SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Preliminaries
  4. Is Sufficientarianism Grounded in a Personal Value?
  5. Is Sufficientarianism Compatible with the Person-Affecting View?
  6. Four Potential Sufficientarian Replies
  7. An Impersonal Sufficientarian Value?
  8. Conclusion

Telic sufficientarians hold that there is something special about a certain threshold level such that benefiting people below it, or raising them above it, makes an outcome better in at least one respect. The article investigates what fundamental value might ground that view. The aim is to demonstrate that sufficientarianism, at least on this telic version, is groundless and as such indefensible. The argument is advanced in three steps: first, it is shown that sufficientarianism cannot be grounded in a personal value. Neither, secondly, is it committed to the person-affecting view, the view that says that nothing can be better (worse) if there is no one for whom it is better (worse). This, in itself, is of interest because some sufficientarians reject egalitarianism precisely for its alleged incompatibility with the person-affecting view. Sufficientarians' disavowal of the person-affecting view implies that their view, similarly to egalitarianism (and, perhaps less famously, prioritarianism), must be anchored in some impersonal value. But crucially, and this is the third step of the argument, there is no apparent value that can fit that role. We must conclude, then, that telic sufficientarianism is groundless.

Here are two central views of telic sufficientarianism:

Strong Sufficientarianism: Benefits that lift individuals above some threshold level T matter more than equally large benefits that don't (whether they occur above T or below it).1

Weak Sufficientarianism: Any benefit below T, no matter how small, and no matter to how few individuals, outweighs any benefit above T, no matter how large, and no matter to how many individuals. Below T equally large benefits matter more the worse off the recipient is.2

I want to argue that both views are groundless, and therefore indefensible.

Sufficientarianism has received much critical attention, with objections roughly divided between those pointing to its counterintuitive recommendations (e.g. ‘threshold fetishism’, see below), and those criticizing the arbitrary location of its threshold.3 My account sympathizes with these criticisms but is independent of them. The objection offered here focuses, instead, on the value that purportedly grounds sufficientarianism. I want to claim that independently of its alleged counterintuitive and arbitrary nature, sufficientarianism is also simply groundless. A second respect in which my critique differs from most previous ones concerns its exclusive focus on telic, rather than deontic, accounts of sufficientarianism. Namely, I shall focus my criticism on that version of sufficientarianism that aspires to say something about the goodness of states of affairs. In contrast, my argument does not bear on deontic sufficientarianism, which I understand as a view about our duties of justice, or about what we owe to each other. Admittedly, it is probably the case that many self-professed sufficientarians advocate sufficientarianism as a deontic principle, say as a requirement of justice, rather than as a telic one.4 The argument in this article need not trouble them (at least not directly). But it should, if successful, trouble those holding (also) a telic version of the view.5 Having said that, I shall employ here a charitable reading and will attempt on occasion to glean a potential telic version of sufficientarianism from known deontic accounts.

If my argument proves successful it would establish that sufficientarianism is not a plausible alternative to other egalitarian patterns in assessing the goodness of states of affairs. This would be a limited but still significant finding. I advance this claim in three steps: first, I argue, sufficientarianism (both the weak and strong versions) cannot be grounded in a value that is merely personal. Neither, secondly, is sufficientarianism committed to the Person-Affecting View, the view that says that nothing can be better (worse) if there is no one for whom it is better (worse).6 This finding, in itself, is of interest because some sufficientarians, we shall see, reject egalitarianism precisely on those very grounds (that is, on the grounds that egalitarians must deny the person-affecting view). Whatever is the case, their forced disavowal of the Person-Affecting View (PAV) implies that sufficientarianism, just like egalitarianism (and prioritarianism, for that matter), must rely on some impersonal value. But crucially (and this is the third and last step of my argument), and unlike egalitarianism and prioritarianism, there is no apparent value that can fit that role. We must conclude, then, that sufficientarianism is groundless.

The article is arranged as follows. The first section sets the ground by making several clarifications about sufficientarianism and about the case I attempt to make against it. Among other things, it explains why it is correct to focus the discussion on the two above-mentioned views (as opposed to other potential formulations of the sufficientarian ideal). Section 2 then examines attempts to ground sufficientarianism, so understood, in a personal value. Having shown that none of these succeed, I then set out, in Section 3, to demonstrate that sufficientarianism must also deny the PAV. Section 4 rebuts four potential objections to that argument. Having shown that sufficientarianism must be impersonal all the way down, I move, in Section 5, to survey relevant impersonal values, concluding that none could fit the sufficientarian bill.

Preliminaries

  1. Top of page
  2. Abstract
  3. Preliminaries
  4. Is Sufficientarianism Grounded in a Personal Value?
  5. Is Sufficientarianism Compatible with the Person-Affecting View?
  6. Four Potential Sufficientarian Replies
  7. An Impersonal Sufficientarian Value?
  8. Conclusion

Let me start with several points of clarification about the case this article tries to make against telic sufficientarianism. First, my critique of sufficientarianism concerns its purported lack of underlying value. I shall therefore be looking at different sufficientarian formulations and speculate as to the value that may underlie them. What I do not assume is that the vast sufficientarian family must have one unifying value underlying it. Second, I said that critics sometimes object to sufficientarianism on the grounds that its specification of T (the threshold) is arbitrary. I want to stress that I shall not pursue that ‘indeterminacy’ objection here (at least not directly). In as much as I scrutinize T it is to question its underlying value, rather than the difficulty of drawing it.7 Third, in the examples that I shall use, 0 represent the line at which life is no longer worth living, and below 0 life becomes worth not living. 20, in turn, represents the level of the sufficientarian threshold (T), that is, the line above which life, according to sufficientarians, is ‘good enough’, however one may interpret that. Fourth, my concern in this article is with the pattern of sufficientarian distribution, and not with its currency. For that reason I shall merely assume, for the sake of argument, that utility is the currency under consideration. Some sufficientarians may restrict their principle to functionings or capabilities, and might not aspire to say anything about wellbeing more generally understood. My argument shall pose no threat to them. But equally, notice, accounts that discard overall wellbeing in favour of the narrower notion of capabilities lose the authority to speak of the overall goodness of states of affairs, which is what we are after here. Consequently, fifth, for the sake of argument, ‘equally large benefits’ (in both formulations above) will be measured in terms of utils. Sixth, I offer no theory of rationality in this article. I shall merely assume for the sake of argument the Von-Neumann-Morgenstern understanding of rationality as maximal expected utility.

