I. Global Justice and Global Institutions: Republicanism Is the Solution . . . Or Is It?
For a long time, the global justice debate has focused almost exclusively on whether or not the egalitarian standards of distributive justice that liberal egalitarian theories advocate at the domestic level also apply beyond borders — and, either way, on what grounds.Footnote 1 Recently, however, other schools of thought have joined the conversation. Republicans, in particular, have been arguing that the ideal of nondomination — namely freedom from arbitrary interference or alien controlFootnote 2 — can make a distinctively helpful contribution to the global justice debate. The problem of the current global order, republicans contend, is not per se the fact that it displays inequalities that we would deem unacceptable at the domestic level, but that it makes domination across borders possible. In our highly interdependent yet underregulated world — so the argument goes — agents can be dominated not just by their own fellow citizens or domestic institutions, but also by other actors (such as other states, transnational nonstate actors, or international organizations). This, crucially, occurs because high global interdependence makes transborder domination possible, yet the necessary institutional mechanisms to prevent such domination are not in place. Thus, for republicans, nondomination is particularly well equipped both to identify what exactly is morally problematic about current global dynamics and to provide sound institutional recommendations.
Upon closer inspection, however, this seems to be all that republicans actually agree about. In particular, republicanism does not speak in one voice when it comes to arguing what exactly a republican global order would look like. The wide spectrum we encounter, ranging from a full-blown global republic to fairly state-centric models, largely replicates the disagreement that can be found in the liberal debate. In this essay, I offer an account for this wide divergence. I suggest that republican institutions should ideally fulfill three different desiderata: 1) avoiding excessive concentration of power; 2) bringing informal asymmetrical power under institutional control; and 3) furthering an active, vigilant citizenry. The three desiderata often pull in different directions, at the domestic as well as at the global level. At the global level, this tension has specific implications: the three desiderata do not easily coincide on a verdict over whether we should switch to a global republic, stick to a world of states, or opt for something else altogether. Therefore, republican global institutional design requires a careful exercise of balancing. My claim in this essay is that existing republican proposals of global institutional design, instead, unduly prioritize one or two of the three desiderata without giving careful consideration to all three. Thus, I conclude that while existing disagreements among republicans do not necessarily undermine the claim that republicanism can offer uniquely promising institutional insights, republicans have so far failed to live up to what makes such insights potentially so promising: the complex, careful balancing of different institutional qualities. The essay concludes with a very brief sketch of what a (very tentative and contingent) balance of the three desiderata might look like under current political circumstances.
Two caveats are necessary before I proceed. First, the essay does not address the questions of whether republicanism is independently plausible or distinctive. With respect to the former, the argument of the essay is internal to the republican enterprise, albeit animated by the hope that my analysis of the institutional sensitivity of the republican project might broaden the theory’s appeal. With respect to the latter, the argument is neutral as to whether or not nondomination can be reduced to other conceptions of freedom. It may be true that freedom as nondomination is nothing but negative freedom in disguiseFootnote 3 — or, put more sympathetically, a shortcut to bring different liberal concerns under a common agenda. The argument put forward in this essay aims at speaking to those who feel addressed by these concerns, whether or not they believe that the language of nondomination constitutes the most appropriate way to capture them.
Second, one might argue that the essay gets the order of tasks of a theory of global justice wrong. One should first concentrate on what a republican theory of global justice requires as a matter of principles (that is, who owes what to whom and on what grounds), and only then focus on which institutional setting can best realize such principles. This has been the traditional modus operandi within the liberal debate: global egalitarians, for instance, have produced a battery of arguments to vindicate the global scope of distributive equality, treating the question of whether this necessarily requires a cosmopolitan institutional order as independent and logically posterior. For republicans, however, such a way of proceeding makes little sense. No republican could plausibly contend that it is okay to dominate those who are not one’s fellow citizens because they are not one’s fellow citizens, in a way that mirrors the liberal anticosmopolitan claim that it is okay not be worried by steep inequalities outside of one’s own polity. For reasons that should be self-explanatory, saying that outsiders may be permissibly dominated (that is, be at the mercy of other agents) would equate to saying that they are not worthy of respect — whereas it is at least prima facie plausible to claim that respecting outsiders is compatible with having weaker or no obligations of distributive justice toward them. There is a fundamental asymmetry between the republican and the liberal egalitarian outlook here. Liberal anticosmopolitans typically argue that, whereas all moral agents are entitled to equal concern and respect, we have more demanding obligations of justice toward our fellow citizens in virtue of some special relation we have with them (a relation which triggers obligations that would not otherwise exist). Republicans who object to cosmopolitan institutions argue, instead, that the same obligation (to avoid domination) might require different things in different contexts; they might contend, for instance, that the best way to ensure the nondomination of outsiders is to secure a robustly enforced system of hard sovereignty for all constituencies, so as to protect smaller and weaker polities. Thus, we might say, all republicans are cosmopolitans at the fundamental moral level,Footnote 4 and the real question for them is what it takes, institutionally, to not dominate all moral agents — what kinds of protections must be in place for that to be possible. This should clarify a crucial insight: not being dominated means enjoying a particular kind of institutional environment — there is no nondomination without institutional protections. I am not dominated if I am robustly shielded from the ability of other agents to interfere with me on an arbitrary basis — that is, if robust protections are present that deprive them of the capacity to do so. This means being in relevant institutional relations with them. Republicans do not disagree about whether or not all moral agents are entitled to nondomination, but on whether we all need to be in the same institutional relationships for that to be achieved. The question is, therefore, an institutional one from the outset.Footnote 5
The essay is structured as follows. Section II introduces the main features of the republican approach and illustrates the ambivalent relationship between republicanism and global justice, largely by drawing a comparison with liberalism. This offers an entry point into both republican theorizing simpliciter and the different institutional concerns that preoccupy republican approaches. Section III introduces the three desiderata that republican institutions should aim to meet and illustrates how these might pull in different directions, by focusing on the domestic level. Section IV argues that, at the global level, the tension between these three desiderata generates pressing questions regarding the allocation of sovereign powers across different institutional levels. Section V concludes by offering a very brief and tentative solution to the problem raised in Section IV — one that is tentative not only because of the sketchy nature of the proposal, but also because, given the nature of the problem at hand, every plausible solution cannot avoid being contingent and open to revision.
