1. Introduction
When we design a community's law-making processes, we have reason to opt for processes that keep social inequalities from affecting the ways in which the community makes its law. Decision-making in the form of a democratic vote is the typical example of a process that responds to that reason. The ‘one agent – one vote’ metric has a strong grip on our intuitions about legitimate political decision-making in modern constitutional democracies precisely because it aspires to exclude social disparities amongst citizens from affecting their citizens’ share in distribution of political (including law-making) power. It is not immediately clear how law-making through customary practices could pass that intuitive test. Social practices and conventions may sometimes be wise or good or efficient, but the process of their formation is not inherently democratic or egalitarian in character. This assessment is partially reflected in the fact that custom plays a peripheral role in what we would regard as well-ordered democratic regimes. Even in systems where constitutional law is largely customary – the United Kingdom is an example – the constitutional practices in question owe their normative force to the fact that they pass some test of democratic legitimacy. Either they are customs developed by, and between, institutions (what Bentham called customs in foro Footnote 1) with sufficient democratic credentials, or they are social practices whose bearing on the law turns on the power of democratic institutions to check (endorse, modify, or ban) them under the light of substantive normative standards.Footnote 2 This makes harder to see how one could justify the normative force of customary practices in communities not only marked by staggering social inequalities, but also lacking legitimate institutional controls over the outcome of the custom-making process.
The international community fits this description all too well. It is rife with arbitrary inequalities.Footnote 3 International agents differ widely in their power and ability to influence how other international agents behave. Few states control disproportionately large parts of the world's natural and technological resources. Fewer still have military capabilities that increase their political leverage manifold. These power disparities and their influence on how international agents act cannot be checked by global institutions with sufficient political legitimacy because – with few subject-specific exceptions – no such institutions exist. Yet customary international practices are typically regarded as a source of general international legal duties, binding on all international agents except those that have persistently objected during their formation. How far can a process so exposed to social inequalities, unfair advantages, and power imbalances be justified to its addressees as generating rules with normative force?Footnote 4 In what follows, I will refer to this concern as the ‘justificatory challenge’ for customary international law-making.
The worry that customary international law-making may lack certain normative credentials is not new, but the stakes of the worry, turning out to be correct, have become much higher in recent decades.Footnote 5 Individuals and groups look to international law more than ever. They invest in it to advance crucial global projects such as the maintenance of international peace and security, the protection of the global commons, and the global environment. They see it as a crucial instrument for the elimination of world poverty and the achievement of better conditions for the world's most vulnerable individuals. They rely on it for protection against governmental practices that violate basic human rights. These hopes and aspirations are channelled through more fora than ever before. International courts and tribunals have proliferated, but so have the occasions in which national political institutions of all three branches are called to interpret and apply norms of customary international law. At the same time, those institutions increasingly find themselves under pressure to refuse to give effect to customary international norms borne out of an apparently illegitimate political process.Footnote 6 To find out how far the investment in international law is worthwhile and whether national institutions have reason to underwrite it, we need to determine not only whether the substantive norms of customary international law are good,Footnote 7 right,Footnote 8 impartial,Footnote 9 or efficient,Footnote 10 but also how far the process for creating that law is capable of meeting the justificatory challenge. That challenge is not ‘apologetic’. Its concern is not to legitimate the current conditions of international society. It is, rather, to see whether international political structures that offend against our moral and political sensibilities, but are not likely to disappear in the near future, might be put at the service of genuine values.
In fact, closer attention to those structures is useful not just to those who want to question the normative force of customary international law-making, but also to those who want to defend it. Take the charge that customary international law-making is undemocratic. One obvious response to it would be that democracy is a virtue of certain particular structures of governance, rather than a virtue of all decision-making in a society, national or international. While we have reason to insist on democratic standards in the way government gets to make law, distribute resources, and use its coercive powers, we are less keen to insist on democratic standards when it comes to other decisions, e.g., decisions about who to be friends with or what art forms to patronize, even when the pattern of those decisions over time has a distinct bearing on the shape and the direction of our community, e.g., it makes our community more or less socially and artistically diverse. Maybe some of the questions that customary international law-making is concerned with are closer in character to those questions. Suppose that the practice of state A has encouraged state B to believe that state B is entitled to exercise a right of passage over state A's territory. Why would the question of whether the past conduct of the two states entitles state B to exercise such passage as a matter of right, in case state A subsequently refuses to grant it, be a matter on which states other than A or B should have a say?
The point also cuts against the charge that customary international law-making is inherently inegalitarian. Consider the fact that customary international law-making accords more weight to the practice of states whose interests are especially affected by an emerging practice. By contrast, egalitarian decision-making in well-ordered national communities either makes a certain subject-matter a question of individual rights, or a question of administration, or a question on which all citizens have an equal say.Footnote 11 Typically, though not always, it lacks the intermediate position of making something a matter of collective decision-making in which some participants are allowed a stronger say in virtue of their higher stakes in the subject-matter of the decision.Footnote 12 But maybe customary international law-making comes out well in this comparison. Perhaps some of the decision-making that customary international law-making involves pertains to questions on which any society should allow some agents to have a stronger say than others, in virtue of their special interest in the question at hand, or the higher stakes that the result of the decision-making process has for them. The point, again, is not that doing things through custom is better than doing things democratically, but that we cannot explain what is good or bad about customary international law-making just by pointing out that it fails to meet the familiar standards of equality and democracy.
The article falls into five sections. Section 2 gives more definition to the justificatory challenge. Sections 3–6 discuss whether that challenge might be met by appeal to the ideas of the common good; consent and ‘framed choice’; the protection of reasonable expectations; and fair play. My core contention will be that although those ideas can justify the force of some types of customary international practices, we have no reason to think that any one of those principles can justify all customary practices that are typically taken to have such force. Accordingly, instead of proposing a unifying justification for all customary international law-making, I will suggest that the impact of past international practices on the normative situation of international agents depends on the nature of the practical problem that those practices are called to resolve. If that is correct, the enquiry as to whether customary international law-making meets the justificatory challenge must proceed on what I will call a ‘disaggregative’ basis.Footnote 13 The conclusion considers how this view relates to the International Law Commission's recent debates on whether different types of customary rule may be formed in different ways.
