1. Introduction
‘A treaty is a disagreement reduced to writing (if one may be permitted to do such violence to an ancient definition of a contract). But so is legislation. The eventual parties to a treaty enter into negotiation with different ideas of what they want to achieve. Negotiation is a process for finding a third thing which neither party wants but both parties can accept. The making of legislation, at least in a society with an active system of politics, is a similarly dialectical process, by which conflicts of ideas and interests are resolved into a legal form which then re-enters the general social process as a new datum. A treaty is not the end of a process, but the beginning of another process. And so is legislation.’
(P. Allott, ‘The Concept of International Law’, (1999) 10 EJIL 31, at 43)
Philip Allott, the international legal theorist, once said that international treaties are best understood as ‘disagreement reduced to writing’.Footnote 1
This was, and still is, an intriguing statement. Indeed, international treaties, but, more broadly, international legal sources in general are predominantly,Footnote 2 and despite occasional normativeFootnote 3 and descriptiveFootnote 4 critiques, (understood as being) ‘based’Footnote 5 on (state)Footnote 6 consent.Footnote 7 To that extent, international law is unlike domestic law where consent to the law has long been considered peripheral or irrelevant to law-making, whether as a criterion of validity or as a ground of legitimacy.Footnote 8 Not the least, arguably, due to the acknowledgement of widespread and persistent reasonable (moral-political) disagreement about and in the law, i.e., disagreement among people who think and converse in good faith and do their best to apply the general capacities of reason pertaining to the domain.Footnote 9
So, is disagreement really that irrelevant to the way we make international law? But for a few exceptions,Footnote 10 international lawyers have not yet addressed the issue of reasonable disagreement, and certainly not from the angle of the central role of state consent in international law-making. Most of them seem curiously resigned to some form of pragmatic endorsement of consent and this even when they reject it on another conceptual or normative ground.Footnote 11 To address this apparent paradox, we need to enquire into the potential role of consent, agreement and disagreement in international law-making. Discussions of disagreement in legal philosophy have almost exclusively concentrated on domestic law so far.Footnote 12 It is time therefore to broaden our legal philosophical scope,Footnote 13 and to venture into a discussion of disagreement in international law.
In short, my argument is that moral and social pluralism is actually greater in international law than it is domestically, thus making the existence,Footnote 14 but also the role of reasonable disagreement in law-making even more central internationally. I submit that it is precisely due to those circumstances of international law-making, however, and to its decentralizedFootnote 15 and non-democratic features additionally, that (democratic) state consent should remain central to international law-making. To do so, it should not be conceived as it usually is through voluntarist, contractualist, or other classical analogies with individual consent and self-determination,Footnote 16 and especially not as a criterion of legal validity or as a ground of legitimate authority of international law. We should set aside the conception of state consent that has predominated in many conflict-averse and dispute-settlement understandings of international law since the early twentieth Century,Footnote 17 whereby international law is understood to signal the end of politics and hence of disagreement.Footnote 18 Instead, it is a revised account of consent attuned to the circumstances of reasonable disagreement and as a way to channel and manage disagreement one should endorse, one that is at home with the legal pluralism, conflicts and fragmentation that characterize international law today.Footnote 19
The proposed interpretation of consent in international law-making fits and justifies the practice of international law, and two of its dimensions in particular. First of all: consent, when it is required in international law-making, does not amount to strict agreement only, but rather to an agreement to disagree further. And second: when consent is not required in international law-making, it is only when that normative process is situated outside of the formal sources of international law.
First of all, then, the proposed disagreement-attuned understanding of consent accounts for the dualistic practice of international law that demonstrates both traits of agreement and disagreement. This can be exemplified on the basis of the current practice of international treaty law (that refers to mutual agreement as a constitutive element of treaties, but, at the same time, entails rules on reservations, on interpretation, and on intractable treaty conflicts that all confirm how disagreement is allowed to continue in the practice of treaties despite the original agreementFootnote 20) or of customary international law-making (that is based on a converging albeit non-unanimous practice as a constitutive element, but, at the same time, entertains the possibility of persistent objection that confirms how agreement matters after all in the disagreementFootnote 21).
Secondly, and even more importantly, the proposed argument for the importance of state consent in international law-making also accounts for the, irritating to some,Footnote 22 resilience of consent in the formal sources of international law. Of course, it has not been uncommon for international lawyers to observe or predict its erosion based on the practice.Footnote 23 What is often mentioned indeed is that treaties have decreased in number since the 1990s, that a lot of the treaties concluded since then have third-party effects and that customary international law, judicial law, and international organizations’ law are becoming more and more majoritarian and less consensual in the way they are produced. All the same, the changes usually identified are mostly located at the periphery and, one may even argue, outside of international law:Footnote 24 it suffices to think of the development of soft law or of unilateral law by powerful states.Footnote 25 Thus, although there are new non-consensual means of international co-operation, neither of them are (yet) regarded as sources of international law stricto sensu. The latter are all still mostly consent-based. Whether one regards sources of international law as a contingent feature of legality or as a conceptual-normative one, doctrinal practice shows that the sources of Article 38 of the 1945 Statute of the International Court of Justice (ICJ) are still regarded as the formal sources of international law. So, if there is a threat to state consent in international law-making, it is external rather than internal to international law. One way to account for this, therefore, and to react is precisely to re-conceive consent, as I propose, so as, first, to accommodate disagreement in agreement in treaty-making and customary international law-making and, second, to enable the revised understanding of consent to do its share of work in the legitimation of international law without driving some matters outside of the scope of international law.
The structure of the proposed argument is five-pronged and so is the structure of the present article. After a first section on the presentation of what is meant by ‘consent’ and ‘disagreement’ in domestic and international law (Section 2), the next section explains why consent cannot and should not be considered either a criterion of legal validity or a ground for the legitimate authority of law, be it domestically or in international law (Section 3). In a second step, further popular myths about the role of state consent in international law-making are dispelled, and in particular in relation to its alleged ties to legal positivism and state sovereignty (Section 4). What the next section argues, however, is that state consent can and should still play a distinctive role in international law (Section 5). Arguably, the role of state consent is best understood and justified by reference to the circumstances of reasonable disagreement about and in international law, and hence to state democracy and equality. This only applies, however, within the limits of what democratic state consent can actually amount to, and this means within the limits of both basic political equality and the free and fair nature of consent. This in turn enables me to explain, in the last section, how consent remains a central element in most sources of international law and to justify that centrality, and in particular to show how agreement and disagreement go hand in hand in the practice of international treaties and customary international law (Section 6).
