the Priority View, the strength of our reasons to benefit people depends both on the size of these possible benefits and on how well off these people are. Since the benefit to Tom would be only slightly smaller than the benefit to Ted, and Tom is much worse off than Ted, we would have a stronger reason to give Tom this benefit. When we have such a stronger reason to benefit one of two people, we can say that, even if this benefit would be in itself smaller, it would have greater moral weight, or be a greater weighted benefit. According to a fuller statement of

I have said that, on this view, we have a stronger reason to benefit Tom because Tom is much worse off than Ted. Such claims can be easily misunderstood. We might similarly claim that, when some mountaineers are higher up than others, they find it harder to breathe. But they do not find it harder to breathe because they are higher up than others. They would find it just as hard to breathe even if there were no other people who were lower down. These mountaineers find it harder to breathe the higher up they are, or when they are at a higher absolute level. In the same way, Prioritarians believe, we have a stronger reason to benefit people the worse off these people are. We have a stronger reason to benefit Tom, and this act would do more to make the outcome better, not because Tom is worse off than Ted, but simply because Tom is at a lower absolute level. Our reason to benefit Tom would be just as strong even if Ted was not better off.

Of the writers who have criticized the Priority View, some suggest that this view may not make sense. If we believe that, by benefiting Tom, we would do more to make the outcome better, this may show that, on our view, the benefit to Tom would be greater than the benefit to Ted. To defend the Priority View, we must explain what it would mean to claim that one of two benefits would be smaller, though this benefit would contribute more to the value of the outcome. These sceptics doubt that we can distinguish between what is better for people and what makes the outcome better.

If these sceptics were right, this objection would also undermine Egalitarian views, since Egalitarians also believe that, if we give smaller benefits to people who are worse off, we may thereby make the outcome better. If there was no distinction between the size of a benefit and the amount by which this benefit makes the outcome better, Egalitarians and Prioritarians could not intelligibly reject Utilitarian beliefs about the goodness of outcomes. These people could only disagree about which benefits would be greater.

This objection raises difficult questions about the concepts good and good for, and about what it would be for some benefits to be greater than others, and for some outcomes to be better. Though I believe that this objection can be answered, I shall not try to show that here. I shall simply assume that we can distinguish between the size of different possible benefits and the amount by which these benefits would make the outcome better, so that we can intelligibly reject Utilitarian beliefs.

Utilitarianism beliefs are in themselves plausible. Though he is not a Utilitarian, Nagel writes:

Nor does this value depend, Utilitarians could add, on how well off different people are. Suffering and happiness are always in themselves good or bad, and it makes no moral difference when, where, or to whom they come. Though this view is plausible, it is not, I believe, true. In cases like One, for example, we would have stronger reasons to give slightly smaller benefits to people who are much worse off. But my main aim here is not to argue against Utilitarianism, but to compare Prioritarian and Egalitarian views.

It may be worth explaining why, in Case One and my other examples, I suppose that we could give slightly smaller benefits to people who are much worse off. I do not mean to imply that, if there was either a greater difference in the size of different possible benefits, or less difference in how well off different people are, Prioritarians and Egalitarians would accept Utilitarian conclusions. I am merely discussing cases in which the disagreements between these views take their clearest forms.

On the Priority View, one of two acts would make the outcome in one way better if this act would give people a greater sum of weighted benefits. On what we can call

This principle could not conflict with the Priority View. If one of two acts would be better for some people and not worse for anyone else, it is only this act that would give anyone any benefits, so this act must give people a greater sum of weighted benefits.

Many cases are unlike Case One, because we cannot predict what effects our acts would have. In such cases, we should often try to act in the way whose outcome would be expectably best. To estimate such expectable goodness, it is often enough to add together the goodness of an act's possible outcomes, multiplied by the chances that this act would produce these outcomes. In some cases, for example, the expectable goodness of an act's outcome roughly corresponds to the expectable number of lives that this act would save. Suppose we know that

and that

On these assumptions, X is the act whose outcome would be expectably better, since X would expectably save 4 lives, and Y would expectably save only 3 lives. We can similarly claim that one of two acts would be expectably better for people if this act would give people a greater sum of expectable benefits, which are the possible benefits to people multiplied by the chances that this act would give people these benefits. In this example, since X would expectably save more lives, X would be expectably better for people.

When restated to cover such cases, our principles could become:

As Wlodek Rabinowicz first pointed out, these principles can conflict.Footnote ⁴ Suppose that, in

or

Utilitarians would again believe that we ought to do Y, since all that matters is the size of these possible benefits. Prioritarians would again believe that we ought to do X, because we would have more reason to do what would give the slightly smaller benefit to the person who is much worse off. Egalitarians would also believe that we ought to do X, because this possible benefit would be only slightly smaller, and this act would reduce rather than increase the inequality between Tom and Ted.

In cases like Two, Utilitarians could defend their view by appealing to the Probabilistic Principle of Personal Good. If we do Y, we would give both Tom and Ted greater expectable benefits, so this act would be expectably better for both these people. If instead we do X, as Prioritarians and Egalitarians claim that we ought to do, our act would be expectably worse for both Tom and Ted, and expectably better for no one. Utilitarians might claim that

1. (A) it cannot be true that we ought to do what would be expectably worse for everyone whose well-being our acts could affect.

This claim may seem very plausible. (A) may seem like

1. (B) It cannot be true that we ought to do what would actually be worse for everyone whose well-being our acts could affect.

Though (B) can be challenged in several ways, this claim is very plausible. Despite their resemblance, however, (A) and (B) are very different claims. In Case Two we know that, though Y would be expectably better for both Tom and Ted, Y would actually be worse for one of these people. As we also know, Y would be worse for the person who is very badly off, whom we would have failed to benefit.

Case Two, we can next point out, is not morally different from Case One. In both cases, we have a stronger reason to give a slightly smaller benefit to someone who is much worse off, and if other things are equal that is what we ought to do. It is irrelevant that, in Case Two, we don't know who is the person who is worse off.

It may seem surprising that, in some cases, it can be true that we ought to do what would be expectably worse for everyone whose well-being our acts could affect. But if that seems surprising, that is because we overlook the fact that, when some act would be expectably worse for everyone in some group, this act may actually be, in predictable ways, better for some of these people, who may also be the people who are worse off. When we consider cases like Two, Prioritarians and Egalitarians can defensibly reject the Utilitarian argument given above, by rejecting claim (A) and the Probabilistic Principle of Personal Good.Footnote ⁵

Suppose next that, in

or

Prioritarians would again believe that we ought to do X. On their view, though X would give Tom and Ted slightly smaller expectable benefits, this fact would be morally outweighed by the fact that Tom and Ted would receive these benefits if they are much worse off.

