How Thresholds Matter: On the Bounds and Demands of Justice

Abstract

Two important tasks for theorists of justice are to determine the bounds of justice, which explain why some claims are matters of justice and others are not, and to determine the demands of justice, which settle conflicts that fall within those bounds. In this paper, we clarify the distinction between bounds and demands, revealing two striking things. First, while thresholds have typically been understood to be demands of justice, their use as such is confusing and arguably implausible. Second, thresholds appear to be better understood as demarcating the bounds of justice, if a suitable explanation for their use can be found. We explore three explanations for why thresholds can demarcate bounds and assess the prospects for seeing thresholds in this new and different role. These are satiability of the value of goods, satiability of justice, and conceptual engineering.

1 Introduction

Some claims that can be made of a political community must meet certain conditions in order to have force. For example, you cannot demand that the political community, or the state, supply you with a fiancé or a new best friend. Nor can someone claim unemployment benefits or pensions from states in which they do not reside or hold citizenship. Such claims go beyond the bounds of justice. That some things are out of bounds, and some things are not, follows from the trivial claim that not everything is a matter of justice. Once we have ruled out all the claims that are outside the bounds of justice, we are left with the demands of justice. When citizens call for higher unemployment benefits, equality of opportunity in education, or an end to the gender pay gap, for example, they are proposing an alternative response to the claims made on such goods. They do so based on a particular view about how to weigh justice’s different demands. Here, the question is how to respond appropriately to them, with reference to (distributive) principles, such as equality, priority, sufficiency, and so on.

In this way, the concepts of bounds and demands allow us to distinguish two important tasks for theorists of justice. One is the task of determining the bounds of justice. To do so, we must identify general rules that explain why some claims must be considered under the banner of justice, but others can be disregarded. Another is the task of settling the demands of justice. To do so, we must identify general rules that explain how to decide between competing claims within those bounds. This paper concerns the distinction between the bounds of justice and the demands of justice. Our aim is to show how, once clearly understood, this distinction highlights both confusion and possibility in distributive justice – and, especially, in theories of distributive justice that employ thresholds. The crux of our paper is that arguments about bounds and arguments about demands are logically independent, require different kinds of argumentative support, and are vulnerable to different types of objections. It follows from this independence that theorists who take themselves, or their critics, to be making an argument about the demands of justice, when they are, or should be, making an argument about its bounds, miss their target, and vice versa. Confusing arguments about bounds and arguments about demands have led to mistakes and the neglect of promising positions. Since some arguments in distributive justice can be interpreted both as bounds and demands, we show that properly distinguishing the two opens up new and interesting theoretical possibilities. In this paper, we focus primarily on debates in distributive justice that include thresholds, namely, sufficientarianism and limitarianism, as this is where the confusion and possibility are most pronounced. We will, however, also sketch a few other ways in which the distinction may be employed, suggesting that its utility may be more general.

The paper proceeds as follows: In Section 2, we distinguish arguments about bounds and arguments about demands in justice theorising and survey several familiar examples from debates in distributive justice. In Section 3, we discuss the ways in which this distinction has been overlooked, leading to unnecessary and problematic confusion in debates about thresholds. In Section 4, we provide the first self-conscious treatment of thresholds as determining the bounds of justice. We explore several ways of supporting thresholds as bounds, some of which are more promising than others. Section 5 summarises and concludes.

2 Bounds and demands

The bounds of justice delimit the range of cases to which a principle applies its powers of adjudication. Principles may be limited in application to only certain recipients (say, citizens of a particular state), to certain distributors (say, the government of one’s state), or only to certain kinds of distribuenda (say, resources or welfare but not, e.g. a fiancé or a new best friend).¹ Demands focus on the appropriate response to the many claims that fall within the bounds of justice, for example, by prioritisation or weighing. The debate about the demands of distributive justice has often centred on the pattern of justice. For example, prioritarians hold that the demands of those who are worse off should be given greater weight than the demands of those who are better off, and egalitarians hold that the demands of any individual are indexed to the extent to which satisfying those demands helps to advance equality. Whereas these patterns are concerned with weighing up and trading off important considerations, bounds of justice set some apparent considerations aside as irrelevant – as outside of the weighing – rather than having little or no weight.

Although they are not acknowledged as such, disagreements about the bounds of justice are recognisable from several major debates within the field of distributive justice concerning the scope of justice, the site of justice, and the currency of justice, respectively. The shared character of these debates and the application to questions of the pattern of justice have not yet been appreciated, nor has the broader significance of the bounds question. In what follows, we discuss these examples to show that arguments about bounds and arguments about demands are logically independent, have different implications, require different kinds of support, and are vulnerable to different sorts of objections. Thus, properly understanding whether an argument is about bounds or demands is crucial to any assessment of it. As we shall see, although the bounds of justice discussion may appear unfamiliar to some,² political philosophers usually place several different principled limits on the bounds of justice. The aim of this section is to illustrate the logical independence of claims about the bounds and demands of justice by showing how familiar debates can be reconceptualized along these lines. By showing that such debates can be reimagined at a more abstract level as containing claims about the bounds of justice, we do not mean to say that the more specific versions – site, scope, and currency – are redundant. Rather, we note a kinship between these claims, showing that they are of the same argumentative kind – and, importantly, that claims about the existence of thresholds in distributive justice may also be understood as being of the same kind.

Global justice debates are often described as being about the scope of justice. This is one way of delimiting the bounds of justice. On statist views, the mere fact that one person is not a compatriot of another means that they cannot make demands of justice against that other.³ Thomas Nagel, for example, argues that obligations of justice arise between people who subject each other to comprehensive coercion through laws written and enforced in their name.⁴ Only under such conditions, when one’s circumstances and opportunities are shaped by policies undersigned by another, can one make claims and demands on them in the name of justice. Only co-citizens of states stand in such relations, and only they, therefore, have obligations of justice towards each other.