Next, I should like to qualify the type of case that this article attempts to make against telic sufficientarianism. My claim, recall, is that the view is groundless. That is, I claim that there is no value that may underlie it. This requires saying something about the type of value we are looking for here. As mentioned, the value underlying sufficientarianism, or any other axiological view for that matter, may be either personal, or impersonal, or both. ‘Personal value’ does not require much elaboration. It simply denotes that we may identify a category of person(s) whose interests or wellbeing is tied up with the value in question. Impersonal value is trickier. Larry Temkin lists the following as examples of impersonal values: freedom, friendship, love, altruism, knowledge, perfection, beauty, rights, duty, truth, virtue, equality (or fairness), and desert.8 Importantly, all of these are not just impersonal values but what we may call fundamental impersonal values. That is, they are not reducible to some other value. We shall address these and others later on. But for now the important point to stress is that we search here for a fundamental impersonal value on which sufficientarianism (just like egalitarianism and prioritarianism, say) may rely on. So, to anticipate some of the points to come, saying that sufficientarianism is good because it strengthens democracy, say, is helpful but crucially incomplete. It is incomplete because democracy is not a fundamental impersonal value. Democracy might be (impersonally) good because it is just, or because it manifests equality, or because it is efficient, or because it reflects desert, etc. And what is true of democracy is true of other second-order values such as the elimination of oppression, poverty, and exploitation. Once again, all of these social goals are undoubtedly valuable. But they are not fundamental. Each of them is reducible to other (more fundamental) values (e.g. equality, desert, or the personal value of welfare).

My case against telic sufficientarianism is restricted to the two above-mentioned views (weak and strong sufficientarianism). Let me make a couple of points in justifying that restriction. Note, to begin with, that the two versions under consideration here, or, more accurately, the weaker one, represent quite a plausible and attractive account of sufficientarianism. This version, for example, is able to avoid the problem of ‘threshold fetishism’:9 benefits way below the threshold outweigh equally large threshold-crossing benefits to recipients who are better off. Weak sufficientarianism (at least as we understand it here) also avoids Thomas Scanlon's famous ‘transmission room’ objection: Aggregation of small (or even large, for that matter) benefits above T, no matter to how many millions of individuals, cannot outweigh benefits below it, no matter how small and to how few.10 These are usually considered attractive features of the view.

Next, notice that in my formulation of the two views above I have adopted a non-lexical version (‘equally large benefits’) with regard to strong sufficientarianism, but a lexical one (‘any benefit, no matter how large or small’) with regard to weak sufficientarianism. This calls for a quick explanation. The non-lexical version of strong sufficientarianism is the weaker, and thus more easily defended version of the view. I shall therefore attempt to refute it, thereby refuting also its more demanding, lexical version.11 Things are slightly more complex with regard to weak sufficientarianism. Here I have employed the stronger, lexical version. I have done so for the simple reason that non-lexical weak sufficientarianism is not a distinctly sufficientarian view. That non-lexical version of weak sufficientarianism would hold that it is better to bestow benefits below the threshold (whether or not they succeed in lifting individuals above T) compared to bestowing equally large benefits above the threshold. But this view is endorsed also by egalitarians and prioritarians. It is not a distinctly sufficientarian view.

I do not claim that the two views I consider here are the only sufficientarian versions worth exploring, all things considered. But I do think that the two versions as formulated entail the telic version of the most plausible sufficientarian views out there. And therefore if I succeed in refuting the two views under consideration I will have also refuted those other views.12

Is Sufficientarianism Grounded in a Personal Value?

  1. Top of page
  2. Abstract
  3. Preliminaries
  4. Is Sufficientarianism Grounded in a Personal Value?
  5. Is Sufficientarianism Compatible with the Person-Affecting View?
  6. Four Potential Sufficientarian Replies
  7. An Impersonal Sufficientarian Value?
  8. Conclusion

In this section I shall use the strong thesis merely as a focal point, but what I have to say applies to both versions. Here it is once again:

Strong Sufficientarianism: Benefits that lift individuals above T matter more than equally large benefits that don't.

Our question, recall, is how to make sense of this view. What value may underlie it? Why does it matter more to lift people above T? The reasons one can find in the sufficientarian literature suggest that the principle is motivated by such goals as the concern to eradicate deprivation and poverty, eliminate exploitation, strengthen democracy, secure the bases of self-respect, and so forth.13 These are all, we said, very worthy social goals, but as answers to our question they won't do, for they are not fundamental values. Alternatively, a typical sufficientarian statement would say that what is important is ‘the hunger of the hungry, the need of the needy, the suffering of the ill’.14 In that, sufficientarians invoke a satiable principle, namely one whose demands are satisfied at a certain threshold and no amount of further input will improve the result. But once again, we wish to know: what fundamental value is underlying that principle, thereby making it satiable in that particular way? So for starters: when sufficientarians say that it is better to lift people above T (compared to equally large benefits that don't), who, exactly, is this supposed to be better for?

In this section I shall survey four answers to that question. I should quickly say that I don't find any of these particularly promising to begin with. Still, the inquiry in this section is no straw man. The nature of the value underlying sufficientarianism has for the most part gone unexamined, and a thorough investigation is much in order. Surveying the following four options is useful if only to get them out of the way. Readers who happen to think that it is blindingly obvious that sufficientarianism cannot appeal (merely) to a personal value should skip to the next section.

Here, then, is the first answer, and it locates the goodness of sufficientarianism, surprisingly enough, with the allocator. Namely, it might be suggested that lifting people above T is good because it is good for the allocator. Something along these lines can be gleaned from one of the prominent sufficientarian accounts, namely Roger Crisp's. Crisp grounds his account in J. S. Mill's theory of moral sentiment, and the requirements of compassion.15 On that account, we (society or the individual allocator) have a stronger reason to respond to a person who is deprived and needy, than to a person who is not. In fact, Crisp uses compassion as a test to help draw that threshold (T is the line above which a person's request for help does not elicit compassion from an impartial spectator).16 Our reason for benefitting a person, then, seems to have something to do with the virtue of showing compassion.17 Now, it is hard to deny that showing compassion to those who are deprived is virtuous. Notice, however, that Crisp's account concerns the allocator and her reasons for action, quite independently of the recipient and her wellbeing. This makes it doubtful that the value in question can ground telic sufficientarianism. Just to take an obvious example, suppose that in Outcome X Smith lies below the threshold, whereas in Outcome Y he is lifted by an accident of nature over the threshold (suppose he has struck a vein of gold). There is no more virtue in Y compared to X, yet sufficientarians unanimously hold that Y is better than X. The moral betterment of the allocator cannot therefore explain what makes certain outcomes better on telic sufficientarianism (it might do so with regard to deontic sufficientarianism).

