The Imperialism of Desert

3.1. Why the Pluralist View is not Obvious

While, as noted above, value pluralism may seem evidently correct when the question is considered in the abstract, it is less attractive when the issue is considered more specifically. Indeed, the just slightly less abstract proposition that inequality is unjust when the relevant persons have exactly what they deserve, seems odd. The pluralist view is even less appealing with regard to more concrete examples. Consider a vicious murderer whose life is less good as a result of the punishment that he received, and specifically, less good than the life of an innocent person. When the punishment is justified, it seems that this is not because a consideration of desert defeats a consideration of equality, priority or even utility that oppose the reduction in the wellbeing of the criminal due to the punishment. Rather, the punishment is justified, when the criminal gets what he deserves, because desert undercuts these other considerations. In this spirit, Larry Temkin (2017: 45–46) explains that “egalitarians needn’t object if criminal John is worse off than law abiding Mary, even if John craftily avoided capture, and so is only worse off because, through no fault or choice of his own, a falling tree limb injured him.” Indeed, it seems that those who justify punishment by appealing to the (non-instrumental) value of desert or retribution, do not typically think that the disutility to the criminal which resulted from the deserved punishment is pro tanto bad or unjust.

Similarly, consider two workers, performing the same task: One of them works 100 hours in one month while the other works 200 hours—and all else is equal (i.e., both invest the same effort for each hour of work, both are able to work more or less, and so on). Suppose that person who worked less deserves X and the person who worked more deserves Y, and (Y>X). If equality or priority for the worse off can coexist alongside desert, then the worker who worked less may claim that although he deserves X, he should get more than X, since desert should be balanced against equality or priority. This claim seems misguided.

The monistic view thus appears normatively appealing, despite its radical nature. Moreover, it has several related theoretical virtues. The most obvious is the virtue of simplicity: the monistic view explains diverse intuitions in light of a single basic value. It also explains the intuitive appeal of the pluralist view, since the implications of other considerations, such as equality and priority, are similar to those of desert when people are equally deserving or when there is no evidence that they are unequally deserving.

Monism about desert seems especially called for if we assume an absolute version of desert, according to which each person deserves a specific level of wellbeing. For according to this view, there is a point above which a person gets more than she deserves. Things are more complicated regarding comparative desert. On the one hand, unlike absolute desert, monistic comparative desert has implausible implications. For example, in a world populated by just one person, comparative desert does not apply and therefore it implies indifference between a situation in which his life goes very well or very badly regardless of whether he is evil or a saint. Similarly, monistic comparative desert seems to have the unreasonable implication that we should be indifferent between a situation in which two people are suffering, and a situation in which both flourish, just so long as the difference between them reflects their relative moral record—even if both are good people and the difference in their moral record is small.

On the other hand, the tension between desert and pluralism also seems to exist with regard to comparative desert. Assume that, in the above example, the level of wellbeing of the criminal and the innocent reflect exactly what they comparatively deserve. It seems odd to suggest that there is also a reason to improve the wellbeing of the criminal that should be balanced against the consideration of (comparative) desert. Such balancing seems odd especially if the rationale of the competing reason is related to equality or priority, for instance, if a criminal claims that there is a reason to give him more than he deserves since he has less than the innocent or since he is worse off.¹⁷ But balancing utility against comparative desert is also somewhat strange if the additional utility is due merely to an improvement in the wellbeing of the criminal who does not comparatively deserve more.

