Complexity Theory and the Horizontal and Vertical Dimensions of State Responsibility

Abstract

This article argues that a gap that has always existed in the law of state responsibility is now becoming more apparent. That gap divides a state from its citizens, making it difficult to justify why state responsibility should be distributed to them. Purely legal approaches to the issue are not likely to resolve the problem, and although the literature of moral collective responsibility suggests some bases for having citizens share the costs of state responsibility, none are completely satisfying. Concepts from complexity theory show why this is so. If the theory is correct, the state is neither a legal abstraction nor reducible to the individuals who purportedly comprise it. Instead, it is an emergent phenomenon that arises from complex interactions among individuals, formal and informal subgroups, and the conceptual tools and structures that individuals and subgroups use to comprehend and respond to their physical and social environments. The theory is consistent with a basic premise of international law that the state as such is an appropriate bearer of responsibility. However, because in a complex system there is no linear connection between the emergent phenomenon and its underlying constituents, this suggests that the divide between a state and its citizens in the distribution of state responsibility may never be bridged.

1 Introduction

From its inception, international law has had to do with collectives, how to think of them, and how to hold them responsible for wrongful acts. This began of course with international law’s initial focus on states and is even more important now, as individuals, international organizations, and multinational corporations have been brought within international law’s ambit as international subjects or participants. This article explores whether it makes a difference for international responsibility if those subjects or participants are complex adaptive systems, a concept taken from complexity theory. Complexity theory posits that interactions between large numbers of individuals can give rise to phenomena that are not predicted from individual behaviour itself. In a form of self-organization, a whole can result that is more than the sum of its parts, distinguishable from its environment but not impermeable to it. What might that mean for individuals in such a system if the emergent system itself incurs liability?

The doctrines of state, corporate, and individual responsibility have generated large literatures, as has complexity theory itself, and it is impossible to do justice to them here. This article thus focuses on certain aspects of state responsibility to see whether complexity theory has anything to contribute to that area of law. It does so in at least two ways. Complexity theory suggests that the state is neither a legal abstraction nor reducible to the individuals who purportedly comprise it. Instead, the state is an emergent phenomenon that arises from complex interactions among individuals, formal and informal subgroups, and the conceptual tools and structures that individuals and subgroups use within their physical and social environments. This means a gap exists between the emergent phenomenon that is the state and the underlying individual interactions of the citizens and other elements from which the state emerges. The state stands apart from its citizens, and thus complexity theory is consistent with a premise of state responsibility that the state as such is the proper subject of legal responsibility; it is not simply a proxy for agents that act on its behalf or for its citizens, nor is it simply an extension of the citizen.

Although complexity theory confirms this aspect of state responsibility, it also sheds light on what might be an irresolvable problem for it. The gap between the state and the underlying interactions from which it emerges corresponds to a gap that is beginning to show itself in the law of state responsibility. That gap divides the horizontal responsibilities states have towards other states from their vertical responsibilities towards their citizens to protect, and sometimes further, such citizens’ human rights.¹ State responsibility tends to ignore this vertical dimension and has put aside questions about the ontology of states; thus it would seem unimportant whether or not a state is a complex adaptive system. However, for all of the practical and conceptual benefits of separating law from ontology, there are practical and conceptual costs too. The burdens of legal responsibility are felt in the real world, at the level on which most of us live, and if indeed there is an ontological and legal separation between the state and the individual, it is unclear why those at the citizen level should shoulder those burdens. Certain changes to the doctrine can be made to address this problem better; however, it may be that the gap between state behaviour and the people who bear responsibility for it will never be fully bridged. If so, any rules of state responsibility that distribute costs to citizens will be to some extent incomplete and arbitrary. As a result, complexity theory provides an ontological explanation of why the problem of distributing state responsibility downward to its citizens may be intractable.²

These arguments are set out in three sections. Section 2 discusses the tensions between state responsibility to other states and to its citizens;³ state responsibility pays little attention to this tension, and because standard legal solutions to the distributional problem are not likely to be fully satisfactory, for guidance the article turns to the literature on moral collective responsibility. That literature indicates that a respons ible state may be justified in curtailing the rights of its citizens if they somehow are complicit in the state’s action. However, although the literature provides some bases for such culpability upon which one might ground legal justifications for diminishing citizens’ rights, those bases are not uncontested, and there are enough differences between moral and legal norms for such justifications to be incomplete. Section 3 introduces certain concepts from complexity theory to explain why the problem is so difficult. As discussed, the theory suggests that in a complex adaptive system, the whole is not the sum of its parts. There will always be a divide between interactions on the individual level and emergent state behaviour. One result is that it will be difficult, if not impossible, to justify distributing the costs of responsibility to citizens, particularly if this involves the violation of their rights. This is obviously problematic, and the article thus responds to jurisprudence that urges that we lay aside ontology and turn instead to purely legal conceptions of the state. Section 4 outlines some of the implications of these findings, as well as steps that could be taken to ameliorate the problem to some extent.

2 State Responsibility as a Problem of Collective Responsibility

A The Problem in Context

In recent years, scholars have raised concerns about the potential for conflict between a state’s responsibilities to other states and other international actors on the one hand and its responsibilities to its citizens on the other. The underlying issue is not new, but has become more salient because of two well-known developments in international law. It is now a given that states are obligated to protect and, to some extent, further the human rights of individuals. There are, for example, 161 parties to the International Covenant on Economic, Social, and Cultural Rights. Among the rights recognized in the Covenant are the rights to work, social security, an adequate standard of living, the highest attainable standard of physical and mental health, and education.⁴ Rights are not always absolute, and under the Covenant some are limited by available resources and are to be progressively achieved.⁵ Nevertheless, such rights continue to evolve: some states have adopted them as part of their internal law,⁶ and the available resources exception implies that rights can take on a harder cast as such resources do in fact become more available.⁷

The growth of human rights law has accompanied a more general concern about the impact of international law on people. This is reflected in criticisms about the democratic deficit in international law, as well as the critique by some Third World scholars that international law has been used to impose particular economic and political structures on the developing world, with resulting doubts about international law’s legitimacy. These impacts are often economic in nature. To home in on the concerns addressed here, several scholars have expressed worries about the possible impact of arbitral awards, often arising from private actions brought under bilateral trade or investment agreements, on the human rights of citizens of respondent states.⁸ Bilateral investment and trade agreements have mushroomed; there are some 3,000 as of this writing, and many of them give private rights of action against the host state. This has led to an increase in actions in which states have found themselves potentially liable for large damages awards.⁹ Then there are other forms of liability. For example, in connection with claims arising from the Iraqi invasion of Kuwait, the United Nations Compensation Commission has awarded compensation of $52 billion, paid through the sale of Iraqi oil.¹⁰

Sometimes the consequences of state responsibility are negligible from the perspect ive of the citizens of a responsible state,¹¹ but when a state has fewer resources and the award is high, it is possible that its citizens will feel their impact:

When a state pays compensation to its victims, the material losses it suffers pass on to its citizens, either directly or indirectly. By direct transfers I mean that the government taxes citizens more, to cover for the assets it lost. By indirect transfers I mean that the government covers for its losses using resources that would otherwise have been used for other goods and services, thus lowering the overall welfare level of the population.¹²

Under some circumstances, the direct and indirect transfers a state makes in satisfaction of its horizontal obligations to other states (or diagonal obligations in the case of payments to private parties) may well interfere with its ability and duty to meet its vertical obligations to its citizens.¹³ Monies paid to satisfy an obligation to another state or the private parties associated with the other state are monies that will not go towards medical care, education, and the like, things that have been recognized as international human rights or are emerging as rights.