Here is the second not-very-promising answer, and it follows from the first. Lifting people out of deprivation is better, it might be suggested, because it is better for these individuals (i.e. for the recipients). There is some urgency in benefiting people who are below T, urgency that stems from the fact that benefits simply make a greater difference to a person who is poor and deprived compared to someone who is not. A hundred dollars, just to illustrate, normally makes a much greater impact on a homeless person compared to a wealthy person. This can be due to diminishing marginal utility. A benefit that lifts a person over T thus matters more to that person. But we can quickly see that this account of diminishing marginal utility cannot motivate sufficientarianism. Suppose you can hand two equal units of utility either to Jones at T-1, or to Smith at T+1. To make this even more concrete suppose (following Paul Weirich) that a spoonful of caviar would produce the same amount of utility for the wealthy caviar aficionado as it would for a hungry homeless person.18 Sufficientarians of every strand are committed to preferring the latter, even though the increase in personal utility is identical. Helping people who are below T cannot be better because it is better for these recipients (compared to other recipients who are above T, or even those who are better off than the former and yet still below T). This is worth stressing, if only because the point often goes unnoticed. Weirich's caviar example shows that sufficientarianism cannot be grounded in a personal value in so far as that value resides with the recipient. The same, incidentally, is true of prioritarianism. Prioritarians, also, hold that is better to give the caviar to Jones (at T-1) than it is to give to Smith (at T+1), even though doing so is not better for Jones (compared to how good it is for Smith). In that sense, prioritarianism, as well, must rely on an impersonal value.

Here is the third possibility. In the example just invoked, the spoonful of caviar would give homeless Jones and wealthy Smith identical amounts of satisfaction (or personal utility). Still, one reason why it might be better to give it to homeless Jones is that doing so would (also) yield indirect benefits to Smith. For example, he (Smith) no longer needs to worry about Jones's hunger, or be anxious about the prospect of Jones's breaking into his house (and stealing his stash of caviar), and so forth. Lifting people over T is good, on that account, because it is good for third parties. It might be, then, that when sufficientarians say that lifting individuals beyond a certain threshold of autonomy, or democratic capabilities say, is better, this is so whether or not it is better for these very recipients. Instead, it is simply better for the rest of us to live in a society where as many members as possible are autonomous, democratically capable, and so forth.

But we can rather quickly see that this cannot be the value motivating sufficientarianism. Consider the following three alternative outcomes:

A (17, 21) B (19, 23) C (21, 22)

(Recall that T is set at 20.) Sufficientarians (of all colours) are committed to holding that C is the best option, even though the welfare of those already above T has (somehow) decreased (from what it was in B, say). Concern for third parties cannot, therefore, motivate sufficientarianism.

The fourth and final answer is the suggestion that sufficientarianism is good for society as a whole. Something of this sort might be gleaned from one of the most intuitively appealing cases of (strong) sufficientarianism, namely triage in battle.19 Triage prioritizes those wounded soldiers whom medical treatment would lift over the threshold of battle-competency, and restore them back to action in a relatively short time. It does not, in contrast, prioritize those whose injuries are the most severe. It is motivated, therefore, not (or not merely) by the good of the treated soldier, but by the good of the group as a whole (be it the platoon, the army as a whole, or the warring nation, say).

But we can quickly dismiss that rationale as well. Consider a slight variation on the above example:

A (17, 21) B (19, 25) C (21, 22)

C, the option unanimously preferred by sufficientarians, contains less aggregate utility than B. Sufficientarianism therefore cannot be grounded in some utilitarian rationale of maximizing the good of the group (or the set).

Sufficientarianism, we see, is not grounded in the personal value to the allocator, to the recipient, to third parties, nor to the group as a whole.

Is Sufficientarianism Compatible with the Person-Affecting View?

  1. Top of page
  2. Abstract
  3. Preliminaries
  4. Is Sufficientarianism Grounded in a Personal Value?
  5. Is Sufficientarianism Compatible with the Person-Affecting View?
  6. Four Potential Sufficientarian Replies
  7. An Impersonal Sufficientarian Value?
  8. Conclusion

If benefitting individuals who are below T is better (than benefitting people above T, say), we are yet to see for whom this is so; if sufficientarianism is grounded in some personal value we have failed to see with whom that value resides. Sufficientarianism, it appears, must rely on an impersonal value. In this section I want to take this claim one step further and show that sufficientarianism must, in fact, deny the person-affecting view. The person-affecting view (PAV), recall, says that a state of affairs cannot be better (worse) if there is no one for whom it is better (worse).20

Notice that our findings in the previous section do not lend themselves to the conclusion that sufficientarians must deny the PAV. For it is prima facie possible that sufficientarianism is grounded in an impersonal value, and yet that the pattern it recommends is always better for someone. Now, finding out that sufficientarianism is incompatible with the person-affecting view would actually be quite surprising. For sufficientarians sometimes object to egalitarianism precisely for denying the person-affecting view.21 (Notice that I say ‘sometimes’ — this is by no means a necessary feature of all sufficientarians.) Egalitarians, allegedly, must do so in order to escape the so-called ‘levelling down objection’ (LDO), more on which below. I should perhaps quickly say that, for myself, I do not find the person-affecting view to be particularly compelling,22 nor levelling down to be much of an objection (least of all to egalitarianism). But the plausibility of the PAV (and the LDO) is not what stands to test here. I seek to show that sufficientarians must (contra some of its proponents' assertions) deny it, whether or not PAV itself is plausible.

Are sufficientarians, then, committed to the PAV? We must examine this question for both versions. Let us begin with strong sufficientarianism, which we understood as:

Strong Sufficientarianism: Benefits that lift individuals above T matter more than equally large benefits that don't.

Consider the following dilemma.23 A person, Agnes, suffers from some medical condition the exact nature of which it would take some months to ascertain. She is informed that she has an equal chance of occupying one of two scenarios. She might currently be at 0 or she might be at 11. (Recall that T, the level of ‘good enough’ life, is set at 20.) Whichever of the states she in fact occupies, Agnes can be given treatment that would improve her condition by exactly 10 units of utility. X and Y are the alternative treatments, where X is only effective if Agnes is indeed at 0, and Y is only effective if she is at 11. Here are the alternative states of the world that Agnes has equal chance occupying. (Notice that [RIGHTWARDS ARROW] signifies ‘can reach’):

Agnes's Dilemma

50% = 0 [RIGHTWARDS ARROW] X [RIGHTWARDS ARROW] 10

50% = 11 [RIGHTWARDS ARROW] Y [RIGHTWARDS ARROW] 21

Unfortunately, Agnes must decide in advance which treatment she prefers (suppose it is part of two different health-insurance packages she must choose between). Agnes, then, must choose between:

X (10, 11)

Y (21, 0)

(The two scenarios within each choice are equiprobable.)

The first thing to notice is that X and Y represent equally rational courses of action, at least on the understanding of rationality (Von-Neumann-Morgenstern) employed here: X and Y have equal expected personal utility. On this understanding of rationality, then, it would not be irrational for Agnes to opt for X. It is easy to see where this is going: it is perfectly rational for Agnes to opt for X whereas the (strong) sufficientarian allocator is forced, in contrast, to opt for Y. Y is the only course of action that has a chance of lifting Agnes over T. We have here, then, a case where sufficientarian recommendations can potentially clash with the perfectly rational preference of the recipient. Strong sufficientarianism, in short, violates here the person-affecting view. It says that a state of affairs (Y) is better, while in fact there is no one (not Agnes, nor anyone else we can think of) for whom it (state of affairs Y) is better.

