LEAP 5 (2017) Shielding Sufficientarianism from the Shift1 L A S SE N I E L SE N University of Southern Denmark ABSTRACT This paper discusses Liam Shields’ sufficientarianism and especially his very innovative construction of the Shift Thesis: that above the relevant threshold there is a significant change in our reasons to benefit people further. The paper argues that, despite its clear advantages, Shields’ view still faces some general problems. First, that it says too little about how different types of reasons to benefit someone should be weighed against each other. Second, and more importantly, that Shields does not provide satisfactory reasons for why we need the Shift in the first place. The paper argues that given the value assumptions that sufficientarians normally adhere to, the upper limit version remains a more promising alternative. Keywords: Shields; sufficientarianism; pluralism; the shift thesis Liam Shields’ development of the sufficiency view – the view that justice is concerned with securing enough for everyone – is among the most promising outlines for a theory of distributive justice in contemporary political philosophy. Shields’ rewritings of the sufficiency principle have, since their origin in 2012, gained much attention and many philosophers and political theorists have found them to improve the general outlook of sufficientarianism. Although Shields speaks into a field of great complexity, the key contribution is utterly simple. In a nutshell, Shields’ main point of argument is that sufficientarianism need not imply that we should ignore inequalities once everyone has “enough”. Instead, he argues, sufficiency implies merely that there is a significant shift in our reasons to benefit people further. This development has now – true to Shields’ own wording – become known as “the Shift Thesis” (2016; 2012). The Shift Thesis effectively offers a very appealing sufficientarian reply 1 For useful comments on earlier versions of this paper, I thank A xel Gosseries, David A xelsen, Anca Gheaus, Danielle Zwarthoed, Pierre-Etienne Vandamme, Chris Mills, Liam Shields, and two anonymous reviewers. D OI : 10. 310 0 9/L E A P. 2017.V 5.12 Shielding Sufficientarianism from the Shift 143 LEAP 5 (2017) to the critics’ concern about how justice applies to situations where inequalities persist but where no one is below the threshold. Thus, the advantages of Shields’ sufficientarianism are obvious. However, the view is not without its limitations. In this paper, I raise some critical questions for Shields’ sufficientarianism and I defend the “upper limit” sufficiency view as a more promising framework. The paper proceeds as follows. Section 1 lays out sufficientarianism generically. Section 2 presents Liam Shields’ amendment to this view in the form of his Shift Thesis. Section 3 raises some critical questions for Shields’ version of the sufficiency view that I believe he needs to answer. Section 4 defends upper limit sufficientarianism as a more plausible version of sufficientarianism than Shields’ account. Section 5 concludes. 1. SUFFICIENTARIANISM Sufficientarians care about individual people’s absolute standing. They do not care about people’s relative standing unless it affects their absolute standing (Axelsen and Nielsen 2015). Here I have no space to unfold this idea, but one plausible way to understand it is to say that sufficientarians use a threshold constraint to discriminate between relevant and irrelevant individual demands of justice (Segall 2016; Hirose 2016)– e.g. similar to Scanlon’s objective criterion for distinguishing between urgent and non- urgent preferences (1975). Without distinguishing between different theoretical specifications within the sufficientarian literature, we can assume the following generic principle (adopted from Nielsen 2017): The generic sufficiency view Justice is concerned with eliminating absolute deficiencies rather than inequalities This generic formulation captures the driving moral statement of any specified sufficiency view , and although critics are sceptical, many find it intuitively plausible.2 However, despite the merits of the generic view, sufficientarians might need to say more about how to set the threshold in order to render the sufficiency view theoretically plausible. This is because all sufficiency 2 See among others Frankfurt (1987), Crisp (2003), Raz (1986: 240); Benbaji (2005) Huseby (2010). 