Arguing against Nagel’s view by emphasising that, say, starving people in sub-Saharan Africa are in greater need than one’s poor compatriots in European welfare states or the USA would be a failure to engage properly with Nagel’s argument. It would be like pointing out how much a particular person would benefit from a fiancé or a new best friend to show that the state should provide it. Doing so assumes that the claims of foreigners and compatriots are all within the scope – the bounds – of justice. Instead, to challenge Nagel’s view, it is necessary to target the bounds claim specifically, disproving the reasons provided for why the claims of foreigners are excluded from adjudication, and why the bounds of justice would be limited to relations of co-authorised coercion. This, of course, is exactly what many cosmopolitans do. They deny that any categorical disqualification can consistently be made on the basis of state borders.⁵

Statism, thus, provides us with a sufficient condition for a claim being outside of the bounds of justice: that it is outside of (nation) state borders. It is qua non-compatriot that foreigners are not owed duties of justice. Arguing about the bounds of justice, in this context, involves deciding whether or not state membership plausibly determines who can make claims of justice; whether a shared national identity or a shared liability to coercive state institutions are the kinds of relations that ought to determine who can make justice claims on one another.

Note that statists and cosmopolitans who disagree on the scope of justice can agree on how demands should be adjudicated (within their preferred scope). For example, they might agree that demands should be met through an equal distribution of benefits and burdens. Being an egalitarian doesn’t make you a cosmopolitan any more than being a statist makes you a sufficientarian. This illustrates the logical independence of bounds and demands. There is more to the distinction, however. The differences between bounds and demands also significantly impact how we ought to engage with these arguments.

Although the global justice debate has mainly centred around claims about bounds, the debate could be reconceptualized into positions about how to adjudicate between the demands of justice. To see that this difference in approach works both ways, consider a brief sketch of what statism as a claim about the demands of justice might look like. Statism about demands does not rule out claims made by foreigners categorically. Rather, it treats claims from foreigners and compatriots differently within its adjudication. This view might look one of two ways. First, the claims of compatriots might have lexical priority over those of foreigners. This leaves open the possibility that once the claims of compatriots are met, then duties of justice to foreigners will follow.⁶ Second, it might mean that we have to give extra weight to the claims of compatriots qua compatriots compared to those of others. Here, justice demands compatriot partiality (or, alternatively, entails a compatriot prerogative) – a thumb on the scale – but could, in principle, be countervailed by overwhelming need on the part of the global poor.⁷ By contrast, as a bounds claim, the statist position identifies the claims of foreigners as irrelevant to justice. Such claims can never countervail the claims of co-nationals, even if all domestic demands are fully satisfied or cannot be advanced further. Claims of non-compatriots can never generate a duty of justice on the part of non-compatriots. They are not ruled out circumstantially or only after careful weighing. Such claims are categorically ruled out.

The bounds and demands variants of scope claims in the debate about global justice are different. They have different implications, require different sorts of arguments in their support, and are vulnerable to different sorts of objections. Thus, it is essential for any clear-headed analysis to observe this distinction.

Much the same issue arises in several other debates about distributive justice, including the site and the currency of justice. With respect to the site of justice, some scholars restrict equality to the basic structure of society, while others apply it to the distribution of benefits and burdens quite generally.⁸ If you negotiate fiercely with your employer for higher than equal pay, have you acted against the ideal of equality and, therefore, unjustly? Or is it, instead, the case that labelling my private choices as claims of justice would be jurisdictional overreach? These questions turn on what kinds of claims qualify to be considered under the demands of justice in the first place, and this turns on different sorts of arguments, and a different set of (hypothetical) examples, than the question of how justice should adjudicate between claims that justice rightly judges. After all, you and I can agree that equality is the correct distributive pattern (demands), while disagreeing about whether the basic structure is the primary subject of justice (bounds). Again, the demands and bounds are logically independent and so the views have different implications, are vulnerable to different sorts of objections, and require different sorts of support.

As an illustration of the logical independence of the two types of claims, consider what happens when we understand the site debate as being about demands. Principles about how to adjudicate about the demands of justice in terms of its site may demand either the lexical prioritisation of claims within the basic structure over those outside, or it may demand that additional weight is given to claims made regarding the basic structure rather than those that are outside it (including, e.g. those pertaining to the egalitarian ethos). As far as we know, this type of view does not exist in the literature, but it is a possible (and coherent) view. This demonstrates the different implications of the views and the need for separate analyses. There is a correct and incorrect way to argue about the site of justice. To engage in argument with someone holding the demands version of the view, we would need to show that this weighing was incorrect or leads to intuitively unappealing or absurd conclusions. The more familiar bounds claim cannot be argued with in this manner. Instead, a critic would have to show that the basic structure is the correct site for justice-concern, that certain claims, by virtue of falling outside of the basic structure, are not valid, regardless of what other features they have. Nothing is implied about how to adjudicate claims, whether egalitarian or otherwise, that fall within the basic structure, by simply accepting this site-based restriction of the bounds of justice.

As a final example, consider the debate about the currency in which claims of justice are to be specified. You may complain that you are not as happy as your constantly cheerful neighbour and submit that we should seek to redistribute welfare from your neighbour to you. I might object that you and your neighbour have an equal amount of resources and that, therefore, egalitarian justice has been realised. Like debates about the scope and the site of justice, questions of currency can be understood as concerning the bounds of justice – about which claims are admissible at the bar of justice.⁹ Those who agree on (bounds) questions of the currency of justice can disagree about the correct (demands) pattern, egalitarian or otherwise. Similarly, those who agree on the correct pattern of justice can disagree on its currency. The bounds and demands versions are, thus, independent, and it is also true that they have different implications and require different arguments.

