Owing Me, Owing You: Sufficiency, Demandingness, and Global Justice

In the global justice debate, our duties to compatriots and foreigners are often held to differ in terms of demandingness. Statists, in particular, think that duties to compatriots are more demanding than duties to foreigners. In this article, we flesh out and scrutinize the main elements of Liam Shields’ considerations about global justice in his recent book, Just Enough . Shields notes that the global justice debate largely overlooks that our duties may be more or less demanding in two distinct respects; in terms of content and in terms of stringency. He suggests that the distinction between content and stringency , combined with his sufficientarian thesis, opens up new and (more) plausible positions in the debate. Here, we flesh out the implications of Shields’ tentative suggestions and consider the viability and novelty of the potential positions it permits. We conclude that his considerations of content provide little new to the debate, as this is already the focus of most global justice theorists. However, stringency brings a much needed concern with how to prioritize conflicting duties to the debate, and potentially opens up a range of new positions on how to make sense of our duties across and within borders as well as allowing us to reimagine already existing theories. The article outlines some new potential positions and novel readings of existing views.


INTRODUCTION
Many of us feel a strong sense of moral outrage and obligation when confronted with news of malnourished children or people fleeing civil war abroad. Many have similar responses when confronted by the realities of inequality that affect their own societies, such as vastly unequal access to higher education, inequalities in wealth and property, and the larger obstacles to political influence faced by racial and cultural minorities. But which of these issues place greater moral demands on us and how do we even compare our obligations in the global and domestic realms?
For the last couple of decades, the debate about global distributive justice has been defined by a stark divide between two overarching sides: statists and cosmopolitans. Statists hold that our duties 1 to compatriots are significantly more demanding than our duties to foreigners. Cosmopolitans, on the other hand, hold that compatriots and foreigners are entitled to (more or less) the same. The two sides often seem irreconcilable. In his book on distributive justice, Just Enough (2016), Liam Shields explores how his sufficientarian account of justice might be applied to the global realm in order to overcome this stalemate. He does so by introducing two variables that make possible new ways of conceptualizing our duties of global justice.
These two new nuancing variables are: 1. The sufficiency threshold: what we owe to others varies according to whether they fall below or above the sufficiency threshold; whether they already have enough. The debate between cosmopolitans and statists has been about whether the domestic and global spheres are different realms to which different reasons apply. Shields' sufficientarianism introduces an additional division of realms: it divides the realms of reason below or above the sufficiency threshold.

Two components of demandingness: content and stringency.
Content is about how much we owe others; stringency is about the urgency of fulfilling the duty when its fulfillment conflicts with other duties. The debate thus far has almost exclusively been about content-demandingness. Shields seeks to apply his sufficientarian reasoning to both content and stringency in the global realm.
Shields' analysis is preliminary, but suggests new ways of nuancing the debate. In this paper, we build and elaborate on his aperçu, investigating how the resulting conceptual map compares to the existing positions on 1 In this paper we use the terms duties and obligations interchangeably

200
Siba Harb, David V. Axelsen LEAP 5 (2017) global distributive justice, and exploring whether it, in fact, opens up hybrid positions between the two ends of the statist-cosmopolitan divide. We conclude that the combination of a sufficiency threshold and a shift in content-demandingness does not produce new viable positions. However, the distinction between content and stringency can provide new perspectives on the debate. Thus, we flesh out how the stringency dimension can inform contemporary debates of global distributive justice. We begin by briefly explaining Shields' view on sufficiency and the distinction between content and stringency within demandingness of duties upon which his analysis turns.

