Doing things by halves: on intermediary global institutional proposals

Doing things by halves: on intermediary global institutional proposals

Luke Ulaş*

Cluster of Excellence ‘The Formation of Normative Orders’, Goethe University, Frankfurt, Germany


Various cosmopolitan theorists offer global institutional prescriptions intended to be understood as residing conceptually between a system of separate domestic states and a federal world state. In this article I assess such ‘intermediary’ models, and claim that they are an unprofitable mix of idealism and misplaced pragmatism: they are ostensibly illustrations of future-oriented institutional ideals, and yet they are infused with concessions to present-day reality. In some cases, the concession is merely rhetorical: we are offered world state visions in intermediary clothing. In other cases, the concession is substantive, with the ironic result that intermediary models are in fact less feasible than the idea of a world state which intermediary theorists quickly reject. The overall aim of the argument is not to fully defend the idea of a world state, but rather only to demonstrate that there are reasons, from a cosmopolitan perspective, to consider a world state superior to intermediary models.

Keywords: cosmopolitanism; global justice; world state; cosmopolitan democracy; dispersed sovereignty

Citation: Ethics & Global Politics, Vol. 9, 2016,

Copyright: ©2016 L. Ulaş. This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International License, allowing third parties to copy and redistribute the material in any medium or format and to remix, transform, and build upon the material for any purpose, even commercially, provided the original work is properly cited and states its license.

Published: 6 June 2016

*Correspondence to: Luke Ulaş, Forschungskolleg Humanwissenschaften, Am Wingertsberg 4, Bad Homburg vor der Höhe, 61348, Germany. Email:


With the perceived deficiencies of a multi-state global order in mind, various cosmopolitan theorists have in recent times offered some variety of what I will call an ‘intermediary’ global institutional proposal. Intermediary proposals are intermediary not because they are an intermediate step toward some other eventual institutional ideal (the intermediary model is reckoned by its author to be the end-state ideal), but because they are to be understood as occupying a conceptual space between a system of multiple states and a federal world state. The intermediary proposal has become very popular, being developed or otherwise endorsed by a significant number of cosmopolitans in both the ‘global democracy’ and ‘global justice’ literatures.1

Using a number of specific examples of intermediary proposals, the case I will make in this paper is that intermediary proposals are a jarring mix of idealism and misplaced pragmatism. To think about the institutional form best placed to realise cosmopolitan ends is unavoidably to engage in an idealistic exercise. In particular, to envisage the existence and success of any such institutional proposal is necessarily to envisage a global population generally prepared to understand themselves as cosmopolitan, and to act upon that understanding. That, of course, is not something that can be said of today’s global population, and so we must understand such institutional proposals as offerings to a possible future global constituency. And yet, so I suggest, intermediary theorists also infuse their proposals with what I understand to be concessions to present day attitudes toward the idea of a world state. In some instances, this concession is only a rhetorical one: despite in fact demonstrating no unambiguous distinction from a world state, the latter descriptor is explicitly rejected. In other cases, the concession is substantive, with the perverse result, I will argue, of offering proposals at once less institutionally ambitious and yet also less plausible than a world state as institutional sites for the realisation of cosmopolitan ends.2 Intermediary proposals do not then obviously function either as the most theoretically promising institutional ideal, or as practical political proposals.

To be clear, the paper does not amount to a fully-fledged defence of a cosmopolitan world state. For one thing, there are a number of critiques that can be raised against a world state that are not addressed here (and which, as I will note in concluding, can in fact also be made of intermediary proposals). The critiques which I concentrate upon here—specifically, the supposed lack of necessity of a global sovereign and the worry about the centralisation of physical power—are those which are relevant to the distinction between a world state and the intermediary model. In addressing those critiques, I suggest only that there are reasons to consider the world state model preferable, from a cosmopolitan perspective, to an intermediary model.

The paper proceeds as follows. In the first section I give some examples of intermediary models before, in the second section, clearly conceptually distinguishing an intermediary model from a multi-state and a world state order. In the third section I argue that some intermediary models are intermediary in name only; they are in fact better understood as world state proposals, despite their authors’ assertions to the contrary. In the fourth section, I make my case against substantively intermediary proposals, demonstrating why they are in fact less feasible than a world state as sites for the realisation of cosmopolitan principles. Across these latter two sections, my overarching claim is that theorists alight (in name or in substance) upon an intermediary model, and reject the world state, because they have one eye on the reception their proposals will receive today. But, to repeat, this is to worry about the wrong constituency. And when we focus instead on a hypothetical future constituency who are widely sympathetic to cosmopolitan principles, it is not evident that the intermediary model is the right one. In concluding, I consider two possible objections to the preceding argument.


For many cosmopolitan theorists, the present global order, made up of a multiplicity of separate states, is fundamentally incapable of dealing with the problems the world now faces in a just, democratic or efficient manner. And yet for the vast majority of these theorists, although the transcendence of the multi-state system is recognised as a cosmopolitan demand, the appropriate institutional solution is not a world state. Instead, something intended to be conceptually distinct from both of these institutional forms is offered.

Consider the following examples. Thomas Pogge outlines an institutional proposal he describes as a ‘vertical dispersal of sovereignty’.3 Pogge says that in our world as it stands, there exists, in nearly every territory of the globe, one government with pre-eminent sovereign authority. Yet ‘from the standpoint of cosmopolitan morality … this concentration of sovereignty at one level is no longer defensible’. Resultantly, Pogge proposes that

governmental authority—or sovereignty—be widely dispersed in the vertical dimension. What we need is both centralisation and decentralisation—a kind of second-order decentralisation away from the now dominant level of the state. Thus, persons should be citizens of, and govern themselves through, a number of political units of various sizes, without any one political unit being dominant and thus occupying the traditional role of state. And their political allegiance and loyalties should be widely dispersed over these units: neighbourhood, town, county, province, state, region and world at large.4

Within Pogge’s institutional scheme, ‘state’ ceases to be a relevant concept: today’s states will apparently no longer be such given the extent of their sovereignty that will be ‘dispersed’; and no other new state forms—at the global level or anywhere else—are to be created as a result of the institutional change.

