Moral and political cosmopolitanism, weak and strong: the distinction(s)

Moral–philosophical argument in support of cosmopolitan principles of justice goes hand-to-hand with political–philosophical critique of the contemporary institutional order. Heavy criticism is levelled especially against state sovereignty as the still fundamental principle of international law, reportedly the biggest obstacle to realization of political justice, as usually expressed by civil and political rights and liberties. Also questioned is the non-existence of global redistributive mechanisms that would ensure at least partial realization of social/economic justice on a global scale – state sovereignty, both in its internal and external aspects, being again the root of the problem (e.g. Beitz Reference Beitz1991; Pogge Reference Pogge1992; Kuper Reference Kuper2004, 43–5, 191–202). Superficial reading would suggest that cosmopolitanism simply calls for ‘exchanging’ the territorial state with a world-state, which would then fulfil virtually identical functions – although on a much wider scope.Footnote ³ Such an image gives rise to weighty concerns; however, this is where the key conceptual distinction comes to the fore, on the grounds of which undesirable institutional consequences are to be avoided.

Cosmopolitans distinguish here between moral and political (sometimes also institutional or legal) cosmopolitanism. The former refers to a set of beliefs that can be variously formulated: That all human beings are ultimate units of moral concern, that they share certain morally relevant features that should always bear upon philosophical considerations about justice, that at least some of our duties and obligations concern all human beings and not only our compatriots, that territorial state borders are morally arbitrary, or that all members of this community of mankind are morally equal in some essential sense. In short, moral cosmopolitanism expresses the belief that all people in the world belong to one all-encompassing moral community. Importantly, this argument is independent of the more ambitious ‘cultural’ claim (and surrounding debates) that a cosmopolitan ‘we-feeling’ grounding some kind of common identity is either a palpable reality or a ready-to-be-introduced potentiality (Hutchings Reference Hutchings1999: Ch. 2).Footnote ⁴

Crucial not only to my argument but to liberal cosmopolitan justice as such is a further distinction between weak and strong (sometimes also thick and thin) forms of MC, referring to the scope or depth of moral considerations, especially with regard to socio-economic justice or the contents of the list of human rights. There is no clear-cut criterion or method for ranking someone strong or weak in this sense, and these terms are anyway probably better thought of as poles of a continuum. Nevertheless, as a rule of thumb, invoking comparative judgements of justice or equality, defending the ‘full package’ of international human rights, or calling for democratization of international/global institutions for reasons of principle (such as equality of status or individual and collective autonomy), are indications of a strong position on MC.Footnote ⁵ Although my primary target are cosmopolitan egalitarians who hold such strong positions either explicitly or implicitly, the importance of the ‘weakness’ or ‘strongness’ of one's position resurfaces in the work of those authors who self-consciously avoid ‘strong’ moral language on these issues, and thus the type of reasoning I present here has probably more general application (I will briefly return to this issue in the penultimate section).

PC then stands for the belief that realization of requirements of justice, as specified by moral cosmopolitanism, should be entrusted to global political institutions with legitimate coercive power, which in turn would rest on a global system of positive law to which jurisdiction of individual states would be subordinated – in other words, a version of a globalized modern state of which all people in the world would be citizens. The first (moral) view, however, does not necessarily entail the second (political) – one can hold the strong moral position and still avoid world-statism. Or so argue defenders of cosmopolitan justice when vindicating its desirability, in order to avoid the controversial idea of a world-state or some other kind of global coercive political authority that would take over basic functions of the state. This argument is regularly used in contemporary writings on the subject.Footnote ⁶ Different models of how the cosmopolitan world should be organized are then available, ranging from moderate reforms within the existing international state system to a complete transformation towards a ‘post-Westphalian’ global order.Footnote ⁷

This is then the refined version of the MC–PC distinction, which I want to show is unconvincing, practically speaking (conceptually, it does make sense to treat MC and PC as separate categories). In its basic contours, my argument expands on certain elements of the position articulated recently by Thomas Nagel (Reference Nagel2005). Drawing on Hobbes and Rawls (and partly Kant), Nagel argues that socio-economic justice is an associative obligation and presupposes a system of positive law backed by monopoly of force; he takes the existence of a state-like entity to be an ‘enabling condition’ of justice (see also Ypi Reference Ypi2010, 178ff.). In turn, coercive enforcement of egalitarian political measures is justified by the political authority's acting in the name of each and all of the citizens. Beyond the state, only weak standards of basic human rights and of justification and conduct of war (plus certain humanitarian duties) are applicable, that is, unless a world-state that would maintain what Rawls called background justice is in place. Nagel's position is sophisticated and has sparked a good deal of philosophical debate; what is interesting for my present purposes is his questioning of the idea that underpins many suggestions to transcend the state vs. world-state dichotomy – namely that of a ‘sliding scale of degrees’ of co-membership in nested or overlapping networks of political institutions, among which the responsibility to enforce various duties of justice should be distributed (Nagel Reference Nagel2005, 140ff.). Not only is such a multi-level scenario unstable, says Nagel, but it does not really ‘make moral sense’ either, because in the absence of a coercive global authority it is unclear from where individual obligations of justice and equal consideration originate.

Two conceptual disclaimers are useful here. I use the term political cosmopolitanism as a generic one covering also the institutional and legal labels, although the authors discussed below usually prefer one or another. Richard Beardsworth makes a case for keeping them analytically separate and even throws in the notion of normative cosmopolitanism as the bridge between theory and practice, arguing that all four of them plus MC can be fruitfully combined in order to arrive at a plausible (weak) conception of cosmopolitan international theory (see also Slaughter Reference Slaughter2010, 184–5; Beardsworth Reference Beardsworth2011, 21–46). Although making finer distinctions of this kind might be useful for some purposes, my general point in this text is that no matter how many of them can be made on the political side, the problem lies in the moral modality of cosmopolitanism, without which no theory of cosmopolitan justice and democracy can get off the ground in the first place.

Also, I associate PC strictly with the idea of a world-state or world government, as do most authors discussed in detail below.Footnote ⁸ Weaker versions of the new institutional order to which they do subscribe are labelled by them as ‘cosmopolitan federalism’, ‘global political structures’, ‘multi-level cosmopolitan democracy’ and so on. This is admittedly a stipulative definition, although one that is supported by usage by cosmopolitans themselves. Another reason why I refuse to identify PC with this plurality of visions of a new world order, including world-statism as its ‘strong’ version (see e.g. Dower Reference Dower2010, 3–4), is precisely my aim of exposing the sliding up and down the scale of institutional demandingness on the part of strong moral cosmopolitans, while they retain the fancy notion of MC and simultaneously avoid the language of world-statism. Be it purposive or unintentional, this rhetoric conceals that strong MC has a much fuller political agenda than its defenders would want to admit. In short, I do not think one can have strong MC and ‘weak PC’ at the same time.

Readers will probably notice that all this is not a contemporary invention: The distinction is to be found in an incipient form in the thought of Immanuel Kant whose moral philosophy reveals an unambiguously cosmopolitan orientation (Kant Reference Kant1991c, Reference Kant1997).Footnote ⁹ Kant links the capacity of exercising rational will (i.e. of reason) to acting under freedom and subsequently to the value of human dignity, which is inherent to all human beings. According to Kant, there is a duty to enter into a civil – in his case, republican – condition and defer to the decision making and coercive powers of political authority, so that impartial and efficient mechanisms of enforcement of ‘freedom for all’ are in place (Kant Reference Kant1991a, 84–7; Reference Kant1991b, 99–102). Importantly, this duty affects both individuals and states alike. Since, according to Kant, the question of justice emerges once there is interaction among individuals or their groups, owing to the need for adjudication of disputes (Kant Reference Kant1991c, 136–43), and because states do and will interact, the requirement to determine principles and institutions regulating international relations appears inevitable.