Consider briefly the demands version of the currency claim, which, again, no one has defended to our knowledge. Such views might hold, for example, that we should prioritise achieving equality in the distribution of well-being, lexically or according to some weighting, over equality in the distribution of resources. So, being more cheerful but less wealthy than my neighbour could be resolved without labelling either of these inequalities as irrelevant because they fall outside the bounds of justice. Larger gains in resources might outweigh smaller gains in well-being or vice versa. This depends on how the view is fleshed out. But the bounds and demands versions of the views are different, and significantly so. To defend a currency in the usual way, as a bounds claim, involves explaining why claims of one kind (say, about resources) are relevant to justice and the others (say, about welfare) are not. To object, we must show the opposite. To defend a currency as a demands claim, on the other hand, we must argue for the relative priority of one currency over another, accepting both as legitimate currencies of justice. These are very different ways of arguing.

One final possibility worth considering is that bounds could be explained by assigning zero weight to some claims. A view that gives zero weight to those outside a jurisdiction, zero weight to those above some threshold, or zero weight to those otherwise outside of what we have called bounds would be extensionally equivalent to the corresponding bounds claims. This raises the question of which framework to use, and since the zero weight versions do not require us to invoke the new framework of bounds, the zero weight approach is more parsimonious. But the decisive advantage of talking of bounds is that it provides us with a deep explanation of why the demands of those outside do not count: justice does not cover those circumstances. By contrast, the zero weight versions have no explanation of why out-of-bounds claims are distinctive.

In this section, we surveyed three of the main questions in distributive justice: questions of its scope, site, and currency. Although their kinship is not explicitly acknowledged, each of these questions have to do with different restrictions to the bounds of justice. In each case, we explained why bounds and demands are logically independent, and we showed that they have different implications, require different supportive arguments, and are vulnerable to different objections. This establishes our general thesis, which emphasises the importance of not confusing the bounds of justice with its demands in any analysis. However, our argument does not establish that one cannot take the same general methodological approach to demands and scope. For example, it is consistent with our argument that a contractualist approach or a consequentialist approach should be taken to both determining the scope and demands of justice. Our point is that in doing so, the contractualist or consequentialist will differ in how they approach scope and demands because they have different roles in the theory. For example, if the contractualist argues for a cosmopolitan scope and egalitarian sets of demands, cosmopolitanism cannot be defended except as the scope, and equality cannot be defended except as a demand. Contracting parties may agree on each, if suitably motivated, but the way they reason to cosmopolitanism will be different from the way they reason to equality, even if the overall approach is contractualist. This can be seen by considering that contractualism appears theoretically open enough to affirm any combination of views, depending on how the contract situation is specified, including the motivation of parties.

3 Confusion about thresholds

We now turn our attention to thresholds because, in our view, it is with respect to thresholds that the failure to distinguish arguments about bounds from arguments about demands has caused confusion and misunderstanding, triggering fruitless debates about objections that have no target. But it is also an area in which neglected interpretations of thresholds as setting the bounds of justice make for promising avenues of theorising.

Thresholds are generally presented as pertaining to the demands of justice. Most often, they are presented as rivals to other principles for settling demands claims, such as equality and priority. What is widely regarded as the original statement of sufficientarianism comes in a paper entitled ‘Equality as a Moral Ideal’, and an early statement of prioritarianism comes in a paper entitled ‘Equality or Priority’.¹⁰ Many more recent papers, both defensive and critical, follow their lead in comparing and contrasting some or all of Priority, Sufficiency, and Equality. In this section, we cast some doubt on sufficientarianism understood as a position about the demands of justice to rival Priority and Equality. We note objections made to its plausibility and point out further shortcomings in terms of its coherence that come to light when we consider the bounds of justice. We go on to show that thresholds appear well-suited to determining the bounds of justice. This further shows the importance of distinguishing the two and of developing the idea that thresholds can be bounds, which we take up in Section 4.

3.1 Sufficientarianism

Sufficientarianism holds that justice is concerned with ensuring that everyone has enough. Almost all self-identifying sufficientarians endorse both the so-called ‘positive thesis’ and the ‘negative thesis’.¹¹ According to the positive thesis, we have weighty reasons to ensure that people secure enough of some good(s). According to the negative thesis, those who have secured enough can make no further claims for additional benefits. The negative thesis is apparently what makes the view both distinctive and controversial. It attracts the most critical attention and is the claim that many sufficientarians are most eager to defend.

The debate about sufficientarianism generally proceeds as if the negative thesis were a claim about demands. This is understandable. After all, sufficientarianism is most often explicitly discussed as a rival to other accounts of the demands or patterns of justice, such as versions of egalitarianism and prioritarianism, even utilitarianism. Sufficientarianism’s most well-known proponents present it as a demands claim. Consider Harry Frankfurt’s example of medicine, where he says:

Thus, suppose that there is enough of a certain resource (e.g. food or medicine) to enable some but not all members of a population to survive. Let us say that the size of the population is ten, that a person needs at least five units of the resource in question to live, and that forty units are available. If any members of this population are to survive, some must have more than others. An equal distribution, which gives each person four units, leads to the worst possible outcome, namely, everyone dies. Surely in this case it would be morally grotesque to insist upon equality.¹²

Frankfurt goes on to say,

Another response to scarcity is to distribute the available resources in such a way that as many people are possible have enough, or in other words, to maximize the incidence of sufficiency.¹³

Here, Frankfurt targets the relative plausibility of sufficientarian ways of weighing demands of justice against egalitarian ways. Whatever we think of the success of this argument, it is presented as an argument about how claims should be weighed.¹⁴ It is not an argument we would expect from someone who understands themselves to be defending a bounds claim about what sorts of claims are categorically ruled out.