II. The Ambivalent Relationship between Republicanism and Global Concerns
Republican freedom is understood as the absence, not of interference, but of domination — that is, of the capacity of other actors to interfere with an agent on an arbitrary basis.Footnote 6 Republicanism defines freedom as a form of structural independence — as the condition of not being subject to the arbitrary or uncontrolled power of another agent. Domination occurs, instead, when the power between two agents is deeply unbalanced, with the result that there are no effective constraints on its exercise: the dominating agent can act with impunity and without being accountable to anyone — first and foremost, not to the dominated agent. The republican literature offers a wide array of definitions of arbitrary power, but a clear unifying feature is that arbitrary power is not subject to appropriate checks — especially by those upon whom it is exercised. It is clear, then, that interference can occur without domination, and domination can occur without interference. A nondominated agent can experience a great deal of interference, but she is able to exercise relevant forms of control over the conditions of its exercise. For instance, if the citizens of a polity, upon consultation, overwhelmingly support minimum wage legislation, the decision will result in interference, but not in domination. By upholding a minimum wage, citizens sign up for a policy that interferes quite a lot with the behavior of private actors in the labor market, but (1) which is compatible with nondomination procedurally (because the citizens have exercised control over the relevant public rule) and (2) whose substantive rationale might be to protect certain actors (in this case, workers) from domination. Conversely, domination can occur without interference. A slaveholder may decide to interfere very little with what her slave does — either because she is benevolently inclined toward him or because she does not need to interfere with him to have him do what she wants, as the mere threat of interference is enough to discipline him and make him anticipate her wishes. In other words, being dominated means being subject to the will of another agent — but what is meant by the agent’s will is the capacity of said agent, not her intention.
Recently, republicans have been arguing that nondomination can ground both a distinctive understanding of freedom, and a distinctive conception of justice. On a republican reading, justice requires either the elimination or (more pessimistically) the minimizationFootnote 7 of domination as a matter of right. On this view, a just society is, first and foremost, not one that distributes goods in a certain way, but one in which everybody enjoys nondomination. The idea of grounding a conception of justice in the ideal of nondomination is based on the republican conviction that domination is both particularly pernicious and particularly insidious. It is pernicious because, when it occurs, it undermines the most fundamental component of free agency. This is the normative rationale for protecting said freedom with justice-based entitlements. It is insidious because domination is not only particularly important, but also particularly hard to avoid. A theme that runs through the republican tradition is that power has a natural tendency to degenerate and become arbitrary and abusive. Hence, domination is a particularly pressing and ongoing danger, and this gives us further (instrumental and rhetorical) reasons for branding the fight against it in terms of justice. Republican theories of justice, unlike liberal and especially liberal egalitarian ones, are primarily concerned with political arrangements rather than distributive patterns; justice as nondomination is typically understood as requiring, in particular, relevant forms of democratic control and certain legal guarantees and protections. Several republicans, however, argue that nondomination also requires certain egalitarian socioeconomic arrangements, since excessive need or inequality increases the exposure of an agent to the possibility of arbitrary powerFootnote 8 and because a society of mutually non-dominated agents is a society of social equals.Footnote 9
Can republicanism, however, offer an account of global justice as well? Until fairly recently, most republicans would have answered this question negatively. Republicanism was mainly considered to be a theory of bounded citizenship. Indeed, if asked to compare the inclinations of liberalism and republicanism toward global politics, one probably would have said that liberalism is the more universal, but also thinner, approach, whereas republicanism combines a more demanding understanding of freedom with a more particularistic take on the scope of justice and democracy. Liberalism is traditionally construed as the universalistic theory par excellence, grounded as it is in the idea that the right to freedom is something that individuals have in virtue of their status as persons. However, classical liberalism traditionally insists not on what political institutions should do, positively, to promote freedom, but rather on how the scope of their agency should be limited in its name. For republicans, things have traditionally been the other way around. On the one hand, republicans support a thicker understanding of freedom. Freedom is not about being “left alone,” but about not being dominated — that is, about being shielded from arbitrary power. Arbitrary power, in turn, is often made possible by an institutional vacuum, and fixed by (non-arbitrary) institutional regulation. For instance, an entirely unregulated economy can arguably lead to opportunities for domination (in the form of exploitation, power and bargaining power imbalances, monopolies and cartels, and so on) much more than an aptly regulated one. Additionally, republicans typically hold that one crucial way in which citizens can be protected from domination is by partaking in political power — power is not arbitrary, and therefore its interference is not freedom-reducing, when those on whom it is imposed control the form and substance of its exercise.Footnote 10 Finally, republican freedom is often considered to be incompatible with excessive levels of inequality: if domination is to be avoided, “no citizen shall ever be wealthy enough to buy another, and none poor enough to be forced to sell himself.”Footnote 11 Republicans, in other words, both endorse a richer account of the sphere of action of public institutions, and see a tighter, conceptual connection between freedom, democracy, and (at least some forms of) socioeconomic equality. On the other hand, however, the republican tradition has always insisted on its civic (indeed, sometimes even local and municipal, rather than national) dimension, by stressing the importance of civic virtues and active citizenship within bounded and cohesive political communities as means to secure nondomination. For nondomination to occur, political power must be genuinely controllable by those upon whom it is exercised, and this is more likely to happen if subsidiarity is ensured as much as possible.Footnote 12 Thus, republican freedom is typically described as being the contingent achievement of a people through its unique historical trajectory and through the ongoing exercise of civic virtues.
