Printer Friendly

Delegating to international courts: self-binding vs. other-binding delegation.

I

INTRODUCTION

One often hears complaints that international courts are undermining national sovereignty. Critics tend to associate a compromise of sovereignty with the claim that courts are exceeding their mandate or running amok. This article explores the linkage between these two notions--sovereignty being compromised, and courts exceeding their mandate--by exploring the distinction between "self-binding" and "other-binding" delegation to courts.

A central claim of this article is that using a single lens to view delegation to courts distorts our understandings of the political role of judges. Courts play four distinct roles within political systems: dispute adjudication, administrative review, criminal enforcement, and constitutional review. In some of these roles, legislative actors delegate decisionmaking authority to courts as an "other-binding" means of social control; through delegation, states primarily bind others actors (citizens, businesses, government employees, administrative agencies, police, et cetera) to follow the interpretation and application of legal rules by courts. In other roles, legislative bodies or states bind themselves ("self-binding"), subjecting their decisionmaking authority to judicial oversight so as to enhance their own credibility as a "rule of law" political system. Self-binding delegations are by their very nature sovereignty-compromising. Other-binding delegations to courts are more frequent and less likely to be sovereignty-compromising. The situation of courts' exceeding their mandate, and thus compromising sovereignty, applies, but rarely: when a court transforms a given role, turning an other-binding authority into a self-binding role.

Part II defines more fully the difference between self- and other-binding delegation to courts, and maps these differences onto the four different roles courts play in a political system. The discussion starts with delegation in the domestic context because international delegation borrows from the domestic model. The discussion identifies when and how delegating the same role to a court will affect national sovereignty differently at the domestic and international levels. Part II also discusses how states try to limit the authority of courts in each role.

Part III examines the empirical record on delegating specific roles to specific international courts (ICs), using as data delegation to all existing ICs, twenty in total. (1) The analysis helps explain an empirical puzzle in the trend of delegating authority to ICs. Since 1990, there has been a proliferation in the number of ICs (2) and in IC usage so that seventy-five percent of the total IC output of decisions, opinions, and rulings (24,863 out of 33,057) have come since 1990. These "new" ICs are not only recent creations; they are qualitatively different entities. (3) Newer ICs are more likely to have compulsory jurisdiction and either private access or access for international nonstate actors to initiate litigation, even though most observers agree that these features make ICs more independent and more likely to rule on cases in which a government is an unwilling participant. (4)

State concerns about national sovereignty have not lessened since 1990. What is this change to "new" ICs about? I argue that the trend towards creating and using ICs with compulsory jurisdiction and nonstate actor access follows from the decision to use ICs in roles other than interstate dispute resolution--namely for administrative review, enforcement, and, less frequently, constitutional review. Most of these additional roles involve other-binding delegation wherein ICs are empowered to review the actions of international actors or the decisions of national administrators tasked with implementing international rules. And it appears that, numerically speaking, most of the increase in IC activity involves ICs playing "other-binding" roles. (5) This analysis explains why ICs increasingly have design features that make them highly independent, yet why relatively few international legal rulings are controversial. It also helps to situate the more sovereignty-compromising examples of delegation to ICs within the larger universe of delegation to ICs.

Part IV concludes by addressing the implications of this analysis for debates about IC independence as it relates to sovereignty costs in delegating to ICs, including debates regarding principal-agent theories and about whether "dependent" international courts are more effective than independent ICs. The article urges a focus on judicial roles, rather than on the design of the court, to understand the extent to which sovereignty becomes compromised. It also suggests that expectations attached to judicial roles, rather than concerns about judges being sanctioned, shape how judges think about deference to political bodies and how audiences react to judicial rulings that upset powerful actors.

II

THE LOGIC OF DELEGATION TO COURTS--DISPUTE ADJUDICATION, ADMINISTRATIVE, ENFORCEMENT, AND CONSTITUTIONAL ROLES

As Curtis Bradley and Judith Kelley note in their opening article, delegation of authority inherently involves sovereignty costs. (6) The heart of the issue is the magnitude of sovereignty costs. What is being delegated to courts is the power to interpret the legal rules. The sovereignty risk in ceding interpretive authority to courts is that judicial rulings can shift the meaning of law in ways that can be politically irreversible. (7) This risk is not just hypothetical. Constitutional review involves nullifying laws passed by legislative bodies, while administrative review involves rejecting decisions made by public actors. Thus, if judicial actors play their intended roles, judges will at times disagree with, rule against, or render interpretations that run counter to what the makers and the enforcers of the law might have wanted, and what the democratic majority might prefer.

Although delegation to courts always risks that the judge will interpret the law in unanticipated and unwanted ways, the risk to national sovereignty associated with delegation to courts varies--not primarily in terms of the design of the court, but rather by the role the court is asked to play. Self-binding delegation to courts involve high sovereignty costs because the defendant in the case will almost always be a state actor, and the legal review will involve asking whether legislative actors violated the law or exceeded their authority. Other-binding delegations to courts involve lower sovereignty costs because the defendants will primarily be private actors, or the court will mainly be monitoring to see that public actors faithfully adhere to the legislative will.

Delegation to courts brings benefits as well. Litigants can hope that a judge ruling in their favor will make it more likely that the loser in the case will change their behavior. Governments and legislatures can hope that judicial rulings in their favor increase their credibility, imparting a "rule of law" imprimatur on public actions. After defining these concepts more fully, this Part identifies the logic of delegation in four judicial roles one finds in domestic legal systems, and how international delegation to courts differs from delegating the same role domestically.

A. Self-Binding and Other-Binding Delegation

In all cases of delegation to courts, judges are delegated the decisionmaking authority to interpret and apply the law to the case at hand. The sovereignty risk associated with this delegation is primarily shaped by the judicial role (dispute adjudication, enforcement, administrative, or constitutional review) because the role defines which actor is likely to be the defendant in the case, the nature of the decision or rule that is subject to review, and whether judges are more likely to defer to legislative will in their interpretations. A stylized historical narrative helps explain this difference.