As I discuss below, the potential conflict between a state’s horizontal and vertical responsibilities can and has been minimized to some extent, but the problem does not completely disappear. The complexities of the issue can be seen in a situation faced by Paraguay. In 2010, the Inter-American Court found that Paraguay’s failure to provide access to water, food, health care, and education to certain members of the Xákmok Kásek Indigenous Community in ‘conditions of special, real and immediate risk’ had violated their right to a decent life, to the detriment of the entire community.¹⁴ The Inter-American Court also found Paraguay responsible for the deaths of 13 individuals who had died from various diseases without medical care.¹⁵ Even under a narrow interpretation, the decision indicates that under extreme circumstances, the failure to provide adequate services to particularly vulnerable groups or individuals can violate their human rights.

During the Inter-American Court proceedings, Paraguay was also the respondent in an ICSID arbitration brought by a private investor under a bilateral investment treaty between Paraguay and Switzerland. In 2012, the arbitral tribunal awarded the claimant $39 million with interest and half the costs of the arbitration.¹⁶ In the Inter-American proceedings, Paraguay did not appear to argue that it lacked the resources to provide the required care to indigenous community members, nor did it argue before the ICSID tribunal that its obligations to the Xákmok Kásek community should be taken into account in the calculation of the arbitral award. As my former colleague, Perry Bechky, points out, it is unlikely there will be a one-to-one match between reparations and rights: given the reality of state budgets, not every amount that would otherwise be paid in reparations will be spent to meet human rights obligations to citizens. The point here, however, is that Paraguay was subject to two sets of international obligations, one owed to a subset of its citizens and the other owed to a private party by virtue of Paraguay’s agreement with another state. It takes no great stretch of the imagination to foresee situations in which the payment of reparations could impair a state’s ability to meet its international obligations to its citizens. This is true even if the impairment is incremental. In the Xákmok Kásek case, Paraguay argued that it was already providing health care and education to the indigenous community through various programmes;¹⁷ nevertheless, the Inter-American Court found such services to be insufficient. A state therefore could be put in a situation of violating existing law. Even if this is not so, to the extent that certain rights are limited to available resources, the diversion of funds from citizens prevents progressive development of those rights and defers their fulfilment. The citizen bears the cost of state responsibility by suffering violations of his international human rights or by forfeiting what might have been gained from their expansion.

At present, the law of state responsibility does not address these issues. The law tends to treat the state as a monolithic, insular whole and pays little, if any, attention to its constituents. The 2001 Draft Articles on Responsibility of States for Internationally Wrongful Acts is the best expression of this law,¹⁸ and the Commentary to the Draft Articles is clear on the unitary nature of the state: ‘[t]he State is treated as a unity, consistent with its recognition as a single legal person under international law’.¹⁹ Of course, state organs and officials play a necessary role in state responsibility; the International Law Commission acknowledges ‘[a]n “act of the State” must involve some action or omission by a human being or group: “States can act only by and through their agents or representatives”.’²⁰ However, such actors perform governmental functions and are under the control and direction of a state, and the rules of conduct and attribution assume that a state engages in the conduct itself through those actors. Indeed, the term ‘attribution’ was chosen expressly to ‘avoid [. . .] any suggestion that the legal process of connecting conduct to the State is a fiction, or that the conduct in question is “really” that of someone else’.²¹ Certainly, there is no sense in which ordinary citizens who do not perform governmental or quasi-governmental functions figure in state responsibility.

This focus on the state level when determining whether a state is responsible for an internationally wrongful act matches a similar focus when remedies are crafted. When a state commits such an act, three consequences follow: the state still has a duty to perform the obligation in question;²² if the wrongful act is continuing, the state must cease such action;²³ and, finally, the state must provide full reparation for injury caused by the act.²⁴ Reparations can take the form of restitution, compensation, satisfaction, or some combination of the three.²⁵ Restitution involves ‘re-establish[ing] the situation which existed before the wrongful act was committed’.²⁶ Compensation covers damage that cannot be covered by restitution.²⁷ Satisfaction addresses injuries that cannot be made good by restitution or compensation and can be given in expressions of regret, apologies, and the like.

As discussed, the Draft Articles do not consider the impact reparations might have on a responsible state’s obligations to its citizens. Under Article 32, ‘the responsible State may not rely on the provisions of its internal law as justification for failure to comply with its obligations’ to cease breaches of international obligations or give reparations; thus, there seems to be no argument that a state’s domestic obligations to its citizens can be a factor in that calculus. There are some limits to reparations: a state is not required to make restitution if it is materially impossible to do so or if it represents ‘a burden out of all proportion to the benefit deriving from restitution instead of compensation’.²⁸ Further, a state is excused from providing satisfaction if it is out of proportion to the injury or humiliating.²⁹ However, although the exception for disproportional burdens is ‘based on considerations of equity and reasonableness’, as worded the restriction goes to determine whether compensation is preferable to restitution (and there is no express limitation on compensation); moreover, preference is given to the injured state if it is unclear whether restitution or compensation should be awarded.³⁰ Similarly, the exceptions to forms of satisfaction are meant to rule out excessive demands and humiliating measures,³¹ but since satisfaction is not meant to be a ‘standard’ form of reparation and is to apply when restitution and compensation are inadequate or inappropriate,³² it is not clear whether these limitations are significant when it comes to human rights. No provision in the Draft Articles expressly requires or allows decision-makers to consider a state’s international, as opposed to domestic, obligations to its citizens when determining the content of state responsibility.³³

B Legal Approaches to the Issue

1 Other doctrinal strands in state responsibility

The tension between a state’s responsibility towards other states and its responsibility towards its own citizens is becoming more salient because of the developments just discussed, but the core problem of balancing a state’s responsibilities towards other states with its responsibilities to other international subjects is not new.³⁴ The drafters of the Articles were certainly aware of the issue; in 1956, F.V. García-Amador, the first special rapporteur for the Commission, wrote:

[I]t is no longer true, as it was for centuries in the past, that international law exists only for, or finds its sole raison d’être in, the protection of the interests and rights of the States; rather its function is now also to protect the rights and interests of its other subjects who may properly claim its protection.³⁵

Amador’s remarks were part of an argument that states must ensure that aliens enjoy the same civil rights as nationals. But they also indicate that the Commission understood that the Draft Articles had to be placed in a broader context in which all subjects were to enjoy some measure of protection under international law. This broader perspective makes room for remembering the rights of the citizens of a responsible state when determining reparations.