On the international/global level, however, Kant ultimately backed off from his seemingly strong cosmopolitan commitments, not least because he grappled with some of the issues that have resurfaced in recent cosmopolitan thought. Instead of a global public law enforced by a global republican political authority, Kant proposed ‘merely’ general principles of international law setting up a federation of sovereign states with republican constitutions, plus the cosmopolitan right (and corresponding obligation) of hospitality (Kant Reference Kant1991b, 105–6; Brown Reference Brown2010). Following Beardsworth (Reference Beardsworth2011, 38), we can speak of a dilemma of global right: Rights and freedom of both individuals and states can be secured only through hard law backed by coercive power of a world republic; however, it is precisely by granting these unprecedented powers to the potential government of the ‘state of states’, or even a ‘universal monarchy’, that there arises the twin danger of global despotism (through abusing power) and paternalism (through disrespecting peoples’ right to autonomy, even if the global authority is led by good intentions). This is why Kant opted for a league (federation) of nations as the second-best solution, hoping that ‘universal history’ will eventually – after gradual evolutionary progress – culminate in a genuine and just cosmopolity (world-state). Until then, notions of socio-economic (distributive) justice have no place in cosmopolitan political thought, and this ultimately ranks Kant among weak moral cosmopolitans. His political philosophy of international justice represents an early attempt to reconcile strong moral cosmopolitanism with a rejection of world-statism.Footnote ¹⁰

Contemporary defenders of liberal egalitarian cosmopolitan justice use the distinction as a basic presupposition of their theoretical reflections, notwithstanding the differences in their substantive argument. Charles Beitz states that ‘there is no necessary link between moral and institutional cosmopolitanism’ and that supporters of moral cosmopolitanism are ‘not necessarily committed to the belief that the world should be reorganized as a unitary or stateless political and legal order’. For this reason, Beitz argues, ‘the widely alleged undesirability of world government is not a good reason to reject the ethical aspiration it represents’.Footnote ¹¹ Brian Barry adds that ‘there is no inconsistency in counting the interests of everyone in the world equally and concluding that those interests will tend to be best advanced by a state-centered system with only weak international authority’. According to Simon Caney, moral cosmopolitans largely ‘think that cosmopolitan moral claims are compatible with, or even require, states or some alternative to global political institutions’, and thus ‘critiques of a world-state or global political authorities do not impugn the moral convictions that all persons are of equal moral worth and that everyone has duties to other human beings’. Thomas Pogge sees the substance of moral cosmopolitanism in ‘a requirement that imposes limits upon our conduct and, in particular, upon our efforts to construct institutional schemes’, which is a ‘more abstract, and in this sense weaker [view] than legal cosmopolitanism’. Instead of identical rights and obligations stemming from citizenship of one universal republic, we can imagine ‘a system of autonomous states and even (…) a plurality of self-contained communities’ (Pogge Reference Pogge1992, 49; Beitz Reference Beitz1994, 125–6; Barry Reference Barry1998, 144–5, 153–6; Caney Reference Caney2005, 5, 152). This sample underpins the claim that the distinction between moral and political cosmopolitanism is embraced by many leading cosmopolitans about justice. In contrast to Kant, however, they adopt almost without the exception of the strong version of MC.

Moral cosmopolitanism and negative duties: proving too little

However, the distinction is problematic in two respects – either MC is weak and proves too little, making far-reaching institutional changes unnecessary, or it is strong and proves too much, meaning that the only alternative capable of realizing its goals amounts to a world-state. The relatively uncontroversial claim that all people are equal in some morally relevant sense could imply ‘only’ a requirement to refrain from actions that would be harmful to others, or that would result in using other people merely as means and not at the same time as ends in themselves (to borrow Kant's phrase); David Miller (Reference Miller2002, 81) speaks of a ‘platitudinous element’ in cosmopolitan thought. To put it differently, moral cosmopolitanism might, as a matter of justice, involve ‘merely’ a set of negative duties, supplemented perhaps by cultivation of selected virtues; this would constitute a radically weak version of MC. Even if the Kantian duty of hospitality was added, this would be still a far cry from full cosmopolitan duties of egalitarian justice. A typical objection builds on the conviction that the retreat to negative duties makes ultimately no difference because upholding even the most basic of these (such as not harming, not exploiting or stealing, or not using others for one's own ends) would necessarily require a radical transformation of international economic and political structures, or at least extensive redistribution of goods from rich to poor countries. Such an argument has been put forward by several authors; here I will discuss in some detail probably the most well-known version elaborated by Thomas Pogge.

Pogge couches his view in a distinct institutional conception of human rights, derived from Article 28 of the Universal Declaration: ‘Everyone is entitled to a social and international order in which the rights and freedoms set forth in this Declaration can be fully realized’. He proposes to conceive of human rights as ‘moral claims on any imposed institutional order’ and thus ‘moral claim against anyone involved in their design or imposition’ (Pogge Reference Pogge2002, 165, Reference Pogge2008a, 52). These ‘coercive social institutions’ or ‘any other social system’ are then bound to ensure secure access to the object of the given human rights, with the qualification ‘insofar that this is reasonably possible’ (Pogge Reference Pogge2008b, 70). What we get, then, is a negative duty not to participate in upholding unjust socio-economic international order; if this order is judged unjust, all who are responsible for or profit from its continuing existence are obliged to compensate those who find themselves on the losing end (since a complete transformation is not usually a real or reasonable alternative) (Pogge Reference Pogge2008a, 56, Reference Pogge2008d, 202–5). Since the present-day international system is geared heavily towards the interests of governments (and citizens) of powerful and rich countries who actually spearheaded the process of setting up the grounds and rules of this order, it is primarily their responsibility to ensure fulfilment of human rights worldwide (Pogge Reference Pogge2002, 175, 185).

To put it emphatically, failing to uphold the given duty means that we become causally responsible for the lot of those whose rights have been violated or have not been fulfilled. The point of Pogge's argumentation is that the cause of global poverty and thus of violation of a number of human rights is to be ultimately located in the international political and economic order and not the actions of individual societies and/or states – not least because we can imagine alternative versions of such order where these moral evils would be absent (Pogge Reference Pogge2007, 34ff., Reference Pogge2008c, 118–22). In this sense, it fulfils the criteria of ‘imposed institutional order’, and rather demanding requirements for the transformation of this order follow, including ‘more world government’ that would regulate a host of global interactions (Pogge Reference Pogge1994, 224, Reference Pogge2007, 28).

I see two possible replies here. First, it should be noted that Pogge's negative approach is in fact dependent on taking the full package of international human rights as the set of criteria for detecting injustice, which is a strong moral position (more on this follows in the next section). There is, I believe, nothing inevitable about arriving at cosmopolitan political structures from the starting point of negative-duties MC. Now it has to be admitted that few theorists approach international (in)justice strictly through the lens of negative duties (think of libertarians of both the right and left sort), and weak MC thus does not map neatly onto negative duties. Still, ND form the bulk of such a weak position – and perhaps of any plausible moral point of view – and I will argue later that one should proceed carefully when adding further positive duties, such as those related to enforcement of international human rights.