144 Lasse Nielsen LEAP 5 (2017) views – including the generic view – imply that there exists, at least in theory, a threshold point above which inequalities are irrelevant (or significantly less relevant) to justice. Thus, for example, sufficiency views, even when very generic, are always vulnerable to objections stressing the intuitive dissatisfaction with the implication that above some threshold T, the inequality between the super-rich and those who barely have enough would not be a concern of justice (Casal 2007). But identifying the threshold is a delicate matter, and critics of the sufficiency view believe sufficientarians face a theoretical dilemma on this issue. If defining a relatively high threshold, such as in terms of welfare satisfaction or contentment, the sufficiency view undervalues the urgency found in the substance of absolute deficiency. That is, if our sufficiency view allows not being perfectly content to be an absolute deficiency, we have certainly undervalued the importance of being released of deficiencies such as hunger, deprivation, suffering, etc. Any reasonable sufficiency view needs to underline the special importance of addressing the latter deficiencies rather than the former. On the other hand, setting a very low threshold – e.g. set at the level of basic needs fulfilment – makes the sufficiency view vulnerable to being ignorant of even quite significant inequalities above this threshold. The difference between the super-rich and people who barely have their basic needs met is simply not, on any reasonable interpretation of justice, irrelevant. Thus, the sufficiency view seems faced with this troublesome dilemma in f leshing out a relevant and plausible threshold level. 2. SHIELDIAN SUFFICIENTARIANISM Liam Shields smoothly solves the above dilemma. He proposes to exchange the strong negative thesis with what he calls “the Shift Thesis”, stating that “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” (2016: 30). Although this interpretation of the view does not f lesh out a much specified threshold definition, it does in an important way render the sufficiency view more plausible. What it does, effectively, is to resolve the dilemma by allowing for a more modest threshold level – set at some non-specified level of resources – than would otherwise have been acceptable, since the Shift Thesis enables sufficientarianism to object to inequalities above the threshold. This softens the hard inegalitarian implications of sufficientarianism while still remaining loyal to the central driving intuition of the generic sufficiency view. We can get a better grasp of what the Shift Thesis involves in figure 2.2, Shielding Sufficientarianism from the Shift 145 LEAP 5 (2017) which Shields presents to illustrate what he calls non-uniform prioritarianism (2016: 32). Figure 2.2. Non-uniform prioritarianism Figure 2.2 displays how non-uniform prioritarianism involves a significant change in our moral reasons to benefit people further, once they reach a certain welfare level (here this level is 2). This captures the shift that Shields builds his sufficientarianism upon. Non-uniform prioritarianism is different from uniform prioritarianism because the former claims that there is a central change in terms of the relationship between how well off people are and the moral importance of benefitting them. This change, Shields convincingly argues, can only be explained in reference to the Shift Thesis. Thus, although non-uniform prioritarians can entail a wide range of other distributive principles, they must rely on some sufficiency principle, because they appeal to the shift. This perspective grounds a much wider relevance of sufficiency principles than normally assumed, because it identifies a very intuitively plausible and common idea – that there are changes in the rate of reasons to benefit people depending on their level of welfare – as a specifically sufficientarian idea. And even more importantly, the appeal to the Shift Thesis does take much of the edge off the most widely shared criticism of sufficientarianism, namely that it is implausible to accept that inequalities above the threshold level are irrelevant to justice. Thus, the prospects of grounding the sufficiency view upon the synthesis of the Positive Thesis and the Shift 146 Lasse Nielsen LEAP 5 (2017) Thesis is promising and fully justifies Shields’ status as among the leading contributors to the development of sufficientarian theory. 