Constitutional Demoicracy: An International Republican Regime of Human Rights

1 Introduction

Contemporary globalisation has created unprecedented opportunities for new global and transnational agents and dynamics to make human rights insecure. For instance, transnational corporations (TNCs) can nowadays move where they please to make profit. This has pushed states to lower taxes or labour standards so as to attract them (Dietsch and Rixen 2014). Consequently, human rights to decent work conditions and health of many have been put in peril. Likewise, anthropogenic climate change is jeopardising (through heat stress, drought, and flooding) the enjoyment of human rights to life, health, and subsistence (Caney 2009). Worryingly, the current state system and international law appear ill-equipped to address such new threats effectively (Buchanan 2013, 279–292). Thus, there is an urgent need to rethink the institutional shape that an appropriate human-rights regime – that is, one that protects human rights both securely and legitimately – should have today.

In this paper, I argue that the human-rights regime we need is a new and mixed one: a constitutional demoicracy. And I show that it is republicanism (Pettit 1997, 2012; Lovett 2010) that provides it – and that is a welcome result for, as I shall show, human rights are essential to the realisation of republican freedom. This human-rights regime has two defining features: it is a demoicracy and it is constitutional. It is a demoicracy for it is, much as the European Union, ‘a Union of peoples who govern together, but not as one’ (Nicolaïdis 2013, 351). That is, in this human-rights regime, sovereign states remain the primary locus where human rights are realised. Yet, they recognise that such entitlements are vulnerable to state power itself and to certain dynamics beyond their control. They thus form a union to govern proper on these issues – and on these issues only. Distinctively, however, the union does not replace the different demoi with a sole global demos, but rather acknowledges their status as distinct demoi, and leaves control over the union in the hands of each state.

This human-rights regime is then constitutional for, to govern together, states establish a very thin, federal, international apparatus, which possesses, to a greater or lesser extent, some features of domestic constitutions (entrenchment and supremacy) and pursues the defining aims of domestic constitutions when human rights are at stake – which are, in effect, to constrain state power and to empower agents to pursue projects that they cannot achieve on their own (Waldron 2009, 270–275). The demoicratic component delivers legitimacy in the matter of human-rights protection and the constitutional component security. And even if they sometimes pull in different directions, the regime minimises well this tension, as we shall see. Importantly, the power of the ensuing international apparatus is constrained in such a way that it is hard for it to dominate states, their demoi, and their citizens, let alone give rise to a world state.

This article contributes, first, to the literature on global justice and institutional design. It introduces a novel institutional entity, which straddles a middle way between internationalism and cosmopolitanism, and improves both. States remain central within it, yet undergo more radical reforms than those commonly recommended by internationalists (e.g., Rawls 1999; Pettit 2014: chap. 6), which however fall short of a full-fledged cosmopolitan agenda: global institution-building, but not a global republic (see Martí 2010); a concern for individuals’ basic socio-economic entitlements, without endorsing egalitarian principles of global distributive justice (see Moellendorf 2002); and supranational democratisation, which eschews the superimposition of a global demos (see Archibugi 2008; Held 2010), thus better respecting state self-determination. In so doing, constitutional demoicracy offers guidance on how to address a wide array of global problems impacting on human rights, thus enriching and connecting diverse subdebates within the global justice literature. And as a republican human-rights regime, it foregrounds the power relations generating them, thus complementing the dominant distributive approaches.^[1] As such, constitutional democracy is akin to Laborde’s and Ronzoni’s (2016) progressive republican internationalism, yet improves it by clarifying that the needed reforms ought to be undertaken for the sake of human rights and within a constitutional framework. This article also advances the literature on demoicracy, by showing the relevance of the demoicratic ideal beyond the European Union (see also Besson 2009) and defending its federal, rather than intergovernmental (Bellamy 2019), institutional implications. It finally contributes to the literature on global constitutionalism by broadening its agenda of reforms beyond the development of international law (e.g., Peters 2009).

The article unfolds as follows. Section 2 outlines a republican view of human rights and argues that an appropriate human-rights regime must protect human rights both securely and in a non-arbitrary way. Section 3 fine-tunes the security requirement by briefly illustrating the main threats to human rights today. Section 4 discusses one essential feature of an appropriate human-rights regime: sovereignty. Section 5 and 6 advance the ideal of constitutional demoicracy and show that it meets well both the security and non-arbitrariness requirements, notably when compared to other ideals. Section 7 concludes.

