Abstract
The article explains the essential features of a theory of global justice that combines justice for individuals with justice for political communities. It holds that arguing within the justificatory framework of cosmopolitanism is compatible with a conditional justification of states that are basically just. The justification rests on an argument I will name ‘the moral path dependency argument’. The article follows its normative consequences into the fields of a justly ordered community of legitimate states and of cosmopolitan principles of distributive justice. Thus, it reconciles the latter with claims to political autonomy of particular communities and with the fact of reasonable disagreement between them.
Keywords
I Inequality and state borders
Egalitarians argue that, from the moral point of view, it is equally important for each human life that it can flourish. The society of states, as it already exists, stands in tension to this basic principle. Its borders, and the features of sovereignty they represent, are among the factors that distribute individuals’ prospects unequally from the beginning of their lives. This fact seems to commit egalitarians to cosmopolitanism which holds that a just distribution is owed directly to individual human beings worldwide, irrespective of their various political belongings (for which only an instrumental role in realizing global norms of justice might remain).
But cosmopolitanism has its contestants, even within the egalitarian camp itself. Authors such as Thomas Nagel, Joseph Heath and Andrea Sangiovanni 1 confront it with a ‘political conception’ according to which there is a discontinuity between a morality concerned with pre-institutional values and a theory of justice concerned with political institutions. They criticize cosmopolitanism as being unduly ‘monistic’. Following Rawls, they argue that norms of justice depend upon a basic structure of society shielded by a monopoly on the power of enforcement. Because both features are lacking at the global level, the world as a whole is not a proper subject of claims of justice, at least in a narrow understanding focused on claims to socio-economic equality. Justice in this sense presupposes special relations among citizens, engaged in mutually advantageous cooperation and sharing the political system of a state. If there is such a thing as global justice, it is largely identical with the basic norms of a reasonable public international law.
I do not want to address this kind of criticism in each and every respect but mainly try to defend cosmopolitanism against the charge of monism. It can and should take care of the existence of a society of states as a fact that bears a moral weight of its own. Cosmopolitans have good reasons to adopt some basic norms of global justice that directly apply to states or nations. They have good reasons to be ‘internationalists’ in some regards. 2 Defenders of state pluralism, on the other hand, should be cosmopolitans at the elementary level of justification. And that counts in favour of some principles of global distributive justice owed to individuals.
I will explain the essential features of a theory of global justice that integrates both perspectives, but within the justificatory framework of cosmopolitanism, based on a luck-egalitarian approach. 3 The sketch of such a framework will be given in the next part of the article (part II). Adopting it is compatible with a conditional justification of states that are basically just. The justification rests on an argument I will name ‘the moral path dependency argument’ (part III). I will follow its normative consequences into the fields of a justly ordered community of legitimate states (part IV) and of cosmopolitan principles of distributive justice, including claims for open borders and freedom of settlement (part V).
For reasons of space, my argument will remain entirely within the realm of ideal theory. Such a theory is exclusively concerned with the moral justification of rules and principles. Methodologically, it supposes that there is a sufficient degree of compliance to any valid moral norms.
II Justifying norms of global justice
Impartiality
Claims of justice call for impartial justification. The conception of impartial justification that I favour has three elements: it is individualist, universalist and egalitarian. It holds that all individual human beings’ fundamental interests in life, well-being and autonomy matter, and matter equally. This does not mean that any morally relevant goods have to be distributed equally. But they have to be distributed in accordance with the equal moral worth of each individual affected.
An unequal distribution of a morally relevant good is just if and only if it can be impartially justified. And if there is no such justification, an egalitarian solution follows by default. The burden of proof rests on the defenders of an unequal distribution. 4 A prominent and, as far as I can see, basically sound principle of distributive justice gives room for inequalities as long as they are the result of free decisions and actions and not the outcome of circumstances beyond the control of the person.
To make the point somewhat more precise: a person is justified in complaining about an injustice against her if (1) she is worse off with regard to at least one morally relevant interest she has, if (2) the disadvantage cannot reasonably be assigned to the domain of her own personal responsibility, if (3) others could have prevented the disadvantage or are in a position to correct it or to compensate for it, through individual or collective action, including serious efforts to build or improve institutions, and if (4) they can make such an effort without serious risk of harm that is morally comparable 5 with the disadvantage in question. This principle is, as Ronald Dworkin puts it, both ‘ambition-sensitive’ and ‘endowment-insensitive’. 6 It is designed to equalize the standing of persons instead of letting them suffer from brute bad luck.