Three caveats. First, although I will propose a way of thinking about the normative force of customary international law-making, I will not make firm claims about whether customary norms are more justified in certain areas of international law, e.g., the law on the use of force, than in others, e.g., the law on human rights. In fact, it is part of my thesis that such claims can be plausibly defended only through close attention to the structure of the practical problem that the each of those practices addresses. Second, I stake no general claim as to whether some of those practical problems are best addressed through the past practices of international agents, or in some other way, e.g., by means of treaties.Footnote 14 However, I will suggest that the resolution of at least some practical problems may require a level of specificity or density of practice that will, as a general matter, only be achievable through the conclusion of a treaty.Footnote 15 Third, I will avoid casting the moral questions which I will consider as questions about the ‘legitimacy’ of customary international law-making. I do this for purely practical reasons. The idea of legitimacy is powerful but malleable, and disentangling the different strands of its use in contemporary legal and political debates would require adding more by way of throat-clearing in what is already a long piece.Footnote 16
2. Re-formulating the justificatory challenge
We may ask: what justifies the law-making force of customary practices? This is a justificatory question. But it is not a very helpful one. For a start, the question implies that customary practices make law and that what we need is the explanation for their law-making character. However, that may be false. Perhaps customary practices do not make law, or make law only sometimes, or under some special conditions.
Secondly, asking whether law-making by way of customary practice is justified or legitimate assumes that a pattern of practice is sufficient to determine how such law gets made and what it requires or allows of international agents. This is a widely held view, but it is not necessarily correct. Perhaps determining the content of customary law, i.e., the output of the customary law-making process, requires us to take account of certain normative considerations too.Footnote 17 We would therefore do well to ask the justificatory question in a way that does not exclude the possibility that normative considerations play a role in the formation of customary international law. One way to achieve this is to ask not whether it is legitimate for customary practices to make law, but whether customary practices ought to be a core determinant of the content of customary international law.
Thirdly, the idea that customary international law ‘binds’ suggests that the process for making such law gives rise to ‘conclusive’ or ‘exclusionary’ reasons for action, i.e., that once it is determined that customary international law requires X, international agents ought to do X no matter what other reasons might apply to their situation.Footnote 18 This claim might be true, but it does not speak to the question of the justification of customary law-making. The fact that a decision-making process makes law is not an argument in favour of that decision-making process. Rather, we take that fact that this process makes law as raising the stakes of justifying it properly, or of getting its design right.Footnote 19 Similarly, the legal authority, if any, of the output of the customary international law-making process is not an argument in favour of that process, but a parameter of the problem of justifying how something as important as law-making could be left to custom. So, I propose that we adopt the more modest claim that customary international law creates reasons for international agents, or – in the phrase I will use here – that it changes their normative situation. We can leave aside for the moment whether those reasons are exclusionary in character, how they are properly characterized (moral, impartial, self-interest based etc.), and how they relate to other reasons that apply to international agents. In fact, this more modest position is in line with many familiar justifications of law-making by way of customary practice. Some say that customary law binds because states have consented to it. Others say that it binds because states ought not to disappoint the reasonable expectations that their past conduct has created in others. But neither of those views says that the reasons identified by the principles of consent or the protection of reasonable expectations carry exclusionary force.
At the same time, we should note that customary international law changes the normative situation of international agents, when it does, in two distinct ways. When customary international law requires X, it is true not only that international agent A ought to X, but also that other international agents may be entitled to take certain practical measures to get A to X, or to make repair for its failure to X.Footnote 20 Similarly, when customary international law allows Y, it is true not only that international agent A is entitled to Y, but also that agents adversely affected by A's Y-ing are not entitled to take certain practical measures to prevent A from Y-ing. In short, customary international law changes the normative situation of international agents by providing reasons for some action (or omission, but I will let this lie), and by providing reasons why agents could or could not legitimately take practical measures to get others to undertake such action. A worked-out justification of customary law-making should therefore come with an explanation of the relationship between those two sets of reasons.
Fifthly, the idea that ‘customary’ international law binds encourages us to think that, in order to count as facts that determine the content of international law, the past conduct and attitudes of international agents must meet some prior test of customariness, e.g., a wide spread over the population of international agents, a measure of external uniformity, and persistence across time. That assumption too is controversial for a number of reasons. First, it is not true that conduct that fails to meet such tests does not make an impact on the normative situation of international agents. Conduct by a handful of international agents might not suffice to effect a general change in customary international law, but it might sometimes change the normative situation as between the agents that make up the handful – to create, as it were, a local or oligolateral customary norm. Second, statements about the ‘uniformity’ or otherwise of the conduct of international agents must employ some criterion that determines which aspects of that conduct are significant for the purposes of customary international law. And insofar as statements of the uniformity of some practice, or the lack thereof, are offered as reasons that support a certain view of the content of customary international law, such statements – and the criteria of significance on which they are based – must be normative in character. It follows that saying that the conduct of international agents can only determine the content of international law if it meets some prior standard of uniformity will necessarily involve normative commitments of the sort that we want to avoid as a starting point.
Finally, while states and international organizations are the most obvious candidates for the position of ‘international agents’ whose past practices affects the content of international law, we have no warrant for assuming that they are the only such candidates. That is, we cannot assume without begging the question that the identification of those agents is what Ronald Dworkin has called a matter ‘exogenous’ to the justificatory challenge itself.Footnote 21 Instead, we can get over the problem of definition by letting the principles that determine why past practices carry normative force to tell us whose practices carry that force. This may, for example, open up the possibility that the content of the law in some areas, e.g., international investment law, is affected by the practices of agents other than states, e.g., by international courts and tribunals, professional associations, individuals, and so on.