Now, consider another one of the most famous examples in sufficientarianism, Roger Crisp’s Beverly Hills case. In this example, you can offer fine wine to different groups of well-off individuals, where some are rich and others are super-rich – in other words, they all clearly have more than enough. Crisp asks us to consider the following numbers, imagining that the threshold of sufficiency is at 25:

We can, thus, either provide 10 rich people with bottles of Lafite 1982, making them two points better off in terms of utility. Or, alternatively, we can provide 10,000 super-rich people with bottles of Latour 1982, making them two points better off in terms of utility. Both groups are clearly far above the threshold of sufficiency. Crisp’s conclusion is that: ‘what the Beverly Hills case brings out is that, once recipients are at a certain level, any prioritarian concern for them disappears entirely’.¹⁵ His own view is compassionate sufficientarianism. It is presented as an alternative to egalitarianism and prioritarianism, both of which he argues are implausible in terms of the distributions they favour.

Understood as a demands claim, sufficientarianism faces many difficulties. First, it invites its strongest and most well-known objection: the indifference objection. The Indifference Objection states that ‘sufficiency principles are implausible because they are objectionably indifferent to inequalities once everyone has secured enough’. It has been made by several theorists, both egalitarian and prioritarian.¹⁶ In response to examples like those used above, Casal provides the following case:

To see why the distribution of such benefits might matter, suppose that having provided every patient with enough medicine, food, comfort, and so forth, a hospital receives a fantastic donation, which includes spare rooms for visitors, delicious meals, and the best in world cinema. If its administrators then arbitrarily decide to devote all those luxuries to justice a few fortunate beneficiaries, their decision would be unfair.¹⁷

Casal goes on to give another example.

when a natural disaster such as a tsunami strikes, it is the wealthiest who should make the greatest donation. Sufficientarians cannot accommodate such plausible convictions. Moreover, since sufficientarians are indifferent not only to inequalities among billionaires and those who have barely enough, they cannot support the preference for progressive over regressive taxes when both are capable of securing sufficiency.¹⁸

The examples are used in a way that makes clear the assumption that the negative thesis of sufficientarianism is a claim about demands. They both assume that the claims of those below the threshold (tsunami victims and patients facing inadequate care) are of the same kind as, and are directly comparable to, claims made above the threshold.¹⁹ But if sufficientarianism is understood as being a way to set the bounds of justice, the point of the two examples is that no patients in the hospital case have claims of justice for additional resources, and that neither billionaires, nor other well-off citizens, can justly retain resources in the face of insufficiency. If the negative thesis were understood as a claim about bounds, it would be strange to try to come up with counterexamples that show that it yields a less intuitive distribution or makes trade-offs in a less intuitive way than egalitarianism or prioritarianism. The implausibility of neglecting a badly off person has no force if the scope of justice does not include them. This is the same point that was made in Section 2 about bounds and demands versions of the scope, site, and currency debates. The intuitive implausibility of ignoring a person’s preference satisfaction has no force if the currency of justice is resources. The implausibility of being unable to condemn those who seek above average pay in wage negotiations has no force if the site of justice is the basic structure. And finally, the implausibility of being indifferent towards the well off in either the tsunami or hospital case has no force if the bounds of justice exclude the claims of the well off because they are well off. Each of these bounds claims requires independent support, but it is of a different kind to the support required for a demands claim. If a good argument for restricting the bounds of justice can be provided, then the alleged counterexamples lack force in each case.

Understanding the negative thesis of sufficientarianism as a view about the bounds of justice therefore offers a possible way of avoiding the indifference objection and rescuing a kind of sufficientarian justice. The indifference objection invites the following question, to which sufficientarians have struggled to convincingly respond: why should we be indifferent to supra-threshold benefits? In light of the need to answer this question, a second problem arises: When understood as a claim about demands, the negative thesis appears incoherent. If the negative thesis is a claim about demands, then it says nothing about bounds, which is a logically independent matter. As we saw, those who disagree on bounds (scope, site, currency) can agree on demands (equality and priority), and vice versa. But if sufficiency is a demands claim, then proponents of sufficientarianism cannot say, as they sometimes imply, that justice is extinguished once everyone has enough. They cannot say that justice is extinguished at all qua sufficientarian because that is a bounds claim, not a demands claim. Indeed, saying that justice is extinguished at some threshold of sufficiency is something that prioritarians and egalitarians can agree on without compromising their position precisely because of the logical independence of bounds claims and demands claims. The only position that is compromised is negative thesis sufficientarianism understood as a demands claim. To see why, consider that proponents of sufficientarianism must accept the claim that justice is not extinguished at the point of sufficiency. But if they also, as the negative thesis states, endorse indifference towards supra-threshold benefits, their view appears incoherent, and there is no longer anything distinctive about the view.

Shift sufficientarianism, a minority position among sufficientarians, is not vulnerable to this objection. The view holds that the threshold marks a change in our reasons, rather than an end to them. It thus rejects the negative thesis.²⁰ Proponents of shift sufficientarianism can maintain that other justice-based demands apply above the threshold that are distinct from those below. This position retains what is distinct about sufficientarianism, without leading to incoherence. Shift sufficientarianism is clearly best understood as a demands claim, as a view about how to adjudicate claims for societal goods within the bounds of justice in a sufficientarian manner. Sufficientarian shifts take place within the bounds.

However, if we interpret the negative thesis as a belief about the bounds of justice, negative thesis sufficientarianism is not incoherent. Defenders can, instead, refer to a distinct argument about why justice does not require more than enough, an argument that can be adopted by egalitarians and prioritarians as well. As a bounds claim, the negative thesis need not rival other patterns of the demands of justice, only those who deny that justice ends at sufficiency. Instead, it rivals the view that the bounds of justice are not sufficientarian in nature. We will return to whether a good argument can be made for drawing the bounds of justice in this way in Section 4, but the objections that have been made to sufficientarianism do not apply to the bounds version. There are promising opportunities for thresholds as bounds claims that are worth exploring here, which have been obscured by the focus on sufficientarianism as a demands claim.

