The cosmopolitan strikes back: a critical discussion of Miller on nationality and global equality

The cosmopolitan strikes back: a critical discussion of Miller on nationality and global equality

Nils Holtug*

Centre for the Study of Equality and Multiculturalism, University of Copenhagen, Denmark


According to David Miller, we have stronger obligations towards our co-nationals than we have towards non-nationals. While a principle of equality governs our obligations of justice within the nation-state, our obligations towards non-nationals are governed by a weaker principle of sufficiency. In this paper, I critically assess Miller's objection to a traditional argument for global egalitarianism, according to which nationalist and other deviations from equality rely on factors that are arbitrary from a moral point of view. Then I critically discuss Miller's claim that there is no culturally neutral currency with respect to which we may reasonably claim that people should be equally well off on a global scale. Furthermore, I critically discuss Miller's claim that cosmopolitanism undermines national responsibility. And finally, I turn to Miller's own sufficientarian account of global justice and argue that it exhibits too little concern for the plight of the globally worse off.

Keywords: equality; cosmopolitanism; David Miller; nationalism

Published: 16 September 2011

Citation: Ethics & Global Politics, Vol. 4, No. 3, 2011, pp. 147–163. DOI: 10.3402/egp.v4i2.5873

© 2011 N. Holtug. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Noncommercial 3.0 Unported License (, permitting all non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

*Correspondence to: Nils Holtug, Centre for the Study of Equality and Multiculturalism, University of Copenhagen, Denmark, Email:

In his recent book, National Responsibility and Global Justice, David Miller develops an account of our obligations towards the world's poor. Crucially, this account involves a distinction between domestic (or social) justice and international (or global) justice, where the former is much more demanding in terms of the claims for advantages other people can legitimately make on us. More precisely, we have much stronger obligations towards our co-nationals than we have towards non-nationals. While a principle of equality governs our obligations of justice within the nation-state, our obligations towards non-nationals are governed by a weaker principle of sufficiency, according to which we are required only to bring about a decent minimum, where people's basic needs are catered for. Against Miller, I here want to defend cosmopolitanism; and more specifically, global egalitarianism.

Like Miller, I shall ask what justice requires with respect to the global distribution of advantages in a world of nation-states.1 Thus, I shall not raise the question of whether nation-states should be replaced, wholly or in part, with supra-national states or organisations. That is, my concern is with ‘moral’ rather than ‘political’ cosmopolitanism, to use Miller's terms for this distinction. Roughly, moral cosmopolitanism (henceforth: cosmopolitanism) is the view that what we owe to others does not intrinsically (i.e. non-instrumentally) depend on such factors as nationality, citizenship, race, ethnicity, cultural and religious affiliation and the like.2 While cosmopolitanism picks out a rather heterogeneous collection of principles, I shall primarily focus on a global egalitarian version thereof.

Miller's case for less stringent requirements globally than domestically relies on two arguments; a negative and a positive one. According to the positive argument, membership of a nation means being collectively responsible for the things this nation does, including things it has done in the past. Amongst other things, this implies that as a matter of justice we should share in both the advantages and the burdens that our national membership brings with it.3 And this, again, means that as members of a nation we are entitled to advantages to which less well off outsiders cannot make legitimate claims, just as less well off outsiders may be collectively responsible for some of their burdens in such a way that we are under no obligation to compensate them. According to the negative argument, independently of Miller's positive argument for national responsibility, there are various reasons why global egalitarianism is a flawed ideal. Here, I shall only be concerned with Miller's negative argument. This is not to diminish the importance of Miller's positive argument concerning collective national responsibility, but merely to narrow my focus.

It is worth stressing, though, that while Miller's conception of collective responsibility rules out cosmopolitanism, the latter view does not rule out that we should sometimes hold people responsible for their shares of goods, and this is one reason why cosmopolitan (or global) egalitarians need not claim that all inequalities are morally regrettable. Thus, cosmopolitan proponents of various forms of equality of opportunity will claim that inequalities are only unjust insofar as people are not responsible for these themselves. And indeed, cosmopolitanism is compatible with (at least some) responsibility-catering principles of justice.4 This is because holding individuals responsible for their choices does not distinguish between them on the basis of nationality, citizenship, race, ethnicity, cultural and religious affiliation and the like (henceforth, for brevity, nationality and the like).

There is also another reason why global egalitarians need not consider all inequalities morally regrettable, all things considered, which will turn out to be important in what follows. This is because global egalitarians may hold other values, besides equality. Thus, egalitarians should be concerned with at least one further value, namely efficiency. If not, they may be required to level down, that is, to move from an unequal distribution to a more equal one in which everyone is worse off than in the former distribution. In fact, unless we introduce a concern for efficiency, we cannot even claim that (perfect) equality is better at higher levels than at lower levels. Therefore, minimally, egalitarians should be paretian egalitarians, and so claim that we are never morally required to level down. But, in fact, many egalitarians adhere to requirements of efficiency that are stronger than that.

As far as possible, I will remain neutral between different versions of global egalitarianism. For example, I will not simply assume any particular currency of egalitarian justice, but will try to answer Miller's objections as generally as possible (that is, in a manner that is compatible with as many different currencies as possible). On some occasions, however, it will make a difference which currency we assume and here I shall distinguish between, for example, equality of resources, equality of welfare and equality of opportunity for, respectively, resources and welfare. My most general term for the currency of egalitarian justice, however, is ‘advantages’ and this term is meant to be neutral between the different currencies.

