Federal world government: The road to peace and justice?

Introduction

In 1793 – the year in which Louis XVI was executed by guillotine and France declared war on Spain and Great Britain – Kant famously wrote that the hope ‘for a perpetual peace based on a federation of nations united in a world-republic’ was universally seen as ‘sheer fantasy’. For most people philosophical chiliasm was just as far-fetched as its theological counterpart – which holds up the prospect of ‘the completed moral improvement of the human race’ (1996: 81). Despite the passing of more than 200 years, the existence of numerous international organizations and the popularity of global governance, things have not fundamentally changed. There is still a widespread dismissal of the idea of a world state. ¹ In the current debate on global justice, for example, the idea primarily has a negative function. It is often used to argue against forms of egalitarian justice and global democracy.

Recently, several theorists not only (1) reclaimed the idea of a world state as a topic ‘worthy of serious considerations’ (cf. Scheuerman, 2011; Tamir, 2000: 242), they also (2) argued for some form of global government. Within the field of International Relations, Wendt and Deudney are probably the best-known exponents of this ‘resurgence of the idea of world government’ (cf. Craig, 2008). Höffe, Cabrera and Tännsjö are exemplary for the domain of political theory, and Yunker for economics. Their focus on the idea of a world state, however, differs considerably. Whereas Wendt claims that such a state is inevitable, the other five theorists merely argue that it is desirable and feasible. ² The emphasis of this article is primarily on the last focus.

I have no quarrel with the first claim of these theorists. There is no reason why the idea of a world state should simply be rejected out of hand (cf. Tamir, 2000: 242). In fact, there are good reasons why it deserves more theoretical reflection than it currently gets. Some of these are discussed below. I am also not unsympathetic to the second claim – that some form of world state is necessary to cope with the challenges posed by globalization. What can be disputed, however, is that the form of world state they uphold is sufficiently elaborated. This article discusses two main problems. The first deals with the shape of the world state, the second with its competences.

Comparing the theories of the aforementioned theorists, one will quickly notice that they all reject the idea of a unified global state and defend a limited and multilayered form of world government. Regarding the distinguishing features of a global state there is broad consensus. A global state is a state without any kind of ‘political realization of difference’ (Walzer, 2003: 171–172). All human beings are directly connected to a single centre of political power. Concerning the defining characteristics of a minimal world state, however, there is no such agreement. Some defend a monopoly on the use of violence, others a system of decentralized sanctions. Some, furthermore, argue for hierarchy, whereas others plead for non-hierarchical world federal government (Deudney, 2007: 20). Finally, some claim that a world state should only guarantee ‘peace’, while others include ‘justice’. ³

It is precisely at this point that the analyses of contemporary proponents of a minimal world state fail to convince. If we want to make sense of the idea of a world state, we have to take care not to relax the criteria for modern statehood too much. Wendt makes this claim when he argues that ‘[a] non-Weberian approach, especially if it relaxes the monopoly on the use of force requirement, could suggest that a world state is already here (The UN?) or just around the corner (…), which would make the argument [for a world state] easier but less interesting’ (2003: 506). There is a limit to how far the idea of state can be stretched. World government is not the same as world governance. Where should the line be drawn?

This is not the main problem that haunts current proposals for a minimal world state, however. Most proponents also have a tough job finding the right balance between: (a) the reasons why such a state is needed, (b) the competences it should have, and (c) the safeguards that are installed to prevent a gradual expansion of these competencies. At first sight, a minimal world state seems better capable of addressing the traditional arguments against a world state than a global state. It offers a more realistic arrangement of international society. Whether or not this promise can be fulfilled, however, depends on the safeguards that are installed to prevent an unrestrained interpretation of the responsibilities of such a state. Can these safeguards prevent such a bloating of government? If not, the model of a minimal world state will move in the direction of a global state and will again be exposed to the traditional arguments – like the fear of global despotism.

The main aim of this article is limited. Instead of arguing for the necessity or inevitability of a minimal world state, it merely discusses two important barriers that current proponents face. The structure of the remainder of this article is as follows. Section 2 presents an overview of the traditional arguments against the world state. Section 3 starts with the presumption that a minimal world state is better equipped to deal with these objections than a global state. It then asks how a minimal world state should be defined. The competences of such a state and the reasons why it is needed are discussed in section 4. Two claims are made. Contemporary models of a minimal world state are not ‘minimal’ at all, and could imply more competences and responsibilities than proponents are willing to admit. Section 5 critically discusses some of the constraints that are installed to prevent a bloating of government.