Sufficientarianism, at least on the strong thesis, is forced to deny the person-affecting view. What about the weaker version? Here is how we understood weak sufficientarianism:

Weak sufficientarianism: Any benefit below T, no matter how small, and no matter to how few individuals, outweighs any benefit above T, no matter how large, and to how many. Below T equally large benefits matter more the worse off the recipient is.

Unlike strong sufficientarianism, weak sufficientarianism succeeds in avoiding the Agnes objection. Agnes's self interest and weak sufficientarianism are in tandem (0 is further below T than 11 is, and as such, doing X is of greater urgency). But weak sufficientarianism may be vulnerable to other dilemmas. Suppose another patient, Bob, is faced with a very similar dilemma to Agnes, only the options for him are:

Bob's Dilemma

50% = 19 [RIGHTWARDS ARROW] X [RIGHTWARDS ARROW] 21

50% = 21 [RIGHTWARDS ARROW] Y [RIGHTWARDS ARROW] 30

(T, recall, is set at 20.) As before, Bob should decide in advance which treatment he prefers to include in his healthcare package. He is faced, in other words, with the choice between X and Y:

X (21, 21)

Y (19, 30)

(Once again, within each choice, the two scenarios are equiprobable.)

Recall that the numbers, once again, represent utility gains, and so a unit of improvement has the exact same size irrespective of where it is located (including whether it is above or below T). Some very risk-averse individuals will no doubt choose X, but it is plausible to think that Y is in Bob's best interests. If asked to choose on Bob's behalf (suppose he is a minor, or in a coma), weak sufficientarians will be forced to opt for X, even though it is not apparently in Bob's best interests. Switching from the strong version to the weak one cannot therefore save sufficientarianism from repudiating the PAV.

Telic sufficientarianism, we must conclude, is incompatible with the PAV. To stress, this does not constitute an objection to telic sufficientarianism. A sufficientarian may very well bite the bullet and say, much like many egalitarians and prioritarians do, that sufficientarianism does make a state of affairs better, even when there is no one for whom it is better. Sufficientarianism, in that respect at least, does no worse than its two rivals.

Four Potential Sufficientarian Replies

  1. Top of page
  2. Abstract
  3. Preliminaries
  4. Is Sufficientarianism Grounded in a Personal Value?
  5. Is Sufficientarianism Compatible with the Person-Affecting View?
  6. Four Potential Sufficientarian Replies
  7. An Impersonal Sufficientarian Value?
  8. Conclusion

But sufficientarians might still want to resist that conclusion. There are at least four particular objections to the argument I just made. The first says that the argument is flawed because sufficientarianism does not, to begin with, apply in intra-personal dilemmas. The second says that even if sufficientarianism did apply to intra-personal dilemmas, it need not reach sufficientarian recommendations in those cases. The third objection says that even if sufficientarianism applied intra-personally, and even if its recommendations then were sufficientarian, still it need not apply to ex-ante dilemmas (but only to ex-post ones). And the fourth objection says that even if sufficientarianism did apply to intra-personal dilemmas and even if it did apply ex-ante, its recommendations in those ex-ante cases need not themselves be sufficientarian. Let us examine each in turn.

4.1.

First, then, one might claim that sufficientarianism, of whatever form, applies only in multi-person dilemmas, and never in intra-personal ones (such as the ones faced by Agnes and by Bob). On intra-personal dilemmas sufficientarianism, or so the claim goes, is simply silent. If this is so, my argument in the previous section proves nothing. A very similar case has recently been put by Andrew Williams with regard to prioritarianism:

[…] 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 life.24

The first thing to notice about this restrictive strategy (as we might call this response), and Williams's phrasing makes it easy to observe, is that it is much more plausible with regard to deontic, and less so with regard to telic, sufficientarianism. In thinking of what we owe to each other we may plausibly restrict a view, including sufficientarianism, to solving inter-personal dilemmas. But telic views, those purporting to say something about the goodness of states of affairs, do not lend themselves as easily to such restrictions. They must be able to say something also about the goodness of states of affairs pertaining to one single person. Moreover, recall that we have failed to find a personal value to which sufficientarians may appeal (think again, for one, of the caviar example). But crucially, a principle that is not grounded in a personal value cannot then be easily restricted to inter-personal dilemmas. To see this more clearly consider the following:

X (Clare) = 0 [RIGHTWARDS ARROW] 10

Y (Agnes) = 11 [RIGHTWARDS ARROW] 21

Strong sufficientarianism (just to illustrate) is clearly committed to preferring Agnes to Clare. But if Agnes is preferable to Clare, and for reasons that are not personal (that is, for impersonal reasons), then Y, in all likelihood, must be preferable to X even when obtaining for different histories of the same person (say, Agnes). (The same exercise can be carried out with regard to Bob and weak sufficientarianism.) It is unlikely that the impersonal value in question, assuming there is one, simply evaporates once other people are removed from the equation.

Since it must rely on an impersonal value, sufficientarians cannot, in all likelihood, simply choose to remain silent on intra-personal dilemmas. In that respect, the sufficientarian value is identical to the prioritarian one (but importantly different from the egalitarian one; see below). As Derek Parfit famously notes with regards to prioritarianism, there is value in helping Jones at level 2, which is greater compared to the value of an equal utility increment to Smith who is at level 4, whether or not Smith actually exits. (Think of his famous altitude example.)25 Precisely the same holds for sufficientarianism. Unless provided with a reason to think otherwise, we must assume that, being impersonal, sufficientarianism must apply also intra-personally. (Such a reason, notice, is provided by egalitarians. The impersonal value they are concerned with, namely equality, obtains, by definition, only in multi-personal dilemmas.)26 I am not aware of any such explanation in the sufficientarian literature (nor can I think of one on their behalf). Sufficientarians, both on the strong and weak thesis, must apply intra-personally, and if so, deny the PAV.27

4.2.

I now want to quickly say that not only can sufficientarians not restrict their views to inter-personal dilemmas, they must also pass identical verdicts whether the dilemma is inter-personal or intra-personal. To illustrate this consider Crisp's famous Beverly Hills example concerning the question of whether a spare bottle of expensive wine should go to Rich or to Super Rich.28 The sufficientarian answer to that question, says Crisp, is ‘don't care’. His reasoning, recall, is that above a certain level of wellbeing, an additional benefit simply does not make a (moral) difference. (This is presented as an advantage of sufficiency over equality and priority who both hold that bestowing the wine on Rich would make the outcome better, something which Crisp considers counterintuitive). That is all fine and familiar. But now the question is this: suppose it turns out that Rich and Super Rich are not, as we initially thought, two different individuals but in fact two instances of one and the same person. Can the sufficientarian vary her answer upon discovering that Rich and Super Rich represent different stages of one person's life (or alternative histories of that person's life)? Notice, the question in that intra-personal dilemma is not whether the extra wine bottle matters to that person. The sufficientarian need not deny that it does. But nor can she rely on this because her view, we already saw, does not turn on such personal value. The question, rather, is whether the extra wine bottle makes a moral (beyond merely prudential) difference. And crucially, it is hard to see how a sufficientarian can maintain that the wine bottle makes no moral difference in the inter-personal case, but does make a moral (and moreover, impersonal) difference in the intra-personal case. It is hard to see, in other words, how the sufficientarian might justify that shift in moral reasoning between inter- and intra-personal dilemmas. Justifying such a shift is not impossible, I concede, but it is genuinely puzzling how this can be done. The view is not incoherent, I concede, but it would be seriously under-motivated.

