In the Face of Austerity: The Puzzle of Museums and Universities


  • I am grateful for written comments on earlier versions of this article from Niko Kolodny, Mike Martin, Mike Otsuka, David Owens, Joseph Raz, Bob Sugden, Jay Wallace, Leif Wenar, David Wiggins, and an anonymous referee. Many of the ideas in this article were developed in my graduate seminars at the University of California, Berkeley and at UCL, and presented to audiences at the LSE, Oxford, Stanford, the Sorbonne, the EHESS, Fribourg, Berkeley, London, Reading, and at the Values workshop organized by Joseph Raz and Larry Temkin at Columbia. I am grateful for the helpful suggestions and comments I received from members of these audiences and many other people, including, in particular, John Broome, Ruth Chang, Joshua Cohen, Tim Crane, Meir Dan-Cohen, Jonathan Dancy, Tom Dougherty, Marc Fleurbaey, Kinch Hoekstra, Tom Hurka, Mark Kalderon, Frances Kamm, Cécile Laborde, Maggie Little, Julia Nefsky, Tom Pink, Peter Railton, Tim Scanlon, Samuel Scheffler, David Velleman, Alex Voorhoeve, Ralph Wedgwood and Jo Wolff. Research for this article has benefited from a grant from the Mind association.

The appearance of a £70,000 polished stone sculpture outside a new London hospital has not gone down well in some circles. The money could have paid for three nurses said the Sun, while the Daily Mail questioned how the ‘gallstone’ could ‘possibly improve health care’. BBC News, 25 August 2005

PUBLIC funding of the arts seems unavoidably controversial. There are, obviously, disagreements as to what is of value. But there is also the matter of whether funds could not better meet more basic needs. Often, this line of questioning is also pressed regarding public funding for universities, for example for minority subjects in Arts and Humanities. Critics do not necessarily deny that these are good in themselves, but they consider that there are more crucial necessities to be met. Indeed, there is something intuitive in this line of reasoning. It is tempting to think that the problem of the just society is focused on the question: ‘how should resources be distributed in order appropriately to meet vital needs?’ Yet this is not the only function that we typically suppose the state and other social institutions to fulfil. For we also commonly recognize that they have a central role in providing social goods such as museums, concert halls, or universities.

The role of the state in fostering some social goods rather than others, and so favouring some values at the expense of others, raises a number of challenges. Most obviously it conflicts with the simplest (and indeed too simple) conception of neutrality. If societies dedicate resources in ways which favour one kind of value over another that would seem to undermine an attitude of impartiality among conceptions of what is of most value.1 In this article, I shall assume that there is some way to meet or dismiss such neutrality-based objections, for I want to focus on a starker problem even than that, one which would remain even if those objections could be met, and on which philosophers have had oddly little to say. Few of us would welcome a life in which the various social goods were to be removed. Yet the allocation of resources to these goods certainly takes away some resources from the meeting of urgent needs. How can the just society pursue the aim of providing for such social goods when some urgent needs are unmet?2

In addressing this question, our focus will be on claims that can reasonably be pressed on one for aid, and claims that it is reasonable for one to resist. I shall address the matter by moving back and forth between the perspective of the individual and that of social policy. Different ethical outlooks have different conceptions of what it is for there to be a genuine demand on an individual or a limit to any demand; and these different conceptions interact in diverse ways with conceptions of the constraints on social policy. My purpose, in this article, is to make some headway on how these balance.

In Section I, I outline the relevant conception of social goods, before sketching the central problem. In Section II, I offer a theoretical framework for the problem and four possible responses—only one of these, I suggest, coincides with our intuitions that current distributions are not unjust while avoiding the implausible claim that certain values have an unyielding importance. But the real problem here is seeing how to maintain that position. Section III raises an interpretative problem about the notion of need and its role in generating or resisting claims. In suitable contexts, where someone has committed to one life over another, or where the opportunities for flourishing are determined by one range of cultural options rather than any other, they will have needs not mirrored in the demands of all humankind. These will be needs nonetheless. Section IV then explains the distinctive character and importance of shared needs. We can recognize the importance of museums and universities in social policy, I suggest, without having to appeal to the aggregation of interests across individuals. Properly conceived they are shared needs for us collectively.

I do not seek to offer here any decision procedure for weighing social goods against individual welfare. The aim rather is better to understand the place of social goods in our deliberations about social policy: we should move beyond the view of them as mere preferences, preserved if at all by purely partial concerns. Social goods are a necessity for us, a necessary means to a full life, but this necessity can only properly be made sense of in a social context. It is only against that background that we can properly grasp why it is not unreasonable to devote funds to such goods even, sometimes, at the expense of meeting some urgent needs.


What do we expect in just societies? Plausibly, a well-ordered society is one in which two traditions of social activity continue and flourish: the raising of funds collectively to meet the needs of the destitute; and the directing of resources towards certain values and ends so as to bring about particular social goods.

Universities, concert halls, and museums are obvious examples of such social goods, but a proper catalogue of them would be much more varied and diverse. For example, one might mention the manufacture of musical instruments; or the way in which a whole culture has grown up on making and drinking coffee in particular ways in Italy. Again, one might point to the development of artisanal trades such as viniculture and the limited production of particular types of cheese. And we can also draw attention to the provision of resources to develop local traditions of playing and watching sports: be that rugby in Wales, football in Spain, golf in Scotland, or baseball throughout the US and Japan.

As diverse as this brief sketch is, one can find common elements which justify the thought that there is a central concern or problem here worthy of attention. First, social goods develop according to a certain degree of local contingency. Second, while none of them is indispensable for a fulfilling life, it seems hard to imagine a good society in which no such goods are provided for, or a bearable society without the existence of at least some of the institutions which sustain them. Moreover, third, co-operation is required and demanded from many in order to provide these goods; so their provision spreads costs across a whole community, just as the meeting of urgent needs does. Finally, we should note that these institutions make things of value available, and, just as importantly, they provide for the transmission of an understanding of how these things are to be valued.

Taking all this together, one may say: while it seems hard to imagine a good society which lacks some such common institutions, still we cannot say of any given society why it must foster the particular set of goods and values on which it does expend shared resources. The brief sketch just given of the special status of these goods should help remind us that much effort is in fact given over in all the societies we know to the development and sustenance of social goods, and that this effort is often in tandem with, or even sometimes prioritised over, the meeting of urgent needs. This is a feature of society which does not immediately strike most of us as problematic. Yet if a concern with the proper functioning of society and its distribution of resources should be exercised solely with an eye on need, then such complacency is problematic.