SUFFICIENCY AND DEMANDINGNESS IN GLOBAL DISTRIBUTIVE JUSTICE
The central claim of Shields' book is that justice makes different demands upon us depending on whether the individuals with whom we are concerned are above or below the sufficiency threshold. This is because our reasons for what individuals are entitled to and what duties we have with respect to meeting those entitlements differ, or shift, once we move from a context in which some have less than enough to a context in which everyone has enough. Furthermore, it is the case, Shields stipulates, that for several central dimensions of societal justice, the primary goal is to ensure sufficiency for everyone, and once someone reaches this threshold, benefitting them further brings about a different kind of value or is supported by a different sort of reason. Reasons that, in this way, apply only up to a certain threshold are referred to as satiable.
For example, our reasons to give a loaf of bread to someone who is starving are different from the reasons we may have to give a loaf to someone who is well fed but collects loaves of bread as a welfare-generating hobby (however passionately). And this might be explained by the fact that when one is below a threshold of basic needs (starvation) our reasons to benefit her are of a different character than the reasons we have to benefit someone above the basic needs threshold (loaf-collector). Reasons to do with basic needs are satiable. This is what Shields calls the shift thesis. 2 This idea underlies the first nuancing variable.
In the book's chapter on Sufficiency and Global Justice, Shields,points 2 As Robert Huseby points out in his article in this volume, there are two ways of understanding the shift; one which concerns the weight of additional benefits above the threshold and one which concerns a shift in the nature of the reasons. We think this second reading is the more plausible one and will, hence, be assuming that here. LEAP 5 (2017) Owing Me, Owing You: Sufficiency, Demandingness, and Global Justice out that there are two ways to characterize the demandingness of a duty; two ways in which one duty can be more demanding than another. First, the demandingness of a duty may refer to the content of the duty. This refers to "the conditions under which the obligation has been successfully discharged" (Shields, 2016: 173). In other words, the demandingness of our duties refers to how much it takes for them to be fulfilled. A duty is more demanding than another content-wise when it requires more of us than the other duty does. For example, if a good friend invites you to a wedding then you are, barring exceptional circumstances, obligated to go. If, on the other hand, a stranger (generously and somewhat surprisingly) invites you their wedding you are not obligated to go. However, you do owe them declining their offer politely. Duties to friends, we normally think, demand more of us in terms of time and effort. In what follows we refer to the content dimension of demandingness as content-demandingness.

201
Second, the demandingness of a duty may refer to its stringency, by which Shields means the priority that is attached to the duty's fulfillment (Shields, 2016: 177). A more stringent duty, then, is more urgent to fulfill. A duty, D1, is more stringent than a duty, D2, when fulfilling D1 takes priority over fulfilling D2. This means that we should fulfill D1 before fulfilling D2 and that if the two duties clash such that we could only fulfill one, we should fulfill D1. In terms of demandingness, whereas content concerns the 'size' of the duties, stringency denotes the 'weight' to assign to the fulfillment of a particular obligation. To illustrate the notion of stringency, imagine you are sitting between a friend and a stranger who are both suffering from heartache (incurred, perhaps, because the weddings to which they both invited you are not going as planned). Imagine further that you are in the position to alleviate their pain somewhat by way of a similar effort; a kind word. In terms of content-demandingness, in other words, the two are the same. You might, further, think you have moral reasons to do both. But the urgency of fulfilling those duties would differ; the stringency of your obligation to your friend would be greater. In what follows we refer to the stringency dimension of demandingness simply as stringency.