2 Republican Human Rights

Two theses define republican freedom or freedom as non-domination. Both shape the republican view of human rights and make it a distinctive one. The first one is that you can be free even if you are subject to extensive interferences. What matters is that they are non-arbitrary. Republicans disagree on how to spell out this requirement (Lovett 2010, 111–123), but the core idea is that interferences are non-arbitrary if they are checked and controlled. And this is best realised when individuals have a say in how the power that may interfere with them is wielded.^[2] The second thesis is that it is others’ capacity to interfere, as such, that should be constrained for individuals to be free. In other terms, individuals are free only if they have their options open not just in the actual social world, but also in all the social worlds accessible from the actual one (Pettit 2012, 66–67). This makes non-domination a robust notion of freedom and republicanism an approach chiefly concerned with social structures, for it is social and legal norms that grant some the capacity to interfere at will (Lovett 2010, 43–47; Pettit 2012, 63).

Importantly here, republicans argue that individuals should necessarily enjoy non-domination over certain options to be free: namely, those designating their rights. That is, individuals cannot command respect and so of their standing among persons (Pettit 2012, 91, 98) if they cannot claim ‘You owe it to me!’ The gist of the argument is that if they lack this authoritative standing, all their requests will only be given a hearing if others kindly concede so or in virtue of some benign social rules (say, some legal norms that restrain others’ power but do not grant rights to people), which however track, and are accountable to, others’ practical will, not theirs. Furthermore, when rights are guaranteed, they capture and deliver what agents can do robustly or invariantly across changes in others’ preferences – if my right to, say, speak my mind is guaranteed, I remain free to do so, even if my fellows start thinking that I should not (Pala 2024; see also Layman 2021).

For republicans, however, the rights necessary for freedom are not preinstitutional or natural entitlements (see Griffin 2008), but rather rights protected by institutions and laws, which are controlled by individuals themselves. Specifically, they are the rights of free citizens. The thought here is that in a preinstitutional scenario rights would lack an authority entitled to specify their content in the name of all (omnilateral will), thus allowing the powerful to impose their unilateral or dominating view of them and to enforce it, implying that individuals would only enjoy some of their putative rights, and so precariously or in virtue of others’ goodwill rather than robustly. Free citizenship remedies these problems, thus making it possible to realise rights. Indeed, it provides individuals with an institutional context that permits the formation of a common view of rights and that is able to enforce the rights of all. Equally important, it grants individuals a say in this institutional apparatus (cf. Valentini 2012, 580–581). Unsurprisingly then, on the republican view you are free only if you enjoy free citizenship and the rights usually associated with it (Pettit 2012, 8). This allows me to discuss republican human rights.

For republicans, citizenship and the rights associated with it can be enjoyed either optimally, in a regime of optimal non-domination, or minimally, when basic non-domination – that is, the ability to set up a republic in the first place – is considered (Laborde 2010, 51). Optimal non-domination is the status of optimal citizenship. It is realised by republics that maximally protect the citizenry from the main vulnerabilities that may lead to domination. Its enjoyment requires the rule of law, so as to replace the arbitrium of persons, fair political liberties, so that individuals can exercise a significant amount of control over the power to which they are subject, and extensive resources and opportunities to protect individuals well from domination in the society at large (Laborde and Ronzoni 2016, 288).

Basic non-domination is, by contrast, a citizenship status which simply demands that individuals be able to set up a non-dominating republic. It just requires sufficient opportunities for political influence, so that, even if some have more influence than others, everyone retains adequate control over government. Likewise, it just requires sufficient socio-economic resources and opportunities so that no one depends on others’ arbitrary will to satisfy their basic needs (nutrition, basic health care, education), thereby addressing poverty and vast inequalities of wealth, while permitting inequalities (say, in income or in the workplace) above this threshold (Laborde 2010, 52–53). Once basic non-domination is guaranteed, individuals should be in the position to shape their society in a more egalitarian fashion, as required by optimal non-domination.