Consider a somewhat artificial example.3 There are some diseases which lead to death in teenage years, or shortly after. We expend a significant amount of resources on researching cures and remedies for such fatal, early illnesses. But it is quite conceivable that were we to devote an even higher level of resources, we might discover a cure for one such disease much earlier; and thereby, perhaps, advance the successful treatment by five to ten years. Think now of a very rare illness: one which affects one in a hundred million people in their early teenage years. There is some teenager somewhere who as things stand will die soon. However, if we expended the resources which communally are directed towards museums, or arts, or pure mathematics in universities, and focused it solely on the relevant medical research, they would survive beyond their mid-teens.

If any individual has an urgent need, surely this teenager does.4 So does this teenager have a claim with priority over social goods? When faced with this example, most people's reaction is that we are not required to abandon resourcing the arts and universities in order that the teenager should survive; however unfortunate their plight. Perhaps the idea implicit in this reaction is that there is a limit to the demands any one person can make on the rest of us, and that that limit is in play even in relation to the distribution of resources towards things which are not needs stringently conceived.5 The difficulty is to find a plausible manner in which to think of this sort of limit. Even if there is no precise boundary or algorithm which one can define, we must be able to give a more helpful account of what reasons are in play in this kind of case. As yet, this seems elusive.

If claims of urgent need are a priority for justice, then we should sacrifice social goods in favour of urgent needs. Most of us reject the consequent, but deny that social goods are more important or have greater priority than the urgent needs of others. How, then, can we justify our rejection? This, briskly put, is our problem.


Let us look at the example more closely. One reaction is immediately to suppose that with social policy, numbers are always germane. As the example is specified, the disease is relatively rare and those who will be adversely affected by re-distribution are great in number. So, one might think, this is simply a matter of balancing the demands of the many against the misfortune of a few.

However, it is not clear that aggregative concerns impact on our social policy judgements in quite so straightforward a manner. Consider T. M. Scanlon's example:

Suppose that Jones has suffered an accident in the transmitter room of a television station. Electrical equipment has fallen on his arm, and we cannot rescue him without turning off the transmitter for fifteen minutes. A World Cup match is in progress, watched by many people, and it will not be over for an hour. Jones's injury will not get any worse if we wait, but his hand has been mashed and he is receiving extremely painful electrical shocks.6

Many of us (if, perhaps, not everyone) feel that even if a hundred million people were watching this game, we ought to interrupt the transmission to save Jones from his hour of severe pain. And it is tempting to move to a more general moral: that we should reject putative moral principles which require us to aggregate benefits to many individuals (even where those benefits themselves may be quite small), and weight the sum against the harms or burdens on others. Whether that is so is a matter of some keen debate. And that matters are not simple in this domain is reflected further in the problem case on which we are focused.

There are important differences between the two examples. Depriving the young of school or university education is not on a par with missing out on a football game, even a thrilling one. There is also a contrast between the one-off occurrence of an accident and adoption of general social policies. However, given the remaining similarities in the distribution of burdens and benefits, and in the ratio of numbers on each side, the question remains: why do our intuitions go in opposite directions in the two examples?

The first thing to note is that Scanlon's example has several layers of complexity, and that might bear on the disanalogy.7 We might feel that giving in to our desire to continue watching the game, demanding no intervention should interrupt the broadcast, is tantamount to actively harming Jones. The thought would be that the difference in attitudes towards Jones and the teenager is underpinned by the distinction between a stringent duty to refrain from harming others, and an imperfect duty to help people in need. However, given the extreme plight of the teenager, this would, at best, be only part of the explanation of our difference in attitudes. If as many as one in two, or three, children died at such a young age, our reaction would change. We might think that our society is not well-ordered if it lets this happen; we would feel that children are victims of neglect. Yet in our actual example, most of us feel disposed to support the status quo. Few would be inclined to consider that this is because the need of the teenager is outweighed by the number of people who enjoy education and art (or indeed by the value of art or education). Yet we doubt that all these funds must be redirected for research to cure his illness.

If the numbers involved across these examples can give rise to such conflicting intuitions, we might do best initially to leave aside considerations of numbers entirely, and focus first on the kinds of objections individuals would have to each of the potential policies. A familiar version of this procedure is often called the ‘Complaint Model’. This compares the respective objections that each of the people submitted to particular policies would have.8 It has at least two appealing features. First, it is individualistic insofar as it examines and weighs the complaints that each person might have against each of the policies available. This model also has the result that we pay particular attention to those who do worst under each of the candidate policies: we try to adopt the policy with the least significant objection to it. So of those available to us, and subject to some constraints such as respect for rights, or considerations of fairness,9 the correct policy is that in which the worst-off still would do better than the worst-off (not necessarily themselves) would do under each available alternative policy.

We therefore operate by pairwise comparisons. In Nagel's words: ‘we try to include each person's point of view separately, so as to achieve a result which is in a significant sense acceptable to each person involved or affected.’ Nagel comments:

Where there is conflict of interests, no result can be completely acceptable to everyone. But it is possible to assess each result from each point of view to try to find the one that is least unacceptable to someone than this alternative is to anyone. The preferred alternative is in that sense the least unacceptable, considered from each person's point of view separately.10

Of course what this model directs us not to do is to aggregate complaints. In our second example, we do not aggregate the cost (in terms of sacrificed pleasure) that the many viewers of the football game are incurring. Their combined displeasure is not allowed to outweigh Jones's hour of excruciating pain. In other words, we build in an Individualist Restriction. Parfit, to whom we owe this expression, defines it thus: ‘[i]n rejecting some moral principle, we must appeal to the principle's implications only for ourselves and for other single people.’11

Using this conception, it is easy to see how Jones would have a reasonable complaint against our continuing to enjoy our Sunday afternoon entertainment at the cost of an hour of excruciating pain for him. However, it is much less easy to see how this approach can lead to a contrasting verdict in the example of the teenager. It seems that the loss of a number of years of life in one's early years is an even greater harm than suffering a traumatic pain for an hour. True, one may suppose that a liberal education or the opportunity to spend weekend afternoons in an art museum is a more significant benefit to people than watching a football match. But still, as we are now to understand the Complaint Model, at least one of the undergraduates must be in a position in which depriving him or her of a liberal education (as opposed, say, to a vocational training which will help medical research), or the enjoyment of the arts, is a worse harm than the loss of years of life is to the teenager. As important as art and knowledge are to some of us, it is difficult to put anyone in the position of claiming that these are more important to them than someone else's life. Yet unless we are prepared to make this claim, how are we to avoid the consequence of the teenager's claim being the strongest, thus rejecting any principle which does not save him or her?