A GLOBAL SHIFT IN CONTENT?
As mentioned in the introduction, statists and cosmopolitans disagree about the comparative demandingness of domestic and global duties of justice. The notion of demandingness that influential statist and cosmopolitan accounts employ is (usually) content-demandingness. 3  main focus for statists against cosmopolitans, thus, is that our duties to co-citizens demand significantly more of us than our duties to foreigners in terms of content. In this section, we investigate whether introducing Shields's sufficientarian shift thesis to the current debate about contentdemandingness brings out new distinctive positions.
Many statists have a sufficientarian component in their theories of global justice and hold that, while we have egalitarian duties to our compatriots, for instance, we are only obligated to ensure that foreigners reach a level of sufficiency (Blake, 2001;Miller, 2007;Sangiovanni, 2007). Clearly, on statist accounts, duties to compatriots are more contentdemanding than duties to foreigners. For statists, the difference between what we owe compatriots and what we owe foreigners stems from the fact that the domestic sphere and the global sphere are two distinct areas of interaction each with different reasons of justice operating within it.
Some cosmopolitans also have a sufficientarian component in their views. Either by defending a high threshold of sufficiency for everyone (Nielsen & Axelsen, 2016;Nussbaum, 2000) or by arguing for a (lower) global sufficiency threshold as part of what is owed to everyone (Caney, 2005: 122;Shue, 1980). For cosmopolitans, unlike for statists, the same reasons of justice are at play within the domestic and global spheres (although, obligations may differ depending on how well-placed one is to fulfill them - Caney, 2011: 514;Goodin, 1988). The duties to compatriots and foreigners are equally demanding in terms of content.
Shields suggests that statists and cosmopolitans each capture one familiar and plausible intuition which he calls, respectively, 'Compatriot Partiality' and 'State is Arbitrary ' (2016: 188). The aim of his sufficientarian global justice exploration, then, is to seek to capture both. One can do this, Shields holds, by applying the shift thesis to our theorizing: how we reason about distributive justice is different for a context where some fall below the sufficiency threshold compared to a context where all are above the threshold. This opens up positions according to which our obligations to compatriots vary depending on whether they are below or above the threshold, making space for some (threshold-dependent) partiality towards compatriots. But it still leaves room to say that this partiality should be contingent on whether (or the degree to which) foreigners are below or above sufficiency, thus including a concern with the morally arbitrary effect one's birth country has on one's life prospects into in the reasoning. This, Shields suggests, opens up new positions in the debate. Most interesting, he thinks, are those potential positions where compatriots or foreigners crossing the threshold of sufficiency leads one to shift from 203 LEAP 5 (2017) Owing Me, Owing You: Sufficiency, Demandingness, and Global Justice being a statist to being a cosmopolitan, or vice versa. 4 Shields outlines one version of a shift from statism to cosmopolitanism: "we could owe prioritarianism domestically and sufficiency globally but once some level of sufficiency is reached for one or both groups, we owe equality to all" (Shields 2016: 176). 5 However, it is unclear to us that a shift in content does, or even could, actually open up any viable or new positions on global justice. Consider the example Shields uses where one shifts from being a statist to being a cosmopolitan once we cross the sufficiency threshold. Now, according to Shields, reasons that can justify the existence of a threshold are ones that are satiable. The main candidates for satiable reasons that Shields explores in previous chapters are basic needs and autonomy. Both are satiable in the sense that they do not provide a normative basis for benefits above a certain level (the threshold) (Shields 2016: 34-37). But it is not clear how such satiable reasons can justify a division between the global and domestic realms in terms of content. Reasons of basic needs and autonomy apply universally; everyone shares the trait that gives rise to the relevant obligations. And, indeed, when theorists in the global justice debate, be it statists or cosmopolitans, claim that we owe basic needs fulfillment to foreigners, they do so on the basis of universal human traits and vulnerabilities, not because the potential recipients are foreigners. 6 Shields, it seems, assumes that a division between the two realms can be drawn. But that is difficult to justify on his account, since no satiable reasons seem to support this divide. It is difficult to see, then, how one can be a statist about content below the threshold. Now, despite this, one might still consider the global and domestic realms to be distinct when reasoning about the demands justice. One might do so even when the content of our duties in both realms is basic needs fulfillment and where our duties in the domestic realm are more demanding than globally for other reasons, and perhaps this is what Shields has in mind. But differences between the two realms are, then, due to the different reasons we have to prioritize duties to compatriots vs. 4 Shields refers to this as radical content shift sufficientarianism (2016: 176). 5 Shields' example identifies distributive rules; priority, sufficiency, and equality. Distributive rules, although they are often built around reasons that provide content, are not in themselves content. This is an issue because several distributive rules have stringency considerations as a constitutive feature. Prioritarianism, in particular, says less about how much we owe to someone than about how urgent it is to fulfill such duties. Although, this makes it more difficult to evaluate Shields' content-position, we disregard this issue here.
6 This issue applies when the threshold is the same for the domestic and the global realm. One might think that two different thresholds govern the two realms. It is possible that Shields has this in mind. If he does, he does not mention it and, in any case, this would give rise to a host of very different and difficult questions.