Now, if justice tells us what is owed to individuals, whereas legitimacy tells us whether a political institution is entitled to rule, we can infer the following. Optimal non-domination is a citizenship status realised by fully just societies. Its underfulfilment means that individuals live in a less than fully just society. Differently, basic non-domination is a citizenship status realised by minimally just societies. The failure to ensure basic non-domination is the denial of the minimal core of justice, which calls into question the very idea of legitimacy - namely, whether a polity has a claim to rule or not. If this is granted, then human rights turn out to be those claims whose fulfilment is necessary to realise basic non-domination. For human rights capture and deliver the normative function of basic non-domination – that is, when systematically violated or left unfulfilled, they call into question a society’s legitimacy and raise international concern. So if non-domination can only be realised when individuals are citizens whose citizenship rights are protected, realising basic non-domination will require that individuals’ human rights be acknowledged and protected.

Three comments are needed. First, as on the republican view human rights are the entitlements necessary for individuals to function, at a minimal level, as citizens, a list of republican human rights will overlap with current international human rights law in some respects and diverge in others. Much as the latter, it will include a human right to robust equality before the law, and a human right to citizenship. Yet, since, as said, a suitable civic status is needed to realise (human) rights, the human right to citizenship will be considered the most fundamental entitlement or, in Arendt’s phrase (1973, 275), the ‘right to have rights’. Accordingly, the republican view will demand that all have secure access to its object, whereas current international law grants states wide discretion on whether and whom to naturalise. Importantly, as I argue elsewhere (Pala, unpublished), this implies recognising that this entitlement grounds a duty to naturalise all those without citizenship – a duty that, surprisingly, falls primarily on the international community and only derivatively on states, and whose beneficiaries are not only the formally but also the de facto stateless, such as refugees.^[3] Such a list, then, will include a human right to a minimum of democratic participation, not just to political participation (see also Bohman 2005b; Ivison 2010).^[4] It will not include, instead, the demanding social human rights contained in several international documents, but just, in line with basic non-domination, human rights to sufficient socio-economic resources and opportunities.^[5]

Second, the republican view of human rights turns out to be political, for it shares with the standard political view (Beitz 2009) the idea that human rights are political constructs that presuppose institutions to be realised, and whose point is to constrain their conduct. As we shall see, however, republican human rights are not as closely tied to states as they currently are. Hence, the republican view overlaps with Pogge’s (2008) institutional view, which understands human rights as moral claims primarily addressed against social institutions – yet it insists that institutions bear human-rights obligations not just because they are capable of discharging them, but also because, as noted, they are entitled to specify and enforce them (Valentini 2012, 584–585).

This leads us to the third comment: republican human rights must be protected robustly, or securely, and in a legitimate or non-arbitrary way. Let me unpack these two requirements, for they help us outline the characteristics of an appropriate human-rights regime. That human rights must be protected robustly means three things. First, it means that individuals should have access to their object invariantly across several changes in the actual social world, including, notably, scenarios in which the duty-bearers fail to discharge their human-rights obligations. In a word, republican human rights should be reasonably secure in the face of non-compliance. Their enjoyment would otherwise be contingent on others’ benevolence or sheer luck. Following Shue’s terminology (1996, 171), they will therefore assign secondary default duties to step in when primary duty-bearers fail.^[6] Second, robust entitlements will require, much as non-domination, a structural protection. That is, the institutions we need must tackle the very structures that enable human-rights violations. Finally, robust entitlements tell us what features the duty-bearers should possess. These features are: (a) capacity: agents should be able to discharge the duties assigned to them; (b) definiteness: we should be able to identify the duty-bearers clearly, for undetermined duty-bearers will hardly deliver a robust human-rights protection; (c) security: the duty-bearers should be forced to discharge their duties; this should not depend on their goodwill. Add to this that, distinctively on the republican view, the duty-bearers should also protect human rights non-arbitrarily (d). And this means, we now know, that individuals should have a say in how they act. Failing this, they would not enjoy human rights as rights proper, but as nice concessions, or as the beneficial outcomes of some rules alien to them (Halldenius 2010). That said, let me fine-tune the robustness requirement by offering a concise taxonomy of the main contemporary sources of vulnerability for human rights.

3 Vulnerabilities

The first agent that may jeopardise human rights today is states. They can do so intentionally or negligently, thus violating them, or they might lack the capacity to discharge the corresponding duties (think of poor states), thus leaving them unfulfilled.^[7] Internally, they may violate human rights by enacting laws that explicitly deny them, by not enforcing the laws that protect them, or by adopting policies that will foreseeably endanger them (e.g., hospitals are privatised and the terms they should abide by do not include a minimum of free health services, thus making it hard for the poor to access healthcare). Externally, states may violate human rights through violent intervention or by employing economic sanctions that will foreseeably have deleterious consequences on people (e.g., the US embargo on Cuba). Whether internally or externally, state capacity to violate human rights is facilitated by the sense of impunity that they enjoy internationally – states know that international law lacks teeth.