Unlike absolute and comparative desert, prioritarian desert—which holds that the reason to promote one’s wellbeing is greater in accordance with the degree to which she is morally deserving—is by definition a pluralist view in itself. It determines the degree of the good in light of two criteria: the degree to which wellbeing is promoted and the degree to which the relevant person is deserving.¹⁸ But here too, there seems to be at least some tension between these two criteria. For example, prioritarian desert entails that the best possible world (one with unlimited resources) is a world in which even the most evil person has an infinite amount of utility.¹⁹

Another implication of prioritarian desert, which appears to be implausible and incompatible with the intuitive understanding of the idea of desert, is that in a world with two people, one of whom is slightly more deserving than the other, the optimal state of affairs is that the more deserving person should get all of the available resources. For, according to this conception, the value of every unit of wellbeing a person has increases the more deserving that person is. In order to mitigate this implication, someone who accepts prioritarian desert is pushed into a view that balances this version of desert with other moral values such as priority to the worse off. In this respect, it is not surprising that Arneson’s pluralist view mentioned above refers to prioritarian desert, and rejects absolute and comparative desert.

It may be objected that at a certain point, when the (slightly) more virtuous person has (much) more than the less virtuous person, the value of promoting the wellbeing of the former is no longer greater than that of promoting the wellbeing of the latter. Assume, for example, that the moral record of the more virtuous person is just slightly better (say ten percent higher), but she has ten times more than the less virtuous person (in terms of wellbeing). Now assume that an additional good is to be allocated. It may be thought that, according to prioritarian desert, the less virtuous but much worse off person “deserves this gain” more. However, this interpretation of prioritarian desert seems to collapse into either absolute or comparative desert. For while prioritarian desert ranks the degree to which persons are more or less virtuous, it is incompatible with the conclusion that one person deserves a certain gain less than another because her level of wellbeing is already higher than she deserves to be compared to another—for this collapses into comparative desert. prioritarian desert is also compatible with the conclusion that one person deserves less than another because her overall level of wellbeing already exceeds what she deserves—for this collapses into absolute desert. That is, prioritarian desert is distinct only if it understands the phrase “the moral value of a gain … is greater the larger the degree to which the person deserves this gain” such that the moral value of a gain is greater the better her moral record is, regardless of her actual level of wellbeing.²⁰

3.2. Kagan’s Argument

The most explicit argument for a monistic view is offered by Kagan who claims that (absolute or comparative) desert should displace rather than be balanced against equality or priority as independent values. Kagan (1998: 300) depicts the value of desert with graphs that represent the degree to which the situation is good “from the standpoint of view of desert” (the Y axis)²¹ as a function of the level of wellbeing that a person is at (the X axis). Thus, the “desert graph” for a person looks like a mountain, whose peak is the level of wellbeing that this person deserves (see Figure 1).

Figure 1. The Desert Graph: goodness from the point of view of desert as a function of a person’s wellbeing. The graph for a person looks like a mountain with a peak at the level of wellbeing that the person absolutely deserves.

Kagan’s argument that a proper attention to desert demonstrates that there is no room for an intrinsic consideration of equality or priority, begins with the following case (the “Twin Peaks” case)²²: A is a sinner, who is doing better than what he deserves. B is a saint, who is doing less well than she deserves. Yet B is still better-off than A (she deserves much more). And we can benefit either A or B. The Twin Peaks case can be graphically represented as follows (Figure 2):

Figure 2. Twin Peaks: A is a sinner whose level of wellbeing is greater than what he deserves, and B is a saint whose level of wellbeing is lower than what she deserves.

In this case, equality favors benefitting A, the sinner, since he is worse-off. However, Kagan claims that it is obvious that we should prefer B, as desert advocates. “After all, B is a saint who is getting less than she deserves, while A is sinner who is doing better than he deserves.” Moreover, Kagan (1998: 305) claims that there isn’t even a pro tanto reason in favor of preferring the sinner.²³ Therefore, he concludes that there is no intrinsic consideration of equality or priority.²⁴

While Kagan directs his argument from desert only against equality and priority,²⁵ it seems to us that this argument suggests conclusions that are even more radical: that a commitment to desert is discordant also with other salient considerations such as sufficiency and utility.

