The Principal–Agent Model
Abstract and Keywords
This chapter argues that external restraints on states’ power will have to make sense of the ambiguous and conflicting rules regarding the exercise of sovereign prerogatives in international law. Fundamentally, the problem of deciding which rules apply to state behavior is an institutional problem. States engage in a wide range of patterns of abuses against their citizens and foreign nationals, from small or haphazard daily indignities, to massive violations of rights such as forceful displacement, mass killings, rape, and widespread torture. Judging those patterns requires procedures for making rules, interpreting the rules in light of specific instances of abuse, and determining the appropriate response in each context. The chapter develops a delegation model of political authority for specific international institutions that are empowered to act as fiduciary agents for people living in separate states with respect to a narrow range of human rights norms, namely jus cogens.
CHAPTER 1 SHOWED that we cannot rely solely on mechanisms internal to states to restrain state abuse of their own citizens. This chapter will argue that external restraints grounded in the collective endorsement of other states will have a better chance of succeeding. Such restraints will have to make sense of the existing ambiguous and conflicting rules regarding the exercise of sovereign prerogatives in international law. Rules defending the unconstrained exercise of state authority within one’s territory, such as nonintervention in the internal affairs of a state, sovereign equality, and national self-determination, coexist uneasily with sovereignty-limiting rules such as those prohibiting genocide, crimes against humanity, and other widespread abuses committed by a state against its own citizens. For example, international relations scholar Ian Hurd has shown that according to different sources of international law, humanitarian intervention is legal and illegal at the same time.1 This is just one instance where international law contains justifications both to restrict states’ sovereignty to protect human rights and for those states to resist any intrusion on their authority to keep peace, prosecute criminals, and provide justice. Article 2(7) of the 1945 UN Charter states that there is no right to “intervene in matters which are essentially within the domestic jurisdiction of any state.” Since what is and is not properly within a state’s “domestic jurisdiction” is not univocally spelled out, state leaders can at any given time draw on conflicting interpretations of state sovereignty to suit their interests. This (p.26) condition leads to “organized hypocrisy” and undermines the viability of any existing constraint on what states can do to their citizens.2
Overcoming these ambiguities about rules and getting states to comply with widely recognized minimal standards requires international institutions with the authority to create rules and bind states to follow them. Fundamentally, the problem of deciding which rules apply to state behavior is an institutional problem. States engage in a wide range of patterns of abuses against their citizens and foreign nationals, from small or haphazard daily indignities, to massive violations of rights such as forceful displacement, mass killings, rape, and widespread torture. Judging those patterns requires procedures for making rules, interpreting the rules in light of specific instances of abuse, and determining the appropriate response in each context.
Not surprisingly, states are reluctant to cede authority to international institutions for a number of reasons. From the perspective of the states’ leaders, international institutions tie their hands and limit their ability to set policy objectives. From the perspective of the citizens, international institutions are distant, elitist bureaucratic machines that undermine democratic self-government. Citizens see themselves as bound to delegate authority, by a process of public deliberation and consent, to domestic institutions only, which they can shape, question, and hold accountable. Delegating authority beyond the nation-state is often regarded as inconsistent with preserving the integrity of domestic political processes.3
This long-standing assumption that collective grants of authority from the citizens of a state should be made exclusively for institutions within the borders of that state must make room for a different picture of the relationship between citizens, their states, and international institutions. Despite worries that the international institutions could undermine domestic democratic control, citizens can divide sovereign authority between states and international institutions consistent with their right of democratic self-governance. The principal–agent model as a modified consent model can address some of the worries related to the legitimate authority of international institutions to make rules for states. According to this model, coercive international institutions should be conceived as agents of people living in separate states, whose task (p.27) is to keep those states within the acceptable bounds of legitimate behavior by filling in governance gaps in the state system. Supranational authorities should function as an insurance scheme against the possibility that states may fail to fulfill their duties.4
The argument is structured in four sections. Section 1 illustrates some of the ambiguities related to the current practice of restricting sovereign authority in international law with reference to the International Criminal Court (the ICC). Section 2 uses the principal–agent model to generate a justification for the authority of international institutions. Section 3 explains how to ensure that international agents serve the interests of their principals. Section 4 defends the principal–agent model as the appropriate way to represent the relationship between citizens, states, and international institutions.
1. Sovereignty-Limiting Norms and Their Current Effects
The ICC is a permanent international tribunal instituted by the Rome Statute that came into effect in 2002.5 The signatories intended to prevent the necessity of multiple, ad hoc tribunals for massive human rights violations. The role of the ICC is to intervene when national courts are not willing or able to investigate genocide, crimes against humanity, and war crimes, leaving the national courts to exercise primary responsibility.6 As such, the ICC represents a significant step toward a permanent system of institutions that holds political leaders and regular citizens accountable for extensive violations of human rights.
(p.28) However, the consensual nature of much of international law, including the Rome treaty that led to the creation of the ICC, limits its authority to states that have expressly agreed to its charter, unless the Security Council refers a case to it for prosecution. In practice, this means that, if the slow and capricious Security Council machine does not set its gears in motion, the states most likely to commit human rights abuses are exempted from ICC jurisdiction.
A Security Council referral explains how the ICC asserted jurisdiction over Sudan, which is not a member state, and more recently how it came to investigate Libya’s president for serious crimes committed against his people during their 2011 uprising. In March 2009 the ICC issued an arrest warrant for the acting president of Sudan, Omar al-Bashir. He was charged with five counts of crimes against humanity, including murder, extermination, forcible transfer, torture, and rape, all committed against Sudan’s citizens while in office. The ICC prosecutor Luis Moreno-Ocampo, stated that Bashir had “masterminded and implemented” a plan to destroy three main ethnic groups in Darfur: the Fur, the Masalit, and the Zaghawa. According to the evidence presented before the court, al-Bashir had “purposefully targeted civilians” belonging to these groups, killing 35,000 people “outright” in attacks on towns and villages.7
An international arrest warrant has been issued for al-Bashir. Not only has he avoided prosecution, but he has been reelected as head of state following a popular election in the spring of 2010.8 The ICC does not have its own enforcement mechanism, and it relies on member states to capture individuals with outstanding arrest warrants. If those individuals do not travel to ICC member states, they remain safe from capture. This weak ICC enforcement capability reflects a great ambivalence about its role in a world of sovereign states.
Take the United States, for instance. Although it has signed the treaty that led to the creation of the ICC in 1998, the United States has refused to ratify it, fearing that its own political leaders and citizens could be brought under its jurisdiction, and claiming that the United States alone should have final jurisdiction over crimes committed by its citizens. The United States has gone so far as to threaten withholding military aid for countries that support the ICC with the passing of the American Servicemen’s Protection Act in 2002.9 (p.29) It has also pressured the Security Council to exempt its citizens from prosecution by threatening to veto the renewal of several United Nations peacekeeping operations.10 Finally, the United States has entered various bilateral “immunity agreements” asking individual countries to exclude its citizens and military personnel from extradition to the ICC.11
The Obama administration’s pledge to cooperate with the ICC has marked a sharp turn in the United States’ policy in recent years. It supported the indictment of Libya’s Muammar Quaddafi, pledged resources in Uganda to fulfill the outstanding arrest warrant for Joseph Kony and his associates, and participated in the last general assembly that led to amendments to the Rome Statute. This has led legal scholar David Scheffer to claim that the United States has all but become a legal member of the ICC.12 Nonetheless, the United States is not taking the necessary steps toward becoming an official member, signaling its continued ambivalence toward the ICC.