204
Siba Harb, David V. Axelsen LEAP 5 (2017) foreigners. In other words, such considerations concern reasons to give the fulfillment of one group's basic needs higher priority. But this difference is not one of content-demandingness, but of stringency; it tells us something about how urgent it is to get people in different realms up to the threshold. We conclude that it is not clear that applying the shift thesis to the contentdimension of demandingness adds viable and consistent new positions to the global justice debate. We now move to discuss the stringency dimension and explore whether it provides us with new positions or insights into the global justice question. We think it does.

STRINGENCY IN GLOBAL JUSTICE
The global justice debate has to a large extent focused on what we owe to compatriots and non-compatriots as a matter of justice: whether and to what extent duties to compatriots make greater demands on us than duties to foreigners. However, as Shields points out, the debate about our duties of global justice has paid little attention to a different dimension of demandingness: stringency. To recall, we say that a duty, D1, is more stringent than a duty, D2, when fulfilling D1 takes priority over fulfilling D2. Shields says little about what influences stringency considerations and how stringency might illuminate the debate on global justice. However, we think the idea of treating stringency as a separate dimension has a number of advantages and generates valuable insights. In this section, we flesh out and explore how it may do so.
Introducing the stringency dimension opens up new possible theoretical positions in the global justice debate whose plausibility can be explored further. Stringency works as a new dividing line that brings some positions closer together and pushes others farther apart than otherwise assumed. A new and different conceptual map emerges when we take the comparative stringency of duties into consideration because one's view regarding the comparative stringency of our domestic and global duties of justice need not necessarily track one's view on the comparative contentdemandingness of those duties. The two dimensions, in other words, come apart. For instance, one might hold, like statists do, that the content of justice-based duties to others depends on whether or not one shares membership in a state; i.e. equality for co-citizens and basic needs fulfillment for foreigners. But one might also think that state membership plays no role in defining the stringency of our obligations. Instead, factors such as how badly off a person is, how urgent their plight is, etc., would then determine this. We can say of such a position that it is statist about  Sangiovanni thinks that we owe more (in terms of content) to conationals than we do to foreigners. This is because, on his reciprocitybased conception of justice, we owe others a fair return on their participation in the cooperative scheme we share with them. As such, we owe co-nationals a fair return on their participation in the cooperative scheme that is the state, and we owe foreigners a fair return on their participation in the cooperative schemes that function globally. 8 According to Sangiovanni both the type of goods produced and the extent of one's contribution to their production are significantly more encompassing domestically than they are globally, and this explains why we owe conationals more. But nothing in Sangiovanni's account commits him to assigning higher stringency to domestic duties of justice over global duties. Although Sangiovanni does not explicitly take a stand on this, his view at least allows for the possibility that the stringency of duties is not membership dependent. A view about content like Sangiovanni's is therefore compatible with a view assigning higher stringency to fulfilling duties to those who are very badly off or to those for whom it is more urgent that duties are fulfilled, independently of whether they are compatriots or foreigners. In policy terms, such a view would entail that we should seek to alleviate global poverty before turning to domestic inequalities.