Consider sufficiency first. While considerations of equality or priority apply to every level of wellbeing, some think that inequality is objectionable only if the worse-off is below a certain threshold of wellbeing (e.g., Frankfurt 1987; Crisp 2003). Kagan does not consider whether his argument applies to sufficiency.²⁶ However, it seems that the essence of his argument applies to sufficiency as well. Consider a sinner who is evil enough so that although he has more than he deserves, this is still somewhat less than the sufficiency threshold, and a saint who has much less than she deserves although she is above the sufficiency threshold. It seems that Kagan’s intuition regarding the Twin Peaks case has considerable force also with respect to this case. After all, here too, the saint is getting less than she deserves while the sinner is doing better than he deserves.

It may be objected that if the sinner is in a very miserable condition, we should help him rather than the saint although desert implies otherwise.²⁷ Thus, desert is not the only applicable reason but rather needs to be supplemented by a sufficientarian reason. One reply to this objection is to hold that no one deserves to be extremely miserable. Thus, desert itself entails that there is a reason to prefer sinners over saints in some cases. Accordingly, sufficiency is not necessary for explaining the conclusion that there is such a reason. Alternatively, a champion of desert may insist that if the sinner is extremely evil such that he deserves his miserable condition (below the sufficiency threshold), there is no reason to prefer him over the saint who has less than she deserves. Either way, Kagan’s argument seems to apply also to sufficiency.

This conclusion is not surprising once we notice that the most plausible versions of sufficiency are close to priority or equality. First, the more plausible versions of sufficiency do not oppose inequality only when one person is below the threshold and one above it, but also when both the worse off and the better off are below the threshold.²⁸ For it is unreasonable to hold that there is no reason to prefer a person who is far below the threshold over someone who is almost at the threshold. Indeed, a version of sufficiency that is only interested in whether people are below or above the threshold, regardless of their distance from the threshold, seems implausible.

Second, plausible versions of sufficiency also adopt a relatively high threshold. Indeed, the main motivation for a conception of sufficiency is that (distributive) justice should not be concerned with differences between people who are very well-off. Versions of sufficiency that include both of these elements are similar to priority and equality, for they incorporate equality or priority for most levels of wellbeing. Finally, there are various versions of priority, depending on how much the value of an additional unit of wellbeing increases where one’s level of wellbeing decreases. And some versions—those that attach a lot of weight to differences in the lower end and less weight to differences at the upper end—are very similar to sufficiency.

Thus, if Kagan’s argument is compelling regarding all versions of priority and equality, it seems arbitrary to claim that it does not extend also to sufficiency. If this is the case, Kagan’s argument implies, roughly, that desert should displace also sufficiency. This suggests, more generally, that desert should displace distributive justice entirely, as equality, priority and sufficiency are salient considerations in this context.

Even more radically, although Kagan himself explicitly denies this, we think that the essence of his argument extends also to utility.²⁹ This suggestion is typically not considered, perhaps because utility is often thought of as an impersonal value. Indeed, utility is associated with maximizing the size of the pie and not with the way it is divided. This may explain why discussions of utility typically do not mention desert³⁰ and specifically do not qualify the claim that (maximizing) utility is valuable with the condition that the relevant persons morally deserve more utility.³¹ However, if we take Kagan’s argument seriously, it is unclear whether anything of value is gained when we benefit a person who does not deserve it. Assume, for example, that we increase overall utility by 10% but all of the additional utility goes to a sinner whose level of utility is already higher than what he deserves. Is the resulting state of affairs indeed morally better? It seems that to the extent that Kagan’s argument is compelling, this is because there is nothing good in increasing the sinner’s wellbeing beyond what he deserves.

It may be insisted that utility is valuable in itself, independently of desert, so that it is in one respect better if the sinner gets more utility. Thus, according to this objection, if desert is also valuable, these values may clash as they indeed do in the case of the sinner. However, this view seems odd. While it is perfectly coherent to accept several independent values that may clash, for example, to hold that it is pro tanto good to increase the wellbeing of a person even if this is bad in terms of equality, it seems less plausible to hold that there is something good in increasing the wellbeing of a sinner who already fares better than he deserves. This is because while the value of equality is clearly independent of the value of utility, the idea of desert seems to entangle the appropriate level of a person’s wellbeing with her moral status.