This ambivalence with respect to the ICC is understandable to some degree, since traditionally, sovereignty has meant that states have supreme authority over their territory.13 The traditional conception of sovereignty has entailed both de facto power over one’s territory and citizens, and the authority to make law and require obedience from the law’s subjects.14 Prior to the Second World War, international treaties rarely, if ever, limited national sovereignty. Treaties were mainly concerned with holding states responsible for trespassing on each other’s sovereign prerogatives, such as invading foreign territory or attacking foreign citizens. However, as a matter of international law, states were not ordinarily held accountable for respecting their responsibilities toward their own citizens. The Second World War and the Holocaust have transformed the perspective of international law by bringing front and (p.30) center the ongoing and large-scale abuses of citizens by their own states. The International Military Tribunals that convened in Nuremberg and Tokyo had for the first time held political leaders accountable for abuses they committed against their own citizens. These tribunals represented a big first step toward a new understanding of sovereignty.15
Since the Nuremberg Trials, ad hoc tribunals and humanitarian interventions have made a considerable dent in the practice of unconstrained sovereignty. But although in particular cases, political leaders may agree that the sovereignty of a delinquent state must be curtailed when its abuses are too glaring, they are reluctant to institutionalize principled, universally binding limits on sovereign autonomy. This explains in part why the United States can be an enthusiastic supporter of ad hoc tribunals, but a steadfast opponent of the ICC, at least until recently.16 Yet only by moving beyond soft, noncoercive international legal arrangements, can we make limits on state sovereignty stick. This entails designing and building coercive international institutions that have the power to enforce norms to restrict state behavior. Only with extended prosecutorial and enforcement capacity can the ICC become more effective in capturing and prosecuting the Omar al-Bashirs of the world.
Assuming the ICC can acquire such powers, what would justify its authority to limit states in their exercise of sovereignty? To answer this question, we must look at the ICC’s mandate. The functional justification for the authority of ICC derives from jus cogens norms. Jus cogens norms belong to any plausible set of norms for setting principled constraints on state sovereignty. The Vienna Convention on the Law of Treaties (VCLT), Articles 53 and 64, defines norms of jus cogens as a special category of norms accepted by the entire community of states and from which no derogation is acceptable.17 Norms of jus cogens exist above the will of states and limit what states can do to each other and to their own citizens. A number of commonly accepted jus cogens norms have emerged despite ongoing disagreements over their scope and content: slavery and slave trade, genocide, prohibition of aggression, torture (as defined by the Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, adopted December 10, (p.31) 1984), racial discrimination and apartheid, the right to self-determination, and the basic rules of international humanitarian law applicable in armed conflict.18 By their nature, jus cogens norms alter the consensual character of international law, by setting boundaries on what substantive rules states can consent to.
The establishment of jus cogens norms indicates a weak but growing convergence on the importance of limits for sovereign authority among political representatives of states in international politics. They establish a normative hierarchy in international law by overriding conflicting rules, treaties, or customs. Yet in practice, the tension between sovereign authority and jus cogens norms remains the source of much uncertainty. Outside of the ICC, only a handful of cases have addressed this tension. In five cases brought before national courts between 1999 and 2010, the views with respect to the authority of jus cogens over sovereign autonomy have been divided, with three national courts claiming that “state immunity is not available as a bar to jurisdiction in cases concerning jus cogens violations,” and the remaining two courts upholding the sovereign immunity for the state leaders whose actions allegedly contravened jus cogens norms.19
What complicates this case law is that these trials take place not in international courts but in state courts under the universal jurisdiction principle, raising questions about the jurisdictional authority of states serving as impartial forums for judging breaches of international law committed by other states. Universal jurisdiction is a relatively new and controversial principle of international law that allows any state to claim jurisdiction over egregious violations of human rights, regardless of the country of residence of the accused and the victims and the location of the crime.20
(p.32) As for international courts, the relevant case law is patchy and uninformative. The International Court of Justice, an organ of the UN, endorsed the jus cogens idea for the first time in 2006 in the Judgment on Preliminary Objections in Armed Activities on the Territory of the Congo (Democratic Republic of the Congo v. Rwanda), although it failed to clarify its legal status.21 The European Court of Human Rights has addressed jus cogens also once, in the case of Al-Adsani v. The United Kingdom, when it rejected the argument that jus cogens trumped sovereign immunity.22 As these cases make vivid, the issue of whether limitations on sovereign authority is justified is still very much in dispute as a matter of international law. The 2004 UN Convention on Jurisdictional Immunities of States and Their Properties stipulates no exception to immunity for violations of peremptory norms, which according to one scholar “reflects the fact that such exceptions do not enjoy broad acceptance in the international community.”23 These ambiguities about what rules states should follow cannot be settled without international organizations that are widely perceived as legitimate by states and their citizens, with the authority to make, interpret, and enforce such rules. Such organizations must meet both the functional and emergent justification. They must both serve appropriate goals and emerge through a process that is acceptable to the people who will end up being subject to their rules.
The necessity of curbing abuses of state sovereignty has received sustained attention in the philosophical literature.24 Increasingly, political philosophers have reflected on the moral basis of the principles that justify restrictions of (p.33) state authority. They provided good reasons to believe international institutions could meet the functional justification by serving legitimate goals. But, with few exceptions, they have spent less time thinking about how the institutions that put those principles in practice emerge and what are the legitimate ways in which these institutions can exercise their authority over states. Resolving many of the moral and legal issues surrounding the limits of state authority entails generating a new account of institutional authority that applies to both states and international institutions, and explains how institutions such as the ICC can operate in ways that do not constrain but rather enhance democratic self-governance. I will borrow the principle–agent model familiar from analyses of domestic bureaucratic politics to provide one such account of the emergence and authority of international institutions over states.
2. Divided Sovereignty: The Principal–Agent Model
2.1. The Origins of the Model
Originally developed as a model for understanding the theory of the firm and the problems of moral hazard in the insurance business, the principal–agent relationship such as lawyer–client, doctor–patient, and broker–investor is typically conceived as a contractual relationship between one party, the principal, and another party, the agent, who is expected to act in the principal’s interest.25 Individuals and organizations enter agency relationships for various reasons, such as lack of specialized knowledge or inability to coordinate on their own with other people. In theoretical accounts of domestic political institutions, a breakthrough occurred in the 1980s when political scientists seized on the principal–agent model to explain congressional oversight of bureaucratic agencies. The model describes congress as a principal when it appoints government agencies to fulfill administrative roles.
Principal–agent theorists have emphasized a problem characteristic of the agency relationship: There is no guarantee that, once appointed, the agent will choose to act in the principal’s best interest, or that it will pursue his or her interests as efficiently as possible.26 The agent has its own interests that may conflict with those of the principal and “is induced to pursue the principal’s objectives (p.34) only to the extent that the incentive structure imposed in their contract renders such behavior advantageous.”27 The agency relationship is shaped by the independent goals of the agent, by the mandate (contract) that gives power to the agent, and by the environment in which it operates––all of which structure the agent’s motivation to stay close to the principal’s wishes or stray from them.28
The relationship between the citizens of a state and its political institutions more broadly can also be modeled as a relation of agency. Social contract theorists such as Jean Hampton have argued that the relationship between the people and their political officials is best understood as a relation of agency.29 Scholars of democratic representation such as Jane Mansbridge, or Adam Przeworski, Susan Stokes, and Bernard Manin in their edited volume, portray the relationship between citizens and their representatives in the language of the principal–agent relationship.30
Strictly speaking, political officials do not stand in a contractual relationship with citizens. The relationship between the citizens and political officials is a fiduciary relationship, in which the citizens as the principal delegate authority to an agent or agents that are supposed to act in the principal’s interest. This relationship is a “conditional grant of authority from a principal to an agent that empowers the latter to act on behalf of the former,” circumscribed by constitutional rules defining the extent of the agent’s authority.31 The citizens, as a collective principal, delegate authority to the executive, legislative, and judicial institutions as their agents. These agents can, in turn, delegate authority to bureaucratic and administrative agencies and elongate the chain of delegation.
Scholars have illustrated the unique challenges of political agency as a special agency relationship. For example, George W. Downs and David M. Rocke have modeled this relationship of agency in order to understand how its nature (p.35) influences the executive’s actions when the latter acts on preferences in the area of foreign policy potentially different from the preferences of the public.32 This special agency relationship is plagued by information asymmetries (i.e., the executive has more information about foreign policy) and power asymmetries (i.e., the executive has coercive power over the principal). In addition, the principal is made up of a large mass of people, the citizens, who exhibit rational ignorance and face collective action problems. The principal has difficulties controlling, monitoring, and structuring the incentives of the agent. This is in no small part due to the fact that this relationship can be characterized by a reversal of control from principals to agents. It is a feature and not a bug of the model in the political context that agents have coercive powers over the principal.