In earlier times and in smaller societies, there was no delegation to judges; chiefs and kings both made law and served as the interpreters of the law. As territories grew, delegation of interpretive authority became unavoidable. Sovereign actors--those with the authority to make law--primarily delegated adjudicative authority, the power to make a decision about a controversy or a dispute. Although sovereign actors were ceding interpretation of the law, they were not themselves subject to the interpretations of their "judges," mainly because no judge would presume to know better than the sovereign what the law meant. This delegation was "other-binding"--sovereigns were subjecting others to judicial interpretations of the law. As the state apparatus grew, the role of judges grew. Cases still appeared as controversies judges were asked to resolve, but when the subject of cases became state actors, judges ended up in a monitoring-and-enforcing role, reviewing whether the king's other agents (for example tax collectors, local rulers, state administrators, et cetera) were faithfully following the sovereign's laws. Neither type of delegation--adjudicative, or monitoring and enforcing--bound the sovereign so long as the king himself was never subjected to the authority of the court.

With the advent of constitutional democracy came self-binding delegation, wherein branches of government agreed to limit their powers by binding themselves to the authority of others--including to the authority of courts. (8) Also, the introduction of increasingly complex delegation chains complicates the story. As states have sought to control more elements of the economy and society, governments have created many types of public actors, including administrative agencies, entire criminal-justice systems, and executive agencies that sometimes have what amounts to delegated legislative authority. States have increasingly subjected the actions of these actors to judicial oversight. Do we call such oversight self-binding, or other-binding? The distinction can be subtle, and the difference can be intentionally or unintentionally blurred as the political roles of judges and of public actors shift. This article considers other-binding contemporary delegation wherein judges oversee implementation of legal rules by public actors, so long as a state's legislative outputs or authority is not being subjected to judicial review.

This article operationalizes the distinction between self- and other-binding delegations by examining four roles courts play in political systems. The dispute-adjudication role is analogous to the king's representative resolving disputes. It pertains when there is a disagreement in a contractual relationship, and the disagreement is brought to a judge to resolve. The defendant in the case is a signatory to the contract, and either the contract itself or the relation of the contract to a larger framework of rules is under review. The other roles are contemporary outgrowths of constitutional democracy. In the enforcement role, a judge monitors police and prosecutors as they use the state's coercive power. In administrative review, a judge checks the legal validity of the decisions, actions, and non-actions of public administrative actors, who themselves rely on delegated authority. Constitutional review checks whether the law created by legislatures or interpreted and applied by governments, or both, coheres with the constitution.

By its very nature, constitutional-review authority has the highest sovereignty risk because by definition it involves judges reviewing the legality of laws and, by definition, judges are supposed to prioritize the constitution over the will of the legislature. The other types of delegation vary in the sovereignty risks involved, depending on whether a public actor is likely to be a plaintiff or a defendant, the scope of judicial review, and whether the court's jurisdiction is compulsory. These factors can vary for domestic and international delegation of the same judicial role, even when the international court's design mimics its domestic counterpart identically.

This Part identifies jurisdictional elements of courts, identifying how to recognize if a specific judicial role has been delegated to an IC. At the domestic level, the legislative origin of a judicial role may be hard to trace, and there may indeed be no explicit legislative grant delegating authority. At the international level, however, judicial roles are defined in the founding treaties of ICs. This Part identifies key jurisdictional features associated with specific judicial roles: whether a role requires compulsory jurisdiction or private access for nonstate actors to initiate litigation; whether the role is other-binding, self-binding, or both; and how these categories differ internationally compared to domestically. Note that this discussion describes each judicial role as a Weberian ideal type, focusing on the function for the state that the court is serving in each role. Ideal types are useful in identifying essential characteristics and drawing distinctions, but by definition ideal types simplify and do not comport with reality. (9) With respect to the analysis here, the ideal types both underemphasize and overemphasize variations one might find within a category. (10) The ideal type approach is nonetheless useful because it allows comparison across roles, revealing how the logic inherent in the delegation act varies by judicial role.

In practice, judicial roles may change, in some cases morphing a court considerably from its original design. The role designations inherent in the original act of delegation likely shape the design of the legal body and, at least originally, the nature of the cases raised. When litigants ask judges questions that push them outside of their original roles, when judges embrace these opportunities to expand legal doctrine, and when such doctrinal shifts are accepted by legal and political communities, the court's role will morph. As judicial roles evolve, the roles become hats judges put on as they decide legal issues. When thrust into a role, the judge dons the role, and with it the logic associated with the role, as he would a hat. As judges change roles, they change hats. In this context, the ideal type role would provide a first-cut "logic of appropriateness" that would set expectations as to what the judge should do in the case. (11) Even if the judge were disappointing powerful actors or compromising sovereignty, so long as judges stay close to the expected role-based logic, the ruling is less likely to be controversial, and it will be harder to pin on the court the charge that it was "exceeding its authority."

B. Delegation of Dispute-Adjudication Authority to Courts

Dispute adjudication in its ideal typical form is private-law adjudication. Two private parties subject to the law bring a dispute to a judge, who renders an interpretation that binds both parties. These disputes are usually conceptualized as arising from contractual disagreements--differences in opinions regarding duties and obligations owed--though the "contracts" are often informal and implicit. Shapiro identifies this judicial role as participating in social control. (12) In delegating to judges the authority to interpret the law, state actors are seizing the parties' desires to have a judicial resolution of a dispute as an opportunity to bring state laws into the private realm--into neighborly disputes, business interactions, and even family decisions. In choosing the legal outcome, judges are choosing the state's desired resolution--that custody of a child goes first to blood relatives, that firms be accountable for their actions, et cetera.