More importantly, as Miller points out, it is one thing to construct hypothetical probabilistic scenarios of alternative international order(s) and quite another to assign responsibility for violating (and therefore the obligation to protect/compensate/etc.) a given set of human rights. Miller cites several works on economic history documenting that roots of poverty are to be found primarily in domestic factors (Miller Reference Miller2007, 242–4). Although the relative weight of domestic and global causes will be disputed, and such data therefore cannot be immediately used as a direct refutation of Pogge's claims, one of the most significant examples of human rights unfulfillment – global poverty – is most likely a multicausal phenomenon. This casts doubts over Pogge's attempt to put forward a definite mechanism of assigning responsibilities for global injustice, and ultimately over the entire institutional scheme he proposes: If responsibility shifts for each case of injustice, it is by no means clear that permanent global political structures are the morally required answer. This can be argued also from the obligation bearers’ perspective: Although it may turn out that present citizens of poor countries cannot be always held morally responsible for the deeds of their ancestors, or that causal responsibility does not directly translate into moral responsibility (Reddy Reference Reddy2005, 123–4), such an argument cuts both ways – that is, neither (citizens of) rich countries can be subjected to such simplifying logic. Normative reasoning seems to be dependent here on difficult empirical calculations, and it may turn out that these do not, or do not always, support cosmopolitan conclusions.

There is a more general point lurking behind. Although counterfactual assessment of a given state of affairs is probably unavoidable in political theorizing, it is doubtful whether we can simply postulate an evil-free or at least morally significantly better hypothetical alternative for the sake of the comparison and proceed with the rather trivial adjudication. Pogge speaks about ‘foreseeable’ negative consequences of an institutional order that were ‘reasonably avoidable through a more even-handed institutional design’ (Pogge Reference Pogge2007, 41), but to state in hindsight that some negative state of affairs had been foreseeable before the larger framework was brought into existence seems to get the logic backwards. Moreover, Pogge himself cannot prove in advance that his preferred alternative scenario would not have equally negative consequences, which would then seem equally foreseeable, especially since he in effect advocates radical structural transformation both domestically and internationally (even if split into several steps): Recalling Kant, and anticipating Barnett and Finnemore's analysis of IOs, one may wonder why these new political and administrative structures should behave according to the philosopher's plan, avoiding both negative unintended consequences of their activity (such as the disappointing effects of international aid, or the uncertain legacy of humanitarian interventions) and the usual expansionary logic of political and administrative structures.Footnote ¹² Kurki and Suganami (Reference Kurki and Suganami2012, 427), otherwise staunch defenders of causal analysis in IR, warn before ‘overconfidence in one's prescriptions based on causal assertions’, and I consider their point to be especially pertinent when facing projects of grand institutional redesign of the globe.

It could be argued that moral cosmopolitanism and corresponding assignment of political responsibility need not depend on Pogge's institutional scheme and that our obligations to the rest of humanity can be defined in a different way – for example, by conceiving of a certain set of human rights, or of certain principles such as equality of opportunity, as unconditional claims to be met, even if there is no apparent chain of causation in their underfulfilment or violation (also Miller Reference Miller1999, 198–204; see Shue Reference Shue1996; Caney Reference Caney2001). In such a case, much, including the cogency of my present argument, would depend on the ambitiousness of the list: My hunch is that a comprehensive (‘strong’) account would take us quite far to the realm of strong ‘positive-duties MC’, against which my ‘proves too much’ objection is targeted.

The claim that negative duties unavoidably morph into positive ones and as such would have to be fulfilled by a system of cosmopolitan political, economic and judicial institution is therefore too quick. Doing away with existing injustices (understood in the negative or minimal sense) does not necessarily imply dispersion of state sovereignty, constructing a multi-level governance system, establishing international courts or launching global mechanisms of goods redistribution (all being typical cosmopolitan suggestions and requirements). It may entail ‘only’ stricter observation of the rules and principles that the existing states have subscribed to, such as eliminating protectionist measures on the part of richer countries or more intensive and more consistent application of current instruments of persuasion and change – for example, diplomatic pressure, economic sanctions, and where necessary, also armed intervention under the United Nation (UN) patronage.Footnote ¹³

The expected cosmopolitan reaction will emphasize that without establishing new global institutions outright, and/or without radical transformation of the existing ones, neither negative nor moderate positive moral goals have any chance of being achieved. It will be restated that the new system of governance is not the same thing as a world-state, and that Kant's fears of despotic or unstable nature of a global political authority, echoed in John Rawls's restrained conception of international order, are either unfounded or limited by the intellectual horizons of Kant's 18th century – mainly thanks to the capabilities of contemporary political imagination and institutional design (Habermas Reference Habermas1997; Rawls Reference Rawls2001, 36; McCarthy 2002, 251). Now, to reiterate my central contention, this argument as a whole depends on the nature and scope of moral claims that are put forward on behalf of cosmopolitan justice. If, for example, we follow authors such as Jack Donnelly or Thomas Pogge in taking the contemporary package of human rights (as defined, roughly, by the two major international covenants on HR – the ICCPR and the ICESCR) as the regulative ideal for cosmopolitics – which is in fact a fairly moderate liberal position, at least when compared with egalitarian visions of global justice – then it surely is difficult to see how this ideal might be approached without some major changes in the international order.Footnote ¹⁴

The question remains, and the answer cannot be simply asserted in accordance with one's preferred institutional vision, whether the full human rights package, or principles of egalitarian justice, can be secured without introducing global political authority with legitimate coercive power that will have to be actually and frequently deployed. To drive the point home, the distinction between moral and political cosmopolitanism appears to have, in itself, little purchase on the desirability or feasibility of cosmopolitan justice: What counts is the substantial content of the two halves of the pair, especially of the first one. If a given conception of MC is or turns out to be fairly or very strong, then MC most likely entails PC, claims to the contrary notwithstanding (see also Hurrell Reference Hurrell2001, 47; 2007, 93, 316–7). This is an unwelcome finding since the distinction was introduced to allay fears of a global coercive power and refute simplistic criticisms of cosmopolitanism about justice.Footnote ¹⁵ If it is (sufficiently) weak, radical change of contemporary international order may not be necessary and cosmopolitan morality loses its normative and transformative bite.

Moral cosmopolitanism and positive duties: proving too much

Avoiding world-statism is even less likely once moral cosmopolitans include in their considerations ‘positive’ duties of comparative justice, that is, those that correspond with positive material claims or entitlements and require ‘positive institutional integration’ on the global level. Let me start with Brian Barry and his outline of four principles of cosmopolitan justice: (a) Presumption of equality; (b) personal responsibility and compensation; (c) priority of vital interests; and (d) mutual advantage (Barry Reference Barry1998, 146–9).Footnote ¹⁶ According to Barry, requirements of cosmopolitan justice ‘would (…) be best satisfied in a world in which rich people wherever they lived would be taxed for the benefit of the poor whenever they lived’ (Barry Reference Barry1998, 153). Barry imagines that collection of taxes would be assigned to individual states while redistribution/local distribution of this global tax fund would have to be subjected to tight external control, so that most of the raised money does not get seized by the elites-in-power. Should violation of principles of justice on the part of domestic elites occur, there emerges ‘a case for international intervention to displace the government and, if necessary, place the country under international trusteeship until more adequate institutions can be created’ (Barry Reference Barry1998, 160).