This picture, however, has been put into question by fairly recent developments in both traditions of thought. Contemporary liberalism from Rawls onward,Footnote 13 has become increasingly preoccupied with securing a system of equal liberty for all, and one where specific attention is devoted to the political and socioeconomic preconditions of freedom. Liberal egalitarianism, quite simply, requires a big state. Additionally, several liberals have increasingly stressed the contextualist and associative aspects of the liberal project. Liberals have thus argued, for instance, that liberalism presupposes the existence of a shared public political culture.Footnote 14 Moreover, as noted earlier, many liberals have argued that egalitarian liberal obligations only apply within the borders of the state because only the unique relations among fellow-citizens trigger demands of social justice to begin with. In sum, contemporary liberalism has become both more demanding and more contextualist. Conversely, republicans have started to argue that the ideal of nondomination is both a more universal political value than liberal equality, and a more modest one. As Lovett puts it, republican injustice occurs when domination is present, period — whether or not those who dominate and those who are dominated are members of the same polity.Footnote 15 Although remedying such an injustice requires (possibly new) institutions, domination can occur outside and across clearly defined institutional boundaries.Footnote 16 On the other hand, it has been argued that republicanism can be a more modest normative standard for global politics than liberal egalitarianism, in that achieving nondomination beyond borders might require more-modest distributive measures than global equalityFootnote 17 or measures other than distributive ones altogether.Footnote 18 Finally, republicanism can be construed as a less controversial ideal for global justice: Who would deny that people ought not to be dominated, regardless of where they live?
For several contemporary republicans, indeed, nondomination is particularly well equipped both to identify what exactly is morally problematic about the current global order and to provide sound institutional recommendations to fix the problem. With respect to the former, republicans have argued that the concept of domination best captures the phenomenology of what is normatively troubling about global political circumstances: global economic interdependence without appropriate institutions, several republicans have argued, is a problem because it creates new opportunities for domination.Footnote 19 With respect to institutional matters, the republican outlook is said to offer better insights as to what a morally justifiable global institutional order might look like. This is the case, arguably, because republican theory focuses from the outset on political power — what it is, how it can be controlled by those who are subject to it, and how it can become arbitrary — and is therefore more sensitive to questions of institutional design. And yet, republicans sharply disagree about what such institutions should be, advocating everything from a global republic to fairly statist solutions. The ambivalence analyzed in this section can perhaps already give us a clue about why this is the case; the next section delves deeper into the issue.
III. The Tensions Within Republicanism
The shift in the debate illustrated in the previous section is not without reasons. Indeed, one can detect a cluster of tensions within the republican approach when it comes to securing nondomination via institutional means — tensions that, I shall argue, can be managed, but whose nature needs to be unpacked first. In this section I explore these tensions in their general form, using mainly examples from the domestic context.
Republicanism is a demand for the elimination or minimization of domination. This entails analyzing what the potential sources of domination might be. I submit that there are three particularly prominent dynamics through which domination may occur — and, as a result, three desiderata that republican institutions must ideally try to meet jointly. Furthermore, privileging one of these desiderata often comes at the more or less direct cost of the others; in other words, these desiderata do not always pull in the same direction, and republican institutional design should therefore be, as much as possible, a matter of careful balancing.Footnote 20 The three desiderata are the following: republican institutions should A) resist the excessive concentration of power; B) bring informal power under rule-governed control; and C) further an active, vigilant citizenry. The first two desiderata address two different (and to some extent opposite) potential sources of domination. The third addresses the omnipresent danger of deterioration of republican institutions into dominating ones. I shall address these in turn in the following subsections.