Furthermore, my discussion of gloegalitarianism is intended to cover not just egalitarianism proper, but also (global) prioritarianism. While global egalitarianism (proper) is the view that we should equalise the global distribution of advantages, global prioritarianism is the view that when distributing advantages, we should give priority to the globally worse off.5 One difference between these two theories is that prioritarianism has a ‘built in’ concern for efficiency and so does not need to be supplemented with further values to satisfy the pareto principle. Nevertheless, since these two positions will have roughly similar implications with respect to global distribution and since most of Miller's objections are no less objections to (global) prioritarianism than to (global) egalitarianism, it makes sense to include both. However, since the answers I give to Miller's objections work equally well for both kinds of theory, I can restrict myself to global egalitarianism in what follows.

In the following, I first critically assess Miller's objection to a traditional argument for global egalitarianism, according to which nationalist and other deviations from equality rely on factors that are arbitrary from a moral point of view. This, in part, involves considering Miller's intuitive case for the intrinsic significance of nationality. Then I consider Miller's claim that there is no culturally neutral currency with respect to which we may reasonably claim that people should be equally well off on a global scale. Furthermore, I consider Miller's claim that cosmopolitanism undermines national responsibility. As pointed out above, I consider each of these points in isolation from Miller's positive account of national responsibility, and in this sense the paper has a relatively narrow focus. In particular, regarding Miller's claim that national responsibility is undermined by cosmopolitanism, I merely assess critically his claim that cosmopolitan egalitarianism makes it impossible to hold nations responsible over multiple generations and indeed implausibly implies that responsible nations should compensate irresponsible nations. Finally, I turn to Miller's own sufficientarian account of global justice and argue that it exhibits too little concern for the plight of the globally worse off.


As Miller points out, cosmopolitans sometimes stress that nationality and the like are factors that are arbitrary from a moral point of view and so factors that cannot justify differences in our obligations to others. Consider Rawls’ argument, according to which, in the domestic case, many inequalities are due to natural and social lotteries and so are unjust because non-deserved.6 For example, inequalities between the sexes in a sexist society may be considered unjust precisely because no one chooses his or her own sex. Luck-egalitarians have generalised this argument and suggested that, quite generally, inequalities that are due to luck or circumstance rather than choice should be eliminated, everything else being equal. And cosmopolitan luck-egalitarians have pointed out that, if inequalities that are due to luck should be eliminated, then egalitarianism has global scope. After all, being born into a family of malnourished peasants in, for example, Ethiopia is no less a matter of (bad) luck than, in the domestic case, being born into a poorly educated, low-income family.

Nevertheless, Miller finds this argument for global equality unpersuasive. He quotes Simon Caney's claim that ‘people should not be penalised because of the vagaries of happenstance, and their fortunes should not be set by factors like nationality and citizenship’.7 He then points out that it does not follow from the claim that people are differentiated only by features for which they are not responsible, that it is wrong to treat them differently. After all, we believe that if one individual has a disability and another does not, where they are not responsible for this difference themselves, that may very well be a reason to treat them differently (e.g. by extending disability benefits only to the former individual).

However, the luck-egalitarian argument for global equality outlined above is not that if people are differentiated only by features for which they are not responsible, then it is wrong to treat them differently. Rather, the claim is that if people are differentiated only by features for which they are not responsible and these features cause them to be unequally well off, then it is wrong to treat them differently in the sense of allowing them to be unequally well off, everything else being equal. In fact, Caney may himself be interpreted along these lines when he states: ‘people should not be penalised because of the vagaries of happenstance’ (my emphasis). These claims do not suggest that, for example, people with disabilities may not be treated differently from people without disabilities; rather, in a sense, they suggest that they should be treated differently, in order to equalise advantages.8

It may be objected that in my framing of the luck-egalitarian argument for global equality, equality is simply assumed as a default position from which deviations are unjust if they are the result of luck. And as Susan Hurley and others have argued, there was never an independent argument for equality in the thought that some distributive factors are matters of brute luck. Thus, if we were to claim that luck should not affect distributive shares (what Hurley labels the ‘luck-neutralising aim’), we would be in the unfortunate position that we would have to consider it unjust if an equal distribution (miraculously) came about by luck.9

However, the argument for equality is not what is at stake here. I believe that the most plausible version of the luck-egalitarian argument will generalise the point that when we consider, for example, inequalities between the sexes of the kind described above unjust, this is because no one is responsible for his or her sex. But that is beside the point. After all, Miller accepts egalitarianism as a distributive ideal for the nation-state himself. So the relevant question is whether egalitarianism should have a wider scope or not. And here the argument of the cosmopolitan luck-egalitarian is that if what makes inequalities unjust within the nation-state is luck, then since global inequalities are also a matter of luck, global inequalities are also unjust. In fact, it is precisely Caney's point that the standard justifications for egalitarianism imply a cosmopolitan version thereof.10

Nevertheless, there may of course be reasons that defeat the transference of the injustice of inequality to the international case. And indeed, Miller goes on to suggest that the intrinsic significance of nationality is just such a reason.11 Miller reminds us that many of us, including cosmopolitans, believe that the parent–child relationship has intrinsic significance such that we have stronger obligations towards our own children than towards other people's children. Thus, I have a stronger moral reason to try to find a missing child if it is my own child than if it is someone else's, everything else being equal. Likewise, he suggests that I have a stronger moral reason to try to find this child if he is from my village than if he comes from some other community, again everything else being equal.12 And finally, everything else being equal, I have a stronger moral reason to try to find him if he is a co-national than if he is a non-national.