However, the public exerts some form of control over the executive as they can remove it from office. If the executive engages in wars that are too risky or irresponsible, the public can express its disagreement during an election. Ideally, the threat of this sanction serves to make the executive align its preferences more closely with those of the electorate. But this is an imperfect mechanism of control of the agent by the principal. The citizens often lack sufficient information about the policies of other states to assess the appropriateness of the executive’s actions, and the executive can engage in risky and counterproductive foreign policy if the threat of removal from office disappears, as it does toward the end of a presidential mandate. Evidently, the less control the principal has over the agent, the less effective the agency relationship is in protecting the interests of the principal, and the more a state’s foreign policy can diverge from the interests of its citizens. In states that lack a democratic tradition and effective constitutional constraints on political power, the chasm between citizens and their political leader is wide. The important question becomes how to make sure political agents’ actions serve their principals’ interest as closely as possible. The solution lies in oversight by international institutions.
2.2. The Principle–Agent Model for International Institutions
When the citizens of a state have difficulties monitoring and structuring the domestic agent’s incentives, one way to ensure that the latter stays within the appointed bounds is by providing external constraints. Supranational (p.36) institutions can provide a second layer of agency and act as an insurance scheme that keeps states in check and ensures that they stay within acceptable bounds of their authority. International institutions can ensure, among other things, that humanitarian interventions proceed from appropriate justifications, criminal responsibility is upheld, a modicum of governance exists in failed states, and mass violence during civil war is restrained.
The principle–agent model is implicit in a number of analyses of the legitimacy of international institutions, but its implications for a normative theory of political authority have not been explored at length.33 Principal–agent models can enable us to better identify and describe the difficulties arising from the relationship of citizens and political elites and to locate solutions for the tendency of elites to act contrary to the interests of the citizens, a condition also known as agency slippage.
Citizens from several states can conditionally empower international institutions to act on their behalf. Similar to domestic institutions, international institutions have a collective principal. But unlike the domestic principal, which is composed of a collection of individuals, the international principal is composed of a collection of peoples. International institutions represent the citizens of different states collectively when such institutions are authorized by a single agreement as in Figure 2.1. Otherwise, they have multiple principals, each of which authorizes the agent to act on behalf of its constituent members for a specific (and different) task.34
The collective principal must be able to define and delegate authority to international agents, but also to revoke or to change it. To exert control, the principal can resort to direct delegation, by say, holding elections to elect representatives to international forums, or indirectly, by relying on appointed state representatives who, in turn, appoint representatives to international forums. In the second case, the chain of delegation of authority is longer, but in this respect it is no different than domestic chains of delegation. Keohane and Nye argue, for example, that while delegation to international institutions (p.37) is indirect, “the chain of connections between elections and the actions of an Undersecretary of the Treasury, an independent Congressional commission, or a federal court is at least as indirect.”35 Thus, delegation to international institutions should not raise questions about democratic authority any more than delegation to bodies indirectly appointed by the legislature or presidency with authority to act in domestic politics does. The citizens as ultimate principals delegate authority to state representatives as their primary agents, who then act as proximate principals to delegate authority to international institutions as secondary agents as in Figure 2.2. In the short to medium term, this is probably the most feasible method of delegation since it builds on existing practices of delegations to international institutions.
2.4. The Role of Consent
Consent serves an important role to legitimate the authority of coercive international institutions on this model, but a very different role than consent currently plays in international law. For the time being, states are only bound by treaties and agreements they consent to. Consent is dispositive. The question arises how to interpret the lack of consent of some states to institutions such as the ICC. The fact that some peoples cannot effectively consent, because their political leaders are not legitimate representatives and because they have no alternate route to authorize international institutions, poses a special problem. Their inability to authorize certain international agents to act on their behalf means they cannot be protected from oppressive rulers. We are left with a dilemma: People in decently functioning political regimes (p.38) can consent and consequently acquire protection from international agents, yet people living in states with oppressive political rule, who need that protection the most, cannot acquire such protection.
Collective (consensual) delegation for the protection of nonconsensual norms can solve this dilemma. Consent is required for international institutions to be legitimately created and authorized. However, there is a category of norms that are nonconsensual, in the sense that their validity does not depend on whether states agree to follow them. Jus cogens norms are certainly part of that category. Collective agreements in which a majority of states, suitably defined, grant authority to a coercive international institution that protects jus cogens norms make that institution authoritative even for nonsignatories. We can further distinguish between strong authorization, where a majority of states grant authority to one international organization with the power to act in a certain area of jus cogens enforcement, such as the ICC, and weak authorization, according to which less than a majority but more than a small number of states authorize an international institution to act on behalf of jus cogens norms. For example, NATO has a total of twenty-eight member states, and it is a security alliance that occasionally intervenes to protect people against massive human rights abuses. Should it take on the mandate of human rights enforcement more explicitly, as James Pattison for example argues that it should, it will be weakly authorized on my account by the international community to pursue its mandate.36 Weaker forms of authorization are valuable where stronger forms are not available. Thus, the fact that its authorization is weak does not disqualify an institution from performing essential functions of basic human rights protections, as long as it meets the functional (p.39) dimension of legitimacy, by providing the right capacity in the right way to enforce goals that are desirable.
Allen Buchanan doubts that consent has any role to play, especially given that many states do not plausibly represent their citizens’ interests.37 But it is not clear how we get any international institutions off the ground without at least some states consenting to their power. International institutions could not meet the emergent justification otherwise. Therefore, consent serves an essential instrumental function for authorization. It does not have to be either a super-norm, or a uniquely appropriate requirement, for the legitimacy of international institutions.38 Consent is necessary not for establishing the relationship of authority between every people and the international institution that is newly appointed to protect their rights, but for coordinated efforts, even from a limited number of states, to create it, define its mandate, and the limits of its authority. The consent of some states for coordinating purposes is instrumentally valuable, and not intrinsic to the relation of authority established between international institutions and individuals living in different countries.
Thus, we can separate the legitimacy of the way an institution acquires authority from the legitimate scope of its exercise. But why extend the authority of international agents over nonacceding states this way? Recall that the characteristic problem of agency is that the agents will not always act in the principal’s interest. Because not all de facto state agents accurately represent the interest of their citizens, their lack of consent cannot exempt them from nonderogable norms such as jus cogens. So, for instance, the fact that the Sudanese president does not sign on to the ICC, because he accurately deems it to be against his own interest to do so, should not mean that the ICC cannot act as a legitimate agent that protects the interests of the Sudanese people. States are not undifferentiated wholes, and when leaders choose not to delegate power to certain political institutions, this cannot mean that those institutions lack authority over them.
For a very narrow type of international agents, that is, those that protect jus cogens norms, international institutions can be legitimate fiduciary agents for people who cannot otherwise consent to their rule. Fiduciary or “surrogate” agency compensates for the substantive failures of agency of the primary agents.39 Thus, international delegation relies on a modified consent model. (p.40) International institutions must have the consent of some states in order to exercise authority at all. Organizations cannot become self-appointed international agents; they must be at least weakly sanctioned by the consent of some of the peoples to whom their rules apply. Institutions must themselves acquire their power through legitimate means, and one such mean is a delegation of authority by a collective principal. Once suitably authorized to protect jus cogens norms, international institutions can act as fiduciary agents for the people living in states that have not directly consented to their rule. This is the consent/fiduciary model of international delegation. International institutions that protect jus cogens norms derive their legitimacy to act concomitantly from the special, nonderogable nature of these norms, and also from the consent of the other states. The upshot is that once they meet this dual standard of authorization, legitimate international institutions are justified in curtailing the sovereignty of abusive states, even if people living in those states cannot adequately consent to the authority of these institutions. This serves mainly to impede dysfunctional state agents from inflicting further harm on their citizens and to provide some measure of order and security to populations having to endure abuse. To sum up, we can helpfully distinguish between the legitimate creation of international agents, which we can call “initial authorization,” and the legitimate scope of the authority of the international agents, which we can call “jurisdictional authorization” (covering both consenting and nonconsenting subjects).