1. From positive idea to negative surrogate?

Suppose we want to defend the idea of a world state, more specifically a global state. What objections do we then face? ⁴ The realist tradition is probably a good starting point to look for some of the objections that are traditionally raised. Despite their often critical – but not necessarily dismissive – attitude towards world government, realists like Morgenthau (1985), Waltz (1959) and Bull (2002) do give it a fair hearing. In Man, the State, and War, Waltz writes the following: ‘the obvious conclusion of a third-image analysis is that world government is the remedy for world war. The remedy, though (…) unassailable in logic, is unattainable in practice’ (1959: 238). Morgenthau makes a similar point: ‘There is no shirking the conclusion that international peace cannot be permanent without a world state, and that a world state cannot be established under the present moral, social, and political conditions of the world’ (1985: 536). ⁵ By rejecting ‘in hypothesi what is correct in thesi’ Morgenthau and Waltz echo Kant’s claim that ‘the positive idea of a world republic’ should be replaced by ‘the negative surrogate of a league that averts war, endures, and always expands (…)’ (Kant 1999: 328).

Why should this be the case? Ten objections to world government are discussed. Realists first of all emphasize the utopian character of proposals for some form of world state. Proponents lack a fundamental understanding of empirical questions and real world politics. ⁶ They fail to live up to the dictum formulated by Rousseau that we should take ‘men as they are, and the laws as they can be’ (1997: 41). Contemporary defenders of a world state can, in this regard, be criticized for precisely the same reasons as the first advocates of a perpetual peace project. The proposals of Jeremy Bentham, William Penn or Pierre Dubois were often seen as little more than the foolish daydreams of utopian theorists.

Twentieth-century theorists like Carr or Waltz, likewise, blame their internationalist antipodes for a similar lack of realism. ‘Utopian thinking’, according to many realists, is at odds with the real world of politics. Given the anarchical nature of the international order, the self-interested motivation of states and the constraints imposed on politics by the egoistic passions of human beings, it is very unlikely that a stable form of world state will come into being (infeasibility objection [1]). And even if some form of world state would one day be feasible, such a state could not be developed without changing the condition of peace into its opposite – without causing civil war. Rousseau was very explicit about this point. The process of creating a world state would, according to him, ‘do more harm in a moment than it would guard against for ages’ (1917: 112) (consequentialist objection [2]).

The argument of infeasibility or impracticability is a very powerful one. But one could also argue against world government because it does not present a very desirable ideal. A world state, according to critics, is undesirable because it inescapably poses the risk of global despotism (tyranny objection [3]) and because it is an extremely weak and inefficient form of political organization (ungovernable objection [4]). The tyranny objection is, of course, the most notorious of all arguments against the world state. It consists of two interrelated claims: (a) the objection that a world state – because of its need for hierarchical centralization – poses a great and maybe even inevitable risk of tyranny, and (b) the claim that, unlike a tyrannical nation-state, a tyrannical world state leaves no exit possibilities.

The ungovernable objection, however, also has a long tradition. Grotius, Rousseau and Kant already argued that the advantages of political centralization are often counterbalanced by even greater disadvantages when political structures become too large. Grotius draws a parallel with a ship to further elaborate this point:

For as a Ship may be built to so vast a Bulk, as to be unwieldly, and not manageable so an Empire may be extended over so great a Number of Men and Places so widely distant from each other, that the Government of it becomes a Task, to which no one Sovereign can be equal. (2005: 1108)

What the picture of the world state as an unmanageable mammoth tanker nicely illustrates is that the scale of government is inversely proportional to the ‘vigour’ of laws (Kant 1999: 336).

Suppose proponents of a world state could prove that their ideal is both feasible and desirable, would that take away all the realist’s doubts? Unfortunately not, realists still have a fall-back. The establishment of a world state is not only infeasible or undesirable, but also superfluous. This is not only the case because global problems and security threats can also – and maybe even more effectively – be dealt with on lower levels of governance (superfluous objection [5]). Another reason is that there are significant differences between the state of nature between individuals and the state of nature between states (no social contract objection [6]). Although theorists like Locke and Rousseau claim that a social contract between states is needed, they – for different reasons – are not prepared to argue for some form of world state. One of these reasons is that no sufficient analogy exists between the state of nature between individuals and that between states. Similar accounts can be found in the works of Pufendorf, Hobbes and many contemporary political theorists. Post-national arrangement of international society, according to the last, should not literally mirror the construction of the nation-state. Within international relations theory, critics like Manning and Bull can be mentioned. Both oppose the ‘domestic analogy’, which Bull famously described as follows:

[T]he argument from the experience of individual men in domestic society to the experience of states, according to which the need of individual men to stand in awe of a common power in order to live in peace is a ground for holding that states must do the same. (Cited in Suganami, 1989: 11)

These six types of objection apply to any type of global state, but there are, of course, also arguments that specifically apply to proposals for a democratic state. Liberal nationalists like Kymlicka (2001) and Miller (2000) are the best-known critics in this regard. Considering that democracy and redistribution depend on a high level of trust and common understanding, liberal nationalists claim that there are strong limits to the scale of political communities. Bounded communities – with a certain degree of homogeneity in culture and language – are needed to further democratic politics and economic justice (bounded citizenship objection [7]). This argument closely resembles the restricted size and the democratic distance argument (both part of the ungovernable objection). The underlying rationale, however, is different. Whereas the restricted size and democratic distance objection address the objective conditions for democratic communities to sustain themselves, the bounded citizenship objection is primarily concerned with the subjective conditions for genuine forms of citizenship and democracy – with the glue that holds democratic political communities together.