The moral structure of Miller's claims here is that although it is equally bad if any child goes missing,13 everything else being equal, our obligations to contribute to the search vary with factors such as parenthood, living in the same village and nationality. Thus, an agent-neutral axiology is combined with an agent-relative account of obligations.

Now, it seems to me plausible to claim that we have a pro tanto reason to promote the good, that is, to bring about the best possible state of affairs.14 Assuming the agent-neutral axiology proposed by Miller, this means that we have a reason, of a fixed strength, to help a child in distress irrespective of these three factors; parenthood, living in the same village and nationality. If so, this means that insofar as these three factors affect our obligations, they must introduce further reasons such as to modify the overall pattern of moral concerns. But this does not change the fact that there is a reason, of a fixed strength, that derives from the distressing situation of the missing child.

However, note that Miller cannot accept this picture with respect to global inequality. We might say that as a matter of axiology, it is equally bad if people are unequally well off, irrespective of their nationalities. If so, it seems that there would be a pro tanto reason to eliminate global inequality, but of course a reason that could be outweighed by other moral reasons, including ones that derive from special obligations towards one's co-nationals. But in that case equality would have global scope, contrary to what Miller suggests. Nevertheless, it seems to me that he does not explain why this is an incorrect picture when it comes to equality. However, I shall not elaborate further on this point here.

The crucial question is of course whether we should accept Miller's claims about special obligations. He argues that there are three conditions an attachment must satisfy in order to ground special duties: (1) the relationship must be intrinsically valuable, (2) the duties must be integral to the relationship such that it could not exist in the same form unless the duties were generally acknowledged and (3) the relationship must not be inherently unjust.15

However, even if a certain relationship satisfies (1–3), this does not automatically imply that we have special duties, at least if by special duties we mean agent-relative duties of the kind favoured by Miller. Let us, for the sake of argument, assume that national membership satisfies (1–3). It will then be true that we can only preserve national membership in its current form if we generally acknowledge such duties. However, granting this does not settle the question of whether we should acknowledge them. First, it may be the case that while something of value is lost if we do not acknowledge special duties, namely because national membership becomes a less valuable relation, justice nevertheless requires us to be impartial and so to make this sacrifice for the greater overall good. Of course, cosmopolitans cannot make this claim, because it involves acknowledging the intrinsic significance of nationality, but nor can Miller's nationalist. After all, impartiality at this level rules out special duties since these favour co-nationals.

Second, cosmopolitans can claim that while, in our criterion of rightness, nationality and the like have no intrinsic significance, this criterion may well justify a decision-procedure that allows for agent-relative duties towards co-nationals, namely insofar as acknowledging such duties in our decision-procedure best promotes our cosmopolitan impartial conception of justice.16 The idea is that acknowledging these duties may do so because it involves an effective division of labour between states.

It may nevertheless be argued that, just as special duties in the parent–child relationship have a certain intuitive plausibility even at the level of our criterion of rightness, so do special duties towards co-nationals. However, I must admit that I do not share Miller's intuitions here, either with respect to the intrinsic significance of nationality or even of belonging to the same village. The case of belonging to the same village seems intermediate between parenthood and co-nationality in the sense that we are inclined to imagine that there may be some personal relations in the village case and that they are weaker than those that obtain between parent and child but stronger than those that typically obtain between co-nationals. And I am inclined to think that to the extent that we do find special obligations towards those with whom we share a village intuitively appealing, this may at least in part rely on a further assumption we make, namely that the relation between members of the same village is characterised by certain personal relations. But note that there need not be any relations of this kind. In fact, in my own case, I know less than 0.1% of the people who live in my city.

Obviously, co-nationality does not rely on personal relations either. I have personal relations only with a tiny fraction of my co-nationals. In light of this, suppose that two children—both perfect strangers to me—have gone missing and I am in a position to help only one of them. Perhaps I can only provide transport for one. It seems to me almost obscene to ask them about their nationalities before I decide whom to help. This would be inappropriate simply because their nationalities do not matter with respect to whom I should rescue. Of course, it may be suggested that while it would indeed be obscene to ask about their nationalities, it would not be problematic in the same way to think that their nationalities matter with respect to whom I should help. However, if nationality does indeed matter in this sense, then why would it be obscene to inquire about it? It would seem to be no different than asking about the children's health conditions to settle whom it would be most urgent to help. And unlike asking about nationalities, asking this latter question does not strike me as inappropriate in any way.

This argument can be strengthened by considering what a relation of co-nationality ultimately consists in. Quite generally, it consists in belonging to a group with a common identity of a particular kind, with which one identifies and which is (perhaps) partly constitutive of one's identity. Nevertheless, what kind of common identity are we talking about here? For a thin notion of nationality, such as that of Miller's liberal nationalism, it involves a common public or political culture, but also recognising special obligations towards one's co-nationals, considering the continued existence of the nation a valuable goal, and aspiring to be politically self-determining.17 Nevertheless, note that a missing child and I may be co-nationals in Miller's sense even if, for example, we differ with respect to the particular aspects of the public culture we endorse (insofar as a child, or this child, does endorse aspects of a public culture), and even if I don't recognise special obligations towards my co-nationals (as in fact I don't).18 What is required is that a sufficiently large proportion of the population do share these features. It seems to me that, in the case of the missing children, the fact that one of them is my co-national in this rather weak sense is simply irrelevant.