A methodological caveat is in order before going forward, however. Unlike previous discussions of consent in international law-making, the present one is both descriptive and normative. Many arguments put forward to defend it or rebut it have been expressed as factual claims about the degree of importance of consent in the practice of international law, submitted to judgments of truth or falseness. Others, on the contrary, have focused on conceptual or normative considerations to establish whether or not consent could be considered as a criterion of validity or as a ground of legitimacy. The difficulty is that both dimensions are needed for any argument about consent in international law-making to succeed, and one cannot be severed from the other or, worse, confused with one another.Footnote 26 Indeed, law is a normative practice in which the conceptual-normative level plays a very important function and cannot simply be replaced by either factual observations or mere normative considerations.Footnote 27
2. From ‘consent’ to ‘agreement’ in international law
A lot of the difficulties with the notion of consent in international law stem from its polysemy. It is used to mean different things that may then relate very differently to the validity or legitimacy of international law.
Because of the link made in this article between consent (Section 2.1) and agreement (Section 2.2), and, by extension, between consent and disagreement, it is important to define both notions.
2.1. ‘Consent’ in international law
Defining consent is by no means easy as most international legal scholars do not explain what they mean by it.Footnote 28
A cursory survey of the use of the term ‘consent’ in international law reveals that consent is used to refer both to the ‘free will’ of states and to their ‘acceptance’ of international law.Footnote 29 This comes close to the uses of consent one encounters in domestic law, more generally. There, consent is often equated with self-determination and can have a voluntarist flavour, on the one hand, while also being used to mean acquiescence and positive voting and can have a participatory colour, on the other. In what follows, consent is understood as the latter: acceptance of a given international legal norm. Its original voluntarist dimension will also be discussed, however, albeit eventually set aside.
In international law, like in domestic law and politics, consent can be express, as in the signature or ratification of a treaty, but also tacit, as in the absence of objection to a reservation or to a custom in the making. There are also contexts, like customary international law-making for instance, in which consent occurs not by expressing or stating, but by doing – as in a converging practice or in participating in international law-making, more generally. As a matter of fact, customary international law-making combines tacit consent in the converging practice of states and explicit dissent in their possibility to object to that practice through a persistent objection.
Two further conceptual points are worth emphasizing about the notion of consent in international law: their subjects and their objects.
First of all, the subjects of consent. Interestingly, the collective nature of the actor consenting in international law, i.e., the state, constitutes a key difference to domestic law where what is usually at stake is individual consent. Curiously, the collective and institutional structure of states has not been discussed by international law scholars as potentially invalidating any analogy between state and individual consent. Of course, such a ready individual analogy fits the broader and widespread approach to states as unitary subjects of international law that pervades international law scholarship and reduces states to individuals in a domestic setting.Footnote 30 I will come back to this question later on as it is at the core of a central tension in most accounts of consent in international law. The collective dimension of state consent is crucial, I will argue, to understand its role in international law-making and especially to understand its democratic justification.
Secondly, the object of consent. Regrettably, the question is not usually broached by international law scholars. If consent is central to international law-making, it should apply to the whole range of objects of international law without restrictions, including moral-political issues such as human rights, climate justice, disarmament, anti-terrorism, or global poverty. In rational choice approaches to consent, however, state consent is usually reduced to consent regarding state interests and hence to self-interested consent. These accounts then necessarily conclude to the so-called status quo bias of state consent and to its disabling effect in common problem-solving, as a result.Footnote 31 This objection need not worry us, however. To start with, the importance of individual interests in political and legal decision-making, besides concern for the moral-political issues at stake, is a well-known feature of political and in particular democratic deliberation processes.Footnote 32 Moreover, the interests of collective and institutional agents like states need not be self-oriented only and could actually encompass more community-oriented concerns. It is likely, finally, that collective and institutional agents like states can actually be designed so as to enhance their ability to consent by reference to other factors than their interests. This is definitely the case of democratic states, and I will come back to this issue later on.
2.2. ‘Agreement’ in international law
There are many understandings of ‘agreement’ (and disagreement qua absence of agreement) in legal and political theory. As I have argued elsewhere, it is best understood as unanimous (intersubjective) acceptance for the same reasons by opposition to mere overlapping consensus on different grounds.Footnote 33 This is also how I will understand agreement in international law in this article. The relationship between consent so defined and agreement becomes clear when confronted with the ‘acceptance’-understanding of consent that has just been discussed.
This actually matches the understanding of consent qua agreement one encounter in international law. Although international law-making by consent is sometimes referred to as ‘consensual’,Footnote 34 it does not necessarily imply aiming at ‘consensus’. Of course, in practice, searching for consensus is a way of then securing unanimous consent by all states, but not necessarily so. Generally, ‘consensus’ is actually used in international law to refer to a (non-unanimous) converging practice, distinct from (unanimous) agreement (by consent).Footnote 35 It is the case, for instance, in customary international lawFootnote 36 – at least in the first phase, as we will see.
The source of international law that epitomizes the understanding of consent qua agreement is the international treaty.Footnote 37 Treaties are actually referred to by Article 2(1)(a) of the 1969 Vienna Convention of the Law of Treaties (VCLT) as ‘international agreements’. Agreement is defined in that context as the mutuality of expressed consents to be bound by a treaty (Arts. 2(1)(b) and 11 VCLT).Footnote 38 The expression of consent can take different shapes and names such as, and not exclusively, ‘signature’, ‘ratification’, ‘accession’, ‘acceptance’ or ‘approval’ (Arts. 11–17 VCLT).Footnote 39 Their exact meaning remains surprisingly open, however.Footnote 40
Importantly, treaties are more than agreements. They are based on mutual consent or agreement, but it is only one of their constitutive elements. They actually amount to agreements ‘governed by international law’ (Art. 2(1)(a) VLCT) or, in more direct terms, agreements binding by reason of international law and qua international law. What this means is still subject to controversy, however.Footnote 41 The best interpretation is that treaties qua agreements are unlike contracts or mutual promises that are governed by themselves or, if the parties so wish, by (domestic) private law. Their being (governed by) international law cannot indeed be excluded by the states parties.Footnote 42 Or else they could not amount to a source of international law. This explains in turn why, contrary to what one may sometimes read, the expression of mutual consent in international treaties is not in itself the ground of international legal obligations, but it is their being (a source of) international law that is. I will come back to this in the context of the discussion of the relationship between consent and legal validity.