Now obviously, for an average child born in Ethiopia, it is brute bad luck not to have been born in a part of the western world instead. Long before the child can even begin to take responsibility for decisions and actions, many parameters of its life are fixed, and mostly to his or her disadvantage, in absolute as well as relative terms. The child faces a high risk of dying young and of remaining poor and illiterate throughout its life. What is more, the existence of state barriers makes it highly unlikely that the child will ever reach one of the wealthier states in the world. No such state accepts refugees simply because they are fleeing an initial situation of inequality and poverty.
The state system, with its barriers and armed guards, stands between the child and an access to advantages it might otherwise have. Thus the advantages turn out to be undeserved privileges of those who enjoy them in other states, similar to privileges under feudalism, which also let prospects of life depend on a person’s place of birth. 7 The child faces all the disadvantages without having made any avoidable mistake for which it can be held responsible. The disadvantages it faces are part of an overall situation which we could in principle modify through collective action and institution-building so that the disadvantage would not occur. What impartial justification, no less binding for the Ethiopian child than for an average child born in Switzerland, can there be for the existence of such a state system?
There can be no such justification, is the obvious answer. It is obvious, we might add, because my argument up to this point has been totally individualist. It has taken the existence of the state system into account, but only insofar as it stands in the way of a more equal distribution of morally relevant goods. It has not given the system any moral weight of its own.
This is no accident: an individualist, universalist and egalitarian conception of morality narrows the scope of possible defences of the state system from the beginning. According to normative individualism, we cannot take the existence of collective entities as given. If they are morally justified at all, their justification stems from the entitlements of their individual members. Some states may be justifiable in that sense, as being in the enlightened self-interest of individuals or as providing a framework of their identity-formation. But whatever an individual might owe to the state of its birth, the state has no right to neglect the independent moral standing of the individual.
With regard to universalism, a moral justification must include all persons affected, irrespective of their different social belongings. The drawing and defending of boundaries across the limited space of the earth affects humankind as a whole. Therefore, humankind as a whole has a stake in determining the lines as well as the degree to which they might be impermeable to outsiders.
Finally, with respect to egalitarianism, our duties to justify moral claims cannot be viewed as concentric circles. Each person affected must be able to accept the justifying reasons as an equal. The difference between citizens and strangers is not relevant at the most elementary level of justification. It does not stand at the beginning of an impartial reasoning; it might at best be its outcome.
Consequently, if we are allowed or even obliged to show some degree of partiality towards our fellow countrymen and countrywomen, our reasoning must satisfy the requirement of ‘second-order-impartiality’. 8 Being partial due to special political relationships must itself be universally acceptable at a logically higher level of impartiality. The ‘right to justification’ 9 does not stop at the borders of nation-states. Because borders play a part in structuring the unequal distribution of goods and ‘bads’ in the present world system, the burden of proof rests on their defenders.
Justification and application
This leads us back to the alternative perspectives on global distributive justice mentioned above: cosmopolitanism and internationalism. On cosmopolitanism, we can start with a definition given by Thomas Pogge:
Three elements are shared by all cosmopolitan positions. First, individualism; the ultimate units of concern are human beings, or persons – rather than, say, family lines, tribes, ethnic, cultural, or religious communities, nations, or states. … Second, universality: the status of ultimate unit of concern attaches to every living human being equally – not merely to some sub-set, such as men, aristocrats, Aryans, whites, or Muslims. Third, generality: this special status has global force. Persons are ultimate units of concern for everyone – not only for their compatriots, fellow religionists, or such like. 10
Clearly, a universalist, egalitarian and individualist conception of impartial justification is basically cosmopolitan in all three respects distinguished by Pogge. We can call this a ‘cosmopolitanism of justification’. The alternative at this elementary level would be an ‘internationalism of justification’. The latter is favoured by authors such as the late Rawls, who take states, nations, or ‘peoples’ as the genuine subjects to whom claims of global justice have to be justified. 11 Rawls also rejects the idea that basically the same justification of norms must be acceptable for all, regardless of their cultural or national belongings. In Rawls’ view, reasoning about norms within and among liberal societies has to be separated from such reasoning within and among non-liberal societies. I find this alternative position unconvincing: it leaves us without a compelling reason for accepting rules of justification that presuppose so much about the world as it already is, even in respects that are responsible for serious initial inequalities.