Barry suggests several general instruments of tax-based redistribution, such as levying some proportion of the Gross National Product (potentially tweaked by an element of progressivity) or ‘user fees’ on the use of natural resources and producing/discharging environmental pollution. While he acknowledges that setting up and launching such a redistributive scheme represent a major challenge, Barry also claims that no need for specialized ‘tax collecting corps’ would arise. Referring to the example of the EU and partially the World Trade Organization, Barry speculates that the advantages of entering such a scheme of international cooperation and staying in there will be enough to ensure compliance via some kind of enlightened Hobbesian reasoning based on self-interest – although moral motivation has to come into play as well (Barry Reference Barry1998, 156). In this way, he apparently wishes to build on the distinction between MC and PC, to which he explicitly subscribes (Barry Reference Barry1998, 144–5).

However, a glimpse over the extensive and even-expanding administrative apparatus of the modern state yields at least a reason for doubting that administration of the global tax-based redistribution scheme would be possible without the parallel existence of global bureaucracy and global political authority, which would provide both legitimacy for and control of this proposed global ‘authoritative allocation of values’.Footnote ¹⁷ Like Barry, we might point to the EU as a ‘laboratory of cosmopolitanism’: For one, there is no shared tax system in the EU, and it is safe to assume that if it was to be introduced at some time in the future, such a system would require a fairly strong administrative, as well as political backing centralized on the EU level. The recent pressures to establish a fiscal union, being a fallout of the protracted economic difficulties within the EU area, are a case in point. Second, historical sociology teaches that modern constitutional states were made possible through taxation, even if the proximate cause for seeking revenue was armed competition between the then-existing polities (Tilly Reference Tilly1975, 73–6; Møller Reference Møller2007). Although historical and political context of early modern Europe was much different from today's EU, the link between taxation and sovereign political authority that grants certain rights to citizens in exchange for their money is clearly conceptual and not just historical. What is not clear is the basis for thinking that a polity will no longer need ‘tax collecting corps’ for collecting taxes. Third, a post-sovereign future of Europe is no more self-evident and perhaps no more obviously desirable than a sovereign one, if certain valued goals such as security or international relevance of the EU are to be pursued (Morgan Reference Morgan2005: Chs. 6–7; Scheuerman Reference Scheuerman2011, 121–5).

Other major cosmopolitan thinkers take positions analogous to that of Brian Barry. Darrell Moellendorf advocates global egalitarian principles of justice in the form of ‘liberal equality’ (which is a combination of Pareto optimum and fair equality of opportunity) and ‘democratic equality’ (basically a globalized version of Rawls's difference principle) while stating that armed intervention, as an exemplary case of encroachment upon state sovereignty, represents ‘a violation of sovereignty if and only if the intervention will not attempt to advance the cause of justice either in the basic structure of the state or in its international policies’, subject to a handful of conditions (Moellendorf Reference Moellendorf2002, 78–81, 126–7). Moellendorf also rejects the idea of a world-state because of insurmountable practical difficulties (recalling Kant's argument), and also judges insufficiently a mere strengthening of democratic accountability of both the existing and future multilateral agreements and/or regimes. The latter model of decentralized international cooperation covering, among others, active monitoring by non-governmental agencies and pressure exerted by ‘global social movements’, cannot and will not provide ‘a satisfactory solution to the problem of global justice’, above all because obligations following from the principles of global distributive justice ‘cannot be met in the absence of global institutions’ (Moellendorf Reference Moellendorf2002, 172–3).

Moellendorf concludes that solution would be provided by a multi-level system of state, regional, and global institutions subject to mechanisms of democratic accountability. Which specific institutions and on what levels they should be established depends on the concrete issues that are to be dealt with: For example, achieving the goals of global distributive justice, curbing environmental pollution, and collective armed reaction to injustices would most likely fit the functions of global institutions, whereas issues of security and judicial protection of human rights would be best handled on the regional level. However, the actual difference between a ‘world-state’ and a system of state, regional, and global institutions of governance is again not quite clear: The fact that some powers might be delegated from above (the global level) onto regional or state institutions does not mean that they are not ultimately dependent on the will of global institutions, as the embodiment of the ideals of cosmopolitan justice. For example, when Moellendorf writes about ‘mechanisms of democratic accountability’ of regional institutions, who are ‘the people’ who these institutions shall be accountable to – the regional people or the global people? If the former, how are we to ensure that the democratic will of the regional people conforms egalitarian principles of cosmopolitan justice, when Moellendorf states at the same time that global political institutions that would enforce a determinate interpretation of the particular principles or norms would not be available (i.e. would not be necessary)? If the latter, then it is very unclear how a ‘global democratic will’ could possibly be aggregated and implemented without cosmopolitan political and administrative institutions being already in place.

A detailed defence of moral cosmopolitanism as opposed to political cosmopolitanism has been put forward by Kok-Chor Tan. He argues, when discussing the coherence of combining (liberal) nationalism with cosmopolitan justice, that selected goals of MC could be most readily achieved within the borders of national communities: ‘Indeed, as we can recall, many liberals have argued that certain moral cosmopolitan goals are best achieved in the context of the national community, rather than in the context of a global state’ (Tan Reference Tan2004, 95; italics in original). Let us recall that, in his previous book, Tan claimed that global expansion of comprehensive liberalism as an umbrella-type moral and political theory (1) is grounded in the principle of moral autonomy and (2) rejects toleration as one of the fundamental values or principles of liberalism that would provide more adequate tools of coping with issues of justice and human rights, as opposed to Rawlsian political liberalism (Tan Reference Tan2000). The belief that such comprehensively defined goals of MC cannot be achieved without a PC-like institutional framework is premature, according to Tan, because it ignores instruments and measures that do not require a world-state in order to be successfully implemented. In other words, realization of cosmopolitan ideals would in practice rather mean the abovementioned ‘clinging’ to the existing institutional structure, supplemented by a yet undetermined set of institutional norms and institutions.

I have two comments. First, this image of MC, as realized through existing communities, necessarily presupposes that all these communities will have acquired, prospectively, liberal egalitarian character – if only for the ‘right’ form of liberal nationalism to blossom, so that the fundamental liberal principle of individual autonomy is secure from violations. As Tan explicitly states, acts of violation of liberal values in no way acquire a special moral status just because they have been committed outside the territory of liberal states themselves; for this reason, cultural and moral pluralism can be and has to be circumscribed (Tan Reference Tan2000, 82–3). True, fully in line with the distinction between MC and PC, Tan adds that enforcement of comprehensive liberal norms and principles is logically dissociable from and independent on the question of moral rightness and wrongness. However, it seems questionable to let such a very strong moral conception float within just weakly changed structure of both domestic and international institutions; doubts quickly emerge whether statements about the ‘limits of moral and cultural pluralism’ would have any practical–political purchase. The contemporary picture does not correspond with this liberal moral ideal; thus, global prevalence of liberal egalitarian moral values becomes an essential – stable and in the long term sustainable – condition.Footnote ¹⁸

The actual preservation of political autonomy also remains dubious. Imagine, Avery Kolers (Reference Kolers2002) suggests, that cosmopolitan theories of justice turn their attention to the issue of property rights. Autonomous exercise of political decision making will most likely lead to varying outcomes, depending on the social conditions, broadly construed. It is neither illogical nor irrational to envision a community of economic libertarians, which – making extensive use of the principle of subsidiarity – agree on an exclusive and inalienable individual right to private property, unlike their communist neighbours who by definition reject such right. Both, however, will be most likely at odds with the liberal egalitarian ideal of cosmopolitan justice; also to be expected is a sharp increase of material inequality between the two communities (or at least between some strata within them). It is not at all clear how a defender of vertical dispersion of state sovereignty could avoid siding with one of the parties to this dispute (global standards of equality vs. local autonomy): Either she prefers local autonomy and thus negates the universal standards that induced her to criticize the territorially bounded state in the first place, or she clings emphatically to regional or global institutions embodying cosmopolitan ideals, by which, however, she abandons the idea of dispersed sovereignty. One of the higher levels will now be granted the ‘right of last word’ backed necessarily by a coercive apparatus of some kind. In other words, in order for the global community of mankind to be reasonably capable of pursuing common moral goals and uphold cosmopolitan ‘societal order’, a coercive global political authority would have to exist that would keep in check the strong states’ whims of power, or overcome the potential resistance of strong regions, as Barbara Buckinx (Reference Buckinx2011, 270) notes.