3. SOME GENERAL PROBLEMS WITH THE SHIFT Although the advantages of the Shift Thesis are clear enough, there are still some questions that need to be addressed. The Shift Thesis is generic and ecumenical in its outline, and although this is of course not in itself a problem, it blurs our perception of what happens after the shift. Shields seems to imply that the Shift Thesis could be compatible with sets of moral reasons that refer to the value of fairness above the threshold (2016: 35). But this raises the question of what the relationship is between the sufficiency-reason and other moral reasons. Shields’ central idea seems to be that once our sufficiency-reason – that is, our reason to benefit a person that stems from this person being below the absolute threshold-level of welfare – stops being salient, because the person is pushed above the threshold, we will turn to the best alternative reason to benefit further. Fairness concerns, such as distributive egalitarianism or responsibility, seem likely candidates. But that gives the impression that the shift is not accurately depicted as a bend on the otherwise nicely linear (prioritarian) graph of the development of our moral reasoning. Rather, it seems that we should think of it as two separate lines. One line, the sufficiency-reason, representing our very strong commitment to bring people above the threshold, and then a separate line, representing our other moral concerns, that so to speak “take over” once we reach the absolute threshold. This is shown in figure 2.2* below. In figure 2.2*, the shift is depicted as the intersection between the red line, representing our sufficiency-reason to benefit people which is based on their (below-threshold) level of welfare, and the blue dotted line, which represents whatever weaker moral reasons we are left with once the stronger moral reasons becomes non-salient. If 2.2* is a fair illustration of the shift, and I believe it is actually more precise than Shields’ own from figure 2.2, then it raises the question of how these two sets of moral reasons relate to each other. In other words, how should we interpret the relationship between the two lines in figure 2.2*? Shielding Sufficientarianism from the Shift 147 LEAP 5 (2017) Figure 2.2* The Shift Thesis as plural sets of reasoning It seems then, that to make the Shields framework plausible, one would have to decide on a reasonable relationship between the sufficiency-reason and other reasons. First, one possibility is to say that the sufficiency-reason should only take some priority over other moral concerns, but in general be weighed against our alternative set of reasons. That is, if fairness is a relevant moral reason, then this reason ought to be given some weight in our moral deliberation. That is, our decision to help Person I rather than Person II to reach the threshold depends not only on their level of welfare but also on the interaction of other moral reasons – say responsibility- sensitive fairness – on the sufficiency-reason. For example, we might say that if Person I and II are faring equally badly (both below threshold at level 1,5), but differ in terms of exercise of responsibility, then responsibility sensitivity tie-breaks our moral deliberation in favour of priority to the prudent. That is, under resource scarcity, we should give priority to helping the prudent over helping the imprudent. But then we might also say that although being worse off than others below the threshold takes more presence in our calculation than responsibility sensitivity, then large differences in responsibility could outweigh the priority given from level of welfare, so that even if Persons I and II are unequally badly off (e.g. I at 1; II at 1,5), then the difference in their exercise of responsibility could be significant enough as to alter our immediate priority. Finally, it could also very likely imply that under given circumstances, where we have very weighty responsibility-sensitive reasons to benefit Person I who is above the threshold (e.g. at level 3), these reasons could potentially outweigh our 148 Lasse Nielsen LEAP 5 (2017) reason to help Person II (et level 1,5) reach the threshold. This seems like a possible way to embrace value pluralism much in line with standard luck egalitarianism (Temkin 2003; Lippert-Rasmussen 2016), but it seems a very unlikely sufficientarian strategy. This is because it downplays the work of the Positive Thesis (at least in theory) to a miniscule degree, although this is so centrally carrying the sufficiency intuition. This leaves the Shift Thesis shifty3 because it makes the sufficiency-reason – stemming from the strong appeal of the Positive Thesis – merely one among a number of moral considerations. Shields might of course decide that the sufficiency-reason should take absolute priority over other moral