From the level of justification we can distinguish a level of application. Defending cosmopolitanism at the first level does not logically commit us to cosmopolitanism at the second. ‘Cosmopolitanism of application’ means that at least some norms of global justice directly apply to individuals, unmediated by states or other collective entities. States might mediate between individuals in a technical sense, for reasons of efficiency, without holding valid claims by themselves. Taken to the extreme, however, there would be no room for justice between and among nation-states as a genuine dimension of global justice. Along this view, there is no such thing as international justice, there is only justice among individuals.
But it is no less possible that free and equal individuals all over the world would agree that at least some globally valid norms directly apply to states or other collective entities. Although being cosmopolitans at the level of justification, they might support an ‘internationalism of application’ in some regards. The parties might even agree that no cosmopolitan norms of distributive justice are applicable at all, for example, because of the lack of a proper institutional structure for their enforcement. Perhaps we should restrict ourselves to some basic principles of public international law regulating only interstate affairs and leaving any matters of distributive justice owed to individuals entirely within the domain of each nation-state. I am convinced that we should reject such a position. But it is logically compatible with a cosmopolitanism of justification.
A ‘political’ critique of cosmopolitanism
This distinction might help us to understand what is at stake when cosmopolitanism is criticized as being unduly ‘monistic’. Advocates of the alternative, ‘political’ conception need to clarify whether they are attacking cosmopolitanism at the level of justification or only at the level of application. Nagel, for example, is ambivalent on this point. On the one hand, he refers to Thomas Hobbes, maintaining that ‘[w]ithout the enabling condition of sovereignty to confer stability of just institutions, individuals however morally motivated can only fall back on a pure aspiration for justice that has no practical expression, apart from the willingness to support just institutions should they become possible’. 12 The problems of coordination and of possible defection Nagel mentions here, however pressing they are, are problems at the level of application. They cannot as such discredit a cosmopolitanism of justification.
Nagel himself defends universalist duties of humanity, such as helping people in dire poverty, as well as a ‘pre-political’ conception of negative rights like bodily integrity and freedom of religion. But with regard to these standards the collective action problems are not fundamentally different to those raised by claims for redistribution. 13 Therefore, the Hobbesian line of argument either leads to a rejection of any norms that cannot be enforced through sovereignty or to an attempt to build or improve international and supranational organizations. A political conception that would reduce the space of the global validity of norms to those that can already be implemented within the semi-anarchic global system as it now exists would be extremely conservative.
Consequently, we should distinguish between two levels of duties with respect to universally valid norms, be they ‘negative’ human rights or ‘positive’ rights to socio-economic goods. Our duties at the first level are directly linked to the objects of norms, like the duty to refrain from violating a right or to provide a certain service. At the second level, we have duties that concern the prerequisites of fulfilling duties at the first level. Among these second-order-duties is the duty to build or to improve institutions necessary for the full realization of globally valid norms. We can therefore counter the argument that some rights are mere ‘manifesto-rights’ 14 because no addressee of the corresponding duties exists: insofar as we are able to establish such an addressee by means of coordination and collective action, and insofar as we are able to coordinate our actions and to act collectively, we have a second-order-obligation to transform the manifesto right into an effective one.
It is then an open question how far we should go in replacing national sovereignty with structures of world government. Even if such replacement would be too costly and we should therefore be satisfied with the second best, it is still possible that we have moral reasons to regret that we cannot undertake large-scale institutional reform. Whether we really have those reasons can only be clarified at the level of justification.
At that fundamental level, Nagel asks himself why we should reject arbitrary inequalities among fellow citizens as a matter of justice whereas we must not reject inequalities based on the arbitrariness of one’s birthplace for the same reasons. Like other ‘institutionalists’, he is convinced that norms of justice apply only on the basis of a common institutional background like a system of cooperation, interdependence, or power relations. 15 He therefore takes into account a bundle of necessary conditions for ‘triggering’ justice such as the coercive nature of a system, the large material effects it imposes on individuals, or the fact that we have neither chosen to join it nor are able to leave it without prohibitive costs. All these features surely characterize the society of states and the global economic order it protects and promotes. 16 But in order to show that even in combination these do not suffice to bring justice on the table, Nagel presents the following example:
The immigration policies of one country may impose large effects on the lives of those living in other countries, but under the political conception that by itself does not imply that such policies should be determined in a way that gives the interests and opportunities of those others equal consideration. Immigration policies are simply enforced against the nationals of other states; the laws are not imposed in their name, nor are they asked to accept and uphold those laws. 17
The last part of the quotation contains the condition that was missing: a state not only imposes laws on individuals, it does so in the name of its citizens and expects that they comply with the norms as if they had imposed them on themselves. Here, Nagel refers to another heroic figure in the history of political ideas: Jean-Jacques Rousseau. As citizens, we are somewhat like ‘participants in the general will’. 18 And the exceptional demands citizens make on one another bring with them the exceptional obligations of distributive justice. Where the demands are absent, there is no room for equal respect and concern above the minimalist level of humanitarian duties.