For these reasons, I think that rather than ask whether customary law-making is justified, we should ask: under which conditions may the past conduct of international agents affect how these agents ought to act and whether other agents may take practical measures to get them so to act? Subsequent references to the justificatory challenge in the paper will be references to this, hopefully not unnecessarily cumbersome, formulation.
The next four sections discuss three different types of answer to this normative question and consider how far each might be able to back up the claims international lawyers typically make about the normative force of customary international law in particular situations. Even without going into that discussion though, it seems to me that we can safely say two things about the reconstructed justificatory challenge.
One is that it would be utterly surprising if there was a single answer to that challenge. To take a simpler setting, if you ask me how my own past conduct may affect what I ought to do when others may take practical measures to get me to do it, I am not sure I could do much better than to talk about examples of particular ways in which all that may happen. Instead of giving you a Grand Theory of the Normative Effects of Past Conduct, I would talk about the effect of promises and other assurances I have given to others; of any expectations and reliance that my conduct has given rise to; of my and/or others’ participation in co-operative schemes that produce shared benefits; of my and/or others’ participation in social structures that promote and sustain a certain distributive pattern etc. Similarly, asking how the past conduct of international agents may affect what they ought to do and when others may take measures to get them to do it does not seem to be the kind of question that admits of a general and comprehensive answer. That, I think, is no coincidence. To put the point in the abstract, we can only hope to estimate how agents’ conduct changes their normative profile against whatever background reasons apply to those agents. I have a background reason to keep my promises and that is why my conduct in making a promise to you has the effect of putting me under an obligation to perform. If I did not have such a background reason, my making a promise would not have resulted in an obligation. Furthermore, unless we have warrant for thinking that the background reasons that I have can be captured in a single and comprehensive normative proposition, we should not expect a single and comprehensive answer as to how my conduct may change the rights and duties I have towards others.Footnote 22 Consider this a first defeasible indication that a satisfactory justification of customary international law-making will need to be ‘disaggregative’ in character.
The other thing we could say about the reconstructed justificatory question is that there is no obvious reason why plausible answers to it must involve an appeal to ideas like democracy or equality of decision-making power. Maybe the background reasons that justify why past conduct has the normative effect that it does relate to some basic moral duties, e.g., the duty not to disappoint the reasonable expectations one has created in others, the duty to keep one's promises, the duty of fair play in schemes of social co-operation, and so on. The next sections turn to some of those basic moral reasons. It is, of course, possible that democracy and equality play a role too, either through those basic moral reasons, or independently of them, but that is a claim that must be defended, not a truth that follows as a matter of course from the inherent appeal of those ideas and the fact that customary law-making involves decision-making.
3. Wisdom and the common good
One answer to the justificatory challenge holds that past international practices change what international agents ought to do insofar as there is reason to think that those practices are wise or conducive to the common good. Indeed, the fact that agents have long followed a course of conduct in their relations to each other can often mean that there is something good (useful, expedient, prudent etc.) about that course of conduct. We have some reason to think that rational agents will, over time and under certain conditions of decision-making independence, settle on terms of interaction that are intrinsically desirable and valuable. John Finnis has defended such a view and the argument in its favour has been iterated elsewhere in the literature.Footnote 23
The appeal to the wisdom or the desirability of iterated decision-making seems a plausible candidate answer to the justificatory challenge. Insofar as we have reason to think that certain standards of international conduct have been the result of reiterated interaction between mutually independent and rational international agents, we can say that the standards in question are intrinsically valuable, and therefore, that international agents have reason to abide by them. If the argument worked, it would show an instance where past international conduct changes what those agents ought to do. Of course, even its own terms, the argument leaves open whether this or that customary practice meets all the necessary conditions to be considered ‘wise’. For example, it leaves unclear whether the vast disparities of power amongst international agents allow for the necessary degree of mutual independence in their decision-making.Footnote 24 The real problem, however, is that, even when the relevant conditions are met, that argument cannot provide a general answer to the justificatory challenge. The fact that, under certain conditions, a generally followed decision-making pattern is wise may give an agent a reason to follow the general pattern, but it does not suffice to justify anyone else taking measures to get that agent to follow that pattern. In other words, the mere fact that one is being unwise, or that one's conduct is not conducive to the common good, does not entitle others to hold one to account for not following the wise course of action.Footnote 25 This is not to deny that the wisdom of the general pattern is relevant for the justification of getting someone to follow it. If the intrinsic value of X is relevant in deciding whether an agent can be legitimately coerced into doing X – and we have good reason to think it isFootnote 26 – and if following a settled pattern P is likely to lead one to do X, then the propensity of P to lead to X must also be relevant in deciding whether an agents can be legitimately coerced into following P. The point is that the existence of P is insufficient to justify coercing agents into conformity with the general pattern. It follows that, at best, the ‘wisdom’ argument can only work in tandem with further considerations.
4. Consent and framing
Consent has long been thought to constitute one such consideration. An agent may be required to conform to a pattern of conduct P and other agents may be justified in taking measures to get that agent to conform with that pattern insofar as that agent has consented to those things. Many accounts of the normative force of past international practices seek to justify that force by appealing to this idea. These accounts propose that what justifies the force of a customary practice is the fact that international agents have chosen to endorse it. They differ, however, on their explanation of why consent matters.