4.3.

Here is the third objection to our finding that sufficientarians must deny the PAV. Both the Agnes and Bob examples, it is clear, concerned comparing ex-ante dilemmas. They asked whether the sufficientarian allocator should do X or Y for Agnes (and Bob) under circumstances of uncertainty. Both cases did not involve comparing ex-post distributions, or, put simply, outcomes. Had we focused on ex-post distributions, the critic might say,29 sufficientarianism would once again be revealed to recommend verdicts that are plausible (it would clearly prefer 21 to 10, for example).

But can sufficientarians restrict their view to ex-post evaluations? I shall try to answer this question by focusing on the strong version of sufficientarianism, merely for the sake of convenience, but what I have to say applies also to weak sufficientarianism. Recall that we understood strong sufficientarianism as:

Benefits that lift individuals above T matter more than equally large benefits that don't.

The motivating idea here, obviously, is that lifting individuals above T is good, and is, moreover, better than benefitting them in ways that do not succeed in achieving such an end. Consider, then, the following dilemmas:

Case A

X: Helping Clare = 9 [RIGHTWARDS ARROW] 19

Y: Helping Agnes = 11 [RIGHTWARDS ARROW] 21

Here strong sufficientarianism clearly prefers Y. Now, when asked to justify this preference the strong sufficientarian reasons as follows. What is important, she says, is to secure as many transfers of individuals from below T to above it. In competing with Agnes, helping Clare, good as it is for her (the strong sufficientarian need not deny this) proves (relatively) wasteful. Helping Clare wastes 10 units that could have otherwise been used to lift a person above T. So far so good.

But now consider the following dilemma:

Case B

X: Helping Clare = 9 [RIGHTWARDS ARROW] 19

 0.01 probability: 9 [RIGHTWARDS ARROW] 19

Y: Helping Agnes = ____________

 0.99 probability: 11 [RIGHTWARDS ARROW] 21

This dilemma is identical to the one in Case A apart from the fact that here there is a minute risk that Agnes will not, in fact, end up above the threshold. If the objection under consideration was correct, and sufficientarianism indeed did not apply ex-ante, the strong sufficientarian must give here the verdict of: ‘don't care’. But this seems very odd indeed. We said that in Case A, strong sufficientarianism prefers helping Agnes because it must consider helping Clare to be wasteful. But how is such a judgment consistent with being neutral with regard to Case B? Isn't that view almost as wasteful as helping Clare in Case A? For a view so committed to lifting individuals above T — so much so that equally large benefits to worse off individuals (who cannot be lifted above T) wilt away in the presence of such an option (think of poor Clare in case A) — to remain neutral in Case B is simply implausible. Once again, it is not incoherent, I concede, but it is quite puzzling, and certainly under-motivated. Sufficientarians, to sum up the point, cannot, in all likelihood simply remain silent on ex-ante dilemmas. At the very least, they must provide a good reason for doing so. As things stand, such silence would be entirely at odds with the underlying motivation of their view.

4.4.

Even if the critic is persuaded by my response to the third objection, and concedes that sufficientarians cannot remain silent in ex-ante dilemmas, such as the ones facing Agnes and Bob, she may still resist my argument. For, she may maintain that the sufficientarian's recommendations in such cases need not themselves be sufficientarian. Sufficientarianism properly understood, the critic may say, is committed to evaluating ex-post outcomes in a sufficientarian way. True, it cannot remain silent on ex-ante dilemmas, but that is not to say it must ply its sufficientarian creed (also) in those ex-ante cases. Instead it can adopt any number of decision-making procedures.30 After all, utilitarians, some people maintain, can be utilitarians merely with regard to assessing outcomes, but adopt a non-utilitarian decision-making procedure for ex-ante dilemmas.31 This objection might be further bolstered once we notice that the person-affecting view, our focus in the previous section, concerns outcomes. Yet in assessing the Agnes and Bob dilemmas we were out of the realm of outcomes and into that of prospects. This is at odds, moreover, with the restriction of my discussion to telic, and not deontic, sufficientarianism (thus focusing on outcomes and not actions). So perhaps my argument purporting to show sufficientarians' rejection of the PAV simply misses the point.

Since my focus here is on sufficientarianism in its telic form I must concede that, formally at least, ex-post sufficientarianism indeed need not be committed to ex-ante sufficientarianism. The combined view is not incoherent. But, once again, it is hard to see how sufficientarians can motivate a view other than sufficientarianism in addressing ex-ante dilemmas. True, one can accept that ‘providing an account of right-making characteristics still is not the same thing […] as providing a decision-making procedure’.32 But it is hard to see how ex-post sufficientarians can simply reinvent themselves as egalitarians, prioritarians, utilitarians (or what have you) once faced with a dose of uncertainty. This is true for sufficiency, but also for all other axiological views (e.g. equality, priority). Axiological views concern the goodness of outcomes. Each view obviously differs with regard to the value it seeks to maximize in those outcomes (be it equality, priority, sufficiency, or utility). But the introduction of uncertainty cannot be a license for forsaking the commitment to that value. In the absence of certainty, telic views simply tell us to pursue the course of action most likely to maximize their favoured value. As Parfit writes: ‘[…] 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 are likely to be better’.33 Ex-ante sufficientarianism can therefore be nothing other than the edict: ‘do what you can to maximize the prospects of whatever it is you hold dear’ (namely, lifting as many people above T, or improving as much as possible to condition of those under T). Recall Case A and Case B just invoked (concerning Agnes and Clare). It is hard to imagine how the introduction of a minute risk can allow sufficientarians to jettison the underlying logic of their view (in this case, lift as many individuals as possible above the threshold at the smallest possible cost). Again, one need not deny that an axiological assessment of outcomes is different from a decision-making procedure. But in devising such decision-making procedures the strong sufficientarian (to take one of the two views) cannot simply abandon her underlying motivation (to lift as many individuals above T with as little waste of resources). At the very least, the burden of proof shifts to the sufficientarian to provide a justification for such a dramatic shift.

An Impersonal Sufficientarian Value?