3.2 Limitarianism

Limitarianism is the view that it is morally impermissible to be too rich.²¹ Rather than setting a threshold up to which everyone must be raised to have enough, like the positive thesis in sufficientarianism, limitarianism advocates an upper limit to the societal distribution. Resources held above the limit, so-called surplus money, have zero moral weight,²² and claims to retain such wealth in the face of competing claims should, therefore, be dismissed. Thus, for example, when some people have too little, if society is marked by democratic inequality, or if resources are needed to overcome major societal challenges (like the climate crisis), claims made on resources to ameliorate these injustices should entirely eclipse those made by the wealthy to retain surplus money.

Both critics and sympathisers have treated the limitarian principle as a means of adjudicating between competing claims within the bounds of justice. They’ve treated it, in other words, as a view about the demands of justice. Critics have argued, for example, that the limitarian principle is redundant because it does not provide any distinctive distributive guidance.²³ We already know, they note, from sufficientarianism that we should redistribute to those who have too little, from egalitarianism that we should rectify democratic inequality, and from intergenerational justice that we should contain the climate crisis and its impact on future generations. Critics also argue that it has counterintuitive implications to insist on confiscating wealth from those that are too rich regardless of the consequences of doing so and the circumstances under which it is done. In certain cases, they argue, doing so may not lead to the most just distribution.²⁴ When critics object that the distributive outcomes of the limitarian principle are counterintuitive or that these results also follow from other principles, they are objecting to the way in which the principle adjudicates between different claims. They are treating limitarianism as a view about the demands of justice.

Some limitarian sympathisers have met these points of criticism by arguing that the limitarian principle operates in a different mode than other distributive principles. It is not meant as an ideal theoretical principle, but rather as one meant to provide political guidance in this world.²⁵ Therefore, its ideal theoretical redundancy does not disqualify the principle from being politically relevant. Furthermore, theorists have sought to modify the core claim of limitarianism from saying that surplus money has zero moral weight to saying that claims to retain surplus money are lexically outweighed by other claims or, simply, that surplus money has little moral weight.²⁶ In this way, the limitarian principle can avoid some of the counterintuitive distributive implications to which critics point. In both cases, the defenders assume that implicatory redundancy and implausibility must be overcome to vindicate the principle. They assume, in other words, that limitarianism is a view about, and must be judged on, adjudication of conflicting claims – the demands of justice.

But limitarianism can, instead, be understood as a view about the bounds of justice. It is, we think, more plausible when understood in this manner. On the bounds version of limitarianism, claims to retain surplus money are simply outside the bounds of justice. Such claims are not up for adjudication. The claims of the wealthy to retain their excess wealth are like claims to have the state provide you with a fiancé or a new best friend or like claims of foreigners on statist accounts of global justice: they are categorically ruled out. They are automatically disqualified. Understanding limitarianism in this way makes it obvious how it is a distinctive view about justice.²⁷ Limitarianism may not say anything new compared to sufficientarianism and egalitarianism about the targets of redistribution – it, too, holds that wealth should be redistributed to ensure that everyone has enough, and so that inequality is ameliorated. But it says something new about the bounds of justice. Namely that claims of justice can only be made by those who do not possess surplus money, those who are not (too) rich. Claims to retain wealth (or to have more wealth) by those who are already rich should never be considered at the bar of justice. They should be ruled out categorically. Such claims, then, should not figure in our adjudication of citizens’ competing claims. Many of the strongest objections to limitarianism, we believe, can be side-stepped if we understand a limitarian threshold as a bounds claim, instead of a demands claim.

Some may find the sharp and sudden boundary drawn at the threshold implausible. While the boundary between when someone is and when someone is not your fiancé is a matter of broad agreement, the thresholds used in distributive justice are often not as clear-cut. Indeed, theorists tend to think of threshold concepts as having a significant grey area around them instead of marking a stark transition.²⁸ Importantly, however, the existence of a grey area is consistent with there being some demands that are out of bounds.²⁹ Even if such grey areas exist in practice, what the defender of thresholds as bounds must defend, is the in principle claim that having more than enough, can fail to be a matter of justice.³⁰ That is what we turn to below.

4 Thresholds as bounds

Thresholds lend themselves well to being interpreted as demarcations of the bounds of justice, and yet much of the discussion of them, including the most prominent objections, assumes they are meant to decide the demands of justice.³¹ This leaves open the possibility of defending thresholds as bounds claims more explicitly and, potentially, more successfully. More sense than they make as demands claims. When interpreted in this way, sufficientarianism is less structurally similar to its supposed rivals, egalitarianism and prioritarianism, and more structurally akin to views that delimit the claims of justice in certain ways – such as those determining the scope, site, and currency. At the very least, the mode of argumentation appropriate to bounds sufficientarianism is the same as that used for site, scope, and currency, and quite distinct from that used for egalitarianism and prioritarianism.

To see why bounds sufficientarianism makes more sense, we can return to the question that exposed the failings of negative thesis sufficientarianism and limitarianism as demands claims and see how much better the bounds version can respond to it. Posed to the bounds version that question can be rephrased as: What makes reaching sufficiency, or being too rich, significant such that it pushes one’s claims for additional benefits outside the bounds of justice?

We consider three possible explanations for why thresholds might categorically disqualify a claim from being a claim of justice, regardless of what else might be true of that claim or the distribution more generally. We take our lead from some subtle suggestions in the literature that have not been explicitly argued for in this way. They are satiability of the value of goods, satiability of justice, and conceptual engineering. In each case, we assess the potential for these explanations. These explanations may double as explanations for why sufficiency as a demands claim is unconcerned with supra-threshold benefits.