The importance of bounded communities, however, is not the only reason why theorists like Miller and Kymlicka argue against a world state. Another critique is that a world state poses a serious threat to the concrete ethical life of political communities. A world state nullifies cultural differences (pluralism objection [8]). If unity requires centralization and pluralism decentralization, then extreme centralization – as is the case in a unified global state – will leave no room for a political realization of social or cultural differences (Walzer, 2003: 171–191). ⁷

Beyond the liberal nationalist objection that democracy requires bounded communities that are united around a common feeling of national belonging, the idea of a world state is finally also objected to in cruder terms. According to this type of argument the identity of political communities depends on a form of ‘closure’ (Walzer, 1983: 39). Unlike the bounded citizenship objection, the argument is not so much that feelings of political commitment are necessarily bound to a specific culture or language, but rather that feelings of belonging to a certain group are essentially feelings of not belonging to another group (collective identity objection [9]).

Schmitt and Mouffe develop a similar objection in their analyses of the domain of the political. Characteristic of this domain is a clear-cut distinction between friend/enemy. No politics without competing collective identities, without a differentiation between ‘we’ and ‘they’. This contrast constitutes its ‘condition of possibility’ (Mouffe, 2005: 16). That is why Schmitt so fiercely argued that a ‘completely pacified globe’ – a cosmopolitan community – would be ‘a world without politics’ (1996: 35; antagonism–agonism objection [10]). Given that Schmitt and Mouffe differ in their interpretation of the we/they distinction, we use this twofold term and not simply ‘antagonism’ or ‘agonism’ objection. Although Mouffe doesn’t deny that the ‘we/they’ distinction could become the ‘locus of an antagonism’, she stresses the importance of a ‘common symbolic space’ within which the conflict between ‘we’ and ‘they’ takes place (2005: 16, 20). Within this space the conflicting parties recognize each other’s legitimacy.

Taking stock of these ten arguments against world government, one will notice that not all of them are fully convincing or irrefutable. Many, in fact, are not. ⁸ The general argument against world government based on plurality and identity grounds, for example, rests on the mistaken assumption that ‘exclusion’ should be thought of in purely spatial terms. But why should political identity formation presuppose a geographically external other? Why can’t we construct difference and otherness imaginatively (Abizadeh, 2005: 58) or, as Wendt argued, ‘through a temporal differentiation between its present and its past delimitation of the temporal dimension’? History, in the last case, becomes ‘the Other in terms of which the global Self is defined’ (2003: 527). A global democratic community would then develop by distinguishing itself from what it is not: a voluntary association of strategically oriented nation-states.

Despite the fact that not all of these objections are fully convincing as they stand, they do present – as a whole – a knock-down argument against the idea of a unified global state. A completely centralized state presents an infeasible and undesirable ideal. This is, of course, neither a surprising nor a novel claim. Other forms of world state, however, seem better equipped to deal with the traditional arguments against the world state than a global state. The model of minimal or limited world government is the most likely candidate. Yunker gives the following, rather general, definition: ‘world government subject to sufficient and adequate constraints to preclude it from establishing an intolerable totalitarian hegemony over the nations of the world’ (2007: 25). World government is government ‘sufficiently limited and constrained to guarantee the natural rights of its component nations and peoples’ (2005: 206).

Central to this model are the dual principles of federalism and subsidiarity. By recognizing the importance of these principles room is created for a multi-level political system in which nation-states can still be a privileged site of democracy. The general objections that are most difficult to counter – even for proponents of limited world government – are the no-exit, the feasibility and the no-social contract objection (cf. Tinnevelt, 2011). Especially the second one will continue to expose the idea of a minimal world state to strong realist critique. But these objections do not necessarily have enough power of persuasion to reject a form of minimal world state. Instead, they show the relevance of further research.

These objections do, however, serve as a warning. If proposals for a minimal world state come too close to the idea of a unified global state they can be discarded. Unfortunately not all theorists have taken this lesson to heart.

2. Which world state?

The idea that alternatives to the idea of a unified global state are possible has a long and distinguished intellectual history. Take Immanuel Kant’s distinction between world republic and universal monarchy. Or consider Dante’s claim that the supreme ruler of a universal monarchy should not take every ‘trivial decision in every locality’ (1995, I, 14). Dante’s reply to the pluralism objection [8] points in the direction of a federal system: ‘For nations and, kingdoms and cities have characteristics of their own which need to be governed by different laws; for law is a rule which governs life’ (1995, I, 14). Other forms of world state, however, can be differentiated. Pojman – following the minimalist–maximalist controversy in the world federalist movement – distinguishes three forms of world government: minimum (states retain complete sovereignty internally), moderate (states retain some internal sovereignty) and maximal (states give up both internal and external sovereignty) (2006: 63).