  1. Top of page
  2. Abstract
  3. Preliminaries
  4. Is Sufficientarianism Grounded in a Personal Value?
  5. Is Sufficientarianism Compatible with the Person-Affecting View?
  6. Four Potential Sufficientarian Replies
  7. An Impersonal Sufficientarian Value?
  8. Conclusion

Sufficientarians must reject the person-affecting view. Should this be embarrassing for them? Not necessarily, I indicated. For egalitarians and prioritarians are challenged in precisely the same way. Egalitarians famously recommend levelling down despite the fact that doing so would be in nobody's interest. Less famously, perhaps, prioritarians recommend alleviating the worse off option in intra-personal dilemmas, even when doing so is not in the best interests of that recipient.34 In recommending a course of action that is not in the recipient's best interests (nor anyone else's, for that matter), sufficientarianism, admittedly, does no worse than its two rivals.

It is no embarrassment for a distributive view to recommend a course of action that is in nobody's interest. But it is potentially embarrassing for a view to recommend such a course of action without being able to provide a justification for it. Sufficientarians, in other words, must be able to explain why a particular outcome or prospect is better if there is no one for whom it is so. Put differently, they must be able to point to an impersonal value on which to hang such a view. Now, whatever else one might think of them, egalitarianism and prioritarianism do provide such accounts. Egalitarians can (and do) hold that equality represents an impersonal value, independently of the personal good it may bring to individuals. Consequently, while gouging the eyes of the sighted in the name of equality with the blind is (probably) in nobody's interests, it nevertheless makes an outcome, in one (impersonal) respect, better.35 (Notice that I am not asking you to accept the content of this argument, but only its structure.) Prioritarians, similarly, can (and in fact, I would maintain, must) hold that a unit of utility has an impersonal value that increases, the worse off its recipient is. That value, crucially, is independent of the interests of the recipient (or anyone else). Egalitarianism and prioritarianism, then, have impersonal values that they can fall back on, which is precisely that which we are asking with regard to sufficientarianism. Still, before answering that question we should be mindful not to hold sufficientarianism to a higher standard of investigation than that which we apply to its two rivals. Since I cannot test, in this article, the impersonal values grounding egalitarianism and prioritarianism, I will also not ask the same of sufficientarianism. In other words, we are interested merely in whether there is some minimally plausible fundamental value sufficientarianism can fall back on, without probing it too deeply.

Reflecting back on some of the things already mentioned, we can say that alleviating destitution, oppression, and poverty may bestow on a given benefit (or utility increment) an added impersonal value. But what value, precisely? Which of the fundamental impersonal values we can think of may account for that sufficientarian value? In Section 1 we listed the following impersonal values: freedom, friendship, love, altruism, knowledge, perfection, beauty, rights, duty, truth, virtue, equality (or fairness), desert. And to these we may add also autonomy and priority-weighted utility (the increasing urgency of benefiting individuals the worse off they are, a.k.a the prioritarian value). Of all these, the last five seem possible candidates for grounding sufficientarianism. I don't know if this counts as an exhaustive list, but it is hard to think of any other value that might be relevant here. Let us examine those that are.

We may first dismiss virtue. It is hard to imagine why practicing sufficientarianism should count as a virtue if doing so is better for no one. For Bob's guardian (just to give one example) to prefer X (21, 21) to Y (19, 30) does not seem particularly virtuous. True, the guardian may say to Bob: ‘I did this because I wanted to secure for you a level of welfare that is good enough. I did this for you!’ This is certainly no vice. But it is also not particularly virtuous, certainly compared to the other, more rational (in terms of expected-utility) option of (19, 30). We can also quickly dismiss desert as the value in question, for sufficientarianism appears to be independent of desert.36 Indeed, sufficientarians are often hostile to desert.37 Autonomy could perhaps be that value, on the ground that a life of destitution, not to mention oppression and domination, lacks the necessary conditions for autonomy.38 This seems initially attractive. We do think that being under T compromises one's autonomy and ability to escape exploitation and domination. Recall, however, that we are searching here for an impersonal value. The difficulty with autonomy in that respect, then, is that it is not quite clear what its impersonal aspect might be. Why does it matter that someone is autonomous in a way that is divorced of that (or any other) person's wellbeing? (Of course, the more autonomous our fellow citizens are the better the prospects for democracy and civil society. But we have already ruled out benefits to third parties as motivating sufficientarianism.) A decisive problem, then, with grounding sufficientarianism in autonomy is that it is underspecified. It is unclear that autonomy has any impersonal value.

The fourth impersonal value that comes to mind, we said, is the familiar prioritarian value, namely the greater weight to benefiting people the worse off, in absolute terms, they are. That could, indeed, be the impersonal value underlying sufficientarianism. But in order to offer a distinctive sufficientarian account, the added value represented by T must be accounted for somehow. Simply holding that benefits matter more the worse off the recipient is motivates prioritarianism, not sufficientarianism.

This leaves us with the final candidate (I can think of), namely equality or fairness. Benefiting people below T, and certainly raising them over T, might be valuable because it is what fairness requires. A sufficientarian might hold, for example, that our duties of fairness simply taper off (or evaporate) above T. (Recall Crisp's somewhat related definition of T as the line above which no compassion is elicited.) Notice, however, that this strategy makes telic sufficientarianism rely on some pre-existing deontic claim it makes about fairness. This risks circularity. On the present strategy, sufficientarians must tell us why fairness has something to do with T (before they go on to tell us that it is more valuable to help individuals the lower they are below it). Crucially, in providing that fairness account the sufficientarian cannot simply restate the familiar accounts we already encountered. That is to say, she cannot simply say that helping people who are under T is fair because it lifts them out of deprivation, oppression, etc. That rationale, on its own, can no longer do. We need to know why it is fairer to do so. To rehearse some of what has already been said: it cannot be fairer to do so because it brings these people greater personal benefit, or because doing so is a prerequisite to a functioning democracy. We already saw that such person-affecting rationales cannot motivate sufficientarianism.

Appealing to the impersonal value of fairness, then, simply shifts around the sufficientarian burden of explanation. Such an appeal cannot, on its own, account for any alleged impersonal value sufficientarianism can fall back on. What that value could be still remains a mystery.

Conclusion

  1. Top of page
  2. Abstract
  3. Preliminaries
  4. Is Sufficientarianism Grounded in a Personal Value?
  5. Is Sufficientarianism Compatible with the Person-Affecting View?
  6. Four Potential Sufficientarian Replies
  7. An Impersonal Sufficientarian Value?
  8. Conclusion

Telic sufficientarianism, as far as our investigation can tell, is groundless. It is not grounded in a personal value. Neither, contra the assumption of some of its proponents, is it committed to the person-affecting view. Nor, finally, is there an impersonal value to which sufficientarians may appeal. I should end, though, with a word of caution and concession. Since our inquiry has been restricted to telic sufficientarianism I cannot claim to have refuted deontic sufficientarianism (as I do think I have done with regard to telic sufficientarianism). Doing so remains a task for another day.39

Notes
  1. 1

    Harry Frankfurt, ‘Equality as a moral ideal’, Ethics 98,1 (1987): 2143, at p. 31. Also: ‘Those below a utility threshold are not necessarily benefited by additional resources that move them closer to the threshold. What is crucial for them is to attain the threshold. Merely moving closer to it either may fail to help them or may be disadvantageous’ (p. 32).