Pogge’s proposal shares similarities with the idea of ‘cosmopolitan democracy’, the most notable proponents of which being Daniele Archibugi and David Held.5 These theorists, like Pogge, are concerned that the domestic state is an outmoded institutional construct that is unable to respond sufficiently to the challenges brought by the cross-border processes of ‘globalisation’.6 In particular, advocates of cosmopolitan democracy are concerned with establishing the conditions for political autonomy in a world in which various issues that affect individuals are transnational or global in nature: to that end they envisage a number of different levels of democratic governance from the local to the global (as well as ad hoc cross-border referenda). These different levels of governance are purportedly ‘not bound so much to a hierarchical relationship, as much as to a set of functional relations’7; they are ‘mutually autonomous but complementary’,8 and are to ‘act within their own sphere of competence’.9 While Pogge makes no reference to the direct involvement of state governments in the political processes of ‘higher’ level political units, the cosmopolitan democrats declare that both individuals and state governments are to have their own representatives at the global level. This could be achieved by adding a second chamber to the United Nations General Assembly consisting of representatives elected directly by individuals.10

Whereas Pogge and the cosmopolitan democrats envision numerous and somewhat indeterminate political ‘levels’ within their institutional schemes, Jürgen Habermas’ institutional proposal is split specifically across three levels, which he calls the supranational, the transnational and the national. The supranational level consists of a suitably reformed United Nations charged with the ‘vital but clearly circumscribed tasks of securing peace and promoting human rights’.11 Habermas’ position with respect to the idea of direct individual representation at the global level has wavered over time, but he has recently come (back) around to the addition of a world parliament second chamber.12 At the transnational level, most of the issues that transcend domestic state boundaries are to be dealt with and resolved. The agents of the transnational level are to be sizeable ‘regional or continental regimes’13—these might be large existing states, like the US and China, or suitably reformed regional non-state institutions like the EU. Where such regimes are missing, they must be created via domestic state integration. These continental regimes together comprise a ‘transnational negotiation system’ in which the content of ‘global domestic politics’—‘the environment and climate change … [regulation of] financial market-driven capitalism, and especially the distributional problems that arise in the trade, labour, health and transportation regimes of a highly stratified world society’14—is to be agreed upon ‘within the framework of permanent conferences and negotiating forums’.15 The national level is constituted by the states that populate the United Nations today.

Intermediary theorists explicitly and repeatedly reject the notion of a world state. Habermas calls his construction ‘global domestic politics without a world government’ and tells us that ‘the democratic federal state writ large—the global state of nations or world republic—is the wrong model’.16 Pogge’s proposal is ‘not the idea of a centralised world state, which is really a variant of the pre-eminent state idea’.17 The cosmopolitan democracy model ‘is not to be identified with the project of a global government’18; a ‘single, unified international state structure ought not to be regarded as an aim’.19 Being so definitive in their rejection of the world state, one might expect to be able to easily locate the definitive theoretical bases of that rejection—and yet, the reasons behind the evident dislike of the world state are rarely articulated at any length. Held’s appeal to authority, in which he says that the world state ‘is impractical and undesirable for many of the reasons Kant gave’ is typical.20 The reader is apparently assumed to already have a firmly held belief that the world state is bad, meaning that the intermediary theorist can pass fairly quickly over it. This assumption, so I will argue below, leads the intermediary theorist down the wrong path.


The intermediary model is an institutional form ‘intermediate between the confederal model and the federalist one’.21 But what does it mean, conceptually, for an institutional model to be intermediate between a confederal vision (i.e. a world still comprising separate sovereign states) and a federal world state? Archibugi, in particular, highlights a large number of ways in which his version of an intermediary model can be distinguished from confederal and federal alternatives,22 but the conceptual distinction can be most clearly made by reference to only three necessary and collectively sufficient features of statehood, which I take to be: sovereignty; centralised physical power; and citizenship.23 After briefly defining these terms, we can illustrate how differently configuring these elements on a global scale delivers a multi-state, world state or intermediary model respectively.


The idea of sovereignty can be broken down into two conceptual parts: internal and external. For an agent to hold internal sovereignty is to enjoy an absence of challenge to one’s authority from other sources within a given territory (so the idea of sovereignty contains within it the idea of a defined territory). To hold external sovereignty is for one’s authority to be similarly independent of any authority residing outside of that territory. It is not uncommon for it simply to be claimed that ‘the state is sovereign’, but such a claim is not illuminating, at least as far as internal sovereignty is concerned. Indeed, such a statement erodes the conceptual distinction between internal and external sovereignty. If ‘the state’—that is, the composite of the body politic and its political system—is internally sovereign, then it can be sovereign over nothing other than itself. But if this means anything it is simply an expression of the state’s independence and collective self-autonomy, and is then indistinguishable from the notion of external sovereignty.

To be meaningful, internal sovereignty must be located somewhere ‘within’ the state; with, say, ‘the people’, or ‘the government’, or ‘the constitution’. But regardless of where specifically we locate the state’s internal sovereignty, it surely is the case, contra Hobbes, that sovereign authority can in practice be split between a number of ‘offices’ without disastrous consequences. Indeed, it is possible, so history has taught us, not only to constitutionally split authority ‘horizontally’ between different branches of government, but also ‘vertically’, as in a federal state: a federal-level government does not have supreme authority over all decisions within its territory, any more than a legislature, executive or judiciary alone do. But just as it is possible to amalgamate the branches of government and simply say that ‘the government’ is sovereign, so we can amalgamate federal-level and sub-unit governments and say that the federal state’s entire political system (or constitution, or people) is sovereign.24

However, not all dispersals of authority can be amalgamated in this way—or to put it another way, not all dispersals of authority are dispersals of sovereignty. The European Union features a vertical dispersal of political authority: the various EU treaties (which, we might say, together form the basis of a European ‘constitution’) have established that certain competences are to be held at the European level, while other are retained by the member states. But nevertheless, there is no amalgamated sovereign here, because the European ‘constitution’ is not of the right kind: each member state is reserved the right to unilaterally extract itself from the European Union—and because of this, each ultimately remains (externally) sovereign. This is in stark contrast to the member units in a federal state: either they will have no formally recognised right to leave the federation or their right to leave will be contingent upon federal-level acquiescence.25

Centralised physical power

In one notorious account Max Weber defines the state as ‘a human community that (successfully) claims the monopoly of legitimate use of physical force within a given territory’.26 While this feature is not in fact sufficient in itself to denote statehood, it is widely recognised to be necessary. Of course, a government cannot always enforce adherence to laws, and it cannot always locate and punish transgressors. But it is able to do these things more often than not: indeed, where it fails to reliably enforce law and order, we speak of a ‘failed state’. Similarly, although the state might claim monopoly control of the legitimate use of physical force, it clearly does not have a genuine monopoly on the use of physical force simpliciter: criminal groups use force, and domestic violence often evades control by the state.27 Nevertheless, it is clear that the political system of any properly functioning state has a preponderance of physical power in comparison to citizens and sub-state groups—to put it otherwise, power is a relational good.

In a federal state, it is specifically the federal-level institutions that possess physical power over both citizens and sub-state units. While the sub-units will likely retain their own police force and perhaps a modest ‘defence force’, military forces are organised on a state-wide level and are directed by the federal-level institutions.


States have citizens. I will here understand citizenship to comprise two aspects. First, citizenship involves an unmediated and wide-ranging legal relationship between individual persons and institutions of government. Those institutions have ‘personal jurisdiction’ with regards to a range of rights and duties (e.g. they may levy taxes, offer benefits, subject individuals to a range of laws, impose sanctions, incarcerate, recognise individuals as citizens, potentially enforce army conscription, and so on). In a democratic state, citizens also have a range of political rights, and are represented in the legislature. This is the case in unitary states, but also in federal states, where citizens directly elect representatives to the federal parliament, even if sub-units themselves are also represented. The institutions of a confederation (such as, for example, NATO), by contrast, have no such unmediated legal relationship with individuals; if there is any relationship at all, this is entirely mediated via individuals’ separate state governments. However, while states have citizens, not all institutional constructs that have citizens are states. The EU is not a federal state, but the citizens of its member states are also citizens of the European Union, who enjoy a direct legal relationship with at least some of its institutions.