It is important, therefore, not only to be sensitive to the traditional arguments against such a politico-juridical arrangement of international relations, but also to be aware of the fact that ideas like world state, global government or global democracy are often used with different meanings and implications. Not every global democrat argues for world government, and not every advocate of global government would welcome a world state.

So how should we define a world state? A major obstacle to answering this question is that no wide agreement can be found within political theory regarding the defining features of the modern state. Not only is the ‘state’ a historical phenomenon but the concept itself has a history (Kukathas, 2008: 1). The word is used in different ways; often even by the same author. Sometimes it refers to the body politic or society, sometimes to a specific organ of society such as the government (Kelsen, 2006: 81). In addition, many rival definitions can be found (cf. Dunleavy, 2007 and Morris, 2004). So how should we define what a state is? Do we take Hegel’s ‘realm of objective reason and morality’ as our starting point, Bodin’s ‘union of the people under the same sovereignty of government’ or Kelsen’s identification of state and legal order?

One way of resolving this dilemma – not necessarily the best or only one – is to stick to the most common and regularly used definition we can find (most proponents of a minimal world state take this route). The definition that Weber develops in the first part of Economy and Society is a good candidate in this regard:

It possesses an administrative and legal order subject to change by legislation, to which the organized corporate activity of the administrative staff, which is also regulated by legislation, is oriented. This system of order claims binding authority, not only over the members of the state, the citizens (…) but also to a very large extent, over all actions taking place in the area of its jurisdiction. (1947, 156)

Most of these features can be organized under the heading of the following criteria: ‘legality, supremacy, territoriality, and comprehensiveness’ (Caney, 2005: 150).

The problem we face at this point is the following. To make sense of: (a) different forms of world state, (b) the many manifestations of the modern state (federal, multinational, etc.), and (c) the fact that the traditional concept of the modern state is not the best source of political imagination within a globalizing world (cf. Walker, 2002), some of these criteria need to be relaxed. Which criteria and how far? As we indicated in the Introduction, such an endeavour is not without its difficulties. Recall Nozick’s view on what it takes for an institution to be recognized as a state. By slightly weakening the classical Weberian conditions – by opting for a de facto monopoly on the use of force and dropping the fact that a state claims to be the ‘sole authorizer of violence’ – Nozick is capable of recognizing even his ‘dominant protective agency’ as a state (1974: 117). So we should be careful not to relax the criteria for statehood too much. If we do the idea of a world state would cover too many arrangements of international relations to still be a meaningful concept.

Which criteria need to be relaxed? The idea of comprehensiveness first comes to mind. According to this feature a political unit should be recognized as a sovereign state ‘if it has ultimate authority over a territorially defined group of people and it has authority over all issues and not simply some’ (Caney, 2005: 150). Clinging to this idea will straightforwardly lead to a global state. Which issues should be dealt with on the national, international and global level partly depends on the types of argument that are being used to defend some form of world state (see below). Finding a solution to this problem, however, might turn out to be more difficult than initially expected. Not only are an increasing number of problems recognized as global, but many national issues turn out to have global effects.

Another criterion that should be weakened is that of (supreme) authority. The rationale behind this is quite simple. Even on the ‘national’ level federal systems of government cannot properly function without some level of shared sovereignty. But again it is very difficult to determine how responsibilities should be allocated within multi-level political systems.

Even less obvious is how the idea of a monopoly on the use of force should be dealt with. Looking at the different arguments that are given for a world state – peace [A], security [B], legitimacy [C], collective action problems [D] and justice [E] ⁹ – one would expect proponents to defend a political arrangement that is both authoritative and effectively capable of securing all these values and goals. Many contemporary proponents of a minimal world state, however, distance themselves from this claim. Yunker’s Federal Union of Democratic Nations, for example, rests on two fundamental provisions:

(1) that member nations retain perpetually the right to secede (…) at their own unilateral discretion; (2) that member nations retain perpetually the right to maintain whatever military forces and armaments under their direct control they desire, up to and including strategic nuclear weapons. (2007: x)

Government, however, is still defined as: ‘a state organization with direct ties to its constituent population (…) with the power to pass laws and levy taxes, and with the power to raise and permanently maintain armed forces’ (Yunker, 2005: 31).

Although this kind of arrangement will bypass the no-exit [3b] and consequentialist objection [2], traditional world federalists will probably balk at this idea. According to their notion of a world state (based on the idea of complete disarmament), the combination of both provisions leads to a situation in which there is no positive security against violence. Positive security not merely implies a state of non-violence but also a strong mechanism ‘for resolving conflicts of interests via negotiations or law’ (Pojman, 2006: 53). By ditching the demand of a ‘monopoly on heavy weaponry’ (Yunker, 2007: 25) the Federal Union must rely on the deterring effect of its own arsenal. Formulated differently, there is a ‘willingness to establish peace’ but no ‘reliable protection of peace’ (Höffe, 2007: 204).