A worrying and subtle way in which states endanger human rights, well detected by a republican perspective, is informal and does not require any action on the part of states. They may simply profit from their superior power and just stand by, with an eye on weak states, ready to intervene, and, possibly, to violate their citizens’ human rights, should these states displease them. Even if the intervention does not take place, human rights are insecure, for their enjoyment is a function of others’ invigilating power (Pettit 2010a, 77–78).

States may threaten human rights through international institutions too. For example, the WTO’s enactment of TRIPS, which has prevented states from producing cheaper versions of patented drugs for 20 years, has met the demands of wealthy states and TNCs. Yet, as global justice scholars have observed, the poorest have consequently seen their human rights to health go unmet, as neither they nor their states – especially if poor – can afford the more expensive original drugs. Hence, powerful states and international institutions qualify as human-rights violators (Pogge 2008, chap. 9; see also Wolff 2012, 100–108). Beyond this, the republican view underscores that this is not just a distributive injustice – some end up living in dire conditions. For republican human rights are violated not just when one lacks access to their object, but also when the institutions affecting them operate beyond their control. This is precisely the case within the WTO, where poor and weak states are often just ‘rule-takers’, vulnerable to the arbitrary will of the powerful (Laborde and Ronzoni 2016, 282; see also Lafont 2016b, 433–436).

As republican freedom has a structural quality, it also sheds light on the structures that make human rights insecure by creating the conditions for violating them or leaving them unfulfilled. Take the international state system. Within it, states should primarily protect the human rights of those within their territory, notably their citizens. Thus, those who live elsewhere and have no citizenship – formally or de facto – have their human rights largely unprotected. Of course, international law assigns states secondary duties to protect these individuals’ human rights, yet it poorly enforces them (Buchanan 2013, 118–119). Moreover, on the republican view it is particularly problematic that, as noted, international law does not bind states to provide citizenship to all those who do not possess it, for in this way the most fundamental demand of basic non-domination remains unfulfilled.

Furthermore, one of the rules underpinning the current state system is the principle of effectivity, which grants international recognition to any group exercising effective control over a territory, irrespective of how power is wielded (Buchanan 2013, 124). Global justice scholars draw attention to the fact that any such group will in this way enjoy not only immunity from external interference, but also the privileges to sell national resources and to borrow money in the country’s name – which, in turn, facilitate human-rights violations. As Pogge argues (2008, 120–121), they do so by incentivising coups and civil wars, entrenching the power of despots, and saddling future democratic governments with huge debts that undermine their ability to fulfil human rights. For Wenar (2016, chap. 11), they also affect peoples' right to own and use their country’s resources, entailed by their human right to self-determine (see, however, Armstrong 2017, chap. 6). Therefore, in sustaining an international order conferring these privileges, states and international institutions contribute to human-rights violations. In addition, on the republican view the principle of effectivity itself conflicts with republican human rights, as it officially condones living under despotic rule, which is in itself a human-rights violation, and contradicts the republican principle of legitimacy, which, recall, requires that all share control over political power.

Consider, then, the global market. For some (notably elites), it has increased opportunities to prosper, yet has also engendered significant harms, including severe inequalities within and among states (Christensen 2017), exploitation (Wollner and Risse 2014), and, relevantly here, greater human-rights insecurity. For republicans, two political dynamics, less emphasised by the dominant distributive approaches, warrant attention. First, the weak regulation of the global market has undermined state capacity to protect citizens’ human rights – e.g., poorly regulated markets may cause financial shocks, whose costs might lead states to cut social provisions needed to fulfil human rights. Second, the global market has increased the power of private agents relative to states. TNCs, in particular, have gained a poorly constrained capacity for mobility, allowing them to relocate where they please to make profit. As republicans straightforwardly point out, the mere threat – or even the possibility – of relocating unless taxes are lowered (Dietsch and Rixen 2014) or environmental and labour guarantees are weakened is often enough to induce such changes (Laborde and Ronzoni 2016, 280). The alternative, after all, is to lose economic competitiveness and bear the ensuing costs. Consequently, citizens see their human rights deteriorate.