The second aspect of citizenship I want to highlight here is an affective aspect: it is important that, on the whole, a state’s citizens do indeed self-identity as citizens of that state, and that there pertains some level of ‘solidarity’ or fellow-feeling that holds between co-citizens. Without this dual-affective aspect, the state’s coercive practices are unlikely to be understood as legitimate by those individuals, and hence unlikely to be stable. So while it might not be strictly conceptually necessary that a state’s citizens display these affective characteristics (we can at least conceive of a state government operating without them), it is widely recognised to be an important element of any plausible and stable state, and certainly of a normatively defensible one.

Indeed, various theorists have stressed the relevance of shared solidarity to the plausibility of sustainable systems of distributive justice and democracy of the sort that cosmopolitans are variously concerned to extend globally. Quite what must ground shared solidarity is a matter of debate. For liberal nationalists, it is fundamental to the smooth functioning of any liberal democratic state that that state be populated by a self-understood ‘national community’, which delivers the necessary ‘we feeling’ and sense of shared trust.28 On the other hand, it has been argued that the examples of states like Switzerland (in which, it is said, no singular cultural nation exists) illustrate that a sense of citizen solidarity may come instead from a shared patriotism,29 or indeed a ‘constitutional patriotism’ in which shared solidarity arises from a shared historical commitment to and struggle for certain political principles rather than a shared national culture.30 Such a conception of political solidarity is understood by Habermas, in particular, to have the potential to ground social democratic practices in transnational, non-statist contexts like the European Union, and indeed the affective aspect of citizenship becomes increasingly important in contexts like the EU the more the latter is expected to perform functions equivalent to the state. I do not need to intervene here into the debate surrounding the necessary or sufficient grounding of solidarity, but the broader matter of citizen solidarity forms a central part of my critique of the intermediary model.

Having very briefly set out these characteristics of statehood, we can clearly (if schematically) conceptually define three global institutional forms:

An institutional design is a multi-state proposal if: domestic states retain their external sovereignty; they retain control of the use of coercive power and are not subject to the greater, legitimate physical force of a ‘higher level’ entity; and no new transnational or global forms of citizenship are postulated. In other words, all three necessary and collectively sufficient conditions for statehood remain solely at domestic state level.

An institutional design is a (federal) world state proposal if: domestic states are stripped of their sovereignty, with sovereignty being globalised; physical power is globally centralised; and we can speak of a global-level polity comprising ‘world citizens’. In other words, all three necessary and collectively sufficient conditions for statehood are transferred to (or replicated at) the global level.

An institutional design is an ‘intermediary proposal’ if: a globally extensive polity is postulated within which individuals are incorporated as citizens but domestic states ultimately retain their full sovereignty and/or global level institutions are denied predominant use of physical force.31 In other words, some, but not all, necessary and collectively sufficient conditions for statehood are transferred from the domestic to the global level, or are replicated at the global level.

As will be clear, the European Union is a real-world example of an intermediary institutional construct at the regional level. Via the mechanism of a directly elected European Parliament and the establishment of European citizenship conferring upon individuals a range of legal rights, it has constructed a tentative European citizenry (tentative both in terms of the extent of personal jurisdiction and extant European solidarity). And yet, as we have already made clear, the EU is not any type of state: its constituent members reserve the right to withdraw from the Union unilaterally, and retain their own militaries.32


Having clearly articulated this schema, we can use it to evidence something interesting: some supposedly intermediary theorists are in fact better characterised as world statists in denial, because their prescriptions transfer all three elements of statehood to the global level. Consider first, sovereignty. In their designs, both Pogge and Held reference the idea of sovereignty: Pogge describes his model as a ‘vertical dispersal of sovereignty’, and Held refers to ‘cosmopolitan sovereignty’. In both cases, sovereignty is to be in some sense globalised, apparently without it being the case that a global sovereign arises. It’s not clear, however, that conceptual sense can be made of this. Pogge pre-empts two objections to his idea of a ‘dispersal of sovereignty’. The first is the Hobbesian concern that unless there is a final authority that sits above all others in a territory, jurisdictional disagreements cannot be settled and peace cannot be ensured. The second is the claim that ‘there are certain vertically indivisible governmental functions that constitute the core of sovereignty’.33 In response, Pogge points out, as we did earlier, that recent history demonstrates that both ‘horizontal’ and ‘vertical’ dispersals of authority can occur without being destabilising.34 Similarly, all manner of government functions that one can imagine can ‘be handled at various levels and indeed are so handled in existing federal regimes and confederations’.35

The problem for Pogge, however, is that pointing out these truths in no way helps establish the conceptual coherence of his institutional model. On Pogge’s own definition of sovereignty, the latter denotes an agent’s ‘unsupervised and irrevocable authority’ over a set of persons with respect to given issue.36 The dilemma is then this: on the one hand, Pogge’s above appeal to confederations is inappropriate because confederations are not dispersals of sovereignty at all—rather, they are examples of delegated, revocable authority entirely consistent with a world of separate sovereign states. On the other hand, although federations are indeed in one sense examples of dispersed sovereignty, they are also, as we made clear above, at the same time conceivable as a coherent, sovereign whole, i.e. as federal states. Thus, neither appeal to confederations nor federations helps demonstrate the possibility of ‘dispersed sovereignty’ without a state. Indeed, the very idea seems conceptually puzzling. How is the ‘unsupervised and irrevocable’ authority of any one unit in Pogge’s model to be established? Pogge doesn’t go into it, but for such authority to be irrevocable it must surely be secured by some form of constitutional enshrinement; such must therefore be implied (even if not explicitly stated) in Pogge’s model. But if this is the case, it’s not clear why, as well as conceiving of that constitution as the guarantor of ‘dispersed sovereignty’, it should not also be appropriate to speak of that constitution in its entirety as a global sovereign. Why should we not aggregate the different units and conceive of a sovereign global whole?37

Held says that ‘[c]osmopolitan law demands the subordination of regional, national, and local “sovereignties” to an overarching legal framework’.38 The basis of this authority is to be the original consent of those who will thereafter be subordinated.39 What Held is then explicitly offering is a blueprint for a global constitution that will function as the fundamental law of the globe to which all other lower-level laws are subject. Furthermore, Held stipulates that in his proposal, following states’ initial election to join his institutional proposal, their continued commitment to the global political system becomes ‘non-voluntary’ subject to the proper working of that system.40 Held is therefore more or less explicit that states are to lose their external sovereignty (note his use of inverted commas (here formatted as speech marks) in the quotation cited at the beginning of this paragraph), and, as with Pogge, there seems no reason why we should not conceptualise his global institutional model as a fully integrated sovereign political system.41

That Held and Pogge’s models might both be understood, in fact if not in rhetoric, as examples of globalised sovereignty would not threaten those models’ intermediary status if they nevertheless refrained from prescribing centralised physical power. But it is not clear that they do this either. In response to the concern that the legal pronouncements of global institutions will be toothless without coercive power—or, as Hobbes has it, that ‘covenants without the sword are but words’ (a phrase which Held repeatedly refers to directly42)—Held floats the idea that

a proportion of a nation-state’s military (perhaps a growing proportion over time) could be ‘seconded’ to the new international authorities and placed at their disposal on a routine basis. Or, better still, these authorities could increase enforcement capabilities by creating a permanent independent force recruited directly from among individuals who volunteer from all countries.43

Ultimately, Held wishes to see the transference of coercive authority up to the global level alongside the demilitarisation of states.44 This amounts to the global centralisation of power.