Moreover, if we accept Kagan’s claim that it is implausible to accept both desert and priority, it raises a question in light of the relationship between desert and utility. Priority includes a component of utility in the sense that it assumes that an increase in wellbeing always has a positive value. It then adds that the degree to which each (additional) unit of utility is valuable depends also on the relevant person’s level of wellbeing. Thus, Kagan’s position is that desert is incompatible with the additional weight that is due to the fact that a person is worse off, but is compatible with the basic reason to promote a person’s wellbeing. Yet, it seems to us that the intuition that there is no reason to prefer the sinner who has more than he deserves over the saint who has less than he deserves, when the former is worse off, has considerable force also when the sinner and the saint are equally well-off but the sinner would gain slightly more utility than the saint from the relevant resource. We think, at the very least, that if one accepts the first claim (about priority) but not the second (about utility), an explanation is called for.

Moreover, as noted above, there are stronger and weaker versions of priority, depending on how much the value of an additional unit of wellbeing increases where one’s level of wellbeing decreases, and the weaker versions are very similar to utility. Since Kagan’s argument applies uniformly to all versions of priority, including the weaker ones, it seems odd to claim, as he does, that his argument applies to priority but not to utility.³²

Kagan’s appealing argument thus suggests very radical conclusions. It is therefore especially interesting to consider whether Kagan’s argument is sound. In what follows we consider what we believe is the most challenging objection to this argument and in favor of an intrinsic consideration of equality or priority.

4.1. The Argument from Cases in which Desert is Inapplicable

A powerful objection to the monistic view is based on the observation that desert is sometimes inapplicable. For example, desert is presumably inapplicable with regard to (some) non-human animals, young children, people who lack free will, and others (perhaps all of us) who are not morally responsible. Based on this observation, we may construct the following argument:

1. Desert is sometimes inapplicable, for example, in the above cases of non-human animals or young children.

2. In these cases, other intrinsic values, such as wellbeing or equality, exist (e.g., there is a reason to promote the wellbeing of babies, or to allocate scare resources among them equally, if there are no relevant differences between them).

3. Therefore, there are intrinsic values (e.g., wellbeing or equality) in addition to desert—hence, pluralism.

Since the premises of this argument are very compelling, it demonstrates that (assuming desert) we should be pluralists. However, this conclusion does not settle the more specific question that we consider in this paper about the “imperialism” of desert, namely, whether we should be pluralists when desert does apply. Indeed, the above argument does not address this question, and accordingly does not undermine the strong intuitions that support the monistic view that, when desert applies, it displaces other values such as wellbeing and equality.

However, while the above argument is not directly relevant to our question, a modified version of this argument is. Thus, given premises 1 and 2 of the argument, we may also infer:

• 3*.

    The best explanation for 1 and 2 is that wellbeing (or equality) is always intrinsically valuable, including in cases in which desert applies.

Indeed, once we notice that a certain value, such as wellbeing, is intrinsically valuable when desert does not apply, it seems odd that it would suddenly evaporate when desert applies.

Furthermore, some argue (following G. E. Moore) that intrinsic values are necessarily unconditional, since, by their very nature, they are valuable in themselves rather than in virtue of their extrinsic (relational) properties.³³ However, this view is controversial. Indeed, many hold that intrinsic values may be conditional. For example, some think that equality is intrinsically valuable except when it is achieved by leveling down.³⁴ Others hold that autonomy is intrinsically valuable only if exercised by choosing a morally permissible option.³⁵ Yet another common view maintains that pleasure is intrinsically valuable but not when it is sadistic (e.g., when it is derived from the pain of others).³⁶ Similarly, it may be argued that wellbeing is intrinsically valuable only when it is not undeserved (namely, on the condition that desert does not require more or less wellbeing).³⁷