But the argument is not convincing. First of all, even if immigration laws are not imposed in the name of foreigners, they nevertheless call for their general acceptance. State borders are institutional facts that we do not simply have to take into account in the way we have to take into account the bad weather in Maine. Respecting these laws is among the constitutive norms of the given world system as a society of states. They are not only relevant for the governments of all other states but for their citizens as well. Respecting them is also not only a matter of prudence; most citizens morally expect that foreigners do not try to cross their borders by means of force, deception, or smuggling. Second, state borders, and the laws allowing or denying entrance, are among the factors that deeply influence an individual’s access to morally relevant advantages from the beginning of her or his life.
Thus, the quest for distributive justice cannot conceptually start with the fact that states are already established. It must apply to the very existence of the system of states and to the special obligations among citizens it might create. Those special obligations have serious consequences for outsiders and therefore must satisfy the requirement of ‘second-order-impartiality’. The existence of a society of states is not just an arbitrary fact underlying the presumption against arbitrariness within each state, as Nagel argues. It is an institutional fact we are collectively responsible for, and the presumption against arbitrariness belongs to the fundamental level of justification. At that level, I see no alternative to being a cosmopolitan.
But the result leaves open the question of the conceptual space my argument allows for some elements of internationalism, that is, justice between and among states as claim holders of their own, at the level of application. One might refer to purely pragmatic reasons here, focusing for instance on the robust fact that states are already given as potential veto-players whose vital interests we have to take into account in any reasonable strategy of global reforms. Or one might argue that people are mostly motivated to share their possessions with their fellow countrymen and countrywomen. 19 Surely, any strategy of transformation has to take into account that feelings of belonging and trust do not change overnight and can at best be extended step by step. 20 We should also be aware of the rather narrow limits for arriving at an intercultural consensus on substantial norms of distributive justice.
For reasons of effectiveness and of efficiency, a careful consideration of established institutions and of widely held feelings and sentiments is required. And any conception of global distributive justice has to be context-sensitive with regard to specific traditions and experiences of particular cultures and political communities. The following two arguments are designed to fulfil these requirements. The first will be shown to be a failure, but a philosophically influential and instructive one, whereas the second will be the one underlying my further argumentation.
III Two arguments in defence of the state system
Moral division of labour
A well-known defence of the society of states refers to the benefits of a moral division of labour. 21 It would be inefficient and dangerous to concentrate political competences at only one point in social space. Nation-states are more comprehensible than a world state would be; they leave some room not only for ‘voice’ but also for ‘exit’; the abuse of power is easier to identify and to correct if necessary, etc. Unfortunately, this argument is not specific enough. It does not distinguish the given system from some functional equivalents we might think of instead.
Furthermore, it is highly implausible to assume that in a situation of true impartiality (an ‘original situation’) we would have chosen just the state system as it now exists. Why should it be most efficient and secure to construe states as large as China, Russia, the United States, or India on the one hand, and as small as Luxemburg or Liechtenstein on the other? Why should it be most efficient and secure to place some states at the upper reaches and others at the lower reaches of big rivers; some on the sea and some in the midst of a desert? For reasons of efficiency and of security only, the present state system clearly is an imperfect solution. It calls for a radical redrawing of territorial as well as functional boundaries.