One school of thought holds that consent matters in the international arena because the absence of a world government always poses a risk to international peace and co-operation. International law is build on a fragile horizontal structure and this entails that the fruition of any important global project cannot rely on the presence and powers of central political institutions (since, with sporadic exceptions, there are not any), but must instead depend on the willingness of international agents to co-operate. Coercive measures taken without the consent of the international community jeopardize those co-operative structures.Footnote 27
This argument is plausible, but it ends up proving either too much or too little. If the argument says that coercive enforcement jeopardizes international peace and co-operation when it is taken without the consent of the international agent who is threatened with it, then it proves too much. International peace and co-operation can be jeopardized by attempts at coercive enforcement even when the agent against whom coercion is used has at some point in the past consented to its use. If the argument says, more plausibly, that coercive enforcement jeopardizes international peace and security when it is taken without the consent of a part of the international community that is sufficiently strong to absorb the shock of conflict in the event of enforcement, then it proves too little, since it still allows that coercive enforcement may be legitimate against international agents who have not consented to its use.
Another view sees consent as drawing its intuitive plausibility from a distinctly egalitarian aspiration: international agents may be unequal in their power and resources, but the customary law-making process treats them as equals in the sense of giving all of them the opportunity to choose to endorse a practice or to opt out of it. As Malcolm Shaw puts it:
Custom (. . .) mirror[s] the characteristics of the decentralised international system. It is democratic in that all states may share in the formulation of new rules, though the precept that some are more equal than others in this process is not without its grain of truth. If the international community is unhappy with a particular law it can be changed relatively quickly without the necessity of convening and successfully completing a world conference. It reflects the consensus approach to decision-making with the ability of the majority to create new law binding upon all, while the very participation of states encourages their compliance with customary rules.Footnote 28
This assessment is open to objection that the notion of consent is too thin to do the required justificatory work. Consent is not always sufficient to change an agent's normative situation. Coerced consent to past practice or consent extracted by fraud are obvious illustrations. Saying that consent to past practice must be ‘free’ or ‘voluntary’ does not improve things much, since it is hard to think of any decision to endorse or to reject a practice that will not have been influenced by pressuring factors.Footnote 29 We want to distinguish between legitimate and illegitimate sources of pressure, and the appeal to the notion of consent cannot achieve that differentiation on its own.
I think we should agree with T.M. Scanlon that our intuitions about the legitimizing force of consent are better accounted under a more nuanced idea. The reason why agents may be required to bear a certain practical burden, such as to conform to a practice on pain of having certain practical measures taken against them, is not that those agents have consented to that burden, but that that institution or decision-making process that generates that burden allows those agents the opportunity to affect their obligations through their choices, and this opportunity is something that those agents have reason to value.Footnote 30 Applied to customary practices, the ‘value of choice’ idea entails that these practices can be a legitimate source of burdens on the part of an international agent when that agent had the opportunity to shape its obligations by taking an attitude towards those practices, and that opportunity was valuable to that agent. This allows us to draw morally important distinctions between the situation of an agent who supports or does not object to an emerging practice for fear of being subjected to illegitimate coercive measures, and the situation of an agent who supports or does not object to that practice on the strength of the benefits that it stands to receive under it, or in order to snuff out an alternative practice that it finds even more objectionable. It also allows us to distinguish between agents who are silent in the face of a widespread practice because they cannot afford the resources to make sustained diplomatic representations against it, and agents who are silent because they are content to follow developments from the diplomatic sidelines. The difference between coerced and financially strained agents, on the one hand, and benefiting and acquiescent agents, on the other, is that while both groups of agents have the opportunity to form an attitude towards the practice, that opportunity is something that only agents in the latter group have reason to value.
This basic setup can help refine our doubts about the justification of customary law-making. The source of those doubts, I think, is that customary practices can allow what I will refer to as ‘unjustified framing’. Let us say that I frame you when I act so as to limit your options or to increase their relative cost, in order to get you to act in a certain way, or to lead you towards or away from a certain choice.Footnote 31 For example, I lower the price of my goods in order to drive out the competition; I set voter registration and identification requirements; or I declare exclusive jurisdiction over a certain part of the sea and its subsoil. I will assume that the following two propositions are true of framing. First, framing is legitimate only if it can be justified towards the framed agent. Second, the fact that the choices left to the framed agent are intrinsically good is not generally sufficient to justify the framing act or practice.
This description of framing is wide enough to apply to both formal decision-making of the sort we find in well-ordered democratic regimes and to the formation of customary practices. In the former setting, the political choices we make as citizens are typically framed in one way or another, from the way the ballot is organized (e.g., in favour of parties rather than specific policies; elections every four years rather than, say, every year etc.) to the availability of choice of particular political parties or candidates (e.g., only parties that have been registered; quotas for female candidates etc.). These measures limit the alternatives open to us and are intended to lead us to exercise our political power within certain confines. We consider them justified not insofar as we have consented to them (almost none of us have), but insofar we have reason to value having a choice on the questions that these measures leave to us. Some of the more complex debates in our democracies are concerned about the value of having certain choices and not having others.
The choices that international agents face in the customary law-making context can be similarly ‘framed’ in a variety of ways. To take an obvious example, the fact that states A, B, and C have embarked on a certain practice may affect the cost of silence for state D, in the sense that D may be taken to have acquiesced in the practice and therefore to have become bound by it. If D wants to avoid being bound, it must declare its objection to the practice at an early stage and to maintain it across time (it must, in the jargon, assume the role of a ‘persistent objector’). But this option may have become more expensive as well, since it carries the risk of alienating A, B, C, and all other agents who may have jumped on the bandwagon, therefore limiting D's ability to co-operate with them.Footnote 32 The frame within which D has to make its choice of political attitude towards the practice could be even tighter: perhaps A and B have indicated that, should D not support the new practice, they will cut down on the aid they are supplying to it or they will increase tariffs on D's exports. In these scenarios, D has to make a choice within a frame set by the practice of other international agents. For D's choice to be taken as a basis of holding it bound by the practice instigated by A, B, and C, it must be the case that the framing of D's choice by means of that practice can be justified towards D, and this will depend on whether having that framed choice in the situation is something that D has reason to value.