At the end of the day, it seems that the above argument from cases in which desert is inapplicable demonstrates that monism involves (an additional) theoretical cost. Thus, the argument provides some support for the pluralist view. However, the force of the argument depends on the resolution of the general controversy about the intelligibility of conditional intrinsic values. And this in turn depends on the strength of the intuitions about specific cases, such as that of sadistic pleasures, which support the hypothesis that some intrinsic values are conditional (and indeed, more specifically, the monistic view that undeserved wellbeing is not valuable).

4.2. The Desert Monster

Another argument against the monistic view is based on the claim that its implications are sometimes implausible. Consider first a monistic view that assumes a simple accumulative version of absolute desert, according to which every good action entitles the agent to an additional amount of wellbeing that reflects the moral worth of that action.³⁸ Now, imagine two persons whose initial levels of wellbeing accurately reflect what they deserve. Assume further that both are decent persons who are, accordingly, reasonably well-off (say, each has 500 units of wellbeing). If one of these persons performs a good action that entitles her to an additional amount of wellbeing (say, 10 units), and there are no more available goods, we should allocate to her some goods at the expense of the other person, who did not perform such an action (although he could have), in order to equalize the distance of the two persons from what they absolutely deserve, i.e., their “peaks” (the new distribution should thus be 495:505).³⁹

This seems plausible. But now assume that the more virtuous person keeps on performing good deeds (which, however, do not increase the overall amount of goods in the world) say, she feeds starving sea turtles.⁴⁰ So we keep on transferring goods from the less virtuous to the more virtuous person. At a certain point (when the latter deserves 1500 units of wellbeing), equalizing the distance between what the persons have and what they absolutely deserve requires that the more virtuous get all the goods (1000 units of wellbeing) and the other get nothing. (Thus, the more virtuous person who deserves everything may be described as a “Desert Monster”).⁴¹ Intuitively, this result seems wrong. The fact that the less virtuous person is very badly-off seems to be something that should be taken into account at a certain point when deciding if additional transfers are just. This means that desert cannot be the only pertinent moral factor with regard to the distribution of goods: we should consider also the degree to which a person is worse (or well) off.

We reach similar results assuming comparative desert. There are several possible versions of comparative desert (e.g., Kagan 2003; Gordon-Solmon 2017). Consider first the “X-gap” view. This view requires that we equalize the distance between what the persons have and what they absolutely deserve, in terms of wellbeing, similarly to the absolute version in the case of limited resources.⁴² Thus, this version entails the same implication with regard to the above case (see Figure 3).

Figure 3. The X-gap view: comparative desert is satisfied since the difference between the levels of wellbeing of A and B is the same as the difference between their peaks.

A second option is the “Y-gap” view, which holds that when everyone has less than what they absolutely deserve, the drop down the Y axis in the desert graph (relative to the person’s peak) is the same for everyone (see Figure 4). In other words, this view requires that the gap in terms of goodness from the perspective of desert, between the optimal state of affairs where a person has exactly what she deserves and the actual share, is equal for everyone (Kagan 2003: 107; 2012: 390; Gordon-Solmon 2017: 376–377).

Figure 4. The Y-gap view: comparative desert is satisfied since the offense against noncomparative desert (i.e., the “Y gap”) is the same for A and B.

In order to implement this view, we need to determine the degree to which each allocation is good (or bad) from the perspective of desert. Kagan considers two options in this regard.

One option is that it is equally important that everyone get what they deserve, regardless of what they deserve (that is, all peaks have the same Y coordinates). The other option that Kagan (1998: 301; 2012: 98–107) considers as plausible is that desert graphs exhibit “bell-motion” regarding the peaks, namely, it is more important that good people get what they deserve compared to bad ones. Accordingly, it is worse if a virtuous person has a certain amount less than she deserves than if a less virtuous person has that same amount less than she deserves (and conversely with regard to the case in which they have more than they deserve).

