David Owen has produced a very interesting commentary on some of the themes in my book National Responsibility and Global Justice. Rather than trying to respond to everything that he says, I have chosen to focus on just three of the areas that he has identified: the relation between nationality and citizenship, the critique of global egalitarianism, and the issue of states' right to exclude immigrants. This in particular means that I have not tried to pursue the contrast between distributive and political justice with which he opens the discussion. Although I sense that something quite important is at stake here, I am not clear enough in my own mind about how this contrast is to be understood, and where the line between the two positions should be drawn. So this is a debate that will have to be pursued on another occasion.

1. Owen points out, quite correctly, that a number of the claims made in the book are claims about nation-states – for example when it discusses how obligations to fellow-members differ from those owed to outsiders. Now a nation-state is a composite entity: it refers to a body of people living as citizens under the common authority of a set of political institutions, but it also refers to the fact that these people share a national identity with the familiar cultural attributes that such identities possess. This composite character of the thing being studied may cause problems. Quite apart from the possibility that nation and state may not coincide – as in the case of stateless nations, or multinational states – there is the difficulty that even when they do very largely coincide, it may not be clear which of the two relationships – citizenship or nationality – is doing the work when arguments about the importance of the nation-state are being advanced.

Owen suggests, in particular, that many of the claims made in the book about the significance of nationality could instead be recast without loss as claims about citizenship. Moreover if citizenship rather than nationality is what really matters, we could avoid the practical problems involved in trying to include minority groups, immigrants and so forth within the national community by devoting ourselves instead to policies that aim to promote common citizenship.

So does nationality matter? What difference does it make to the functioning of a democratic state that its citizens share a national identity rather than simply an identity as fellow-citizens, sharing loyalty to a constitution (Habermas) or a polity (Mason)? This is a hard question to answer empirically, not least because we have no working examples of democratic states that are tied together only by a sense of common citizenship. In another article, my co-author and I have surveyed a range of research that attempts to measure national identity and determine its consequences both for public attitudes towards, for example, inequality and redistribution, and for the actual performance of states in this domain.Footnote ¹ Predictably, perhaps, no simple pattern emerges, so we cannot at this point claim that the national identity argument, construed as a general thesis about the difference it makes if citizens share such an identity, has been confirmed by hard evidence of this kind. It remains a plausible hypothesis, supported as well by micro-level studies of social identity and some macro-level studies of, for example, the causes and consequences of interpersonal trust.

My particular reasons for pursuing the theme of national, rather than citizen, responsibility in the book under review were rather different. First, even if we think that collective responsibilities can most cleanly be attributed in cases where conationals act together through the shared institutions of the state, we would presumably wish to be able to say something about the responsibilities of stateless nations – for example in the case where they carry out a form of ethnic cleansing precisely in order to constitute a territorial state of their own. Second, leaving that special case aside, if we confine our attention to citizenship, then our attributions of collective responsibility will have to rely wholly on what I call in the book ‘the cooperative practice model’ and forgo the support of the alternative ‘like-minded group’ model. To show that individual people can be included in the scope of collective responsibility, we will have to be able to prove that they participated in a mutually advantageous practice that meets certain conditions of fairness and democratic accountability. There is a fairly large gap between the example that I use to introduce the model – an employee-controlled firm that distributes its rewards fairly among members – and even the best version of a modern democratic state. So any attributions of collective responsibility in the latter case – particularly to those who oppose the state's current policies – will be correspondingly fragile. Now of course you may think ‘so much the worse for collective responsibility’. But if you believe that our understanding of global justice is enhanced by asking the question ‘and who is responsible?’ – both when looking at economic inequalities between societies and when looking at the historical impact of one society upon another – then it may be important to explore the way in which shared national cultures can affect states' behaviour, even in cases where there is dissent among citizens about the policies to be followed. This is the aspect of collective responsibility that the ‘like-minded group’ model attempts to capture.