Other political theorists accept that a world state would imply a hierarchical scheme, but deny that the sanctioning system should be centralized. Kelsen’s pure theory of law is a case in point. Decentralization and a strict monopoly on the use of force are, according to Kelsen, not in contradistinction. Primitive and international law are characterized by the ‘complete lack of a particular organ charged with the application of legal norms to a concrete instance’ (2006: 338). Individuals and states take the law into their own hands. Nevertheless, both primitive and international law can be seen as law:

A social order which has not progressed beyond the principle of self-help may produce a state of affairs leaving much to be desired. Nevertheless it is possible to consider this state a legal state, and this decentralized order a legal order. (Kelsen, 2006: 339)

Within the current political theoretical debate, Cabrera also denies that ‘the highest-level governing bodies must have a monopoly on the use of force’. By applying the principle of subsidiarity to the ‘question of armed forces’, he is able to opt for a ‘dispersal and balancing of global coercive authority’ (2004: 114). The main rationale behind his defence of a decentralized system is the tyranny objection [3a]. But Cabrera also touches on a related topic: whether compliance can only be enforced on the basis of the threat of the sword. International legal orders, according to some authors, need not necessarily be enforced on the basis of independent sanctions. This can also be done through some form of cooperative interaction like the imposing of costs (cf. Scott and Stephan, 2006).

Some proponents of a minimal world state, finally, not only fiddle with the idea of a monopoly on the use of violence but also with the idea of a hierarchical system as such. Deudney’s plea for a ‘fully non-hierarchical, fully republican, world federal government’ (2007: 20) is very interesting in this regard. On the one hand, Deudney clearly argues for a form of substantive world government. Anarchy characterized by high levels of violence interdependence is incompatible with security from violence (Deudney, 2007: 4, 274). On the other hand, he does not straightaway argue for some form of hierarchy. Following the tradition of republican security theory, Deudney points to the hierarchy-restraint problem: ‘governments can themselves pose as severe a security threat as first anarchy’ (Deudney, 2007: 46).

The tyranny objection [3] is again the guiding principle. Deudney replaces the realist dualism between anarchy and hierarchy with the trinity anarchy–negarchy–hierarchy. Unlike actors in hierarchies, actors in negarchies are not in an ‘ordinate and subordinate relation’ but are ‘authoritatively ordered by relations of mutual restraint’ (2007: 48). By defending a ‘federal-republican image of world government’ and criticizing the federal state model, however, Deudney seems to return a more nuanced version of the traditional realist dictum of a balance of power – a balance between different republics.

Combining the positions of Yunker, Cabrera and Deudney we can return to our original question. What are the defining features of a world state? If we defend a decentralized and fully non-hierarchical arrangement of international relations – one in which states have a right to secede and retain the right to their military forces – are we then still arguing for some form of world government or state? Or are we merely supporting a system of global governance? With the exception of Höffe, no proponent of a minimal world state is prepared to offer such an analysis.

3. ‘Every kingdom divided against itself shall be laid waste’

However, even if we are able to give a satisfying account of the defining features of a world state, a second problem still looms large. Are contemporary proponents of a minimal world state able to keep a balance between: (1) the reasons why a world state is needed, (2) the competences such a state should have, and (3) the safeguards that need to be installed to prevent a bloating of government? As is argued below, most of them are not. ¹⁰ Here we focus on two reasons. The advocates of a minimal world state: (a) use a very broad set of arguments to argue for a world state (consequently assigning extensive responsibilities to it), and (b) are heedless of some of the implications of these arguments – some competences naturally lead to others.

The easiest way to maintain a stable balance between these three elements is by limiting our ambitions to something similar but not identical to an ultraminimal or night-watchmen state, ¹¹ and only argue for some form of minimal peace between states. An ultraminimal state, according to Nozick, ‘maintains a monopoly over all use of force except that necessary in immediate self-defence, and so excludes private (or agency) retaliation for wrong and exaction of compensation; but it provides protection and enforcement services only to those who purchase its protection and enforcement policies’. The night-watchmen state differs to the extent that it is redistributive, ‘it compels some people to pay for the protection of others’ (1974: 26). An ultraminimal world state, accordingly, would have minimal competencies and responsibilities (cf. Höffe, 2007).

Looking at the most straightforward and most often used case for world government, one might think that this is what proponents of a minimal world state have in mind. Many deduce the need for some form of world government from the idea of a Hobbesian state of nature and the importance of the idea of peace (A). Whether this argument is valid depends partly on the content of ‘peace’ and partly on the validity of the no social contract or disanalogy objection [6]. ‘The suggestion to secure international peace through World State’, as Kelsen rightly remarks, ‘is based on the analogy assumed to exist between a World State and the national State by which national peace is so effectively secured’ (1944: 5 f.). Although Kelsen primarily reminds us of the way in which the national state has come into existence – i.e. on the basis of conflicts – the disanalogy between the Hobbesian state of nature and international anarchy runs deeper.