There is some jurisprudence that allows such consideration. In his separate opinion in CME v. Czech Republic,³⁶ Ian Brownlie argued that a state’s obligations to its citizens should be a factor in determining reparations. The Czech Republic had been found to be in breach of various provisions of a bilateral investment treaty with the Netherlands. Brownlie argued that one of the factors to be taken into account when setting damages is the economic impact they will have on a state. The Czech Republic was a sovereign state ‘responsible for the well-being of its people’. Given that responsibility, it was doubtful that in entering the treaty the Czech Republic assumed the risk of ‘national economic disaster’ that would result from a large award.³⁷ Moreover, for Brownlie the Czech Republic should enjoy the ‘benefit of civilized modern standards in the treatment of States’, in which even states held responsible for wars of aggression and crimes against humanity ‘are not subjected to economic ruin’.³⁸ By way of contrast, Brownlie pointed out that the Japanese Peace Treaty expressly recognized that although Japan should pay reparations, at the time of signing it did not have the resources to do so and maintain a viable economy.³⁹ There, Japan had been the aggressor; in the present case, the Czech Republic had been the victim of aggression through invasion. How much more appropriate should it be to consider the impact a damages award would have on the Czech economy, and by extension on its people.⁴⁰

The examples used by Brownlie and the concurring opinion itself provide some evidence that state responsibility can and does take into account a state’s obligations to its citizens when reparations are assessed.⁴¹ Those precedents could be juxtaposed with the widespread acceptance of individual human rights to make the case that state responsibility can take into account the impact reparations may have on a state’s ability to protect those rights. However, such a case is somewhat weak because, as discussed above, the Draft Articles do not expressly provide that the rights of the citizens of a responsible state can, let alone must, be taken into account in awarding reparations, and what limitations are placed on reparations do not appear to apply to the issue here. Because of the weight given to the Draft Articles as an articulation of state responsibility, it would be relatively easy to consign Brownlie’s position to the minority view.⁴²

2 Accounting for human rights obligations among other primary obligations

In section 4, I discuss ways in which the law of state responsibility might be changed to address the problems being discussed. Here, I foreshadow that discussion by arguing that purely legal responses are unlikely to be completely satisfactory. The Draft Articles distinguish between primary obligations that stem from specific treaties, customary international rules, and other sources of international law on the one hand, and secondary obligations that arise from the law of state responsibility proper on the other.⁴³ Reconciling a state’s horizontal and vertical responsibilities could happen at either the primary or secondary level. Human rights could be accounted for and protected among the primary rules of obligation, as courts and arbitral panels construe specific state obligations and as they apply the secondary rules of responsibility and craft remedies for wrongful acts. There is some evidence that this is already happening; arbitral panels do take human rights law into account and support such law.⁴⁴ Arbitrators have used human rights law as they applied substantive and procedural rules: these include defining the scope of regulatory expropriation, the exhaustion of local remedies, the availability and assessment of damages and allocation of costs, retroactivity of law, and the right to water on the substantive side, and the admission of amicus curiae, the setting aside of arbitral awards, and general procedures on the procedural side.⁴⁵

On the strength of these precedents an arbitrator could, for example, use Inter-American Court of Human Rights jurisprudence on the right to life to limit a state’s obligation to provide full and timely compensation for an expropriation, and thus find no violation at all. An argument could further be made that if an investment treaty incorporates general international law, under Article 31(3)(c) of the Vienna Convention on the Law of Treaties, a panel must take all other relevant law, including human rights obligations, into account. However, at least some arbitral panels do not seem to be swayed by that argument. James Fry identifies two cases in which arbitral panels have been asked to consider human rights norms but have declined to do so. The second is of interest here.⁴⁶ In Biloune v. Ghana Investment Centre,⁴⁷ an investor brought claims for expropriation, denial of justice, and violation of his human rights.⁴⁸ The claims arose from the investor’s arrest and deportation from the country. The tribunal agreed with the claimant that international law establishes fundamental human rights that were relevant to the investment dispute; however, it found that it had no jurisdiction over those claims because it was neither competent nor authorized to pass judgment on human rights issues.⁴⁹

Fry argues that Biloune does not undermine his claim that arbitration does not conflict with human rights concerns; he points out that the tribunal did not reject or denigrate human rights, and indeed stated expressly that there were certain human rights a government was not allowed to violate.⁵⁰ This, however, is a matter of the glass being half empty or half full. Biloune may not imply that a decision-maker is free to ignore, let alone contravene, human rights norms, but the decision does support a separate line of jurisprudence, one in which the decision-maker declines to adjudicate on human rights claims for lack of competence or because she is restricted by terms of reference or by the primary rules themselves.⁵¹ A decision-maker could follow an approach suggested by Gabrielle Marceau in the context of WTO jurisprudence;⁵² she could take human rights into account and yet cabin them, first, by relegating human rights law to the background rules against which treaties and other primary obligations are interpreted and applied instead of using a human rights norm as rule of decision; secondly, by minimizing conflicts between human rights law and primary obligations and to construe the terms of a treaty in a manner consistent with human rights norms; and thirdly, in the resulting rare case when conflicts between the primary obligations in question and human rights obligations appear unavoidable, declining to exercise jurisdiction over such claims. Under this approach, human rights norms run the danger of being ignored as a matter of substance⁵³ or, at best, becoming part of an uneven patchwork of law in which human rights play an important role in some international regimes and a lesser role in others.

This result could be mitigated somewhat through recourse to the various human rights regimes and their courts. Citizens of a responsible state could claim that a state’s payment of reparations violated one or more of the rights provided in the foundational documents of a particular human rights regime. Human rights norms obviously would not be in danger of being given short shrift by such courts; however, this would simply reinforce the patchwork nature of the jurisprudence and conflict between regimes. ⁵⁴ The problems of consistency and legitimacy caused by multiple regimes are of course much broader than the particular concerns here, but in this context, a responsible state could find itself caught between an international regime that orders payment and another that forbids it, thus running the danger of incurring further responsibility that would be distributed to citizens yet again.

3 Giving priority to one regime over the other at the level of secondary obligations

Another way to address the issue would be to establish rules of priority at the level of secondary obligations, so that one could give state responsibility priority over human rights or vice versa. On the side of state responsibility, one could take the position that since legal responsibility is foundational to international law, the regime of responsibility should control if there are unavoidable conflicts with human rights law. Indeed, it could be argued that without state responsibility human rights itself would suffer, because it is the state that is responsible for protecting human rights. It must also be kept in mind that the human rights of the injured state’s citizens might also be at stake, so that as between the citizens of the injured state and those of the responsible state, the equities would seem to weigh in favour of the former. However, these arguments are not entirely convincing. International law does recognize jus cogens and ergo omnes norms,⁵⁵and the case can be made that state responsibility approaches some priority status. However, short of that there is no sense in which one set of international norms controls other sets; indeed, some human rights norms can make equal or better claim to priority. Further, in human rights there appear to be no rules that allow states to discriminate between citizens and non-citizens when it comes to protecting human rights, except perhaps in favour of particularly vulnerable individuals.