Dispute adjudication can involve a mix of public and private actors, yet still involve an other-binding, social-control logic. States want their interlocutors to follow general contractual rules so that private actors will be willing to enter into trustful relationships like signing contracts, letting school buses bring their children to school, and so forth. When public contractors are held accountable in the same terms as private contractors, the social-control logic is still at play--states are binding their interlocutors to follow a set of common rules. (13)

How does one know if an IC has dispute-adjudication authority? Although it is easy to identify administrative review or enforcement authority, dispute adjudication is a catchall category. Every "concrete" legal case has two parties who disagree (otherwise the parties would have settled out of court), leading to a judge interpreting and applying the law to render a ruling. Given its ubiquitous nature, judicial dispute-adjudication authority has to be identified in terms of what it is not. International courts with dispute-adjudication authority have a formal jurisdiction to "interpret the meaning of the law" in concrete cases brought before them. A judge stays entirely in a dispute-adjudication role when there is no question about the legal validity of the law itself, or about the validity of a public actor's action executing the law. Dispute adjudication is also not enforcement when a public prosecutor is charging the defendant with violations of the law.

Within this definition, domestic delegation of dispute-adjudication authority is primarily other-binding delegation, and thus minimally sovereignty compromising. Dispute adjudication does not require that a court's jurisdiction be compulsory. But since states are binding others--firms, citizens, et cetera--they usually have no qualms about making the judge's jurisdiction for this role compulsory. Because delegation of dispute-adjudication authority is other-binding, the interests of both the state and the judge are aligned. When there are questions about the meaning of the law, the judge should be deferential to the legislative body that wrote the law. Since both the judge and the state want the parties to follow the law as it becomes legally defined, it is no surprise that states lend their coercive mechanisms to the task of enforcing judicial decisions.

At the international level, however, delegation of the same type of authority can be self-binding because a state's public policies might themselves become the subject of international judicial interpretation. Given that state policy might be subject to review, governments have historically been ambivalent about granting ICs compulsory jurisdiction in this role so as to be able to decide on a case-by-case basis whether they will submit to legalized dispute adjudication. (14) This ambivalence has been overcome it seems: increasingly, disputead-judication authority is coupled with compulsory jurisdiction that turns an IC's dispute-adjudication role into a sort of decentralized enforcement role. (15) With compulsory jurisdiction, states may get a more usable international disputead-judication system, one that is less easily blocked. But now the IC can be ruling on cases where a state is an unwilling party, and thus unsurprisingly, compliance with such rulings is more problematic. Compulsory dispute adjudication thus differs fundamentally at the international level compared to the domestic level primarily in that the interests of the state and of the IC are not aligned--the losing state may not want to participate in legal proceedings let alone comply, and since compliance is not per se in the interest of the noncomplying state (or other states in the system for that matter) there may not be any coercive support to enforce a ruling. (16)

C. Delegation of Enforcement Authority to Courts

Although it is commonly said that courts "enforce the law," it is always states, with a monopoly on the legitimate use of force, that enforce the law by punishing those who violate the law. States can enforce the rules on their own, using their extensive coercive power to punish those who violate their rules. In a rule-of-law system, however, the task of overseeing the legitimate use of coercive power is delegated to judges. In this "enforcement" role, the judge essentially monitors the state's use of its coercive power and thereby helps convince the public that the state is not abusing its power.

For the enforcement role, a court is given jurisdiction over a body of law, and a public prosecutor or enforcement body that charges a defendant with violating the law brings cases before the judge. If the prosecutor manages to convince the judge that the defendant violated the law, the judge can authorize a public actor to do what would otherwise be illegal and illegitimate--to deny a person his or her liberty, to seize his or her property, or to violate the law in retaliation. Since guilty parties are unlikely to submit themselves voluntarily to judicial proceedings about their behavior, enforcement roles require that courts have compulsory jurisdiction.

At the domestic level, the judicial-enforcement role is largely other-binding in the sense that defendants are likely to be private actors and judges are mainly being asked to hold police accountable to following the rules legislative actors set. But, delegating this role to judges is also self-binding to the extent that states are subjecting their use of police powers to judicial oversight. States minimize the sovereignty implication of this self-binding dimension by controlling the prosecutor. Criminal courts rule on cases only at the prosecutor's request; victims are not allowed to trigger legal proceedings. By making the prosecutorial office a political office, governments have a big say over which cases are brought to court for review.

At the international level, there are two very different types of judicial-enforcement roles. Criminal enforcement mimics its domestic counterpart--there is a public prosecutor, the court has jurisdiction over an enumerated list of crimes, and convicted criminals face prison terms. Clear examples in which international criminal enforcement is delegated in an other-binding way include victors' justice war-crimes trials, and ad hoc international tribunals set up by the UN Security Council (UNSC). These delegations are "other-binding" because the states delegating authority to judges knew they would not themselves be subject to the ad hoc criminal court's jurisdiction. The International Criminal Court (ICC) stands in sharp contrast, representing potentially self-binding delegation since its jurisdictional reach is not limited geographically or (post-2001) temporally. At the domestic level, the way to limit the sovereignty costs of delegation is by political bodies influencing or controlling the prosecutor. Appointment of international commissions or prosecutors can be influenced by powerful states, and the UNSC can put a six-month stay on a prosecutorial investigation. Moreover, the international prosecutor will need resources (financial and informational) to investigate crimes and compile cases. By withholding resources, rich states can greatly undermine the functioning of the ICC system.

A second international-enforcement role concerns law violations without violence, wherein the stigmatizing term "criminal" is intentionally not used. In the international context, one finds "infringement" mechanisms whereby an international commission triggers a legal proceeding and the judge determines if a state's behavior is incompatible with the requirements of the treaty.

Both forms of delegated enforcement authority can be harder for a single state to control at the international level, compared to the domestic level. International prosecutorial bodies see it as their job to pursue legal violations. Collectively, states are usually able to block prosecutions from proceeding. (17) But a single state may be unable to block a prosecutor or commission from proceeding. Given the risk, international safeguards have been added to international delegations of enforcement authority. For criminal enforcement, an international prosecutor may not proceed with a case if a domestic court has already given serious consideration to the crime. Thus, a state can escape ICC authority by prosecuting the crime in the domestic legal system. States manage international infringement authority by limiting the nature of the sanctions associated with legal violations. Sometimes international bodies can levy a fine or authorize financial retaliation against a state maintaining an illegal policy, and other times the legal ruling itself is meant to evoke social opprobrium by identifying an action as "illegal." In both situations, review of infringements is prospective--illegal behavior only becomes seriously costly should a state persist in violating the law.