What then if we lift the restriction? Those attracted to maximizing models of moral considerations certainly find the Individualist Restriction puzzling. Among the many critics of this idea, Parfit puts the matter particularly sharply: using cases where we can either save one person or several people from a similar type of harm, he argues that the Individualist Restriction leads to the wrong kind of result, and that a more plausible understanding of moral duties would do away with it.12 Might the same be said with our problem? We can explain our intuitions in the case simply by removing the Individualist Restriction and recognizing the conclusive force of aggregated burdens or benefits which groups of people together bear.

Before we rush to endorse this solution, we should reflect on whether we really are happy to offer this kind of justification of the policies we pursue to the teenager whose life is at stake. Advancing aggregative considerations to explain our policies to the teenager just seems formidably callous. And this doesn't seem to be a matter of mere politeness or diplomacy in spelling out the real reasons. If the sacrifice for each seems so superficial in relation to the teenager's plight, piling up the number hardly makes our resistance more acceptable.

The problem is compounded because the alternative (i.e., to claim that it is unjust not to make these sacrifices) doesn't seem the right thing to say either: we do not feel that a just society must systematically redirect all of its available resources to meet claims of need for survival.

So we may doubt that the Complaint Model, even with due modification, would allow us to distinguish between the case of the teenager and the case of Jones. The question is: beyond their structural differences, are the two examples the same; are we led to the conclusion that it would be wrong to refuse to sacrifice certain cultural goods in order to give a few more crucial years to the teenager?

As suggested, one option is to claim that for each student currently enjoying a liberal education, the benefit he or she gains from this aspect of life just greatly outweighs the cost to the teenager. Any policy which led to the deprivation for any of them of this education would leave them worse off than the teenager is now under the status quo. Access to art and knowledge are just too important to be sacrificed in this way.

Such an insistence on the importance of these values indubitably sounds as odd as the demand that the resources should be redistributed. We surely don't want to insist, even given our position in the business, on the central importance of learning and knowledge. It seems as if, when we have to compare the importance of the opportunity to read a significant piece of literature, to enjoy a superb painting, or have one more life saved, there is little choice but to opt for the third of these.

And at this stage, it may then look as if the only way to rescue the status quo without over-inflating the claims of universities or museums is to appeal after all to some aggregative considerations. Perhaps it does matter in the end that the teenager is one of so few such afflicted members of our society, and those with the opportunity to benefit from education or from the arts are so many. As already stressed, the difficulty of resting here is the recognition that we already have from Scanlon's example that if aggregation can come to bear, it cannot do so directly in all cases.

So far, then, we seem to have three options:

  • (a) to recognize our current attitudes towards the fostering of liberal education and museums as unjust, and to direct all resources towards activities which can best meet the urgent needs in our societies;
  • (b) to resist the claims of the teenager by insisting on the importance of art and knowledge and claiming that these values have a centrality which trumps that of an individual human life; and
  • (c) to drop the Individualist Restriction, and say that the number of claims to museums or universities outweighs the claim of need of the teenager.

As I have already indicated, though, each of these options comes with a certain cost; none seems entirely compelling or to respect the complexity of our intuitions here. So the question remains, is there some other way of explaining our attitudes in this case? That will set the parameters of our remaining discussion. In particular, I want to look for an account of our considered judgments here which is:

  • (i) sensitive to the Individualistic Restriction;
  • (ii) not such as to claim the greater importance of values such as education and art over life; and
  • (iii) consistent with the kind of policies actually pursued by current state or social institutions; that is, policies under which a significant number of resources are directed towards such social goods as education and the arts, and in a way that does not guarantee benefit for all individuals from whom costs are levied within society.


Let us now make more explicit a couple of elements left intentionally unclear in the Complaint Model. When we engage in a pairwise comparison of individuals' situations in relation to different policies, there are different things we could be comparing.13 On one conception, working out the cost of a policy is a relative matter, contrasting an initial situation and the outcome. So one might think that an individual had a complaint against endorsing the new policy because the resulting situation would leave them too badly off relative to their actual possession of resources, and so would be too demanding of them. If we take this way of thinking seriously, then the grounds for reasonable rejection will in part be comparative across outcomes.

On the other hand, we might measure a person's welfare relative to some appropriate standard. We do not first have to compare a situation with any other possible outcome (i.e., an outcome determined by some different policy about distribution from that which holds in the status quo) in order to determine whether the person is in an unfortunate position and has claims on us. We need only to know that in this situation the person in question is below some given standard. And it is this fact about the situation which would ground the verdict that a policy which resulted in this outcome would be reasonably rejectable by that individual. Talk of rejectability may however mislead, and suggest a particular decision procedure. The point, rather, is that there is an absolute standard according to which we rate the urgency of a person's claim.14

In turn, a question arises as to what the baselines used for comparison are. On one way of doing this the baselines used are determined just by average welfare, so that costs and benefits are a matter of how a subject moves relative to the overall average welfare. An alternative way of conceiving of them, though, builds in more structure to the idea of a complaint: claims, we may suppose, are distinctively grounded in the needs of individuals, not in their preferences. In identifying the claims of individuals we need to look at their opportunities to lead a flourishing life, and to possess the means necessary for that. And here we help ourselves to a substantive conception of what it is to lead a good life, rather than thinking of it simply as a formal notion filled in by appropriate values for an individual's preferences. Such a substantive conception may draw on the contingent possibilities that arise given the specifics of culture and history that individuals find themselves in. Hence, the claims that any individual has, in the actual world or in any possible world resulting from a given policy, are always relative to a potential good life, actualisable in their historical circumstances.

Given this richer conception of the baseline, we arrive at a different picture of the costs that individuals face. Rather than simply asking whether an individual will end up with fewer benefits under this policy than that, we can ask whether they will be left with the means themselves to live a good life. That is, rather than simply conceiving of them as beings with a certain set of preferences and resources, we should think of them as having needs which form the basis of claims against others, including claims which balance or block the claims of others. With respect to any bundle of resources in excess of the minimum required for living life well, we might think that an individual has no good complaint against a policy which would deprive them of those resources. On the other hand, we can, in the context of a pairwise comparison, point out that any policy which deprived a particular individual of the necessary means to a good life in order to provide another individual with their needs would be self-defeating from the point of view of such distributive policies. An individual faced with such a demand could reasonably reject the policy. The basis of this rejection we might think of in terms of the very moral concerns which would give such a policy its point in the first place: that such policies answer to the demands of meeting needs which all of us can recognize.