Bull’s analysis of the argument from international anarchy is of interest in this regard. He convincingly argues that the argument has three major flaws: notions of right and wrong are not absent from modern international relations, the society of states looks more like Locke’s state of nature than Hobbes’s, and the domestic analogy is limited in several regards (Bull, 2002: 45–47). Although these objections do not entirely refute the argument from international anarchy, they do function as a warning. Arguments focusing on the scheme anarchy–order–government can only be taken seriously if they are based on a searching analysis of the differences between the national and international stage. Unfortunately, not many proponents of a minimal world state develop such an analysis. Deudney’s analysis of pre-state anarchy, the anarchic state system and violence interdependence forms an exception (2007: 33 ff.).

The importance of peace between states, however, is not the only argument given for a minimal world state. A second type of argument is security [B]. Depending on the content of this concept – how narrow or wide ‘security’ will be interpreted – the competences of a minimal world state will differ. Does security, for example, only imply a negative ideal (such as security against physical violence) or does it also refer to a more positive notion (such as security against basic rights violations or the protection of basic goods)? In the second case, the range of competences and responsibilities of the world state will widen.

The main problem is that ‘security’ is a container concept. Many theorists, for instance, claim that subsistence rights should be seen as part and parcel of the category of basic rights. Others even claim that the differences between security and subsistence rights ‘are far too fine to support any weighty conclusions’ (Shue, 1996: 37). Both types include the same duties: ‘to avoid depriving’, ‘to protect from deprivation’ and ‘to aid the deprived’ (1996: 52). The wide definition of ‘human security’ in the Human Development Report of 1994 is also a case in point. Human security does not only mean safety from ‘hunger, disease, and repression’ but also protection ‘from sudden and hurtful disruptions in the patterns of daily life – whether in homes, in jobs or in communities’ (UNDP, 1994: 23).

An example of how a similar ‘chain of responsibilities’ can lead to a form of hierarchical centralization is the following. If we want the WTO to monitor the use of unfair trade barriers, should we then not also make the WTO responsible for monitoring monopolies and cartels? Both, after all, unjustifiably prevent free trade. So a minimal range of duties could entail more responsibilities than initially anticipated (cf. Hank and Berthold, 2002: 310). Within international law one could point to the discussion on the idea of ‘implied or implicit powers’. To what extent do international organizations also have the competences that are needed for effectively exercising their functions? How restrictive or wide should international agreements be interpreted? These questions need to be dealt with.

Proponents of a minimal world state, however, do not merely argue that basic rights need to be secured. (Nuclear) war and human security are not their only points of reference. They also want global decisions regarding these rights to be legitimate [C]. Although political legitimacy need not necessarily imply a claim for full-blown democratic decision-making – supranational decisions might, for example, be indirectly legitimized on the basis of the consent of the democratic constituent member states – even limited forms of legitimacy will further increase the distance towards the model of an ultraminimal world state. Some proponents, however, argue that democratic legitimation can only be effected on the basis of the interplay between the strong and weak global public sphere. Political legitimacy, in this case, does imply a form of global democracy (Marchetti, 2008: 7; Tännsjo, 2008).

But if a minimal world state not only depends on the ideas of peace and security but also on that of democracy, more extensive questions regarding design and feasibility arise. The question of solidarity and communicative embedding, for example, becomes acute – and the bounded citizenship objection [7] emerges. Again it is possible to argue that it is very difficult to defend a truly ‘minimal’ world state and that contemporary proponents pay too little attention to these and related problems.

Similar difficulties apply to the last two arguments for a minimal world state: collective action problems and global distributive justice. Depending on the definition of basic rights and the notion of distributive justice that is used, one – or maybe even both – of these arguments could be subsumed under the heading of ‘security’. Extreme poverty, for instance, can be seen as a form of physical violence.

The rationale behind the argument of collective action problems [D] is well known. Humanity faces transboundary problems (like global warming or economic crises) that are ‘intractable if tackled on a less than global level’ (Tännsjö, 2008: 1). International agreements regarding these issues can only be effective if they are coercively enforced by a centralized agency. Whether this argument is valid depends, among other things, on the reply given to the superfluous objection [5]. Again more research into the specific character of a minimal world state is needed. What type of effectiveness do proponents of a minimal world state have in mind and what kind of responsibilities are required? Many proponents simply accept the effectiveness argument and assume this efficiency has no downside in terms of the necessary competences of their minimal world state.

The rationale behind the global justice argument [E] will vary according to the specific approach taken. Cabrera and Höffe offer a right-based theory, whereas Marchetti starts from a consequentialist one. Although favouring a utilitarian theory, Tännsjö tries to remain neutral regarding the correct approach. While sceptical of a global ‘redistribution of economic resources’, he is convinced that world government ‘creates better conditions for local struggle for global justice’ (2008: 50) – for example, by removing trade barriers. Most other advocates of a minimal world state, however, defend robust forms of redistribution of welfare and other primary goods. Global distributive justice implies more than just the fulfilment of basic wants (such as food and housing). It also entails ‘freer movement among states’ and a ‘just distribution of life opportunities’ (Cabrera, 2004: 4).