Pogge states that ‘non-proliferation and gradual abolition of weapons of mass destruction presuppose a substantial centralisation of authority and power at the global level’.45 This is ambiguous, in two respects. First, it is not entirely clear whether by ‘power’ Pogge means to refer to physical power, although it seems reasonable to infer as much given that a potential synonym, ‘authority’, is itself used in the sentence. Second, what Pogge means by ‘substantial’, and in particular how the extent of global-level power will compare relative to lower-level units, is not specified. But as we pointed out previously, physical power is in any case a relational good—one only has it to the extent that one has more of it than others—and so for there to be a substantial centralisation of power at the global level must mean that we can speak of the global-level system as physically dominant.46 I consider it reasonable to conclude, therefore, that both Held and Pogge also globally centralise physical power in their models.

It is interesting to reflect on why it might be that they are nevertheless keen to cloak world state proposals in intermediary clothing. My suggestion is that this occurs because they—and other theorists who have replicated or endorsed their models—are engaged in something of a methodological equivocation: on the one hand, they are in substance quite clearly engaged in an enterprise—the construction of an institutional ideal designed to realise the tenets of a highly ambitious moral theory that is not currently accepted by the global population—that has little short- to medium-term political import. On the other hand, one gets the sense that they are not fully content to understand their proposals in such terms, and so make a rhetorical concession to current political reality—i.e. explicit rejection of the ‘world state’ bogeyman—designed to make their proposals more politically palatable. It is certainly plausible that such theorists may have one eye of the present-day political landscape; Pogge has been engaged in various initiatives that aim to make a direct practical contribution to the alleviation of global poverty, and in one book Held remained noncommittal regarding whether or not there should be a global parliament populated by representatives of individuals, possibly because that book explicitly sought the attention of global practitioners.47

To be worried about how today’s population might react to one’s ideal institutional prescription is, however, to be worried about the wrong constituency. To reiterate, the question these theorists are answering, that of the most propitious institutional site for the realisation of cosmopolitan ends, is a question of ‘ideal theory’, in two senses.48 First, the matter at hand is the long-term, ‘end-state’ prescription for global institutions, rather than ‘transitional’, practical, short-term reform.49 Second, in addressing this question we are compelled to make the clearly counterfactual and idealising assumption that a significant reserve of ‘cosmopolitan solidarity’—the affective element of ‘world citizenship’—is extant in the world, by which I mean that the global population recognises itself as a global political community, is correspondingly motivated to approach global politics in a democratic, cosmopolitan spirit, and is amenable to at least the most modest tenets of cosmopolitan distributive justice. One might characterise this, in a Rawlsian vein, as an assumption of cosmopolitan ‘compliance’. What is of relevance when considering the end-state institutional model, therefore, is whether it is appropriate for a potential future population, a population among whom a sense of cosmopolitan solidarity does indeed pertain. And, as I will now argue, it is not at all obvious that the intermediary model is the most appropriate for such a population.


Substantively intermediary theorists also bring present-day, practical considerations to bear in the development of their prescriptions. Ironically, however, the result is to render those prescriptions more psychologically ambitious—and hence less politically feasible—as sites for the realisation of cosmopolitan ends than even a world state.

An intermediary design, recall, declines to globalise sovereignty and/or centralise physical power. An intermediary model that declines to globalise sovereignty can do so by retaining states’ unilateral right of secession (this is what Archibugi does50), and/or by investing states with a unilateral veto at the global level (this is what Habermas does51). But what is the case for doing so? It is difficult to find explanation in Archibugi’s case. He has stated that ‘it is preferable not to proceed beyond a certain degree of centralization of power, and in particular, of means of coercion, on such a large scale as that of the entire planet’.52 Even if that were true, however (and I will turn to that shortly), it doesn’t obviously function as an argument against global sovereignty, since conceptually speaking one could have the latter without centralising physical power. For his part, Habermas conceives of domestic states as historic achievements that deserve to be represented at a global constitutional convention, and thereafter to continue to operate as collective actors at the global level. But similarly, neither of these caveats should obviously preclude global sovereignty—all that is required is that (former) states, as well as individuals, are represented in a global legislature.53

An alternative reason that comes readily to mind for declining to globalise sovereignty is that it enables states to avoid being bound by laws they do not wish to be bound by; and the content that comes to mind for a type of law that a state wouldn’t want to be subject to is a law that would go against its own self-interest. But this hardly stands as a justification, from a cosmopolitan perspective, for states to retain their sovereignty, since it is a concession to the straightforwardly strategic state behaviour characteristic of much of today’s international politics that the intermediary theorists seek to transcend.

Nevertheless, I suspect that the rejection of global sovereignty is indeed born of this kind of concession to current global reality: as with the theorists above who rejected the world state descriptor even though it seems entirely appropriate, I suggest that substantively intermediary theorists have one eye on the reception their proposals might receive from today’s global population, as if there really might be a global constitutional assembly sometime soon at which the proposals will be under consideration. Perhaps it is this that also explains intermediary theorists’ tendency toward ‘premature specificity’, in which specific details regarding, for example, the allocation of seats in a global legislature, are set out.54 Such detail only seems to be required where one understands oneself to be engaged in some form of short-term political intervention to which global practitioners may pay attention.55 Of course, if one does partly understand their prescription in such terms, it will seem necessary to retain states’ right of veto and/or secession—today’s states are hardly likely to look favourably on proposals to give up their sovereignty, after all.

However, again, to conceive of intermediary institutional prescriptions in this manner is a methodological mistake. Retaining a right of veto/secession is not sufficient to render intermediary proposals politically plausible—they remain infeasible if understood as proposals for present-day, or even medium-term, reform. And in fact, the ironic effect of refraining from globalising sovereignty is to render such intermediary models more infeasible as a propitious site for the realisation of cosmopolitan ends than a world state, since to reject global sovereignty is to rely upon a standard of cosmopolitan solidarity stronger than that which would be required if sovereignty were globalised. That this is the case, I submit, undermines the criticism sometimes made against the idea of global sovereignty that it is unnecessary to the realisation of cosmopolitan ends.