It may seem that Egalitarians would here take a different view. Whether we do X or Y, Tom and Ted would be equally well off, so there is no inequality between these people that our acts might either reduce or increase. This fact may suggest that we have no egalitarian reason to do X, so that Egalitarians would here believe that we ought to do Y, because this act would be expectably better for both Tom and Ted.

For Egalitarians, however, there is no significant moral difference between cases like Two and Three. Egalitarians would agree that, when we must choose between acts that would affect the well-being of certain people, we can often ignore the question whether there are other people who are worse off or better off. That may be true, for example, when we must choose between acts that would benefit our close relatives or friends, or other people to whom we stand in various special relations. But we are here discussing acts that would benefit strangers. Egalitarians believe it to be in itself bad whenever some people, through no fault or choice of theirs, are worse off than other people. If Tom and Ted are very badly off, they may be worse off than most other people. If that is true, it would make no moral difference that, in cases like One, Two and Three, our possible acts would affect the well-being of only two of these many people. When some people are much worse off than others, and we can reduce this inequality by benefiting these people, our egalitarian reason to act in this way does not disappear when, and because, these are the only people whose well-being our acts could affect.

It may help to redescribe these cases, using figures to suggest the differences between various levels of well-being. We can say that, if Tom or Ted are very badly off, they would be at level 10, and we could benefit them by raising them to 20. If they are very well off, they would be at level 80, and we could give them a slightly greater benefit by raising them to 91. Most other people are at intermediate levels of well-being. We can suppose that these people are at level 50.Footnote ⁶ In both cases Two and Three there are two possible states of nature, or SNs, which are equally likely to be what is actually true. In Case Two the possible outcomes of our acts would be these:

In Case Three, the possible outcomes would be these:

It is true that, in Case Three, there would be no inequality between Tom and Ted. But in both cases, if we do X rather than Y, we would reduce rather than increase the inequality between Tom and Ted and most other people. This fact, most Egalitarians would believe, would morally outweigh the fact that X would give Tom and Ted slightly smaller expectable benefits.

Suppose next that, in

or

This case is just like Three, except that our acts would affect the well-being of only one person. For Prioritarians and Egalitarians, this fact would not make Four morally different from Three, so these people would again believe that we ought to do X.

I have now claimed that, in cases like Two, Three, and Four, Prioritarians and Egalitarians would have similar beliefs. These people would agree that we can have stronger reasons to do, and in some cases ought to do, what would be expectably worse for people, by giving these people lesser expectable benefits. The disagreement would be only about which are the facts that would give us these reasons. Prioritarians believe that we can have stronger reasons to give people lesser expectable benefits when these people would receive these benefits if they are worse off, in the sense of being at a lower absolute level. Egalitarians believe that we can have stronger reasons to give people the lesser of two expectable benefits when these people would receive these benefits if they are worse off than other people, since such acts would reduce inequality. Utilitarians would reject these claims, since they believe that, in all these cases, we ought to do what would be expectably better for people, by giving them greater expectable benefits.

We can next note how this disagreement is related to another, wider disagreement. According to what we can call

On one version of this view, which we can call

In some of the cases that we have been considering, our acts could benefit people who are equally likely to be either very well off or very badly off. According to the Equal Chance Formula,

1. (C) we ought to treat these people as it would be rational for them to choose to be treated, if these people made this choice for self-interested reasons and they knew that they had equal chances of being either very well off or very badly off.

When Rawls and Harsanyi appeal to their versions of Veil of Ignorance Contractualism, they claim that the Equal Chance Formula supports the Utilitarian Average Principle, which requires us to act in ways that would maximize average utility, by producing the greatest sum of expectable benefits per person. This is the principle whose choice would be rational, in self-interested terms, for people who have equal chances of being in anyone's position.Footnote ⁸

We can plausibly reject this argument, because we can reject this version of contractualism. As Rawls points out, Utilitarianism is, roughly, self-interested rationality plus impartiality. If we appeal to the choices that would be rational, in self-interested terms, if we were behind some veil of ignorance that made us impartial, we would expect to reach conclusions that are, or are close to being, Utilitarian. But this argument cannot do much to support Utilitarianism, because this argument's premises are too close to these conclusions. Suppose that I act in a way that imposes some great burden on you, because this act would give small benefits to many other people who are much better off than you. If you object to my act, I might appeal to the Equal Chance Formula. I might claim that, if you had equal chances of being in anyone's position, you could have rationally chosen that everyone follows the Utilitarian Principle, because this choice would have maximized your expectable benefits. As Scanlon and others argue, this would not be a good enough reply.Footnote ⁹ You could object that, when we ask whether some act would be wrong, we are not asking a question about rational self-interested choice behind a veil of ignorance. Acts can be wrong in other ways, and for other reasons. If we are contractualists, we would do better to ask which are the moral principles that no one could reasonably reject, or which are the principles that it would be rational for everyone to choose, with full knowledge of their own and other people's situations, and without a restriction to self-interested reasons. We could also plausibly reject contractualism, and appeal directly to various beliefs about which acts are wrong and what makes them wrong.

Let us return now to the disagreement between Utilitarians, Prioritarians and Egalitarians. In Case One, we could either

or

Utilitarians might argue that we ought to do Y, since that is how it would be rational in self-interested terms for Tom and Ted to choose to be treated, if they were behind a veil of ignorance, and had equal chances of being in anyone's position. But, as I have said, Prioritarians and Egalitarians can justifiably reject this contractualist argument. They can claim that we ought to do X, giving the slightly smaller benefit to the person who is much worse off. If we are not Utilitarians, most of us would agree.

Things are in one way different in cases Two to Four. When they consider these cases, Utilitarians do not need to appeal to this version of contractualism, by asking us to suppose that Tom and Ted are behind some veil of ignorance. In these cases, Tom and Ted do in fact have equal chances of being either very badly off or very well off. To defend their belief that, in cases like Two to Four, we ought to do Y, Utilitarians could appeal to the Probabilistic Principle of Personal Good, claiming that an act would make the outcome expectably better if this act would be expectably better for some people and expectably worse for no one. And they could appeal to the fact that, if we do X, this act would be expectably worse for everyone whose well-being our acts could affect. Utilitarians could claim that

1. (A) It cannot be true that we ought to do what would be expectably worse for all these people.

As I have also said, however, we can justifiably reject (A). We can claim that

1. (D) Even if one of two possible acts would be expectably worse for people, this act may actually be better for these people. We may also know that this act would be better for these people if they are worse off. This fact may be enough to make this act what we ought to do.

Since we can justifiably reject (A) and appeal to (D), we can justifiably reject the Probabilistic Principle of Personal Good.