A. Resisting excessive concentration of power
The most obvious way in which power can become dominating is when there is too much of it. Excessive power means that those who hold it can act with impunity and that those who do not hold it are powerless. The first, and most paradigmatic, case of domination in terms of excessively concentrated power is that of an unduly powerful sovereign. If the authority of the sovereign is not subject to controls, it is dominating, because it can in principle interfere with those over whom it has sovereign power with impunity and without accountability. The paradigmatic nemeses of republican freedom are therefore monarchical or tyrannical regimes, and the demand for nondomination is first and foremost the demand for the establishment of a republic. The republican sovereign is one whose power is dispersed and fragmented, and ultimately lies in the hands of those who are subject to it. Preventing domination means preventing too much power from concentrating in one source, and ensuring instead that the power of different institutional and individual actors is mutually counterbalanced. This, according to traditional republican institutional strategies, can be achieved through a variety of means, of which three are particularly prominent. First, democratic accountability ensures that those in power are accountable to those over whom power is exercised. Republicans are therefore almost unilaterally democrats, although (as we shall see below) they diverge quite widely in how demanding and meaningful their conception of democracy is.Footnote 21 Second, the separation of sovereign powers into mutually independent judiciary, legislative, and executive branches ensures that the sovereign is not a monolithic actor with a set of perfectly aligned interests, but that different branches counterbalance each other. Otherwise, even a democratic sovereign can become tyrannical. Third, devolution and federalism ensure subsidiarity, that is, that decisions be taken at the lowest possible level that is compatible with a competent, effective, and fair solution. Devolution additionally ensures that the highest level of authority, the one that is both potentially most threatening and farthest away from direct democratic control, is suitably limited in its powers. Not all republicans are jointly or equally committed to a strong conception of democracy, the separation of powers, and devolution — and indeed, it has been argued every now and again that these three institutional qualities are somewhat in tension with each other.Footnote 22 However, all three are ultimately motivated by the republican desideratum to counter the excessive concentration of power.
B. Bringing informal power under rule-governed control
So far we have mainly focused on the threats to nondomination posed by excessive institutional power. However, and somewhat paradoxically at first, republicanism does not have the same instinctive skepticism toward public institutions that characterizes classical liberalism. On the contrary, as we have already seen, nondomination can only be secured through institutional means, for it is a demand for robust and guaranteed protections. In other words: whereas excessively concentrated institutional power is certainly a classical source of domination, so is informal asymmetrical power, because it is unaccountable. Indeed, at closer inspection, there is a certain conceptual analogy between the two. Informal power is a problem for republicans because, when an actor has asymmetrical power over another, not because of an official role she covers, but because of her social or economic superiority, such power is not governed by rules, is not controlled by those upon whom it is exercised, and cannot be held accountable in a proper manner. But so is excessive institutional power — when the sovereign is absolute, she is above the law. And the point of the republican project is that nobody should be above the law — not the sovereign, not the large corporation, not the church leader, not the head of the family. Thus, the second desideratum that republican institutions must strive to meet is to bring informal power under rule-governed control. This explains why republicans are less hostile to state regulation than are classical liberals. An example that republicans often give is that market regulation might be called for in order to prevent monopolies or asymmetrical bargaining situations from arising. Another might be regulation of family life to prevent some family members from exercising arbitrary informal powers over others (typically, but not only, men over women and children).
Now, whereas it is quite clear that republicans have reasons to worry both about excessively concentrated institutional power and informal power, tackling both might generate some tensions. The kind of public power that can reliably and effectively bring informal power under control must be fairly strong indeed. Strong, powerful institutions are necessary to effectively gain such control. Whereas a strong state is not logically incompatible with robust systems of checks and balances, ongoing democratic control, and the separation of powers, tensions become more likely. Fairly concentrated public power establishes effective control more easily than informal power does; but fairly concentrated power is also more likely to exercise its own prerogatives in an arbitrary way. An excessively strong state is also one that is most likely to develop a self-serving bureaucracy, lose accountability, and ultimately issue arbitrary forms of regulation as a resultFootnote 23 — the machinery of the state apparatus of countries in the Eastern Bloc (independently of what one thinks of the ideology they embodied) is arguably the clearest example in this regard.
Conversely, dispersing power across different bodies is a traditional republican strategy to meet the first desideratum. This, however, might generate uncertainty in allocating responsibility to solve specific disputes — and indeed even encourage institutions to escape their responsibilities by each claiming that the others have competence over the relevant domain, thus leading to never-ending disputes.Footnote 24 To put it bluntly, constitutional crises and uncertainties generate institutional vacuums where informal power, rather than rule-governed power, is likely to take the upper hand. This problem can, in principle, be resolved by having clear rules regarding who is the competent authority to settle what disputes, without necessarily leading to excessive concentration of power at one level. However, fragmented power remains by and large more likely to generate these uncertainties. Thus, the tension between the first and the second desideratum might not be devastating, but it can hardly be denied.
One final example of how tensions might arise is constituted by judicial review. Fundamental rights susceptible to judicial review are one crucial way of diminishing the vulnerability of agents who are at the lower end of relationships of asymmetrical informal power because, for example, they belong to a vulnerable social class, gender, ethnic group, or religious association. Constitutionally entrenched rights can protect the powerless against the powerful. Indeed, there is a strong tradition of constitutional thinking within republicanism. However, constitutional entrenchment constrains democratic processes, and therefore constrains the capacity of the people to control the means through which power is exercised over them. Secondly, while constitutional entrenchment is often defended as a way of protecting vulnerable minorities against the will of the majority, strong mechanisms of constitutional review arguably encourage the accumulation of power in the hands of an unelected and possibly unaccountable judiciary. Some republicans therefore object to constitutional review on the grounds that it gives arbitrary law-making power to courts.Footnote 25
In sum, the desideratum of preventing an excessive concentration of power and that of bringing informal power under institutional control pull in different directions, and meeting both to a satisfactory degree is a matter of careful balancing in institutional design. Things are complicated further once the third desideratum is introduced.