We thus operate with a different conception of reasonable rejection. We start from the fact that we tend to think that someone's claims of need issue from their situation in the actual world—how far their resources fall below a certain baseline. That is, someone has claims on the rest of us where they have need of things that are necessary for life to go well. In mirror form, we might contrast the resources others have beyond what they need with that which they have which meets their own needs. If another's need is stringent enough, it is not reasonable to resist their claims if one has suitable surplus resource; but where it would deprive one of the necessary means for flourishing oneself, it is. (Again, there is a further question here whether, in terms of policy making, there can be further comparative ranking of urgent claims.)

One element of the puzzle with which we closed the last section was the question of how any of our undergraduates could resist the demands of the teenager without claiming the greater importance of philosophy or art over the teenager's life. The supposition that we would be committed to such a claim in the face of the teenager's needs itself presumes a particular conception of the Complaint Model introduced above. We can now see that we need not be committed to any such conception, and that on reflection it should be rejected.

If we just think that the cost in question is determined by some quotient, then the deprivation of education for one of the undergraduates must be greater in disvalue than the loss of life for the teenager. And it is natural to think that this calculation can only hold if impersonally (allowing now for comparison across people) the value of an education to someone is greater than the extended life of a teenager. But we can avoid this demand if we focus instead on the grounds of rejectability that an agent may have. For the basic idea is that it is reasonable for someone to reject a claim against them where meeting the claim would deprive them of the necessary means to live life well. This consideration does not entail that the ground of rejection have greater weight or value than the claim it forms the resistance to: what matters is the relation it bears to the person resisting the claim, that it be a need for them.

The point can be made vivid by focusing on certain special cases involving psychological impossibility. We find it entirely understandable if someone cannot bring themselves to act in a way which costs them a finger to save another's life. Their attitude here is not well-modelled by saying that they value their little finger more than the life of the other. Rather, the kind of sacrifice here, breaching the integrity of one's body, is something that is psychologically impossible, or near to, for most people. We have a background concern for the preservation and integrity of our bodies, and this concern is not normally up for comparison with other things we value; rather it structures what is possible or impossible for us to do.

Note that this is not to say simply that the value of one's finger is incommensurable with the value of a stranger's life. Nor is the thought that there is a privileged realm for partial or prudential concerns which the demands of others cannot breach. The psychological impossibility of sacrificing part of one's body holds even when we are concerned solely with the prudential. After all, this is why we find the action of a man who sawed off his own arm in order to avoid starving to death both fascinating and heroic.15 For as prudentially essential as the action might be, and as rationally commanded on one conception of decision theory, still we recognize that so acting may well be beyond what we are normally capable of doing.

So the idea here is that questions about what it is reasonable to ask of someone must take into account what is practically, and hence psychologically, possible for them to do. Someone may resist a certain demand because it requires giving up on something that one cannot sacrifice, that is the flipside of its being a necessity for them. And these aspects of the reasonable and the unreasonable cannot simply be reduced to questions about how much we value or disvalue certain outcomes. Hence we cannot conclude from the fact that someone reasonably resists a sacrifice that the cost to them of the sacrifice is of greater value than the benefit which would accrue to others.

The prospect of sacrificing a finger offers a vivid case for grounding the resistance of an otherwise worthy claim. But it seems to fit the model of basic or urgent needs as a ground of resistance rather better than the social goods that we are principally concerned with. So how can we generalise the moral of this discussion?

One who cannot sacrifice their finger for the life of another is not claiming the indulgence of caring more about a finger than about a human life. The integrity of the body puts a limit on what someone can do. This illustrates a point which has wider application: that just as the grounds for making a claim are principally one's needs rather than preferences, so too one's grounds for resisting a claim are needs rather than preferences. While we may rank preferences in terms of their importance or urgency, there may be no such ranking for needs—claims for redistribution occur in circumstances of scarcity, but not such scarcity that there are no resources which otherwise would be superfluous to some people leading a flourishing life. The contest of claims is therefore over this superfluous element.16

We have here a contrast between the resources that an individual has which are superfluous to their leading a flourishing life, and those which, in the circumstances, they cannot flourish without. One undermines the shared purpose of policies aimed at meeting the demands of the benevolence we all share, if in succumbing to those policies one is required thereby to sacrifice the good life one seeks to help others achieve.

So we can avoid the move which seemed compelling given the account above of the Complaint Model. It is just not true that any of the undergraduates who would reject the policy of redistribution to the teenager must see his or her education as more important than a human life. Rather, the thought is that there are normally limits to the claims that others in need can make on us. Where meeting the claims of another would entirely disrupt one's life, it is not generally taken to be a requirement that one answers to that claim. And it is in this spirit that an undergraduate can resist the claim of the teenager without having in addition to insist that there is more value in his or her gaining an education than in the teenager gaining more years of life.

We should note, however, one important issue which will arise in the broader case. When we are concerned with those needs protected through our psychological makeup, it makes little sense to claim that we have chosen to value these things over the concerns of others. What our discussion above illustrates is that on the whole these are things which just aren't options of choice for us at all. The same cannot be said in general in relation to any specific need one has which is variable across people. For often such needs arise in the context of someone having pursued one kind of life over another. Hence we can locate within a life choices made where things could have turned out differently; on that alternative history the thing in question would be no need. This raises a further challenge: the opponent may try to locate the question of policy acceptance or rejection at the point of opting for the life in question, and ask how the individual can opt for that life over another possibly rewarding one without supposing that the goods it involves have a greater value than, say, the life of the teenager. In turn, in responding to this challenge we will see the significance of social goods being peculiarly social.


I have suggested in the discussion above that we move beyond the Complaint Model. That approach envisages the picture of reasonable rejection as a kind of decision procedure, ranking people's claims through pairwise comparison so that we can sort these claims from the most to least pressing. In the Complaint Model, an increased level of urgency gives one claim more weight than another, and hence allows us to see how competition may be settled between different perspectives. I have proposed instead that we might think of the pattern of reasonable rejection at one further remove from policy, as telling us about the normative grounds from which decisions would be drawn.