This perspective can help us to understand better the character and to assess the force of familiar complaints about the customary international law-making process. One familiar complaint is that new states may not legitimately be bound by customary law that was in place before those states were created.Footnote 33 Under the account I am proposing, that complaint would be justified insofar as the opportunity to participate in the customary law-making process would be something that new states do not have reason to value. It seems to me that the position of new states is rather different. Consider the well-documented objections of developing states to the requirement of ‘prompt, adequate, and effective’ compensation in the context of nationalization of natural resources.Footnote 34 The usual way of understanding these objections is to say that developing states claim not to be bound by customary law made before their ascent to independence because they were not afforded the opportunity to express their consent to or dissent from it. Under the account proposed here, we should understand developing states as putting forward a more nuanced claim: that the process of customary international law-making frames their choices in an illegitimate way, to the extent that it treats their objections as efforts to change customary international law (thus placing on them an unfair onus of having to convert other – possibly recalcitrant – agents towards their viewpoint), rather than as contributions to be weighed equally alongside older practice.
Similar considerations may account for the objections international agents sometimes voice against the idea that widespread support for certain formally non-binding resolutions in the context of global international organizations may give rise to generally binding norms of customary international law.Footnote 35 The basis of that objection seems to me to be that such a view would allow the choices of these agents to be framed by the practice of others just in virtue of their sheer majority. After all, majorities have no intrinsic claim to be followed, and there is no general reason why their views must be privileged over those of dissenting agents, unless there are good substantive reasons why one must be held to be committed to the result of the vote.Footnote 36
Having recast the justificatory challenge to customary international law-making as a challenge against illegitimate framing of choice, in the following sections I try to see how that challenge shapes the conditions under which such framing might be justified. I consider two candidate principles that might do the required justificatory work: the principle of legitimate expectations and reliance, and the principle of fair play or fairness.
5. The protection of reasonable expectations and reliance
The most widely endorsed account of the normative force of customary international practices locates its source in the basic duty to take care not to defeat the reasonable expectations one has led others to form. As the International Law Association has put it:
a rule of customary international law is one which is created and sustain by the constant and uniform practice of States and other subjects of international law in or impinging upon their international relations, in circumstances which give rise to a legitimate expectation of similar conduct in the future.Footnote 37
The principle of protected legitimate expectations seems to account reasonably well for the actual process of formation of a customary practice. On the one hand, it registers the fact that, as is typical in conventional settings, constant and uniform practice on the part of many international agents will tend to create some general presumption that other agents too will follow suit. On the other hand, the principle does not protect any expectation of similar future behaviour; any such principle would clearly be reasonably rejectable on the part of international agents who have yet to commit to the practice. The principle only protects ‘legitimate’ expectations, i.e., only expectations that international agents are justified in having.
It could be objected that this last feature of the principle renders it circular: after all, the principle appears to say that an international agent is entitled to claim certain customary international rights as long as that agent is entitled to expect that it will enjoy such rights. That circularity can be avoided through a more relaxed reading of the condition that agents be ‘justified’ in having a certain expectation. We should understand this condition as requiring that agents engaged in a customary practice have some reasonable grounds to interpret the fact that other international agents have not opposed the practice as an endorsement of that practice. So understood, the condition should be relatively uncontroversial. Barring any special circumstances, the fact that, despite knowing about it, you have not objected to the shortcut I have been taking through your farm every day for the last two years gives me reasonable grounds to believe that you have licensed my actions and leads me to entertain a reasonable expectation that you will continue to do so in the future. What constitutes a reasonable ground for an expectation will, of course, differ from case to case, so we should not demand that the principle of protected expectations produce a complete specification of ‘legitimizing’ circumstances. However, as long as the general statement of the principle avoids the charge of circularity, it looks a plausible as a candidate normative basis for the binding force of customary international practices.
Nevertheless, I want to suggest that the principle of protected legitimate expectations cannot bear this justificatory burden for two related reasons. First, the duty imposed by the principle is too wide to be normatively appealing. Second, a narrower and more plausible version of the principle would not justify some of the most typical claims about the binding force of customary international practices.
Consider one of the best-known instances of ‘local’ or ‘special’ international custom, the Right of Passage case.Footnote 38 Portugal claimed that India was bound by a local custom to allow civilian transports between two Portuguese enclaves through its territory. The International Court of Justice held that the custom had been created through a long history of interaction between Portugal and British India, during which the passage of Portugal's convoys through Indian territory had gone unopposed by the local authorities.Footnote 39
How did India's lack of protest towards the passage of the Portuguese civilian convoys generate an obligation on its part to continue to allow such passage? The legitimate expectations principle provides an intuitive explanation. India's failure to object to the frequent passage of Portuguese civilian convoys had led Portugal to expect that it had the option of channelling civilian traffic between its two territorial enclaves through Indian soil. Therefore, it would have been wrong to allow India to defeat the expectations it had led Portugal to form.
This explanation leaves an important question unanswered. The Court's judgment assumed that Portugal's practice had the effect of imposing on India a ‘duty to speak’ and that India's eventual silence was a legitimate basis for Portugal to expect that it could claim civilian passage as a matter of right. Putting the issue in terms of framing, we would say that the Court's view of local custom allowed Portugal to frame India's choice by making its silence more costly: the silence now carried the implication that India had consented to allow passage as a matter of duty. How could framing of this sort be reasonably justified towards India?