C. Furthering an active, vigilant citizenry
One of the chief pillars of republicanism, over and above its institutional underpinnings, is the idea of an active citizenry — one that is willing to cultivate civic virtues by way of participation in decision making and/or the monitoring of political power. For republicanism, even the best institutions will degenerate without an active and vigilant citizenry to uphold them. Civic engagement can take the form of enthusiastic participation or of monitoring, contestation, and vigilance — both are necessary to keep republican institutions alive. This is the case for two reasons. First, one key way to ensure that institutions do not exercise arbitrary power is by making sure that they are, as much as possible, the product of the authorship of those to whom they apply. In other words, and as we have already seen, republicanism has strong reasons to support democracy. In turn, if democracy must be meaningful in this way (if it must be a genuine act of authorship, and not just a symbolic ritual), it requires an active, committed, vibrant citizenry. The second reason points to a further distinctive feature of the republican tradition. As we have already seen, a theme that runs through the republican tradition is that power has a natural tendency to degenerate and become arbitrary and abusive. There is, therefore, no institutional crystallization that can guarantee nondomination once and for all. Nondomination is an ongoing struggle, a fragile achievement that must always be monitored and nurtured. Without an active, vigilant citizenry, even the best institutions will rot by republican standards. Therefore, institutional barriers or political disaffection should not stand in the way of an active civic culture of participation and/or vigilance.
Republican institutions, therefore, should be designed with an eye to being conducive to such civic culture. For many republicans, this simply means furthering the democratic nature of the polity; but even less radically democratic versions of republicanism must be committed to the idea that, in order not to degenerate into arbitrary ones, institutions need to be held in check. Even if citizens are not meaningful authors of democratic rules, they must be active and vigilant controllers for a regime to be meaningfully republican. This is what Pettit means when he says that citizens should be scrupulous editors of rules and laws even if they are not authors proper.Footnote 26 Even if we live in a republic à la Madison, where the emphasis is less on power being held by the people and more on it being appropriately fragmented and subject to checks and balances, the editorial activity must remain regular and vibrant — and for that we need an active citizenry.
Again, this third desideratum might generate tensions when considered jointly with the other two. A strong state, capable of avoiding informal power, might end up leaving little space for active citizen engagement. The latter requires an openness to change and revision which might go against the very reasons why we want to do the former (namely, secure protection against informal arbitrary power). If regulations are too easily open to change on the basis of citizen activism and democratic will, it is not clear that they will deliver on the standard that justifies them to begin with (indeed, this is one of the reasons why Pettit prefers editorial activity, and considers the authorial activity of law-making by the citizens too volatile). In other words, getting the regulation right has a tendency toward “expertocracy” which goes against the third desideratum. A fragmented and multi-leveled polity, in turn, might also fragment our allegiances, in a way that makes it difficult for us to decide at what level we wish to channel our political activity (a problem which, as we will see, becomes particularly salient at the supranational level), and which might encourage apathy. Finally, careful institutional design might succeed in making a state strong enough where it needs to be strong, but sufficiently protected against excessive concentrations of power, thus meeting the first two desiderata. One such state, however, might be one that, again, needs the careful design of policy-makers and experts and is preferably not disturbed by ongoing and potentially volatile democratic contestation. It might be possible to design the perfect system that satisfies the first two desiderata — but civic participation and contestation must not get in the way of that, letting the experts do their work of getting the balance exactly right. However, if the third desideratum is necessary, then no institutional mechanism is impeccable over time without an active and vigilant citizenry.
In sum, the three desiderata of republican institutional design are equally motivated by a concern for domination, but might pull in different directions. The conflicts between them are real, and there is no straightforward roadmap as to how to balance them. The concentration of power can give rise to domination, yet might have its advantages. Regulating informal power is crucially important, but must be balanced against the danger of creating unaccountable regulatory bodies. An active citizenry is crucial to hold political power in check, but might be in tension with the careful and complex institutional design that the first two desiderata jointly seem to require. Balancing the three desiderata against each other is, thus, the predicament of republican institutional design.
IV. The Tension within Republicanism – The Global Case
This section explores how the tension between the three republican desiderata reappears at the level of global institutional design, generating specific challenges. In what follows, I shall first provide a brief overview of existing republican models of global institutional design (IV.A) and then illustrate how they all fail to give proper consideration to all three republican desiderata (IV.B).
A. Three models of republican global governance
As I noted at the beginning of this essay, republican authors widely disagree on the institutional form that a global republican order should take. All proposals are motivated by the conviction that cross-border domination is a pressing issue, but attempt to address it in very different ways. Existing proposals can be grouped under three broad categories: 1) statist models; 2) cosmopolitan models; and 3) transformative models.
1. Statist models. Statist models embrace Skinner’s slogan that one can only be a free person in a free state.Footnote 27 In order to be free from domination, individuals must live in polities that are not themselves dominated. The best way to achieve the goal of nondomination globally is to strive toward a world order of mutually nondominating states. As we have seen, these statist models are institutionally deeply anti-cosmopolitan without necessarily rejecting the claim that nondomination is global in scope. They do not deny that every moral agent is entitled to nondomination, but that nondomination is best realized through a cosmopolitan order.