Our conception of a baseline, constituted by what is needed for a good life, allows us a first step in the direction of the explanation of our puzzle. It provides a suitable negative response back to the teenager: we don't have to find a more pressing claim in order to resist the demand that resources needed elsewhere should be directed towards his cure. The claims we can make and the claims we can resist each draw on what people need in order to live their lives.

Each of us has a concern with his, or her, own well-being, and something which is necessary to that well-being can be protected from the claims of others. We do not suppose that this holds because each person has to think of him or herself as more important or more valuable than others. We simply have to recognize that there are limits to what can be asked of people. So far we have suggested that a parallel line of defence should be available for the arts and humanities: it is not that we have to suppose that these goods are more valuable than a teenager's life in order for us to defend the resources that flow to them. Rather we need only argue that they form a necessary component of the good lives that we can have, and hence that the sacrifice of them is demanding too much of us.

At best, though, this gives us a stand-off between the claim of the teenager which, I have argued, can be legitimately resisted, and the claims that students have for education, which they cannot be forced to forego, but still are claims of need which too might be found too demanding (as indeed the teenager presumably would find them). That is, just as the picture I have offered does not provide a decision procedure, so too it doesn't offer us sufficient to answer why we are happy with the status quo.

Furthermore, our strategy will succeed only if we can both show that the arts and humanities are a need for us, and also show that in treating them as such we do not have to suppose that they are more valuable than human life. Otherwise we will be back with the problems with which we first began. Now, there is a line of thought which makes it sound very odd to suppose that culture or education really are needs or practical necessities in the kind of stringent manner one might suppose is required to explain away resisting the demands for sacrifice from others. In what sense does one need to have visits to the Louvre, or need to learn about the poetry of Walt Whitman?

At least initially, one might think that this response misconceives the range of needs we have. As we noted at the outset, there are invariant needs among human beings which pretty much all members of the species homo sapiens share. And these needs seem closely connected to the kinds of psychological incapacities we highlighted in the last section. But invariance across human beings should not be thought of as the key test of need: what rather we want to isolate is how, in the relevant circumstances, an individual is pursuing a worthwhile and flourishing life. Those things which are a necessary means to that kind of life will indeed be among the individual's needs. To that extent, simply given the kind of life in question, the status of art and culture as necessary explains why they are not a required sacrifice.

All the same, the resistance so far highlights merely that we could not require, as things stand, of any of the undergraduates this sacrifice. But it doesn't rule out thinking that the sacrifice would be the heroic thing for them to do. And hence, doesn't rule out us thinking that in an important sense the world would be a much better place if the students could only bring themselves to the right kind of motivation that would direct them towards lives as medical orderlies or insurance accountants, dedicated towards the health industry's goal of extending the life of each to an adequate length. It may not be possible to require this heroic sacrifice from the present generation, but maybe the expectation to study at publicly supported arts and universities would begin to disappear if many young people chose in this way, and that would lead to the possibility of switching funds towards meeting urgent needs rather than supporting liberal arts education, without anyone having to give up what is a necessity for them. Inspired by this vision, we can imagine one of the students campaigning for the following proposal: ‘in fifty to a hundred years we will withdraw public support, and encourage all charitable donations away from museums, professional operas and ballets, and so on, and instead use the money for medical research.’

Yet, merely to describe this outcome, I think, is to sketch a position that we do not find attractive at all. It is not merely that it would require heroism of the young now at university to sacrifice their education, and that this is beyond the call of any duty for them. It is that there would be something misguided and ill-conceived in the sacrifice of all social goods. And the account that we have offered so far does not explain why one should have that reaction.

The extra element which we have yet to capture seems to be one directed towards the status of the social goods themselves: the goods of liberal education and the arts. The method of pairwise comparison captures these concerns by factoring in for each individual the benefit they receive from that good. And this itself, we should note, is a complex benefit. For there is both the benefit that a given individual will receive through his or her education, or through the visits he or she makes to various museums throughout the world, and, in addition, there is the benefit that accrues to one simply through living in a society in which people gain such an education or go to museums.

Nonetheless, the added concern we have here is that even when one takes these into account for any given individual, one still will not properly have captured the way in which social goods matter to us. In addition to the individual benefits distributed through society, there seems to be a good which we share collectively through there being educational establishments, museums, fine buildings, or fine wines. The pairwise comparisons as they are envisaged in the Complaint Model of necessity would leave any such value out of the account for determining acceptable or unacceptable policies for a society to follow.

If this is right, then one element of the case that we would want to retain and explain is the thought that it is unreasonable of the teenager to demand these resources from us, when the consequence of so directing the resources would be the loss of all of these social goods. Our concern here is that the grounds of rejectability are not exhausted by the concerns of any of the individuals who face a life choice between either partaking in social goods such as higher education, or devoting themselves to the urgent needs of the few. But where, in an individualistic picture, can we make room for any such additional concerns?

The difficulty we saw with insisting on this is that the grounds for this verdict would seem to require the aggregation of the numbers involved to have a value weighty enough to balance the urgent needs of the teenager. We noted two problems with this. First, as in the Jones case, it seems that there are cases where such aggregation is not intuitive, so we need to know what the key differences are between the examples. Second, appealing to the sheer numbers involved didn't seem the right kind of ground to offer back to the teenager.

The appeal of aggregation depends on the thought that all that can bear on the question of permissibility for the distribution of resources are the costs and benefits to individuals. If one doesn't value art and education more than the life of a teenager as such, then the only way that the benefits to any given individual of a liberal education could excuse the cost of life to the teenager is through being aggregated with the benefit accrued to other individuals as well, or so the thought goes. But that seems a narrower base of concern than is really in play. We do not value liberal education solely for the benefits that individuals separately enjoy, even when we consider the aggregate of this benefit. We also think that there is a shared social benefit in life being so organized that people are inducted into such education, and others are dedicated to preserving these institutions of learning. So the question rather is whether there are social goods or values, to be justified to all regardless of whether they benefit from them, and relative to which we can ask whether the claims of an individual impose too much on us.

Again, now in a social context, we can question whether this is solely a matter of comparing values. Are we claiming that considered jointly we value art and education more than the life of a teenager? If so, why is that more acceptable than any one of us insisting on the importance of the good in his life over the length of the teenager's life? In the individual case, we avoid having to press this last question by drawing a contrast between what is necessary for someone, and so beyond the claims of others, from the other resources that they possess. Is there a comparable contrast that we can make at a social level?