Second, when introducing examples of measures that do not require a world-state in order to be successfully implemented, Tan mentions the Tobin Tax, Pogge's Global Resource Tax or the ‘bit tax’ covering digital transmissions on the Internet (also Pogge Reference Pogge1994; Tan Reference Tan2004, 81 fn. 26, 95; Caney Reference Caney2006, 737).Footnote ¹⁹ However, even if we leave aside the worry that speculative capital controls will not work unless there are control and coercive mechanisms of comparable scope, such modest measures conceal that Tan's moral ideal is actually constituted by a global extension and application of Rawls's principles of justice – especially the Difference Principle but also the Principle of Fair Equality of Opportunity (Tan Reference Tan2000, 159–80; compare Caney Reference Caney2001; Tan Reference Tan2004, 55–61; Caney Reference Caney2005, 122–5; Moellendorf Reference Moellendorf2009, 68–89). And even if the former instruments could be implemented and enforced by the existing institutions, it is quite reasonable to expect that institutional changes and transformations matching the fuller requirements would far exceed the moderate official suggestions. Take, for example, the Difference Principle, which cannot be isolated from the principle of just savings (i.e. from intergenerational justice). Rawls is adamant that the rate of just savings depends on the experiences, habits, etc. of each particular community (Rawls Reference Rawls1999, 242–62) – in other words, we cannot expect the rates to converge, unless there is a central authority that sets and enforces them universally.

It is tempting to ask whether this is a case of neglect or a trick of argument: While ‘moral cosmopolitanism’ would cover all the ambitious moral ideals and principles of justice, which have not yet been realised even on the domestic level, explication of desired institutional arrangements would seemingly support the asserted realistic nature of cosmopolitan theory by listing only moderate and relatively uncontroversial (regarding the necessary institutional arrangements) examples.Footnote ²⁰ To put it differently, the admittedly extensive aspirations of strong MC render the proposed ‘middle way’ between statism and PC (=world-statism) inadequate.

Simon Caney has pursued still another way of defending the practical relevance of the MC–PC distinction, offering two argumentative strategies. One stems from a critique of the concept of state sovereignty. Caney distinguishes four definitional criteria of full-fledged sovereignty: Legality, that is, legal (as opposed to power or economic) foundations of political authority; supremacy, or the non-existence of any further institution with overriding authority; territoriality; and comprehensiveness, or having the last word in all affairs concerning the given territorially demarcated community. Caney argues that by variously combining (and discharging) these criteria of state sovereignty, we obtain several alternatives both to a state-centric system and a world-state entrusted with full sovereign powers (Caney Reference Caney2005, 149–52).

Within this alternative framework, Caney employs his political imagination to support the case for ‘global political institutions’ or ‘cosmopolitan political structures’, in essence a version of Held and Archibugi's or Linklater's models of cosmopolitan democracy (Archibugi Reference Archibugi1995; Linklater Reference Linklater1998, 179–220; Held Reference Held2006, 295–359).Footnote ²¹ Democratic political institutions, economic super-organisation for coordinating the workings of the International Monetary Fund (IMF), the World Trade Organisation and the World Bank, global volunteer force (perhaps under the UN auspices) and international courts of justice dealing with not only human rights but also corporate grievances and injustices – this is the institutional network that would, according to Caney, open the way towards achieving global political and economic justice. At the same time, absence of supremacy and comprehensiveness as its definitional criteria should guarantee that the spectre of a ‘sovereign world-state’ is ruled out (Caney Reference Caney2005, 160–4).

Caney supports his case with another argument, tantamount to saying that whatever moral ideal one is strongly committed to, it does not follow that she has to think there is a duty to bring it about – that is, to establish the requisite institutional structure. He calls this a ‘purist instrumental view’ of cosmopolitan justice, and adds in refutation that above a certain minimal threshold, there is deep disagreement among ‘reasonable and reflective persons’ about what socioeconomic justice requires (Caney Reference Caney2006, 730). Above this basic minimum defined by protection of ‘fundamental interests’ of human beings, conditions should be secured for fair mediation among rival ideals of global justice. Caney then gathers eight arguments why a system of ‘international institutions’ (either reformed or brand new) is desirable and necessary, in order to ensure that the mixed conception of cosmopolitan justice will be realized. Finally, he offers eight more specific proposals regarding which virtues these institutions should promote.Footnote ²²

There is first the possible objection that the two strategies are not compatible, since the former includes as it moral bedrock controversial principles of justice that the latter explicitly rejects – namely global equality of opportunity, and the right to equal pay for equal work (Caney Reference Caney2005, 122–3). I will, however, leave it aside and focus on more general issues. As should be clear, the first way of defending of cosmopolitan justice is susceptible to objections brought up on the previous pages – among others, the example of libertarians and communists casting doubt on the possibility of doing without supremacy; and the example of a global Difference Principle and global Principle of Equality of Opportunity on the possibility of doing without a world-state as such.

The second argumentative strategy exemplifies the general concern behind this paper. On the one hand, Caney speaks about ‘international institutions’ or a ‘basic minimum’ of cosmopolitan justice in order to signal his opposition to overambitious institutional proposals. On the other hand, however, he loads his global system of governance with a plethora of sanctioning and enforcing tasks in all the mentioned areas, openly conceding that these institutions will have coercive powers to impose sanctions, levy fines, punish wrongdoers, and so on (Caney Reference Caney2006, 747, 753). It is implausible to assume that we can expect consensus of all decisive agents, be it states, non-governmental organizations (NGOs), IOs, or multinational corporations (MNCs), on almost any of the hot issues, as there will always be ‘ruthlessly instrumental’ or plainly stupid or irresponsible actors – not to mention those who ‘merely’ pursue their national interests (Price Reference Price2008a, 201; Scheuerman Reference Scheuerman2011, 147). Given that so much is at stake for Caney – not least fundamental human interests – and keeping in mind how meagre is the contemporary record of protecting even the barest human rights in many parts of the world, it is puzzling why he would expect the resulting institutional structure to become anything less than an all-powerful, coercive, although perhaps legitimate global political authority.

Finally, if we are to follow Caney's earlier suggestions, the notion of ‘fundamental human interests’ seems to be substantively at par with the full contemporary package of international human rights, and it is difficult to see how this might be considered a ‘basic minimum’ of morality (Caney Reference Caney2005: Chs. 3–4).