Third, insofar as we appeal to collective responsibility to show that some international inequalities, emerging over time, are just, the responsibility-generating activities in question are unlikely to be confined to what fellow-nationals have done when acting as citizens. Equally important will be practices and forms of behaviour in economic and social life that reflect the shared culture of the society. So, for example, if we give credit to the Japanese for their swift and effective response to the damage inflicted by the recent earthquake and tsunami, the credit goes to a culture that values cooperation and mutual aid in daily life. When governments take decisions and citizens vote to support them, this will also reflect the national culture, but the influence of the latter is more pervasive than that. So, assuming that we can establish that the conditions for collective responsibility hold in these cases, the form of responsibility we appeal to will indeed be national responsibility.

2. Owen raises questions about one of the arguments that I deploy against global egalitarianism (the other argument, having to do with the problems involved in finding a suitable metric by which to assess global inequality, can be left aside here). This argument invited readers to imagine two autonomous political communities, each endowed with the same per capita quantity of resources, pursuing different economic or population policies which lead over time to per capita resource levels diverging. Why, I asked, should we think that such inequality of resources was unjust, or in some other way problematic? If we confine our attention to those who were adult (and eligible to participate in politics) at the time when the different policies were chosen, this later inequality appears fair since: a) it reflects a collective choice; and b) the people who later on have fewer per capita resources are compensated in other ways, for example by having larger families. In contrast, if we look at those who enter the political community some time after the policy direction has been set, they will experience unequal starting points (measured in term of resources) and we cannot assume that the choices they would have made will always be the same as those that their predecessors actually made. To preserve equality across the societies, it seems that we have to deny them political autonomy by preventing policy choices that would lead to such a divergence.

Owen argues that this dilemma can be avoided through an international tax mechanism that would redistribute between the societies. This would not nullify autonomy any more than, say, the freedom of an individual to earn money by taking a high-paying job is eliminated by a system of progressive taxation at domestic level. But the argument is developed in terms of protecting future generations against disadvantage. However the original example does not suppose that anyone's per capita level of resources actually diminishes over time. Suppose society A, all of whose members are already comfortably above any plausible absolute poverty line, decides upon a no-growth policy, so resources levels remain constant over time (meanwhile its members enjoy ample leisure opportunities). Society B, meanwhile, decides to grow at the rate of 5 per cent per annum, allowing its members to consume more while still accumulating additional resources. What justification could there be for imposing a tax on B and redistributing to A such that resource levels in both societies increase at the same (lower) rate? Society B, though not prevented from growing economically, cannot now grow at its chosen pace; whereas Society A is being awarded resource increases that its members chose to forgo in the name of increased leisure. Now it is perfectly possible to argue that global equality should prevail as a matter of principle, so the tax mechanism is justifiable. But the example was meant to suggest not only that the members of society B would very likely resent being taxed for this purpose, but that they have good reason for their resentment, since the inequality that is being corrected for arose from collective choices made by the two societies. It is not, in the familiar though slippery phrase, an ‘arbitrary’ inequality.

What if society A were to go further still and pursue policies that had the effect of reducing per capita resource levels over time? Would redistributive taxation from B to A then be required? Although Owen's suggested ‘principle of ensuring that future generations are not worse off in terms of access to advantage than the current generation’ is a good one for nations to follow, it is not clear to me that when one nation defaults on this principle, other nations have an obligation to redistribute in its favour. The hard case will be when resource reductions mean that members of future generations do not have the opportunity to lead a decent life, measured in absolute terms. But here the case for respecting political autonomy comes into question. Just as we do not think that a political community that systematically violates the human rights of many of its current members deserves to enjoy an ongoing right of self-determination, so we should think that a community whose policies threaten the human rights (to subsistence and so forth) of its future members loses that right. What justifiably places limits on national self-determination, therefore, is not the principle of global equality, but the principle of basic human rights.

3. Owen raises a number of challenging questions about my treatment of the rights of prospective immigrants, clearly one of the most contentious issues in the current debate over global justice. One unanticipated question concerns states with national minorities, where Owen asks whether it would be permissible for a minority nation to restrict entry of fellow-citizens to its territory in the name of preserving the national culture. Suppose, for example, that large numbers of English chose to relocate to Wales, thereby threatening to dilute further the remaining Welsh culture (including areas in which Welsh is still the primary spoken language): would the Welsh assembly be justified in seeking to prevent this?