  2. 2

    See Roger Crisp, ‘Equality, priority, and compassion’, Ethics 113,4 (2003): 745763 ; Liam Shields, ‘The prospects for sufficientarianism’, Utilitas 24, 1 (2012): 101117 ; Robert Huseby, ‘Sufficientarianism: Restated and defended’, The Journal of Political Philosophy 18,2 (2010): 178197, at p. 180; Yitzhak Benbaji, ‘Sufficiency or priority?European Journal of Philosophy 14,3 (2006): 327348, at p. 329.

  3. 3

    On the arbitrary nature of the sufficientarian threshold see Paula Casal, ‘Why sufficiency is not enough’, Ethics 117,2 (2007): 296326, at p. 312ff; Richard J. Arneson, ‘Perfectionism and politics’, Ethics 111,1 (2000): 7793, at p. 56; Why justice requires transfers to offset income and welfare inequalities’, Social Philosophy and Policy 19,1 (2002): 199200, at p. 185; Democratic equality and relating as equals’, Canadian Journal of Philosophy 40, sup. (2010): 2552, at p. 32; Dale Dorsey, ‘Equality-tempered prioritarianism’, Politics, Philosophy, & Economics (2013), online first, doi: 10.1177/1470594X13483479, p. 6 . Dorsey says that sufficientarianism ‘lacks rationale’, and in that sense comes very close to the criticism that I want to level. But his argument is really one about arbitrariness.

  4. 4

    Martha Nussbaum, for example, writes, ‘moving all citizens above a basic threshold … should be taken as a central social goal’: ‘Women and cultural universals’, in her Sex and Social Justice (Cambridge: Cambridge University Press, 1999), p. 43 . Cited in Dale Dorsey, ‘Toward a theory of the basic minimum’, Politics, Philosophy, & Economics 7,4 (2008): 423445, at p. 423.

  5. 5

    Dorsey writes: ‘I believe that the most perspicuous way of characterizing the basic minimum is as a feature of a teleological account of political morality’ (ibid., p. 437). According to Huseby, also, sufficientarianism is primarily a telic view ( Huseby op. cit., p. 179). Other telic sufficientarian accounts include Roger Crisp's and Itzhak Benbaji's. See Crisp 2003a op. cit.; Yitzhak Benbaji, ‘The doctrine of sufficiency: A defence’, Utilitas 17,3 (2005): 310332 .

  6. 6

    On the person-affecting view, see Derek Parfit, ‘Equality and priority’, Ratio 10,3 (1997): 202221, at p. 219. Larry Temkin famously calls this ‘the slogan’ in Inequality (New York: Oxford University Press, 1993), p. 256 .

  7. 7

    This also means that I shall leave aside the question of how the threshold of sufficiency is be drawn, including whether its delineation entails only absolute or also relative factors. On the latter see, Elizabeth Anderson, ‘Fair opportunity in education: A democratic equality perspective’, Ethics 117,4 (2007): 595622 ; Debra Satz, ‘Equality, adequacy, and education for citizenship’, Ethics 117,4 (2007): 623648 ; Ideals of egalitarianism and sufficiency in global justice’, Canadian Journal of Philosophy 36, supp. (2010): 5371 .

  8. 8

    Larry S. Temkin, ‘Egalitarianism defended’, Ethics 113,4 (2003): 764782, at p. 779.

  9. 9

    Arneson 2002 op. cit., p. 188 ; Arneson 2000, op. cit., p. 56 ; John E. Roemer, ‘Eclectic distributional ethics’, Politics, Philosophy, & Economics 3,3 (2004): 267281, p. 279.

  10. 10

    Crisp 2003a op. cit., p. 754 ; Casal op. cit., pp. 319320 .

  11. 11

    An example for someone who advocates lexical strong sufficientarianism is Dale Dorsey: ‘When rank-ordering states of affairs, the state of affairs with more rather than fewer individuals obtaining the basic minimum is, no matter the arrangements below and above the minimum, better’: Dorsey op. cit., p. 437 . The original quotation ends with the word ‘worse’, but this is surely a typo. Also, although Dorsey says that this version is a bit too strong for his taste, and that he prefers a non-lexical version, he never provides the latter, as far as I can see.

  12. 12

    I can think of at least two other versions that are prominent in the literature, Yitzhak Benbaji's and Liam Shield's. Yitzhak Benbaji offers the following as his preferred sufficientarian thesis:

    Benefitting a person is of special moral importance only if she is badly off. When the distribution in question is among people who are on the same side of the priority line, the purely utilitarian solution of the distributive dilemma should prevail ( Benbaji 2005 op. cit., p. 312 ).

    Notice two features of this principle. First, it shares the following similarity with weak sufficientarianism as we have understood it above. It holds that benefitting recipients below T outweighs any benefits to people above it. This is the only way to make telic sense of the claim that ‘it is of special moral importance’ to benefit people who are badly off, and moreover in a way that is distinctively sufficientarian rather than prioritarian. Second, the second part of Benbaji's formulation implies, effectively, that it is indifferent with regards to dilemmas involving equal amounts of utility on either side of T. So, if we were to give Benbaji's view a telic dressing, it would read:

    Any benefit below T, no matter how small, and no matter to how few individuals, outweighs any benefit above T, no matter how large, and no matter to how many individuals. Below T as well as above it, it does not matter where equally large units of utility are placed.

    This view is a distinctly sufficientarian view, and moreover, it is distinct from the two views I formulated above. However, Benbaji's, we can see, is a special case, and a more stringent one, of Weak sufficientarianism. It is more stringent in that it sees no added value, in dilemmas that occur either below or above T, in placing benefits with the worse off. As such, and this is the crucial point, it is for all intents and purposes entailed in weak sufficientarianism. If I refute the latter I will have also refuted Benbaji's principle, which is the more demanding one. Here is the other prominent view that might seem to diverge from the two views under consideration. Liam Shields offers the following version:

    Once people have secured enough there is a discontinuity in the rate of change of the marginal weight of our reasons to benefit them further ( Shields op. cit., p. 108 ).

    Notice that our inquiry is restricted to telic sufficientarianism. Arguments about our reasons to benefit recipients appeal to deontic notions of sufficiency, and are as such irrelevant for a telic assessment of sufficientarianism. So think, in contrast, of the following formulation: ‘Once people have secured enough, the remaining benefits ought to be distributed equally (or to the worse off, or in a way that maximizes utility)’. This (modular) version, notice, is distinctively sufficientarian. In fact, it is compatible with weak sufficientarianism as I have defined it above. It therefore falls under our purview here. Anything I shall say about weak sufficientarianism shall therefore apply to this version as well.

  13. 13

    Elizabeth Anderson, ‘What is the point of equality?Ethics 109,2 (1999): 287337 ; Madison Powers & Roth Faden, Social Justice: The Moral Foundations of Public Health and Health Policy (Oxford: Oxford University Press, 2006); Shields op. cit., pp. 115116 .