We can now add that, like the Equal Chance Version of Veil of Ignorance Contractualism, this Probabilistic Principle has a built-in bias towards Utilitarian conclusions, and can therefore be rejected in similar ways. According to Prioritarians, we have reasons to benefit people which are stronger the worse off these people are. According to Egalitarians, we have reasons to reduce rather than increase inequality between people. The Probabilistic Principle assumes that we have no such reasons. If we appeal to what would be expectably better for people, that is like appealing to the choices that it would be rational for people to make, for self-interested reasons, if they had equal chances of being in anyone's position. Since this principle appeals only to self-interested or prudential reasons, it ignores the possibility that we may have impartial reasons, such as reasons to reduce inequality, or reasons to benefit people which are stronger the worse off these people are. We can object that we do have such reasons.

When Rabinowicz pointed out that, in cases like Four, Prioritarians must reject the Probabilistic Principle of Personal Good, he did not regard this fact as counting against the Priority View. That, I believe, was the right response. Rabinowicz could have added that similar claims apply to Egalitarians, and to cases like Two and Three.

In some later papers, Michael Otsuka and Alex Voorhoeve defend partly different conclusions.

III. COMPETING EXPLANATIONS

In Case One, we could either

or

Utilitarians would believe that we ought to do Y. When Otsuka and Voorhoeve consider cases like One, they reject this view. In such cases, they write, ‘We ought . . . to give priority to the worse off even when this leads to lower total utility’.Footnote ¹⁰

In Case Four, we could either

or

Prioritarians would believe that

1. (E) We have a stronger reason to do X, though this act would give Tom the smaller of these possible benefits, because Tom would receive this benefit if he is very badly off. If other things are equal, X is what we ought to do.

When Otsuka and Voorhoeve consider cases like Four, they reject this view. With one qualification, to which I shall return, they claim that Prioritarians cannot reasonably believe (E).

There is a great moral difference, Otsuka and Voorhoeve claim, between cases like One and Four. Utilitarians and Prioritarians are both mistaken because they fail to recognize this difference, though in opposite ways. Utilitarians believe that, in both kinds of case, we ought to give the slightly greater benefit or expectable benefit to someone who is or would be very well off. Prioritarians believe that, in both kinds of case, we ought to give the slightly smaller benefit or expectable benefit to someone who is or would be very badly off.

Both views, Otsuka and Voorhoeve suggest, fail to recognize the moral importance of personal identity. When Utilitarians claim that, in Case One, we ought to give Ted the greater benefit, they ignore the fact that this act would be worse for a different person, Tom, who is much worse off than Ted. Utilitarians treat such cases as if they involved only a single person. In Rawls's phrase, Utilitarians do not take seriously the distinction between persons, and the fact that one person's burdens cannot be compensated by benefits to someone else. Prioritarians do, in one way, recognize the distinction between persons, since they give priority to benefiting people the worse off these people are. But Prioritarians fail to recognize the importance of the unity of each person, and the fact that one person's burdens can be compensated by benefits to this person.Footnote ¹¹ When Prioritarians claim that, in Case Four, we ought to do X, because this act would give Tom the lesser benefit if he is badly off, they treat this case as if it involved two different people. Prioritarians ignore the fact that, if we do Y, this act would be expectably better for Tom himself.

In this issue of Utilitas, Otsuka adds the following remarks. Suppose that we do Y in Four, and have bad luck, because Tom turns out to be very badly off, and if we had done X instead, we would have benefited Tom. Despite this fact, Otsuka claims, Tom could not object to our doing Y. We could give Tom a prudential justification of this act, because Y gave Tom an equal chance of a greater benefit. No such claim applies to Case One. If we do Y in One, giving Ted the slightly greater benefit, Tom could reasonably object, and we could not justify our act to Tom by giving him any such prudential justification. Y had no chance, in One, of benefiting Tom. Prioritarians, Otsuka claims, overlook this difference between cases like One and Four. When they consider cases like Four, Prioritarians are insensitive to the possibility of such prudential justifications, and thereby fail to take seriously the unity of a person, and the fact that different possible lives may both be lives of the same person.

This striking and original objection to the Priority View can, I believe, be answered. Remember that, in

or

As in Case Four, Prioritarians and Egalitarians would believe that we ought to do X, and Utilitarians would believe that we ought to do Y. Suppose that we do Y, and Tom turns out to be the person who is very badly off. Tom might object that we ought to have done X, giving the smaller benefit to the person who is very badly off. Utilitarians might reply that we could give Tom a prudential justification of our doing Y. We could point out that Y was expectably better for Tom, by giving Tom a greater expectable benefit. Utilitarians might add that, in both cases Two and Four, Prioritarians and Egalitarians are insensitive to the possibility of prudential justifications, thereby failing to recognize the importance of the unity of a person.

Otsuka and Voorhoeve would reject these claims. When Otsuka discusses a case like Two, he claims that we ought to do X, giving the lesser expectable benefit to the person who is badly off.Footnote ¹² On Otsuka's and Voorhoeve's view, with the qualification to which I shall return, we ought to do X in Two but not in Four.

This view, I believe, is hard to defend. We can first note that, from Tom's point of view, there is no difference between cases Two and Four. In both cases, we could either

or

These cases are in one way different, from Ted's point of view, since it is only in Case Two that our acts could affect Ted. But these effects on Ted would be just the same as the effects on Tom. If we do X, this act would benefit Ted if he is very badly off, and if we do Y, this act would give Ted a slightly greater benefit if, as is equally likely, he is very well off. Since these two acts would have these same effects in both cases Two and Four, it is hard to see how it could be true that we ought to do X only in Case Two.

When Otsuka discusses a similar objection, his remarks suggest the following reply.Footnote ¹³ In Case Two, we know that either Tom or Ted is very badly off. We therefore know that

1. (F) If we do X rather than Y, this act would be certain to benefit someone who is very badly off rather than giving a slightly greater benefit to someone who is very well off.

In Case Four, since there is only a one in two chance that Tom is very badly off, there is only a one in two chance that our doing X would have the effect described by (F). This difference, Otsuka might say, explains why we ought to do X in Two but not in Four.

This reply, I believe, would fail. In cases of the kind that we are considering, if we ought to act in one of two ways because the effects of this act would certainly be better, we ought also to act in this way if this act's effects would be likely to be better. For example, if we ought to do A rather than B because A would certainly save 10 lives and B would save only 5 lives, we ought also to do A rather than B if A would have a one in two chance of saving 10 lives, and B would have a one in two chance of saving 5 lives. Since Otsuka believes that, in Case Two, we ought to do what would certainly give the lesser benefit to someone who is very badly off, he must give some other reason why it is not true that, in Case Four, we ought to do what is likely to have this same effect.