On the other hand, human rights could be given priority over state responsibility, perhaps out of a belief that the individual is the fundamental unit of concern, or because the state is ultimately an extension of the human person. Under this approach, if there is an unavoidable conflict between regimes, a responsible state would be relieved of its obligations to make reparations. But, among other things, this approach would also fall foul of the question of priority as between the citizens of an injured state and those of the responsible state. And as discussed above, unless cabined, a human rights exception to state responsibility could have the unintended consequence of undermining the human rights regime, which requires that states as well as individuals be held responsible for their actions. As often happens when rules of priority are used, a choice of one regime over the other will be somewhat arbitrary.⁵⁶

C Collective Responsibility in Ethics and Lessons for International Law

1 Moral collective responsibility

The issue of distributing state responsibility to citizens is relatively unexplored in international law, but there is an emerging literature on this issue in ethics and it is worth considering whether any guidance is available there.⁵⁷ Much of the field centres round four areas of inquiry. ⁵⁸ The first is whether a collective as such can be subject to moral judgements or whether such judgements are really aimed at a collective’s members. Secondly, if a collective can be subject to such judgement, the issue arises what kinds of collectives can bear responsibility. Thirdly, when, if ever, is it appropriate to distribute group responsibility to the members of the group? Finally, even if a collective is morally responsible for some wrongful act and there are grounds for finding members in a collective responsible as well, what consequences should follow, particularly when those consequences will be felt by members, not the collective?

All four questions have implications for state responsibility, but the third and fourth are of particular interest here. The literature tends to agree that since ‘judgments about the moral responsibility of [a collective’s] members are not logically derivable from judgments about the moral responsibility of a collectivity’,⁵⁹ there must be some culpability on an individual’s part before moral judgements about the collective can be transferred to her. For our purposes this suggests that the best justification for having citizens pay for state responsibility, even if it means the curtailment of their human rights, is that they are somehow implicated in the state’s internationally wrongful act. Several grounds have been raised in this regard. Some ethicists believe that if a citizen commits herself to the objectives of the state it is appropriate that she share responsibility for the state’s actions to further them. Some argue that because a citizen benefits from the state it is fair that she share its burdens. Further, it may be that citizenship itself is grounds for holding an individual directly, as opposed to derivatively, responsible for what a state does. Amy Sepinwall argues that citizenship implies a commitment to the nation-state and to fellow citizens, more particularly a commitment to face judgment with fellow citizens and to be held accountable for a nation-state’s transgressions, ‘in recognition that the nation-state is his as well as theirs’.⁶⁰ Finally, another approach is to focus on the group, as opposed to the individual, as the fundamental unit of concern. Each member would be responsible for group wrongdoing as a matter of course. ‘[W]hen one member of a community commits a wrong against a member of another, all members of the wrongdoer’s community are equally responsible for that wrong, for each member of the community is an expression of its moral center’.⁶¹ If the state is our primary concern, worries about citizens fade into the background.

The question of consequences is conceptual and pragmatic. Whether a group itself should suffer consequences for a wrong will depend in part on the purposes of moral sanctions and whether such consequences serve them. ⁶² Sanctions are used for retribution, societal condemnation, or to deter future wrongs. Determining which purpose should be served and whether a particular sanction will be effective in furthering it is hard enough in the case of individuals, but becomes even more difficult when groups are involved, first, because of the nature of the group (it ‘has no soul to be damned and no body to be kicked’⁶³), and secondly because those consequences often devolve to group members. The result can be seen as a balancing of the objectives of group sanctions with the fairness of distributing those sanctions downwards. Thus, the type of group sanction or the specific consequence matters. For example, Avia Pasternak rejects any distribution of criminal punishment, which she sees as an expression of anger and moral judgement, to citizens of a state. Nevertheless, she supports the distribution of liability to citizens, which in her view does not carry the same sense of wrongdoing as punishment.⁶⁴ Liability also imposes costs on citizens, but in her view liability is not based on the culpability of a particular party; rather, it is based on the sense that when someone has suffered harm it is appropriate to provide compensation.⁶⁵

It seems obvious from even this cursory review of the literature that ethics and law are navigating similar terrain. However, although the moral literature is helpful in pointing out general directions international law might follow in resolving these problems, the conceptual difficulties posed by collective responsibility and the distinctions between ethical norms and legal rules and doctrines indicate that ethics cannot help to solve them completely. This can be seen by examining the moral justifications for finding citizens responsible for state wrongdoing: a citizen’s support for a state’s action; the citizen’s enjoyment of benefits derived from the state; citizenship as a commitment to joint accountability; and a focus on the state as the fundamental unit of concern. Could these serve as justifications for the curtailment of a citizen’s human rights? I treat them in reverse order.

2 The state as fundamental unit of concern

It could be argued that it is entirely appropriate that the state be considered the fundamental unit of concern (this is state responsibility after all) and it should not be surprising that current law sees the state as a monolithic whole. One could thereby retain the atomistic view of the state and leave the law of state responsibility as it is. With respect to the distributive problem, it could be maintained that questions about citizens must fade into the background because, as suggested above, without state responsibility international law would be toothless, and the state continues to be the major actor in international affairs. It would be worse for a citizen if a state, including her own, were not held accountable to other states or parties.

These arguments have obvious merit; however, even if the state remains the predominant subject of international law because of its influence, it does not follow as a moral or legal matter that it should be privileged in the allocation of rights and responsibilities among the various subjects of international law. To use the state as the fundamental unit of concern to justify distributing costs to citizens entails an implicit and, as proposed above, arbitrary decision to privilege state responsibility over human rights. As to the argument that responsibility supervenes all areas of law, such reasoning leads to the odd result that human rights need to be minimized in this regime to be preserved in the other, again with the implication that some legal regimes and subjects take precedence over others. Such an approach risks illegitimacy. If state responsibility does not concern itself with vertical dynamics and impacts, it is unclear why citizens of a responsible state should concern themselves with state responsibility, that is, support reparations, or at a minimum acquiesce to them.⁶⁶ Bertram Malle points out that the ordinary person is not an academic ethicist or legal specialist: people use the same folk theories when blaming group agents for wrongs as they do when blaming individuals. There is a widespread aversion to guilt by association: people are reasonably comfortable with assigning responsibility to groups, but less so to individuals as members of groups.⁶⁷ In these circumstances, law risks becoming ineffective or irrelevant if it strays too far from those normative values and basic understandings of morality; absent further justification, it is hard to explain why the citizens of a state should support bearing the burdens of responsibility simply because the state has incurred it.

3 The citizen as beneficiary of the state and as committed member of the state

Another justification for distributing responsibility to group members is that a person who receives group benefits should share group burdens. There are strong arguments that citizens do benefit from the state because of the economies of scale made possible through a common government, territory and population. If a state that provides such benefits incurs responsibility, it seems fair that those citizens bear some of those costs. This argument is enhanced if citizenship does in fact entail a commitment to stand accountable with one’s fellow citizens for their state’s transgressions.