What are the implications of the global justice argument for the competences of a world state? It all depends on whether world government is only called for to guarantee distributive justice within states or also between them and their citizens. In the first case the minimal world state only has a subsidiary and hopefully temporary function regarding social standards. In the second case a far more elaborate and permanent system of government is needed. Moving in this direction, though, will forcefully subject the model of the minimal world state to the bounded citizenship [7] and pluralism objection [8]. The only acceptable solution exists in grounding such a model in a multi-level ethical position. But this solution, as we will see in section 5, has not been adequately carved out yet.

Accepting all – or only some – of these five different types of argument for a world state can, therefore, lead to a substantive form of world government that leaves the model of an ultraminimal world state far behind. The arguments of distributive justice and democracy, for example, could imply a state with a world parliament, centralized taxation system, a strong economic programme to ‘transfer large quantities of new investment resources from the rich nations to the poor nations’ (Yunker, 2007: 13), and a high court. Given this mechanism it is not only difficult to determine what ‘minimal’ in minimal world state refers to, but it becomes increasingly important to pay attention to the safeguards that need to be installed to prevent an even further shift in the direction of a global state.

4. A bloating of government and bureaucracy?

Limited world government, according to Yunker, is ‘subject to sufficient and adequate constraints to preclude it from establishing an intolerable totalitarian hegemony over the nations of the world’ (2007: 25). Above, we discussed what type of constraints Yunker has in mind. What he shares with most other proponents of a minimal world state is the ideal of a federal world republic. Compared with a unified global state, such a multi-level system of world government has many advantages.

By recognizing the importance of the dual principles of federalism and subsidiarity the model of a minimal world state could, in principle, counteract the pluralism (8), infeasibility (1) and bounded citizenship objection (7). Federal world government constitutes a multi-level political system in which states can remain privileged sites of democracy. Given the nature of such a multi-stage and nested system, no close unity is needed between all people on the global scale. Thinner forms of legitimation might suffice. Habermas, for example, makes such a claim when he discusses the democratization of the UN. A ‘negative consensus would suffice in place of civic solidarity, namely, shared outrage at the aggressive warmongering and human rights violations of criminal gangs and regimes or shared horrors over acts of ethnic cleaning and genocide’ (2006: 109 f.). Depending on the competences and responsibilities of the different political levels, different forms of trust, solidarity and commonality are needed.

From a philosophical point of view, however, three problems arise from the notion of ‘federal world government’. The first is a rather formal one – no comprehensive normative theory of global federalism currently exists. Without such a theory, however, it is difficult to argue for the desirability and feasibility of some form of federal world government.

Federalism has traditionally been defended for different reasons, ranging from promoting political stability, securing individual freedom and enhancing self-government to the protection of minorities’ rights. All these features seem to make federalism a very promising candidate to deal with globalization-related problems in the fields of justice and democracy. One of these problems, for instance, concerns the balance that states need to strike between the need for global competitiveness and the internal pressure to maintain the unique character of their national identity (Marychurch, 2003). Another is the way we should deal with the many transboundary problems we face in our globalizing era.

Can federalism, however, really be adapted to our new constellation? Is world federalism a feasible option? The relevance, but also difficulty, of such an endeavour is apparent to both political and legal theorists. Although the idea of federalism has a long history and has been discussed by a wide range of philosophers going from Althusius and Montesquieu to Rousseau and the Federalists, there is still a clear divide between federalist theory and practice. The idea has been relatively neglected by political philosophers in the 20th century (cf. Norman, 2006) – even with regard to the national and multinational level.

Within legal theory things are not all that different. Although federalism has not escaped the attention of legal theorists – many fine studies have been written in Germany and the US (for example Riker, Friedrich, Forsyth, Kirchhof and Stern) – the ensuing framework is not sufficient to construct a normative theory of (global) federalism. To some extent the divide between national constitutional law and the international legal order testifies to this problem. Theorists of international law and international organizations have, so far, only paid relatively little attention to what has been accomplished within the theory of constitutional law.

The second problem with the idea of federal world government is that currently no comprehensive pluralist theory of global justice exists that could sustain it. The best chance advocates of a minimal world state have of developing a feasible and convincing arrangement of international relations is if they avoid the pitfalls of monism and dualism and opt for a multi-level ethical view of global distributive justice and democracy. Although attempts have been made to develop such a view, these have mainly focused on non-pluralist dualism regarding distributive justice (Cohen and Sabel, 2006; De Bres, 2012) or need to be further elaborated (Fraser, 2009). The real challenge at the moment is to connect the debates on global distributive justice and democracy. Should we aim for ‘democratic justice’ (Shapiro, 1999) or for ‘just democracy’ (Van Parijs, 2011) at the national, international and global level? And, if so, how should this ideal be conceptualized?

The third, more fundamental, problem is whether the model of a minimal world state is able to bypass some of the pressures for hierarchical centralization that are characteristic of the national state. At first sight this seems to be the case. A minimal world state, as Deudney rightly emphasizes, ‘would not require a foreign policy and would not face the competitive pressures of an interstate anarchy’ (2007: 276; see also Marchetti, 2008: 158). A minimal world state, therefore, lacks some important external pressures for hierarchical centralization.