Still, one may object that in the case of decent, liberal democratic states, international agents are redundant, and consequently, the motivation of those states to sign onto an international agreement that limits their sovereignty will be lacking. Why should these states incur the cost of having their sovereignty encumbered while enjoying minimal benefits for their citizens? To answer this question, we must separate the perspective of citizens from that of the ‘state,’ or state leaders. From the citizens’ point of view, international institutions have a purpose even in the case of basically decent states. Democracy gains in most Western democracies are fairly recent, and although the probability of political leaders in those states engaging in massive violations of their citizens’ rights is smaller comparatively speaking, it is not zero, and the costs of such violations can be overwhelming. International institutions can act as impartial and authoritative judges and can hold political leaders accountable for failures of the responsibilities entrusted to them by their citizens, even when the likelihood of such failures is small. Thus, even if states such as France withhold consent for the ICC, this fact should not have implications for whether the ICC can exercise jurisdiction over French leaders or its citizens.
(p.41) This modified consent model is different from the current practice in international politics for authorizing international institutions and deciding what function any given institution adopts, which is state consent. According to the state consent model, states are presumed to be equally entitled to authorize international law and institutions. In practice, this means, of course, that state consent is the consent of state leaders.40 However, there is no distinction between state leaders who actually respect and represent the interest of their citizens and those who do not. The difference between the state consent model and the principal–agent model can be represented schematically as in Figure 2.3.
In international law, additional mechanisms can be made available to pressure governments to be legitimate representatives of their citizens. Buchanan recommends that a process of recognition be formalized at the international level to recognize new states as legitimate participants in international politics only if they meet certain minimal internal and external justice requirements. Among these criteria, the most important are refraining from committing widespread abuses of human rights among its citizens, acquiring power without a major breach in international law (such as military conquest), and respecting the rights of other states’ citizens and their sovereign prerogatives.41 This seems like the right approach, yet the absence of such additional mechanisms should (p.42) not, for the time being, impede the formation of organizations that can act as surrogate agents for the protection of individuals’ basic rights to life and bodily integrity.
Consequently, a hybrid conception of legitimate authority for international institutions is necessary to overcome the problem of countries that refuse to consent on illegitimate grounds. International institutions need the consent of legitimate political representatives to get off the ground, but for certain such institutions, the authorization to act need not be limited only to those that consent. This is because their legitimacy comes simultaneously from consent, and from the special, nonderogable character of the norms protected by the new institutional body. The ICC is not, and perhaps cannot be, meaningfully authorized by the Sudanese to be their agent, but it can act to protect interests that they share with people in signatory countries, interests in not being persecuted, terrorized, or murdered by their political leaders. This is why it would be legitimate to extend the jurisdiction of the ICC even to states such as Sudan, or the United States, which chose not to sign or ratify the Rome treaty.42 The Sudanese leader may refuse to ratify it precisely because such refusal is in his interest, but the ICC, acting as a special, nondelegated type of agent that protects the interests of the people of Sudan, can prosecute him because of dramatic failures in his responsibility to protect their interests. The prosecutorial power of the ICC should therefore be extended over nonconsenting states. This extended prosecutorial function can replace the unreliable efforts of the Security Council to authorize prosecutions in cases involving these states.
Legitimacy on this account is different from de facto power. The idea of “recognition legitimacy” contains normative criteria that tell us if a state is legitimate, independent of whether it exercises effective control over its population and territory. Not all states that have effective control over their citizens and territory should enjoy normative legitimacy. Normative legitimacy is also different from sociological legitimacy, meaning that not all states that are perceived as legitimate by their citizens enjoy normative legitimacy according to this view.43 Citizens can support illegitimate governments based on questionable or self-serving reasons. A majority of citizens can support a country’s government even if it is engaged in massive violations of the human rights of a minority, or if it exploits the minority for the majority’s benefit.
(p.43) Absent more effective forms of direct authorization for international institutions from each state’s citizens, state leaders or legislative bodies may still be the most effective vehicle for delegating authority upward to international institutions. But this conception of legitimate agency is preferable to the status quo for a number of reasons. The fact that the principal–agent model would not accept as dispositive the lack of consent of corrupt political officials has important practical consequences. The authority of international institutions such as the ICC can extend meaningfully to states that refuse to sign its charter. This new understanding of the authority of international institutions as agents of the citizens, directly committed to protecting their interests rather than those of their state representatives, would provide a normative base to give less weight to the lack of consent of unjust political leaders who are not themselves proper primary agents and more weight to the nature of the mandate these institutions adopt.
When applied to international politics, the principal–agent model leaves open both the actual mechanism of authorization of international institutions, and their shape, geographical spread, or domain of authority. However, what makes international institutions different is the fact that there is significantly more scope for a kind of contractarian mode of authorization of international institutions than there is (or was) in the case of states. Since consent to international institutions is an indirect, attenuated form of consent, granted through political representatives rather than directly, it is parasitic on the legitimacy of state agents. Still, individuals living in different countries have broader opportunities to define the mandates of particular international institutions, participate in institutional design, and create mechanisms of accountability for these secondary agents, than they do in the case of states. This process, which allows people to correct agency problems at home by holding severely abusive leaders accountable or removing them from office completely through the actions of international bodies, can enhance, rather than undermine, domestic democratic self-governance. The International Criminal Court is already contributing to that process when it enforces criminal liability for corrupt political leaders.
Coercive international institutions protecting jus cogens norms developed according to the authorization model described here meet the two justificatory criteria outlined in the introduction of the book: functional and emergent. They meet the functional justification when they receive mandates to enforce jus cogens norms against state abuses, and more generally when they mediate between states and their citizens in cases of conflict. Particular international agents also satisfy the emergent justification when authority is expressly delegated to them by a collective principal composed of the representatives of people from different states.
(p.44) The principal–agent model can serve as a template for a vertical division of sovereignty between states and international institutions. Yet it can also help articulate a horizontal division of sovereignty. International institutions can be responsible to citizens from more than one state, because several principals may have the same agent. But there is no reason why multiple principals should all have one agent, such as the global government, as their supranational authority. International institutions can be responsive to different constituencies, have different functional goals and areas of expertise, as well as different geographical foci, as in Figure 2.4.
Due to the special nature of the agency relationship that forms between the citizens of a state and an international institution, this relationship can impose specific agency costs. International organizations lie at the end of a long chain of delegation that further complicates monitoring, compliance, and control by the principals. The interest of the principal and the agent are never identical—there can be slippage between what the principal’s interest requires and what the agent does, and this is just as true for international agents as for domestic ones. Consequently, a defender of divided sovereignty based on the principal–agent model has to confront head-on the following problem: If international institutions are secondary agents meant to correct the agency failure of primary agents, how do we make sure that international agents serve the interests of the principal rather than someone else’s? How can individuals from different states delegate authority to international institutions without losing control?
3. Agency Costs at the International Level
International institutions are often regarded as runaway trains whose conductors have lost contact with the station and drive on a destination of their own choosing without due regard for their passengers or freight. While there is some truth to this picture, it misrepresents the extent to which international (p.45) institutions are responsive to inputs from states and are representative of their interests. There is variance in this regard across international institutions, and while most existing ones are not coercive in the sense used here, the principal–agent model can explain the variance and provide diagnostic tools to make the agency relationship more responsive to principals for coercive and noncoercive institutions alike. Coercive institutions raise specific agency problems that I will address at the end of the section.
The sources of agency slippage can be traced to three major causes: (1) principals that overwhelm international agents with multiple and conflicting goals, (2) institutional design features that fail to maximize agent responsiveness, and (3) environmental factors. The first is not a problem with the agents themselves, but with unrealistic, ambiguous, or incompatible requirements that principals place on them. International bureaucrats face “mission creep,” when they are asked to take on new mandates that do not fit easily with their initial ones, and when they are asked to take on more than the organization’s capabilities can handle. While they may appear to go against the wishes or interests of their principals, agents may simply be struggling to define their mission in accordance with the demands placed on them and failing to do so because of the nature of the demands or the paucity of resources and powers at their disposal.44 Principals can make conscious efforts to avoid this problem by delegating a narrow, well-defined set of goals and empowering agents to enact them by endowing the organizations with appropriate resources and powers for accomplishing them.