These factors are persuasive, but not completely so. As Richard Vernon points out, the costs of any sanction may far exceed any benefit conferred by the state, and in most cases a citizen does not voluntarily receive benefits; she is born to them. Further, it is not the case that a person who benefits from the state is always asked to bear its costs: the state protects certain vulnerable people without expectation of return. Moreover, in a globalized world, benefits sometimes come from sources (a regional agreement, for example) that cross state boundaries and thus do not easily fit into a state–citizen, benefit–burden schema.⁶⁸ Further, given differences in the wealth of states, it is hard to compare the benefits enjoyed by the citizens of one state with those of another. Finally, the content of a citizen’s positive duties to her fellow citizens and to her state is not uncontested, and in any event, most citizens are so by birth; they do not voluntarily commit themselves to joint responsibility, and emigration is not a realistic option for most people.

Even if the benefit or commitment justifications for distributing ethical responsibility from a group to members did not have such limitations, to be workable in law they would need to be transformed into a set of legal principles used to determine whether and when a responsible state is justified in curtailing the human rights of its citizens. This raises the problem of commensurability: it is not clear that benefits can be traded for deprivations of human rights. Further, a set of legal principles based on the premise that citizenship entails a commitment to a form of joint liability with the state would require international law to move beyond a focus on the rights of an individual vis-à-vis the state to the duties she owes to it. Of course, some human rights instruments already refer to the responsibilities of individuals to the community,⁶⁹ and it may be that the problems raised here will give more impetus to their articulation. Even so, it is not certain, given current moral understandings, that an elaboration of citizen responsibilities at the international level would include a principle that a citizen has a responsibility to stand legally liable with all other citizens for a state’s actions. That would be tantamount to committing a citizen to a form of liability by association, even though in most cases she has no choice in her citizenship.

4 Citizen participation in government

Most ethicists agree that a citizen who supports state wrongdoing can be held liable for what the state has done. International law already accounts for this: a citizen-agent who participates in the violation of an international norm, particularly a human rights norm, is individually responsible for that wrong, even if the citizen was acting in an official capacity.⁷⁰ Indeed, when a state agent commits a wrong, it can be argued this is no longer a question of derivative responsibility because individual responsibility is invoked.

More interesting is whether the distribution of state responsibility is justified when citizens participate in representative democratic processes. In a democracy, legislatures and officials, whose actions under the law of state responsibility are attributable to the state, are elected by citizens and deemed to act on their behalf. The state, via various state organs, thus becomes an agent for the citizen principal. Anna Stilz argues in this regard that citizens of a democratic state can be held responsible for state action because a democracy is understood to take citizens’ individual rights into account whenever it acts.⁷¹ The balance between state accountability to other states on the one hand and its obligations to its citizens on the other happens on the domestic level, and such balancing has been authorized by citizens themselves. There is therefore no need for further weighing of state and individual rights on the international level, or at least there is a rebuttable presumption that such weighing has already taken place.

It has become so widely accepted that democracy is the primary source of legitimacy of government and of law that some scholars speak of an emerging international right to democracy.⁷² For now, the law of state responsibility does not take into account whether a responsible state has a democratic form of government but, given the almost universal acceptance of democracy, it could be modified to do so. This, however, would raise a number of issues. To begin with, one reason why state responsibility does not now pay great attention to a responsible state’s form of government is a belief that a state should be held responsible for an internationally wrongful act irrespective of its form of government; state responsibility should not vary with the waxing and waning of a particular kind of government, hence in part the distinction between a state’s government and the state itself. If one did consider the democratic nature of a government in distributing responsibility, a possible result would be that a non-democratic state would be excused from international responsibility because its citizens did not consent to its actions,⁷³ while a democratic state would not be excused because its citizens had (thereby creating an odd set of incentives). Further, demo c racy as an idea might be well accepted, but upon closer examination states have different understandings of the term. A decision-maker would nevertheless need some standard for democracy, creating the problem of consensus for such a standard, and thereafter there would be questions about how closely a responsible state’s government must match that standard to justify its citizens bearing that responsibility. Issues also arise from the nature of human rights themselves. Such rights are often understood as inalienable and anti-majoritarian and not subject to balancing; a majority cannot waive the rights of a minority. Finally, the idea that citizens have authorized state action and thus bear responsibility for it is abstract, given the socio-political realities of the modern state. As John Dunn puts it, ‘the structure of modern representative democracy . . . provides . . . a practical basis through which to refuse to be ruled unaccountably and indefinitely against your will’.⁷⁴ He adds, ‘Less steadily and on far less egalitarian terms, it also provides a framework through which to explore what people should and should not attempt to do as a community.’⁷⁵ If this is true, the argument that citizens in a democracy have consented to a responsible state’s actions and thus have at least implicitly agreed to bear the costs of that responsibility becomes tenuous.

3 Complexity Theory, the Ontology of the State, and their Implications for State Responsibility

A The State as Complex Adaptive System

I have argued that the literature of moral collective responsibility suggests various bases international law could use to justify distributing state responsibility downward to citizens, but at the same time none of them appears to be completely satisfactory. In my view, a major reason for this quandary has to do with the ontology of the state, which complexity theory suggests is a complex adaptive system.

Complexity theory is relatively new, a set of related concepts that have arisen from several disciplines. Two concepts are pertinent here. First, under the right conditions, the interactions of individuals can give rise to ‘higher-level’ phenomena that would not be predicted from the individuals themselves. The flock of birds is an often-cited example: individual birds need not be ‘programmed’ to flock in Vs; instead, when a sufficient number of individual birds, programmed to follow simple instructions, such as to keep up or to avoid collisions, interact, such patterns simply ‘emerge’. Something, in this case the V, comes about that is greater than the sum of its parts.⁷⁶ With the flock of birds, it is the micro-level interactions among individual birds that give rise to complex behaviour, but in more complex systems, such behaviour can also arise from the interaction of individuals and the components of a system.⁷⁷

A second concept from complexity theory is that sometimes an emergent phenomenon like the V will persist in its environment as a complex adaptive system. Melanie Mitchell proposes two definitions for the term. A complex system is one ‘in which large networks of components with no central control and simple rules of operation give rise to complex collective behavior, sophisticated information processing, and adaptation via learning or evolution’.⁷⁸ Put another way, it is ‘a system that exhibits nontrivial emergent and self-organizing behaviors’.⁷⁹ According to the theory, such systems are readily observable in nature: hurricanes, the immune system, the economy, and human societies have all been described as complex adaptive systems.⁸⁰ Such systems are distinct from the individuals from which they emerge and from their environment, although such systems are impacted on by that environment. They are thus ‘inherently anti-reductionist’.⁸¹ At this point, the theory has limitations. It is often hard to define the boundaries and components of a complex system. Nor is it always clear how to describe learning, self-organization, and adaptation. Further, and perhaps most importantly, ‘[a] conceptually coherent view of a complex system is hard to link to reality’.⁸² Complexity theorists have tried to work through these problems by using computer simulations; linking ideas of self-organization to postmodern and poststructuralist views of language and epistemology; using simulated agents to understand institutional and organizational behaviour; and by studying cellular automata.⁸³ But it remains to be seen whether complexity theory describes actual social groups or if its concepts are applicable only by analogy.