  14. 14

    Joseph Raz, The Morality of Freedom (Oxford: Oxford University Press, 1986), p. 240 .

  15. 15

    Crisp 2003a op. cit.

  16. 16

    Ibid., p. 762 .

  17. 17

    Notice, this is only one possible interpretation of Crisp's account, and he may, in contrast, intend ‘compassion’ to serve only to delineate T, not to ground it. Still, notice, Crisp would, in that case, have to offer some other rationale for his sufficientarian view. We shall soon see that this is particularly tricky for Crisp for he implicitly endorses the person-affecting view. Also, I doubt that compassion can be used merely to draw T, while keeping silent on the question of the value that grounds sufficientarianism to begin with. In other words, if compassion does not motivate the view that benefitting people below T matters more than benefitting people above it, it is hard to see why it (compassion) is helpful in drawing T to begin with. If ‘compassion’ is helpful in drawing T because it accompanies (or tends to accompany) some other moral consideration (say, justice), then surely it is that other moral consideration that we should rely on in the first place. This might be indicated in Crisp's reply to Temkin. There, Crisp seems to give the game away by saying that he does not intend compassion, pure and simple, to be doing the work of drawing T, but rather a ‘modified form of compassion that tracks just claims’: Roger Crisp, ‘Egalitarianism and compassion’, Ethics 114,1 (2003b): 119126, at p. 119 (emphasis added).

  18. 18

    Paul Weirich, ‘Utility tempered with equality’, Noûs 17,3 (1983): 423439 . See also Campbell Brown, ‘Priority or sufficiency … or both?Economics & Philosophy 21,2 (2005): 199220, at p. 199.

  19. 19

    See Arneson 2002 op. cit., p. 185 ; Roemer op. cit., p. 273 .

  20. 20

    The person-affecting view may come in narrow and wide versions. On the narrow reading, an outcome cannot be better if it is not better for any of the persons in question. On the wide version, an outcome cannot be better in ways that are independent of people's welfare, more generally. [See Nils Holtug, Persons, Interests, and Justice (Oxford: Oxford University Press, 2010), pp. 158161 .] The move from the narrow to the wide version, it is easy to see, helps save the person-affecting view from cases involving non-identity. Since my argument shall not tread on non-identity, it should hold for both the narrow and wide versions of the view.

  21. 21

    Crisp 2003b op. cit., pp. 124126 . Although he does not explicitly invoke the person-affecting view, Benbaji's case for sufficiency and against equality begins and ends with the levelling down objection. See for example, Benbaji 2005 op. cit., p. 332 .

  22. 22

    I happen to be convinced by Temkin's refutation of it. See Temkin 1993 op. cit., chapter 9.

  23. 23

    I am inspired here, it is evident, by Michael Otsuka and Alex Voorhoeve's somewhat similar objection to prioritarianism. See ‘Why it matters that some are worse off than others: An argument against the priority view’, Philosophy and Public Affairs 37,2 (2009): 171199 .

  24. 24

    Andrew Williams, ‘The priority view bites the dust?Utilitas 24,3 (2012): 315331, at p. 323. This objection was anticipated in Otsuka & Voorhoeve op. cit., pp. 188190 .

  25. 25

    Parfit op. cit., p. 214 .

  26. 26

    ‘Equality is a relationship between the lives of different people’: Dennis McKerlie, ‘Equality and priority’, Utilitas 6,1 (1994): 2542, at p. 25. Some people may think that equality considerations obtain also in assessing different time slices of one person's history, but I think this is confused. [See also Dennis McKerlie, Justice between the Young and the Old (New York: Oxford University Press, 2013), p. 128 .] I cannot, however, hope to do justice to that issue here.

  27. 27

    Why can't the sufficientarian simply say, someone might respond, that she gives weight not only to sufficiency but also to considerations of autonomy, say? She could respond to me by saying that in Bob's case (say), sufficientarianism does indeed require opting for X (21, 21). But that respect for Bob's autonomy would imply opting for Y (19, 30). Sufficientarianism, in other words, has elbow-room also for other values, such as autonomy. There is nothing wrong with this value-pluralist strategy in reply to my Bob (and Agnes) argument. Indeed, egalitarians (and potentially also prioritarians) adopt the same strategy all the time (say, in reply to the levelling down objection). The reply, in short, shows why the argument does not succeed in demonstrating sufficientarianism to be counter-intuitive, all things considered. But crucially, that was not at all my complaint (in the previous section). My claim is not that sufficientarianism is counterintuitive; only that it denies the person-affecting view.

  28. 28

    Crisp 2003a op. cit., pp. 755ff.

  29. 29

    I am grateful to Nils Holtug for putting this objection to me.

  30. 30

    This objection was put to me by Nils Holtug.

  31. 31

    See R. E. Bales, ‘Act utilitarianism: Account of right-making characteristics or decision-making procedure?American Philosophical Quarterly 8, 3 (1971): 257265 .

  32. 32

    Ibid., p. 261 .

  33. 33

    Derek Parfit, ‘Another defence of the priority view’, Utilitas 24,3 (2012): 399440, at p. 415.

  34. 34

    Otsuka & Voorhoeve op. cit. The challenge appears even earlier in McKerlie 1994 op. cit., p. 42 . See also Ingmar Persson, ‘Equality, priority, and person-affecting value’, Ethical Theory & Moral Practice 4,1 (2001): 2339 .

  35. 35

    Temkin 1993 op. cit.; Temkin 2003 op. cit.; Ingmar Persson, ‘Prioritarianism and welfare reductions’, The Journal of Applied Philosophy 29,4 (2012): 289301, at p. 291.

  36. 36

    That is not to say, of course, that the two cannot be combined. Casal espouses ‘luck egalitarian sufficientarianism’, for example ( Casal op. cit., p. 321ff. See also Arneson 2002 op. cit., p. 192 ). The point rather is that sufficientarianism at its core is independent of desert; not that it cannot be coupled with it.

  37. 37

    Anderson 1999 op. cit. Crisp says of desert that it is ‘deeply problematic’: Crisp 2003b op. cit., p. 125 . See also Casal op. cit., pp. 300301, on this.

  38. 38

    This seems to be Anderson's emphasis. See also Arthur Ripstein, Equality, Responsibility, and the Law (Cambridge: Cambridge University Press, 1999), chapter 9; Dorsey 2008 op. cit.

  39. 39

    I have benefitted from stimulating feedback by the participants of the Equality of Opportunity workshops in Jerusalem and Copenhagen, and the Society of Applied Philosophy Annual Conference in Zurich, where I have presented earlier drafts of this paper. I am especially grateful to David Enoch, Nir Eyal, Alon Harel, Iwao Hirose, Nils Holtug, Kasper Lippert-Rasmussen, and Alex Voorhoeve for very useful written comments.