Finally, there is anthropogenic climate change. There is extensive literature on how it affects human rights (e.g., Caney 2009; Bell 2011). Suffice it to note that, first, it may make it harder for states, notably the poor ones, to protect their citizens’ human rights – e.g., if drought results from climate change, states will face further difficulties to protect human rights. Second, it may produce so-called climate refugees (Lister 2014), who are exposed to other states’ arbitrary will to have their human rights redressed. Finally, republicanism shifts the prevalent focus in the literature from how to fairly distribute the burden of combating climate change to the equally urgent political issue that states whose citizens face the greatest threats to human rights from climate change often have little influence in international fora where collective action is decided. With this in mind, we should ask: what human-rights regime is needed to address the threats outlined, and to protect human rights both robustly and non-arbitrarily?

4 Sovereignty

Given the overlap between the present view and Pogge’s view, one option immediately comes to mind: his idea of a multilevel system of dispersed sovereignty (2008, chap.7) and its republican development (Bohman 2004, 2005a, 2005b). I now consider this option. I do not analyse it comprehensively though, but just use it to draw one important lesson dealing with sovereignty and its central role in realising republican human rights.

For transnationalists, the chief threat to human rights is state sovereignty, given the concentration of power it involves. This motivates their core proposal: its fragmentation and redistribution in a residual form to various conventional and non-conventional actors, disposed on a vertical axis (‘neighbourhood, town, county, province, state, region’; Pogge 2008, 184), or on a vertical and horizontal axis (organisations and transnational groups are included too; Bohman 2005b, 300). Each actor, then, will have some, but not any final, authority over human rights and will not occupy a dominant position, in contrast to traditional states.^[8] This should result in a decentralised, non-hierarchical system of multilevel governance that should protect human rights both securely and legitimately: securely because the system should promote a dynamic and synergistic cooperation among the different stakeholders each time involved; legitimately for each actor’s power will be counterbalanced by the power of other actors, and each will have an equal say in any decision that may impact on their human rights.

Transnationalists are right in underscoring the excessive and so dominating power that current states hold over human rights. I also agree that some of their sovereignty should be ceded, as I will argue. Yet the proposal of widely dispersing sovereignty is problematic. If no actor has final authority over the protection of human rights, but rather an ever-changing constellation of agents may claim competence in each case, constitutional conflicts may rise frequently (Bellamy 2019, 89). Delayed action or inaction might follow, and so human rights might not be enjoyed securely. Of course, the parties may find an agreement on a case-by-case basis (Pogge 2008, 185). Yet they either possess roughly equal power, and so reaching an agreement will be a function of their willingness to do so, thus leaving the protection of human rights dependent on it; or one party will be more powerful, and so might impose their view on others, thus dominating them. Moreover, recall that republican human rights must arise from an omnilateral, not unilateral, will. Yet again, without an authority forcing agents to form such an omnilateral view, there is a risk that, given real-world power imbalances, some may succeed in imposing their view on others. The ensuing human-rights protection may then not satisfy the non-arbitrariness condition.

Crucially, both risks can be averted by retaining what transnationalists reject: sovereignty. Sovereignty is the supreme authority exercised over a territory. To have it is to have a say that is final and binding (Philpott 2020). On the non-Kantian understanding I endorse, sovereignty is not necessarily absolute and indivisible: that is, it covers all domains and cannot be constrained or divided. It rather comes in degrees (S can be more or less sovereign), or it is a set of powers that can be limited (standardly by a constitution), or transferred, without thereby disappearing altogether (Pavel 2015, 20–23). Constitutional democracies and the European Union support this understanding. On the republican view, the value of sovereignty resides in it allowing collective self-determination, that is, the ability to govern oneself without others’ arbitrary interference. It does so by shielding collectives from external, alien wills; and internally, by unifying the diverse, competing, and partial views of collective questions into a common view and deciding ‘who has to do what’ to address them (competence-competence), thus solving coordination problems. Given this, transnationalists have failed to see that sovereignty is uniquely placed to address the problems of both constitutional conflicts and the unilateral views of human rights. It is thus able to deliver security in matter of human-rights protection and to satisfy the first necessary – but not sufficient – condition of legitimacy (omnilateralism). It should then have a place in an appropriate human-rights regime rather than be discarded.