4.1 Satiability of the value of the goods

One reason that might be given for thinking about thresholds as bounds of justice is that the good or goods that justice is exclusively concerned with are satiable. According to this explanation, the threshold denotes a point at which people can derive no value from obtaining further resources and opportunities that justice is concerned with because the value of such resources and opportunities is sated.³² A person’s need for vitamin K, for example, is satiable – it can be fully met and, once met, providing this person with greater levels of vitamin K does not result in further value. If the distribution concerns vitamin K and 100 represents the level at which no more has value, there is no reason of justice to worry about the inequality between Gianfranco (at 100) and Jimmy Floyd (at 200). There are good reasons to worry about the inequality between Ole Gunnar (99) and Gianfranco (100), however, even though this inequality is much smaller. Jimmy Floyd’s claim to retain his excess vitamin K when faced with Ole Gunnar’s vitamin K deficiency is outside the bounds of justice. As is Gianfranco’s claim to equalise the distribution of vitamin K between himself and Jimmy Floyd. The threshold denotes a level of satiation, which categorically disqualifies claims above it.

Consider again Casal’s objection against sufficiency views: ‘when a natural disaster such as a tsunami strikes, it is the wealthiest who should make the greatest donation. Sufficientarians cannot accommodate such plausible convictions’.³³ On the face of it, this is quite a powerful objection to sufficientarianism, attacking its indifference to how burdens and benefits fall above the threshold. However, if we imagine a natural disaster that creates vitamin K deficiencies for a number of people, it is no longer obvious that, if two groups both have much more than enough vitamin K supplements, the shortfall must be alleviated by those who have the most. There appears to be no basis for prioritising among those with more than enough – everyone above the threshold simply has no claims to retain their additional supplements, regardless of how far above the threshold they are. So, it is appropriately a matter of indifference who must relinquish their supplements. The force of Casal’s objection, then, comes from conceiving of sufficiency as a view about the demands of justice. The objection loses force when the threshold is conceived as demarcating the bounds of justice.

However, for satiability of the goods to set the bounds of justice at a threshold of sufficiency or affluence, like those adopted by sufficientarians and limitarians, it must be shown that all the relevant goods of justice are satiable. One version of this argument holds that justice is concerned fundamentally, and exclusively, with ensuring everyone’s happiness. Proponents of this argument understand happiness to be satiable. In other words, they believe that there is a point where a person is fully happy, and beyond which she cannot become any happier, even by receiving more resources or opportunities.³⁴ A similar argument might be built around autonomy.³⁵ Several scholars have argued that autonomy is satiable or has an upper limit.³⁶ Its satiability rests on the idea that the elements of which autonomy is made up, those that contribute to a person being able to shape their life path, such as options, information, and deliberative powers, are satiable. In both cases, it would be a conceptual confusion to talk about inequalities above the threshold; if everyone were at the threshold of full happiness or autonomy, they could not be unequal and the idea of benefitting those who are fully happy or autonomous would make no sense. Doing so is impossible. The bounds of justice are constrained by the nature of the goods to which the demands of justice apply. The argumentative work is done entirely by the claim that the currency is satiable.

But this is a controversial claim. Improving people’s health and minimising the threats and risks they face, for example, are part of any plausible conception of justice. But these values are not satiable in the relevant sense. Even at high levels of health and security, devoting more resources could further improve, say, the average life expectancy of a given individual or provide them with even better protection from threats and risks. So, the answer to the question: ‘why should we disregard this person’s claims for further health or security improvements?’ cannot be that their health and security are sated and would not, in fact, be improved by more resources.

On its own, therefore, relying on the satiability of the values or goods in question does not seem very promising as a defence of bounds sufficientarianism. In a further blow to this form of defence of sufficientarianism, we can see that what is allegedly distinct about sufficientarianism disappears if it is only the goods that are satiable rather than the principle of justice. A prioritarian, for instance, can accept that justice is exclusively concerned with autonomy and that autonomy is satiable, without needing to refer to sufficiency at any point. People who have the least autonomy must be given priority in the distribution of further autonomy, but the bounds of justice end when everyone has full autonomy. A person has enough autonomy when they have full autonomy, but full is enough, so it leaves no room for thresholds or sufficiency in the description of the position. A substantive view about the demands of sufficientarian justice regarding autonomy (or happiness) would indicate a way to adjudicate between those that do not have full autonomy.³⁷ Understood as a demands claim, then, the only genuinely distinctive sufficiency thresholds are those that demarcate a shift, rather than a categorial ruling out.

Limitarians, who conceive of the limit in terms of monetary wealth, face an even greater task in justifying the threshold as a bounds claim based on the satiability of the good in question. They have to show that wealth is satiable in the sense that, above a certain limit, it cannot improve any justice-relevant metrics because they are already maxed out. Some economists indicate that, above a certain limit, additional wealth does not contribute significantly to a person’s happiness.³⁸ However, it seems deeply implausible that no further value could be gained from additional wealth at this point in terms of improving someone’s health, security, education, social status, or political influence (indeed, the ability of the wealthy to increase their political influence is one of the reasons for which limitarians object to great concentrations of wealth). The value to be gained from wealth, therefore, is not satiable in the relevant sense, and it is implausible to understand limitarianism along these lines. Indeed, in her more recent work, Ingrid Robeyns has moved away from the view that the value of wealth for the person is satiated at the limitarian threshold. Instead, she argues, there is no public or political value to retaining wealth above a certain limit.³⁹ As we shall see, this makes her argument more similar to the second type of bounds claim for thresholds, which we discuss in the next section.

Overall, the defence of these thresholds as bounds claims that rely only on the satiability of goods, therefore, appears unsuccessful. Satiability of goods may form part of the story of why no further justice claims can be made once someone is above the threshold – specifically, it may explain why inequalities above the threshold are unproblematic for some justice-relevant goods. On its own, as a full defence of bounds sufficientarianism or limitarianism, however, this is not a promising route.