Importantly, this understanding applies to so-called ‘treaty-contracts’ as much as to ‘legislative-contracts’. Those labels are unfortunate, however, to the extent that, on the proposed reading, treaties are never contractual. What these labels capture in fact is another difference between treaties: some are reciprocal in the structure of the rights and obligations they give rise to, arguably due to their content, and hence can be compared to contracts to that extent, while others cannot.Footnote 43 As a result, while it is common among international law scholars to refer casually to the terms ‘contractual’ or ‘contractualism’ to capture the role of state consent in international law-making,Footnote 44 I suggest it should be avoided. This is not to say, of course, that states may not conclude contracts and hence use consent to do so. As I explained before, however, when they do so, they do not create international law and the source of their obligations is not a treaty.
3. Consent and the validity and legitimacy of international law
A lot of the resistance to consent in legal theory stems from its inability to be a criterion for the validity of law, on the one hand, and to amount to a ground for the legitimate authority or legitimacy of law, on the other. This also applies, by extension, to its role in international law. In what follows, I will address both issues in turn. Of course, they are intimately related to the extent that the law should be made (its validity; Section 3.1) in such a way that it can claim to be legitimate and hence to bind/give rise to obligations to obey (its legitimacy; Section 3.2).Footnote 45
3.1. Consent and the validity of international law
One understanding of the role of consent in international law-making is that consent amounts to a, or even the, source of international law or, at least, to one of its dimensions. In short, it is seen as a criterion for the validity of international law qua law.
There are two critiques one may make to this approach to state consent, however: one conceptual-normative and the other descriptive.
As Herbert Hart pointed out early on, first of all, this understanding of state consent, and in particular of treaties, is conceptually confused. It begs the question of what makes treaties international law in the first place, and hence binding qua source of law.Footnote 46 What explains that treaties or any other source of international law gives rise to norms of law, i.e., norms that bind qua law, cannot be that their subjects (i.e., states) have consented to them. On the contrary, it is a rule of law according to which treaties (that states have consented to) are a source of law and bind states as such. This rule is pacta sunt servanda (Art. 26 VCLT). Importantly, that rule cannot itself be explained by reference to consent as this would imply a fatal regress. Of course, treaties could be binding as mutual promises or contracts,Footnote 47 but that is not what their being a source of international law amounts to. Contracts are merely agreements that are legally enforceable: entering into a contract affects my legal obligations, but this does not mean that I have made law.Footnote 48
Of course, the next question would be how to explain the principle pacta sunt servanda, the rule of recognition qua general criteria of validity of international law and, more generally, the existence of a legal system in international law, if it is not by reference to state consent. These questions need not be settled in this article. Hart argued pacta sunt servanda merely amounts to a principle of customary international law.Footnote 49 He also argued, however, that custom could not amount to a separate source of law for lack of a rule of recognition and hence of a criterion of validity determining when a rule of customary international law is in force.Footnote 50 Since then, others have rightly argued to the contrary. Although Hart was right about the distinction between the criterion for the validity of treaties and consent, he was wrong about the absence of a rule of recognition in international law and about the lack of general criteria for the validity of international law.Footnote 51
The second objection to state consent as a criterion for the validity of international law is descriptive. It suffices to look closely at international treaties and customary international law, i.e., the two most important sources of international law, to see that while consent is a requirement in order to be bound by the norms arising from those sources, it is not used as a criterion of validity. (Multilateral) treaties are concluded validly without a particular state's consent, and although they cannot bind that state, they can bind others. The same may be said about a customary norm that cannot bind a persistent objector, but is valid and can bind other states. In sum, state consent does not work as a criterion for the validity of international law that is general. The connection is more concrete, and it is to the legal obligations stemming from valid international law for a particular state. There seems therefore to be a link between state consent and the grounds for international legal obligations. I will now turn to this question.
3.2. Consent and the legitimacy of international law
Another understanding of the role of consent in international law-making is that consent amounts to a ground or justification for the legitimate authority of international law so produced or identified.
As Joseph Raz has pointed out, however, while consent may explain why promises are morally binding, it does not explain how the law can bind and, more specifically, give us moral reasons for action. One may indeed consent to do wrong.Footnote 52 Importantly, it is not only the plausibility of actual or express consent to the law that is at stake in this rebuttal, but its conceptual-normative inability even qua hypothetical or tacit consent to justify the law's authority.Footnote 53 This is also the case for the justification of the authority of international law.Footnote 54 True, consent is a content-independent reason for action, and we will see later that content-independence matters for the legitimacy of international law, but it cannot in itself be the justification for the authority or moral bindingness of law.
Of course, as Raz recognizes, consent may constitute an additional reason to respect the law or have trust in it. However, this does not make it a primary ground for its authority unless the law respects autonomy and satisfies an independent test of legitimacy.Footnote 55 This matters particularly for international law because consent may contribute to enhance the de facto authority of international law by strengthening respect for it in practice.Footnote 56 Additional reasons for trust and respect are significant given the pluralist circumstances of international law.
This rejection of consent as a ground of legitimacy also extends to democratic legitimacy. Democracy is understood here as method of collective decision-making characterized by equality among the participants, on the one hand, and by their inclusion in the decision-making process, on the other.