To start with, Portugal's expectations would draw justification from the plausible general idea that agents should ‘guard’ their rights against act or practices that impinge or credibly threaten to impinge on those rights. The repeated passage of civilian convoys through Indian soil without prior authorization clearly impinged on India's right of territorial sovereignty. If India considered this practice to be a violation of that right, it ought to have made a protest to that effect. However, it seems equally clear that India could have reasonably rejected a principle of local custom that allowed Portugal to frame its silence as incurring a duty to allow passage, if the cost of protest against Portugal's practice had been significant, e.g., if protest would have exposed India to some real threat of suffering adverse consequences. In the circumstances, there was sufficient evidence that protest would not have been too costly for India: when India protested at the passage of armed forces through its territory, Portugal proceeded to ask for permission for subsequent passages.Footnote 40 At the same time, it would be misleading to suppose that India's silence created a right of passage on the part of Portugal simply because the latter reasonably expected passage to go on unopposed. Such a duty would be subject to reasonable objections on the part of both agents. On the one hand, it would not allow any ‘right of exit’, i.e., room for the agent that has encouraged the expectation to escape being obligated through timely notice or some other equivalent gesture.Footnote 41 Similarly, a principle of local custom that justified Portugal's right in terms of its expectations would presumably hold India bound only as long as the expectation of unopposed passage persisted. Such a principle would therefore allow India to shake off its duty by announcing (with advance notice) its intention to prohibit passage for the future. However, Portugal could have reasonably opposed such principle on the ground that it had relied on India's conduct in arranging its administration of the two enclaves, which were completely surrounded by Indian territory, and would incur a significantly increased cost in finding alternative routes of civilian transport between them. A principle of local custom protecting Portugal's reliance would in turn be justifiable towards India only to the extent that the duty of allowing passage was necessary to meet the costs of Portugal's reliance. On the facts of the case, this condition was apparently satisfied, since Portugal claimed a right of passage only ‘to the extent necessary for the exercise of its sovereignty over the enclaves’.Footnote 42 In that regard, it seems to me to have been a contingent fact of the case that Portugal's reliance, in the circumstances, was of such nature that, having no other means of transit between the enclaves, it committed India to continue providing a right of passage for as long as Portugal retained sovereignty over those enclaves. Had Portugal been able to access alternative routes, India could have reasonably claimed to be released from its duty to give passage once it had given adequate advance notice of its intention to do so.
Here, then, is what I think we should take as the justification of the kind of framing that the principle of local custom allows. Instead of invoking a generic duty to honour expectations an agent has led others to form, we should endorse the more nuanced and demanding principle that a practice intended to frame another agent's options and impinging on that agent's interests can become the basis of a duty for the framed agent when that agent has a reasonably inexpensive option of avoiding coming under that duty, and the framing agent has reasonable grounds to rely on the framed agent's choice not to exercise that option.
This principle of protected reasonable reliance justifies why an agent's practice may become the source of duties for another. Note, however, that the range of cases that it covers is quite limited. The principle does not justify any general statement to the effect that an international agent is committed to follow a certain practice either on the ground that many other agents regard the practice as obligatory, or on the ground that other agents might expect that agent to follow suit. In fact, the principle of protected reasonable reliance does not even allow international agents to ‘read’ another agent's silence as acceptance of a duty to follow the practice, unless that agent had a reasonably cheap option to object to the practice and these other agents have somehow relied on its choice to forego that option.
The most obvious upshot of the limited ambit of the protected reasonable reliance principle is that it does not offer a justification for the most central tenet of customary international law, namely that widespread international practice can create generally binding law. In fact, the situations where international lawyers typically affirm the existence of a generally binding customary international practice do not seem to fulfil any of the principle's main requirements. Widespread practice is regarded as binding ‘silent’ agents even when the practice impinges only on potential interests of those agents: for example, general practice on the right of military ships to innocent passage can bind landlocked states that happen to acquire naval forces only after the practice has been formed. Furthermore, the fact that an international agent may incur significant costs in protesting against the practice is not regarded as a ground for exempting that agent from the binding force of the practice: for example, a state that fails to object towards the practice of demanding ‘prompt, adequate, and effective’ compensation for the expropriation of foreign investment for fear of undermining its chances to obtain funding from the World Bank will be regarded as no less bound by the practice than the states that instigated it. Finally, agents claiming that a ‘silent’ agent is bound by widespread and constant customary international practice do not normally need to demonstrate that they have relied in any way on that agent's silence in arranging their affairs: for example, a state that considers establishing a consulate in a foreign country is typically entitled to request that its consular staff enjoy the privileges and immunities provided by customary international law, even if the receiving state has expressed no clear attitude towards those customary practices. Any justification for these normative features that customary international practices are generally regarded as possessing would therefore need to be grounded on different considerations.
6. Fair play
A plausible example of a principle that might fit this bill is what John Rawls called the principle of ‘fair play’. Developing a suggestion by Herbert Hart,Footnote 43 Rawls describes the principle as follows:
Suppose there is a mutually beneficial and just scheme of social cooperation, and that the advantages it yields can be obtained if everyone, or nearly everyone, cooperates. Suppose further that cooperation requires a certain sacrifice from each person, or at least involves a certain restriction of their liberty. Suppose finally that the benefits produced by cooperation are, up to a certain point, free: that is, the scheme of cooperation in unstable in the sense that if any one person knows that all (or nearly all) of the others will continue to do their part, he will still be able to share a gain from the scheme even if he does not do his part. Under these conditions a person who has accepted the benefits of the scheme is bound by a duty of fair play to do his part and not to take advantage of the free benefit by not cooperating. The reason one must abstain from this attempt is that the existence of the benefit is the result of everyone's effort, and prior to some understanding as to how it is to be shared, if it can be shared at all, it belongs in fairness to no one.Footnote 44
The fair play principle covers situations where an agent's choices are framed by a co-operative scheme or practice instigated or supported by other agents and explains how far that framing practice can be legitimate towards agents that have not participated in the practice. The essence of the principle is that agents who accept the benefits of the co-operative efforts of others have a duty to undertake a fair share of the cost for producing those benefits. A typical example is the duty to pay the train fare in a public transport system that operates an ‘honour’ scheme: as long as an agent has accepted the benefits of the system by using public transport, that agent has a fair play duty to pay the fare and not free-ride. Furthermore, this duty is clearly distinct from the duty to protect reasonable reliance, since it arises in virtue of the practical success of the co-operative scheme, the fairness of the way its costs and benefits are distributed, and the acceptance of those benefits, whether or not any particular agent has specifically relied on any other to do its share.