Second, principals can also resort to institutional design mechanisms that thwart organizational objective, which also creates the impression of misalignment of the agents’ incentives with the principal. The principal–agent literature identifies at least five different ways in which principals make design choices that enhance the control of their agents: defining the scope and means of the organization, “screening and selecting” the personnel to reflect their interests, engaging in monitoring and oversight, leveraging credible commitments to punish, and instituting procedural checks and balances to keep agents in control.45
None of these mechanisms works perfectly. People living in separate states can exercise some forms of control over international institutions to ensure that the latter are responsive to their interests and are accountable to them, but there will always be a distance between what principals want or need (p.46) and what agents do. This is inherent in the nature of the agency relationship. Agents can diverge from the wishes of their principals and engage in slack. Slack is a cost for the principals, which must decide, given the possibility of slack, whether the benefits of having agents are worth it.
Empirical analyses of existing international organizations document the ways in which states use control mechanisms successfully to increase their ability to reign in international organizations and the circumstances in which they may fail to do so. In the impressive volume Delegation and Agency in International Organizations, editors Darren G. Hawkins, David A. Lake, Daniel L. Nielson, and Michael J. Tierney conclude that the evidence shows that in many cases state principals do control their international agents, and delegation problems are not greater at the international level compared to the domestic one.46 For example, in a chapter in this volume, Andrew P. Cortell and Susan Peterson show that the World Trade Organization (WTO) and the World Health Organization (WHO) are able to engage in slack but do so infrequently, and the extent to which they do can be explained by staffing decisions as well as by the voting mechanism for decision-making procedures. The more the staff is chosen for its special expertise, the more autonomy it demands and receives and the more it can depart from the states’ explicit mandates for the organization. Similarly, Mona M. Lyne, Daniel L. Nielson, and Michael J. Tierney find that multilateral development banks are responsive to states that are quite adept at controlling them.47 Lisa Martin’s chapter on International Monetary Fund (IMF) conditionality also explains that the extent to which international institutions gain autonomy is often the result of conscious decisions of member states that find it in their interest to give their international agents leeway to pursue their goals. None of these cases show that international organizations act well, all things considered, only that they can be responsive to their member states; and there is also variance in the degree of responsiveness depending on particular institutional design conditions and choices made by principals, both at the moment of initial delegation and later on while monitoring and reacting to the actions of their agents. Yet these examples dispel the more extreme myth that international institutions are “institutional Frankensteins terrorizing the global countryside.”48
(p.47) Electoral control is the most direct method for “screening and selecting,” and it ensures that political decisions are made by individuals authorized by those to whom the decisions apply. Citizens can vote directly for their representatives to international institutions in general elections (for instance, they could vote for their representative to the UN assembly or the ICC), a practice adopted by member states of the European Union, or resort to indirect electoral representation, when their direct representatives nominate officials (ambassadors, delegations, permanent representatives) in international institutions. This form of indirect representation was typical in the United States before the adoption of the Seventeenth Amendment to the US Constitution, when state parliaments elected representatives to the US Congress.49 Forms of electoral authorization could be imagined along nonterritorial lines as well.50
Yet, electoral control can be so indirect as to ensure no meaningful link between the citizens’ needs and interests and those representing them in international institutions. Electoral control of international institutions takes place to a limited extent even now, indirectly. Citizens elect domestic representatives who, in turn, nominate officials (ambassadors, delegations, permanent representatives) in international institutions. But these existing intermediary links are themselves weakly responsive or nonresponsive to the preferences of citizens with regard to the character of international institutions. Therefore, it is not hard to imagine examples in which political leaders are themselves under a weak form of electoral control and appoint officials to international agencies based on whether their interests are served rather than the interests of the citizens.
Even if these obstacles of direct representation are overcome, elections are not always the most adequate route to representation. Indeed, Andrew Kuper, building on previous work in democratic theory, argues that direct electoral representation is a weak control mechanism to ensure reliability, accountability, and receptivity in general.51 Elections guarantee accountability if representatives, in anticipation of the judgment of the electorate, will act to approximate these interests more closely in order to retain office. But if the representatives prioritize their own interests for reelection, they can be rent (p.48) seeking, by pursuing self-interested policy choices up to the point at which doing so threatens their chances of reelection. Elections cannot prevent rent seeking, so the pursuit of citizens’ interests is likely to be suboptimal.52 Accountability through elections is simply punishment, and it is not likely to act as a very good deterrent. Representatives have more information, and can discern opportunities for their own gain and act on them in ways that the electorate cannot evaluate. The voters do not know the motivations and cannot accurately monitor the actions of their representatives, or can do so at prohibitive costs. Moreover, the receptivity of the representatives to the principal’s preferences is measured only retrospectively in elections, and retrospective evaluation means that the representatives can diverge quite widely from their constituents in the meantime.
I believe Kuper is generally correct. He highlights problems that come largely from studies of domestic electoral representation, but, if anything, they will only be magnified in the case of international institutions. Even in functioning democracies, the link between citizens and their international representatives is so attenuated that there is a serious question if the latter count as representatives at all. Direct representation poses its own special problems. First, it places large demands on citizens to acquire knowledge about the regular agendas, purposes, and effects of international institutions. In addition, deliberating with citizens from other countries prior to electoral choice in order to refine one’s interests and take others’ interests into account is difficult and costly. Finally, coordinating elections in multiple states is logistically cumbersome. Electoral control is also unavailable as a mode of representation for people living under oppressive regimes. These people cannot adequately participate in the selection of representatives for international institutions.
Nevertheless, if electoral control is a largely blunt instrument, it must not be abandoned, but complemented with other mechanisms and procedures that can make institutions effective at representing the interests of the people affected by their decisions. For instance, the courts serve the interest of individuals in a justice system because they are constrained by norms and procedures that ensure due process, fair and balanced weighting of the evidence, and the presumption of innocence.53 The ICC combines electoral with procedural constraints. Judges are nominated by states’ members of the Assembly of State Parties who have signed and ratified the Rome Statute, and they are elected (p.49) by the Assembly. In addition to indirect electoral authorization, procedural rules and norms can render the actions of the Court just and fair. For example, Articles 55, 66, and 67 contain procedural rules and principles that grant prosecuted individuals many of the standard protections available in liberal democratic societies, such as requiring that they be provided with counsel, that they be free of arbitrary arrest or detention, and that they enjoy the right to a fair and impartial hearing.54
Procedural and substantive norms as control mechanisms for international institutions work best when it is relatively easy to discern objective interests, interests that people have independent of whether they can or do express them publicly, and that do not change with changing circumstances. The interests not to be persecuted, terrorized, or murdered, or the protections of due process are objective interests that can be identified in the absence of anyone explicitly defending them. But a system that relies solely on the identification of all interests qua objective interests through independent political agencies is not going to be acceptable, precisely because not all interests are of this kind, and doing so denies individuals the ability to identify their own interests.55 There must be room in the global institutional scheme for the subjective and direct identification of interests, which is why institutional design mechanisms that focus on building good procedures can supplement, but never fully replace, ongoing forms of agency control by the people over whom their authority is exercised.
All of these mechanisms are endogenous to the agency relationship. There are, however, exogenous environmental factors that can create a third path for improving the alignment of the principal and agent’s interests. In international politics, third-party actors are intermediate between individuals, states, and international institutions and can be a potential source of control on the latter by collecting and sharing information and thus enabling monitoring and oversight. Nongovernmental organizations (NGOs) and other semiprivate actors can pressure international institutions to justify their operations publicly.56 NGOs can collect and make available information about bad behavior, as Transparency International does. They can provide new information about the preferences and needs of the individuals who are not able to supply that information themselves, such as political prisoners, and can supply ideas about the most appropriate means to protect people’s interests.
(p.50) Some international NGOs are afflicted by corruption and capture, or lack an understanding of what is needed on the ground to help people.57 And the less legitimate, active, or salient third-party actors are, the fewer the opportunities to monitor and pressure international agents to comply. This fact should not disqualify the capacity of NGOs as a distinct form of social organization to offer a unique channel of control and accountability for international institutions, just as the existence of corrupt cops should not disqualify the police as an institution for protection and law enforcement. In addition, there is a burgeoning debate on whether NGOs should be considered representative of the people whose interests they serve or not.58 Whether they count as proper representatives is a separate question that I will set aside; but NGOs clearly can exert a form of control with the purpose of making international institutions and the system as a whole more responsive to the interests and needs of the people they serve.