The disconnection between democratic legitimacy and consent may seem counterintuitive to some. Because democratic decision-making implies participation and voting, indeed, it is often assumed that democratic legitimacy is consent-based.Footnote 57 Nevertheless, this objection can be easily brushed aside by reference to the majoritarian principle qua core democratic principle (by contrast to the requirement of unanimity that would derive from consent).Footnote 58 What the principle of majority actually shows is that democracy protects and is justified by reference to equality, and not to self-determination (that would justify, on the contrary, the recourse to consent).Footnote 59 Moreover, consent and autonomy qua grounds for democracy would be difficult to reconcile with the circumstances of democracy and democratic deliberation in particular, i.e., disagreement.Footnote 60 The publicly equal advancement of interests requires indeed that individual participants’ judgments be taken into account equally when there is reasonable disagreement.Footnote 61 This applies whether consent is actual or hypothetical given how reasonable and hence persistent disagreement can be.Footnote 62
All this also holds true for democratic accounts of international legitimacy.Footnote 63 However, as I will argue, consent plays a distinct role in that context precisely because the democratic legitimacy of international law cannot be approached in the usual way. There is no international polity of equals and this means that it is the domestic democratic legitimacy of states that needs to be taken into account.Footnote 64
Of course, this leaves us with the need to explain the importance of participation and inclusion in democratic legitimacy. Participation seems indeed to imply some form of consent. As a rejoinder, one should stress that disqualifying consent as a ground for the objective legitimacy of law does not imply setting it aside as a factor of subjective legitimacy. The latter, and generally acceptance of law, remain important in practice and arguably even amount to a dimension of objective legitimacy.Footnote 65 The same may be said about the importance of participation and subjective acceptance for the objective legitimacy of international law. Of course, it is not a ground for the latter and should not be conflated with it,Footnote 66 but amounts to an important dimension of the legitimacy of international law all the same.
Last but not least, consent is often invoked in the context of compliance with international law, especially within rational choice accounts of international law-makingFootnote 67 albeit not only.Footnote 68 Of course, reasons for compliance are not the same as justifications of authority. Compliance matters for our purpose, however, because the effectivity of international law is one, albeit not the sole, dimension of its legitimacy to the extent that international law should be such that it can be obeyed in practice. If consent affects compliance and the effectivity of international law, as a result, then it is somehow relevant to its legitimacy. In reply, one may refer to evidence of other sources of effectivity propounded in rational choice theoretical accounts of non-consensual international law-making.Footnote 69 Other factors of compliance are mentioned there such as power or common problem-solving. Moreover, compliance with international law is actually secured largely through domestic law in practice and this is often missed in international law scholarship.Footnote 70 We should not as a result try to locate the reasons for compliance only in international law and the way it is made.
Once consent is severed from the potential grounds of legitimacy of international law, the next question, of course, is what those grounds actually are and how we should account for the legitimacy of international law. This question need not be settled fully here. In short, I have argued elsewhere for a revised Razian conception of legitimate authority whereby international law binds to the extent that the (content-independent and exclusionary) reasons it provides enable its subjects (states and individuals in those states) to comply better with the reasons that apply independently to them (service conception).Footnote 71 On that objective conception of legitimacy, the justification of the authority of international law is piecemeal and its grounds may range from volitional, expressive, epistemic, to co-ordinative reasons. However, my account of the legitimacy of international law is a revised Razian account to the extent that the main basis of the demand for legitimate authority is reasonable disagreement about how to structure common action when such action is required morally or practically. Accordingly, as I have argued elsewhere, the justification of legitimate authority lies mostly in the co-ordinating ability of international law in circumstances of reasonable disagreement.Footnote 72 Moreover, my revised Razian account of the legitimate authority of international law accommodates the importance of public and hence democratic authority by recognizing co-ordination as a ground for general and not just piecemeal authority. I will revert to legitimate authority later in this article.
Although consent is neither a criterion for the validity of international law nor a ground for its legitimacy, this does not mean that it should play no role in international law-making. As a matter of fact, international legal obligations are never imposed on states without their consent in practice.Footnote 73 Unlike other authors, we should not consider this as an unhappy feature of practice that we should be pragmatic and resigned about. On the contrary, I would like to claim that there is a normative argument to be made for the role of state consent in international law-making. But, first, let me show how that argument for state consent is not among the ones usually made by international law scholars.
4. Two myths about consent in international law
There are two myths about the normative role of consent in international law-making that may explain why it sticks in the practice and scholarship of international law despite being disqualified as a criterion of validity and as a ground of legitimacy of international law. Those myths pertain to the relationship between consent and legal positivism (Section 4.1.), and state sovereignty (Section 4.2.). It is important to debunk those connections to be able to focus in the next section on what could actually be the justification for the role of consent in international law-making.
4.1. Consent and legal positivism
The first myth about state consent is that consent in international law is related to legal positivism and necessary to the latter's success as a legal theory of international law. It is common indeed among international lawyers, as it is domestically, to oppose natural law theories to legal positivist ones. However, the specificity of that distinction in international legal scholarship is that legal positivism is often reduced to ‘legal voluntarism’ whereby the grounds of law are not only regarded as a matter of fact, but a matter of will.Footnote 74
The explanation for this conflation is, of course, the role state consent plays in international law-making, but it also goes deeper historically.
The first explanation for this conflation is the role of state consent in international law-making. In reply, one should state that voluntarism is not implied by the idea that the grounds of law are matters of (social) fact and hence that the law is posited. There are other ways of positing international law than consent. Even if consent plays a central role in that process, as it does, it is not a necessary condition thereof. As a matter of fact, even when state consent is required as in treaty law-making, considering consent as a criterion for the validity of international law is question-begging, as I have argued before.
A second explanation for the association between legal positivism and voluntarism, however, is historical.Footnote 75 Until Hart's account of legal positivism, legal positivists distinguished positive law from morality and natural law by reference to will. They considered states’ will as the relevant fact behind positive law. This was the case in Grotius’ understanding of international positive law as ‘acts of will’,Footnote 76 but also later on, and well into the late nineteenth Century, in the context of German statutory positivism and its notion of ‘legislative will’.Footnote 77 This historical connection between voluntarism and positivism was severed by Hartian legal positivism, however.Footnote 78 Today, it has only subsisted, it seems, amidst a few international legal positivists, mostly stemming from rational choice theories of international law.Footnote 79
So, international legal positivism does not require state consent to succeed as a legal theory. Nor, conversely, does state consent require legal positivism as a legal theory to matter in international law-making. One last comment is in order now that the necessarily reference to state consent has been severed from legal positivism. It is common, as we have just seen, among international law scholars to refer casually to the term ‘legal voluntarism’ to capture the role of state consent in international law-making.Footnote 80 What the discussion in the first section has shown, however, is that consent is not just the expression of the will, but much more. Reducing the argument for its importance to voluntarism, as a result, is misleading.
4.2. Consent and state sovereignty
The second myth about the role of consent in international law-making is that it is tied to sovereignty. This goes back to the early days of international law and especially to the personification of state sovereignty (e.g., as a king). The command of the sovereign came close to the expression of its will, therefore.