Otsuka also suggests that, in Two but not in Four, there would be a competing claims complaint.Footnote ¹⁴ Suppose that we do Y in Two, and Ted turns out to be the person who is very well off, so that our act gives Ted the greater benefit. Tom might here complain that we ought to have done X, which would have given him the lesser benefit, because he is the person who is very badly off. In Case Four, since Tom is the only person whose well-being our acts could affect, there are not two people who might have such competing claims. This might be why we ought to do X in Two but not in Four.

This reply, I believe, would also fail. Suppose that we do Y in Four, and Tom turns out to be very badly off, so that our act does not benefit him. Tom might here complain that we ought to have done X, which would have given him the lesser benefit because he is very badly off. Though there are not two people in Four who have competing claims, Tom's complaint in Four about our doing Y could be just the same as Tom's complaint in Two. Otsuka might reply that we could answer Tom's complaint, in Four, by pointing out that Y gave Tom a greater expectable benefit. But we could similarly claim, in Two, that Y gave Tom a greater expectable benefit. Otsuka would agree that, in Two, this claim would not answer Tom's complaint, which would be justified. If Tom's complaint would be justified in Two, why would the same complaint not be justified in Four?

When Otsuka and Voorhoeve state their objection to the Priority View, they also make some remarks about inequality. For Prioritarians, they write, the moral importance of benefiting someone ‘is unaffected by the presence or absence of other people who are worse off’.Footnote ¹⁵ Prioritarians do not distinguish between cases like Two and Four, because they ignore the fact that, in cases like Four, there is no inequality, since ‘there is only one person, who is not worse off or better off than anyone else’.Footnote ¹⁶ As I have said, however, there is no such difference between these cases. Egalitarians would believe that, in Case Four, we would have an egalitarian reason to give Tom the smaller benefit if he is very badly off, since this act would reduce the inequality between Tom and most other people.Footnote ¹⁷

We can now turn to the way in which Otsuka and Voorhoeve qualify their claims. As we have seen, Otsuka and Voorhoeve make remarks which may seem to imply that there are important moral differences between cases like Two and Four. These remarks suggest that Prioritarians are wrong to believe that, in both kinds of case, we ought to reach similar moral conclusions. Otsuka and Voorhoeve do not, however, intend to claim that there are such differences between these cases. They write, for example that, Egalitarians are

This remark may suggest that, in these one-person cases, inequality cannot be present. But that is not what Otsuka and Voorhoeve mean, which is why this sentence ends considered in isolation from others. Otsuka and Voorhoeve mean that inequality cannot be present in one-person cases when we consider these cases in a way that brackets, or sets aside, facts about other people's well-being. They agree that, in one-person cases such as Four, inequality can be present, since the one person whom such a case involves may be either worse off or better off than other people. And such inequality, they agree, may be morally relevant. In the passage just quoted, Otsuka and Voorhoeve mean that inequality cannot be present in our moral thinking about one-person cases when we consider these cases in a way that brackets, or sets aside, thoughts about inequality.

Similar remarks apply to other passages. When they consider a case like Four, Otsuka and Voorhoeve write

They also write that, in such cases, Prioritarians cannot reasonably believe that we ought instead to do what would give this person a slightly smaller benefit if he is very badly off. These remarks may suggest that, on Otsuka's and Voorhoeve's view, we could justifiably do Y in Four, and we cannot reasonably believe that we ought instead to do X. But that is not what Otsuka and Voorhoeve mean. These remarks apply, they explain, only when we consider cases like Four in a way that brackets facts about inequality. They would agree that, when we take into account the inequality between Tom and other people, we might justifiably conclude that we ought to do X. As Otsuka writes

Otsuka and Voorhoeve are not objecting to Prioritarian beliefs about what we ought to do. They are objecting only to Prioritarian explanations of why we ought to do these things. What is unreasonable is the Prioritarian belief that

1. (G) We ought to do X simply because this act would give Tom the slightly smaller benefit if he is very badly off.

Otsuka and Vooerhoeve believe instead that

1. (H) we ought do X because this act would give Tom the slightly smaller benefit if he is very badly off, and would thereby reduce the inequality between Tom and other people.

When two theories give competing explanations of the same facts, it can help to consider artificially simple cases. That is why Otsuka and Voorhoeve suggest that, to assess the Priority View, we should consider ‘the fate of one person in isolation from the fate of others’. As Voorhoeve writes:

These remarks may suggest that, to find out what the Priority View implies, it helps to ask how this view would tell us to treat one person if we ignore facts about other people's well-being. But there is no need to ask this question. As Otsuka and Voorhoeve point out, the Priority View explicitly claims that the strength of our reason to benefit someone depends only on how badly off this person is, or might become, and not on whether there are other people who are either better off or worse off. To use Voorhoeve's analogy, the Priority View is like a version of Aristotle's theory which claims that heavier objects would fall faster even if there was no air-resistance, so that all objects were subject only to the force of gravity. There is no need to ask what this theory would imply if there was no air-resistance, so that all objects were subject only to the force of gravity.

As that remark suggests, Otsuka's and Voorhoeve's objection to the Priority View could be stated in a different way. They could claim that, when we consider cases in the real world, the Priority View seems more plausible than it really is, because Prioritarians and Egalitarians would reach similar conclusions. Prioritarians claim that we have more reason to benefit people the worse off these people are. That seems plausible, Otsuka and Voorhoeve might say, only because by benefiting people who are worse off than others we reduce the inequality between different people. To assess the Priority View, we ought to consider imagined cases in which Prioritarians and Egalitarians would reach different conclusions. That would be like dropping both a hammer and a feather near the surface of the airless Moon, to find out which of these objects would fall faster.

IV. THE ISOLATION TEST

In applying this isolation test, we might consider an imagined Universe in which only a single person ever exists, and ask how this person ought to treat himself. Otsuka and Voorhoeve suggest that we can instead suppose that, though different people exist, these people's fates are in some way tied together, so that it would be impossible for anyone to be either better off or worse off than other people.Footnote ²² In such a Universe, there could not be any inequality between different people. These imagined cases might help us to choose between Egalitarian and Prioritarian views. If the Priority View has implausible implications when applied to such cases, this fact might, as Voorhoeve suggests, falsify this view, or at least provide a strong objection.

Suppose that, in

or

Prioritarians would believe that we have a stronger reason to act in the first way, though this act would give everyone slightly smaller expectable benefits. On this view, we would have a stronger reason to reduce suffering than to increase happiness, even if our act would not reduce inequality. That, I believe, is true.