The structure of the system, characterized by vertical and horizontal division of authority, induces institutions to check each other’s power and reduce asymmetries of information. Institutions that have narrow, well-defined goals that potentially overlap will tend to monitor and constrain one another. Institutional pluralism can act as an international separation of powers in which competing institutions pressure each other to reveal information about themselves. Through this competitive mechanism, voters will gain easier access to information politicians hold, which enhances democratic control.59 In addition, states can keep international institutions in check. Even small, developing states have been increasingly powerful in influencing existing multilateral international arrangements.60 Developed states such as the United States have both dominated the creation of these arrangements and influenced their rules and norms. And in some areas of international cooperation, such as financial policy, the powerful states’ voice is still louder due to weighted voting rules. However, many international institutions have entrenched the principle of sovereign equality, and have granted small states a much greater (p.51) role in decision making through formal institutional design rules. Thus, small states have been able to pressure large states to adjust their positions to better take into account the former group’s interests, influencing the design of voting rules in their favor and successfully blocking undesirable substantive positions advanced by their more powerful counterparts.61 This does not mean that developed and developing states are functional equals in terms of their influence on the character and goals of international institutions, but it does mean that developing states are able to exert increasingly powerful forms of control and checks on the ability of great powers to get their way.
Coercive international agents raise special challenges, because they (1) can coerce the principal through compulsory jurisdiction, and (2) they receive their authorization from proximate principals, state agents, who may or may not act in the best interest of the ultimate principal, the citizens. The second problem is attenuated, however, when delegation to international authority is grounded in jus cogens norms. International agents are valuable for their coercive capacity, because they deploy it to reign in misbehaving state agents. Coercion, nevertheless, can have widespread and extreme consequences in death and destruction. Thus, responsiveness and accountability are critical for these special agents. As in the case of other agency costs, the solution to this potentially large agency cost is to use all of the institutionally designed mechanisms available, particularly to delegate narrow and well-specified goals, to monitor and evaluate conformity to those goals, and to leverage a credible commitment to punish the officials of these institutions when they misappropriate the authority to act. Such officials must be themselves bound by norms of behavior that are enforceable, and the organization as a whole made subject to the possibility that its mandate will be revoked or its authority dissolved.
To sum up, international agents can be controlled by their principals through direct, electoral modes of authorization. However, electoral control is likely to have a limited impact on those agents’ accountability and responsiveness. Other forms of control over the agents can be ensured through institutional design, which means defining and constraining the function of a given agent in advance, and by instituting procedural norms and rules to ensure that agents pursue their functions guided by fairness and an even-handed (p.52) appreciation of the varied interests of their constituents. The pluralist structure of international institutions would also augment the control of principals. International institutions can check and constrain one another to ensure that no one agent is dominant all the time.
4. Why Use Principal–Agent Theory?
Principal–agent theory is not new either to domestic or to international politics. While political theorists have occasionally resorted to the language of agency, principal–agent theory has not made its way into the common vocabulary of political philosophy in the way it has in American or comparative politics. Several advantages make the principal–agent model appropriate for representing the relationships between citizens, states, and international institutions.
First, models of political institutions serve distinctive functions. They focus on the essential features of relationships or phenomena, leave out detracting detail, and make assumptions about the behavior of different parties to the model.62 As such, they help us understand important features of reality and inform our decisions about whether to accept or change them. The principal– agent model can help us better understand and explain, in particular, the nature and difficulties of political delegation. The model makes vivid the misalignment of interests between citizens and political leaders, and casts light on the inadequacy of the state as the final authority for protecting the individual’s rights.
Second, agency theorists have developed detailed accounts of when and under what conditions agents engage in slack, thereby giving scholars and practitioners alike analytical tools to asses solutions to agency slack, to determine when the gains from delegation can compensate the losses, and which trade-offs are worth pursuing.
Third, the principal–agent model has enabled those engaged in institutional design to better understand how to structure the incentives of the agents ex ante in order for them to carry out the desire of the principals ex post. Ultimately, the goal of the principal–agent model is to tell us how particular institutional forms can increase the likelihood of compliant behavior by political officials and bureaucrats.
(p.53) The normative payoff is plain. Nielson and Tierney make clear that the type of principal–agent model one uses to explain an agency relationship makes the difference between showing that the agent is effective in protecting the principal’s interest or not, and therefore makes the difference as to whether we are willing to accept the outcome or mobilize to change it.63 For example, Karen J. Alter identifies a puzzle in the fact that international judges routinely rule against member states, yet they are not sanctioned by these states for their punitive behavior, as one would expect principals to do with disobedient agents.64 We need to ask if this is a feature in need of correction, as she suggests, or not. When evaluating whether international judges stray from the interests of states, it is important to explain whether by “state interest” one means the interests of political leaders or the interests of citizens. Straying from the directives of state agents is not the same thing as acting against the best interest of the citizens. Alter demonstrates convincingly that international judges are not subject to reappointment pressures from states, and the latter do not deploy effectively the screening and recontracting tools available to them to prevent judges from imposing costly decisions on them. She implicitly states that this is an undesirable feature of international courts due to careless institutional design and unreflective imitation of domestic legal structures.65 But whether the relative independence of judges counts as agency slippage cannot be evaluated simply on the basis of how state agents react to them. Even if the latter find the decision of international courts unfavorable, these courts may be doing a good job of imposing costs on otherwise misbehaving state agents. Of course, Alter may be amenable to the suggestion that the normative implication of an agency relationship depends on the model used to explain the relationship. Yet the problem goes deeper still. Whether there really is a puzzle that needs to be addressed—that of agency slack in the case of international courts—or whether judges simply act autonomously to preserve the interests of the ultimate principals, depends on the frame of analysis. Thus, principal–agent theory is a productive way to make transparent the connection between a set of normative aims and the institutional means of realizing them, as Chapter 3 will show we must.
(p.54) This particular principal–agent model for international institutions helps to address another important objection. James Bohman, for example, worries that the principal–agent model legitimates forms of delegation that end up subverting, rather than enhancing, democratic will. International organizations as agents multiply the centers of authority, thereby extending the antidemocratic tensions within the modern administrative state. In addition, organizations such as the WTO and the World Bank reverse the locus of control from principals to agents. Instead of representing individuals, these agents dictate the terms of the relationship between states and individuals.66 Principal–agent hierarchies, he claims, exacerbate legal domination by distancing the locus of control from individuals who suffer the consequences of their policies.67
Despite Bohman’s reasonable worries, nothing in the principal–agent model entails a dominating relationship. Whether the agents dominate the principals or not is not a feature inherent in the agency relationship but depends on the terms of the delegation of authority, the specific design mechanisms that keep the agent’s power responsive to the principal’s interests and revisable, and factors from the environment that help or diminish the ability of the principals to monitor and control the agents. The principal–agent model is nonspecific about the details of delegation, and there is nothing in the principal–agent theory to suggest that delegation cannot enhance, rather than undermine, popular sovereignty if properly directed. The principal–agent model is simply an abstract way of representing the individuals’ relationships with political authorities, whether domestic or international. Delegation can result in agents who act on democratic principles, are reflexive by allowing the principals to change the terms of the delegation, and who are inclusive and even-handed. Choices about the scope of agent authority, the means by which agents can accomplish their goals, and the options for principals to reign them in will shape the possibilities within which international agents act and their likelihood of engaging in antidemocratic, dominating practices.
International institutions can allow individuals to participate in the making of rules at the international level and fortify the normative powers of citizenship. They can do so both by increasing observance of human rights at home and by creating processes of deliberation and inclusion in international agencies themselves. Indeed without such an account of political agency, it (p.55) is hard to know what to make of the institutional features of Bohman’s own account, that of a decentralized transnational democracy. Bohman proposes multiple transnational demoi as inclusive publics that can define and redefine the rights of their members. Far from being an alternative to his account, the principal–agent model can actually be a friendly complement to it. Otherwise, how can we tell whether the different demoi are made institutionally effective by their publics? How do they come into being and what kind of powers do they have over what issues? Only a principal–agent account that explains how power is delegated by individuals and domestic publics at a higher level and how different institutions and publics interact with each other can sort out the confusion that can result from ushering in multiple and overlapping demoi. A guiding model of political agency is necessary for many other groundbreaking accounts of institutional pluralism that have emerged in recent years. Without such a model, these new accounts lack a more fine-grained appreciation of the challenges of institutional design at the international level and of the condition for effective performance in pluralist systems.