Mark Johnston has called this type of argument an ‘argument from below’.19 It claims that since co-nationality (in the relevant liberal sense) just consists in certain lower-level facts (a shared public culture, etc.), co-nationality only matters if these lower-level facts matter. Alternatively, one might mount an ‘argument from above’, according to which, while these lower-level facts do not intrinsically matter, the higher-level fact may matter (and so the lower-level facts matter derivatively, because they constitute the higher-level fact). Thus, nationality may intrinsically matter even though the lower-level facts that constitute it do not.

However, I believe that we should accept the argument from below. If the relevant lower-level facts obtain between two people, it is not a further fact about them that they are co-nationals. This fact just consists in these other facts. The fact that they are (also) co-nationals is a conceptual fact, rather than a further fact about the relations between them. And if co-nationality matters intrinsically, it must be in virtue of the lower-level facts this relation consists in, rather than in virtue of the conceptual fact that when these lower-level facts obtain between two individuals, they are co-nationals.20 Therefore, if I am right in claiming that even if I share a public culture, etc. with one of the missing children, this does not intrinsically matter, nor does the fact that we are co-nationals.


In this section, I want to consider a more specific argument of Miller's against global egalitarianism. According to this argument, there is no currency with respect to which we can plausibly claim that individuals should have equal shares on a global level. The general problem is that there is no culturally neutral way to define the relevant currency.

Consider, for example, Dworkin's conception of equality of resources. Here, a hypothetical auction in which individuals have equal purchasing power is used to determine an equal distribution of heterogeneous resources amongst individuals with heterogeneous sets of preferences. Certain background conditions need to be settled beforehand, because people need to know what uses they may put the available resources to, or they cannot possibly asses the value of such resources to themselves.21 Dworkin's solution to this problem is to claim that individuals should have the greatest possible freedom to use resources in any way they see fit (subject, of course, to the usual liberal constraints). However, Miller argues that such a liberal background is inappropriate in a global context because not all cultures embrace liberal ideals.22

Furthermore, equality of opportunity (of the Rawlsian kind) fares no better. If equality of opportunity requires individuals to have identical opportunity sets, then in order for a child from rural Mozambique to have an equal opportunity to become president of the USA (which requires American citizenship), we must imagine a ‘borderless world in which everyone speaks Esperanto’.23 If, on the other hand, equality of opportunity requires individuals to have equivalent (i.e. equally valuable) opportunity sets, then we again face the problem that there is no culturally neutral way of fixing the relative value of different such sets. For example, assuming that people in Iceland have better educational opportunities than people in Portugal, but people in Portugal have better leisure opportunities than people in Iceland (sunny beaches, etc.), how do we compare their opportunity sets all told?

I am going to assume that the problem of cultural neutrality Miller is getting at here is not a problem of moral relativism, but rather a problem of impartiality. More precisely, the problem is that of finding a currency that is sufficiently sensitive to differences between cultures to provide an impartial standard for assessing how well off members of these cultures are relative to each other.

However, first, it seems to me that Miller does not fully appreciate the extent to which these objections, if successful, would tend to undermine not only an ideal of global equality but also equality in the domestic case. Consider, for example, his case of Iceland and Portugal. While people in Iceland may well disagree with people in Portugal on the relative importance of (the relevant) educational and leisure opportunities, surely people in Iceland will disagree amongst themselves on this same issue (and I see no reason to believe that they will disagree less domestically than with the Portuguese).

Furthermore, nation-states are increasingly becoming culturally heterogeneous as a result of migration and so are increasingly facing the problem that cultural neutrality may be difficult to come by even in the domestic case.24 Of course, Miller may claim that equality is an appropriate ideal only for culturally homogeneous states, but this would seem to imply, implausibly I think, that equality is becoming less and less important in a world of large-scale migration.

Second, I believe there are things that may be said on behalf of equality of resources and equality of opportunity in response to Miller's objection. Dworkin's liberal background assumption, according to which bidders at the auction have the freedom to use resources as they see fit, is designed to allow them to express their preferences in the auction as fully as possible. Furthermore, the idea that equality is defined in terms of people's total sets of preferences for bundles of resources is the very notion of impartiality equality of resources encapsulates. Of course, one may disagree with this understanding of impartiality, but the fact of disagreement does not in itself make it unfair. So equality of resources egalitarians, I take it, will argue that insofar as some people reject the background assumption, they are rejecting what is in fact the most plausible conception of impartiality. And simply to claim at this point that Dworkin's conception of impartiality is not shared in all cultures is to move from a concern for impartiality to cultural relativism, and I am assuming that this is not Miller's intention.

Consider also equality of opportunity. Miller objects to the idea that individuals should have identical opportunity sets because this would require a ‘borderless world in which everyone speaks Esperanto’.25 Now, many egalitarians would in fact consider a world of free migration preferable to the actual regime of restrictive immigration policies and therefore may not be so worried about at least some of these requirements.26 But it is also worth stressing, as I have done above, that egalitarians need (and should) not have equality as their only moral concern and so there may be other concerns that imply that, for example, a global language policy requiring people to speak Esperanto cannot be justified, all things considered.