But what about internal pressures? To what extent does a multi-level political system that is based on federalism and subsidiarity introduce its own pressures? Here we return to the safeguards that proponents of a minimal world state install to prevent unrestrained interpretations of its responsibilities and a ‘bloating of government and bureaucracy’ (Höffe, 2007: 223). Are these principles able to avoid a gradual expansion of these responsibilities? What I would like to argue – but cannot fully develop for reasons of space – is that they might not be. ¹²

A first problem is related to the principle of subsidiarity. The only reasonable way in which one can claim that powers should rest with lower-level sub-units is if there exists a context of reasonably fair background conditions – such as an ‘initial fair, global distribution of resources’ (Gosepath, 2001: 153). Without such a context it makes no sense to ‘demand self-help from those who unfairly have less than others, but are still able to help themselves’ (ibid., p. 154). This, however, has serious consequences for the set of competences of a minimal world state. This set would not only include ‘peace’, but also ‘global distributive justice’. Although this broader range of duties de facto might only remain in play as long as a fair, global distribution of resources hasn’t been guaranteed, it still points to the fact that subsidiarity is certainly not a neutral principle.

A second problem concerns the ‘all affected principle’ as a criterion to solve the boundary problem within the multi-level political system that proponents of a minimal world state defend. The boundary or demos problem in political theory basically deals with the following question: ‘who are eligible to take part in which decision-making processes’ (Arrhenius, 2005: 14). The ‘all affected principle’ seems to offer a rather straightforward answer to this problem. All those affected by political decisions have a right to participate in the decision-making process. Within our globalizing era, however, this solution loses its apparent simplicity. Given the process of de-territorialization it is no longer possible (if it ever really was) to assume that the all-affected principle coincides with the state-territorial principle (Fraser, 2009: 24). So new sites of representation need to be developed to deal with transboundary issues.

The main difficulty with the ‘all affected principle’ is its tendency to lead to a more centralized form of government. Tännsjö’s radical defence of democratic world government nicely illustrates this point. He accepts neither the importance of subsidiarity nor that of shared sovereignty. The world parliament, according to Tännsjö, should have supreme authority over all decisions. It and it alone should decide on what level decisions should be taken (2008: 123). The power of nation-states, then, is merely a delegated one. Tännsjö gives two reasons for his radicalism. The first is related to the fact that there is no principled line between the problems that do or do not belong to the responsibility of the world state, the second to the indeterminacy of the category of ‘all affected’.

The traditional distinction between national and international affairs, according to Tännsjö, is vague, ‘politically contested and insufficient when it comes to problems to do not only with war and peace, but also with the environment, international justice and so forth’ (2008: 125). So how do we solve disputes regarding the correct delineation of national and international affairs? And who is responsible for deciding where those boundaries lie? One way of revolving possible disputes is to let states decide one-sidedly which decisions should be taken on the national level and which on the international or global level. Such a solution, however, runs counter to the basic intuitions of the model of a federal union of democratic states and gives individual states the opportunity to withdraw from international institutions at any moment they see fit and for any kind of self-centred reason. These decisions will therefore need to be taken by higher-level political institutions.

If we add to this the fact that globalization will lead to a further increase in transboundary issues and a potentially rapid widening of the circle of ‘all affected’, the need for more centralized political decision-making will quickly arise. Although I do not agree with Tännsjö’s radicalism, his criticism does reveal a dynamic of centralization that is inherent to the model of a federal world republic that is defended by theorists like Höffe, Cabrera, Yunker and others. Their limited and multilayered form of world government keeps a very fragile balance between a simultaneous need for centralization and decentralization.

5. Concluding remarks

The aim of this article has, of course, been fairly limited. It merely wanted to give an analysis of certain shortcomings of the model of a minimal world state (based on the principles of federalism and subsidiarity). A necessary consequence of this line of approach is that many relevant topics have not been addressed. The idea of ‘world community’ or ‘world society’ is a case in point. This does not imply, however, that the link between world state and society is not an important one. Amstutz, for example, rightly points to the following dilemma of world government: ‘the international system needs world government to reduce the threat of war, but the precondition for world government is world community, which can only be solidified through the political transformation of the anarchic world system’ (cited in Yunker, 2005: 48). ‘Constitutionalization’ is another important issue that has not been discussed. Some authors argue that in one form or another something like ‘world statehood’ already exists (Albert and Stichweh, 2007).

Despite its limited aim, two important challenges for proponents of the currently dominant model of a minimal world state have been elaborated. They are related to the line that can be drawn between global governance, minimal world state and unified global state. Regarding the left side of the middle position, we are confronted with a problem of definition. Contemporary proponents pay great attention to what a minimal world state is not (a form of global governance), but too little attention to what it is. Regarding the right side of the middle position we are confronted with a problem of design. Contemporary proponents have a tough time retaining the ‘minimal’ character of their world state and preventing it from moving towards a unified global state.