There are many ways in which people can choose to live their lives. Among such are those which require joint endeavours and the pursuit of common ends. It cannot be a requirement that the only end we maintain together is the pursuit of urgent needs. That we have a society together involves the pursuit of commonly recognized goods such as the developing of traditions of education, violin making, cheese making, or sporting excellence. Given that in fact we live in a society which has flourished through engaging in these activities, we jointly have an interest in these kinds of life carrying on. There can be no requirement on us from the urgent needs of any individual that we do not live in this kind of way. That this is a proper way of living together must be something that any reasonable person can recognize. It is not just that, in asking of any given individual that they should opt for a life of a hospital orderly or of middle management, one would be demanding more than it is proper to ask. It is also that, in hoping for a world in which individuals all made that heroic sacrifice, one would be looking for a world in which the proper means of living together would have been taken away.

To put things this way is still to hold to the Individualist Restriction, that all policies must be justifiable to each individual. But what we have resisted is one conception of what such justification consists in: that it should merely track individual benefits and costs. We have modified this picture in at least two ways. First, we consider not just whether something is a cost to an individual, but whether that cost involves depriving them of a necessary means to flourishing. Second, in understanding the demands that each of us can make of others we look not just to the needs of individuals, but also to the needs we have jointly or collectively.

Now we can return to Jones and his one hour of excruciating pain. In that example, people are enjoying a football game, but it is only their individual hour of pleasure which is in question—not the very continuance or disappearance of such shared activity. Only an aggregation of individual pleasures can outweigh Jones's pain, and this explains the overwhelming feeling that it would not be right to prolong his agony in order to see the conclusion of the game. The structural differences between the two cases turn out to be multiple, but each in different ways to reflect important aspects of the social goods involved. What matters is not mere benefit to individuals on a given occasion, however much this might be aggregated. Rather, what matters is what, in the circumstance, turns out to be an essential element of life's going well. While this may focus on aspects of any individual's flourishing, there are also elements of this which we can only properly bring into focus from a social and a plural perspective.

Social goods vary across different societies, and people engage with them differently. A particular social good becomes a need for someone once they are already set on some type of life. To exploit an example by Joseph Raz, a concert pianist with broken fingers may be able to make a new life as business accountant, but they nevertheless lose the life they have. Preserving their fingers is directly dictated by a concern for their personal needs.17 How does the same thought play out at a social level? The variety of good lives available is more restricted once we take this level into account. Some goods are neither identified nor accessible in certain contexts. What may seem optional at an individual level is not so at a social level: past choices have already committed all resources to some goods and not others. We need to allow that there is no other than a historical justification as to why these particular goods matter around here. The particular goods and values that we are so engaged with need have no special status beyond being the ones that we here have come to sustain. But for us to flourish together, we must be so committed to some goods. The question we are concerned with is: once we recognize this fact at a social level, how can we accommodate it in an individualistic perspective.

We have so far talked in terms of what we collectively or jointly have a commitment to. This might suggest the picture of supposing that entities other than individuals are relevant to political or ethical enquiry: that we need to think about groups or communities and their rights. A commitment to such entities does not involve aggregative methods: we need not think that a group has benefits just through the fusion or aggregation of the benefits to each of its composing members. But still, thinking in terms of such entities and, moreover, supposing them to have an appropriate ethical status of making or resisting claims would seem to take us beyond a properly individualistic perspective consonant with the Individualist Restriction.

An alternative approach to these inflations is just to pay proper heed to the significance of the grammatical category of number, and recognize the role that the contrast between what one does singly, just as an individual, and what one does jointly, with others, is marked in our ordinary thought and language. That is, some things hold plurally of us, in a way not to be understood simply in terms of what is true of each of as us a single individual. It is the bleakness of a society whose institutions aim solely at servicing the several needs of individuals without attention to what we can jointly engage with which we should recoil from in the teenager's demands.

How is this talk of things holding of pluralities, and of something being plurally true of some of us, rather than true of each of us singly, a way of avoiding a commitment to groups as entities? Consider an example. We form a circle uniting our hands around the castle. Of none of us individually can it be said that ‘x surrounds the castle’ holds of them. The predicate of the verb ‘surround’ is intractably plural. In analogous manner, the importance of social goods cannot be fully explained by the benefit that each of us derives from them. In recognizing that something holds of some us together, namely that we surround the castle, we do not introduce some new entity into the story, but only recognize a range of ways in which things can be true of each of us, or of us all together.

Perhaps the most direct way to spell out what the idea that we are to consider ourselves as a plurality gives us in this context is to contrast ‘this is good for us’ and ‘this is a need for us’. From an individualistic perspective, ‘this is good for us’ can only be true if the sentence is read distributively, and the good in question is taken to be good for each of us individually. For were we to suppose the sentence collectively true while distributively false, that would imply that the good would be present without having to be good for any individuals in particular.

On the other hand, it is more plausible that ‘this is necessary for us’ can be true when read collectively consistent with individualism. That is, it might be the case that something is necessary for us taken together collectively even though it is not true of each single individual that they can say that the thing in question is necessary for them. For, with respect to each individual we might deny that for them on their own it is strictly speaking true that the social good in question is a necessary means to a flourishing life—they could opt to live differently, and the social good would still flourish through others' engagement with it. That the social good is a necessary means to individuals' pursuing the good life may be reflected only at a social level where we can consider the coordination of different individuals' plans. Acknowledging this is not to grant that there are any entities other than individual human beings whose good we need to take into account. So there are no more entities in the story than those with which we started.

An individualistic perspective does not ascribe any distinctive moral status to social groups, corporations, or collectivities as entities which can make claims. In common with this impulse, we have not introduced any such entity. But the individualistic perspective will go wrong, I suggest, if it fails to recognize that there is an important status to those things which arise from our common or joint action, when we act collectively or co-operatively. In taking as primitive talk of us ‘plurally committing to something,’ we can properly recognize that level of fact.18 ,19

There is, furthermore, a difference between individual sacrifice and sacrifice at a social level. Once a student has decided to make the sacrifice to give up higher education, such an education is no longer a need for that particular individual. They now live differently, according to what is their reasoned life-choice. But there still is for us as a whole the need of that good, and that reflects how that good bears on us, independently of any person's given choice.