This kind of worries may be generalized with reference to Jürgen Habermas's defence of ‘constitutionalization’ of global politics, facilitating on the intermediate (regional) level what he labels ‘global domestic politics’ while retaining an important role for nation states. The argument has analogous structure to Caney's: Although Habermas (unlike Caney) avoids strong moral claims on the global level, he also uses a multi-layered conception of political agency, dispersing political decision making among the supranational, transnational, and national levels (Habermas Reference Habermas2008). However, as Scheuerman repeatedly emphasizes, Habermas ‘appears to want to have the cake and eat it as well’ because even if the reformed United Nations (UNs) would be charged only with maintaining peace and securing human rights and would be refused formal monopoly of legitimate use of force, it would need ‘substantial political and military muscle’ to bring uncooperating states into line (Scheuerman Reference Scheuerman2008, 159; also Scheuerman Reference Scheuerman2009; 2011, 117–9). At this point, my earlier comment regarding the actual ‘strongness’ of weak MC resurfaces: Whether a given conception of a just international order is actually feasible or not depends on its specific demands and requirements, and cannot be simply read off the label. Habermas's idea of constitutionalization of world polity/politics seems inadequate for this reason, for it is couched in terms of universal civil and political rights, and as such presupposes not only that existing polities become constitutional democracies [which Habermas (2002, 202)] himself admits is presently a rather tall order), but most likely also a particular liberal conception of individual moral agency (Shapcott Reference Shapcott2002).

Scheuerman rightly generalizes his critique to all cosmopolitan authors who seem to believe that goals such as (necessarily centralized) control of weapons of mass destruction can be achieved in a decentralized world of dispersed sovereignty and power. The same goes for the ideals of socio-economic justice, which are ordinarily cast in comparative terms. How is it to be provided, without actually extending core elements of statehood onto the global level? How can the power of powerful, unjust, instrumental actors be dispersed, and power of the weak and vulnerable strengthened, unless possessing even greater power? Major powers, be it the United States, Russia, China, or even the EU, do not seem to be eagerly anticipating the opportunity to subject themselves to ambitious cosmopolitan moral ideals. This and the affiliated ‘dilemma of global right’ are issues Kant grappled with in relation to securing individual rights and freedom beyond the constitutional state, and I do not see any reason to regard his core insight as outdated.

Some observations on IOs and the EU

Michael Barnett and Martha Finnemore (Barnett and Finnemore Reference Barnett and Finnemore2004) studied the internal structure and functioning of three major bodies – the IMF, the United Nations High Commissioner for Refugees (UNHCR), and the UN Secretariat, and have come up with conclusions that can be read as supporting my overall argumentation. All IOs need efficient bureaucratic structures, or to be more precise, they themselves are bureaucracies that pursue given political ends. Importantly, for the present purposes, the defining feature of bureaucratic structures is their hierarchical nature, together with continuity, impersonality and expertise. From this stems their authority and legitimacy as actors who are able to identify and deal with problems that may exceed capacities of individual states, and to promote ‘socially valued goals’ globally.

Bureaucracies are thus by nature hierarchical, even if subject to some principle of subsidiarity (such as the one that formally co-defines the EU). Now one of Barnett and Finnemore's principal contentions is that international bureaucracies acquire agency, that is, they become partially autonomous actors independent on the will or interests of their founders (usually states) – the extent of their autonomy of course varying according to the given goal. As agents possessing legitimate authority and capable of defining ‘problems’ and mobilizing resources for their solution, IOs wield power: Creating and enforcing obligatory rules regarding human rights or humanitarian intervention are two prominent examples of this power. The authors emphasize that this is not only the power of constitution (which is an echo of their largely constructivist approach to international relations), but also the power of domination, that is, the ability to force other actors behave in a way they otherwise would not. As they put it, ‘[t]he bureaucratization of world politics means that international organizations have more authority than ever before and therefore have more power over other actors than ever before’ (Barnett and Finnemore Reference Barnett and Finnemore2004, 165).Footnote ²³

My general argument can be thus recast in the following terms: If we set ‘huge, aspirational tasks’ for global political structures, as strong MC unambiguously and consciously does, then these structures stand in need of much greater power than is presently the case with IOs. This is, however, at odds with claims that such structures may operate in some kind of post-sovereign manner, or even that states will retain most of their authority and power. The problem with a government-free cosmopolity lies in the fact that it is not at all clear from where the global bureaucracy will derive its political legitimacy: Both historically and conceptually, bureaucracy in modern democratic states is the backbone of the executive branch of government, which, as the term suggests, executes rules passed by the legislature. The key notion here is political accountability, and many sophisticated constitutional and political procedures have been devised to ensure that a domestic government that loses the parliament's confidence can be peacefully (democratically) replaced or reconstructed. Global bureaucracy, however, cannot lose confidence in this sense because questions of political trust are by definition irrelevant for it. Here another concern emphasized by Barnett and Finnemore becomes important, namely that high aspirations will almost unavoidably lead to organizational pathologies to which bureaucracies universally incline; this much is well known both from social scientific research and day-to-day experience (Barnett and Finnemore Reference Barnett and Finnemore2004, 34ff., 44, 163ff.). Who is to take blame and how if certain grand goals and corresponding strategies (think of reducing poverty) turn out to be bust?

In other words, it is natural to ask what use will be a ‘world legislative body’ and ‘global bureaucracy’ without a world executive body: As Michael Goodhart notes, ‘[p]arliaments without power are no guarantee of democracy’ (Goodhart Reference Goodhart2011, 193). The aversion to the idea of a world government sounds artificial if it is followed by, or if it follows, enthusiastic advocacy of a world legislative body and judicial globalization. It is tempting to conclude that there is no systematic account available of how a political system could work in which government, understood as one of the three branches of constitutionalized political power, has been replaced by whatever is understood under the heading of ‘governance’. This would, however, be premature: Many authors point to the EU as the example of an emerging post-sovereign stateless polity with a cosmopolitan vision (e.g. MacCormick Reference MacCormick1999; Weiler Reference Weiler2003; Beck and Grande Reference Beck and Grande2004; Brunkhorst Reference Brunkhorst2005: Ch. 7; Benhabib Reference Benhabib2006, 170–4; Habermas Reference Habermas2006; Brown Reference Brown2008: 436; Cabrera Reference Cabrera2010, 525; Stone Sweet, Reference Stone Sweet2012).

I see at least four reasons why the EU is most likely not a good example. First, the EU does have a strong and centralized executive body, that is, the Commission, which at the same time has a de facto monopoly over legislative initiative. Second, EU law, as interpreted by the Court of Justice of the EU (CJEU), is based on principles of supremacy and direct effect, meaning that if certain procedural conditions are met, European laws take precedence over member states’ legislation. In theory, the principle of subsidiarity pulls in the opposite direction; however, it has to be noted that (a) it does not apply to areas where the EU has exclusive competence, and (b) may be overridden even in areas of shared competence if ‘the objective of its action could not be achieved sufficiently by the Member States’, pending the decision of the CJEU (CJEU 1997). As is widely known, the Court has acted as perhaps the main vehicle of Europeanization, understood here as expanding the boundaries of EU's political and legal competence, for example, in the important field of human rights protection (Golub Reference Golub2000, 187ff.; Smekal Reference Smekal2009). There are probably good reasons for upholding a clear judicial hierarchy – principles of legal certainty and consistency come to mind – but again, what is the point of establishing hierarchy in one branch of government (or two if legislature is included) and not in another?