Nevertheless, I agree with Miller that, ultimately, equality of opportunity requires equivalent sets of options, not identical ones. So how do we assess the relative value of options in a way that is appropriately neutral between the different commitments of members of different cultures? There could of course be different answers to this question, but I believe that the best answer is that people's sets of options should be equally valuable in terms of welfare. In other words, the most plausible version of equality of opportunity is equality of opportunity for welfare.27

Miller does not consider this possibility although it does, in an important sense, capture the claim that the currency of egalitarian justice should be impartial between the commitments of members of different cultures.28 Thus, if Danes and Malaysians value a particular option differently, this option may very well give rise to different welfare levels for them and so they are not necessarily equally well off in the relevant sense, even if they have identical sets of options. Nor are they necessarily equally well off even if they have equal shares of resources. More generally, equality of opportunity for welfare is more sensitive to cultural differences than equality of resources because cultural preferences are more directly related to welfare than to resources.29 For example, while a ban on ritually slaughtered meat does not give rise to unequal resource levels for members of different cultures, it may very well give rise to different welfare levels, given the differences between their cultural (or religious) preferences.

Of course, equality of opportunity for welfare needs to be worked out in much greater detail than it is possible here. But the point is simply that this principle offers a currency of justice that is, in an important sense, sensitive to cultural differences between individuals and more so than the egalitarian principles considered by Miller.


Another objection Miller raises to global egalitarianism is that this principle undermines national responsibility. Consider two societies, Affluenza and Ecologia. From an equal starting point, Affluenzians decide to use up most of their resources for current consumption whereas Ecologians conserve their resources by adopting a policy of sustainable development. After a while, perhaps in the next generation, people in Ecologia will be better off than people in Affluenza because they have adopted a sustainable environmental policy. So does justice require that, in the next generation, resources are to be redistributed between Affluenza and Ecologia?

I have suggested that global egalitarians may well claim that people should sometimes be held responsible for their choices, but the problem here is that not all the people in Affluenza are necessarily responsible for the choice of depletion. This is most obvious if we consider the next generation, which was not around when that policy was adopted. However, if we do not hold Affluenza and Ecologia responsible for their choices, this has two very unfortunate consequences, according to Miller. First, there will be very little incentive for states and citizens to behave in responsible ways. And second, it is simply unfair to tax Ecologia to restore Affluenza to a position of equality.30

Miller then suggests that the solution to these problems is to adopt a principle of collective responsibility, according to which Affluenza and Ecologia can be held responsible even in the next generations. And it seems that such a principle must rely on a claim about the importance of nationality, citizenship or the like to justify that the descendents of the decision-makers are also to be held responsible. Therefore, such a principle is incompatible with cosmopolitanism egalitarianism.31

To address Miller's first objection to global redistribution first, I believe that it attacks global egalitarianism at the wrong level, so to speak. We need to separate questions of what global justice ultimately consists in (a criterion of rightness) from questions of which principles to follow in order best to achieve global justice (a decision-procedure). The question of incentives pertains to the latter question.

Suppose, as Miller does, that redistribution to compensate for irresponsible policies creates incentives to adopt these very policies. Then such redistribution may be a bad idea, even in terms of global equality. It will tend to render the next generation worse off than the present generation, because the present generation will deplete resources. Of course, if other states do not deplete their resources, then perhaps they may compensate the next generation of Affluenzians such that they are no longer worse off, or at least not much worse off, than their parent's generation. But presumably, assuming the incentive structure described by Miller, other states will also be tempted to deplete. Likewise, the next generation may be tempted to further deplete resources, at the (further) expense of the generation after that, etc.

The consequence is that of grave inequality between generations. Therefore, to the extent redistribution to compensate for irresponsible policies creates incentives to adopt these very policies, global egalitarians have a reason to resist such redistribution. That is, they have reason not to implement a policy of equalising Affluenzians and Ecologians. A reason that simply follows from their commitment to global equality. Therefore, global egalitarianism does not imply a perverse incentive structure.

However, note that the consequence that their children and grandchildren, as well as future generations more generally, will be much worse off than themselves, even if there is redistribution in the next generations, may itself be a factor that disinclines the present generation towards irresponsible policies. Thus, the incentive structure imposed by a scheme of redistribution may not coincide with that described by Miller.

Furthermore, as I pointed out in the introduction, egalitarians are concerned not only with equality but also with efficiency. And arguably, in terms of efficiency, it will not be a good idea to create incentives for adopting irresponsible policies of depletion. In order to make this claim, egalitarians will need a stronger conception of efficiency than merely pareto optimality. This is because if the present generation does not deplete resources, this will reduce their level of welfare. Nevertheless, since the resources will then be more evenly spread over generations, the principle of diminishing marginal utility implies that these resources generate a higher total of welfare, everything else being equal.

Therefore, insofar as redistribution to compensate for irresponsible policies is likely to lead to further adoption of such policies, global egalitarians have a second reason to resist this kind of redistribution. Thus, global egalitarians may allow global inequalities to create an efficient incentive structure just as in the domestic case, egalitarians are usually inclined to allow inequalities insofar as this is necessary to promote incentives for productivity.32

What about Miller's second objection, then? Is it really fair to tax Ecologia to restore Affluenza to a position of equality (insofar as this is required by global egalitarianism)? First, note that by adopting a policy of depletion, the citizens of Affluenza have adopted a grossly unjust policy at the expense of the next generations who will pay the price for the self-indulgence of their forefathers.33 The injustice, of course, consists in the fact that the next generations will be worse off than the present generation. Therefore, according to global egalitarians, it may be appropriate for Ecologia and other states or organisations to sanction Affluenza to prevent them from committing such a grave injustice.