In response to this example, Voorhoeve has said:

Otsuka has made similar remarks. As these remarks show, we should draw some more distinctions. According to

According to

Some Prioritarians combine these views. We have stronger reasons to benefit people, these Double Prioritarians believe, the worse off these people are both in their lives as a whole, and when they receive these benefits.Footnote ²⁴

Suppose that, in

or

This case could take different forms. Suppose first that, in

According to all Prioritarians, these facts would give us a stronger reason to do X. That, I believe, is true.

Suppose next that, in

Different Prioritarians may here disagree. According to pure Whole Life Prioritarians, we ought to do Y, giving the slightly greater benefit to the person who is on the whole worse off. According to pure Temporal Prioritarians, we ought to do X, giving the slightly smaller benefit to the person who is worse off at the time. According to Double Prioritarians, we have two conflicting prioritarian reasons. On this view, if Jane is much worse off than Ruth in her life as a whole, we would have a stronger reason to do Y. If the whole-life difference is less, and Ruth is now, while she is suffering, much worse off than Jane is now, we would have a stronger reason to do X, giving Ruth the smaller benefit.

In the remarks quoted above, Voorhoeve may have in mind a case like Six B. Voorhoeve's comment would then be that, even if it were true that we would have a stronger reason to do X, this truth would not be implied by, and could not support, the Priority View. That is not so. We may believe that we have a stronger reason to give Ruth the smaller benefit because Ruth is worse off now, while she is suffering. Though we are not pure Whole Life Prioritarians, we may be Temporal or Double Prioritarians. And if our belief is true, that would support such views.

Suppose next that, in

This might be true, for example, if Ruth is now in great pain, caused by the freezing wind on the summit of the mountain that she has just climbed. Though Ruth is in great pain, this pain's badness is outweighed by her sense of achievement, and by her seeing the sublime view. Despite the fact that Ruth is suffering and Jane is not, Ruth is the person who is now better off.

Suppose we believe that, even in this case, we ought to do X, reducing Ruth's suffering rather than giving Jane her greater benefit. Voorhoeve would be right to claim that, if we have this belief, that would not show that we are Prioritarians. More exactly, we would be Prioritarians of a different kind, who believe that we have stronger reasons, not to benefit people who are worse off, but to reduce suffering rather than promoting happiness.

Return now to the imagined Universe in which there could not be any inequality. In Case Five, it would be equally likely either that everyone will suffer greatly or that everyone will have no suffering and much happiness. We could either

or

Prioritarians would here believe that we ought to do X. Egalitarians might disagree. If we do X rather than Y, we would not be reducing inequality. Egalitarians might claim that, since we have no egalitarian reason to do X, we ought to do Y. This example does not, I have claimed, provide a strong objection to the Priority View. We can plausibly believe that we ought to give everyone the slightly smaller of two expectable benefits, by doing what would reduce their suffering, when and because these people would receive these benefits if they are very badly off.

Otsuka and Voorhoeve might say that, even if we ought to do X, that would not support the Priority View. As before, they might claim, Prioritarians reach the right conclusion for the wrong reason. If we ought to do X rather than Y, that might be because we have stronger reasons to reduce people's suffering than to add to their happiness, however well off these people are. But when we consider Case Six C, we can plausibly reject this view. We can therefore claim that, in Case Five, we ought to give everyone the slightly smaller benefits because these people would receive these benefits if they are badly off.

Consider next two hellish possible worlds. Suppose that, in

Though the total sum of suffering would be slightly greater in Hell Two, it is clear, I believe, that Hell One would be worse. Prioritarians could explain this fact. Since the fifteen people would each be much worse off than any of the billion people, by suffering more than fifty million times longer, these people would together endure a much greater sum of weighted suffering. In neither of these worlds, however, would there be any inequality between different people, so Egalitarians could not explain why Hell One would be worse.

These hellish worlds provide the simplest version of Otsuka and Voorhoeve's isolation test. Since it is only Prioritarians who can explain why Hell One would be worse, this test does not count against but supports the Priority View.

V. MORALITY AND PRUDENCE

In a different version of Case Four, which we can call

or

Suppose first that Dick is unconscious, and that we do not know how Dick would prefer us to act, so that we must make this decision on Dick's behalf. If we are Prioritarians, we would believe that we have a stronger reason to do X, and that, if other things are equal, X is what we ought to do. As I have said, it may seem implausible that we ought to do X, since this act would be expectably worse for Dick, by giving Dick a smaller expectable benefit. But if this claim seems implausible, that may be because we are being misled by the word ‘expectably’. Though X would be expectably worse for Dick, there is a one in two chance that X would actually be better for Dick, and this act would be better for Dick if he is much worse off.

It might be objected that it could not be true that we ought to do X, since Dick could rationally prefer that we do Y, which would give him the slightly greater benefit if he is much better off. But even if Dick could rationally prefer that we do Y, that does not show that we could justifiably act in this way. When we have to make some decision on someone else's behalf, and we don't know how this person would prefer us to act, we may believe that we ought to be cautious, or risk averse. On this plausible and widely held view, it would be wrong to take, on this person's behalf, some of the risks that this person could rationally choose to take, if he or she made this choice for purely self-interested reasons. As Prioritarians claim, we ought to give more weight to avoiding possible outcomes in which this person would be worse off.Footnote ²⁵

Suppose next that, in a different version of Case Seven, Dick is conscious, so that we can tell him the relevant facts about treatments X and Y, and ask him what he would prefer us to do. Rabinowicz writes that, in such cases, there is ‘a striking divergence between prudence and prioritarian morality’.Footnote ²⁶ Prioritarians would believe that, if we do X, we would make the outcome expectably better, and that, if other things are equal, we ought to do X. These beliefs may seem to imply that, if other things are equal, Dick ought to prefer that we do X, and that it would be wrong for Dick to choose that we do Y. This conclusion may seem clearly false. Since Y would be expectably better for Dick, and would not be worse for anyone else, it may seem absurd to suggest that Dick is morally required to choose that we do X, and that choosing Y would be wrong.

If Prioritarians accept a pluralist moral view, of the kind that Sidgwick calls common sense morality, their view has no such implausible implication. On such a view, though X would make the outcome in one way better, other things are not equal. First, we can plausibly believe that, in cases like Seven, we ought to treat people in the ways in which they do, or would, prefer to be treated. As we might say, we ought to respect these people's autonomy. Even if we believe that X would make the outcome better, we can believe that, if Dick prefers that we do Y, that is what we ought to do. That is why, in some of my earlier examples, I have supposed that we have to make decisions on other people's behalf, and without knowing what these people would prefer.

We can next claim that, even if X would make the outcome better, it would not be wrong for Dick to choose that we do Y. When we are choosing between acts that would affect our own well-being, we are not always morally required to choose the acts that would make things go better. We may not be morally required to choose certain acts, even if these acts would both make things go better, and be better for us. We may not be required, for example, to refrain from various acts that would damage our health, thereby shortening our lives.