The relationship between state sovereignty and consent carried on well into the early twentieth Century, however. This is mainly because of the way in which the Permanent Court of International Justice in SS Wimbledon, and many international lawyers back then, solved the so-called sovereignty paradox.Footnote 81 Their aim was to explain how states could incur obligations under international law, and especially international treaties, and still remain sovereign or, in other words, how they could be free and not free at the same time. The simple answer was that ‘the right of entering into international engagements is an attribute of state sovereignty’.Footnote 82 This in turn was only possible because under the Lotus doctrine, ‘the rules of law binding upon states . . . emanate from their own free will’.Footnote 83 From then on, consent was linked to state sovereignty in international law scholarship, and vice-versa.
In reply, one should state, following Endicott's analogy with individual freedom in Mill,Footnote 84 that state sovereignty is best understood as state autonomy in an objective and thick notion of autonomy. This limits what a state can consent to as a sovereign to what actually enhances its autonomy. It therefore severs sovereignty away from state consent. In turn, what this implies is that sovereignty seems to amount to a large extent to what legitimate international law says it is, and not the other way around, contrary to what consent-based accounts of sovereignty have long defended.Footnote 85
Of course, this does not mean that whatever international law says lies outside the scope of sovereignty. There may indeed be restrictions to sovereignty by legitimate international law that are deemed incompatible with the (objective) autonomy of sovereign states.Footnote 86 The legitimacy of international law and the limits to state sovereignty do not therefore match entirely. The importance of self-determination in sovereignty can actually be explained along the lines of the autonomy-based exceptions to the prima facie legitimacy of law in the Razian account. This is also what Raz calls the ‘independence condition’.Footnote 87 According to that condition, there are circumstances in which autonomy requires determining oneself (literally ‘self-determining’) what to do despite the fact that one would comply with one's own reasons better if one did not. It is in those circumstances that state consent can play a role for the respect of a state's sovereignty, as I will explain in the next section.
5. International legal disagreement and democratic state consent
International legal obligations are never imposed on states without their consent,Footnote 88 even though it is not consent that explains their validity or legitimacy. The time has now come to make a normative argument for state consent in international law-making. After presenting a disagreement-based argument for democratic state consent (Section 5.1.), I will discuss some of the latter's inherent limits (Section 5.2.).
5.1. Justifications of democratic state consent in international law-making
The central role of state consent in contemporary international law-making is justified, I would like to argue, for democratic reasons in circumstances of reasonable disagreement. This revisits and justifies state consent as ‘democratic state consent’.
At first, this may read as a paradox. Indeed, domestically, the role of consent has long been disparaged from a democratic perspective precisely because of reasonable disagreement. Actually, as I explained before, the equality-based justification of democracy accounts for majority-voting instead of unanimity, thus making consent even less relevant procedurally. Why would it be different internationally? How could democracy justify the role of state consent in international law if not domestically? After all, reasonable disagreement is likely to be even more widespread and persistent in international law-making.
The first thing to say is that democratic justifications and conditions in international law should not be conflated with domestic ones.Footnote 89 First of all, states should not be treated as equal individual members of an international democratic polity in the same way as individuals are equal members of a domestic democracy.Footnote 90 There is no democracy of states, but should not be one either. States cannot, and should not be too readily identified with individual subjects whose basic equality actually justifies democracy; the same justification could not arise from the equality of states. Secondly, the conditions of a global democracy for individuals in which states would merely act as their officials are not (yet) given either. Indeed, in the absence of the egalitarian pre-conditions for global democracy, and in particular of equal and interdependent stakes shared by all of us internationally, there cannot (yet) be a global democracy.Footnote 91 One may even argue that we should not aim at global democracy in any case, in particular by reason of its total(itarian) scope.Footnote 92 As a result, the legitimacy of international law cannot and should not strictly speaking be democratic.
All the same, the importance of democratic legitimacy domestically implies that we should try to find a way to respect domestic democracy in the way we make international law, and especially the political equality of the members of democratic states – actual ones, of course, but also democratic states in the making. After all, to quote Thomas Christiano, states remain ‘the most important institutional mechanism for making large scale political entities directly accountable to people’ and the sole locus of democracy, as a result.Footnote 93 The way to link international law-making processes to domestic democratic legitimacy, arguably, is to respect the equality of each democratic state qua statespeople.Footnote 94 Of course, this gives rise to well-known and largely irreducible tensions, because state equality and global individual equality are not transitive and may not be reduced to one another.Footnote 95 Still, there is no better way (to date, at least) to protect domestic democratic self-determination on the global scale than to protect democratic states’ sovereign equality.
The way to protect the equality of states(people) qua collective equality is to consider their consent as a requirement in international law-making. The requirement of the equal consent of states provides a way for small and weak states to resist the domination and the hegemony of large and powerful states or coalition of states.Footnote 96 In turn, states’ equal consent protects the individual members of those states’ right to an equal voice in the collective decision-making process they are participating in through their states.Footnote 97
The upshot is that only democratic states may invoke their consent as a ground not be bound.Footnote 98 Of course, democracy is incremental. As a result, all states could be considered as democracies in the making. However, as long as they have not attained the minimal democratic standards that have arisen from democratic state practice in international law,Footnote 99 their individual members are not treated as equal and the justification for democratic state consent should not work as an exception to the (independently justified) authority of international law. Importantly, as I will explain below, there are limits on democratic state consent that are meant to protect existing democracies and individual political equality within those democratic states. These limits are important to curb the potential anti-democratic consequences of state consent in practice. They may be used both against democratic states themselves but also, and most often, against non-democratic ones.
Importantly, considering democratic state consent as a requirement of international law-making does not mean considering it as a ground for its legitimacy: as we saw before, it simply cannot be such a ground, whether tout court or in a democratic context. As a matter of fact, the legitimacy of international law is not strictly speaking democratic.Footnote 100 As I explained before, its legitimacy is substantive and, actually, mostly justified on grounds of co-ordination in circumstances of reasonable disagreement. Other substantive grounds may, of course, also apply concurrently, but co-ordination is the most important one due to reasonable disagreement. All the same, it is precisely because reasonable disagreement among states is widespread and persistent in international law, on the one hand, and because of the centrality of democratic states in making international power accountable to their people,Footnote 101 on the other, that democratic state consent should work as an exception to the prima facie legitimate authority of international law.