What sort of a moral division of labour would the parties in an original position choose, given that they are free and equal representatives of individuals? 22 This is not easy to determine, because we have to know much more about the general information accessible behind the ‘veil of ignorance’. But let us suppose that the parties are well aware of the problems highly abstract and anonymous social relations might pose for moral motivation, trust and control of the performance of power, that they are informed about the role of specific cultures in shaping a person’s identity – although this need not be national cultures – and that they see some pragmatic advantage in determining the boundaries of citizenship on a territorial, instead of a purely functional, basis. At the same time, they realize the – however ambivalent – role natural resources play in enabling economic development. 23 Moreover, the parties consider the possibility that some person’s conceptions of the good might best be realized in a political, social, cultural, or natural environment other than she or he were born into. By considering this, the parties differentiate between the role communities might play in shaping an individual’s identity and the fact that the identity of a person is never reducible to the influences of any specific community. 24
Given all this kind of information, the parties would surely vote in favour of some moral division of labour, and one form among others might be a plurality of territorially defined nation-states. Nations could help to shape a person’s identity as well as her or his sense of belonging, they could foster the person’s ability to participate in democratic deliberation, and his or her motivation to agree to fair schemes of redistribution. 25 But the states would be much more equal in territorial scale and in access to natural resources than the states in our real world. (Perhaps they would be more like the emerging regional organizations above nation-states such as the European Union and the Association of South East Asian Nations.)
Yet, while favouring some moral division of labour, the parties would also consider some redistribution, because even in an ideal world not all regions could provide equally good conditions for human flourishing and development for all and under all circumstances. First of all, the parties would consider natural resources to be a common asset of humankind that would have to be allocated to the greatest benefit of all and especially to the advantage of those least well-off in terms of natural endowments. For basically the same reason, they would also agree to generous regulations of freedom of movement, and not only with regard to emigration but to immigration as well. 26
We must not go too far in speculating about the exact shape of such a world. There are too many dimensions of indeterminacy for arriving at definite conclusions beyond the highly abstract ones I have just mentioned. But whatever the exact outcome might be, the result would clearly be negative regarding the given society of states as a favourable form of a moral division of labour.
Moral path dependency
Let us now take a look at a second argument that might be more promising. I will call it the moral path dependency argument. 27 ‘Moral path dependency’ means that a fact which is morally arbitrary in its origins gains some moral importance thereafter, in the course of history. Among those facts is the shape of the existing state system. While it is the outcome of practices such as forced marriages, post-feudal warfare and colonial drawing of boundaries, later on some states have developed into institutionally thick and effective promoters and protectors of basic principles of justice. They have incorporated universally valid moral claims, yet in specific forms based on specific interpretations of their content.
Although no state in the real world is fully just, some are procedurally open to an ongoing approximation of unrestricted justice. States that are basically just in this sense have institutionalized a reasonable version of sovereignty of the people, a regime of basic rights including social rights like subsistence, basic health care and elementary education, and a minimally acceptable version of equality of opportunity in order to take care of the fair value of civil and political liberties. In doing so, they provide all of their members with justifying reasons to identify with them as their states. 28
To be sure, there is no logical connection between political autonomy and a plurality of nation-states – as opposed to one single world state. But as a result of political struggle, at some places ‘the particular’ has provided space for roughly fair terms of social and political participation based on shared feelings of responsibility and belonging. 29 Under the auspices of political powers over bounded territories some political communities now satisfy standards of political morality above a certain threshold. And what is more, these standards are provided with a concreteness they would lack in a ‘pure’ form, untouched by any particularities.
The institutions and procedures of any basically just states are in part responses to historically specific challenges and experiences. Without this specificity, the principles of human rights, sovereignty of the people and fair equality of opportunity would suffer from too much indeterminacy. In order to realize universally valid norms, citizens have to grasp their meaning in the light of what they take to be the morally most significant facts of their and their ancestor’s common history. 30 So, realizing them is always more than simply applying an already fixed content to a specific context. Consequently, the universal validity of basic principles of political morality should better be interpreted in terms of an ‘iterative universalism’ than a ‘covering-law universalism’, to use Michael Walzer’s famous distinction. 31
In the real world a regime of justice always bears the signs of its historic origins. Inevitably, this leads to some pluralism within the realm of justice, even on the basis of shared moral principles. Within this realm, there is room for ‘reasonable disagreement’. 32 As the individual person’s views are in part shaped by her or his different national belongings and the histories of his or her political communities, we have reason to apply the principle of tolerance on the fact of state pluralism.
To sum up, the argument from moral path dependency starts with the arbitrariness of the given boundary lines, but it results in tolerating differences within the realm of justice. Borders are acceptable or even required as long as they allow free and equal citizens to pursue the flourishing of basically just political communities. Part of the common project of a citizenry might be to realize a reasonable version of justice in a historically and culturally adjusted form. Such a form deserves toleration. Therefore, justice also requires tolerating a plurality of political communities.