Statist solutions, in turn, can come in strong and weak varieties.Footnote 28 Strong statism defends a system of both hard sovereignty in politico-legal matters and autarchy in socioeconomic matters. Hard sovereignty ensures that polities are robustly protected from external interference, while autarchy ensures that polities are socioeconomically self-sufficient and thus protected from the risk of asymmetrical informal power in international market relations. No republican, however, explicitly endorses strong statism.Footnote 29 Moreover, both economic autarchy and hard sovereignty seem unfeasible under the conditions of significant global interdependence that we currently live in. Finally, even if polities did manage to achieve hard sovereignty and economic autarchy, the complete absence of an international scrutiny over the internal activity of states seems incompatible with the republican ideal. Skinner might be right that one needs a nondominated polity to be a nondominated person, but the reverse certainly is not true: externally nondominated polities can certainly become internally dominating — and a republican must plausibly be committed to establishing at least weak forms of controls against that, such as a regime of international human rights and international criminal and humanitarian law. I shall therefore leave this model aside.
Weak statism, instead, assigns a central role to the ideal of the free state in a republican world order, but also endorses a weak global regulatory framework to ensure the mutual nondomination of polities and allows for a more qualified and conditional account of state sovereignty, so as to protect citizens (via, for instance, a system of international criminal law and international human rights law) from abuses committed by their own states. In other words, while states remain the central features of a republican global order, global quasi-constitutional arrangements must be in place.Footnote 30 A republican order is a world of free states, but one where freedom is understood in a republican sense — that is, as compatible with some interference. The most prominent example of weak statism is Pettit’s Republican Law of Peoples.Footnote 31
2. Cosmopolitan models. Cosmopolitian models start from opposite assumptions, and argue that, since the current global order is one where opportunities for cross-border domination are abundant, strong global republican institutions must be established to counter them. There is only one solution to global domination, and that solution is called global republic Footnote 32 — for there “cannot be freedom as nondomination outside a legal framework that gives security to individuals.”Footnote 33 If nondomination, domestically, needs a carefully designed system of republican institutions to robustly guarantee the absence of alien control, it is simply not clear, so republican cosmopolitans argue, why things should be any different when global domination is at stake. In a globalized world, individuals can be dominated, not only by their own state and fellow citizens, but also by states other than their own, citizens of different states, and transnational non-state actors such as large corporations. The only way of countering that is by establishing a robust global rule of law and final mechanisms to settle transnational disputes. This can be achieved through a largely decentralized and federal global republic — but a global republic nevertheless.
3. Transformative models. Finally, transformative models are based on the idea that global domination indeed needs a fairly deep institutionalization of the global order, but not necessarily one that replicates the model of the state at world level. What we need is, instead, a complex system of multi-level transnational governance where global issues are dealt with globally, transnational ones transnationally, national ones nationally, and local ones locally. Two transformative models are particularly salient — I shall call them “static” and “dynamic” respectively.
The most paradigmatic example of the static model is Pogge’s proposal of a dispersal of sovereignty over vertical lines.Footnote 34 Under this model, the regulation of each policy area is assigned to its most appropriate level (local, national, regional, global, transnational), with no level of authority playing a central role. Within such a model, nation-states as we know them would cease to exist, but they would not be replaced by a world state: sovereignty would be dispersed across different levels with no privileged locus of concentration. The “dynamic” model proposes something fairly similar, but in a “liquid” rather than crystallized fashion.Footnote 35 What does this mean? It means, in a nutshell, that sovereignty is not only vertically dispersed, but that, additionally, no sovereign competence is assigned once and for all to a specific level. There is no prior and fixed decision about what the right constituency for a given issue is. Transnational authorities, instead, represent, and are accountable to, a multiplicity of constituencies, which are labeled as “functional,” “deterritorialized,” “transnational,” and “overlapping.”Footnote 36 The characterizations “functional” and “deterritorialized” indicate that decisional authority ought to be flexible depending on the issue at stake, and is thus not set by clear territorial boundaries (whether vertically dispersed or not). Every issue has its own “demos,” or population that is affected by it, and the issue itself indicates what the demos should be. “Transnational” and “overlapping” mean, instead, that the relevant constituencies for different decisions will cut across borders in complex ways, because such is the nature of global domination. In other words, there will be in-built flexibility, rather than the neat system of nested constituencies envisaged by the static model. This regime might also be called transnational governance. It is transnational because the relevant constituency that should inform decision making on a given matter might cut across borders, and might change depending on the issue. It is a form of governance because a model of this kind cannot possibly have the rigid institutional channels and chains of command that government proper requires. It will instead be based on flexible networks of individuals, states, stakeholders, and representatives of civil society who come together and attempt to find common solutions – like all processes of governance do. Collective outcomes would be likely produced through the fluid modes of governance rather than through the legal and political instruments of government. In both systems, however, sovereignty would be fragmented and dispersed, rather than concentrated in either a state or in a global republic.Footnote 37
B. The three desiderata ride again
How can republican institutional blueprints of global institutional design be so widely different? They can, I suggest, because the three broad models described in the previous subsection fail to give due consideration to all three republican desiderata. Whereas it might not be possible to satisfy all three desiderata equally well, and whereas one source of domination might be particularly pressing and urgent at a given juncture, the three models above fail to even consider the possible tension between the three desiderata with due care. This is particularly salient because the three desiderata pull in different directions at the global level as well. What is more, they do not converge on a verdict over which form (cosmopolitan, state-based, or something else altogether) the global order should take. Let us take a closer look at why this is the case.