There is a caveat, though. Sovereignty is worthy of protection only if it allows people to live on their own terms. Its value decreases the more collectives depart from this legitimacy threshold, for below it sovereignty is just an instrument of domination. Yet recall that what distinguishes a legitimate kind of self-determination from an illegitimate one is the protection of basic non-domination and human rights – which also include, as argued, the right to have a say in the omnilateral will that the sovereign enables and enforces. Hence, the task for an appropriate human-rights regime is to reflect and enforce the only conditional value of sovereignty.

This raises the further question of where sovereignty should be allocated. One hypothesis is to place it where it is needed the most: at the global level in a global republic, as argued by political cosmopolitanism (e.g., Martí 2010).^[9] Yet this proposal raises many worries.^[10] I shall mention two. First, sovereign states are the only loci where, in modern conditions, (basic) non-domination and human rights have been realised (Laborde and Ronzoni 2016, 285). Moreover, despite globalisation, many states still protect these them reasonably well from certain threats (Hirst et al. 2009, chap. 8). The withdrawal of state sovereignty and self-determination implied by the advent of a global republic would thus represent a loss of value. Second, as republicans have long argued, a global republic is exposed to the risk of global despotism, which would endanger everyone’s human rights.^[11]

If sound, these remarks suggest that an appropriate human-rights regime should remain centred on sovereign states as the fundamental unit where basic non-domination and human rights are realised, but that they must be reformed. A similar agenda has been advanced by Pettit (2010a, 2014, chap. 6), who, providing a republican version of the Rawlsian law of peoples (Rawls 1999), has argued that states should establish a voluntary supranational association, so as to foster interstate deliberations on how to counter cross-border domination and fora where the weak can band against the strong (see also Christiano 2010). As should be clear, however, this agenda is not sufficiently transformative. For the reflections on sovereignty above indicate that to protect human rights securely and in a non-arbitrary way, we need not just a voluntary association, but, rather, an international binding apparatus with some sovereign competences – which are however very limited and distinct from those of states. In short, a dualistic human-rights regime is needed – that is, one with an internationalist basis bolstered with some political cosmopolitanism. The rest of this paper expands on these claims by developing an ideal of constitutional demoicracy. I first analyse its demoicratic feature, and then its constitutional feature.

5 The Demoicratic Feature

‘Demoicracy’ is a term employed to describe and recommend how to improve the European Union, often by employing the republican language of non-domination. The idea is that the European Union is and should better realise ‘a Union of peoples who govern together, but not as one’ (Nicolaïdis 2013, 351).^[12] In detail, for demoicrats European states and their demoi face common problems (negative externalities, climate change, financial instability), which affect their capacity to self-determine democratically, yet each is unable to address them individually. They thus become duty-bound to govern on these issues. For demoicrats, this governing together should not take the shape of either an intergovernmental forum or the European macrostate advocated by federalists: the former relies on poorly accountable executives, whereas the latter would dominate the different European demoi by replacing them with a single macro-European demos, rather than recognising their democratic agency. A third way is thus needed. In institutional terms, this means, following Ronzoni (2017b), either a moralised intergovernmental union – that is, not just driven by the pursuit of mutual benefit – or a very thin federal union. In democratic terms, this middle ground is provided by a union whose citizens are not the European citizens unified in a macro-European demos, but the different peoples of Europe, which would therefore retain control and the constituent power – that is, the power to constitute a political order and to give oneself a constitution (Cheneval and Schimmelfennig 2013, 342).

This search for a third way resonates with the previous section’s conclusions – in fact, there is more than just a consonance. If we abstract from the European Union, we can see that a demoicracy is needed anytime three conditions conjunctively obtain: (i) agents are interdependent on critical issues affecting their capacity to self-determine democratically; (ii) to be handled effectively, these issues must be governed proper; (iii) the agents want to remain distinct rather than merge, and this claim is justified. My contention is that these conditions also hold internationally, and so all states and their demoi must establish a demoicratic union, which, in turn, meets well the demands of basic non-domination.

That condition (i) is met follows from Section 3: states may violate the human rights of citizens of other states or weaken other states’ capacity to protect human rights. Either way, they may impinge on the ability of other demoi to set up and maintain a republic, and thus on their capacity to self-determine democratically. Moreover, agents such as TNCs and structures like the global market endanger human rights everywhere, yet states and their demoi are incapable of addressing such threats each on their own. Likewise, addressing the plight of the stateless requires structural reforms, which in turn need cooperation and coordination among states. Condition (ii) is also satisfied: as argued, republican human rights are robust entitlements, and so the sources of vulnerability that threaten them should be governed proper, not left to the willingness of agents to address them or not. Finally, do the world’s demoi have a justified claim not to merge into one macro demos (iii)? Below, I answer this question by discussing the idea of a global demos defended by political cosmopolitanism.