The divided sovereignty model is predicated on a certain understanding of state sovereignty as the agent of the people under its rule. But like any agent, the state does not always act in the best interest of its citizens. To ensure that the state remains within legitimate bounds of its authority, its citizens can delegate some authority for the protection of their interests to international institutions as secondary agents. The ICC is one such secondary agent, because it has jurisdiction to try individuals for genocide, crimes against humanity, and war crimes when states are unable or unwilling to do it and is the first standing international institution with the express authority to do so.68
In international law, the commitment to limiting state sovereignty for egregious violations of human rights is selective. This selective commitment is akin to a domestic situation in which individuals claim that the autonomy of other individuals should be limited on an ad hoc basis when the latter cause harm but that there should be no general, autonomy-limiting rules that apply (p.56) to all. To overcome this condition of haphazard norm creation and enforcement, states such as the United States cannot consistently uphold the idea that other states’ sovereignty should be limited when super-crimes are committed within their territories or by their citizens, but that its own citizens should be exempt from prosecution. All states must submit to universal, principled limits on state sovereignty.
The argument does not settle which other international institutions, apart from the ICC, fit the model I describe here. Therefore, in thinking about the ICC and ICC-like institutions, we should be careful not to idealize the potential of international institutions. I do not advocate for the proliferation of international institutions, precisely because setting up formal institutional arrangements is a process ridden with uncertainty, and high expectations about the performance of international institutions are often thwarted. Design choices and conditions from the larger environment in which they operate determine how effective institutions are in reaching their goals; but as it turns out, we still do not understand very well what kinds of institutional features will produce the best results and under which circumstances.69
The agency model developed here has a number of advantages. First, it explains the authority of at least some international institutions as a means to correct the problematic relationship of agency between the citizens of a state and their political leaders. Second, the reliance on a collective principal that represents a majority of states explains why lack of consent from individual countries to the authority of international institutions should not be dispositive. International institutions can and should protect the interests of people living in repressive regimes, despite their leaders refusing to grant authority over them to those institutions. Third, this understanding of international institutions as agents of people living in different states can serve to broaden rather than subvert domestic democratic authority. Hence, it adds to and extends recent arguments that show the ways in which international institutions can be an expression, rather than a diversion, of democratic will.70
(1.) Ian Hurd, “Is Humanitarian Intervention Legal? The Rule of Law in an Incoherent World,” Ethics & International Affairs 25, no. 3 (2011): 293–313.
(2.) Stephen D. Krasner, Sovereignty: Organized Hypocrisy (Princeton, NJ: Princeton University Press, 1999).
(3.) Robert A. Dahl, “Can International Organizations Be Democratic? A Skeptic’s View,” in Democracy’s Edges, ed. Ian Shapiro and Casiano Hacker-Cordon (New York: Cambridge University Press, 1999), 19–36; Jed Rubenfeld, “Unilateralism and Constitutionalism,” New York University Law Reivew 79, no. 6 (2004): 1971–2028.
(4.) I use the term “insurance” here in a loose sense. This follows other references in the literature for international institutions as insurance. For instance, Andrew Moravcsik uses the term “democratic insurance” to explain why emerging democracies committed to the postwar human rights regime in Europe, while Michael Goodhart deploys the idea of insurance or lock-in to discuss constraints that global institutions place on states. Andrew Moravcsik, “The Origins of Human Rights Regimes: Democratic Delegation in Postwar Europe,” International Organization 54, no. 2 (2000): 217–52; Michael Goodhart, “Human Rights and Global Democracy,” Ethics & International Affairs 22, no. 4 (December 2008): 395–420.
(5.) Arash Abizadeh, “Introduction to the Rome Statute of the International Criminal Court,” World Order 34, no. 2 (2002–03): 19; Jamie Mayerfeld, “The Mutual Dependence of External and Internal Justice: The Democratic Achievement of the International Criminal Court,” Finnish Yearbook of International Law 12 (2001): 71–107.
(6.) The ICC jurisdiction can be triggered in three situations: by a state party to the treaty, by the ICC prosecutor, or by the Security Council. The Security Council is the only agency that can refer a case to the Court for states in which crimes were committed that were not parties to the statute. Abizadeh, “Introduction to the Rome Statute of the International Criminal Court,” 22–23.
(7.) Marlise Simons and Neil Macfarquhar, “Court Issues Arrest Warrant for Sudan’s Leader,” The New York Times, March 5, 2009.
(8.) Jeffrey Gettleman, “Bashir Wins Election as Sudan Edges Toward Split,” The New York Times, April 26, 2010, sec. World/Africa.
(9.) Robert Perry Barnidge, “The American Servicemembers’ Protection Act and Article 98 Agreements: A Legal Analysis and Case for Constructive Engagement with the International Criminal Court,” Tilburg Foreign Law Review: Journal on Foreign and Comparative Law 11, no. 4 (2004): 738–55.
(10.) Thom Shanker and James Dao, “U.S. Might Refuse Peace Duties Without Immunity,” The New York Times, July 3, 2002, sec. World; Carsten Stahn, “The Ambiguities of Security Council Resolution 1422 (2002),” The European Journal of International Law 14, no. 1 (2003): 85–104.
(11.) Brian Knowlton and Thomas Fuller, “Europe Opposes World Court Exceptions: After War, a New Rift between U.S. and EU,” The New York Times, June 11, 2003.
(12.) David Scheffer, “America Embraces the International Criminal Court,” Jurist, http://jurist.org/forum/2012/07/dan-scheffer-us-icc.php (accessed February 15, 2013).
(13.) Dan Philpott, “Sovereignty,” in The Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta, Summer 2010, http://plato.stanford.edu/archives/sum2010/entries/sovereignty/.
(14.) Krasner, Sovereignty: Organized Hypocrisy; Jean L. Cohen, “Rethinking Human Rights, Democracy, and Sovereignty in the Age of Globalization,” Political Theory 36, no. 4 (2008): 593.
(15.) David Held, “Law of States, Law of Peoples,” Legal Theory 8, no. 1 (2002): 1–44.
(16.) According to Schabas, in the face of strong reservations from the international community, the United States was the decisive force in establishing the International Criminal Tribunal for the former Yugoslavia. William A. Schabas, The UN International Criminal Tribunals: The Former Yugoslavia, Rwanda and Sierra Leone (New York: Cambridge University Press, 2006), 19–20.
(17.) VCLT, “Vienna Convention on the Law of Treaties,” 1155 U.N.T.S. 331 (1969): arts. 53, 64.
(18.) Alexander Orakhelashvili, Peremptory Norms in International Law (New York: Oxford University Press, 2008), 50–54; Stefan Kadelbach, “Jus Cogens, Obligations Erga Omnes and Other Rules—the Identification of Fundamental Norms,” in The Fundamental Rules of the International Legal Order: Jus Cogens and Obligations Erga Omnes, ed. Christian Tomuschat and Jean-Marc Thouvenin (Leiden, Netherlands: Martinus Nijhoff, 2006), 21, 27.
(19.) Lorna McGregor, “State Immunity Jus Cogens,” International & Comparative Law Quarterly 55, no. 2 (2006): 439.
(20.) The United States declined to assume universal jurisdiction. In Samantar v. Yousuf, in which five foreign nationals of Somalia sought damages from the former Somali defense minister Mohamed Ali Samantar, for the crimes of torture, rape, arbitrary detention, and extrajudicial killing, the Supreme Court decided that the claim of state immunity does not hold under a domestic law, but it refused to pronounce on whether the Somali’s defense minister’s behavior complied with international law. Sevrine Knuchel, “State Immunity and the Promise of Jus Cogens,” Northwestern University Journal of International Human Rights 9 (2010–11): 149.
(21.) Evan J. Criddle and Evan Fox-Decent, “A Fiduciary Theory of Jus Cogens,” Yale Journal of International Law 34 (2009): 346.