Republicanism is concerned, on the one hand, with informal power inequalities; it therefore has reason to worry about forms of power that are not yet bound by juridical relationships and institutional regulation, such as, arguably, those that take place within a world characterized by high levels of interdependence but low or patchy levels of institutionalization. Most republicans concerned with global issues are convinced that global economic integration poses specific challenges in this respect. The example is often given that, when a powerful transnational corporation manages to impose certain regulatory standards (such as corporate taxation, poor labor standards, thin health and safety regulations, and so on) in a host country with weak regulatory capacity by threatening to leave otherwise, said country and its citizens are arguably dominated.Footnote 38 The corporation is not accountable to any institutional agent in a satisfactory manner, and the host country has no control over the forms through which power is exercised over it. Indeed, what happens in most such cases is a pattern of behavior that seems to come straight from a republican handbook: countries with weak regulatory capacity and a fragile economy tend to adjust their labor regulations in ways designed to avoid conflict with powerful non-state actors (most notably, transnational corporations) on which a large part of their economy depends. They anticipate the preferences of the dominator, without actual interference being necessary. Thanks to poor or uneven regulation of the global order, some actors come to hold informal power that is dominating in nature. Therefore, at the global level, the desideratum to bring informal power under rule-governed control grounds not only an argument in favor of thicker, more extensive regulations where institutions are already in place, but also of new institutions able to regulate actors currently operating in a rule-free zone.Footnote 39 In other words, the desideratum to regulate informal transnational power grounds reasons to establish forms of supranational authority when these are not yet in place. In short, the second desideratum pushes in the direction of (at least some) institutional cosmopolitanism.
On the other hand, however, republicans have traditionally been skeptical toward cosmopolitan institutional agendas, for the dangers of global despotism that these might generate.Footnote 40 There are nontrivial reasons why republicans endorse a civic, rather than cosmopolitan, understanding of citizenship, which are largely, if not entirely, grounded in the first desideratum. If informal power should be subjected to rules, we should be careful not to establish such rules by creating forms of institutional power that are excessively concentrated — for these too are dominating. As we have seen in the previous section, democracy, the separation of powers, and systems of checks and balances are key instruments through which this aim is approximated at the domestic level. It matters to republicans to keep political power as close as possible to where its authority is relevant and to those who are most entitled to control it. Arguments made time and again by democratic theorists of all schools of thought about the IMF, the WTO, the World Bank and, most notably, recent developments in the EU show that supranational governance is very exposed to the risk of technocratic distortions and poor accountability. It could be that some supranational governance is necessary for nondomination’s own sake; however, the republican principle of subsidiarity encourages us to transfer only what needs to be transferred at all costs. Finally, within a cosmopolis, active citizenship and the ongoing exercise of civic virtues and institutional monitoring would be very hard indeed to achieve. In sum, both the first and the third desideratum pull against the model of a global republic.
This tension could perhaps be mitigated by devising a complex multi-level model, where neither states nor a global republic would be the primary unit of political authority, but where different issues would be addressed at the level that is most compatible with the minimization of domination. This model would be based on a republican principle of subsidiarity of sorts, according to which political decision-making authority should not be transferred to the next level up unless doing so is necessary — but should be so transferred if necessary. In other words, the transformative model might offer a way to reconcile the first and the second desideratum. I would like to suggest, however, that this goal can plausibly only be achieved at the cost of sacrificing the third. Both the static and the dynamic transformative models would require fragmenting sovereign authority — and therefore the need to control it — across different levels. One reason why republicans have historically advocated forms of civic patriotism, however, is that civic engagement is strongly facilitated both by processes of identification with a polity (be it a state or city) and by the clustering of most (if not all) salient political issues at one institutional level — as a way to channel both stakes and solidarity. Now, both static and dynamic transformative models are, admittedly, explicitly devised with the aim of allowing for the control of informal power while not diminishing democratic accountability. It is far from clear, however, that political participation could reach the necessary level of robustness in such a multipolar world. People might simply disengage out of sheer exhaustion (too many demands of participation from too many sides, none of which seems to matter quite enough), and institutions might become technocratic, unaccountable, and ultimately dominating as a result. This is largely an empirical matter, but EU elections and EU public culture more generally are a fairly robust indicator of the fact that it is difficult for citizens to sustain a genuine democratic spirit of activism and vigilance across different levels, each with partial competence. Indeed, some would argue that the risk is to demotivate interest at the national level, as well, because citizens might start thinking that the “real” decisions are being made elsewhere — or indeed because, with the exception of a highly intellectual elite, it is hard to figure out where exactly each important class of decisions is being made to begin with. Again, EU member states seem a relevant example here. In sum, there is a concrete danger that multi-level systems might generate apathy, political disaffection, and ultimately distrust. Of course, issue-based activism is possible and vigorously on the rise in recent years, and multi-level systems could arguably be more compatible with that sort of activism. The dynamic transformative model in particular (but not only) — with its emphasis on constituencies which might change according to the issue at stake — might be able to capture precisely this kind of activism. However, issue-based activism tends both to be fairly minoritarian and fairly volatile (with some notable exceptions). These forms of activism tend to rise and fall in fairly steep curves, and might therefore lack the sustained and widely spread form of engagement and vigilance that republicans aims to achieve through active citizenship and civic engagement. In a nutshell, a world of fragmented political levels might well end up being a world of fragmented, and fairly thin, political allegiances — and this is quite hard to square with the third desideratum.