It might next be claimed that, since Y would be expectably better for Dick, by giving Dick a greater expectable benefit, Dick could not rationally choose that we do X. If this claim were true, that would count against the view that, if we have to make this decision on Dick's behalf, we ought to do X. It would be implausible to claim that we ought to treat people in ways in which they could not rationally choose to be treated.

Dick, however, could rationally choose that we do X. Dick is not rationally required to choose that we act in the way that would be expectably better for him. As I have said, though X would be expectably worse for Dick, there is a one in two chance that X would actually be better for Dick, and this act would be better for Dick if he is very badly off. Given these facts, Dick could rationally choose that we act in this way.

VI. THE SHIFT IN WEIGHT CLAIM

Otsuka and Voorhoeve state other objections to the Priority View. They suppose that, as Prioritarians, we have stronger reasons to benefit people the worse off these people are. But they argue that, even if that claim is true, the Priority View fails in a different way.

Suppose that, in

or

Suppose next that each extra year of life would be an equal benefit. On that assumption, if we do Y rather than X, this act would give Jack a greater expectable benefit. Jack's possible gain of living for twenty more years would be greater than Jack's possible loss of living for ten fewer years, and this gain would be as likely as this loss. On the Priority View, however, gains and losses matter more, or have greater moral weight, if they come to people who are worse off. If Jack dies at 60, he would be worse off than if he dies at 70, in the sense that his life would go worse. On this view, if we give ten more years of life to someone who would otherwise die at 60, this act would give this person a greater weighted benefit than giving only eleven or twelve more years of life to someone who would otherwise die at 70. But compared with ten extra years for someone aged 60, there would be some rough number of extra years for someone aged 70 that would be a greater weighted benefit. We can suppose that, as Prioritarians, we believe this number to be twenty or more years. When we consider Case Eight, we would then believe that

1. (I) we would have a slightly stronger reason to do Y rather than X, since Jack's possible gain of living to 90 rather than 70 would marginally morally outweigh Jack's possible loss of living to 60 rather than 70.

On this view, if other things are equal, we ought to do Y.Footnote ²⁷

Suppose next that, in

or

If we are Prioritarians, Otsuka and Voorhoeve would claim, we would here reach similar conclusions. If we believe that,

we would also believe that,

Otsuka and Voorhoeve argue that, in failing to distinguish between cases like Eight and Nine, Prioritarians make a serious moral mistake. We are supposing that, as Prioritarians, we believe that

1. (J) in Case Eight, Jack's possible gain would only marginally morally outweigh Jack's possible loss.

Otsuka and Voorhoeve claim that, if we believe (J), we cannot also defensibly believe that

1. (K) in Case Nine, Jill's possible gain would morally outweigh Jack's possible loss.

We ought instead to recognize that

1. (L) When we turn from one-person cases like Eight to multi-person cases like Nine in which people's interests conflict, there is a shift in the moral weight of such possible gains and losses.

On this

Prioritarians are mistaken, Otsuka and Voorhoeve claim, because these people fail to recognize that there is any such shift in the moral weight of such gains and losses. They also write that this

This objection to the Priority View Otsuka and Voorhoeve call ‘the crucial argumentative move of this paper’.Footnote ²⁸

VII. OUTCOMES, PROSPECTS AND EXPECTABLE LEVELS OF WELL-BEING

Otsuka and Voorhoeve intend this objection to apply to all versions of the Priority View, including the telic version which makes claims about the goodness of outcomes.Footnote ²⁹ So we can first ask whether, in our beliefs about the goodness of outcomes, we should accept the shift in moral weighting that Otsuka and Voorhoeve describe.

Suppose we believe that, in Case Eight,

would do slightly more to make the outcome better than

If we accepted Otsuka's and Voorhoeve's Shift in Weight Claim, we would then believe that, in Case Nine,

would do less to make the outcome better than

If (1) would be in this way better than (2), but (3) would be in this way worse than (2), (1) would have to be better than (3). It would have to be true that

But this cannot be true. Jill's well-being matters just as much as Jack's. It would not be worse if it is Jill rather than Jack who lives for an extra twenty years.

Otsuka and Voorhoeve might now restrict their claims about the required shift in moral weighting. They might say that, when we turn from cases like Eight to cases like Nine, we should reach different conclusions not about the goodness of outcomes, but only about what we ought to do. In Case Eight we could justifiably do Y, thereby making the outcome expectably better. But in Case Nine, though doing Y would again make the outcome expectably better, this act would be wrong.

If Otsuka and Voorhoeve made these claims, they would no longer be stating an objection either to the Telic Priority View, or to Telic Egalitarianism. These views make claims that are only about the goodness of outcomes.Footnote ³⁰ But if it would be wrong, in cases like Eight, to do what would make the outcome expectably better, that would be an important conclusion. So we should ask whether Otsuka's and Voorhoeve's claims support this conclusion.

Otsuka claims that, though the Priority View

this view

In both Cases Eight and Nine, someone's life might go well, because this person would live to 90, and someone's life might go badly, because this person would live to only 60. But in Eight it is the same person, Jack, whose life might go either well or badly, and in Nine, though Jack's life might go badly, it is a different person, Jill, whose life might go well.

In Case Eight, Otsuka and Voorhoeve write, we could justifiably expose Jack to the risk of his lesser loss for the sake of his greater possible gain. But in Case Nine,

Otsuka also writes that, according to Prioritarians:

This claim is too strong, since Prioritarians may not be Act Consequentialists.Footnote ³⁴ But these remarks suggest how it might be wrong, in cases like Nine, to do what would make the outcome expectably better. Otsuka and Voorhoeve might claim that

1. (M) in deciding what we ought to do, we ought to take into account not only the goodness of outcomes, but also the goodness of people's prospects, in the sense of their chances of receiving benefits or burdens, and their expectable levels of well-being.Footnote ³⁵

On this view, it matters whether, in Otsuka's words, those who ‘lose out’ had ‘any prospect of gain’.

In considering (M), we can start with an example that does not involve chances or merely expectable benefits. Suppose that, in

or

Suppose next that each extra year of life would be an equal benefit. Utilitarians would then believe that we could justifiably do either X or Y, since both acts would give these people the same total of forty more years of life. Egalitarians would disagree. These people would believe that, if we do X, so that Jack lives to 80 and Jill lives to only 40, the outcome would be in one way worse, since there would be more inequality between different people. For that reason, Egalitarians would claim, we ought to do Y.

Though Prioritarians do not believe that inequality is bad, they would also believe that we ought to do Y. To apply the Priority View, we should compare the benefits of

1. (1) living to 60 rather than 40,

and

1. (2) living to 80 rather than 60.