Complexity theory is thus a work in process with unanswered theoretical and empirical questions. If true, however, it offers a particular understanding of the state that has implications for the issues discussed here. The theory suggests that the state is a complex adaptive system that has emerged from the interactions of individuals, families, and groups, and the various conceptual, social, and cultural tools such as language, writing, law, ethics, religion, and other social institutions that support cohesion among significantly large populations who occupy what in some instances are large geographic territories. The state as complex adaptive system has causal effects in its environment and on the citizens and interactive processes from which it emerges. A state can be influenced from time to time by powerful individuals, but it tends to persist despite them. As such, the state is much more than the sum of its citizens; it is capable of doing far more than what any one individual or group of individuals can do separately; and there is no clear link between the two.

B Complexity Theory and Responsibility

The ontology of complex adaptive systems has several implications for collective responsibility in general and state responsibility in particular. Because it accepts the reality of complex adaptive systems, complexity theory appears to side with those who argue that collectives are ontologically distinct from their members; thus, to the extent that responsibility requires that there be some actor to which responsibility can be attributed, complexity theory suggests that the state qualifies as such an actor because the state itself has causative effects and does not dissolve into constituent parts. Complexity theory is thus consistent with a basic assumption of international law: that the state is an appropriate unit to which to ascribe legal responsibility; it is not simply an extension of the individual.

At the same time, the ontology of complex adaptive systems suggests that the gap between citizen and state is real in at least two senses. First, complexity theory confirms that relatively innocuous individual behaviour can contribute to unwanted results at the group level. David Bella argues, ‘To merely blame individuals . . . is to avoid . . . the essential claim of emergence: that the character of wholes should not be reduced to the character of parts. . . . [E]vil . . . outcomes can emerge through the efforts of normal, competent, and well adjusted people much like ourselves.’⁸⁴ Bella uses as an example scientists who worked for a government organization responsible for producing biological weapons. He shows how the interactions among individual scientists under relatively simple working rules and incentives could nevertheless result in the production of these weapons of mass destruction.⁸⁵ If Bella is right, complexity theory again confirms that collectives as such must be the focus of collective responsibility, if only because individual actions that contributed to emergent state behaviour can seem unremarkable.

However, if an individual’s actions are indeed innocuous, yet unlawful behaviour at the state level can result from them, the distributive problem becomes even more vexing. The non-linear relationship between the complex interactions of individuals and the phenomena that emerge from those interactions means that in most cases it will be impossible to trace direct connections between an individual and the complex adaptive system of which it is a part and the impacts that system may have in the world. Since the emergent phenomena that make up and occur in complex systems are products of emergence, how can an individual be said to have caused them? This suggests that even if an individual agrees with the complex system’s behaviours, there may still be no meaningful connection between that agreement and what the system has done. The non-linearity between member and group suggests that the gap between the two cannot be bridged, and to the extent that the distribution of responsibility depends on some connection between a group and its members, the theory explains why the problem of distribution is not easily resolved.

Complexity theory also has implications for responsibility itself. We think it is appropriate to hold an individual responsible for an action when she has committed a proscribed act with some appreciation of the consequences and did so freely. But if either action or inaction can lead to unanticipated consequences in the long run, how can a person be responsible for them? ‘Self-organization . . . strongly counsels for a wider denotation for the term cause, one reconceptualized in terms of “context-sensitive contraints” to include those causal powers that incorporate circular causality, context-sensitive embeddedness, and temporality.’⁸⁶ Klaus Mainzer agrees that in a linear model of causation, ‘the extent of an effect is believed to be similar to the extent of the cause. Thus a legal punishment of a punishable action can be proportional to the degree of damage effected.’⁸⁷ However, since under complexity theory many phenomena result from random events, the belief in proportionate responsibility is called into question. Complexity theory suggests that most of the issues of concern to us have non-linear characteristics. ‘As the ecological, economic, and political problems of mankind have become global, complex, nonlinear, and random, the traditional concept of individual responsibility is questionable.’⁸⁸ The result for states might be similar: assuming arguendo a state has intentions and can ponder potential outcomes, its actions taken over the short term may have little to do with the effects that are eventually felt in the world over the long term.

At a minimum, responsibility must be understood as being highly sensitive to context. Alicia Juarrero, who writes about individual responsibility in general, and William Frederick, who is concerned with the responsibility of members of business corporations, argue that the individual retains certain degrees of freedom to act in the short term, such that she can be held responsible for her actions.⁸⁹ However, she is also constrained by her environment and by internal dynamics that can lead to emergent behaviour. These constraints lead other scholars to doubt whether an individual does enjoy much freedom: ‘[w]e are shaped by environment, genetics, and experience in a way that affects what we perceive as reasons and narrows the horizon of possibilities for action. Environmental, genetic and psychological factors all shape what count as reasons for a person. Recognizing this should challenge our confidence that a given wrongdoer was morally capable of doing better.’⁹⁰ It is for reasons like these that some who want to retain some concept of responsibility are led to conclude that such responsibility must be constructed. ‘Responsibility . . . does not exist prior to its assignment.’⁹¹

For the purposes of this article we need not decide whether a state has enough freedom or some other set of characteristics to justify assigning responsibility to it or whether it is so constrained that any such assignment must be constructed. In my view, however, the ontology of complex systems is such that any system of law or ethics that distributes the responsibility of a complex system like the state downward to individuals who are part of that system will necessarily be a product of construction, and hence arbitrary to some extent because there is no linear connection between the state and the citizen. The ontology of the state suggests that when a state is held responsible for a wrongful act and required to make reparations, there is no satisfactory way to avoid the criticism that its citizens are suffering injury by association if those reparations adversely affect them.

C The State as Legal Concept or as Ontology

It seems that complexity theory and its ontology of complex adaptive systems have led us down a blind alley. If so, why pay attention to that ontology, particularly when some jurisprudence urges that questions of state ontology be put to the side? The Draft Articles themselves appear to be based on this premise. The Commentary provides, ‘[t]he State is a real organized entity, a legal person with full authority to act under international law’.⁹² James Crawford, the last rapporteur for the Commission, argues that ‘[a] State is not a fact in the sense that a chair is a fact; it is a fact in the sense in which it may be said a treaty is a fact: that is, a legal status attaching to a certain state of affairs by virtue of certain rules or practices’. ⁹³ Crawford traces an early parting of ways between legal accounts of the state on the one hand, and socio-political-economic accounts on the other; as early as Vattel, states in the natural order were distinguished from states under law.⁹⁴ Roland Portmann argues that contemporary international law practice shares that earlier understanding.⁹⁵ Following Crawford, Portman urges that international practice is largely informed by, first, a formal conception of international personality in which legal rules set out the criteria for such personality and, second, an ‘actor conception’ of personality in which legal personality comprises actors who participate in decision-making processes on the international level.⁹⁶ Based on this ‘blend’ of conceptions, ‘the state is a legal status, the individual has to be protected from state power rather than being entirely subjected to it, and there are general norms of international law transcending state consent, including overriding legal principles of preemptory norms which cannot be derogated from as a result of policy considerations’.⁹⁷