In forgoing the bolster of sovereignty, intermediary theorists apparently anticipate a form of global solidarity that is itself sufficient to underpin the realisation of cosmopolitan ends. But note that such high aspirations are rarely held for political solidarity in the domestic context; sovereign institutions are understood to remain necessary. Consider, for example, the provision of public goods, a classic ‘collective action problem’. Nationalist theorists, for instance, do not suppose that the fact of national solidarity would enable the stable provision of such goods absent coercive institutions. Even if we generously assume that parties will always themselves be initially disinclined to ‘free ride’ (which, as I note in the next paragraph, is itself highly questionable), it is an extra step to imbue each party with the secure belief that all of the other parties are similarly disinclined. In order for collective action problems to be reliably solved without the aid of institutional coercion, a deeper level of relationship, of friendship, seems to be required.56 A sense of political solidarity is not qualitatively equivalent; it is not a reiterative interpersonal familiarity with, trust of, and commitment to each of one’s co-nationals. Without such interpersonal relationships, sovereign coercive authority remains central to the provision of public goods.

Similarly, in the domestic context, it is not the case that theorists understand social solidarity to mean that individuals are purged of all self-serving behaviour, and of weakness of will, and hence that, for example, compulsory and coercible tax contributions are an unnecessary supposition of sustainable welfare practices. I take it as plain that any such claim would be incredible. As Brian Barry has remarked, ‘if contributions to the government coffers had to be raised by voluntary subscription, even tax rates of 10% would no doubt be regarded as quite visionary and utopian, in the same way as they are in the international context’.57 This is surely true of any presently or historically existing complex political community, no matter the extent of political solidarity that pertains between the individuals within it.

Earlier I described the assumption of cosmopolitan solidarity to be similar to the Rawlsian assumption of compliance that underpins his understanding of ‘ideal theory’—but of course, for Rawls, the context of a sovereign state was presumed, and moreover he is explicit that an assumption of compliance does not mean that sovereign state institutions are unnecessary:

The sense of justice leads us to promote just schemes and to do our share in them when we believe that others, or sufficiently many of them, will do theirs. But in normal circumstances a reasonable assurance in this regard can only be given if there is a binding rule effectively enforced … The need for the enforcement of rules by the state will still exist even when everyone is moved by the same sense of justice.58

Political solidarity, then, can be understood as a necessary attitudinal ingredient that potentially allows state sovereignty to be understood as legitimate by those who are subject to it, and that therefore potentially enables the institutions of government to sustain redistributive welfare schemes. Solidarity is no replacement for such authority.

One who nevertheless sought to reject the need for global sovereignty in the delivery of cosmopolitan ends might put pressure on the analogy I have been drawing between the domestic and global contexts, highlighting the difference between states and individuals. After all, state parliaments—unlike individuals—can pass laws to legally bind themselves, and so if states pass laws binding themselves to actions designed to further cosmopolitan ends, it won’t be as easy to renege on those actions as it is for an individual in the state of nature to do so. But this assumes that such other-regarding laws will in the first place be passed, which I take to be a challenging assumption. Consider that the economically advanced members of the UN General Assembly first pledged to give 0.7% of yearly GDP in international aid in 1970, which we may charitably consider an expression of (extremely modest) international solidarity: very few countries have ever met this target, and only one, the UK, has (very recently) enshrined the commitment in law.59

This may be on account of the following: in the democratic domestic context, the worse-off (along with everyone else) have the institutional opportunity to lobby for their own interests and to vote for those interests in elections, meaning that, for instance, somewhat redistributive policies are often politically plausible. However, in the global context the self-interest of those same individuals pulls only one way: toward the aggregate self-interest of that particular state. There is no equivalent opposing self-interested pull, which in this case would be the self-interest of the globally worse-off, because those persons are not represented in the relevant domestic legislature.60 Other-regarding self-binding law is thus particularly difficult for states to enact.

Another way in which the international ‘state of nature’ might be thought disanalogous with the individual variant is that in the international case some form of legal condition has already been realised between states. As Habermas puts it, states, as political–legal standing achievements, have been able to realise a legal condition between themselves ‘horizontally’ in a way that individuals could not. What’s needed in the international case is to bolster international law with power, whereas in the domestic case, power came first and needed curtailing by law.61 But it does not follow from this disanalogy that global sovereignty is not needed: as Habermas himself says, what’s missing in international law is ‘supranational power above competing states that would equip the international community with the executive and sanctioning powers required to implement and enforce its rules and decisions’.62 One can’t avoid the suspicion that such a supranational power would need to be sovereign in order to be effective by pointing out the disanalogy above.

A recognition that states and persons will inevitably often be tempted to act in their own interests, even in conditions of cosmopolitan solidarity, also throws the advisability of states retaining their own militaries into question. Archibugi tells us straightforwardly that domestic states are to ‘retain their own armed forces’.63 Habermas is similarly clear that there will be no centralisation of power, suggesting simply that if the global-level polity needs to employ force it ‘would draw upon the sanctioning capacities “lent” to it by the able and willing members’.64 This, however, seemingly offers the opportunity, at least for some more powerful states, to retain a de facto right of veto even where, on cosmopolitan grounds, such a right should be (or has been) denied de jure.

Is there then a positive case, from a cosmopolitan perspective, for separate militaries, as endorsed by intermediary models? That case cannot, of course, be the enabling of acts aggression, or the provision a de facto veto. Instead, the justification must be defence of a state’s own citizens and/or, in a cosmopolitan vein, of world citizens generally via acts of humanitarian intervention. But, with regard to the latter, a centralised, independent force will conceivably be far more effective and financially efficient than a series of separate forces. Habermas writes that, ‘given the decentralised monopolies on the use of violence enjoyed by individual states, the [global-level] executive must be reinforced to the point where it can guarantee the effective implementation of resolutions of the Security Council’ (my emphasis).65 It is not clear what this means—how could the executive be reinforced such that implementation is guaranteed without centralised force? The idea seems to be that the Security Council should be better financed by, and given greater access to the troops of, states, while at the same time acting ‘independently of national interests in its choice of agenda and its resolutions’.66 But Habermas’ stipulation that means of coercive force should remain with domestic states sits uneasily, to say the least, with his assertion that a reformed UN should be capable of ‘effective, and above all non-selective’ fulfilment of that task, since, for example, human rights violations occurring in powerful states with large militaries are clearly going to represent less palatable sites for humanitarian interventions than small, weaker states.67

When it comes to the defence of a state’s own citizens, it’s plain enough that a state does not need a military to protect itself against other states if those other states do not themselves have militaries. It’s then not clear what the positive case for continuing separate militaries in a cosmopolitan institutional model might be. One suspects, indeed, that there is no positive case to be made, only a negative case against centralising power, namely the prospect, voiced by Kant and repeated ever since, of its leading to global tyranny. However, the force of this fear is, I suggest, largely parasitic on the notion that centralised power would come about, as Kant imagined, via the imperial conquest by one state of all others, and would resultantly be aggregative (i.e. centralising the military power that exists in the world today) rather than subtractive (i.e. substantially reducing the extent of existing military power) in nature.68 But this is not the only (nor, today, given the existence of nuclear weapons, even the most likely) route to the centralisation of power. It should also be evident that to conceive of such a scenario is to assume a level of international antagonism and enmity that would equally undermine any intermediary model’s cosmopolitan prospects.