concerns. This is the standard sufficientarian move. And this, I should stress, is what I believe he ought to say. But there are two problems involved for Shields in taking this path. First, if fairness (or another egalitarian concern) is fully outweighed by the sufficiency-reason below the threshold, but takes the lead above the threshold, once our sufficiency-reason becomes non-salient, then Shields’ sufficientarianism is not distinctive from pluralist telic egalitarianism such as Temkin’s comparative fairness egalitarianism (2003; 2017). Shields might of course just say that this is because, on his account, Temkin is a sufficientarian, but this seems strange because the dispute between sufficientarians and egalitarians is not about accurate labelling but about the value of distributive equality. Hence, if that is the case, it seems more correct to withhold that Shields is no sufficientarian. Second, if Shields gives absolute priority to the sufficiency reason, then his synthesis of the Positive Thesis with the Shift Thesis makes his view vulnerable to the same objection he presents against upper limit sufficientarianism, which he contrasts himself to. To see how Shields’ view departs from upper limit sufficientarianism let me employ an example, also borrowed from Shields (2016: 23). Table 2.3 Upper limit sufficientarianism I II III I V V V I A 1,000 1,000 1,000 1,000 1,000 3 B 5 5 5 5 5 5 C 5 5 280 5 5 4 3 I am indebted to Jens Thaysen for this catchy, although admittedly slightly tacky, punchline. Shielding Sufficientarianism from the Shift 149 LEAP 5 (2017) Table 2.3 shows three different hypothetical scenarios (A, B, and C) entailing very different distributional shares for different groups or persons (e.g. in a given society). As Shields rightly points out, upper limit sufficientarianism would prefer B to both A and C (which headcount sufficientarianism would as well), when the threshold is set at 5. This contrasts it with weighted prioritarianism, which would rank the scenarios A, C, B (from best to worst). Shields favours the weighted prioritarian reply on intuitive grounds, and if this was not the case, it would be unclear on what grounds he would dismiss upper limit sufficientarianism in the first place. But if the Shift Thesis is to be understood to entail absolute priority to our moral reason to benefit people who are below the threshold, then Shieldian sufficientarianism – synthesising the Positive Thesis with the Shift Thesis – gives the same ranking as upper limit sufficientarianism. That is, it would favour B over both A and C, when the threshold is at 5. This is surely no embarrassment. Maximin and leximin prioritarian views would also prefer B over A and C, as would telic egalitarianism. That is, on further ref lection, it is not at all obvious that our intuitions about this case work in favour of weighted prioritarianism or Shields’ own account, and against the other theoretical standpoints. In fact, in section 4 defend upper limit sufficientarianism against this intuitive strike. My point here is merely to highlight that if Shields wishes to stick with the weighted prioritarian view in this case – and therefore rank A, C, B (from best to worst) – then he is left with giving away the sufficientarian commitment to the strong priority of the Positive Thesis. 4. DEFENDING THE UPPER LIMIT This section defends upper limit sufficientarianism. As suggested in section II, the sufficiency view could be understood in a negative form, as a generic principle saying that, justice is concerned with eliminating absolute deficiencies rather than inequalities. If this principle is accurate, then justice would be fulfilled once absolute deficiencies are eliminated, regardless of whether inequalities still persist beyond this point. This adheres to the ranking B over A and C in table 2.3 (when the threshold is 5), because only in B are deficiencies eliminated. As mentioned, this is the same guidance as maxmin and leximin prioritarianism as well as egalitarianism, so it need not be counterintuitive, but as Shields mentions, upper limit sufficientarianism also gives counterintuitive guidance in cases where everyone has secured enough (2016: 23). To see this, we can take a look at table 2.3* 150 Lasse Nielsen LEAP 5 (2017) Table 2.3 Upper limit sufficientarianism (w ith no one below the threshold) I II III I V V V I A 1,000 1,000 1,000 1,000 1,000 5 B 5 5 5 5 5 5 C 5 5 280 5 5 5 Table 2.3* is similar to table 2.3 except that the worse off in A and in C are now lifted to the threshold level (at 5). Thus, no one is below