It is through coming together and partaking in idiosyncratic but necessarily joint enterprises that we find meaning together. It cannot, therefore, be a proper demand of justice or of morality that this should be forsaken in other than the most extreme circumstances.

Where the example of Jones reflects that we do not have simple aggregative impulses in our moral thought, the example of the teenager shows that we nonetheless have a keen eye on what goods society must provide if we are to live well together.


The questions raised here at a level of great theoretical abstraction echo concrete and persistent political complaints. It is a common thread of journalistic commentary, for example, that NASA and in particular the race to the moon was an inappropriate use of federal funding given the level of destitution within the USA. And more recently in London, there has been the sharp criticism of a hospital for its purchase of a polished stone as a monument at the public entrance of the hospital. Critics have been shocked at the relatively modest cost of the work of art and have suggested that the resources expended on this could better have been directed at the saving of lives.

So far, one moral to draw from our discussion, I suggest, is that these criticisms do not really wear their reasoning on their sleeves. Although the debate seems to be at a certain level of abstraction about how we should use social resources, we cannot really see it as that. Those who criticize the Space Race simply feel that this was not a form of engagement which had value in itself. At best, instrumental political and technological advantage could be had through these means, but the activity of walking on the moon had nothing in itself which ought to have been of appeal. Likewise those who detest the polished stone do not really reject all forms of public art in institutional settings: they do not recommend the removing of Nelson's Column or the Tower of London. Rather they find the work of art in question impenetrable or of low quality and hence not appropriate for the hospital. So the critics do not uniformly accept some general principle about the proper use of public funds which neatly excludes just the examples they highlight for criticism. They do recognize the need for some activities which do not meet the narrowly defined purpose that they suppose public funds should be directed at.

That, of course, does not explain or explain away the political force of challenging some social enterprise through contrasting it with unsatisfied needs. And this political challenge is one that we need to take seriously independent of the theoretical puzzles we have been concerned with here. It is notable that in many Western countries traditional social goods such as higher education come under scrutiny and challenge, and in public debate it is increasingly common to see both critics and defenders make the assumption that the only appropriate justification for such goods rests on the benefits accrued to a wide variety of individuals, or the ways in which an increasing number of needs may be met. If universities or museums are to survive, a better way has to be found to articulate publicly the distinctive values to which they answer and with which they are engaged. And one can succeed in that political aim only with a better theoretical purchase on the role of social goods in the well-ordered society.

My hope, in this article, is to have begun addressing these problems. In closing, let me summarise how the understanding of social goods as plural needs developed in the last section helps.

Our initial question was whether a just society is one in which urgent needs have absolute priority. It is surprising that there is little discussion of the problem in these stark terms: the absence of any resources left for valuable activities such as art, philosophy, or pure mathematics. Moreover, this is not solely a practical dilemma for how to distribute public funds raised from taxation, but is also a question facing private donors: should they contribute to cancer research, or offer a large amount of money to acquire a rare painting for a worthy museum?20

Our problem, then, was how to make sense of the urgency that some needs have while leaving room for social goods. My strategy has been to focus on understanding social goods. In sum, I have offered two elements as part of our solution. The first is the notion of need as the basis of reasonable rejection. This rejects the model of balancing value against value, and so avoids the counter-intuitive claim that opera or football are more valuable than a human life. The second is the idea of social goods as collective needs. If I think only of myself, then it is difficult to claim that any aesthetic good is a need for me. But this is the wrong perspective from which to evaluate the claims of distinctively social goods on us: we should ask whether they are a need for us taken collectively.

Now, we should not overplay the result that these two additional elements have given us. There is no question that the teenager's interest in a prolonged life reflects a need the teenager has, and hence a claim on the rest of us. I have provided no calculus by which needs can be traded against needs. So we have here a stand-off: given the relative scarcity within the world we cannot meet both needs. The outcome, then, is one in which a tie is broken at best by inertia: the social goods we have are actual, as is our collective need for them; the needs of the teenager are pressing, but not such as to require us to sacrifice the goods we collectively need. So we haven't exactly explained why it should be the right thing to build universities and museums rather than more hospitals and accountancy firms. What I have provided is an explanation of why the basic needs of others do not always impose a requirement on us to dismantle these goods.21 In sum: in raising here the question of social goods, I hope to have provided a crucial, if not invariably conclusive, argument for devoting efforts and resources to preserving them.


  1. 1

    The judgment that one value is more important, or more beneficial, than another is not the only ground for funding the realization of some values rather than others, but the funding of some activities over others is often presented as originating implicitly in such a verdict. For a liberal pro-neutrality justification of state funding of the arts, see Ronald Dworkin's short discussion ‘Can a liberal state support art?A Matter of Principle (Cambridge, MA: Harvard University Press, 1985), pp. 221233. Harry Brighouse challenges Dworkin but accepts the need for a ‘neutrality constraint’ in his ‘Neutrality, publicity, and state funding of the arts’, Philosophy & Public Affairs, 24 (1995), 3563. Contrast Joseph Raz, The Morality of Freedom (Oxford: Oxford University Press, 1986), esp. chs 5 and 6; Steven Wall, Liberalism, Perfectionism and Restraint (Cambridge: Cambridge University Press, 1998). See A Theory of Justice, rev. ed. (Cambridge, MA: Harvard University Press, 1999), sec. 50, esp. pp. 288291 for Rawls's discussion of funding of arts and sciences within the context of his rejection of perfectionism; see Justice as Fairness: A Restatement (Cambridge MA: Harvard University Press, 2001), esp. pp. 151152 for his mature view.

  2. 2

    In this article, I will use ‘urgent needs’, ‘basic needs’, ‘needs stringently conceived’, and ‘vital needs’ interchangeably, to refer to those needs which are invariant between people, no matter the lives or cultures they are in. Of course by ‘urgent’ or ‘basic’ I do not mean to suggest that these needs take absolute priority since this is precisely what is at issue.

  3. 3

    See my ‘The distribution of numbers and the comprehensiveness of reasons’, Proceedings of the Aristotelian Society 105 (2005), 207233. The example was introduced to highlight differences among various cases of aggregation of claims; I also suggested the possibility of a limit to claims of need, an idea which I intend to explain and explore further here.