Assuming instead, as one must, a widespread sense of cosmopolitan solidarity, centralisation of power could potentially be carefully negotiated and circumscribed, and could, indeed, largely consist of demilitarisation and disarmament rather than an overwhelming build-up of force. Mutual demilitarisation gives concrete expression to a commitment to solidaristic global politics; the global centralisation of power would represent the practical culmination of that demilitarisation. By contrast, continued desire on the part of states to maintain separate militaries appears expressive of a level of distrust not easily reconcilable with an assumption of cosmopolitan solidarity, and indeed contains a logic liable to foster such distrust, for it is surely difficult for any one state to understand itself as genuinely engaging in collaborative global politics with other states when those other states have their guns trained on it. There is then a case to be made that the continued existence of separate militaries is a more dangerous, more expensive, and less efficient state of affairs than centralised power could be.

To understand the centralisation of power in aggregative terms is to fail to pay due heed to the idealising assumption of cosmopolitan solidarity that must be made when discussing cosmopolitan institutional alternatives. That the centralisation of power is resisted by intermediary theorists in their institutional ideals could be because this has been overlooked, or could be, as I suspect, because those theorists recognise that the notion of the centralisation of power will as a matter of fact today be understood in aggregative terms, and resultantly feared, by the public. But if the latter is part of the reason for rejecting centralised power, there is again a methodological error in play—what is relevant is not how today’s public react, but rather how a hypothetical future global public, a public with the requisite cosmopolitan solidarity, might react. And there is reason for such a public to be more amenable to the idea.


My argument in this paper has been that the intermediary institutional model popular among cosmopolitans is an unprofitable mix of the idealising and the pragmatic. On the one hand, to engage in thinking about the appropriate institutional design to best realise cosmopolitan ends is an unavoidably idealistic enterprise: it requires the ambitious, counterfactual assumption of widespread cosmopolitan solidarity or ‘compliance’. At the same time, however, intermediary theorists appear to make concessions to present-day reality. Sometimes, this concession is ultimately only rhetorical: we are offered world state proposals, and yet, perhaps out of concern for the way such a concept is typically received today, the world state descriptor is explicitly rejected. In other instances, the rejection of the world state is more substantive, and yet the motivation underlying that rejection seems to be the same: an implicit concern for how the world state would be received by, and what it might mean for, today’s global population. In both cases the concessions are inappropriate. In the first case the genuine content of the proposals is rhetorically obscured without rendering the proposals any more politically feasible. In the second case the perverse result is to render intermediary models less politically feasible than a world state, on account, in particular, of the more excessive demands placed upon a sense of cosmopolitan solidarity.

I should conclude with consideration of two potential critiques of my argument. The first is that I am drawing an overly sharp distinction between ideal theory designed for some future constituency and theory designed to speak to today’s population: doesn’t my argument ignore the transformative potential of theory itself? Can’t the production of ideals be part of the process that stimulates change in the direction the theory advises? Setting aside the concern that the ideal itself is not the right one, I have two responses. The first is that I am sceptical that political theory, ideal or otherwise, often has much transformative effect, not least because it is in the first place rarely read widely enough to have the chance. Second, even if ideal theory can play this transformative role, to offer cosmopolitan institutional ideals before the global population has internalised the cosmopolitan moral ideal itself, would seem to be a strange strategy: if people have not yet been convinced by direct moral arguments, it’s unclear why they should be receptive to institutional visions for the realisation of those moral principles. If cosmopolitan theory has a transformative intent, ‘non-ideal’ approaches, proceeding from where we are now, would seem likely to offer a better, more efficacious vehicle for that intent.69

The second potential critique is that I have here hardly fully demonstrated that a world state is itself defensible as a cosmopolitan ideal. After all, as I have said myself, the world state remains largely theoretically under-investigated.70 It is true that what I have said here does not amount to a full defence of the idea of a world state—and nor is it meant to. But I believe it does indicate that cosmopolitan theorists, when operating in an idealising mode, at least need to take the idea of a world state far more seriously than they have tended to thus far. We have seen that although intermediary theorists are explicit in their rejection of the world state model, they are less explicit about their reasons for that rejection. This is unfortunate because a more direct engagement with the idea of the world state would clarify that many of the criticisms of that idea function at the same time as criticisms of intermediary proposals. For example, although in this paper I have only meant to compare two ‘end state’ institutional proposals rather than think about how we get to them, we should note that the argument that there is no route ‘from here to there’ can be made as easily against an intermediary model as against a world state.71 Similarly, one cannot assert that democracy is infeasible on a global scale without implicating both the world state and intermediary models; and being a global institutional construct, individual persons would be no more able to unilaterally exit an intermediary model than they would a world state. Moreover, any worry about the ‘homogenising’ effects of a world state must apply also to intermediary models, given that both require a world populated by cosmopolitans (most typically, liberal egalitarians) to be plausible.

Upon considered analysis, one may end up rejecting the world state. But to recoil from the world state does not change the fact that the intermediary model is troublesome for the reasons I have outlined. Even on the ambitious assumption of the prevalence of a sense of cosmopolitan solidarity, sovereignty and power remain important. The idea of a world state is often disparaged at utopian, but it is cosmopolitans who deny or ignore this importance who are the real utopian theorists.


An earlier version of this paper was presented at the workshop ‘Does the world need more than one state?’ held at the London School of Economics and Political Science in June 2014; I would like to thank the participants, and Simon Caney in particular for acting as respondent. This paper was completed while I was a postdoctoral research fellow at the Centre for Advanced Studies ‘Justitia Amplificata’ at Goethe University Frankfurt.


1. See, amongst others: Daniele Archibugi, The Global Commonwealth of Citizens (Princeton, NJ: Princeton University Press, 2008); Simon Caney, Justice Beyond Borders (Oxford: Oxford University Press, 2005); Jürgen Habermas, The Divided West (Cambridge: Polity, 2006); David Held, Democracy and the Global Order (Cambridge: Polity, 1995); Charles Jones, Global Justice: Defending Cosmopolitanism (Oxford: Oxford University Press, 1999); Thomas Pogge, ‘Cosmopolitanism and Sovereignty’, Ethics 103, no. 1 (1992): 48–75; Jan Aarte Scholte, ‘Reinventing Global Democracy’, European Journal of International Relations 20, no. 1 (2014): 3–28; Kok-Chor Tan, Toleration, Diversity and Global Justice (University Park, PA: Pennsylvania State University Press, 2000); Iris Marion Young, Inclusion and Democracy (Oxford: Oxford University Press, 2000); and Lea Ypi, ‘Cosmopolitanism Without If and Without But’, in Cosmopolitanism versus Non-Cosmopolitanism: Critiques, Defences, Reconceptualisations, ed. G. Brock (Oxford: Oxford University Press, 2013), 75–91.