These political safeguards do not apply to the morphed role of decentralized judicial enforcement. When international dispute resolution is coupled with compulsory jurisdiction, dispute adjudication easily morphs into an enforcement role that may actually be more sovereignty-compromising than explicit delegations of enforcement authority. Prosecutors can be politically dissuaded from raising a case, and their burden of proof is higher; they must show beyond a reasonable doubt that a legal violation occurred. It can be harder to dissuade a plaintiff-state from bringing a case than it is to dissuade an international prosecutor, and the plaintiff-state need convince the judge only that its interpretation of the law is correct, making the case perhaps harder to stymie. Thus compulsory dispute adjudication may, along with delegation to the ICC, represent the most sovereignty-compromising examples of ICs with explicit and de facto enforcement roles.

D. Delegation of Administrative Review Authority to Courts

Administrative review is the main judicial means to hold the actors implementing legislative policies accountable. This delegation is other-binding in that the actors who write the law (legislatures) are using judges to monitor the actors that implement the law (administrators, or "the government"). The administrative-review logic tells judges to be deferential to the legislative body and to defer to the will of the legislative body over that of the public administrator as they interpret and apply the law. (18)

One can recognize a court with administrative-review authority from its jurisdiction. Courts with the authority to hear cases regarding the legality of a government action, policy, or regulation, or to hear "actions to annul" or "failure to act" charges regarding decisions or nondecisions of public implementers of the law, have administrative-review powers. Administrative-review courts have compulsory jurisdiction and private access so the actors affected by government decisionmaking can challenge arbitrary decisions. Administrative-court rulings generally do not substitute a specific judicial decision for the contested administrative decision; rather, they remand the case back to the administration so that it can try again to make a decision that will not be rejected by the court. Thus administrative review tends to be a fire-alarm system of oversight, akin to Bradley and Kelley's category of delegation-of-over-sight authority. (19)

Administrative review differs from constitutional review in that judges are not ruling on the validity of the law itself, but rather on whether a particular government decision or policy is congruent with the law, or whether the policy has been implemented in accordance with the law, or both. Admittedly a fine line separates constitutional and administrative review, and in political systems in which all courts feel free to practice constitutional review, the lines can become quite blurred. But the difference between administrative and constitutional review has also been made distinct in both domestic and international contexts. (20)

There can be great variation in the extent of administrative check created through administrative review. Some administrative-review systems have a narrow standard of review where courts check only that proper procedure was followed or that the decision was not "arbitrary and capricious" in its application to the litigant, or both. This narrow standard of review tells courts to grant administrators significant deference in how they interpret and apply rules. Some standards of review are broad, with judges checking the facts and the interpretations of rules to see if the administrator made the correct decision. (21) The broader the standard of review, and the harder it is to change laws underpinning administrative rules, the greater the sovereignty costs associated with administrative-review authority.

International administrative review is in large part the "other-binding" tool of divided government that one finds in the domestic realm. When ICs are reviewing only the decisions of international institutions--like the Seabed Authority, the Andean General Secretariat, the General Secretary of the Common Market of Eastern and Southern Africa (COMESA), or the European Commission (EC)--international delegation of administrative-review authority is not sovereignty-compromising even when coupled with private access so that the subjects of IO (international organization) administrative decisionmaking can challenge IO decisions. But international administrative review can also be sovereignty-compromising. The main implementers of international regulatory law are states, not international organizations. Examples of ICs' explicitly granted review authority over domestic administrations include the North American Free Trade Agreement (NAFTA) Chapter 19 panels (which are not permanent international courts). These are, by design, intended to review whether American, Canadian, and Mexican administrations and administrative courts have made correct decisions in subsidy and anti-dumping cases. The European Court of Justice (ECJ) and Andean Tribunal of Justice (ATJ) were also, from inception, designed to review decisions of supranational administrative actors and whether national administrators were implementing common policies correctly. (22) Delegation in these contexts was meant to both help implement complex international rules, and reassure other states that countries would not practice favoritism or undermine the meaning of their commitments during implementation. Explicit grants of international administrative-review authority tend to be coupled with private access, even though wider access rules can make delegation more sovereignty-compromising because they limit state latitude in interpreting legal rules.

E. Delegation of Constitutional-Review Authority

Although the rule of law requires that governments (like private actors) be held accountable to law, it does not require checks on law-making power. Indeed, philosophers like Thomas Hobbes and Jean Jacques Rousseau considered any check on sovereign power to be inherently problematic. (23) Political systems embodying this view include the United Kingdom, which has no constitution and no constitutional court, though it is certainly a "rule of law" country. Other philosophers (like John Locke) (24) believed that sovereign power ends up being exercised better when it is subject to checks and balances. Those who believe in checks and balances create constitutional political systems with constitutional-review mechanisms. Constitutional-review authority entails the power to nullify laws and policies that contradict the constitution. Committing to constitutional review is both a self-binding precommitment on the part of the legislature, and an other-binding choice made to bind future legislative actors and units within the political system to the constitutional bargain. (25)

Like administrative-review- and criminal-enforcement authority, constitutional-review authority can work only when the court's jurisdiction is compulsory. Unlike administrative review, constitutional review does not require private access. Indeed, in France, constitutional review exists without any right of private actors to instigate review. Delegation of constitutional-review authority is always sovereignty-compromising and by design, it shifts power away from those with majority control of the political apparatus so as to provide a check against majority rule. (26)

We can recognize an intentional delegation of constitutional-review authority to international courts through the grant of jurisdiction to nullify laws. As in the national realm, the delegation reveals an intent to limit what the international institutions can do in the future. The European Union (EU), the Andean Community, and the Common Market of Eastern and Southern Africa have political bodies that are, in essence, legislative bodies capable of creating rules, policies, and even laws that are directly binding on member states. The international courts in these political systems (the ECJ, the ATJ, and the COMESA court) were explicitly empowered to hear challenges to the collective decisions raised by member states or private actors. In these cases, raised either directly or for some ICs referred from a national court, the IC determines whether acts taken by these legislative actors are ultra vires (exceeding the authority of the bodies). (27) If a law were ultra vires, it would be nullified. In this example, states are self-binding against their own potential desire to use an international body expansively. Since European laws can be created based on qualified majority voting, supranational constitutional review can also be a means for the minority to challenge decisions of the majority. (28)