the threshold level and, moreover, the worst off in A and in C are just as well off as the best off in B. In other words, in economic terms, A dominates B and C; and C dominates B, so that moving from B to C; and from C to A would be Pareto-efficient moves. However, upper limit sufficientarianism would be unable to prefer A over B (or C over B), simply because everyone is above the threshold, and because upper limit sufficientarianism accepts the “upper limit claim” that “no distributive principles apply to benefits among those who have secured enough” (2016: 22). It is not that it necessarily needs to prefer B, but the problem is that it cannot in itself capture that we should not be satisfied with B, although dominantly better alternatives A and C exist. This clearly seems to put a stark challenge against upper limit sufficientarianism; but one that Shields’ Shift Thesis can enable us to tackle. On my account, there is nothing wrong with the guidance of upper limit sufficientarianism even in this extreme scenario. In fact, on further ref lection it is not even clear that our intuition works to count against it. The central problem with the illustration of upper limit sufficientarianism above is that it gets lost in what I call the Illusion of Numbers against Sufficientarianism.4 The illusion of numbers against sufficientarianism assumes that the difference between how well different people are doing is meaningfully captured by the numerical distance between larger and smaller numbers. But sufficientarianism properly understood should reject this assumption. The remainder of this section explains why and thereby argues against the illusion of numbers against sufficientarianism. To see how Shields’ illustration of upper limit sufficientarianism gets lost in the illusion of numbers against sufficientarianism, let’s consider the content of the example in Table 2.3*; which, you will recall, is a hard case against upper limit sufficientarianism. In Table 2.3*, the threshold is set at 4 I adopt this from Nielsen (unpublished paper). Shielding Sufficientarianism from the Shift 151 LEAP 5 (2017) welfare level 5. Since the figure “5” alone does not in itself tell us a lot about why we choose that level rather than any other as the relevant threshold, we assume that it is the content (of resources, opportunities and welfare) that this figure stands for that constitutes the relevant cut-off point. Following Shields’ example, let’s say that “those who pay the top rate of income tax have enough after tax” (2016: 22), and therefore let’s assume that these people are at level 5. What that must mean is that they have sufficient resources and opportunities for their welfare level to be considered a “5”. This seems intuitively appealing. These people have a stable monthly income; they lead autonomous lives; they can afford decent or very decent housing; they have access to decent social insurance; they have a stable health together with a health care system that is prepared to assist them if they fall ill; they are also mentally healthy; their offspring face good social opportunities and have access to good quality education etc. All these things are tacitly put them into the figure “5”. The problem arises because if all these welfare goods are contained in the number “5”, then what can possibly be the content of the number “1000”? In instance, it follows that, if the threshold (5) contains all the above mentioned welfare goods, then the best-off (at 1000) have all the same times two-hundred. Or, more accurately, they would have the welfare level that you gain from having all these goods times two-hundred. But that is not only hard to grasp, but simply meaningless. You could of course imagine a case in which everyone has all the before mentioned welfare goods and then still some have 200 times as much money as others, but then the inequality in question is solely expressed in material resources, and this is useless because “money” alone is very rarely the currency that critics of sufficiency would employ. For one thing, it is evident that very rich people could be worse off than less rich people in other value-metrics. The illusion of numbers against sufficientarianism stems from the fact that numerical comparisons are simplistically scalar and potentially infinite, whereas real life comparisons are not only more complex than that, but also simply incompatible with that way of making interpersonal comparisons. Upper limit sufficientarians reject the simple numerical comparison assumption on which these comparisons are made. They are