2. Throughout this paper, the reader can assume that all references to a world state are references to a federal world state specifically.

3. Pogge, ‘Cosmopolitanism and Sovereignty’, 61.

4. Ibid., 58. For something similar see Scholte, ‘Reinventing Global Democracy’.

5. Daniele Archibugi and David Held, eds., Cosmopolitan Democracy: An Agenda for a New World Order (Cambridge: Polity, 1995); Daniele Archibugi and David Held, ‘Cosmopolitan Democracy: Paths and Agents’, Ethics and International Affairs 25, no. 4 (2011): 433–61; Daniele Archibugi, ‘Cosmopolitan Democracy and its Critics: A Review’, European Journal of International Relations 10, no. 3 (2004): 437–73; and Archibugi, The Global Commonwealth of Citizens; Held, Democracy and the Global Order; and David Held, Global Covenant: The Social Democratic Alternative to the Washington Consensus (Cambridge: Polity, 2004).

6. David Held and Anthony McGrew, eds., The Global Transformations Reader (Cambridge: Polity, 2003); and David Held and Anthony McGrew, Globalization/Anti-Globalization: Beyond the Great Divide (Cambridge: Polity, 2007).

7. Archibugi, ‘Cosmopolitan Democracy and its Critics’, 446.

8. Archibugi, The Global Commonwealth of Citizens, 97.

9. Held, Democracy and the Global Order, 234.

10. Richard Falk and Andrew Straus, ‘Toward Global Parliament’, Foreign Affairs 80, no. 1 (2001): 212–20.

11. Habermas, The Divided West, 136.

12. Jürgen Habermas, ‘The Constitutionalisation of International Law and the Legitimation Problems of a Constitution for World Society’, Constellations 15, no. 4 (2008): 444–55.

13. Jurgen Habermas, Between Naturalism and Religion (Cambridge: Polity, 2008), 324.

14. Jürgen Habermas, The Crisis of the European Union: A Response (Cambridge: Polity, 2012), 57.

15. Habermas, The Divided West, 138.

16. Ibid., 134.

17. Pogge, ‘Cosmopolitanism and Sovereignty’, 58.

18. Archibugi, ‘Cosmopolitan Democracy and its Critics’, 454.

19. Held, Democracy and the Global Order, 230.

20. Ibid., 230.

21. Archibugi, The Global Commonwealth of Citizens, 109.

22. Ibid., 103–6.

23. William Scheuerman accuses intermediary theorists of articulating a ‘conceptual straw-man’ in their typical definitions of statehood, by appealing to ‘ideal types drawn from Hobbes and Weber’ that do not track ‘the modern state’s diverse manifestations’. Scheuerman argues instead that ‘stateness’ is a matter of degree. See his ‘Cosmopolitanism and the World State’, Review of International Studies 40, no. 3 (2014): 419–41. While recognising this argument, I here take up the ideal-type understanding of the state.

24. Simon Caney understands sovereignty to involve ‘comprehensiveness’, by which is meant ‘authority over all issues and not simply some’: Justice Beyond Borders (Oxford: Oxford University Press, 2005), 150. On Caney’s reading, it must be only from the amalgamated perspective that we could locate sovereignty in a state (federal or otherwise) at all, since no one element of government holds ‘comprehensive’ authority.

25. For example, Quebec has held referenda asking voters whether Quebec ‘should become sovereign’—if either of those referenda had returned a positive result, it would not automatically have followed that Quebec in fact became sovereign. Negotiation with the Canadian federal government would have been required. Following the referenda, the Supreme Court of Canada ultimately ruled unilateral secession illegal.

26. Max Weber, ‘Politics as a Vocation’, in From Max Weber, ed. Hans Gerth and Charles Wright Mills (London: Routledge, 2009), 75.

27. Christopher Pierson, The Modern State (London: Routledge, 2004), 8.

28. Margaret Canovan, Nationhood and Political Theory (Cheltenham: Edward Elgar, 1996).

29. Will Kymlicka, Multicultural Citizenship: A Liberal Theory of Minority Rights (Oxford: Clarendon Press, 1995).

30. Jürgen Habermas, ‘Citizenship and National Identity: Some Reflections on the Future of Europe’, Praxis International 12, no. 1 (1992): 1–19; and Jan-Werner Müller, Constitutional Patriotism (Princeton, NJ: Princeton University Press, 2007).

31. Strictly speaking, this is not the only form an intermediary model could take. For example, we can conceive of a system in which sovereignty and physical power were globally centralised, but the global level government only existed to moderate interactions between states, and hence no ‘world citizens’ were created. Bertrand Russell prescribed something along these lines in his Common Sense and Nuclear Warfare (Oxford: Routledge, 2001). However, no one offers such prescriptions today, perhaps for reasons of democratic legitimacy. All contemporary models hold world citizenship constant; the purported intermediary character of their prescriptions comes from resisting global sovereignty and/or centralised physical power. For that reason, I will not be further concerned with Russell’s model, beyond affirming that I do not consider it to be a world state, precisely due to the lack of world citizenship.

32. Notwithstanding its expository value, I won’t be making any further reference to the EU in what follows. Suffice it to say that the EU cannot be appealed to as evidence of the cosmopolitan plausibility of the intermediary model. First, it is not of a democratic standard that the theorists considered here recognise as sufficient, and it is an open question whether or not it could be brought to that standard without European federalisation. Second, the EU is not delivering on distributive justice in a ‘cosmopolitan’ manner: not only is the redistributive impact of the EU budget limited, the beneficiaries of the various ‘structural funds’ are rarely individuals (instead of states or regions), meaning that the redistribution in question is not cosmopolitan in character. Furthermore, the EU might instead be brought as evidence against the intermediary model, given its current instability. Archibugi remarks that the real-life examples that come closest to the intermediary blueprint have been transient affairs: ‘confederations that took on the essential characteristics of cosmopolitan democracy in the move toward federal arrangements’; The Global Commonwealth of Citizens, 110. The EU’s current turbulence may be demonstrating a reason for this transience.

33. Pogge, ‘Cosmopolitanism and Sovereignty’, 60.

34. Thomas Pogge, ‘Cosmopolitanism and Sovereignty’, in The Cosmopolitanism Reader, ed. David Held and Garrett Brown (Cambridge: Polity, 2010), 121.

35. Pogge, ‘Cosmopolitanism and Sovereignty’, 60.

36. Ibid., 57.

37. Similarly, Simon Caney believes he is endorsing an intermediary model because no one political unit possess ‘comprehensive’ authority. But no one unit in a federal state possesses comprehensive authority either, so the case for intermediary status cannot be made on that basis.

