II. THE CASE AGAINST THE PRIORITY VIEW

Having set the stage, we now turn to rehearse the two main arguments employed by Otsuka and Voorhoeve, the first of which I shall term the Cure versus Damage Limitation Objection.

Consider initially the following Intrapersonal Conflict. Suppose it is known that some young adult, Unlucky, will develop one of two equally probable medical conditions which, if left untreated, will produce, respectively, disabilities characterized by Slight Impairment or Very Severe Impairment.Footnote ¹⁰ A ‘morally motivated stranger’ can now provide Unlucky with only one of the two possible treatments available,Footnote ¹¹ namely

and

Suppose that because of the specific benefits Cure and Damage Limitation provide Unlucky's own preferences and the best account of her utility, or what she has reason to value for her own sake, coincide in favouring indifference between the two treatments. Under these conditions, Otsuka and Voorhoeve conclude that the Stranger has an undefeated reason to share Unlucky's indifference, at least when considering her condition in isolation from the level of advantage of other individuals. More generally, they claim it is in this sense reasonable to provide the ‘treatment that maximizes the expected increase in the utility of the recipient’.Footnote ¹²

Now consider a second related Interpersonal Conflict involving a group of young adults half of whom are identifiable as facing Very Severe Impairment whilst the other half faces Slight Impairment. Individuals’ preferences are constant across the two conflicts and once again provide accurate measures of their utility. Each individual, therefore, correctly regards Damage Limitation for Very Severe Impairment and the Cure for Slight Impairment as increasing utility by the same amount. Nevertheless Otsuka and Voorhoeve insist that it would not be reasonable for the stranger to be indifferent between providing the two treatments. She must instead protect individuals from the most serious condition. As they conclude, ‘Our considered judgment is that your only reasonable option is to provide the treatment for those that will develop the very severe impairment.’Footnote ¹³ Furthermore, they point out that various surveys by economists suggest that their conclusion is shared since there is a widespread preference to give significant priority to the worse off in decisions about interpersonal allocation of health care.Footnote ¹⁴

Having alleged that a contrast exists in the reasons governing the Intrapersonal and the Interpersonal Conflicts, Otsuka and Voorhoeve then offer the following explanation of how their description refutes the Priority View. The explanation relies on attributing the following two assumptions to the Priority View,Footnote ¹⁵ namely

1. (a) Diminishing Value of Benefits: The weight of our reasons to benefit others decreases as the level of advantage of the recipient increases,

and

1. (b) Invariance across Conflicts: The same weightings apply in both intrapersonal and interpersonal conflicts.

Appealing to the first assumption, Otsuka and Voorhoeve argue that the Priority View is attractive insofar as the assumption mandates treatment for Severe Impairment rather than Slight Impairment in the Interpersonal Conflict. The Priority View's implausibility becomes fully apparent only when we realize that the second assumption of invariance favours extending our greater concern for the less advantaged from the Interpersonal to the Intrapersonal Conflict, thereby also mandating treatment for Severe Impairment in the second scenario. Such an implied mandate, however, conflicts with what Otsuka and Voorhoeve regard as the more plausible assumption that there is an undefeated reason to treat slight impairment in the Intrapersonal Conflict since doing so maximizes expected utility. The implausibility of this implication, they conclude, provides powerful counter-evidence against the Priority View.

As should be apparent, the Cure versus Damage Limitation Objection depends heavily on an assumption about the reasonableness of expected utility maximization in intrapersonal conflicts. Otsuka and Voorhoeve note that others have already pointed to a divergence between the Priority View and expected utility maximization in such cases but have not concluded that it constitutes a reason to reject the View.Footnote ¹⁶ Facing the Objection, then, an advocate of the Priority View might dig in her heels, and insist that its major assumption is insufficiently compelling for the Objection to provide decisive counter-evidence. Those sceptics, however, face a further but related criticism, which Otsuka and Voorhoeve regard as their ‘crucial argumentative move’.Footnote ¹⁷