To be sure, this justification of the existence of borders is strictly conditional: it rests on normative individualism all the way down. It does not presuppose any normative collectivism. States are never ends in themselves, only individuals are. Nonetheless, individuals might have good reasons to develop a political sense of belonging with regard to their states. The moral path dependency argument can thus serve as an indirect justification for the existence of basically just states.
But only of basically just states: with regard to the given society of states as a whole the results are still clearly negative. The existing society of states includes legitimate as well as illegitimate members. Explaining the normative consequences of the all-too-obvious fact that not all states in our real world are even basically just would draw us deeply into problems of non-ideal-theory, from incentives to morally justified reform to ‘humanitarian intervention’. In the case of refugees and stateless persons, realizing cosmopolitan principles between as well as within states is a matter of life and death. In the case of gross violations of basic human rights on a large scale, claims to sovereignty must be overridden in order to save huge numbers of individuals. 33 But for reasons of space, I will leave such problems aside and simply conclude that a society of states is not as such incompatible with norms of justice. This raises the question of the relative moral weight the claims of states can have compared with those of individuals when the former are at least basically just.
IV Internationalism
Following the moral path dependency argument, individuals can have good reasons to identify with their states. Basically just states are qualified to represent individuals in the international sphere. As true representatives of natural persons, states are ‘moral persons’ with claims of their own. This gives international justice, justice between and among states, some legitimate room at the level of application: international justice is an integral part of global justice. I will now turn to its main implications.
In the world of ideal theory, we would first find basically just states only. I differ from Rawls in this regard, for I do not think we should include ‘decent hierarchical societies’ in ideal theory. 34 Such societies would not provide all citizens with sufficiently good reasons to identify with them as their societies. A decent hierarchical society includes different classes of members; for example, the ‘orthodox’ at the top and the ‘unbeliever’ at the bottom. And because it neither guarantees rights to democratic participation – as opposed to a ‘decent consultation hierarchy’ granted to groups – nor a minimally acceptable version of equality of opportunity, it is not procedurally open to an ongoing approximation of real justice.
Rawls himself concedes that ‘decent hierarchical societies’ are not as just as ‘well-ordered’ liberal societies. His ultimate reason for including the former in ideal theory seems to be that he wants to avoid the suggestion that any society that is not truly liberal is in principle a candidate for non-military or even military intervention. 35 But even admitting as much, we are not forced to include non-liberal societies in ideal theory. We can morally criticize a society – for example, for not being democratic or sufficiently fair to women or members of minorities; and we can even disadvantage it – for example, by excluding it from certain agreements and privileges like the ‘most favourable’ principle or the international borrowing privilege; 36 and we do so without disregarding its sovereignty.
Second, in the world of ideal theory, there would only be fair relations between and among the states. All basically just political systems are entitled to free and equal membership in a community of legitimate states. The community is effectively regulated by five principles of a just public international law.
The first one, which is logically prior to, and a necessary component of, all the other principles, holds that every treaty is binding upon all its parties: pacta sunt servanda. The second principle forbids any interference in the internal affairs of other states against the will of the demos concerned. It requires mutual recognition of inviolability in the essential domains of sovereignty of the people. The third principle forbids any formal or informal discrimination; for example, in international trade. All peoples should be free, without pressure or fear of disadvantaging, to take decisions following their own reasonable public conception of political morality; for example, regarding workers’ rights, public health care, animal welfare, or environmental protection. 37 The fourth principle prescribes an equal right of any state to participate in the making, application and enforcement of the global rules of conduct. Finally, the fifth principle guarantees any state the effective freedoms and resources needed to retain just political institutions and to function as a free and equal member in the international community, especially with regard to the fourth principle.
The fifth principle obviously is a requirement of prudence, because failed states could easily destabilize the entire community of states. But it is also a precept of justice, at least in the following version: if a basically just state, through no fault of its own, is unable to play its role as a free and equal member in the community of states, it is entitled to international assistance. This is similar to the duty to assistance justified by Rawls. 38 Unfortunately, Rawls does not go substantially beyond this duty. He does not accept cosmopolitan principles of distributive justice, that is, a just distribution directly owed to individuals, apart from a minimalist right to subsistence. And he does not say a word about claims to global freedoms of movement and of settlement.