In the above examples, supranational constitutional-review authority does not necessarily compromise national sovereignty, so long as the court is reviewing the validity of supranational rules. But, international courts have also assumed a sovereignty-compromising constitutional authority to review the compatibility of national and international rules. Legal scholars call the phenomenon the "constitutionalization" of an international treaty, by which they mean that the treaty is elevated to a sort of constitutional (supreme) status by the rulings of the court. The ECJ's declaration of the supremacy of European law (mimicked by the AT J) was one such constitutionalizing act because it gave the ECJ the de facto authority to render national rules that conflict with European laws inapplicable. (29) Some see the creation of the WTO (World Trade Organization), and the WTO appellate body's jurisprudence, as constitutionalizing the WTO Treaty because it makes incompatible national laws too costly to maintain (though others disagree because countries can accept retaliation instead of changing conflicting laws). (30) Design changes undertaken and under discussion regarding the European Court of Human Rights (ECHR) have, according to some scholars, increasingly turned the ECHR into a supranational constitutional court that reviews the compatibility of national laws and practices with European human-rights rules. (31)

Constitutionalization of international treaties represents a case in which a court expands its initial authority and in doing so compromises national sovereignty. Whether constitutionalizing acts of ICs have the intended effect depends mostly on the reaction of the country whose policy is condemned. In many countries, governments are bound to international law, but there is no corresponding domestic rule or legislation to make international law, or IC rulings, binding within the national system. If governments choose to change "illegal" policies, or if IC decisions that rule national policies illegal have no national effect, in fact there may be no compromise of national sovereignty. If, however, national governments do not act but their courts accept an international decision as authoritative within the domestic realm, the international legal ruling can have a constitutional effect in the domestic system.

F. The Fundamental Risk of Delegation to Courts

This Part has defined four roles that courts play in the international political system. For some roles the authors of both the law and the delegation contract are binding others--using courts to help ensure that other actors (police, national administrations, private actors) follow the rules they created. Other-binding delegation is based on efficiency logic--states are using the legal process to monitor compliance with the law, expending their coercive resources only when a legal ruling on its own is insufficient to induce compliance. Self-binding delegation, on the other hand, is based on credibility-enhancement logic. Whenever publics might be suspicious of self-serving interpretations of the law by public actors, governments and legislatures can gain credibility by entrusting the interpretation of the rules to independent courts. (32) But it is likely impossible to make a delegation wholly other-binding. In making courts the keeper of "the law," governments create a rival body with the authority to say what the law means, and in exchange governments perhaps get some credibility as being committed to a rule of law. The key distinction is whether delegation to courts will be primarily other-binding in that it is mostly other actors--private actors, or state interlocutors--that will be subject to the decisions of courts. Table 1 below summarizes which delegations to courts tend to be primarily self-binding or other-binding, examining the domestic context separately from the international context. The international column shows that delegation can be designed to be both other-binding and self-binding. It is an empirical question whether specific delegations to ICs in practice end up more self-binding or other-binding.

If one compares the domestic and international columns of Table 1, it is clear that delegating the exact same tasks involves a greater sovereignty risk internationally than it does domestically. Here the limits of a domestic analogy are evident; diplomats making assumptions about ICs based on their knowledge of domestic courts may end up with unintended outcomes. But, regardless of whether a judicial role is primarily other-binding or mostly self-binding, delegation to courts involves a risk that judges will interpret the law differently than governments or legislative bodies might want, and a risk that judicial roles will morph over time. These risks are more problematic at the international level because international rules are very hard to rewrite, making legal rulings harder to reverse, and because any finding against a national law inevitably strikes at the heart of national sovereignty. Of course this is the whole point of international judicial review--to make it more costly for a country to defend the legitimacy of policies labeled "illegal" by an authoritative international legal body.

The discussion above identified a number of ways in which authors of the delegation contract can influence the likelihood that sovereignty will be compromised, meaning that states will find themselves bound by judicial interpretation. States can create restrictions on who can bring cases, on the types of legal arguments that can be raised, and on whether and how sanctions are associated with a finding of a legal violation. Dispute adjudication can be noncompulsory, requiring both parties to consent before a case proceeds to court. Prosecutors can be tightly controlled to limit the extent of enforcement delegation. Although administrative delegation requires compulsory jurisdiction and private access, legislators can create broad or narrow standards of review and broad or narrow rules of standing to bring a case. Access can be limited in constitutional review, thereby limiting the number and types of cases that can be raised. Finally, public actors can be exempted from certain types of legal challenges (for example, sitting government officials can be exempt, or states can be exempt from cases involving national security).

The differences in sovereignty costs domestically and internationally are captured graphically in Diagram 1, which also highlights some of the ways in which sovereignty costs are regulated. Note that the sovereignty costs are only "potential" costs--usage of the court combined with the willingness of judges to assert their authority will determine the extent to which delegation actually becomes sovereignty-compromising.

[TABLE OMITTED]

This diagram contrasts to some extent with the diagram in this symposium's introduction in which the function of the judicial roles--the monitoring of administrative and enforcement roles, and the adjudication of a dispute-adjudication role--are seen as relatively sovereignty-compromising compared to policy implementation or research and advice roles. (33) The difference is that I do not see binding others as per se sovereignty-compromising--especially if the "others" being bound are international as opposed to domestic actors.