4.2 Satiability of justice

A different argument for a threshold determining the bounds of justice also draws on the idea of satiability. Here, the argument focuses on the satiability of justice itself, not on the goods that justice is concerned with (which may be insatiable). Compared to the previous version, this argument is built around a more comprehensive satiable concept, which encompasses all justice-relevant values. Here, then, justice excludes claims because justice isn’t about more than enough: there is a limit to what you can demand from others. Those who insist on receiving additional benefits or retaining their wealth above this threshold insist on more than justice allows. Martha Nussbaum, for example, argues that a person can flourish if she has robust access to ten central human capabilities, which include health and security. Nussbaum’s argument does not run into the problem described above, however, as it does not claim that humans who have the capabilities to flourish could not, for example, be made healthier or more secure. Rather, her argument holds that improving someone’s capabilities in these areas means that they do not need further improvements in health or security in order to flourish – and that is all justice is concerned with.⁴⁰ The concept of flourishing is satiable.

Robeyns’ limitarian threshold, above which wealth has zero moral weight, is also grounded in the value of flourishing, and can be understood to play a similar bounds-defining role – particularly, in its more recent form.⁴¹ Huseby’s reasonable contentment and Axelsen and Nielsen’s freedom from duress work in similar ways, applying a satiable concept onto the values with which justice is concerned.⁴² Roger Crisp’s compassionate sufficientarianism could also function in this way.⁴³ All these aims are, ex hypothesi, satiable and they may be thought to set defensible limits to the bounds of justice, even when more of the currency of justice can, in principle, be obtained.⁴⁴

For these views, satiability is assumed in the definition of the master concept, justice, and imposed upon the underlying values. It is the conception of justice being applied to the values in question that is satiable, not the values or goods themselves. It rests on the idea that even if further value can be gained above the threshold, it is no concern of justice to obtain it. Someone who asks for further advantages above this threshold – once they are already flourishing, reasonably content, or free from duress – may gain value from such advantages, but they are not advantages one can demand of others in the name of justice.

A critic can always ask: ‘Why is justice exclusively concerned with compassion/ flourishing/freedom from duress/contentment?’ The burden of proof may seem to pose a problem for this version of the bounds argument, but, conceptually, underpinning the concept of justice in this manner is fairly common. For example, some hold that justice is, ultimately, about equal concern (Dworkin), equal freedom (Steiner), giving each person their due (Aristotle), fairness (Rawls), relating as equals (Anderson), or negating the effects of bad luck (Cohen).⁴⁵ Here, too, a critic could reasonably ask why justice is concerned exclusively and fundamentally with this value. Perhaps, it is a necessary and fundamental task for any theorist of justice to answer this deeper question, and some ways of underpinning justice may well lend themselves to understanding the bounds of justice in the form of a threshold. Just how promising the view is, therefore, depends on these further arguments, but there doesn’t appear to be any blanket objection to conceiving of justice as satiable. On the contrary, a satiable account of justice may be quite attractive, affirming as it does, that justice can be fully realised, in principle at least. Denying that justice is insatiable also seems to line up with intuitions about the need for our duties, in this case of justice, to be in principle actionable. If ought implies can, then an insatiable ought is a contradiction in terms or, at the very least, raises problems. So, the basic idea of justice being satiable is one that has some promise.

4.3 Conceptual engineering

A third possible motivation for thresholds as bounds is that it is an instance of conceptual engineering. Conceptual engineers seek to (re)define key concepts, so that these concepts better help us realise certain aims. These aims may be directed at natural kinds, as when marine biologists redefined the concept of ‘fish’ to no longer include whales or when astronomers redefined the concept of ‘planet’ to no longer include dwarf planets (like Pluto).⁴⁶ They may also be directed at social kinds, such as ‘race’, ‘parent’, ‘marriage’, ‘woman’, or, in our case, ‘justice’. Conceptual engineers are not merely trying to change the meaning of particular concepts, but also how people use this concept: how they classify the world, speak, think, and act.⁴⁷ When scholars and activists, for example, called for the concept of ‘marriage’ to be broadened to include unions between same-sex partners, they were not making a mere semantic point about what marriage properly is and means. They are also seeking to change how people think about same-sex marriage and how the concept of marriage is used. They are aiming for a worldly change.⁴⁸

As Sally Haslanger argues, concepts are tools that serve particular purposes.⁴⁹ They categorise and demarcate the world in ways that further certain forms of action and particular types of critique. Conceptual engineers seek to redefine social concepts because they believe this to be a necessary step in changing the social world. As Koch puts it: ‘for people to entertain different thoughts, form different desires, and act differently, it is crucial that they actually change the way they think about certain categories and how they mentally and linguistically classify objects, people, events, etc.’⁵⁰ Conceptual engineers accept that words and concepts have powerful connotations in the real world. It matters whether we call something a civil union or a marriage – even if the two were to entail the exact same rights and entitlements. And it matters, for example, whether we call a transwoman a woman and an adoptive parent a parent. What we call things matters because particular words and concepts come with strong associations and deep undertones: marriage, woman, parent, democracy, justice, etc.⁵¹