Third, if the EU was to actively promote and enforce certain ambitious principles of justice, such as those proposed by strong moral cosmopolitans, the given policy arenas (such as fiscal, social, or immigration policy) would have to be moved to the group of EU's exclusive competence – otherwise it is hard to see how binding yet costly or otherwise demanding decisions pushing for what has been termed ‘positive integration’ could be carried through against resistance; the often discussed open method of coordination has mixed results depending on the policy area, countries in question, and many other variables – besides not being particularly open (Borrás and Radaelli Reference Borrás and Radaelli2010). Even in the EU, it took a ‘long period of time’ among constitutional and democratic states to establish a robust system of human rights protection (Mayerfeld Reference Mayerfeld2011, 237ff.).

Fourth, it is at best controversial to take the EU as a model case of transnational democratic polity. Democratic legitimacy of EU-level decision making has been questioned by many, although it is fair to admit that it has also many outspoken defenders (see e.g. Beetham and Lord Reference Beetham and Lord1998; Scharpf Reference Scharpf1999; Lord and Magnette Reference Lord and Magnette2002; Bellamy and Castiglione Reference Bellamy and Castiglione2003; Moravcsik Reference Moravcsik2001; Eriksen and Fossum Reference Eriksen and Fossum2004, Reference Eriksen and Fossum2007; Follesdal Reference Føllesdal2006; Dufek Reference Dufek2009). Nevertheless, it is plausible to assume that if there are problems with input legitimacy in the EU, they will only be aggravated on the global level if the EU is to represent a model case of cosmopolitan polity. Recently, political theorists have been busy devising novel accounts of the possibilities of transnational political representation, tapping into the resources of both deliberative and cosmopolitan democracy (Held Reference Held1999; Kuper Reference Kuper2004; Dryzek and Niemeyer Reference Dryzek and Niemeyer2008; see Saward Reference Saward2009). However, these are still more or less in their theoretical infancy, and I doubt that a trial and error method is the appropriate approach when stakes are, politically, that high.

Towards a proper middle ground

If my argument is right, then friends of liberal egalitarian cosmopolitan justice face a dilemma: Either bite the bullet, retain strong MC and admit that the institutional solution points to a centralized global political authority, or back off from deep moral demands in the first place and perhaps keep flirting with the idea of a multilayered system of global governance where states remain the principal actors. This dilemma also haunts global democrats such as David Held or Daniele Archibugi who pursue the agenda of global democratic governance while defining the desirable model of democracy in egalitarian (e.g. social democratic) terms; as in the case of strong moral cosmopolitans, the list of tasks to be taken care of by transnational and global institutions is ‘anything but minimal’ (Scheuerman Reference Scheuerman2011, 116). The question of wanting to have the cake and eat it accompanies any attempt to argue that the complex tasks and functions of the modern state can be redistributed to a stateless post-sovereign/multilevel/non-hierarchical polity.

This might seem like a dead end. On the basis of my argumentation in the article, however, I would maintain that it is precisely the overblown moral aspirations of strong MC that lead to such discomfort. There is an ever denser network of IOs, NGOs, commercial, and other private subjects at work on the transnational level, and although their achievements (e.g. the ban on landmines) may seem fragmentary and slow to come when compared with models of cosmopolitan justice and/or democracy there is hardly any ‘sticking to the status quo’. What it comes down to, as Bellamy and Jones argue, is that the existing interstate system of governance should be supplemented, not subverted or supplanted by whatever institutional design comes with international (weak cosmopolitan) morality (Bellamy and Jones 2000, 212). Although states may no longer be the sole sources of legal norms in the international sphere, they are still central political actors – as well as the only actors that can efficiently deploy coercive power, even if under the patronage of IOs. If there is any hope of ‘transnational democratization’, it should mean transnational pressure to democratize individual states. After all, as Barnett and Finnemore (Reference Barnett and Finnemore2004, 164) remark, the occasional democratizing success of today's IOs consists in ‘creat[ing] durable, modern nation-states that are organized around democracy and markets’ (Barnett and Finnemore Reference Barnett and Finnemore2004, 164; also O'Donnell Reference O'Donnell2010, Ch. 10).

Here my position overlaps with a group of weak cosmopolitans who call for ‘bringing the state back into cosmopolitanism’, as Garrett W. Brown nicely put it (Brown Reference Brown2011). Authors such as Brown, Beardsworth (Reference Beardsworth2011), or Shapcott (Reference Shapcott2008) do aim for a self-consciously weak international morality, taking seriously the worry that strong versions of MC leave in dark how to translate strong moral argument about justice to real-world political practice, which is – and for the foreseeable future will be – still dominated by states. I believe that my own position (for which I have no catchy label, unlike Beardsworth's ‘cosmopolitan realism’ or Brown's ‘responsible cosmopolitan states’) is ultimately more restrained in its normative load, not least because some of these authors are reliant on the Poggean causal argument that I criticized above, or slip towards unrealistic institutional schemes (such as Beardsworth's plea for ‘uniform and cosmopolitan’ capital controls, or his pronouncing of Held's model of cosmopolitan democracy as ‘feasible and practical’). But this is a topic for another time: The main and common point is that weak MC is the only reasonable option if one wishes to keep the distinction functional and navigate between statism and world-statism, that is, to avoid the first horn of the dilemma.

It may be replied that it is not clear as to why this (dis)solution should be preferred to biting the bullet in favour of a world-state. In fact, I have been mostly agnostic in this regard, since my main aim has been to highlight a significant incoherence in the existing literature. My argumentation, however, provides at least two fundamental reasons: First, it would return the issue one step back to the Kantian dilemma of global right and the corresponding concerns about despotism and/or paternalism. For liberals of all sorts, this is a big worry: There may be both principled and consequentialist arguments in favour of the desirability of a world-state, but they most likely work with overly optimistic scenarios of the new world order, and perhaps also a naïve version of the domestic analogy (here I would recall my earlier considerations on unintended consequences).

Second, as many weak cosmopolitans have also pointed out, it is unclear where states actually fit in – conceptually and normatively – within the world-statist picture, especially with respect to the fact that it will be mainly the existing states that will have to carry the burden of transforming the world as we know it and, effectively, abolishing themselves as states. Wendt (Reference Wendt2003, Reference Wendt2005, Reference Wendt2011) believes that, in the long run, states as agents have actually good reasons to want a world-state to arise, since it would enhance both collective and individual agency, as well as ensure better recognition from others. This, however, requires accepting his peculiar structural–teleological view of international politics; besides that, it seems to point to a truly long run. Insofar as one understands the point of political theorizing to provide general guidelines for political action, this sounds troubling, and at any rate strong moral cosmopolitans are not the most patient bunch.

It follows from these considerations that the morality to be sanctioned and enforced by the system of international institutions should be decidedly weak or thin. In order to avoid pitfalls that I have attributed to strong MC, this means, for example, selecting a core subset of human rights – such as those put forward (in ascending scope of demands) by Michael Walzer, John Rawls, or David Miller – as a medium-term moral ideal, knowing that strict and consistent enforcement of even the most basic rights is a gargantuan task.Footnote ²⁴ This holds especially if combined with humanitarian duties, which, according to interpretations by weak cosmopolitans themselves, carry pretty demanding obligations – only without the ‘justice’ trademark (Shapcott Reference Shapcott2008). What I consider ‘middle ground’ would be nevertheless rejected as unacceptably minimal, conservative, or excessively thin by strong moral cosmopolitans and probably by some weak moral cosmopolitans as well.