III

DELEGATION TO INTERNATIONAL COURTS: THE EMPIRICAL RECORD (34)

This Part assesses the empirical record in delegating the four roles to ICs. Table 2 below lists the existing ICs that meet PICT's definition of an international court, organized by the year they were established. (35) The table indicates whether the court has compulsory jurisdiction, whether private actors have access so as to initiate litigation, and the number of cases the court has litigated. Where courts existed before 1990, the table breaks out the judicial activity since then. The PICT definition is stringent, requiring that a court be permanent to count as an IC. This table would be longer if it included quasi-judicial bodies or legal bodies that are not permanent (like NAFTA). Also missing from the table are seven African courts, (36) which mimic in design their European counterparts but which exist mostly on paper. If African courts were added in, along with legal bodies that are functionally equivalent to permanent courts, the trends discussed below would mainly be reinforced; there would be more delegation of administrative, enforcement, and constitutional roles to international legal bodies, and more often than not these international judicial bodies would have compulsory jurisdiction and allow private actors to initiate litigation.

Table 2 shows a proliferation in the number of ICs and in international litigation since 1990. Although other issues, like security, can be litigated before the ICJ--a general-jurisdiction court--the delegation pattern reveals that states have greatest comfort in delegating to ICs the interpretation of trade commitments and human-rights issues, including war crimes. Table 2 also shows a change in the design of courts over time. European courts account for the majority of international legal outputs. Usually the usage rates by European courts are attributed in part to the design of European courts--European courts have compulsory jurisdiction, and private actors can initiate disputes. (38) Table 2 paints a picture of ICs increasingly resembling the European design model of compulsory jurisdiction and private access. Twelve ICs allow private parties to initiate legal suits against state actors. Six allow nonstate actors--international commissions or prosecutors--to initiate disputes against state actors. The last two columns of data support the notion that ICs with compulsory jurisdiction and nonstate-actor access hear more cases, (39) but it also shows that not all ICs with compulsory jurisdiction and private-access ICs are equally active.

In fact, much of the design trend can be explained by the roles delegated to ICs. Table 3 below summarizes my findings on roles delegated to ICs, and on the design of the ICs for the given role. These classifications are based on an analysis of the Court Treaties defining the jurisdiction and design of the IC. To categorize a court, I looked for the grant of jurisdiction identified in the first column. For example, to be classified as having administrative-review authority, a court needed explicit jurisdiction in cases regarding the "legality of any action, directive or decision" of a public administrative actor (which often included authority to hear appeals for nonaction). Note that most ICs have been delegated more than one role, with each role defined in separate treaty articles. The rules regarding compulsory jurisdiction and access can also vary by role. For example, the International Tribunal for the Law of the Sea (ITLOS) has an interstate dispute-resolution role that lacks compulsory jurisdiction. ITLOS also has an administrative-review role with respect to the Seabed Authority for which it has compulsory jurisdiction, and access for private litigants. (40) The results are as follows.

A. International Delegation of Dispute-Adjudication Authority

At first glance, delegation of dispute-adjudication authority appears to be the most common form of delegation to ICs. But this appearance may mainly be a result of the "catch-all" nature of dispute adjudication--the fact that the other three roles need explicit definitions of jurisdiction or design elements to be classified in the role. Indeed, if one labeled as "decentralized enforcement mechanisms" all dispute-resolution mechanisms with compulsory jurisdiction, then enforcement would be the most prevalent role delegated to ICs (see the discussion of delegation of enforcement authority that follows).

The ICJ is the oldest IC on Table 2, and from inception it has served as a default international dispute-adjudication body, meaning that many treaties designate the ICJ as the dispute-adjudication body rather than create a new body for the specific treaty. The ICJ lacks compulsory jurisdiction, but there is an optional protocol whereby states can commit to compulsory jurisdiction (42) and countries can decide a la carte to make the ICJ's jurisdiction compulsory for specific treaties. (43) Thus, the ICJ's lack of compulsory jurisdiction is not per se a reason not to rely on the ICJ. But the ICJ is a general-jurisdiction body, with judges who could come from anywhere in the world. Although litigants have the option of appointing ad hoc judges for a specific case, if an agreement involves substance that requires specific expertise and pertains to only a handful of countries, the ICJ's general design may make it unattractive. Indeed, all of the other ICs with dispute-adjudication roles either cover more specific issues, include only a small group of states, or both, suggesting that these courts were created with the intent that they have a narrower, more-specialized jurisdiction than the ICJ.

It appears that states are increasingly abandoning the ICJ model of noncompulsory dispute adjudication. The new ITLOS court and the Organization of Arab Petroleum-Exporting Countries (OAPEC) court lack compulsory jurisdiction, but every other dispute-adjudication body has been given compulsory jurisdiction. Why have states agreed to a more sovereignty-compromising delegation of international dispute-adjudication authority? An analysis of which ICs have compulsory jurisdiction suggests an answer. States appear to make dispute adjudication compulsory primarily in economic agreements: seven of the ICs with compulsory authority are part of trade unions, (44) and two others handle primarily corporate investment disputes. (45) Of course, not all economic agreements have international dispute-resolution mechanisms, let alone mechanisms with compulsory jurisdiction. Analyzing trade agreements, James McCall Smith found that trade unions are more likely to be associated with compulsory dispute-adjudication compared to free-trade zones. McCall Smith reasons that the desire to capture the benefits of trade is driving decisions about the type of dispute-adjudication mechanism chosen. (46) Indeed, the concentration of dispute-resolution mechanisms with compulsory jurisdiction in economic agreements suggests that states especially want economic commitments to be enforceable. Having these agreements enforced through interstate dispute resolution, as opposed to private actors or an international-commission litigation, helps ensure that only cases the member parties really care about are litigated. When dispute-resolution agreements allow for suspending trade access as remedy (the Andean Community and WTO), reciprocity becomes the main force for compliance.