The principle of fair play has famously been the subject of intense discussion in the context of the justification of political obligations in national communities. One aspect of that discussion concerns the proper formulation of the principle, or the proper specification of the circumstances in which it justifies the framing of non-participating agents. The other, more controversial, aspect concerns the principle's justificatory scope, especially its ability to account for national political obligation. With regard to the first, it is disputed whether the duty of fair play kicks in when an agent accepts the benefits of social co-operation (as Rawls's formulation suggests), or whether it is sufficient that that agent has received those – potentially unwanted – benefits (as Hart's original proposal implied).Footnote 45 With regard to the second, it is argued that the acceptance of the benefits of social co-operation cannot ground a general political duty to obey a community's laws, since very few citizens can be reasonably held to have signalled acceptance of those benefits: most of us are ‘born into’ the benefits that political institutions provide, e.g., basic infrastructure and security, and disclaiming them is not a practical option.Footnote 46
For reasons that will become clear, it is not important for our present purposes to take sides on those familiar debates. It is enough to note that there is little dispute that a tightly formulated principle of fair play can justify the binding force of certain social practices on agents who have not (yet) participated actively in them. I therefore propose to concentrate on the tight formulation of the principle which requires acceptance rather than mere receipt of the benefits of social co-operation, and examine how far it can justify the binding character of customary international practices. I will leave open the possibility that a more relaxed formulation might have even broader justificatory scope.
It seems to me that several customary international practices, and certainly some international treaty regimes, might be reasonably described as schemes of co-operation that produce common goods, the acceptance of which generates a duty to undertake a fair share of the costs of producing them. The case might be at its strongest in respect of customary practices regarding the global commons, such as the high seas, the deep sea-bed and the subsoil thereof,Footnote 47 the outer space and celestial bodies,Footnote 48 and perhaps certain aspects of the global environment. Maintaining those common resources for the benefit of current and future generations requires an international co-operative effort which is only likely to succeed if most international agents restrict their liberty to exploit those resources individually. Once a scheme of co-operation that can achieve a reasonable measure of success in this aim has been put in place and generates benefits, it is fair to require all international agents who accept those benefits to do their fair share in supporting it by similarly restricting their liberty. So understood, the principle of fair play also makes space for the idea of persistent objection, since it allows international agents to avoid becoming bound by a co-operative practice as long as they clearly choose not to accept its benefits, although the precise limits within which persistent objection may be available will depend on whether one endorses the narrow or the wider formulation of the fair play principle.
Note, furthermore, that a fair play account of the general binding force of those practices could also explain why theorists are sometimes prepared to hold that they give rise to ‘instant’ customary international law.Footnote 49 Under the fair play principle, the length of co-operation in time is not critical: the practice becomes the source of duties as soon as a reliable scheme of co-operation is put in place and begins to deliver its benefits.
Accepting that the principle of fair play might be employed to justify the binding force of certain customary international practices does not entail that all such practices will pass the principle's normative test. A ‘silent’ agent will be bound by customary international practice only if that practice fulfils the following conditions: it must create a scheme of co-operation; the scheme must produce goods that are free in the sense of it being possible for agents to obtain them without paying; the costs of the scheme must stand in reasonable proportion to the benefits achieved through it; and the costs of the scheme must be fairly distributed.
The last condition is particularly demanding, since it reserves normative force only for cases where the costs and the fruits of co-operation are fairly distributed amongst international agents. The problem is that setting on a fair distribution will normally be very difficult in a community without central political institutions. That may add to the explanation why the international community has found it difficult to form generally binding customary practices in respect of, say, environmental protectionFootnote 50 : any schemes of co-operation in this field are bound to impose very different costs from international agent to international agent, e.g., between developed and developing countries, while a fair distribution of those costs would probably require setting-off or means-testing mechanisms of a level of sophistication that the customary process could not possibly be expected to deliver.Footnote 51
Yet, despite the demanding nature of the tests set by the fair play principle, it is clear that at least some customary international practices may be able to meet them. For reasons just given, these will most likely be customary practices that require international agents to abstain from exploiting resources or goods that lie outside their exclusive jurisdiction and, in that sense, are common. However, customary international practices on other areas too might be justifiable under the fair play principle. It may, for example, be possible to extend the principle to some basic customary international practices regarding the use of armed force or other coercive measures, say on the ground that these practices produce the common good of international peace and security. I venture no view as to the merits of that particular argument and I have no grand theory to offer on what might count as an international common good. I can only suggest that, as a general matter, the extensibility of the principle of fair play to any customary international practice will depend on whether it is plausible to regard that practice as producing common goods that non-participating agents might be able to, but should not, free-ride.
Herein lies an important problem. Many customary international practices cannot plausibly be said to produce common goods that non-participating agents might free-ride. Consider, for example, the classical customary practices on jurisdiction, sovereign immunity, or the law of treaties. These practices purport to lay down rules applicable on a basis of reciprocity between any two subjects of international law. They are not co-operative schemes in the sense required by the fair play principle, since their success does not depend critically on the number of international agents that take them up. While a co-operative scheme of, say, abstention from unilateral exploitation of the moon's minerals, that happened to be shunned by most international agents, would not be able to generate any benefits at all, a practice relating to state immunity can yield substantial benefits to however few or many states accept it.Footnote 52 By the same token, these practices do not appear to produce any good that non-participants in the practice might be able to enjoy without being committed to the practice. In other words, these practices cannot really be free-ridden, since agents who do not participate in them cannot enjoy their benefits. It must follow that the fair play principle cannot justify their binding force.