If each extra year of life would be an equal benefit, (1) and (2) would be equal benefits, since they would both involve having twenty more years of life. But someone who will live to only 40 is relevantly worse off than someone who will live to 60. Prioritarians would therefore believe that (1) would be a greater weighted benefit than (2), since (1) would come to someone who would otherwise be worse off. On this view, we ought to do Y rather than X because we would thereby give this greater weighted benefit not only to Jack but also to Jill.

Suppose next that, in

or

If Jack and Jill are in other ways relevantly similar, as we can here assume, these outcomes would be equally good, and we would have equal reason to save either Jack's life or Jill's.

Some people would believe that we would be morally permitted to act in either of these ways. Others accept what we can call

On this view, rather than simply saving either Jack's life or Jill's, we ought to toss a coin, or use some other randomizing device, thereby giving Jack and Jill such equal chances of being saved.

Of those who reject the Equal Chances View, some believe that all that matters is what actually happens. These people might claim that if we toss a coin, and Jill loses, that would be just as bad for Jill as if we had simply gone ahead and saved Jack rather than Jill. This claim is true, but is no objection to the Equal Chances View. When we don't know what the effects of our acts would be, we often ought to do what would have a greater chance of having good effects. Suppose, for example, that if my doctor gives me treatment A, this would have a 1 per cent chance of saving my life, and that treatment B would have a 99 per cent chance of saving my life. These facts would give both me and my doctor strong reasons to choose treatment B. It is true but irrelevant that, if my doctor gives me treatment B, I may have bad luck, and die. In Case Eleven, Jill would have similar strong reasons to prefer that we give her an equal chance of being saved, rather than simply going ahead and saving Jack's life.

Other people reject the Equal Chances View with the different claim that if, in cases of this kind, we merely go ahead and save one of two people, that is already close enough to giving these people equal chances of being saved. If we save the person who is nearer to us, for example, each of these people would have been equally likely to be in this position. In some cases, however, no such claim is true. If we save the person whom we know to be much richer, because this person is more likely to give us some reward, the poorer person may have had no chance of being in this position. Nor could such claims show that we could justifiably reject the Equal Chances View. Unless we must act immediately, so that we have no time to toss a coin, we have at least some moral reason to give Jack and Jill equal chances, and we have no reason not to act in this way. If either Jack or Jill claimed that we ought to give them equal chances, we could not reasonably reject that claim.Footnote ³⁶ We ought, I conclude, to accept the Equal Chances View.

This view is in one sense Egalitarian, since it requires giving people equal chances. But this view is not distinctively Egalitarian, since Prioritarians could also accept this view. As I have said, when Otsuka and Voorhoeve argue against the Priority View, their remarks suggest that

1. (M) in deciding what we ought to do, we ought to take into account not only the goodness of outcomes, but also the goodness of people's prospects, in the sense of their chances of receiving benefits or burdens, and their expectable levels of well-being.

Prioritarians should accept (M). According to another statement of the Probabilistic Priority View, which we can call

We are considering cases in which our acts would affect the length of people's lives. In such cases, the goodness of people's prospects, or their expectable levels of well-being, roughly corresponds to their life expectancy, in the sense of their expectable number of years of life.

In Case Eleven we could

or

If we do Z rather than X, this act would be expectably worse for Jack, whose life expectancy would fall from 80 to 60. But this act would be expectably better for Jill, whose life expectancy would rise from 40 to 60. Jill's expectable gain of twenty more years of life would be as great as Jack's expectable loss. And if we did X rather than Z, Jill would be much worse off than Jack, by dying forty years younger. So if we do Z rather than X, we would give an equal expectable benefit to someone who would otherwise be worse off, thereby giving this person a greater expectable weighted benefit. PP2 therefore implies that we ought to do Z rather than X. For similar reasons, this view implies that we ought to do Z rather than Y. So this view implies the Equal Chances View.Footnote ³⁷

If Prioritarians take into account, not only the goodness or expectable goodness of outcomes, but also the goodness of people's prospects, and how well off people expectably are, they must answer several questions about the relative importance of these facts. Here is a partial answer to one of these questions. Prioritarians may believe

Suppose next that, in Expanded Nine, we have a wider range of possible acts. We could

or

These last three acts would have similar effects. Whether we do Y, Z, or T, there would be equal chances that either

or

When we assess the goodness of these outcomes, it makes no difference whether it is Jack or Jill who might have these longer or shorter lives, so these three acts would have outcomes that would be expectably equally good. But these acts differ in their effects on people's prospects, or their expectable levels of well-being. Jack and Jill's prospects correspond to their life expectancies, which would be these:

Whether we do Y, Z or T, we would give Jack and Jill the same total sum of expectable benefits, since their combined life expectancies would be 145. But these acts would distribute these benefits in different ways. If we do T rather than either Y or Z, we would give more of these expectable benefits to someone who would otherwise have a life expectancy of only 65, thereby being expectably worse off. PP3 therefore implies that, if other things are equal, we ought to do T rather than either Y or Z.

We can now return to what Otsuka and Voorhoeve call their main objection to the Priority View. According to

Prioritarians are mistaken, Otsuka and Voorhoeve argue, because these people must deny that there is any such shift in the moral weight of such gains and losses.Footnote ³⁸ A similar objection applies, we can add, to Telic Egalitarians.

If Prioritarians make the claims that I have described, they can answer this objection. To make our cases easier to compare, we can suppose that, in Case Eight, Jill exists, but will be unaffected by our acts, and will live to 70 whatever we do. The life expectancies would then be these:

In both cases, our doing Y would give Jack and Jill the same total sum of expectable benefits, since their combined life expectancies would be 145. But these acts would distribute these expectable extra years of life in different ways. If we do Y in Case Nine, this act would give fewer of these years to someone who would be expectably worse off, by having a life expectancy of only 65. PP3 therefore implies that

1. (N) Compared with our reason to do Y in Eight, we would have a weaker reason to do Y in Nine.

If we do Y in Eight, Jack's life expectancy would be 75 because Y gives Jack equal chances of living to 90 or to 60. We have supposed that, on the Priority View, Jack's possible gain in living to 90 rather than 70 would only marginally morally outweigh Jack's possible loss in living to only 60 rather than 70. Prioritarians would therefore claim that

1. (O) Since this possible gain would only marginally morally outweigh this possible loss, our reason to do Y in Eight would be only slightly stronger than our reason to do X.

They can add that

1. (P) Our reason to do X would be equally strong in both Eight and Nine.

Prioritarians can therefore claim:

1. (Q) Since our reason to do Y in Nine would be weaker than our reason to do Y in Eight, which would be only slightly stronger than our reason to do X, our reason to do Y in Nine may be weaker than our reason to do X. If our reason to do Y in Nine is in this way weaker than our reason to do X, and other things are equal, we could justifiably do Y in Eight, but it would be wrong to do Y in Nine.