Indeed, the grounds upon which Sangiovanni's account is built are particularly well-suited for this interpretation, since there is nothing inherent in his conception of reciprocity or the content of the particular duties that justifies granting one precedence over the other. It would be perfectly compatible with such an account to say that duties to the badlyoff ought to take precedence regardless of whether those suffering this plight are co-nationals or foreigners. Thus, for an account like the one proposed by Sangiovanni, a whole range of positions on stringency is available. And this includes the one sketched here which, as far as we know, is an unoccupied seat in the global justice debate chamber.
The space opened up by introducing the stringency dimension becomes clearer if Sangiovanni's position is contrasted with one that does not allow 7 We also note that Thomas Nagel (2005) makes a number of comments that suggest sympathy to the view that factors related to urgency of need influence the stringency of duties. He writes, for instance, that "[t]he urgent current issue is what can be done in the world economy to reduce extreme global poverty" (118, emphasis added). We thank an anonymous reviewer for pointing this out. 8 See Sangiovanni (2007) for similar interpretations. The way in which David Miller's statist account is grounded, for example, seems to commit him to assigning higher stringency to our duties to co-nationals, to be, in other words, statist about content and about stringency. Miller thinks that our duties to co-nationals are more demanding in terms of content than our duties to foreigners. To Miller, this is because co-nationals share a relationship that is intrinsically valuable. Furthermore, having and acting on special commitments to each other are constitutive elements of what makes the relationship between co-nationals valuable in this manner. In order to maintain the intrinsic value that flows from such relations, then, co-nationals must give (some, although not absolute) priority to fulfilling their duties of social justice over those of global justice (Miller 2007: 40;2013: 175-179). Such relations, on the other hand, do not exist globally and so similar priority is not required. Unlike for Sangiovanni, then, on Miller's version of statism, higher stringency to domestic duties is constitutive of the account.
In our world of massive global inequalities, statist views strike many as morally objectionable for asserting that duties of domestic justice eclipse duties of global justice. But as we have pointed out, when statists talk of demandingness, most often they are talking about content-demandingness. Taking note of the fact that one's view on stringency can come apart from one's view on content-demandingness renders some statist views less objectionable from this point of view. A statist position as Sangiovanni's, for instance, seems more plausible if combined with a cosmopolitan take on stringency; a view, that is, which assigns higher stringency to fulfilling our duties to the worse off or those most urgently in need of help regardless of their membership. And, in that, it is importantly different from a view such as Miller's which, if we are right, is committed to assigning both higher content-demandingness and higher stringency to our domestic duties of justice.
The stringency dimension of our duties of justice has potential implications for cosmopolitan positions too. For just as it is open for statists to be cosmopolitan about stringency, it is open for cosmopolitans to be statist about stringency. Cosmopolitans can, for instance, maintain that domestic and global duties of justice are equally demanding but submit that fulfilling our duties to compatriots takes priority. Or they can be cosmopolitan through and through, maintaining that both the content and stringency demandingness of our duties to compatriots are on par with our duties to foreigners. Reasons for why a person can have a more stringent duty to fulfill D1 (e.g. domestic duties) than to fulfill D2 (e.g. global duties) include that she is in a better position to fulfill D1, or that she has created the expectation in targets of D1 that she will fulfill D1, or even LEAP 5 (2017) Owing Me, Owing You: Sufficiency, Demandingness, and Global Justice plain partiality towards the targets of D1 on account of their special relationship. Several factors, thus, might impact our judgement of the comparative stringency of our duties, some have to do with the level of wellbeing of the target of the duty, others have to do with the capacity of the duty holder, and others still with the relationship between the duty holder and the target.