Figure 5. Bell Motion: it is better to give the less virtuous (B) too little, by a given amount, than it would be to give the virtuous (A) too little, by that same amount—thus, the more virtuous a person is, the more steep her western slope; Likewise, it is better to give the more virtuous (A) too much, by a given amount, than it would be to give the less virtuous (B) too much, by that same amount—thus, the more virtuous a person is, the more gentle her eastern slope. Accordingly, for a given difference in levels of virtue between A and B, the difference in their levels of wellbeing is larger if we assume bell motion (the solid lines in the figure, whereas the dashed lines assume that there is no bell motion).

In the Desert Monster example, the first version has the same implication as the non-comparative version, namely, when the more virtuous person performed enough good actions so that she deserves 1500 units of wellbeing, while the other still deserves only 500 units, the former should get all the goods (1000 units of wellbeing) and the latter nothing. And the only difference with regard to the second version, is that we reach the same outcome even more quickly. For this version holds that it is more important to benefit a more deserving person. Accordingly, when both persons are equally distanced from what they deserve in terms of wellbeing, the loss of value (in terms of desert) is greater with regard to the more virtuous person (see Figure 5). Thus, in order to equalize the loss of value, this person should be closer to her peak and thus should get all the resources even sooner.

A third option is the “ratio” view which requires that, in the case of scarcity, each person has the same percentage of what she deserves (in terms of wellbeing).⁴³ In order to apply this view to the Desert Monster case, assume that for every good action, the agent is entitled for positive (additional) points that reflect the moral worth of that action.⁴⁴ In the case of scarcity, the relative share of a person is determined by her overall number of points divided by the total sum of points that the relevant persons are entitled to. Consider again the two persons from the previous case: in the initial state, the two are equally deserving and accordingly equally well-off (say, each has 50 points which entitle her to half of the available 1000 units of wellbeing, namely 500 units). If one of these persons now performs an action that is morally praiseworthy (that entitles her to, say, 2 more points), and there are no more available goods, we should transfer to her some goods from the other person (the new distribution of wellbeing should thus be approximately 510:490).⁴⁵ This may seem plausible. But now assume that the more virtuous person keeps on performing good deeds, which do not increase the overall amount of goods. So we keep on transferring goods from the less virtuous to the more virtuous person. Eventually, the former will be left with (almost) nothing.⁴⁶ Again, this seems inappropriate. Indeed, in this respect, the outcome of the ratio view is the same as the outcome of the former versions of comparative desert.

Our analysis thus suggests that in order to accommodate our intuitions concerning the above examples we need to appeal to an intrinsic consideration such as priority or equality in addition to the concern for desert. That is, Kagan’s monistic view should be rejected.

It may be thought that the best explanation for the intuitions in the Desert Monster case is not a concern for equality or priority but rather a concern for sufficiency, namely, that desert does not undermine the reason to aid people who are in a very miserable condition. We think, however, that this is not the case. First, we believe that the idea of sufficiency is theoretically suspicious in general,⁴⁷ and would thus reject the suggestion to accommodate the Desert Monster example by appealing to sufficiency instead of equality or priority. More importantly for our purposes here, we believe that (modified versions of the) Desert Monster example has a considerable intuitive appeal also if the worse-off (less virtuous) person is slightly above the sufficiency threshold. Assume that initially both persons flourish. Now, since the virtuous person begins performing good actions, she deserves more and more of the available resources. So that, if she receives what she deserves, at a certain point her situation becomes incredibly good—she literally enjoys eternal bliss—while the wellbeing of the less virtuous declines to a point just above the sufficiency threshold. This extreme inequality is arguably unjustified. Thus, we believe that the most plausible way to accommodate our Desert Monster case requires supplementing desert with a consideration of priority or equality and not (only) with sufficiency.