My answer comes in two parts. First, national identities in such situations are likely to be complex, if we accept the theory of ‘nested’ national identities that I have elaborated elsewhere.Footnote ² If most of those on both sides of the argument see themselves as British as well as Welsh or English, then this grounds a claim to reside anywhere on British soil on nationality grounds as well as on grounds of liberal citizenship. Second, nonetheless, since it is Welsh identity and only Welsh identity that is being threatened, there is a case for restricting the English inflow. This need not take the form of setting light to cottages owned by English expatriates! A differential property tax aimed at discouraging people who were not Welsh from buying houses in the territory might be permissible. In general I do not see that public policy aimed at controlling internal migration within liberal states is forbidden by the principles of liberal citizenship, even if physical travel restrictions would be. When governments locate some of their departments in regional offices in an attempt to boost the local economy and discourage migration to the capital, this does not invade anyone's liberal rights.

Turning now to international migration, it is important to stress that the discussion in National Responsibility and Global Justice is intended to show that it is not a requirement of justice that states should open their borders to all-comers. The aim is not to say what the best, most desirable, immigration policy for a particular state should be, since the whole point of the argument is that within certain limits this is something that should be a matter of democratic debate and decision. It is worth saying, therefore, that an ideal world might be one in which all states opened their borders but the volume of migration was relatively low – that way, those who wanted to migrate would be free to do so, but the problems associated with migration, having to do with the integration of immigrants, and so forth, would very largely be avoided. So when Owen says that various individual people – frustrated lovers, religious believers, would-be opera-singers, and so forth – should have the right to move to their chosen destination, one can agree that it would be good if states felt able to allow this. The real question, however, is whether a decision to restrict immigration violates the human rights of those who are excluded.

Owen's case against such restrictions amounts to the view that in individual cases, people may have a deep interest in pursuing their chosen goal in a particular form (life-partner, religion, opera). But could they have a human right to have this interest satisfied? One immediate obstacle is that in each case the interest requires the willing participation of other people, who have rights of their own and may not wish to collaborate – the love of my life may not wish to marry me, the religious community I ardently wish to join may decide to dissolve, there may turn out to be no call for opera-singers. That is one good reason for specifying the scope of human rights in the way suggested in the book, namely in terms of generic opportunities rather than specific ones. A rights-respecting state must allow religious practice, for example, but need not guarantee the success of any particular denomination no matter how fervently some may believe in its teachings. Owen's strongest card is the case of the lover who has actually found a partner willing to reciprocate, and certainly it would be a breach of human rights for a state actively to interfere with Romeo's attempt to get together with Juliet. Whether it must in all cases grant Juliet the right to immigrate and reside is less certain, if some other state is willing to admit the happy couple, but family reunion will certainly feature strongly as grounds for admission in any just immigration policy even if it is not strictly a corollary of human rights.

The human rights of refugees are a special case, and I am not sure that Owen and I are far apart at all on what ought to happen – namely, the creation of an international regime for managing refugee flows that distributes bona fide refugees fairly as between receiving states, instead of the present pass-the-parcel system by which states adopt strategies to try to ensure that refugees never arrive at their borders in the first place in the hope that they will attempt to enter somewhere else. It may be that in the book I was over-impressed by the difficulties involved in setting up such a scheme – particularly in agreeing on the relevant criteria of fairness – and I should have insisted more strongly on the duty that states have to establish such a regime. The problem that remains is to say what duties with respect to refugees fall on the members of a political community when no such regime exists. In principle they are required as a matter of justice to admit as many refugees as they would have to admit were the scheme in place – in other words to do their fair share of shouldering the refugee burden. Taking in more than that would at best be an (unenforceable) humanitarian obligation.Footnote ³ The problem is that there can be reasonable disagreement about what a ‘fair share’ would mean in these circumstances, partly because the criteria that should govern the distribution are open to interpretation, partly because no one can say in advance how many bona fide refugees there actually are world-wide (since it is only when a state takes somebody in for investigation that one can know whether he or she fulfils the UN Convention's definition of a refugee, or whatever other definition one prefers). It was the likelihood, indeed inevitability, of reasonable disagreement on this question that led me to the conclusion from which Owen dissents, namely that in the absence of an international agreement, each state must have the right to decide how many refugees it will offer sanctuary to within its borders. That it can abuse this right, for the reasons that he gives, is not in dispute between us.