  4. 4

    There is, though, a further delicate issue to raise about this example: to suppose that there is a special claim here is to suppose that we consider the extent of lives as a whole (and someone who dies as a teenager thereby has a smaller such life). An alternative view is that we should just consider the perspective of each agent now: and so the concern is just with how many more years we can expect to have. This perspective does not privilege the teenager over the nonagenarian. For the sake of the example, I will just assume, and not seek to establish, that we think that on the whole the loss of life in one's mid-teens is a misfortune considerably greater than death late in life.

  5. 5

    Note that this is not to say that it is an absolute demand that funds be directed to universities, or museums. Rather most people doubt that the teenager has a claim that all these funds be redirected for research to cure his illness.

  6. 6

    T. M. Scanlon, What We Owe To Each Other (Cambridge Mass.: Harvard University Press, 1998), p. 235.

  7. 7

    As Frances Kamm points out in ‘Owing, justifying, and rejecting’, Mind 442 (2002), 323354 at p. 340.

  8. 8

    The idea of a Complaint Model is in Scanlon, What We Owe To Each Other, p. 229. Scanlon says that he takes the expression from Parfit. The idea of comparing level of complaints (as well as possible departures from such strict weighing) is in any case already present in Scanlon's 1982 ‘Contractualism and utilitarianism’, reprinted in T. M. Scanlon, The Difficulty of Tolerance (Cambridge: Cambridge University Press).

  9. 9

    For qualifications introduced to a simple Complaint Model see Scanlon, What We Owe To Each Other, pp. 196, 229. In particular Scanlon stresses that effects on a person are not solely effects on her well-being, but also on how costs are imposed. Unfairness and violations of rights and entitlements, for example, constitute grounds for rejection even of principles which would not have adverse effects on a person's well-being. I shall suggest in Section III further ways of building more structure into the idea of a comparative complaint.

  10. 10

    Thomas Nagel, ‘Equality’, in Mortal Questions (Cambridge: Cambridge University Press, 1979), pp. 106127 at p. 123.

  11. 11

    Derek Parfit, On What Matters (Oxford: Oxford University Press, 2011), vol. II, p. 193. One might consider Parfit's picture here too consequentialist since it is conceived as a restriction on an aggregative end while many who find the case problematic take it to show that we don't think in aggregative terms at all.

  12. 12

    Derek Parfit, On What Matters, vol. II, ch. 21.

  13. 13

    The question of whether the Complaint Model imposes too severe a restriction on individuals' lives has been raised before. In Equality and Partiality, (Oxford: Oxford University Press, 1991), Nagel claims that there may not be a general principle of conduct that no individual will have a legitimate complaint against. Either a principle will be insufficiently attentive to the claims of those in need, or it will be too demanding from the perspective of those called upon to make a sacrifice, given that they have their own lives to live. More recently Elizabeth Ashford, ‘The demandingness of Scanlon's contractualism’, Ethics 113 (Jan. 2003), 273302, has argued that contractualist theories face a problem of moral demandingness. Using roughly the Complaint Model, she argues that contractualists no less than utilitarians find themselves committed to extremely demanding obligations to help those in need. She further comments that any impartial moral theory will be similarly demanding; hence, she concludes, any plausible moral theory will be much more demanding than we are initially and pre-theoretically inclined to find acceptable.There is a plausible element in what both of these otherwise different approaches agree on: reasonable demands to help those in need will in many contexts be more demanding than some of those called upon to help would initially find acceptable. Nonetheless before we can assess the importance and extent of this resistance, and its implications for ethical principles justifiable to individuals, a more detailed understanding of the Complaint Model is needed.

  14. 14

    For a similar point see Raz, The Morality of Freedom (Oxford: Oxford University Press, 1986), ch. 8. Cf. also Parfit, Equality or Priority (Lindley Lecture: University of Kansas, 1991), which offers the comparison with the difficulty someone at a high altitude has to breathe. There is an absolute standard (of oxygen necessary to breathe well) by which we judge that this person is badly off. I delay until Section V below discussion of what might follow in terms of social policy.

  15. 15

    See Angie Wagner, ‘Climber amputates own arm to save his life’, Albuquerque Journal, 3 May 2003.

  16. 16

    I am assuming throughout that our focus is on situations where there are conflicting claims on scarce resources, but where there is no excessive scarcity (as in cases of societal emergency or breakdown in extreme famine or total war). That is, the core assumption is that what Rawls terms ‘the circumstances of justice’ obtain: those circumstances under which it is both necessary and possible to settle matters of distributive justice.

  17. 17

    Raz, The Morality of Freedom, p. 153.

  18. 18

    The idea that we should take the contrast between plural and singular predication seriously is central in recent discussion in the philosophies of logic, mathematics, and language; see George BoolosTo be is to be the value of a variable (or some values of some variables)’, Journal of Philosophy, 81 (1984), 430450; Keith Hossack, ‘Plurals and complexes’, British Journal for the Philosophy of Science, 51 (2000), 411443; Alex Oliver and Timothy Smiley, ‘Plural descriptions and many-values functions’, Mind, 114 (2005), 1039–18; Fraser MacBride, ‘The particular–universal distinction: a dogma of metaphysics?’, Mind, 114 (2005), 565614. For opposition see, James Higginbotham, ‘On higher-order logic and natural language’, Philosophical Logic, ed. Timothy J. Smiley, Proceedings of the British Academy, vol. 95 (Oxford: Oxford University Press, 1998).

  19. 19

    Of course this leaves open what commitments we have to the existence of group agents and the ontology of the social world.

  20. 20

    The example is from Joel Feinberg, ‘Not with my tax money’, Problems at the Root of Law: Essays in Political and Legal Theory (Oxford: Oxford University Press, 2003), p. 123.

  21. 21

    How does this account fit with a liberal conception of justice? For Rawls public funding is justifiable for some research and cultural activities because they are instrumental to the meeting of needs, for example, health care, but their study and practice for their own sake falls outside the scope of justice. Where citizens' claims to social resources are grounded in their common needs, the study and practice of pure sciences, non-applied philosophy, or the arts belong to partial conceptions of the good. Moreover, for Rawls, justice takes priority: ‘Fundamental justice must be achieved first. After that a democratic electorate may devote large resources to grand projects in art and science if it so chooses.’ ( Justice as Fairness: A Restatement, p. 152 ). The liberal might speculate that a just society could only decide to build universities devoted to research in the liberal arts and pure sciences once justice has been achieved. Nevertheless, given that such institutions have in fact come into existence, even if through decisions that could not be justified on Rawlsian grounds, we still face a question about the attitude we should take towards the needs that these institutions serve.