This result leaves a range of typical claims about customary international law without adequate normative support. For example, international lawyers consider it trivially true that, when supported by widespread practice, the customary international regimes on jurisdiction, immunities, or the law of treaties become binding on all international agents (save for persistent objectors) even when those agents have not benefited from those practices in any way. To take an earlier example, a state planning to send its first ever consular mission to Ruritania is entitled to demand that Ruritania extends to its consular staff all the customary diplomatic privileges and immunities, whether or not Ruritania has ever claimed such immunities for its own low-ranking consular officials and, indeed, even if Ruritania has recently declared that it will never claim or recognize diplomatic immunities for such persons. This demand might perhaps be justified as long as Ruritania has led other states to expect that their low-ranking consular officials will be accorded immunity, but that justification will, as we have seen, have its own conditions and limits. What could not be said is that Ruritania is bound because it is somehow gaining a benefit from the practice of immunity and fails to pay the fair share of its cost.
One might argue that even if this or that customary practice does not put in place co-operative structures that generate free-rideable goods, customary law-making as a whole creates such a structure. The argument could go: in governing their relations by reference to their past practices, states put in place a scheme of social co-operation; this scheme generates goods such as peace and stability that are available and useful to both participants and non-participants in that scheme; the cost of conforming to customary practices stands in reasonable proportion to those goods; and the costs in question are fairly distributed.Footnote 53
I think that this broad way of deploying the fair play principle goes awry in its second stage, when it claims that the regulation of international affairs by reference to the past practices of international agents produces the goods of international peace and stability. The problem is not that this argument is necessarily false, but that it risks proving too much. Old practices do not necessarily make the world a more peaceful place. Sometimes they can be a continuing source of tension in it. Again, the appeal of that argument depends on seeing it applied in a disaggregated fashion.
I conclude that, like the protected reasonable reliance principle, the principle of fair play can justify the generally binding character of only some customary international practices, namely those that display the structure of co-operative schemes that generate ‘free-rideable’ common goods.
7. Conclusion
We are not short of perspectives from which to measure the capacity of international law to live up to fundamental moral and political values. We can ask how far the content of international law could be justified on an impartial basis. We can discuss how international law stacks up against the moral demands of basic humanitarian concern for the most vulnerable and, perhaps, of international distributive justice. But we can also focus our critical gaze on the main process for making general international law and ask whether we have reason to think that its outputs carry normative force.
This article is part of that last project. I began by noting that both the defence and the critique of the normative force of customary international law-making must be sensitive to the particular conditions of international society, and the kind of problems that customary law is called to resolve in that society. With that constraint in mind, I reconstructed the ‘justificatory challenge’ as a question about the conditions under which past international practices may change what international agents ought to do and whether other agents may take practical measures to get them to do so. I then considered some possible answers to that challenge. I argued that our doubts about the justification of customary international law-making are best accounted for as concerns with a generic wrong I called ‘unjustified framing’. Accordingly, I distilled those doubts into the following question: under which circumstances can an international agent's choices be legitimately framed by the practice of other agents, so that the framed agent will come to owe duties under that practice? Here, I suggested that the protected legitimate reliance principle accounts reasonably well – give or take the problem of ‘exit’ – for local or special international customs, though not for general practices. The fair play principle accounts for general practices and the idea of persistent objection, but only insofar as these have the structure of a co-operative scheme for the production of a ‘free-rideable’ common good. In the end, instead of offering a unifying answer to the justificatory challenge, I have offered a messier or ‘disaggregated’ account, the upshot of which is that we should not apply the same moral measure to all customary practices. Local practices must be judged for their capacity to protect reasonable reliance, but their force will be similarly limited to the protection of such reliance; co-operative schemes must be judged for their cost-benefit efficiency and the fairness of the ways they distribute those costs and benefits, and so on.
I cannot claim to have exhausted the kinds of customary practices that international law features, or the normative principles that each one engages. I certainly have not made any case to the effect that customary international law may be more justified in some areas than in others. That disclaimer is particularly applicable to areas such as international human rights law which is not obviously covered by any of the principles I have considered. Still, I hope that the disaggregative view points to a way of carrying that discussion forward: it tells us that the significance of past practices for the international human rights obligations of international agents will depend on the nature of the problem that those practices are meant to resolve.Footnote 54 We can therefore expect different formulations of that problem to support different views about how much practice is required to make customary international human rights law, or about the proper distribution of weight between the conduct of international agents and their expressed attitudes.
If the disaggregative view is right, the oft-heard complaints that customary international law-making falls short of basic standards of fairness, democracy, and equality are much too blunt. Given that the past conduct of international agents may affect their normative profile in different ways and in different contexts, any objection we have to statements of the sort ‘states are bound by customary practice P to do X’ will need to be grounded on the particular normative principles. Sometimes, our complaint may be that this statement is false because P has not given rise to the required pattern of expectations and reliance. At other times, our complaint may be that the statement is false because P does not satisfy the requirements of a duty-generating scheme of co-operation etc. The mere fact that the ideas of democracy and equality have such a strong grip on our intuitions does not suffice to make them appropriate standards against which to measure customary international law-making, except when those standards are internal to the problem that this process is trying to resolve.
This makes it easier to explain why the formation of customary law may be subject to different standards, or differences in the application of some more general standards, depending on the nature of the practical problem that a certain customary practice responds to. The recent and ongoing deliberations of the International Law Commission on the formation of customary international law have demonstrated a degree of openness to this idea. On the one hand, the discussions of the Commission make clear that almost all its members consider the two-element conception of custom to apply to all areas of international law.Footnote 55 On the other hand, its members appear content to leave open the possibility that the precise weight of each element may vary depending on the type of rule of international law under discussion. For the reasons I have given, I believe that the disaggregative account shows why the Commission is right to entertain those possibilities, insofar as what explains the differentiation is that some areas of international law and some types of rule may deal with practical problems that display a different normative structure.
Philip Allott has said that self-centred international agents would see in customary international law only ‘an unintimidating ragbag of law-like ideas’.Footnote 56 I have argued that Allott is right that international custom is indeed a ragbag of ideas rather than a normatively unitary entity. However, this ragbag is not altogether unintimidating, as long as each of the elements that make it up has a robust justification.