The second Intervention versus Non-Intervention Objection focuses directly on an alleged shift in the moral importance of various benefits and burdens that takes place when moving from intrapersonal to interpersonal trade-offs. To understand the Objection, consider the following two cases.Footnote ¹⁸

III. THE CASE FOR RELATIONAL EGALITARIANISM

As the title of their paper suggests, Otsuka and Voorhoeve claim that the examples used in the previous two Objections not only refute the pure Priority View but also support the rival relational view of equality. The examples are alleged to have this implication because they illustrate an important difference between aggregating benefits within lives rather than across lives which relational views are uniquely capable of explaining. Summarizing this ambitious positive conclusion, Otsuka and Voorhoeve state that:

Here Otsuka and Voorhoeve mention two quite different ways of drawing upon relational factors in order to explain the shift under consideration.

The first explanation appeals to a form of Telic Egalitarianism according to which it is bad (because unfair) for some individuals to be unavoidably disadvantaged relative to others. In the examples involving intrapersonal conflict this factor plays no role since the examples involve only a single individual who is considered in isolation. Inequality becomes relevant only in the interpersonal conflicts where its presence allegedly explains why the balance of reasons shifts to provide decisive support to treatment for severe impairment and to intervention, namely because those two options now benefit individuals in ways that also diminish the bad of relative disadvantage.Footnote ²²

Even if the telic explanation is successful it has an important strategic disadvantage for those hoping to defend a relational view of equality because it relies on the controversial assumption that inequality makes outcomes bad, and so provokes the Levelling Down Objection. The disadvantage arises because the explanation's proponents must rebut the Objection if their explanation is to succeed. Doing so is a formidable task, however, and one which if accomplished renders additional arguments in favour of relational egalitarianism much less significant. It would be advantageous, then, for relational egalitarians to explain the shift in a way that bypasses the Objection. As Otsuka and Voorhoeve note,Footnote ²³ their additional explanation eschews reliance on any reasons to level down in order to promote equality, and so possesses this significant advantage.

Instead of invoking any telic assumptions about the badness of unequal outcomes, the second explanation relies on a more complex deontic account of why inequality matters.Footnote ²⁴ The account assumes that decisions to distribute benefits to some rather than others should be justifiable to each affected party in a way that depends upon the strength of each individual's separately assessed claim to be benefited. In addition, the account assumes that distributive decisions should satisfy the strongest claims, and that, crucially, the strength of one individual's claim compared with another's claim depends at least in part upon those individuals’ relative positions. As Otsuka and Voorhoeve explain,

Thus construed, the deontic account of why inequality matters provides a ready explanation of how the balance of reasons might shift decisively in favour of damage limitation for severe impairment and intervention in situations of interpersonal rather than intrapersonal conflict. In interpersonal conflicts some individuals have comparatively strong claims on those options in virtue of their relative disadvantage compared with others with weaker claims on the cure for slight impairment and non-intervention. In contrast, in the intrapersonal conflicts nobody can voice the same strong claim since all individuals are similarly situated. Note too that since the deontic account assumes that the object of a claim is always some benefit the explanation just offered never favours pursuing equality by levelling down to nobody's benefit. As mentioned, then, the deontic explanation has a significant strategic advantage over the telic explanation.

IV. THE PRIORITY VIEW RIDES AGAIN

Having outlined the case against the Priority View and in favour of the relational alternative, we now turn to develop one prioritarian response. The response involves drawing a distinction between telic and deontic versions of the Priority View parallel to the distinction Otsuka and Voorhoeve rely upon between telic and deontic forms of relational egalitarianism. Parfit himself alludes to such a distinction when he writes that:

In his Lindley Lecture Parfit does not indicate whether he favours the telic or deontic form but claims only that ‘for most of my discussion, this difference does not matter’.Footnote ²⁷ I shall argue, however, that the distinction is relevant in assessing the force of Otsuka's and Voorhoeve's critique of the Priority View. More specifically, I shall show how one deontic version of the Priority View escapes the critique, and refutes its authors’ attempt to vindicate relational egalitarianism.