Here a counterfactual scenario about humanitarian intervention comes in handy: Assume, Walzer suggests, that each child in the world has a right to education and non-discrimination and that these rights are violated in a given country (e.g. girls are officially prohibited from attending school classes). Walzer asks, ‘Should a foreign legion of Teachers with Swords forcibly open the schools to boys and girls alike?’ (Walzer Reference Walzer2007, 261). What if the same problem arises with religious liberties? Old age security? Collective bargaining? Democratic voting? What does it mean that people have rights to this or that, when the international community is not able to secure their very right to bare life? To see why this scenario supports the claim that such minimalism is quite different from status quo conservatism and is far from empty, recall that the specific rights mentioned above are instances of universal human rights included in the two International Covenants. Since there are many countries where some or all of these rights are not protected (not to mention what I call core HR such as prohibition of torture and cruel, inhuman or degrading treatment), it seems puzzling why this minimalism should be ever considered empty.

Besides interstate treaties, there are two main roads towards a transnational system of law: one is the hierarchical model of judicial and legislative globalization addressed above, which promises robustness, consistency, and predictability; and the other is legal polycentrism and/or pluralism, which draw on various extra-statal or extra-political political sources of law (such as private or commercial law), at the expense of irregularity and inconsistency. For strong MC, only the first option is available; weaker international morality can come to terms with the latter, assuming that the traditional mode of international lawmaking stays in place – including its achievements, such as the lex ferenda status of human rights. In any case, the move towards transnational system of governance endowed with binding and enforceable rules should proceed in a gradual and evolutionary manner. Research into emergence of transnational norms of governance such as those related to international regimes suggests that while there is space for gradual establishment of social norms that might be considered or conceptualized as law even if their source is extra-statal and their scope reaches beyond the nation-state and is more or less independent of it, their area of application is presently rather limited – mostly to commercial interactions, as in the case of the lex mercatoria, and to some basic human rights – and their stability and legitimacy still very much depends on whether they conform to the actors’ expectations and normative convictions (Calliess and Renner Reference Cabrera2009).

Conclusion: lessons for political philosophy

What is the theoretical and normative substrate of this sketch? First of all, my argument should not be taken as implying that a move towards some kind of transnational political authority backed by a system of law is intrinsically wrong. The promise of constructivism in IR lies exactly in the conviction that some moral and political progress can be made, for example, on the back of a novel understanding of international regimes, or the claim that values, norms, and ideas have primarily a constitutive role with regard to interests, goals, and identities of various actors qua potential agents of cosmopolitan justice (Hurrell Reference Hurrell2007, 70). As construed by a host of contemporary scholars, constructivism in IR and in social sciences in general promises to serve as the theoretical bridge between normative reasoning and empirical research, as well as between normative scepticism and utopianism (e.g. Beardsworth Reference Beardsworth2011, 88ff., 152; Barnett and Finnemore Reference Barnett and Finnemore2004, 6, 18; Caney Reference Caney2006; 738ff.; Price Reference Price2008a, Reference Price2008b; Wendt Reference Wendt1999, Reference Wendt2011).Footnote ²⁵ Even if implicit, the essentially constructivist assumption of the malleability of actors in world politics underpins a large part of weak MC works (including those I have cited).

My worry with constructivism is that, if embraced uncritically, it too easily lends to endorsing strong forms of MC. Although constructivism as an approach to IR studies can be said to be fundamentally agnostic regarding desirable states of the world, it may be readily adopted into thick normative positions (Price Reference Price2008c, 320). Both outside commentators and leading constructivists themselves concur that many, if not all, members of the constructivist camp lean towards ‘progressivist’ international ethics, such as that of liberal cosmopolitanism or Critical theory (Wendt Reference Wendt1999, 87; Erskine Reference Erskine2012, 461–4; Price Reference Price2012, 485; Rengger Reference Rengger2012, 474). Here we are back on a familiar terrain of the strong vs. weak MC, and my point is that the normative self-reflexivity that, for example, Richard Price and his collaborators display (Price Reference Price2008b) may be missing on the part of many cosmopolitan political philosophers: It is tempting to incorporate constructivism within a strong normative perspective, and voilà – we have a theoretical approach that is firmly rooted in empirical IR research, incorporates the idea of moral progress as one of its central notions, delimits the spectrum of alternative scenarios to a feasible set, and carries theoretical space for cosmopolitan moral principles that ought to be realized.

This enthusiasm should be checked both from outside and from inside. As the realist scholar Stephen Krasner notes, real-world disagreements are more often than not still resolved by power and interest, not least because international actors understand the relevant norms differently. These disputes will not go away unless and until there is an authoritative structure for adjudicating them; for this reason, ‘logic of consequences’ should not be underestimated in favour of ‘logic of appropriateness’, for ethically loaded rhetoric international relations is presently rather a sign of ‘organized hypocrisy’ (Krasner Reference Krasner1999, 10–11, 181–2, Reference Krasner2009, 5ff.). This is not to say that hypocrisy cannot be ‘leveraged’ for purposes of progressive change (Risse and Sikkink 1999; Lynch Reference Lynch2008, 169–76), only that we should be careful about what kind of moral demands are thrown at otherwise self-interested, hypocritical yet powerful actors. From within, a balanced assessment is needed, one that is aware of the dilemmas present in both international politics and normative political theory – such as moral regress accompanying moral progress, double-sided developments, or conflict of fundamental moral or political principles (e.g. protection of human rights vs. self-determination as constitutive norms of the international order).

Discovering a proper middle ground between status quo realism and cosmopolitan utopianism presupposes in turn a more careful examination of both the promise and the limits of political philosophy. Here I propose a dual inspiration: On a general level, from authors who emphasize the mutual interdependence of politics and morality, where morality becomes immanent to political practice and not simply a teleological philosophical blueprint to be reproduced in political reality: We should avoid becoming ‘…moralists who already know where we must be going’ (Walker 2003, 285, Price Reference Price2008a, 218; see also Mason Reference Mason2010; cited in Beardsworth Reference Beardsworth2011, 226; Buckinx Reference Buckinx2011, 273). More specifically, there is still much to be learned from the pluralist approach in international relations theory, as exemplified by the pluralist strand of the English school, which attempts to strike a balance between ethical (moral) concerns and the persistent logic of interests and power, or between the realms of justice and order (also Hurrell Reference Hurrell2007, 287–316; Cochran Reference Cochran2009, 204). Normative self-restraint is based here on a combination of epistemological scepticism and priority of the actual (real-world politics) to the ideal (philosophical conceptions of justice); normative content is then provided by Hartian rules of respect for life, property, truth, keeping promises, etc., that is, rules without which no society can possibly function (Cochran Reference Cochran2009, 212). Epistemology typically translates into the issue of normative justification, a grey area of political theory that cosmopolitans about justice prefer to circumvent, not least because the ‘problem of justification’ is said to divert attention from substantive issues to sterile metatheoretical debates.

Nevertheless, normative self-restraint rooted, among others, in epistemological concerns may prove to be an advantage: If there is no thick normative consensus among the key actors of international or global politics, it seems prima facie reasonable to work our way up from some kind of minimal normative foundations – rather than trying to ram demanding moral principles down the throats of global humanity. Substantively weaker moral claims about the nature of a cosmopolitan political community are a more promising departure point when looking for an institutional ‘middle road’ between the spectre of the status quo and the spectre of the world-state.