Moreover, as we have claimed above, we believe that the essence of Kagan’s argument extends also to sufficiency (and perhaps to distributive justice more generally). Therefore, even if sufficiency is the best explanation for the intuition in the Desert Monster case, this suggests that at least Kagan’s argument is flawed, and desert cannot displace distributive justice completely.⁴⁸

The above cases are irrelevant as counterexamples to a monistic view that rejects utility, since they assume that there is no loss of utility even if all resources are transferred to the desert monster. However, a different variation of these cases appears to be a counterexample also to monism regarding utility, namely, to the view that if we accept desert, we should reject utility as an independent value. Assume a variation of the Desert Monster example in which the virtuous person is much less efficient in terms of transforming resources to wellbeing. As she begins performing good actions, she deserves more and more of the available resources but the utility that she gains from these resources is very small while the level of wellbeing of the less virtuous person drops rapidly. Thus, at a certain point, the more virtuous person gains almost nothing from transferring more resources whereas the less virtuous person remains with almost nothing. This implication also seems implausible, and it appears that at least part of the problem is in terms of utility, namely, that one loses a lot while the other gains almost nothing.

4.3. The Conflicting Intuitions

It thus seems that different cases generate different intuitions. Kagan’s Twin Peaks case suggests that there is no place for equality or priority once we accept desert, whereas our Desert Monster case suggests that even if we accept desert there is also room for other considerations, such as equality, priority, and utility. The questions are therefore how to explain this intuitive difference and which intuition is more reliable.

One explanation for the different judgments may be related to the difference in the level of abstraction between the cases, which seems to affect the intuitions about the interaction between desert and other (potentially) morally relevant factors. Kagan considers an abstract case in which an unspecified, overall level of wellbeing is assumed to be appropriate in virtue of a person’s moral record. In his Twin Peaks case, it is stipulated that the worse-off person fares better than she deserves while the better-off person fares worse than she deserves. Kagan suggests that it is quite clear, intuitively, that there is no reason to prefer the worse-off in such a case. Indeed, as we have noted above, this judgment reflects the principle that a person should not get more wellbeing beyond what she deserves. Accordingly, inequality is not unjust when the relevant persons have exactly what they deserve.⁴⁹

However, if we decrease the level of abstraction a bit, as we did in our Desert Monster example, the intuitive judgment changes: there is a respect in which it would be better to prefer the worse-off over the better-off person in our example, even though the latter is more deserving. While our example is also quite abstract—we too do not specify exactly what the relevant desert base is (and hence the intuition in favor of the worse-off cannot be explained by the implausibility of the desert base)—our case is more concrete than Kagan’s in terms of how the appropriate level of wellbeing is determined from the perspective of desert alone. In particular, we add the plausible assumption that, from the perspective of desert alone, every virtuous action entitles the agent to some additional wellbeing (however small).⁵⁰ That is, the entitlement due to an additional good action is not discounted completely when the number of good actions the agent already performed becomes large. This additional assumption may explain the intuitive difference between Kagan’s case and ours.⁵¹

Interestingly, if we increase the level of abstraction beyond Kagan’s cases, the verdict also seems to be different than Kagan’s judgment. When we ask ourselves whether what a person should have is to be determined solely by facts about her moral record, the more intuitive answer seems to be “probably not.” Kagan’s way of framing the question therefore seems to generate a unique intuitive response—different from the ones generated by considering the question both more and less abstractly. The fact that formulating the question at different levels of abstraction brings about different intuitive judgments thus casts doubt on Kagan’s inference from the Twin Peak case to the denial of equality or priority as independent considerations.⁵²

Second, Kagan’s examples make it difficult to disentangle the concern for desert from other morally relevant factors and specifically from equality or priority. Clearly, when Kagan stipulates that a person deserves a certain unspecified level of wellbeing, he means that this is what she deserves in virtue only of her moral record, as opposed to what she deserves period (all-things-considered). However, it is not so easy to focus our intuitions in the required manner without a (de re) reference to a concrete level of wellbeing (which is appropriate from the point of view of desert alone).