not concerned with numbers. Their only concern is deficiency. They care about eliminating material deficiency such as hunger, deprivation, illness, suffering etc., and they care about social deficiency such as oppression, dominance, discrimination etc. (Frankfurt 1987; Raz 1986). Upper limit sufficientarians reject the simple numerical comparison assumption in favour of another assumption; namely, that there is an absolute level of well- being (broadly conceived), above which additional resources will not benefit 152 Lasse Nielsen LEAP 5 (2017) people further in any way relevant to justice, regardless of the strength of people’s personal desire to possess more resources. On that assumption, it seems that there is nothing wrong with perceiving B as incomparable with A and C. But from that assumption, we can derive a rather interesting and strong claim about the nature of the value of distributive goods, which is controversial, but which I think we have good reasons to accept. We can say that, no distributive good (or bundle of goods), that is relevant to justice, can have a comparable value if given to people below the threshold than if given to people above that threshold (Nielsen 2016). This claim is incompatible with the simple numerical comparison assumption, because it implies that the difference between 996 and 1000 is incomparable to the distance between 1 and 5, although these distances are clearly comparable in a numerical sense captured by the mathematical fact that 4 equals 4. But translated into what these numbers stand for in terms of real goods and their value, it is far from implausible to accept it. Four loaves of bread is of course equivalent to four loaves of bread, but the value of that bundle of goods surely hinges on whether one faces an absolute deficiency in food supplies. 5. CONCLUSION L ia m Sh ields’ w r it i ngs have su rely had a g reat i mpact on t he t heor i z i ng w it h i n d ist r ibut ive just ice. A nd a lt houg h t h is is on ly for good rea sons, i n t h is paper I have a rg ued t hat h is ma i n cont r ibut ion, consist i ng i n t he of fer i ng of t he Sh i f t T hesis a s a n a lter nat ive to t he Negat ive T hesis, is a n u n necessa r y detou r for su f f icienta r ia ns. In fact, adopt i ng t he Sh i f t T hesis need lessly leaves su f f icienta r ia n ism open to a nu mber of cr it ica l quest ions, because t he a lter nat ive i n sta nd i ng up for upper l i m it su f f icienta r ia n ism seems much less t heoret ica l ly t roublesome. I conclude t hat where t he Sh i f t T hesis leaves su f f icienta r ia n ism “sh i f t y ”, t he upper l i m it seems to do good enoug h for su f f icienta r ia n ism to ma i nta i n it s st rong potent ia l for bei ng t he lead i ng idea l of d ist r ibut ive just ice. BIBLIOGRAPHY Axelsen, D. and Nielsen, L., 2015. “Sufficiency as Freedom from Duress”, Journal of Political Philosophy 23: 406-426. Benbaji, Y. 2005, “The Doctrine of Sufficiency: A Defence,” Utilitas 17: 310-32. Casal, P. 2007, “Why Sufficiency is Not Enough,” Ethics 117: 296-326. Crisp, R. 2003, “Equality, Priority, and Compassion,” Ethics 113: 745-63. Shielding Sufficientarianism from the Shift 153 LEAP 5 (2017) Frankfurt, H. 1987, “Equality as a Moral Ideal,” Ethics 98: 21-43. Hirose, I., 2016, “Axiological Sufficientarianism”, in What is Enough? Sufficiency, Justice, and Health, C. Fourie and A. Rid (eds.), 51-68, Oxford: Oxford University Press. Huseby, R. 2010, “Sufficiency: Restated and Defended,” Journal of Political Philosophy 18: 178-97. Lippert-Rasmussen, K., 2016. Luck Egalitarianism, London: Bloomsbury. Nielsen, L., 2016. “Sufficiency Grounded as Sufficiently Free: A Reply to Shlomi Segall”, Journal of Applied Philosophy 33: 202-216. Nielsen, L., 2017. “What Is Our Real Concern with Real Inequality?” Policy Studies Journal, early online view (DOI: 10.1111/psj.12205). Nielsen, L., Unpublished paper. ”The Illusion of Numbers Against Sufficientarianism”. Raz, J., 1986, The Morality of Freedom, Oxford: Oxford University Press. Scanlon, T.M., 1975, “Preference and Urgency”, The Journal of Philosophy 72: 655-669. Segall, S., 2016, “What is the Point of Sufficiency?”, Journal of Applied Philosophy 33: 36-52. Shields, L., 2012, “The Prospects for Sufficientarianism”, Utilitas 24: 101-117. Shields, L., 2016. Just Enough: Sufficiency as a Demand of Justice, Edinburgh: Edinburg University Press. Temkin, L., 2003, ”Equality, Priority, or What?”, Economics and Philosophy 19: 61-87 Temkin, L., 2017, “Equality as Comparative Fairness”, Journal of Applied Philosophy 34: 43-60.