Conceptual engineering can take the form of broadening a concept. When feminists argued in political debates of the last two centuries, that non-consensual sex within marriages should be considered rape, for example, they were attempting to broaden the concept of rape. They sought to reconceptualise the social world in order to enable particular goals – here, ending sexual violence against women. Conceptual engineering can also take the form of limiting or narrowing a particular concept. For example, scholars of affirmative action or critical race theory sometimes argue that being a member of the relatively powerful racial majority (e.g. being white in many European societies), you cannot be a victim of racism.⁵² Similarly, many feminists argue that men cannot be victims of sexism.⁵³ This is not to say that white men cannot experience disadvantages that arise because they are white or male. The point, rather, is that such disadvantages ought not to be labelled racism or sexism because such a labelling would dilute or undermine the usage of the relevant concept.⁵⁴ Conceptual engineering ensures that concepts are defined in a way that serves specific goals, such as social justice. That they avoid conceptual inflation.⁵⁵ By targeting the way people categorise the world around them (regarding, e.g. what constitutes rape), the concept of gender serves as a tool ‘in the quest for sexual justice’.⁵⁶ Including specific instances of male disadvantage within this concept, ‘drains the concept of its political potency for diminishing or eliminating the historical control of women by men’.⁵⁷ On this approach to theorising, the concepts we use to describe the world should serve the aim for which we want to demarcate the world along these lines in the first place.⁵⁸

Thresholds in distributive justice can be understood as bounds claims along similar lines. On this interpretation, thresholds serve to narrow or limit the concept of justice by ruling out particular claims or disadvantages as not being appropriate for the label ‘justice’. Something akin to what is said about racism and sexism, then, may be said about the concept of (in)justice. We may deprive the concept of justice of its critical potential, or else diminish it, by using it in certain ways and in certain contexts. Sufficientarians and limitarians demarcating the bounds of justice might take a similar stance on the concept of justice – or, rather, injustice. If the concept of justice is to be conceptually engineered and serve the purpose of providing critical potential for the real world, then the claims of the very wealthy can never be claims of justice – simply for the reason that they are very wealthy. They have too much. Similarly, white people cannot invoke racism when they are disadvantaged due to their race because this would stretch and strain the critical concept of racism so that it would no longer be usable for its purpose. Its purpose is to criticise and politically oppose a certain form of systematic oppression that occurs on a background where, as Sylvia Law puts it: ‘Black people invariably note their race and white people almost never do’.⁵⁹

Here, we are reminded of Elizabeth Anderson’s ambition to link theories of justice with the concerns of social movements, such as the civil rights movements. If we are inspired by those movements directly, then we should be dismissive of any concern for the most advantaged in society – like complaints about racism made by white people, they should be dismissed out of hand, categorically ruled out, rather than considered and weighed at the bar of justice. Critics of sufficiency and limitarianism sometimes propose hypothetical cases in which it seems intuitively plausible that one should confer additional (enormous) benefits on the rich, rather than conferring (tiny) benefits on those just below the threshold.⁶⁰ Or, alternatively, they point to the intuitive implausibility of not distinguishing between the very rich and the mega-rich, normatively speaking.⁶¹ The conceptual engineering argument for bounds sufficientarianism and limitarianism, however, holds that these critics miss the point. Such critics assume that claims that could be made in a hypothetical world where everyone has more and few (or no one) have too little should be used to inform the way in which justice is understood in this world in which many have too little, and some far too much. But in the hypothetical world that critics imagine, in which we have genuine reason to worry about the plight of the very wealthy, the principle simply has no application: our concept of justice should be tailored to this world.⁶² Its aim, after all, is to serve our present purposes.

The limitarian principle seems particularly well-suited to this function. On this reading, it says: to best promote distributive justice in this world, the way in which people categorise, think about, and act upon particular claims must be altered. Specifically, they must dismiss, rule out categorically, any claims to retain or further their fortune made by the extremely wealthy out of hand. And they should do so without pausing to weigh and inspect those claims – just like the concept of sexism should not lead us pause to weigh and inspect the claims of disadvantages suffered by men. This is important because, in the real world, wealthy people make such demands frequently and passionately – and they have the resources to make these demands heard. To bring about social justice, to retain the critical potential of the concept of justice, we must change the way people think about and use this concept by conceptually engineering the underlying meaning and its use.⁶³

Critics will argue that the conceptual engineering response begs the question. How, after all, does it strengthen the case for disadvantages suffered by wealthy people not being injustices to add that this is so because such disadvantages should not be called injustices? For this argument to succeed, then, one element in particular needs elaboration. Conceptual engineering arguments must provide an account of why the critical potential of the concept of justice can be retained when including disadvantages below the threshold, but not when including disadvantages above the threshold. This does not necessarily mean that sufficientarian or limitarian conceptual engineers must provide a detailed account of why changing the concept in this way is likely to change the way people think. It is plausible that philosophers should be more concerned with designing and evaluating concepts than they are with the strategic implementation of a particular usage.⁶⁴ We will not attempt to settle the question of how and to what extent strategic aims (say, to promote social justice in the real world) are in conflict with pursuing the philosophical truth (say, to determine the most ‘correct’ meaning of a concept).⁶⁵ Nonetheless, a plausible account of such a conceptual basis and demarcation could provide the grounding for an argument for the bounds of justice to be limited, which differs in important ways from the two arguments outlined above.⁶⁶

The argument for conceptual engineering cannot stand alone. If certain disadvantages are to be deemed outside the bounds of justice (i.e. the disadvantages of the well off) because they might otherwise distract from actual injustices (i.e. the disadvantages of those below the threshold), we need an account of why certain disadvantages are more important than others. The argument from conceptual engineering for the bounds of justice to be limited, then, will need to be combined with other arguments – for example, one of the two outlined above. This may be a plausible and necessary component in a more ameliorative project in distributive justice, in terms of conceptual engineering.⁶⁷

5 Conclusion

In this paper, we have developed and explored a new way of carving up debates about distributive justice, separating arguments that demarcate the bounds of justice from principles regarding what justice demands. We have shown how this distinction maps on to several key debates in political philosophy. The distinction also helps illuminate debates about sufficientarianism and limitarianism. In particular, we have explained how thresholds in sufficientarianism and limitarianism can be understood as bounds claims, rather than demands claims. Understanding thresholds in this manner calls for fresh analysis, sidesteps some of the most prominent objections, and may promise a new way forward for our understanding of distributive justice, both on its own and in combination with sufficientarianism about demands.