38. David Held, ‘Law of States, Law of Peoples: Three Models of Sovereignty’, Legal Theory 8 (2002): 1–44, at 32.

39. Held, Democracy and the Global Order, 231.

40. Ibid., 231. It is not clear who Held foresees as deciding whether or not the global political system is continuing to meet the normative demands of his ‘cosmopolitan democratic law’. There does presumably need to be a definitive decision-maker here who can adjudicate between legitimate and illegitimate rejections of global-level decision-making.

41. Takeshi Nakano, ‘A Critique of Held’s Cosmopolitan Democracy’, Contemporary Political Theory 5 (2006): 33–51.

42. Democracy and the Global Order, 276; and Global Covenant, 113.

43. David Held, ‘Democracy and the New International Order’, in Cosmopolitan Democracy: An Agenda for a New World Order, ed. Daniele Archibugi and David Held (Cambridge: Polity, 1995), 190; see also Held, Global Covenant, 113.

44. Held, ‘Democracy and the New International Order’, 111.

45. Pogge, ‘Cosmopolitanism and Sovereignty’, 123.

46. Furthermore, for power to be ‘substantially centralised’, it must surely be the case that nuclear weapons are held, if anywhere, only at the global level. This means that the global centralisation of power would in fact not be a presupposition of disarmament; instead, it would be the culmination of domestic state disarmament.

47. Held, Global Covenant, 111n6.

48. Laura Valentini, ‘Ideal versus Non-Ideal Theory: A Conceptual Map’, Philosophy Compass 7, no. 9 (2012): 654–64.

49. It is true of Held in particular that he has also addressed the transitional question, in Democracy and the Global Order and elsewhere.

50. Archibugi makes clear that domestic states’ inclusion in the global political system remains ‘voluntary and revocable’: The Global Commonwealth of Citizens, 106.

51. In his recent writing on Europe, Habermas approvingly notes that the European-level political system lacks the authority to amend the European constitution, or ‘the competence to decide about its own competence’ (The Crisis of the European Union, 26). This marks a distinction from a federal state order: while amendments of, say, the US constitution are ‘contingent on the agreement of the legislative bodies of a qualified majority of the states, amendments to the European treaties require … unanimity among the member states’ (ibid: 40). Federative units in the US thus lack a veto; states in the EU retain one. Habermas say of his global institutional model that

No structural analogy exists between the constitution of a state that can determine what political competences it claims for itself (and hence possesses supreme constitutional authority), on the one hand, and the constitution of an inclusive world organisation that is nevertheless restricted to a few, carefully circumscribed functions, on the other. (The Divided West, 134)

I take Habermas here to be proposing that states would retain a veto over any proposed alteration to the UN body of law, and thus they would ultimately retain their sovereignty (this is presumably the reason that Habermas thinks the role of the UN will remain ‘carefully circumscribed’).

52. Archibugi, The Global Commonwealth of Citizens, 110.

53. Ronald Tinnevelt and Thomas Mertens, ‘The World State: A Forbidding Nightmare of Tyranny? Habermas on the Institutional Implications of Moral Cosmopolitanism’, German Law Journal 10, no. 1 (2009): 63–80, at 75.

54. Richard Falk, A Study of Future Worlds (New York: The Free Press, 1975).

55. For one chronic case, see James Yunker, The Idea of World Government: From Ancient Times to the Twenty-First Century (Oxford: Routledge, 2011). Ironically enough, Yunker characterises his genuinely intermediary model in world state terms. In my view, intermediary theorists’ specificity is premature not just because theorists are getting ahead of themselves, but also because issues like global parliamentary representation just can’t sensibly be decided yet. By the time, if ever, that the creation of a global parliament with law-making power is politically feasible, who is to say how big the global population will be, how that population will be distributed, what persons’ primary political identities will be, and indeed which states will exist?

56. Bonaventura Majolo et al., ‘Human Friendship Favours Cooperation in the Iterated Prisoners’ Dilemma’, Behaviour 143, no. 11 (2006): 1383–95.

57. Brian Barry, ‘Can States Be Moral?’, in Ethics and International Relations, ed. Anthony Ellis (Manchester: Manchester University Press, 1986), 61–83, at 70.

58. John Rawls, A Theory of Justice (Cambridge, MA: Harvard University Press, 1999), 236.

59. In 2003, Belgium introduced a revised Law on State Accountancy determining that the government must include a ‘solidarity note’ with its budget justification. This solidarity note explains ‘the measures [the government] envisages taking with a view to attaining, at the latest from 2010 onward, 0.7% of gross income for the funds dedicated to Belgian ODA’. Belgium has to date never met the target. See for data on overseas development assistance by country.

60. I address this issue at further length in Luke Ulaş, ‘Cosmopolitanism, Self-Interest and World Government’, Political Studies 64, no. 1 (2016): 105–120 (online early doi:

61. Habermas, The Divided West, 129–30.

62. Ibid., 132.

63. Archibugi, The Global Commonwealth of Citizens, 95; see also ‘Cosmopolitan Democracy and its Critics’, 454.

64. ‘The Constitutionalisation of International Law’, 450.

65. The Divided West, 173.

66. Ibid., 173.

67. Habermas, Between Naturalism and Religion, 322. At an earlier point in time Habermas himself appeared conscious of these types of problems. See Jurgen Habermas, ‘Kant’s Idea of Perpetual Peace, with the Benefit of Two Hundred Years’ Hindsight’, in Perpetual Peace: Essays on Kant’s Cosmopolitan Ideal, ed. James Bohman and Mattias Lutz-Bachman (Cambridge, MA: MIT Press, 1997), 113–54, at 127–8.

68. That Kant’s worry about ‘soulless despotism’ is explicitly linked to the idea of a ‘universal monarchy’, which is a form of world state that arises as a result of domestic states ‘being absorbed into a single strong hegemonic state’ is frequently overlooked by those who appeal to Kant to fast-track their opposition to the idea of a world state. See Pauline Kleingeld, ‘Approaching Perpetual Peace: Kant’s Defence of a League of States and his Ideal of a World Federation’, European Journal of Philosophy 12, no. 3 (2004): 304–25, at 313.

69. For an example see Adrian Little and Kate Macdonald, ‘Pathways to Global Democracy? Escaping the Statist Imaginary’, Review of International Studies 39, no. 4 (2013), 789–813.

70. See, however: Luis Cabrera, Political Theory of Global Justice: A Cosmopolitan Case for the World State (London: Routledge, 2004); Raffaele Marchetti, Global Democracy: For and Against (Oxford: Routledge, 2008); and Danilo Zolo, Cosmopolis: Prospects for World Government (Cambridge: Polity, 1997).

71. For an example of that argument, see Barbara Buckinx, ‘Domination in Global Politics’, in Global Governance, Global Government: Institutional Visions for an Evolving World System, ed. Luis Cabrera (Albany, NY: SUNY Press, 2011), 253–82.

About The Author

Luke Ulas
Goethe University, Frankfurt.

Article Metrics

Metrics Loading ...

Metrics powered by PLOS ALM

Related Content