As a preliminary, we should grant that some notable versions of the Priority View are telic insofar as they are based upon a requirement to promote valuable outcomes, and understand the prioritarian conviction that benefiting people matters more the worse off these people are as a claim about how to evaluate different outcomes. One such example is the distribution-sensitive consequentialism found in the early work of Samuel Scheffler, and, in different forms, in the more recent work of Richard Arneson and Brad Hooker.Footnote ²⁸ Like their purely aggregative utilitarian ancestors, these versions of the Priority View are inclusive insofar as they apply the same weightings to benefits in cases of intrapersonal and interpersonal conflict, and so face Otsuka's and Voorhoeve's challenge that they cannot accommodate the shift in our reasons that allegedly arises across such cases.

Suppose, however, that we question whether proponents of Parfit's Priority View need to assume Invariance across Conflicts, and consider a restrictive deontic version of the Priority View. It assumes the individuals have claims on each other's beneficence but that the content of those claims differs significantly depending on whether we need to resolve normative conflicts within rather than across lives. In cases involving only intrapersonal conflicts this Restrictive Priority View assumes individuals have a claim that conflicts are resolved in a way that maximizes their expected utility, at least assuming individuals’ preferences and utility coincide. In contrast, in cases of interpersonal conflict individuals have claims to be benefited that become stronger as their absolute position worsens. On the Restrictive View, then, the conviction that benefiting people matters more as they become worse off does not apply to all our decisions but instead is triggered by a specific context in which we face interpersonal conflicts or must choose between promoting personal interests and impersonal values.

Since it affirms their twin convictions that unity within persons’ lives favours aggregating benefits whilst the distinction between persons’ separate lives places moral barriers on aggregation the Restrictive View can withstand both of Otsuka's and Voorhoeve's objections. Thus, the Restrictive View can easily accept that it is reasonable to prefer Cure to Damage Limitation in the first intrapersonal conflict whilst also treating the reasons in favour of Damage Limitation as decisive in the corresponding interpersonal conflict because the second case triggers a concern to give priority to the less advantaged. Given that the Restrictive View affirms a discontinuity between the factors that should govern the distribution of benefits within lives and across lives, it also has no difficulty in accommodating a shift in our preferences when choosing between intervention and non-intervention in cases of intrapersonal and interpersonal conflict. Since it explains that shift by appealing to the difference in our reasons to aggregate benefits within lives and our reasons to give priority to the less advantaged when distributing benefits across lives but grounds the weightier claims of the less advantaged on their absolute rather than relative position, the Restricted View also undermines any suggestion that relational views of equality are uniquely well-placed to account for the shift. In short, then, the Restricted View sails past Otsuka's and Voorhoeve's objections, and refutes their vindication of relational egalitarianism.

V. NAGEL'S VIEW

Before addressing one possible rejoinder, let me briefly attempt to dispel the suspicion that the Restrictive View plays no role in the Lindley Lecture and lacks any rationale. To do so we can note some of its affinities with the form of egalitarianism advocated by Thomas Nagel in his 1977 Tanner Lecture on ‘Equality’ and in Equality and Partiality, the book based on his 1990 John Locke Lectures.Footnote ²⁹

In both works Nagel appeals to an ideal of unanimity which, in situations where others have conflicting claims on our concern, favours the outcome that is least unacceptable when considered from each person's separate perspective. He also assumes that, ceteris paribus, unacceptability increases as a person becomes ‘worse off’. Although in his Tanner Lecture Nagel claims to be assessing outcomes and fails to distinguish clearly between individuals being worse off on an absolute scale rather than relative to others, his later remarks clearly indicate that he thinks unacceptability varies with absolute levels of advantage. For example, referring to his conviction that ‘the proper form of equal concern for all will sometimes favour benefit to the worse off even when numbers or quantity go the other way’ Nagel endorses a description of his view as non-relational; doing so, he writes that in a manuscript entitled On Giving Priority to the Worse Off Derek Parfit ‘calls this form of egalitarianism the Pure Priority View, to distinguish it from an attachment to equality . . . which he calls Relational Egalitarianism’.Footnote ³⁰ Parfit himself makes the same suggestion in his Lindley Lecture when he writes that ‘Nagel is one writer who sometimes uses the language of equality, when he is really appealing to the Priority View.’Footnote ³¹ It is puzzling, then, that Otsuka and Voorhoeve interpret Nagel as affirming relational egalitarianism,Footnote ³² and claim to draw on his view in their second explanation of why the balance of reasons shifts between intrapersonal and interpersonal aggregation. In the absence of an argument for their interpretation, I think the non-relational reading of Nagel's view remains at least an eligible interpretation, which suffices for my purposes.