Recall, first, that some existing demoi have realised basic non-domination reasonably well. Merging them into one would not fully respect their moral worth qua distinct demoi. Second, a demos is, historically, the product of processes of nation-building, which usually involve many human-rights violations, thus being undesirable, especially on a global scale. Granted, the nexus between nationalism and the constitution of a demos is perhaps just contingent. Let us then shift to a functional understanding and define a demos as a group of people able to be organised democratically in such a way as to function as a group agent (List and Archibugi 2010, 89). Even so, to display this capacity people should have some internal cohesion (2010, 95–97, 100–103). This requires, following Miller (2009, 208–213), some mutual concern and trust, as well as some common reasons, all of which are facilitated by the presence of a common public sphere. Without them, others might just be seen as an obstacle to the pursuit of one’s interests, apathy might prevail, and disagreement might be extreme. However, achieving this at the global level remains difficult, even if our focus is only on human rights: individuals are widely dispersed; human rights are specified in different, and even conflicting, ways; and it is hard to become informed well about complex issues concerning them, such as climate change. In fact, what we have witnessed so far is just the temporary coalescence of some (NGOs, activists) around negative norms prohibiting atrocities (List and Archibugi 2010, 105, 109). This issue-based activism should not be underestimated,^[13] yet falls short of the committed and vibrant citizenry, traditionally invoked by republicans (Pettit 1997, chaps. 4, 8), needed to control and contest, on an ongoing basis and over time, the power of a supranational union.^[14] Such power might thus easily degenerate and dominate. Conversely, as Section 4 suggests, free states have unique resources to provide the needed cohesion, and some already provide it – not only sovereignty and democratic institutions allow the formation of a popular will, but civic education and a common, historically nurtured, cultural and political horizon foster also solidarity and active participation (Ypi 2008). State-based demoi may thus count on these resources to perform their international function of control. Condition (iii) is then met too.

To clarify the idea of demoicratic control, I will now outline some criteria that a demoicratic union must meet. I will propose some reforms on how such criteria might be fulfilled, though alternative reforms may work equally well.

1. All demoi. Since basic non-domination concerns everyone, all the demoi must be part of the union and share control over it. Two exceptions apply though. First, despotic states: as they are not representative of the citizenry, and would introduce domination in the union, they must be excluded from membership (see also Pettit 2010b)^[15] – though the union must monitor and sanction them. Second, there could be undemocratic states that, much as Rawlsian decent societies (Rawls 1999, 62–77), are accountable to their citizens and protect all the other human rights necessary for basic non-domination. This makes them relatively good candidates for membership in a union aimed at protecting human rights. Yet full inclusion could still introduce domination. To address this, their membership and the access to the privileges conferred by the union could be made conditional on transitioning to democracy.

   One problem is that some democratic demoi might be reluctant to fulfil their duty to establish and join a supranational demoicratic union. As republican demoicrats argue (Bellamy 2014; 2019), they cannot be forced to do so, or, in Rousseau’s words ([1762] 2012), ‘forced to be free’ (42), for this would violate the demoicratic principle that each demos should retain the constituent power and would thus engender domination. Yet, note that the claim not to join the union is a weak one, for it amounts to a claim to put at risk human rights without appropriate constraints, whilst enjoying impunity. Hence, although compliant states cannot coercively compel non-compliant ones to join the union – say, through annexation – they will still be authorised to apply pressure through diplomatic efforts, economic incentives, and sanctions.

2. Robust power of influence. Each member must possess control over the union. The ensuing human-rights regime would otherwise be alien to some. Each demos, then, must have robust opportunities to influence common decisions. Since, however, the demoi vary greatly in terms of their size and stakes in different decisions (arguably, Bangladeshi have a greater stake in addressing climate change and its impact on human rights than German citizens), the union must account for these differences. One possible solution is to introduce votes proportional to each demos’s size and stakes (Brighouse and Fleurbaey 2010).^[16]

   Importantly, the electoral mechanisms to adopt should avert the risk of persistent minorities. For as republicans such as Pettit (1999) insist, I live according to others’, not my own, will, not only when I have no say in common decisions, but also when, although enfranchised, I am constantly outvoted. I am, in effect, subject to the ‘majority tyranny’ or, as Madison calls it (Madison et al. [1787–8] 2006), ‘elective despotism’ (300). Simple majority voting should then be avoided, given its high exposure to this risk. A strong qualified majority vote and consensual decision-making should be preferred: the former makes it harder for a majority of demoi to unionise and dominate a minority; the latter theoretically excludes this risk altogether.