V Cosmopolitanism
But even in ideal theory, cosmopolitan principles of distributive justice should find a place, along with the principles of a just public international law. As I argued earlier, the parties in an ‘original position’ would accept some redistribution due to the fact that even in an ideal world not all regions could provide equally good conditions for human flourishing and development for all and under all circumstances. They would also agree on generous regulations on migration and would consider the possibility that a person might choose to follow his or her conception of the good in a political community that is not his or her own from the beginning of his or her life. By introducing both sorts of principles, the parties would do justice to the fact that human beings are always more than members of collective entities. They remain individual persons with claims irreducible to those of states.
This is even more important if we accept the moral path dependency argument, which leads to the conclusion that at least some of the states in our real world are basically justified – including their existing borders. But these borders separate large states from small ones and territories rich in natural resources, including access to the sea and large rivers, from those that are poorly endowed by nature. The living conditions in rainy Norway would still vary significantly from those in dry Mongolia if both states were well-ordered and had sufficiently just constitutions.
We should therefore introduce some elements of a ‘cosmopolitanism of application’ into ideal theory. They would include entitlements to just distribution of economic goods and resources as well as claims to freedom of movement and of settlement. Taken to the extreme, cosmopolitanism would lead to a world in which the fate of birth no longer influences any individual’s access to advantages but is totally replaced in that regard by criteria such as need or merit which, other than the place of one’s cradle, are not morally arbitrary. 39 According to the principle outlined at the beginning, this would require redistribution in order to exclude, or to compensate for, brute bad luck.
Unfortunately, we will never be able to neutralize the effects of someone’s birth and upbringing in a particular country and culture up to the point of unlimited equality of opportunity for all the children of the world. What is more, even trying to do so would lead to serious tensions with the valid claims of basically just states. Without some control over borders and some space for democratic decision-making, particular political communities could not play any independent role. We also have to take into account the fact of reasonable pluralism within the realm of justice. The universalist principles of political morality are much too abstract to be realized in a ‘pure’ form, devoid of particularities.
For both of these reasons, political autonomy and reasonable disagreement, we should favour a conception of global distributive justice that allows for morally acceptable differences. Consequently, I propose to deviate from a strictly individualist understanding of cosmopolitan global justice in two main ways.
Open borders
First, we should introduce a rule of priority: global freedom of movement and settlement is a prima facie right of all citizens of the world but one that is ranked lexically lower than the right to retain sufficiently just political communities. Only insofar as a community is able to preserve its basically just institutions are its members obliged to let newcomers settle among them and to grant them full rights to citizenship after an acceptable short period of settlement.
This proviso entails what Joseph Carens calls the ‘public-order-restriction’. 40 But it also rests on a deeper consideration: opening borders must be compatible with due respect to the particular ‘roads to justice’ leading through a plurality of political communities. Immigration should not destroy the social bonds and expect too much of the political institutions in which a particular yet reasonable interpretation of justice has taken shape. 41
In any case, given just background conditions, it is not too likely that the amount of migration would pose an existential threat to particular communities. Most people prefer to stay in countries or at least in regions they are already familiar with. In our real world, the largest proportion of migrants wants to leave behind seriously restricted conditions. This applies not only to those accepted as recognized refugees. Even among ‘voluntary’ migrants many would stay in their native lands if only the global structure were more egalitarian. In short, important aspects of the problems open borders might pose for political communities would surely disappear under conditions that we are already obliged to foster for reasons of justice.
Economic goods and resources
The second aspect of how my argument deviates from a strictly individualist understanding of cosmopolitan global justice is my view that although some redistribution of economic goods and resources is directly owed to individuals, their political communities should still play an important role as its direct recipients, and not only for reasons of efficiency. Political communities provide the space for public deliberation necessary for determining the goods and the principles of a just distribution owed to individuals.
Due to the fact of reasonable pluralism, people in different communities and cultures will differ in both regards to a certain degree. They might, for example, attach different value to economizing and busyness on the one hand, pleasure and leisure on the other. They might differ about where exactly we should draw the line between a person’s responsibility and that of her or his community, which goods should be publicly available and which should be at the disposal of individuals, etc. Therefore, a global welfare state, unmediated by particular communities, would not only in all probability turn out to be inefficient, it would also surely act unjustly, for it would impose some particular, and reasonably contested, interpretation of distributive justice all over the world.