Second, assuming that this injustice has already taken place, or cannot be prevented with politically acceptable means, is it really clear that it is unfair to tax citizens of Ecologia to restore global equality? Why should the new citizens of Affluenza have to be worse off than everyone else because of the unjust decisions of their forefathers, for which they are not responsible?34 To claim that they should seems like claiming that, in the domestic case, it is unfair to tax people to secure equality for children who face deprivation because their parents have made irresponsible choices in the past; a claim that egalitarians are usually quite sceptical of.

Furthermore, if the burdens imposed by their forefathers were distributed on only half the citizens of Affluenza, we—and, I take it, Miller—would not consider it unfair to tax the other half to restore internal equality, although this other half is no more responsible for the policy of depletion than the citizens of Ecologia are. Therefore, it seems to me that redistribution would not be unjust but would in fact be required (again, assuming that considerations of the impact of incentives on global equality and efficiency do not speak against such a measure).

Finally, note that even Miller must be prepared to limit national self-determination to impose a fair scheme of global distribution, since he believes that, in the global case, there is a decent minimum to which everyone is entitled. So the question is not whether or not it may be legitimate to limit national self-determination to secure a fair global distribution, but rather how far we may go in this regard. This must be determined on the basis of a theory of global justice. And I have argued that Miller's arguments against egalitarianism as a theory of global justice are not persuasive.


In this section, I want to consider the global principle Miller wishes to replace global egalitarianism with. According to this principle, there is a minimum level of advantages to which everyone in the world is entitled, and insofar as people are below this minimum it triggers what he calls ‘remedial responsibilities’ on others.35 What this means is that others are obligated to assist, everything else being equal, irrespective of whether they have themselves contributed to the harm suffered by those in need. If others have themselves contributed to the harm, on the other hand, they may have more extensive duties than this.

The minimum to which everyone is entitled is set on the basis of people's basic intrinsic needs, where the list of basic intrinsic needs includes (but is not exhausted by) food and water, clothing and shelter, physical security, health care, education, work and leisure, and freedoms of movement, conscience, and expression.36

Miller's global principle is sufficientarian since it sets up a threshold of advantages and requires us to ensure that everyone at least meets this level. However, once the threshold is met, it requires no more of us, regardless of how unequal the global distribution of advantages is. And in this respect, of course, it differs from global egalitarianism.

Since Miller proposes (at least) two principles, one for domestic distribution and one for global distribution, there is an issue of how to weigh them against each other in cases of conflict. He has several things to say about this but I want to briefly focus on one particular aspect. This is his claim that ‘all obligations of social justice towards fellow nationals should take precedence over international obligations that arise from failures of responsibility by third parties—this despite the fact that the condition we are responding to may be much worse in the case of outsiders’.37 Suppose that people in Sudan starve because the British have not compensated them for past injustices of colonialism, or simply because the British have been assigned special responsibility for Sudan in a just international assignment of obligations but have not done their part. If I understand Miller correctly, the Scandinavian countries should rather further increase their already high levels of domestic equality—assuming, plausibly, that social justice requires this, everything else being equal—than prevent the Sudanese from starving to death. I find this hard to swallow.38

But even if we restrict ourselves to Miller's sufficientarianism, and so ignore the potential conflict with domestic justice, his account seems to me unsatisfactory. First, we may ask what his sufficientarian principle implies, in practical terms. At one point he writes:

…. no one can doubt that undernourishment, low life expectancy, lack of access to elementary education or health care, and the other components of extreme poverty add up to a life that is less than minimally decent. So our moral response to these facts should also be clear: it is morally intolerable that we live in a world where somewhere between 15 and 20 percent of people live in dire poverty as defined by these indicators.39

Now, it is not entirely clear whether Miller is claiming that between 15 and 20% live below his sufficiency threshold, or merely that it is clear that at least between 15 and 20% are below it. But if he is making the former claim, then it seems to me that he is setting his standard remarkably low. Consider, for example, the fact that 2.5 billion people or more than 40% of humankind live below the World Bank's $2/day poverty line.40

However, suppose Miller sets the threshold much higher. In fact, suppose for the sake of argument that he was to claim that what a decent minimum involves is a standard of living roughly equivalent to the average standard of living in, say, France. Now consider the following case.41 Suppose that in two neighboring nation-states, everyone is at the sufficiency threshold. However, in their part of a forest that stretches over both states, citizens of one state have just found some very special fruit trees. The fruits contain vitamins that cause people to live, say, 400 years and to enjoy exceptionally high levels of welfare. Having heard the good news, citizens of the second state have searched their part of the wood but, sadly, only to find that they have not been blessed with any of those marvellous fruits. Citizens of the first state could of course decide to share their fruits with citizens of the second state, but are they required to do so as a matter of justice? According to Miller's sufficientarianism, they are not. Not even if sharing would not in any way deprive them of the benefits bestowed on them by the fruits.