In considering his view, it is also notable that Nagel stresses that his concern to give priority to the less advantaged is a view about ‘how to settle conflicts among the interests of different people’, and speaks of pairwise comparison as ‘the natural way to deal with conflicting claims’.Footnote ³³ It is clear that other philosophers assume that there are distinctive principles governing the resolution of interpersonal conflicts. The most obvious example is John Rawls, who characterizes principles of ‘social justice’ in part by appeal to their role in governing the distribution of advantage in the face of separate persons’ conflicting interests.Footnote ³⁴ Of course, Nagel's previous two remarks are formally consistent with the claim that prioritarian reasoning governs the resolution of intrapersonal as well as interpersonal conflicts. When interpreting Nagel's position, however, we should bear in mind that recognition of the dissimilarity between these two types of conflict is a long-standing feature of his thought.

Thus, in his first book, The Possibility of Altruism, Nagel argues that the unity of a single person's life favours one principle for intrapersonal conflict resolution whereas the separateness of different persons’ lives favours a quite different principle for interpersonal conflict resolution. Expressing this insight, Nagel writes that:

Echoing Rawls, Nagel goes on to write that:

Given these earlier statements by Nagel and the absence of any explicit statement that his prioritarian approach to interpersonal conflict resolution is also applicable to intrapersonal conflict, I conclude that the Restricted View has a clear Nagelian and Rawlsian pedigree.

Nagel's statements are also important because they suggest a natural rationale for the Restricted View, and consequently a reply to the inevitable suspicion that it is objectionably ad hoc to reply to Otsuka and Voorhoeve by applying prioritarian reasoning to interpersonal trade-offs but not to intrapersonal trade-offs.

The rationale starts with the assumption that when an individual or set of individuals makes trade-offs within their own lives considered in isolation from the lives of others each has undefeated reasons to maximize her own expected utility; there is no distinct additional requirement to attach greater weight to benefits that might fall at a lower absolute level of advantage. The rationale then insists that the normative situation does not change when the same decision must be made by a second decision-maker acting on behalf of the previous individuals. Some may balk at this second stage but the rationale insists that what we owe to others when we need to resolve their purely intrapersonal conflicts, and consider their lives in isolation, involves conforming to the reasons they should and would have acknowledged. Assuming those individuals' preferences and reasons are known to coincide, the second decision-maker should therefore also maximize expected utility.

According to the rationale, the normative situation then changes when we move from intrapersonal to interpersonal conflict resolution. It does so because when a purely intrapersonal conflict is resolved by withholding some benefit to an individual such a decision does not necessarily come at any particular individual's expense. The decision-maker can always appeal to the possibility that withholding some benefit produces a greater benefit for that very same individual that he himself had reason to prefer, and perhaps actually does prefer. In contrast, when a decision-maker resolves an interpersonal conflict by withholding some benefit to an individual such a justification appealing to the unity within a life is not always available. Instead the decision comes at the expense of some individual for the sake of some separate individual, and some other type of justification is necessary to show that sufficient weight was attached to the losing individual. The rationale insists that these specific conditions result in the weight of our reasons to benefit individuals diminishing as the recipients enjoy greater benefits on some absolute scale. Perhaps further investigation will show such insistence to be unsound, but the pedigree of the rationale does at least suggest that the Restricted View is not a merely an ad hoc manoeuvre.