3. Demoi, not states, must be in control. The demoi should exercise their power of influence through the members of national parliaments rather than the executives. To this end, national parliaments could enjoy, domestically, a greater deliberative role in what is decided in supranational institutions and a stronger power of revision (Bellamy 2019). Alternatively, supranational institutions could be parliamentarised. An example is a reformed WTO that includes a parliamentary assembly where national MPs check that the policies on trade under discussion do not endanger human rights (Shaffer 2005). The latter option is preferable, for if MPs deliberate with the MPs of other states, they will likely be less exposed to a chauvinist understanding of their national interest.

4. Right of exit. The union should envisage an element of no return, yet it should also comprise a right of exit. For I am unfree if I cannot exit relationships which have gone wrong – being able to divorce an abusive spouse is one obvious example (Pettit 2012, 157–158). Yet this right cannot be a robust one, ‘ready to be exercised’, as some demoicrats envisage it (Cheneval and Schimmelfennig 2013, 342). Indeed, a strong opt-out clause would give the demoi a unilateral power over human rights (‘either we do that or I shall leave’) and make their protection too reliant on the demoi’s goodwill. It should then be a thin one, that is, it can only be employed when certain conditions occur. Here is one proposal for specifying these conditions: (a) there is a serious and systemic injustice caused by the union at the expense of human rights; (b) all other means to solve the problem (procedural means, juridical checks, and even resistance) have repeatedly failed; (c) the other demoi will not suffer a grave injustice because of this opt out. The right of exit would then work similarly to how the right to secession works within states.

6 The Constitutional Feature

The point of the demoicratic union, I now argue, is to set up a constitution. Below, I clarify what this means and shed light on the core institutional ingredients of the needed constitutional apparatus. I also stress how the proposed ideal improves comparable ideals.

7 Conclusions

In this paper, I have done three things. First, I have outlined a republican approach to human rights and shown that such entitlements are essential to enjoying a minimum of republican freedom – that is, basic non-domination, defined as the ability to set up and maintain a republic. I have then offered an overview of the main threats to human rights today and focused on new global and transnational agents and dynamics, such as TNCs and the rules underpinning the current state system. Republicanism, I have highlighted, allows us to foreground the underlying, sometimes insidious, power relations generating such threats, thereby complementing the dominant distributive approaches to global justice. I have also noted that current states and international law are ill-equipped to address these threats either effectively or legitimately, thus necessitating substantial reform.

Second, I have suggested that the necessary reforms should not lead either to a transnational human-rights regime in which sovereignty is dispersed or to a global republic. Instead, a middle ground is needed, where sovereign states remain central but undertake more radical reforms than those usually envisaged by internationalists.

Lastly, I have advanced such reforms and argued that the human-rights regime we are looking for is an international constitutional demoicracy. In this new mixed human-rights regime, sovereign states remain the primary units where human rights are realised. Yet they acknowledge that certain threats to human rights escape their control and so form a union to address them – and them only. The union, however, does not replace the demoi with a sole global demos, but rather leaves to each demos control over the union. In institutional terms, it does not give rise to a political cosmopolis, yet it establishes a binding supranational apparatus that is thin, for it is only concerned with the human rights necessary for basic non-domination, and constitutional because it pursues distinctively constitutional aims: it constrains state power over human rights and empowers them against global and transnational threats to such entitlements that each demos is unable to address on its own.

The ideal of constitutional demoicracy makes various contributions. It contributes to the literature on global justice and institutional design by introducing a new and different institutional entity, which protects human rights both robustly and in a non-arbitrary way, and so better than other comparable human-rights regimes; to the republican literature by expanding its internationalist agenda, and improving Laborde’ and Ronzoni’s (2016) progressive internationalism; to the literature on demoicracy by extending the democratic ideal beyond the European Union; and finally, to the literature on global constitutionalism, by broadening its agenda of reforms.

Of course, further, interdisciplinary, work is needed to fully articulate all the reforms necessary to protect human rights today. What I hope to have shown is that the ideal of constitutional demoicracy gives us useful guidance on how we ought to think about them.