That international law should consider the state as a legal construct and no more is attractive, if only because it avoids conceptual problems involved in viewing the state as a socio-political entity. Further, it has direct relevance to responsibility itself. As discussed above, much of the moral literature struggles with whether a collective can properly be subject to moral judgement in the first place because it is unclear whether the concepts we use to make such judgements about individuals apply to groups. The same holds in the political science and international relations literature. ⁹⁸ A purely legal conception of the state skirts the issue; by definition, a state is understood as capable of engaging in relations with other states and of fulfilling international obligations, so that, in a sense, international law is already comfortable with a construction of the state as capable of bearing legal responsibility. As Portman puts it, ‘[T]he only consequence directly stemming from international personality is the capacity to invoke international responsibility and to be held internationally responsible.’⁹⁹

Whether a state should be viewed as a legal entity may also have implications for individual responsibility under international law. Roberto Ago, the second rapporteur for the Commission, believed that a legal conception of the state was almost required by its ontology. Like others, Ago argued that a state needs individuals and groups to act for it. In turn those individuals’ actions are imputed to the state. Ago insisted, however, that the entity to which those actions are imputed is a legal one, not a socio-political entity:

There are no activities of the State which can be called ‘its own’ fro[m] the point of view of natural causality as distinct from that of legal attribution. . . . In describing the State as a real entity—and it is such an entity, like any other legal person—one must nevertheless avoid the error of giving an anthropomorphic picture of the collective phenomenon, in which the individual-organ would have his personality absorbed and annulled in the whole and would be an inseparable part of it, rather like an organ of the human body.¹⁰⁰

Ago was thus fully aware of the conceptual difficulties associated with collective responsibility and he anticipated the problems for individual responsibility if the state is viewed as something that subsumes the individual. He wanted to preserve some degree of autonomy, and thus responsibility, for the individuals and groups who act on the state’s behalf. In his view, such autonomy is preserved by rejecting an understanding of the state as an organism that would subsume individual choice.¹⁰¹

Viewing the state in exclusively legal terms thus has its strengths, but it has its shortcomings too. As Portman acknowledges, to maintain that the state becomes a state by meeting certain legal criteria assumes there are legal principles that pre-exist the state, but modern positivism views the state as the sole origin of law. One could respond, as Portman does, by arguing that ‘at least some fundamental legal principles precede the existence of states as international persons’.¹⁰² However, unless one begins with an abstract legal principle such as a grundnorm, the issue of which came first is a bit of chicken-and-egg. Law certainly existed before Westphalia, but Westphalia is shorthand for processes in which law and social-political entities interacted well before 1648 and thereafter, not just a single event. The question of which is first, law or socio-political entity, leads to infinite regress, with the implication that neither view can take precedence, even though contemporary international practice may have opted for one view over the other.

It is therefore just as important to pay attention to, to use Crawford’s phrase, the ‘state of affairs’ as to the legal status that attaches to it; to do otherwise leads to an odd insularity. Territory, population, government, and the ability to engage in international relations are legal terms (as is indeed, the term ‘citizen’), but they also purport to say something meaningful about the socio-political and geographic state of affairs to which legal analysis applies. One might follow Luhmann and maintain that as an autopoetic system, law always and perhaps must view the world through its own terms, even states of affairs. But even so, the state of affairs is not ignored. As with the relationship between law and ethical norms, if legal characterizations stray too far from reality, however perceived, two dangers are posed. Law risks becoming ineffective because it cannot deal adequately with that reality or it risks becoming irrelevant for the same reason.¹⁰³ Both exact a toll on international law’s legitimacy. As discussed, this felt reality includes citizens who sometimes feel the adverse effects of state responsibility. Once, however, we acknowledge that the state of affairs is a relevant part of any legal analysis, we make room for ontology. In the case of the socio-political entity that is the state, complexity theory indicates that that ontology is one of an entity with a population, people who now enjoy certain human rights, but who by virtue of being part of a complex adaptive system, are not linearly connected to that state.¹⁰⁴ This dilemma is part of the state of affairs with which law has to contend.

4 Constructing Downward Responsibility: Second-Best Responses

Complexity theory thus suggests that there is no satisfactory way to justify distributing state responsibility downwards to citizens based on some relationship between the two; that relationship is non-linear, and unless a citizen has directly engaged in some internationally wrongful act one cannot establish culpability on a citizen’s part that would justify the distribution of state responsibility to her, let alone a violation of her human rights. Any justifications will therefore be constructed, an imposition of a kind of strict liability, and arbitrary to some extent. In this last section I discuss what might be done to address this. J.B. Ruhl and James Salzman argue that the best approach to complex problems is to ‘whittle away’ at them. ¹⁰⁵ Complex problems do not lend themselves to linear, top-down, global solutions, so a second-best response is to make small-scale fixes to components of the problem.¹⁰⁶ In that vein, I discuss a number of small-scale responses to the problem here, all of which, however, will be second-best.

A Human Rights Assessments at the Level of Secondary Obligations

That reparations by a responsible state may be felt by citizens should make a decision-maker think twice before awarding them. This involves insuring that the injured state or private party has proven its damages, but it also requires including expressly in the secondary rules of state responsibility an assessment of the impacts reparations will have on the human rights of the citizens of a responsible state. Several factors would be considered, among them the gravity of the internationally wrongful act; the degree of harm to the injured state and to its citizens; the impact of reparations on a responsible state; the specific human rights obligations, if any, that would be violated in the event that reparations are made; and the availability of less impactful remedies.

Such an assessment would take better account of the obligations a responsible state owes to the injured state and to its own citizens than is done now. At the same time, such an accounting will be imperfect. For example, it seems obvious that the degree of harm to the injured state and to its citizens is relevant in deciding whether and to what extent a responsible state should be required to make reparations. There is of course the injury suffered by a state itself and its need for compensation, and the international community would want to set an example to discourage future wrongful acts. Although under international law the very finding that a state has committed an internationally wrongful act is a form of relief, restitution or compensation will be likely to have greater deterrent effect.¹⁰⁷ Here it would seem that the equities and instrumental considerations weigh in favour of the injured state. Moreover, as Pasternak points out, that innocent family members will be adversely affected does not prevent the state from punishing individuals convicted of serious crimes;¹⁰⁸ so too, that citizens might be harmed should not prevent the international community from holding a state responsible: to take into account all adverse impacts on third parties would render any system of responsibility unworkable.