VI. REJOINDER

Although Otsuka and Voorhoeve do not consider the Restricted View, they do explicitly anticipate critics asking ‘Is the Priority View Meant to Apply to the One-Person Case?’Footnote ³⁷ Their treatment of the question suggests at least one possible response to my objection worth considering.

The response draws on Parfit's characterization of the Priority View through an analogy, which he states as follows.

Drawing on this statement, and especially its final italicized sentence, Otsuka and Voorhoeve argue that ‘Parfit's use of the altitude analogy . . . to characterize the Priority View implies that this view applies to a case in which one must choose how to aid a single individual.’Footnote ³⁹ They might, then, argue that even if my objection does show that one non-Parfitian formulation of the Priority View is immune to their central argument the objection does not refute their conclusion that if the shift they identify is justified then ‘the Priority View as formulated in Derek Parfit's Lindley Lecture would be unsound’.Footnote ⁴⁰

When assessing this possible rejoinder it is worth noting at the outset that Parfit uses the altitude analogy to support his conclusion that the ‘chief difference’ between relational egalitarianism and the Priority View is that ‘Egalitarians are concerned with relativities: with how each person's level compares with the level of other people. On the Priority View, we are concerned only with people's absolute levels.’Footnote ⁴¹ However, the italicized last sentence in Parfit's statement of the analogy, on which Otsuka and Voorhoeve rely heavily, plays no essential role in grounding this conclusion. Instead the sentence plays a rhetorical role by reiterating the previous essential premise that on the Priority View whether individuals are worse off than others has no relevance in determining the weight of any reasons to benefit them. It is debatable, then, whether so dispensable a remark should play a major part in identifying how the Lindley Lecture construes the Priority View.

Suppose, however, that we should not simply disregard Parfit's remark that according to the Priority View benefits to the worse off ‘would matter just as much even if there were no others who were better off’. Even then it would be hasty to conclude that Parfit's formulation of the Priority View is always inclusive, and so unable to accommodate the shift Otsuka and Voorhoeve affirm.

One reason to question such an inference is that whilst the restricted version of the Priority View attaches priority to benefiting worse off recipients only in cases involving interpersonal conflicts, that version does still supply some guidance in cases involving only one recipient; it does so because here it favours expected utility maximization.Footnote ⁴² It is possible then for proponents of the Restricted View to accept Parfit's remark because they believe that the reasons for benefiting a badly off individual are just as conclusive in both types of case despite some variation in what makes this true.

Suppose, however, that Parfit's remark does show that he sometimes understands the Priority View in the inclusive manner Otsuka and Voorhoeve propose. Even then, it does not follow that Parfit never implies that the Priority View can also be understood in the manner I have proposed. The inference would be secure if, as Otsuka and Voorhoeve perhaps also assume, the Lecture suggests only one formulation of the Priority View. But it seems more likely that a work as fertile as Parfit's Lecture contains several distinct versions of the Priority View. Indeed, Parfit's reference to telic and deontic forms of the Priority View and his description of Nagel as a proponent of the Priority View supports this conjecture, at least on my reading of Nagel, and suggests that Restricted View is one of the formulations present in the Lecture.

I conclude then that Parfit's remarks about the altitude analogy show at most that Otsuka and Voorhoeve have cast doubt upon only one formulation of the Priority View present in the Lindley Lecture. If so, their paper constitutes a powerful challenge that telic prioritarians need to address. (I sketch one way for them to do so in the next section.) Nevertheless the response gives no grounds to reject my suggestion that another formulation of the Priority View present in the Lecture escapes their challenge. The critical impact of their argument is, therefore, far more limited than it may initially appear: the argument threatens distribution-sensitive consequentialist variants of the Priority View rather than every version of the View present in the Lindley Lecture. Moreover, and most importantly, the response does nothing to support their main positive conclusion that it matters that some are worse off than others because only relational egalitarians can explain why a shift in the balance of reasons takes place in the move from intrapersonal to interpersonal trade-offs. Even if Parfit had never alluded to a restrictive formulation of the Priority View, the availability of that formulation shows that it is possible to accommodate such a shift without affirming relational egalitarianism.