While fanciful, the case of the vitamin-rich fruits is meant to bring out the implications of sufficientarianism in a particularly lucid and stark way. But this should not conceal the fact that Miller's sufficientarianism has real-life implications for the globally poor. Nevertheless, are there ways in which Miller's sufficientarian may avoid the strongly anti-egalitarian conclusion in the fruit case scenario? It is not clear that there are ways to avoid this. The minimum to which everyone is entitled is set on the basis of basic needs and unlike what Miller calls ‘societal needs’, basic needs are ‘conditions for a decent human life in any society’.42 This means that even though the expected life-length and welfare has now been hugely boosted in the lucky society of fruit-tree owners, members of the other, less fortunate nation-state cannot make claims for resources (fruits) on the basis of their much lower expected life-length and welfare. A longer, better life than that of the average French person is not a need in the relevant sense.43

This implication of sufficientarianism strikes me as highly counter-intuitive. We may imagine citizens of the second state asking, at the end of their lives, why they should have to die now when there are fruits enough to go around and citizens of the first state have another 320 excellent years to look forward to? It seems to me that we should set our standards for global justice somewhat higher.


In previous sections, I have considered a number of Miller's objections to global egalitarianism. First, I have critically examined his objection to the luck-egalitarian argument for global equality. I have argued that while indeed luck does not imply equal treatment, this is not what luck-egalitarians claim. Rather, if a person's being worse off than others can be attributed to bad luck, then this inequality is unjust, everything else being equal. Also, I have suggested that Miller's intuitive case for the intrinsic significance of nationality may be less intuitive than it may initially appear. Second, I have challenged Miller's claim that there is no culturally neutral currency on the basis of which we can assess claims for global equality. Amongst other things, I have suggested that equality of opportunity for welfare may be a currency that is sufficiently sensitive to cultural differences. Third, I have considered Miller's suggestion that global egalitarianism undermines national responsibility. In response, I have argued that global egalitarians may have reason not to compensate irresponsible nations if compensation leads to perverse incentive structures, and that there is nothing unfair about such compensation if it does not lead to perverse structures and if the people compensated have not participated in the irresponsible acts. Finally, I have argued that Miller's global sufficientarianism does not sufficiently cater for the interests of the globally poor.


I would like to thank Simon Caney, Gro Hellesdatter Jacobsen, Signe Kjær Jørgensen, Kasper Lippert-Rasmussen, Sune Lægaard, two anonymous reviewers and not least David Miller for helpful comments on an earlier version of this paper.


1. David Miller, National Responsibility and Global Justice (Oxford: Oxford University Press, 2007), 20.

2. Therefore, this form of cosmopolitanism should also be distinguished from what Samuel Scheffler calls ‘cosmopolitanism about culture’, which is a view about the prudential significance of being a member of a determinate culture. See Samuel Scheffler, ‘Conceptions of Cosmopolitanism’, in his Boundaries and Allegiances. Problems of Justice and Responsibility in Liberal Thought (Oxford: Oxford University Press, 2001), 112.

3. David Miller, National Responsibility, 265.

4. The reason for the parenthesis is that some forms of responsibility, including Miller's conception of national responsibility, will include non-eliminable references to nationality and the like.

5. For further clarification of the distinction between egalitarianism and prioritarianism, see Derek Parfit, Equality or Priority? (The Lindley Lecture, University of Kansas, 1991; reprinted in The Ideal of Equality, eds. Matthew Clayton and Andrew Willliams, Basingstoke: MacMillan Press Ltd., 2000). I defend prioritarianism against various objections and argue that it is the superior distributive view in Nils Holtug, ‘Prioritarianism’, in Egalitarianism. New Essays on the Nature and Value of Equality, eds. Nils Holtug and Kasper Lippert-Rasmussen (Oxford: Clarendon Press, 2006); and in Nils Holtug, Persons, Interests and Justice (Oxford: Oxford University Press, 2010), Chap. 8.

6. John Rawls, A Theory of Justice (Oxford: Oxford University Press, 1971), 71–2.

7. David Miller, National Responsibility, 32; Simon Caney, ‘Cosmopolitan Justice and Equalizing Opportunities’, Metaphilosophy 32 (2001): 115.

8. Caney has confirmed, that this is the interpretation he himself intended (pers. comm.). See also Caney, ‘Cosmopolitanism and Justice’, in Contemporary Debates in Political Philosophy, eds. T. Christiano and J. Christman (Oxford: Blackwell, 2009).

9. Susan Hurley, Justice, Luck, and Knowledge (Cambridge, Mass: Harvard University Press, 2003), 146–80.

10. Simon Caney, Justice Beyond Borders. A Global Political Theory (Oxford: Oxford University Press, 2005), 107.

11. As I pointed out in the introduction, Miller has a positive argument for the intrinsic importance of nationality, namely for national responsibility, that I shall not consider here. However, prior to developing this positive argument, he argues that cosmopolitanism “conflicts with an intuitively plausible picture of agent's responsibilities”, and it is this intuitive picture that I engage with here (Miller, National Responsibility, 50).

12. Ibid., 29.

13. Ibid., 29.

14. For an argument to this effect, see Shelly Kagan, The Limits of Morality (Oxford: Clarendon Press, 1989), 47–56.

15. David Miller, National Responsibility, 35–6.

16. See e.g. Robert E. Goodin, ‘What Is So Special about Our Fellow Countrymen?’ Ethics 98. Miller is himself aware of this argument and discusses it in On Nationality (Oxford: Clarendon Press, 1995), Chap. 3.