This corresponds to what I explained earlier about state sovereignty and the exceptions to the legitimacy of international law. There may indeed be restrictions of sovereignty by legitimate international law that are deemed incompatible with the autonomy of sovereign states and hence with the justification of sovereignty. The importance of self-determination in sovereignty can be explained along the lines of the autonomy-based limits to the legitimacy of law in the Razian account of the legitimate authority of law. The so-called ‘independence condition’ to legitimate authority is even more important when justifying the authority of law over states than over individuals. This matters particularly for states whose autonomy reflects their individual members’ autonomy, and especially for democratic states whose sovereignty is closely linked to democratic self-determination and whose sovereign equality protects the self-determination of a statespeople.Footnote 102 Importantly, however, this should not turn democratic state consent into a general exception of sovereignty to the legitimate authority of international law; the former only corresponds to an instantiation of the latter when the state is democratic and the democratic limits inherent to the justification of the exception are respected, as we will see.Footnote 103
In those conditions, the way in which democratic state consent ties in is merely as an exception to the legitimate authority of international law that is justified on other grounds. When a state consents, the legitimate authority of international law and the corresponding duties are confirmed. However, when it refuses its consent to a treaty or objects to a new customary norm, the (legal) duty to obey international law may arise from what has become valid international law for that state just as it does for others, but the norm does not actually bind it (as a moral duty). Importantly, if the state decides to consent later on to an international legal norm and to lift the exception, it is bound. This is the case also when its consent cannot be considered valid, as I will explain in the next section. This is because state consent is not the ground for the legitimate authority of international law that arises independently from it, but merely an exception to that authority.Footnote 104
Democratic state consent qua exception to the legitimate authority of international law may take different forms. It may be explicit, but, as I explained before, it may occur tacitly as well. This could be through statements, but also through practice or participation. Thus, state consent may, for instance, take the shape of the principle of subsidiarity when international human rights courts’ decisions defer to the democratic implementation and review of international human rights law as long as there is no European consensus on a given question.Footnote 105
The proposed account of the role of consent as a democratic exception to the legitimate authority of international law should be distinguished from presumption-based approaches to state consent. Jan Klabbers, for instance, understands state consent in international law as being presumed.Footnote 106 That presumption of consent may always be reversed, however. This presumptive role of state consent is puzzling. Of course, other authors have also drawn on the ICJ's odd formulation in the Lotus case according to which ‘restrictions upon the independence of states cannot . . . be presumed’Footnote 107.Footnote 108 This is, however, a legal procedural feature whose application to state consent is difficult to understand. It is not because consent is tacit, for instance after a certain time, that it was presumed in the first place. Moreover, the term ‘presumption’ is used very loosely and it is difficult to know what it refers to exactly. It is also applied, for instance, to the actual legality of an international legal norm or to its bindingness, which are very different from state consent itself being presumed. Finally, even if consent is presumed, it remains important to know what role it plays in international law-making. Thus, although Klabbers disparages consent as a ground for the legitimacy of international law, he retains it as ‘main criterion of validity’.Footnote 109
It should be clear by now how different the proposed understanding of the nature and role of democratic state consent is from considering it as a ground of democratic legitimacy. This explains the paradox I started with away. First, it accounts for how democratic state consent can require unanimity while democratic decision-making is associated to majority. Second, it also explains how democratic state consent is actually required by the circumstances of reasonable disagreement while the latter usually makes unanimity impracticable.
Of course, this is not to deny the other helpful features of state consent in international law when establishing the latter's validity or strengthening its legitimacy. First of all, what democratic state consent does in treaty-making is allow for the salience of the co-ordinating option.Footnote 110 To that extent, it contributes to the process of validation of international law in the absence of central institutions and law-maker.Footnote 111 This is particularly important in the circumstances of moral and social pluralism and hence of substantive and epistemic disagreement that apply in international law.Footnote 112 This is even more important as states are collective agents: their co-ordinating intent may be difficult to identify. As a matter of fact, democratic states are the officials of their people and their institutions are themselves made of officials, thus making for layers of potential disagreement in any given disagreement. Secondly, as I alluded to before, consent amounts to a clear and public method for the creation of content-independent reasons to obey international law.Footnote 113 It is also a way to strengthen respect and trust in international law, which may explain why the de facto authority of international law is quite strong in practice.Footnote 114
5.2. Limits to democratic state consent in international law-making
While state consent is an important dimension of the validation and legitimation of international law, there are limits to consent inherent either to its democratic justification or to consent.
The first set of limits on democratic state consent pertains to its democratic justification itself: the protection of democratic statespeoples. Those limits are well-known in domestic constitutional theory, and include the protection of the basic conditions of democracy, i.e., basic political equality. Without abiding by those values and principles, there can be no meaningful democracy and they justify constraining the democratic process, as a result.
More concretely, inherent democratic limits usually take the shape of non-discrimination rights, on the one hand, and of absolute or minimal human rights duties, on the other. In international law, those limits to democratic consent are sometimes referred to as jus cogens norms, i.e., imperative norms or norms that may not be derogated from (Art. 53 VCLT).Footnote 115 Thus, one could not allow for the exception of democratic state consent to be opposed to absolute rights or duties in an international human rights treaty, like the prohibition of torture, and to specify reservations to those provisions, as a result.Footnote 116 The same may be said of the limits to opt-outs of newly independent states from treaty regimes accepted by their predecessors: international succession law usually excludes human rights treaties from their scope.Footnote 117 This is also how one should understand the justification of so-called ‘objective regimes’ and of third-party obligations in international treaty law.Footnote 118
A second set of limits inherent to democratic state consent have to do with consent itself. There are mainly three: the voluntariness of state consent, its fairness, and its informed nature.Footnote 119
First of all, democratic state consent can only be invoked as an exception to the legitimacy of international law provided consent is free and unconstrained.Footnote 120 This is also the point of some of the provisions of the VCLT pertaining to the conditions bearing on the quality of state consent to a treaty and to its potential invalidation/invalidity on those grounds (Arts. 46–52 VCLT; constitutional concerns, error, fraud, corruption, coercion).Footnote 121 These limits matter particularly in international law-making where the power imbalances between states can be very important and threaten the voluntariness of state consent. Secondly, state consent should be expressed and potentially exchanged in a fair and egalitarian fashion. The procedures should be such as to guarantee that fairness, and in particular the respect for the sovereign equality of states that grounds the democratic state consent exception. This implies adopting procedures that give states equal participation and voting rights in particular. Finally, state consent should be duly informed to be valid. This matters particularly in pluralistic international relations where epistemic disagreement can be more important than domestically.