Finally, it is worth noting that we do not even know whether power could be sufficiently effective in such a fragmented system. The dynamic model, in particular, poses specific challenges in this respect, for it is not even clear that informal power could be sufficiently tackled within it. A multi-level system where the identification of the right level is constantly open to negotiation clearly has the virtue of flexibility, but might also be easily manipulated by those who hold high levels of informal social and economic power. Thus, it could be a system lacking sufficient reliability to meet the second desideratum as well.Footnote 41
We can now see why the three republican desiderata pose specific challenges when it comes to global institutional design. At the global level, they do not only raise questions of institutional form, but also fundamental questions of institutional level.Footnote 42 Should states relinquish their sovereignty to allow for a better regulation of global affairs?43 The second desideratum pushes in that direction, because it demands that currently informal power be brought under institutional control. The first and third pull away from it. The first does so because it envisages global despotism as the most dangerous form of concentration of power — indeed, a concentration of power that would leave us with nowhere else to go. The third does so because a global republic is hard to square with active civic engagement and control. The statist model, in turn, does not seem sufficiently sensitive to the danger of informal international and transnational power. Both transformative models, finally, seem hard to square with a culture of sustained civic participation and vigilance, and might prove ineffective in bringing informal asymmetrical power under robust and reliable control.
In sum, if it is true that all three desiderata matter, and that republicanism is a particularly promising theory of institutional design precisely because of the balanced approach which joint concern for all three desiderata generates, then all three republican models of institutional design — the statist, the cosmopolitan, and the transformative — fail to live up to such a promise in their current form.
V. A Tentative Solution: Thin But Hard Regulation of States
In this essay, I have mainly brought bad news to the republican project of global institutional design. Is it possible, however, to try to honor all three republican desiderata when devising a model of global governance? In this concluding section, I would like to suggest that we perhaps have reasons not to despair.
This is grounded in two sets of considerations. First of all, the essay has only argued that the three models of republican global institutional design, in their current formulations, do not exhibit due consideration for all three desiderata. It is still possible that, on balance, one of the three models will prevail after a proper balancing of all three desiderata. It could be, for instance, that, say, weak statism is, on balance, the best way of achieving sufficient protection against both excessive formal power and asymmetrical informal power while keeping an active citizenry. This case, however, needs to be made on the ground. My argument in this essay has been that the three desiderata matter, that the tensions between them are real, and that no republican institutional model can disregard them. In this sense, the essay is a conversation starter rather than a conversation stopper: it intends to identify the challenges that each of the three families of models needs to meet in order to be faithful to the republican ideal. One of them might still be the right one, but it will prove to be so by showing to provide the best balancing of the three desiderata, rather than by unilaterally focusing on one or two of them only.Footnote 44
Second of all, perhaps the three models do not exhaust our possibilities. Taking up this challenge, I would like to end the essay by quickly mentioning a very tentative alternative.Footnote 45 The solution is tentative in the basic sense that it is only a preliminary idea. It is, however, tentative also in a further and more interesting sense: all attempts to jointly meet the three republican desiderata cannot avoid being tentative and provisional. What strikes us as a good balance today might go out of balance in light of social, economic, and political changes further down the line.
I suggest that it would be wisest to start, like statist models, from states, and take seriously the idea that states are not only the institutions we are most familiar with, but also the only ones within which we have seen some real-world approximations of the ideal of nondomination. They are the institutions within which we have negotiated compromises and balances between the three republican desiderata. Relinquishing them would therefore be too risky. We should then ask, in an incremental spirit, how much the current state system can be improved so as to cater to the second desideratum (which would pull toward state elimination) without neglecting the first and the third. It is possible, for instance, to take seriously the idea that international and transnational informal power undermines the reasons why we want states to begin with. Skinner highlights how, from its very origins, the republican tradition put the emphasis, not only on the freedom of citizens within a law-bound regime, but also on the importance of living in a polity which is itself “capable of acting according to its own will — that is, according to the general will of its citizens — as a result of not living in dependence on the will of anyone other than the citizen-body as a whole.”Footnote 46 We want states because we want to achieve republican self-government; however, for states even to try to achieve this goal, they must themselves not be dominated. Citizens of a dominated state cannot sustain robust mutual relationships of nondomination, because their national institutions are more responsive to external dominating actors than to their own democratic control. As we have seen, transnational informal power can lead to the domination of states. If we sufficiently care about these cases — that is, if we sufficiently care about the second republican desideratum of bringing informal power under control — we can start from states, and yet seek robust regulation of these phenomena for states’ own sake. In other words, we can conceive of the task of the international order as that of protecting the nondomination of states — and while this clearly disallows the creation of a world state or a fully decentralized multi-level world order, it also calls for a fairly demanding regulation of global economic and political dynamics. Regulation can only occur to protect the nondomination of states, but we must do all that it takes in that respect; we must not only, for instance, establish a system of international law, but also regulate areas of the global market which allow certain actors to hold informal dominating power over (some or all ) states. Such an agenda entails thin, but hard, regulation.Footnote 47 The regulation is hard, because some areas are delegated to global authorities as a result; but it is thin, because it is both severely limited in scope and aimed at empowering states, rather than replacing them. It is, in particular, different from both cosmopolitan and transformative models because it is necessarily conducted in an incremental spirit: we 1) start from states but 2) ask what can be done to protect them from transnational forms of asymmetrical power (second desideratum), 3) while trying to limit global institutionalization to this aim only (first desideratum), thus 4) keeping a plurality of states within which a sufficient level of civic engagement can be sustained (third desideratum).
This might minimize, if not eliminate, domination — until new challenges arise. Republicanism is an ongoing project, one that we should constantly be vigilant about, rather than an objective we can take care of once and for all and then move on to more interesting tasks.