(22.) Ibid., 347; McGregor, “State Immunity Jus Cogens,” 443. The United Kingdom’s standing to hold Kuwait responsible for jus cogens violations committed against the latter’s own citizen, Al-Adsani, rested on the controversial “universal jurisdiction” principle. The European Convention on Human Rights did not, however, make a judgment about the standing of the United Kingdom with respect to Kuwait, but only pronounced on the relative force of different principles of international law. In this case, it claimed that jus cogens cannot supersede sovereign immunity.
(23.) David P. Stewart, “Immunity of State Officials under the UN Convention on Jurisdictional Immunities of States and Their Property,” Vanderbilt Journal of Transnational Law 44 (2011): 1064.
(24.) Larry May, Crimes against Humanity: A Normative Account (New York: Cambridge University Press, 2004); Andrew Altman and Christopher Heath Wellman, “A Defense of International Criminal Law,” Ethics 115, no. 1 (October 2004): 35–67; Allen Buchanan, Justice, Legitimacy, and Self-Determination: Moral Foundations for International Law (New York: Oxford University Press, 2007); John Rawls, The Law of Peoples: With “The Idea of Public Reason Revisited,” (Cambridge, MA: Harvard University Press, 2001).
(25.) Terry M. Moe, “The New Economics of Organization,” American Journal of Political Science 28, no. 4 (November 1984): 739–77.
(28.) Barry R. Weingast and Mark J. Moran, “Bureaucratic Discretion or Congressional Control? Regulatory Policymaking by the Federal Trade Commission,” The Journal of Political Economy 91, no. 5 (October 1983): 765–800; Kenneth A. Shepsle and Barry R. Weingast, “Policy Consequences of Government by Congressional Subcommittees,” Proceedings of the Academy of Political Science 35, no. 4 (1985): 114–31.
(29.) Jean Hampton, Hobbes and the Social Contract Tradition (Cambridge, UK, and New York: Cambridge University Press, 1988).
(30.) Jane Mansbridge, “A ‘Selection Model’ of Political Representation,” Journal of Political Philosophy 17, no. 4 (2009): 369–98; Adam Przeworski, Susan C. Stokes, and Bernard Manin, Democracy, Accountability, and Representation, Cambridge Studies in the Theory of Democracy (Cambridge, UK: Cambridge University Press, 1999).
(31.) Darren G. Hawkins et al., Delegation and Agency in International Organizations (New York: Cambridge University Press, 2006), 7.
(32.) George W. Downs and David M. Rocke, “Conflict, Agency and Gambling for Resurrection: The Principal Agent Problem Goes to War,” American Journal of Political Science 38, no. 2 (1994): 362–80.
(33.) Allen Buchanan and Robert O. Keohane, “The Legitimacy of Global Governance Institutions,” Ethics and International Affairs 20, no. 4 (2006): 405–37.
(34.) For more on the distinction between multiple and collective principals, see Daniel L. Nielson and Michael J. Tierney, “Delegation to International Organizations: Agency Theory and World Bank Environmental Reform,” International Organization 57, no. 2 (2003): 241–76.
(35.) Robert O. Keohane and Joseph S. Nye Jr., “Between Centralization and Fragmentation: The Club Model of Multilateral Cooperation and Problems of Democratic Legitimacy,” Harvard Kennedy School working paper, 2000, 20.
(36.) James Pattison, Humanitarian Intervention and the Responsibility to Protect: Who Should Intervene? (Oxford: Oxford University Press, 2012).
(37.) Buchanan, Justice, Legitimacy, and Self-Determination, 307–09.
(39.) Jennifer Rubenstein, “Accountability in an Unequal World,” The Journal of Politics 69, no. 3 (July 2008).
(40.) Allen Buchanan, Human Rights, Legitimacy, and the Use of Force (New York: Oxford University Press, 2010), 313.
(41.) Buchanan, Justice, Legitimacy, and Self-Determination, 266–67.
(42.) Or signed but chose not to ratify the treaty.
(43.) Allen Buchanan and Robert O. Keohane, “The Legitimacy of Global Governance Institutions,” Ethics & International Affairs 20, no. 4 (2006): 405–37.
(44.) Tamar Gutner, “World Bank Environmental Reform: Revisiting Lessons from Agency Theory,” International Organization 59, no. 3 (2005): 780.
(45.) Nielson and Tierney, “Delegation to International Organizations,” 246; D. Roderick Kiewiet and Mathew D. McCubbins, The Logic of Delegation (Chicago: University of Chicago Press, 1991), 22–38.
(46.) Hawkins et al., Delegation and Agency in International Organizations, 365.
(47.) Mona M. Lyne, Daniel Nielson, and Michael J. Tierney, “Who Delegates? Alternative Models of Principals in Development Aid,” in Delegation and Agency in International Organizations, ed. Darren G. Hawkins et al. (New York: Cambridge University Press, 2006), 73.
(48.) Hawkins et al., Delegation and Agency in International Organizations, 4.
(49.) Wendy J. Schiller, Charles Stewart III, and Benjamin Xiong, “U.S. Senate Elections before the 17th Amendment: Political Party Cohesion and Conflict 1871–1913,” The Journal of Politics 75, no. 3 (2013): 835–47.
(50.) Andrew Rehfeld, The Concept of Constituency: Political Representation, Democratic Legitimacy, and Institutional Design (New York: Cambridge University Press, 2005).
(51.) Andrew Kuper, Democracy Beyond Borders: Justice and Representation in Global Institutions (New York: Oxford University Press, 2004), 90–96.
(53.) Michael Goodhart, “Democratic Accountability in Global Politics: Norms, Not Agents,” The Journal of Politics 73, no. 1 (2011): 45–60.
(55.) Terry Macdonald, Global Stakeholder Democracy: Power and Representation Beyond Liberal States (New York: Oxford University Press, 2008), 118.
(57.) Michael Maren, The Road to Hell: The Ravaging Effects of Foreign Aid and International Charity (New York: Free Press, 1997).
(58.) Macdonald, Global Stakeholder Democracy; Andrew Rehfeld, “Towards a General Theory of Political Representation,” The Journal of Politics 68, no. 1 (February 1, 2006): 1–21.
(59.) Kuper, Democracy Beyond Borders, 101.
(60.) Robert O. Keohane, After Hegemony: Cooperation and Discord in the World Political Economy (Princeton, NJ: Princeton University Press, 2005); John Gerard Ruggie et al., Multilateralism Matters: The Theory and Praxis of an Institutional Form (New York: Columbia University Press, 1993); Stephen D. Krasner, Structural Conflict: The Third World Against Global Liberalism. (Berkeley and Los Angeles: University of California Press, 1985).
(61.) Miles Kahler, “Multilateralism with Small and Large Numbers,” International Organization 46, no. 3 (Summer 1992): 701–02; Krasner, Structural Conflict, 230–63; the formula of weighted voting in the IMF and the World Bank, according to which member countries acquire a number of votes corresponding with their relative share in the global economy, clearly favor economically powerful states. Ngaire Woods, The Globalizers: The IMF, the World Bank, and Their Borrowers (Ithaca, NY: Cornell University Press, 2007), 22.
(62.) For David Schmidtz, models are like maps. They don’t tell you where to go but how to get from point A to point B. See his The Elements of Justice (New York: Cambridge University Press, 2006), chapter 1.
(63.) Nielson and Tierney, “Delegation to International Organizations,” 241–76; Lyne, Nielson, and Tierney, “Who Delegates? Alternative Models of Principals in Development Aid,” 43.
(64.) Alter, “Delegation to International Courts and the Limits of Re-Contracting Political Power,” in Delegation and Agency in International Organizations, ed. Darren G. Hawkins et al. (New York: Cambridge University Press, 2006), 312–38.
(66.) James Bohman, Democracy across Borders: From Dêmos to Dêmoi (Cambridge, MA: MIT Press, 2007), 7–8.
(68.) This distinguishes it, for instance, from the International Court of Justice, an organ of the UN, which only has jurisdiction over states. The ad hoc criminal tribunals have had jurisdiction over individuals; however, the ICC is the first permanent international institution with the authority to do so.
(69.) Michael N. Barnett and Martha Finnemore, “The Politics, Power, and Pathologies of International Organizations,” International Organization 53, no. 4 (1999): 699–32.
(70.) Robert O. Keohane, Stephen Macedo, and Andrew Moravcsik, “Democracy-Enhancing Multilateralism,” International Organization 63, no. 1 (2009): 1–31.