Finally, additional non-inherent limits to state consent could be developed based on criticisms made by international law scholars to the justification of the role of state consent in international law-making. Importantly, all these critiques can be accommodated within the democratic justification defended in this article. I will discuss two of them here: limits in the personal scope of democratic state consent and limits in its material scope.
First of all, the need for experts. It is a critique that has been developed mostly in the area of global administrative law (GAL) where consent has actually lost in importance in practice.Footnote 122 Of course, as I explained earlier, GAL has not been developed using the formal sources of international law and does not, to that extent, contribute to the erosion of the role of state consent within those sources. However, it is true that it has contributed to the sidelining of those sources tout court in international law-making and hence of state consent overall. The reasons for this need to be identified, therefore, and accommodated if democratic state consent is to play its role again in these areas of international law-making.
In reply, one should stress, first of all, that the expertise critique has been made against the justification of democracy in domestic law as well. This is not the place to address one of the most difficult objections in democratic theory and scope precludes doing so. Let me say, however, that part of the answer lies in adapting one's conception of political equality by reference to the epistemic pre-requirements of democracy. This implies in particular organizing forms of political representation for the implementation of the choices and objectives identified by citizens.Footnote 123 The same could be done in international law. A rejoinder, however, would be that GAL can be even more complex than the corresponding domestic legislation. It suffices to think of international environmental law or financial law. Again, there are ways to accommodate the need to resort to experts while justifying the role of democratic state consent. One could make sure the aims and principles that guide the experts called to contribute to international law-making are themselves identified by formal sources of international law and hence respect democratic state consent.Footnote 124 Ultimate democratic state control is key if one is to avoid global technocracy.
Secondly, the exclusion of matters of common concern. A second critique made to the justification of state consent in international law-making has to do with the co-ordination-based justification of the legitimacy of international law. In short, the argument is that the kind of collective or common action required of states to protect global public goods or community interests, such as environment or health, is conceptually incompatible with the role of democratic state consent in international law-making. This requires excluding state consent from some areas of international law-making, as a result.Footnote 125
In reply, one should, first, emphasize the broader and more encompassing understanding of co-ordination on moral issues used in this article. It is very different from the kind of game-theoretical co-ordination encountered in rational choice theory. It relies in particular on a pre-existing duty to co-ordinate on moral issues in circumstances of reasonable disagreement. Moreover, the co-ordination argument for the legitimate authority of international law applies outside of (transnational) co-ordination problems stricto sensu and extends to conflict and partial-conflict co-ordination over any moral concern over which there is a disagreement and where having one common take is better than none.Footnote 126 It therefore extends to trade treaties as much as to human rights treaties or counter-terrorism resolutions. In those conditions, on the proposed account of democratic state consent, there actually is a moral duty to abide by international law norms justified by their ability to co-ordinate before state consent can be invoked as an exception. This makes an important difference.
A second reply has to do with the motivation of democratic state consent qua exception to co-ordination-based international law. As I explained before, it is not (merely) about the promotion of states’ interests and welfare, as argued by rational choice theorists but also by others.Footnote 127 Not only, for the sake of analogy, are individual citizens not necessarily motivated solely by their self-interest in a democracy, as I explained before, but states are institutional constructs that can actually be designed to decide in a non-self-interested fashion. State consent extends to our moral reasons as members of those democratic states, including our community-oriented reasons. There is no reason therefore to fear that democratic state consent necessarily be invoked to obtain self-interested exceptions to international law pertaining to the protection of community interests and to collective international action aiming at protecting them. One should also stress the potential of international institutional design in the regimes of international law that face collective action problems.Footnote 128
6. Consent and disagreement in international law-making
The proposed account of democratic state consent fits and justifies the contemporary practice of the formal sources of international law where consent still plays a central role. With its revised understanding, justification and limits, it may even contribute to bringing states back to these sources in the future, and away from informal law-making, to the benefit of domestic democracy and accountability of international law to statespeople themselves.
What I would like to show in this section, more specifically, is that the proposed disagreement-attuned understanding of consent accounts particularly well for the dualistic, and prima facie paradoxical, practice of international law. Indeed, that practice actually demonstrates both traits of agreement and disagreement. Like domestic law, international law-making has ways to channel and manage disagreement so as to make the most of it, on the one hand, without being undermined by it, on the other. This can be exemplified on the basis of the current practice of treaty law-making (Section 6.1.) and of customary law-making (Section 6.2.).
6.1. Consent and disagreement in international treaty law-making
While international treaty law-making is often considered as the epitome of law-making by consent and hence by agreement, a closer look shows that reasonable disagreement amounts to a permanent concern of the secondary rules of treaty law. If it is true that treaties are the expression of mutual consent and cannot be concluded without consent, consent need not imply complete agreement. Instead, treaties are arguably best approached as incompletely theorized agreements or, at least, as agreements to disagree. This may be exemplified by the rules on reservations, on interpretation and on intractable treaty conflicts in the international law of treaties.Footnote 129
First of all, disagreement and the rules on treaty reservations (Arts. 19–20 VCLT).Footnote 130 Treaty reservations are forms of bilateral sub-agreements within a multilateral agreement that allow for pockets of disagreement to subsist on the detail. They make the conclusion of treaties possible despite remaining disagreements on issues other than the object and aim of those treaties (Art. 19(c) VCLT). Of course, such disagreements within the agreement are only allowed provided the states parties agree to disagree (Art. 19(a) and (b) VCLT). That agreement is usually tacit, however (e.g., Art. 20(5) VCLT). Interestingly, reservations are never definitive and may always be withdrawn (Art. 22 VCLT), thereby extending consent later on to the whole treaty – a treaty that was actually valid and binding, except for the reserving state, since its first entry into force.