207
Some cosmopolitans have noted this possibility. Simon Caney, for example, is of the view that both compatriots and foreigners are equally entitled to equality of opportunity but entertains the possibility that: " [o] ne has a 'special' duty to protect the (cosmopolitan) entitlements of one's fellow citizens, as well as a 'general' duty to protect the cosmopolitan entitlements of everyone" (2008: 511). Caney is vague on what could justify uncoupling entitlements from duties. And it seems to us that the best way to make sense of Caney's view would be to understand him as highlighting exactly the distinction between the content of duties of justice and their stringency: the content of what we owe to compatriots and foreigners is the same; but in terms of stringency, what we owe to compatriots might be more demanding. This is an interesting potential position on global justice, one that may in some of its variants be attractive to those who worry that standard cosmopolitan views do not leave adequate space for ethical partiality towards those with whom one shares special bonds such as one's family or, in the case at hand (which is not as similar to that of families as some theorists would have us believe), one's compatriots.
Besides separating the two dimensions of demandingness, Shields mentions the possibility of applying the shift thesis to stringency. Shields suggests that our reasons about the comparative sufficiency of duties shifts according to whether some are below the sufficiency threshold or all are above. It seems plausible to us to hold a view according to which stringency is determined by level of wellbeing when some are below the threshold but then shifts to being determined by other considerations such as legitimate expectations and ethical partiality when all are above the threshold. 9 We merely want to note that if Shields is correct then this quickly multiplies the possible positions in the debate. Here is one possibility: one might be statist about content and cosmopolitan about stringency when some are below the sufficiency threshold (like in our reconstruction of Sangiovanni's view), then back to being statist about stringency when all are above the threshold. Another possibility would be that one is cosmopolitan about content and stringency when some are below the sufficiency threshold and 9 It may seem as though there are no duties left to fulfil after everyone is above the sufficiency threshold if one is a statist. However, recall that Shields' account is shiftsufficientarian and, thus, places some (diminishing) value on adding benefits above the threshold.

208
Siba Harb, David V. Axelsen LEAP 5 (2017) statist about stringency when all are above the threshold. 10 Here, we have sketched a range of new positions that become possible with the introduction of stringency as a dimension of the demandingness of our duties of justice. It goes beyond the scope of this piece to evaluate the sketched positions. What we have done, instead, is to show how reconceptualizing and fleshing out Shields' notion of stringency casts the global justice debate in a new light.

CONCLUSION
Shields introduces two new ideas to the global justice debate. First, he applies the notion of a sufficiency threshold and suggests that this could apply to both the global and domestic realms; that we might have different obligations to both compatriots and foreigners, depending on whether they have enough. Second, he differentiates between two ways in which our obligations may vary in demandingness: content and stringency. In this paper, we have cast doubt on the usefulness of applying the contentdimension of Shields's sufficiency thesis to the global justice debate in which the global and domestic realms are separate. When spelled out clearly it turns out that it opens no new, viable positions regarding the content of our duties of global justice. The stringency dimension, however, does illuminate the global justice debate in new and interesting ways.
While global justice theorists have sometimes hinted at considerations of stringency, it is indeed surprising that so relatively little attention has been paid to this aspect of our justice-based duties. Picking up on Shields's suggestion that the content and stringency dimensions of our duties of justice are distinct and influenced by different considerations, we have tried to show how introducing the dimension of stringency can provide an alternative map of the global justice literature.
But this is not just about conceptual possibilities. Understanding demandingness not only in terms of content but also in terms of stringency, allows us to see that some positions have more similar implications and others more dissimilar implications than otherwise thought. Consider two discussed statist accounts, Sangiovanni's and Miller's. While both agree LEAP 5 (2017) Owing Me, Owing You: Sufficiency, Demandingness, and Global Justice that our duties to compatriots are more demanding than duties to foreigners (albeit for different reasons) Sangiovanni's account opens up the possibility of giving priority to the fulfillment of the basic needs of poor foreigners over social justice obligations to compatriots (even if the latter are more demanding in content). Miller's account, on the other hand, does not seem to allow assigning the same stringency to basic need fulfillment. In this way, Sangiovanni might be closer to a cosmopolitan who assigns higher stringency to basic needs fulfillment, while Miller might be closer to a cosmopolitan who assigns higher stringency to fulfillment of domestic duties of justice. This reshuffling of positions can be useful in pulling the debate about global justice out of the stalemate in which it has, arguably, landed. And, no less importantly, thinking about stringency points us towards important discussions about what to do first, rather than merely speculating about where we should end up.