In giving this argument, Prioritarians would both accept Otsuka's and Voorhoeve's Shift in Weight Claim, and provide a plausible explanation of this claim's truth.

When Otsuka and Voorhoeve defend this claim, some of their remarks suggest that

1. (R) If possible gains and losses would come to the same person, any chance of some loss could be outweighed by an equal chance of any greater gain. If such gains and losses would come to different people, that is not true.

If we are Prioritarians, we would reject (R). As we have just seen, however, when we consider cases like Eight and Nine, Prioritarians can accept Otsuka's and Voorhoeve's claims about what we ought to do. We can agree that

1. (S) though we could justifiably do Y in Eight, giving Jack his possible gain at the risk of a lesser loss, it may be wrong to do Y in Nine, giving Jill the same possible gain at the risk of the lesser loss to Jack.

As Prioritarians, we can defend (S) without appealing to (R). We can appeal to

We can then give the argument summarized in (Q). If we do Y in Eight, Jack's life expectancy would be 75 and Jill's would be 70. If we do Y in Nine, Jack's life expectancy would be only 65, and Jill's would be 80. Though both acts would give Jack and Jill combined life expectancies of 145, Jack would be expectably worse off in Nine than either Jack or Jill would be in Eight. This fact is enough to explain the shift in moral weighting. If we could justifiably do Y in Eight, but it would be wrong to do Y in Nine, that need not be claimed to be because Eight is a one-person case and Nine is a multi-person case. It could be because, as PP2 claims, we have stronger reasons to give people expectable benefits the worse off these people expectably are.

In proposing and defending their Shift in Weight Claim, Otsuka and Voorhoeve have pointed out some important moral truths. These truths are not less important if, as I have argued, they are no objection to the Priority View, and can be better explained in this Prioritarian way.

VIII. COMPETING CLAIMS

Otsuka and Voorhoeve present one more objection to the Priority View. This objection applies to cases that do not involve chances, risks and expectable benefits. In these cases we must decide how to distribute various resources between different people, thereby giving some of these people various benefits. When we make such decisions, Otsuka and Voorhoeve write, we need to invoke interpersonal considerations that are essentially relational, such as the intrinsic badness of inequality or the comparative strength of the claims of different individuals. This last phrase refers to what we can call the Competing Claims View.Footnote ³⁹ Otsuka and Voorhoeve then write:

This objection does not, I believe, succeed. There are different versions of the Competing Claims View. According to those whom we can call

These Egalitarians do appeal to what Otsuka and Voorhoeve call essentially relational facts. On this view, the strength of anyone's claim to receive some benefit depends in part on whether, and by how much, this person is either better off or worse off than other actual people. But Otsuka and Voorhoeve should not assume that, in comparing the strength of people's claims, we need to appeal to such essentially relational facts. According to those whom we can call

On this view, the strength of anyone's claim to receive some benefit depends only on this person's level of well-being. It is irrelevant whether this person is better off or worse off than other people. Claim Prioritarians could, however, restate their view as

Otsuka and Voorhoeve seem to assume that claims like PP4 appeal to essentially relational facts. That may be why they write that the Priority View view rules out ‘such essentially comparative considerations’. But this assumption misunderstands this view. In a passage that Otsuka and Voorhoeve quote, I wrote:

When Otsuka and Voorhoeve discuss the comparative strength of different people's claims, they may seem to be asserting the Prioritarian PP4. They write:

The word ‘relatively’ seems to make no difference here. This sentence might mean

Prioritarians would accept this claim, which is implied by PP4. Otsuka elsewhere writes:

This last clause may seem to add nothing, since we could instead say

or, more simply

This is another Prioritarian claim.

In these quoted sentences, however, Otsuka and Voorhoeve were intending to imply that

1. (T) the strength of an individual's claim to receive some benefit depends on whether, and by how much, this person is better off or worse off than other actual people.Footnote ⁴⁴

Since they accept (T), Otsuka and Voorhoeve are Claim Egalitarians. Prioritarians would reject (T), since (T) appeals to essentially relational facts.

To apply (T), we must answer various questions about the relative importance of such relational facts. Different versions of (T) may correspond to different Egalitarian views about the badness of different patterns of inequality. According to some Claim Egalitarians, for example,

1. (U) people have stronger claims to receive some benefit if their level of well-being is further below the average level of everyone's well-being, or further below the median level, or the level of the best-off people.

Claim Egalitarians must also answer questions about who are the people whose being at different levels of well-being constitutes morally relevant inequality. On the simplest answer, the word ‘everyone’ in claims like (U) refers to everyone who ever lives. There are several other more complicated answers.

Otsuka and Voorhoeve believe that, compared with the Prioritarian version of the Competing Claims View as stated by PP5, some Egalitarian version, such as the view stated by (U), is more plausible.

Such Egalitarian views are not, I believe, more plausible. When we compare the strength of two people's claims to receive some benefit, it is often enough to know how well off, or badly off, these two people are. In such cases, we do not need to know how these people's levels of well-being compare with the levels of other people, in the various ways described by (U). The strength of each person's claim depends only on how well off this person is. One of two people would have a stronger claim if this person is worse off; but this person's claim would not be stronger because there are other people who are better off.Footnote ⁴⁵

In some other cases, we would need to know more. People's claims to receive certain benefits may depend on other facts, such as facts about how much different people have contributed to some cooperative scheme. There are also many cases in which, for egalitarian reasons, we ought to give people equal shares, or do what would make them equally well off. We need not discuss these other cases here. My aim has only been to show that Claim Prioritarians can answer Otsuka's and Voorhoeve's objection to their view. People can have stronger claims to receive some benefit the worse off these people are.

The Priority View, Otsuka and Voorhoeve write, cannot be defensibly claimed to be ‘a “distinctive view” that provides a genuine and attractive alternative to views that are essentially comparative in nature’.Footnote ⁴⁶ That, I have argued, is not so.

There are some other relevant questions that I have not discussed. It is sometimes claimed, for example, that we have prioritarian reasons to have children, since we would thereby benefit some of the possible people who would otherwise be badly off, by never existing. We might be similarly claimed to have egalitarian reasons to have children, since we would thereby reduce the inequality between some possible people and those actual people who have lives worth living. But when we apply these distributive principles, we should not include, among the people who are badly off, possible people who never exist. Like the Principles of Personal Good, or Pareto Principles, the Prioritarian Principles that I have considered cannot be applied to cases in which, in the different possible outcomes, different people would exist. When we consider these cases, we need other principles.Footnote ⁴⁷