One of the controversies in this context is what should happen to the consent to the treaty in case a reservation is deemed incompatible with the treaty (Art. 19 VCLT). The rule is that the parties usually ought to decide individually on the approval of reservations and objections thereto (Art. 21 VCLT). This implies that state consent should always be respected even when a reservation is deemed invalid and that the question of the bindingness of the treaty as a whole arises again. In the context of international human rights treaties, however, international human rights courts and bodies usually impose the conclusion of the treaty, with the provisions that were the object of the reservation, onto all states parties, including the reserving states and the others, without requesting their consent.Footnote 131 This has been criticized for violating the principle that international legal obligations are never imposed without state consent.Footnote 132 However, based on the argument I made earlier about the limits on democratic state consent, and especially absolute or minimal core human rights duties, applying the human rights treaty in question in its entirety, and without the invalid reservations, on the mere grounds of the original state consent may be justified.Footnote 133 After all, democratic state consent only amounts to an exception to the obligations arising out of international treaties on other grounds and an exception whose justification is not necessarily granted, especially in the context of imperative and core duties of international human rights law.
Secondly, disagreement and the rules on treaty interpretation (Arts. 31–32 VCLT).Footnote 134 Treaty interpretation is another context in which disagreement may re-appear despite the primary consent to be bound and agreement to the content and wording of the treaty. Articles 31 and 32 provide various canons and methods of interpretation for states to use in the process of ‘auto-interpretation’ of their treaties. They are meant to manage potential disagreements about their meaning, but remain very broad and pluralistic in the constraints they set on the interpretation of treaties.
There is one principle, in particular, that draws attention in this context, and that is Article 31(3) VCLT. It envisages different ways in which the subsequent agreements of the state parties, when different from the original one, may be used to revise the content of the primary agreement (the treaty in question), whether they actually pertain to the latter (a) or more generally (b). This is a confirmation, to quote Klabbers, of the idea of treaty practice as one of ‘continuing negotiations’.Footnote 135 Further, the reference to ‘any relevant rules of international law applicable in the relations between the parties’ under Article 31(3)(c) VLCT confirms that the kind of concurrent and continuing agreement at stake in treaties (but also outside treaties) amounts to more than a concurrence or sequence of mere overlapping consensus or strategic compromises. Or else it could not be squared with the requirement of normative coherence in interpretation across international law.
Finally, disagreement and the rules on intractable treaty conflicts (Art. 30(4) and (5) VCLT). Treaty conflicts are another area of resurgence of disagreement once a treaty has been concluded. While the VCLT entails rules of conflict (e.g., Art. 30(1) to (3) VCLT) that may tame fears about the ‘fragmentation’ of international law,Footnote 136 it also refers, in Article 30(4) and (5) VCLT, to the possibility of intractable conflicts between treaties with non-identical parties or deliberately created treaty conflicts.Footnote 137 Importantly, these provisions offer no clear rule of conflict and hence no way to resolve those conflicts. They acknowledge rather that consent may sometimes be used to agree to disagree and not to agree completely on a given issue. To quote Surabhi Ranganathan, treaty conflicts of this kind are best approached as ‘conduits for legal change’.Footnote 138
6.2. Consent and disagreement in international customary law-making
By contrast to treaties, customary international law is often regarded as less ‘consensual’, at least to the extent that the practice and opinio juris need only be general and not unanimous for a custom to arise.
This does not mean, of course, that the two elements, i.e., convergent practice and opinio juris, are not consent-oriented: they are.Footnote 139 Still, consent need not be secured completely for custom to arise out of that ‘consensus’. As a result, customary international law may seem prima facie to bind some states without or even against their consent. Importantly, however, this is only the case provided a state does not object expressly and persistently to the emerging consensus (persistent objection). It may therefore dissent, and so-doing it can withdraw the consent it was otherwise giving tacitly.Footnote 140
To that extent, like international treaties, customary international law cannot impose obligations on states without their consent.Footnote 141 However, unlike treaties, customary international law-making does not start with an agreement which may then be nuanced through various agreements to disagree. It starts despite disagreement, or with some form of tacit agreement, but that disagreement/agreement may then be set aside, however, through an express disagreement (to agree). One form of consent is active or express, while the other is passive or tacit. Another difference is that the expression of state consent in customary law-making can be unilateral (and juxtaposed) and not necessarily mutual, whereas it is often, albeit not necessarily, mutual in treaty law-making.
These differences between the place of democratic state consent in the two sources lies arguably in their respective scope. Because customary international law has a universal scope, individual state agreement cannot realistically be expressed and hence should not be. In treaty law-making, by contrast, the scope of states parties is more limited or at least, in the case of multilateral treaties, spread out over time, thus making the actual expression of agreement feasible and hence required from the start.
7. Conclusions
This article opened with a paradox: international law-making is ridden with reasonable disagreement and yet no state can be bound by international law without its consent and hence without agreement.
Breaking away from the pragmatic resignation that prevails among international law scholars on this question, I have proposed an interpretation of the role of state consent that both fits and justifies the central role of state consent in the practice of international law-making and, hopefully, may help strengthening the latter's legitimacy in the future. The justification put forward actually lies in the circumstances of reasonable disagreement among democratic states. It thereby dissolves the paradox.
I started by arguing that, in international law as domestically, consent is neither a criterion of validity of law nor a ground for its legitimate authority. I then explained how state consent is related to and required by neither legal positivism nor state sovereignty. Unlike what was the case earlier in the history of international law, both can be defended today without endorsing state consent as a basis for international legal obligations and vice-versa.
Instead, the role of democratic state consent, I argued, is that of an exception to the legitimate authority of international law and hence to its bindingness in a concrete case. While the legitimacy of international law is not democratic, the democratic nature of states and their democratic accountability to their people need to be taken into account in international law-making. This is especially the case in the circumstances of widespread and persistent reasonable disagreement that prevail among and within democratic states in international law. This is best done, I have argued, by respecting the sovereign equality of democratic states and hence their consent as the way to grant an equal voice to their people. Of course, there are limits to the democratic state exception that are inherent to both its democratic dimension (and have to do with basic political equality) and its consensual dimension (and have to do with consent being expressed in a free, fair and informed fashion). When the limits apply, the exception of state consent is lifted and the state is bound by the international law norm whose validity and legitimacy are justified independently.
I concluded by showing how the proposed disagreement-attuned account of democratic state consent fits and justifies the main international law-making processes, i.e., treaties and custom. Despite being both based on state consent, albeit in an active or express way for treaties and in a passive or tacit way for custom, they both feature traits of disagreements to agree that cannot be accounted for otherwise. They also provide ways to manage reasonable disagreement, as it is the case in domestic law, and those mechanisms are in need of further exploration.