Only three out of the thirteen ICs with dispute-resolution authority also allow private access for this role. These cases appear to be designed to allow relatively small disputes to be handled outside of diplomatic channels. The Common Court of Justice and Arbitration for the Organization of the Harmonization of Corporate Law in Africa (OHADA) is mainly an appeals body for national rulings applying common corporate laws and regulations; allowing private actors to appeal national court rulings creates a non-diplomatic outlet to handle investor disputes. (47) The ITLOS generally lacks compulsory jurisdiction, but it has compulsory jurisdiction for disputes involving the seizing of vessels (48) and for contractual disputes between private actors and the Seabed Authority, perhaps so these issues do not become diplomatic controversies. The exception to both arguments above is the Central American Court of Justice (CACJ). It is a general-jurisdiction court pertaining to the countries of Costa Rica, El Salvador, Guatemala, Honduras, and Nicaragua; its jurisdiction is compulsory; and private actors can raise cases. (49)

Although dispute adjudication may be a prevalent role delegated to ICs, it is not per se the most-important or most-frequently activated international judicial role. It is hard to assess what percentage of each court's docket actually involves dispute adjudication. The most active ICs with this role--the ECJ and ATJ--break down the type of legal case by how the case was referred, creating categories of preliminary rulings cases (cases referred by national courts), infringement suits raised by the Commission or Secretariat, nullification suits (administrative review of IO outputs), and direct actions (cases raised directly in front of the IC). There appear to be very few straight-up interstate dispute-adjudication cases in the ECJ and ATJ. This is not really surprising. Dispute-adjudication cases may reach the IC as infringement suits, with states asking the Commission or Secretariat to pursue the issue instead of bringing a case themselves. The other dispute-adjudication courts are rarely used--with the notable exception of the WTO. (50)

B. International Delegation of Enforcement Authority

The enforcement role involves public prosecutors bringing criminal or infringement suits against states or their agents (government administrative actors or government officials). Fifty-five percent of ICs (eleven of twenty) have been delegated explicit enforcement roles. Delegation of enforcement roles are found in the three central areas of IC authority--international criminal law, trade law, and human-rights law. In Part II of this article, I argued that compulsory dispute adjudication can easily morph into a decentralized enforcement role, in which aggrieved states rather than central prosecutors bring suits to enforce the international agreement. (51) If ICs with compulsory dispute-adjudication authority but no international prosecutor to help enforce the agreement (for example, the WTO, (52) the Economic Court of the Common-Wealth of Independent States (ECCIS), (53) the Benelux Court of Justice (BCJ), and the Caribbean Court of Justice (CCJ)) were included, the number of ICs created with some enforcement role in mind would expand to seventy-five percent of all ICs.

At the international level, delegation of enforcement authority can become self-binding in principle and thus bring sovereignty risks. But a deeper look at the record of delegation shows that eight of the eleven delegations were designed to minimize risk. For two ICs, delegation of enforcement authority was coupled with political-control mechanisms: Inter-American Commission on Human Rights (IAHCR) countries can opt out of the court's compulsory jurisdiction and pressure the Commission not to pursue a case; for the COMESA court, a council of states must sign off before an infringement suit can be brought. Three of the delegations of enforcement powers are other-binding delegations to ad hoc criminal courts (Yugoslavia, Rwanda, and Sierra Leone Tribunals). Another three delegations couple enforcement authority with fairly weak sanctioning systems so as to minimize the cost of a legal loss--CACJ, COMESA, and European Free Trade Area Court (EFTAC) courts can find a violation but cannot authorize sanctions.

The exceptions to these statements include the ECJ, the ECHR, the ATJ, and the ICC. In each of these cases, submission to IC authority gradually developed over time through a series of sequential changes that each involved greater sovereignty costs. The ECJ's enforcement mechanism was originally combined with a "toothless" sanctioning system. (54) When national courts started enforcing ECJ rulings, Europe found itself to have, de facto, an uneven enforcement mechanism--countries with more-robust national judiciaries were held accountable to European law and were more likely to follow ECJ decisions, compared to countries with weaker national judiciaries. In 1998, European states added a European-level sanctioning mechanism to address this imbalance. Although any European state can be fined for noncompliance, the change was made mainly to confront the chronic noncompliers. (55) To date there have been very few fines levied, and there is little to suggest that this change has significantly affected compliance levels with ECJ rulings--mainly because national courts tend to enforce ECJ rulings directly. (56)

The ECHR has never had large sanctioning capabilities--its rulings primarily carry social stigmas. (57) The ECHR was originally designed to be politically controllable; the ECHR's commission was set up both to investigate charges of human-rights abuses, and as a gatekeeper to ensure that frivolous cases did not reach the court. Although at first reticent to refer matters to the court, (58) over time the commission became willing to refer more cases. By 1998, the commission was no longer gatekeeping--it referred nearly every plausible case to the ECHR. At that point, the commission mainly created an extra step in the process. Thus, states decided to abolish the commission as a first step to reaching the ECHR. Eventually membership in the Council of Europe, and submission to the ECHR's authority, became a signal that a government is committed to following a liberal democratic path (and thereby is a candidate for accession to the EU). (59) Thus, European states now willingly submit to the ECHR's authority so that they can obtain the benefit of being part of the European liberal democratic club.

The ATJ was created in 1984; (60) it took until 1996 for the Junta to be authorized by member states to bring an infringement suit against a state. In 1996, the Andean Pact adopted a number of changes to make the institution more accessible, including allowing private actors to raise infringement suits directly with the ATJ should the General Secretariat refuse to raise a suit. (61) This change made it harder for states to keep the General Secretariat from pursuing infringements, and it meant that the ATJ's sanctioning system was finally usable. As of June 2007, the ATJ has found sixty-one formal infringements of Andean rules, leading to thirty authorizations of retaliation. (62)

In contrast to the evolutionary development of Europe's courts and the ECJ, the ICC began as a true departure for international criminal justice. Historically, international criminal justice was victor's justice--other-binding delegation wherein the losers of the war were held accountable for their violations despite the victors having committed similar crimes. (63) Ad hoc international criminal tribunals were other-binding delegations in the sense that the actors supporting legal redress knew that they would not themselves be subject to international criminal-justice efforts. The ad hoc courts were nonetheless path-breaking delegations because they introduced a new model, one in which all sides were held accountable for their crimes. (64) Once the "all sides equally accountable" model was adopted, it was not clear why certain atrocities could have legal remedies (for example, crimes committed during the Yugoslavia, Rwanda, and Sierra Leone crises), while other crimes escaped punishment. The ICC is meant to generalize the "all sides accountable model," and it has been met with stiff resistance by some because the ICC self-binds states.