However, the analogy is not completely apt. For all of the moral and pragmatic reasons why the balance of harms should weigh against a convicted person’s family members, there usually is no sense that the state is violating those members’ rights. Further, in wealthier states, adverse effects on the family can be reduced through government assistance. Here we are concerned with situations in which a state is required to pay reparations that will cause it to violate its international legal obligations to its citizens. To a lesser extent, there is also the danger of perverse incentives. I have argued that the ability of citizens to influence state behaviour is probably minimal at best. However, if citizens believe they will suffer the consequences of their state’s actions even if they are not culpable, they will have no incentive to use what little influence they do have to prevent a state from engaging in a wrongful act in the future.¹⁰⁹

B A General Law of International Responsibility

The Draft Articles provide that doctrines associated with state responsibility are not in derogation of other international obligations;¹¹⁰ thus, the law itself could be said to call for an overarching set of principles that would encompass and balance the responsibilities owed respectively by states, individuals, international organizations, and all other entities that emerge as subjects of international law. This general law would have the advantage of addressing all distributional issues that arise in group responsibility, and, given the coordination problems that arise when there are multiple actors, it is perhaps inevitable that such a law will be developed.¹¹¹ That law would be likely to incorporate some of the factors discussed in the previous subsection, as well as others, because multiple subjects are involved. But for that reason, the conceptual and practical difficulties just discussed would not disappear; they would be raised to a higher level of abstraction and thereby be made even harder, if only because more entities must be taken into account. Moreover, concerns about legitimacy at the level of state responsibility would be exacerbated because decisions about the allocation of responsibility among all subjects of international law would be even further removed from ordinary citizens.

C Targeted Distribution of Costs

Care can also be taken when the costs of responsibility are distributed. Pasternak argues that costs can be distributed from a group to its members in three ways:¹¹² they can be allocated in proportion to a member’s involvement in the wrongful act; they can be imposed randomly on a small group of members; or they can be shared evenly. She believes that proportional distribution is the most morally defensible method, but the most difficult to implement because levels of involvement are often hard to assess. A random distribution has the least impact on most members but is hardest to defend because it is a form of scapegoating. Pasternak concludes that an even distribution of costs is the best compromise.

With this framework, a responsible state could be required to distribute costs in a similarly principled way. As discussed earlier, reparations mean that funds paid to the injured state are funds not available to fulfil an obligation owed to citizens. In what sense, then, can the failure to provide, for example, health care be shared among citizens? Presumably this involves allocating fewer resources to health care than would have been allocated without reparations. It is unlikely that the costs of a significant award could be distributed by withholding expenditures from only those citizens who can be said to have participated in the proscribed state act (and in many cases identifying such responsible people is not possible). The state could therefore allocate those scarcer resources to those who need them most;¹¹³ one could argue in this regard that a state’s obligation to provide certain socio-economic rights does not require it to be as proactive when a person can secure them on her own. This will not be possible when the right involved is such that a state’s obligations cannot be differentiated in that way, so in that case the allocation of costs would be distributed evenly.

This approach has the advantage of reducing the negative impact an award might have on a citizen’s rights. But it also is a second-best solution for at least two reasons. First, unless the right is such that a state’s obligations vary with a citizen’s personal resources, individual rights are still being violated, even though the impacts of that violation are spread in some principled way. This contributes to a weakened view of human rights, with the implication that human rights violations must be tolerated for higher ends, state responsibility in this case. Secondly, when an award is large, an international decision-maker would oversee major distributional decisions usually left to states. One could respond that any award has distributional implications, and further, in human rights jurisprudence, international courts have become involved with internal policies whenever they vindicate such rights. The Xákmok Kásek case is a good example. However, these practices are not uncontroversial because such distributional decisions represent an erosion of sovereignty and are made without citizen participation.

D Citizen Consent

Earlier I was critical of justifying the distribution of the costs of state responsibility through citizen agreement, because it is unclear whether certain rights are alienable and, if they are, whether a majority of citizens willing to waive their rights can speak for the minority who are not; because, given the size and nature of the modern state, the ability of citizens to give meaningful consent to state wrongdoing is limited; and because it is unclear whether state responsibility should depend on the form of government of the responsible state. Yet, in the last three subsections citizen consent has served as a benchmark for evaluating other approaches to the distributional problem. Does this mean that the democratic nature of a responsible state serves as a second-best response to this quandary?

On balance, the legitimacy of a responsible state’s distribution of that responsibility would be enhanced if it could be said that its citizens had agreed to that distribution. But such consent does not overcome the problems just listed and, given the current state of international law, can lead to further cynicism about international responsibility and domestic consent. Laurence Boulle argues that the law of economic globalization, which he understands as an amalgam of hard and soft norms that emerge from international institutions, states, and private actors, far from eliminating domestic law, depends on it for legitimacy.¹¹⁴ To use a well-known example, multilateral financial institutions often require states to make structural adjustments to their economies before receiving financial assistance. Those measures must be approved and incorporated at the domestic level. This is necessary to implement the recipient state’s international obligations, but it also strengthens those obligations by supplementing them with domestic ones so that the people who will feel the brunt of restructuring can be said to have consented to them. International norms thus piggy-back onto and find legitimacy through domestic norms. However, recall that a state’s internal law cannot excuse it from international responsibility. Thus, on the one hand, domestic law created through democratic processes plays an important role in ensuring that a state complies with its international obligations, but on the other, such law cannot be used to excuse a state from those obligations. Given that, it is open to question whether citizen consent does in fact increase the legitimacy of international norms.

5 Conclusion

As stated in the Introduction, international law has always had to do with groups, and this article has explored an important implication of that fact. It has discussed how certain ideas from complexity theory affect the law of state responsibility, in particular, the problems that arise as the distribution of state responsibility conflicts with human rights norms. That problem is a subset of issues that concern collective responsibility more generally. Complexity theory contributes to the debate by providing an ontology of groups that jibes with international law’s focus on the state; it is a real entity, not dissoluble to its individual citizens, so that if international law is to be effective it must deal with the state as such. However, the theory also explains why any system of law or ethics that tries to link individual and group behaviour will be fraught with problems: as a complex adaptive system, a state would not exist without the interactions of the individuals within its borders, but a state is more than the sum of those individuals, and the relationship between the citizen and the state cannot be directly drawn. International law ignores that ontology only at the risk of losing effectiveness or relevance.

These findings do not necessarily preclude state responsibility as such, but make it hard to account for it, and with regard to justifying the distribution of state responsibility to its citizens, almost impossible. Any attempt to balance a state’s international obligations to other states with its international obligations to its citizens will be imperfect at best because of the non-linear relationship between state and citizen. This does not mean that such attempts should not be made; factoring human rights into reparations, a law of general responsibility, targeting the distribution of the costs of responsibility, and citizen consent can ameliorate the problems discussed to some extent. But we should not expect any of these approaches to be completely satisfactory. This article may thus be disappointing; like much scholarship, it purports to demonstrate why an important problem in international law is intractable and why any solutions will only half-succeed. However, the adverse impacts of the distribution of state responsibility are often felt by real people and can rise to the level of a denial of their rights. Thus, in any given instance it is at least worth asking whether the benefits accruing from state responsibility are worth those negative impacts and whether there are ways to minimize those effects, since there may be no good way to justify them.