In this respect, the intuitions regarding Kagan’s cases may be less reliable than the more abstract intuitions or the intuitions regarding more concrete cases. Let us demonstrate this worry with the following example. Suppose that we have two persons, where the virtue level of the first is 20% greater than the second. Since the principle (that the relevant distributive factor is the agents’ moral records) is an open formula, which leaves unspecified the translation from moral record into wellbeing, it is very difficult to determine, even roughly, what is reasonable from the point of view of desert alone. Moreover, even specifying exactly what each person should get won’t help. For example, we may be inclined to think that the first should get 20% more wellbeing. Still, this judgment may not reflect only their level of virtue but also other moral considerations, including their (absolute or relative) level of wellbeing. When you deliberate about how to allocate available goods in virtue of what someone morally deserves, you take account of facts such as that she is a human being, that she can feel pain and suffering, and that pain and suffering are prima facie bad. Yet these facts are relevant not only to desert but also to other considerations (such as priority and utility). Relatedly, if you intuitively judge that a certain worse off and less virtuous person should be preferred over another better off and more virtuous one, it is difficult to know if this is due to the fact that their (relative) shares do not properly reflect their moral record alone, or also to the greater force of the reason to aid someone who is worse off. Of course, we can abstract from these issues in a stipulatory way, for example, by assigning every action a certain amount of wellbeing that reflects (only) its moral worth, as we did in our example. But once we do so, as our example suggests, the view that the (comparative) moral record of persons is the only relevant distributive factor, seems much less plausible.

More generally, there seem to be two reasonable options. Either to disentangle desert from priority and equality—in which case the implications of desert are implausible unless they are mitigated by an additional consideration of priority or equality—or to incorporate priority or equality within desert. In either case, desert is not the only distributive factor.

In response, it may be argued that our Desert Monster case too is flawed as a counterexample to the monistic view according to which desert is the only distributive value. According to this objection, desert itself does not entail that the more virtuous person in our case should have everything and the less virtuous nothing, since from the point of view of desert, the worse-off person does not deserve so little.⁵³ This amounts to rejecting the assumption that every (additional) good action entitles the agent to an additional amount of wellbeing that reflects the moral worth of that action. For example, one may argue that there is an upper cutoff above which additional good actions do not entitle the agent to a higher level of wellbeing.⁵⁴ Yet it seems that, from the point of view of desert, the more plausible assumption is that a person who performs more good actions deserves more. Indeed, any cutoff—such that an additional good action does not entitle the agent to more wellbeing—seems not only arbitrary but also incompatible with the idea of desert.

This seems clear assuming that the desert base is actions. It may be objected that the desert base is character rather than actions, and that one’s character is not affected by one’s actual actions, but at most by the tendency to perform actions. Therefore, the objection continues, the assumption that a person (such as the desert monster) who performs more good actions deserves more wellbeing is false. However, if two people can perform a good deed at the same cost and only one of them does, then this must be (assuming all else is equal) due to a difference in their character. Therefore, we think that the assumption that the desert monster is entitled to more wellbeing for every additional good action is plausible also if the desert base is character rather than actions.

Indeed, the assumption that the more virtuous deserve more is a common assumption among those who endorse desert. For example, Kagan, after considering several alternatives in this regard, concludes that “it seems more plausible to hold that every increase in virtue will result in at least some increase in absolute desert. Which is to say: the more deserving deserve more, and the mapping function is always increasing” (2012: 253–254).

To sum up, it seems that the only way to accommodate the conclusion that the worse-off person does not deserve so little in the Desert Monster case is by considering, as part of the desert function, the degree to which a person is worse-off. And this amounts to abandoning the monistic view that desert is the only basic distributive consideration.