17. David Miller, National Responsibility, 124–6.

18. Ibid., 125.

19. Mark Johnston, ‘Human Concerns without Superlative Selves’, in Reading Parfit, ed. Jonathan Dancy (Oxford: Blackwell Publishers, 1997), 167. Johnston, however, is not concerned with the significance of nationality but with the significance of personal identity.

20. For a similar argument, albeit in the context of personal identity and in response to Johnston, see Derek Parfit, ‘The Unimportance of Identity’, in Identity, ed. H. Harris (Oxford: Clarendon Press, 1995), 28–41.

21. Ronald Dworkin, Sovereign Virtue (Cambridge, Massachusetts: Harvard University Press, 2000), 65–183.

22. David Miller, National Responsibility, 61–2.

23. Ibid., 63.

24. And indeed, acknowledging this, Miller argues elsewhere that in the domestic case, equality of opportunity requires the costs to individuals of taking advantage of opportunities to be equalised, where differences in costs are due to differences in culture. See David Miller, ‘Liberalism, Equal Opportunities and Cultural Commitments’, in Multiculturalism Reconsidered, ed. Paul Kelly (Oxford: Polity, 2002), 46.

25. Brian Barry sometimes writes as if he believes that equality of opportunity requires identical opportunity sets. See e.g. Brian Barry, Why Social Justice Matters (Cambridge: Polity Press, 2005), 22.

26. See e.g. Joseph Carens, ‘Migration and Morality. A Liberal Egalitarian Perspective’, in Free Movement. Ethical Issues in the Transnational Migration of People and of Money, eds. Brian Barry and Robert E. Goodin (New York: Harvester Wheatsheaf, 1992).

27. As Richard Arneson has coined this view, see Arneson, ‘Equality and Equal Opportunity for Welfare’, Philosophical Studies 56 (1989).

28. For a much more detailed discussion of this point, see Nils Holtug, ‘Equality and Difference-blind Rights’, in Nationalism and Multiculturalism in a World of Immigration, eds. Nils Holtug, Kasper Lippert-Rasmussen and Sune Lægaard (Basingstoke: Palgrave Macmillan, 2009).

29. Ibid.

30. David Miller, National Responsibility, 70–1. For a similar argument, see John Rawls, The Law of Peoples (Cambridge, Mass.: Harvard University Press, 1999), 117.

31. Actually, this is not quite true. Our cosmopolitan criterion of rightness may imply that, in the world as it is, we should adopt a decision-procedure that includes collective responsibility of the relevant kind. However, I shall not consider this possibility any further here.

32. Such incentives, of course, are Rawls’ reason for substituting egalitarianism (proper) with the difference principle. See A Theory of Justice.

33. For an account of inter-generational egalitarian justice, see e.g. Larry Temkin, ‘Intergenerational Inequality’, in Philosophy, Politics, and Society, eds. P. Laslett and J. Fishkin (Sixth Series, Yale University Press, 1992).

34. Miller will of course object to this way of phrasing the question, because of his commitment to the notion of trans-generational national responsibility.

35. David Miller, National Responsibility, 167.

36. Ibid., 184.

37. Ibid., 50.

38. Robert van der Veen has suggested that such conflicts between Miller's domestic and global principles do not arise in rich societies, because there is a ‘surplus of opulence’ that is not related to social justice and from which benefits to poor nations may be financed [Robert van der Veen, ‘Reasonable Partiality for Compatriots and the Global Responsibility Gap’, Critical Review of International Social and Political Philosophy 11 (2008), 422]. However, unlike sufficiency, the combination of equality and efficiency is most plausibly characterised as an open-ended position in the sense that there is in principle no limit as to how many benefits we may bestow on people to promote domestic justice [see also Miller, ‘A Response’, Critical Review of International Social and Political Philosophy 11 (2008), 562]. A different objection to my argument is this. An anonymous referee has suggested that while Miller here speaks of “all obligations of social justice towards fellow-nationals” taking precedence, he may be thinking only of the human (basic) rights of fellow-nationals. However, it would then be strange to refer to all obligations of social justice. However, if indeed Miller is only thinking of human rights, the objection I raise will not apply.

39. David Miller, National Responsibility, 232.

40. UNDP, Human Development Report (New York: UNDP, 2005), 24.

41. I develop this argument against sufficientarianism in greater detail in Nils Holtug, ‘Prioritarianism’, 149–54; and in Persons, Interests, and Justice, Chap. 8. There, I also present a number of other arguments against this principle. In the present paper, I have changed my example slightly to match Miller's specific brand of sufficientarianism.

42. David Miller, National Responsibility, 182.

43. In response to my argument, Andrew Mason has offered a different defence of sufficientarianism [Mason, Levelling the Playing Field. The Idea of Equal Opportunity and its Place in Egalitarian Thought (Oxford: Oxford University Press, 2006), 125–6]. He suggests that this view may be combined with a quasi-egalitarian principle that rules out (sufficiently large) inequalities above the threshold. However, as we have seen, Miller explicitly rejects the notion that egalitarianism has global scope and so this option does not seem available to him. Miller has himself suggested that there may be reasons other than justice why it would be wrong not to share the fruits (pers. comm.). However, intuitively, this case seems to me to be a paradigm case of a distribution that is wrong because unjust.

About The Author

Nils Holtug
Centre for the Study of Equality and Multiculturalism, University of Copenhagen

Director, Centre for the Study of Equality and Multiculturalism
Associate Professor, PhD

Article Metrics

Metrics Loading ...

Metrics powered by PLOS ALM

Related Content