What is more, citizens of basically just states have reasons to attribute intrinsic value to their membership. Good citizens appreciate having political autonomy in a state they reasonably accept as their own. And given the plurality of basically just states, we should encourage individuals to take seriously their political responsibility as citizens of particular communities. Because the justice of the world order depends on the adequate functioning of its constitutive elements, and states take part in constituting a just world order, each individual citizen should take care of the flourishing and ongoing improvement of at least one particular state. And that requires having some local democratic competence, even with regard to cosmopolitan norms of distributive justice.
Finally, a consideration of efficiency goes in the same direction: we should avoid setting incentives that are self-defeating from the point of view of justice. Governments of states shall ensure good governance as far as they can. But the incentives to do so would be diminished if a global welfare state automatically guaranteed a just distribution to individuals no matter what the performance of their local states. The states could then simply rely on institutions of global governance instead of deciding and acting carefully by themselves.
But it is equally reasonable to suppose that all citizens, wherever they live, want to determine the courses of their communities under fair background conditions. Assuming that any state is a fully just mediator of its citizens in the global realm, and any individual is a citizen of at least one fully just state, fully just background conditions would suffice to conclude that any individual, as a ‘citizen of the world’, receives her or his fair share of economic goods and resources. On the other hand, if a particular individual does not receive his or her fair share, then, given fully just background conditions, it follows that at least one state is not fully just. In any case, fully just background conditions would guarantee that the ‘outcome-responsibility’ 42 for a remaining injustice belongs to a particular political community. If the community is at least basically just, as we suppose in ideal theory, it is procedurally open to cure the injustice by means of democratic deliberation and decision-making. And that, to repeat, is part of the responsibility each political community should bear by itself.
Consequently, each individual has a cosmopolitan entitlement to fair economic background conditions to his or her democratic engagement as a citizen. According to our general principle of distributive justice, the fairness of the background conditions should be determined in a way that is both ambition-sensitive and endowment-insensitive. But contrary to the ‘luck-egalitarian’ principle of distributive justice I started from, the resulting principle of global distributive justice entails collective responsibility on the part of the entitlement bearer.
Across the borders, an individual is justified in complaining about an injustice against her or him if (1) the citizenry is worse off economically, if (2) the disadvantage cannot reasonably be assigned to the domain of the citizenry’s democratic responsibility, if (3) others could have prevented the disadvantage or are in a position to correct it or to compensate for it, through individual or collective action, including serious efforts to build or improve institutions, and if (4) they can make such an effort without serious risk of harm that is morally comparable to the disadvantage in question. In short, if a particular society’s economic competitiveness is inferior to some other, it must be due to its own democratically accountable decisions and actions and not due to influences from outside the domain of its collective responsibility.
One of the influences a particular society cannot initially control is its access to natural resources. Even in ideal theory, states face unequal starting conditions in this regard. We should therefore introduce a scheme of progressive taxation, such as Thomas Pogge’s Global Resources Dividend. 43 Whatever the technical details of the scheme, it should ensure that the unequal distribution is to the greatest advantage of all states, especially those worst-off in terms of natural endowments. On the other hand, it should provide any state with incentives to diminish the dependency of its economy on resources that are not renewable.
A possible objection to my proposal holds that it is not truly cosmopolitan because it does not necessarily do justice to individuals: why should an individual suffer from a decision of her government with which she might disagree? 44 Even in the most democratic systems the best we can hope for are decisions backed by the will of a majority. And the whole point of cosmopolitanism is that we shall not reduce the valid claims of individuals to those of states and their representatives. 45
It is worth concluding with two considerations. First, according to my ideal theory, borders and societies are generally open to immigration and freedom of settlement. This norm shall secure that an individual who does not identify with his initial political community does have a fair chance to become a member, endowed with equal rights, elsewhere in the world. Second, among the universally valid claims of individuals there are claims to democratic citizenship as well as claims to distributive justice. But in order to fulfil claims of the first sort it is impossible to pass over the distinct responsibility of states. That is why cosmopolitans should concentrate on just background conditions for any local democratic deliberation and decision-making. 46
To be sure, even in combination, just economic background conditions and generally open borders cannot guarantee just results in any individual case. Therefore, part of the proposal is a moment of moral compromise. But hopefully, this concession is outweighed by the central strength of my approach: that it combines internationalist with cosmopolitan norms of global justice and that it reconciles the latter with claims to political autonomy of particular communities and with the fact of reasonable disagreement between them. Some cutbacks on justice to individuals are a price we have to pay for being cosmopolitans but not monists.