One could look at the gradual strengthening of certain international enforcement mechanisms and argue that there is a trend towards creating and using international enforcement mechanisms. Indeed, the enforcement roles of the GATT system, the Andean Community, the European Union, and the European Human Rights system were all beefed up over time to increase the opportunity and capacity of these ICs to hold states accountable to their legal obligations. (65) And, in the 1990s many states took the biggest plunge of all by committing to the ICC's general jurisdiction over all war crimes.

But if commitment to international legal enforcement is a sign of linear progress, one must note how lumpy and unequal the commitment often is. Enforcement mechanisms are the strongest for the issues the West cares most about--trade and mass human-rights atrocities. Outside of Europe, delegation of enforcement authority tends to bind the weak more than it binds the powerful. For example, the WTO's and ATJ's sanctioning system of allowing winning states to retaliate against states maintaining illegal trade barriers allows the rich to essentially buy their way out of compliance by accepting retaliation rather than complying. (66) The ICC allows states to escape its authority by prosecuting their own violators--which Western states are likely to do.

Although powerful actors have escape mechanisms to deal with IC enforcement authority, it is noteworthy that wherever countries have pre-committed to an IC's enforcement authority (explicitly or as general compulsory dispute adjudication), powerful and weak states have willingly participated in legal suits that are brought. This fact stands in contrast to the ICJ, in whose proceedings some countries have refused to participate (in cases for which the ICJ's jurisdiction was compulsory), forcing the ICJ to continue the case with no defendant present.

C. International Delegation of Administrative Review Authority

Forty percent of ICs (eight of twenty) have been delegated explicit administrative-review authority as indicated by the IC's authority to review the legality of any action, regulation, directive, or decision of a public actor, and its authority to question failures to act. The OHADA court is categorized as a dispute-resolution body since it does not have the explicit authority to hear challenges regarding the legality of a public decision, but it will primarily hear appeals of national-court rulings when the case involves a challenge to a public decision regarding a private firm. (67) If this court is added into the calculations, forty-five percent of ICs play an administrative-review role.

Most of the ICs with administrative-review authority are embedded in economic agreements (the exception to this rule is the ITLOS Seabed authority). All eight agreements with explicit delegation of administrative-review authority also create supranational administrators with the power to issue binding decisions. Thus, the delegation of administrative-review authority appears to be a direct attempt to extend to the international level the sort of legal protections found within domestic administrative states, and it appears to be primarily other-binding delegation (true for the ATJ, the BCJ, the CACJ, the Court of First Instance (CFI), COMESA, the ECJ, the EFTAC, and the ITLOS contexts). Sometimes, however, there is a self-binding dimension to this delegation too--found for the ECCIS, the OHADA, the ATJ, and the ECJ. In these cases, domestic actors end up applying international agreements, which has led to a concern that rules will be unevenly applied. This concern has led in turn to a decision to submit national administration of the specific international agreements to international supervision.

Administrative review requires compulsory jurisdiction and private access, so as to allow those affected by administrative rulings to challenge them. All of the ICs with administrative-review authority (both the eight ICs with explicit administrative delegations, and the two with implicit administrative-review delegations) have compulsory jurisdiction and private access for this role. Thus, administrative-review powers can account for ten of the twelve ICs that have private access and compulsory jurisdiction.

All ICs on this list play roles other than administrative review--be it dispute resolution, constitutional review, or enforcement roles. But it is noteworthy that the fewest caveats are placed on ICs in an administrative-review role as compared to other roles. One sees a lack of caveats in terms of access rules--administrative review and labor-dispute roles (68) tend to be among the few places in which private actors are allowed direct access to the IC. (69) One also sees a lack of caveats in that legal standing is rarely denied (in contrast to constitutional review for the ATJ and ECJ in which private actors must show that the law in question directly affects them). (70)

The busiest courts--the AT J, the ECJ, and the CFI--find themselves most occupied with respect to administrative-review cases--be they reviews of supranational administrative rulings, or reviews of national efforts to implement supranational regulations. This means that numerically speaking, administrative-review cases account for the lion's share of all international litigation (all CFI cases, all ECJ direct-action cases, most ECJ preliminary-ruling cases, all ATJ preliminary-ruling and nullification cases, all OHADA cases, and all BENELUX cases) (71)--thus roughly 20,032 of the existing 33,057 cases brought to all ICs). If one adds in the reality that seventy-two percent of ECHR rulings involve "access to justice" claims--charges that the national administration of justice is either too slow or insufficiently respectful of plaintiff's due-process rights (72)--it becomes clear that numerically speaking, most international litigation involves reviewing the actions of public implementers of rules and policies. Although international review of national administrative actions can compromise national autonomy, administrative review is mostly other-binding and thus not deeply sovereignty-compromising. In any event, when review is limited to the actions of IOs, no national sovereignty is compromised.

D. International Delegation of Constitutional Review Authority

This analysis considers only explicit delegations of authority to ICs, thus not the actions of ICs to expand or constitutionalize their authority. Four ICs (twenty percent of all ICs) have been granted explicit authority to review the legality of legislative acts. In all of the cases, the subject of review is designed to be IO outputs--and thus the delegation is primarily other-binding. This review role fits with Bradley and Kelley's delegation of "oversight authority"--providing states a means to oversee the actions of an international organization. (73) The ICs with explicit constitutional-review authority are primarily common-market bodies--the ATJ, COMESA, and the ECJ. (The CACJ also has constitutional-review authority.) All of these institutions have supranational political bodies that can exercise delegated legislative authority in that they can promulgate rules that are legally binding within domestic systems. International constitutional-review authority subjects this rulemaking power to constitutional review--which in most cases will involve reviewing the legality of actions of supranational legislative bodies. The circumscribed nature of this constitutional-review delegation perhaps explains why there are relatively few caveats limiting access to, or IC authority in, this role--all ICs with this delegated role have compulsory jurisdiction and private access.

The two other potential ICs in this category include the CCJ--whose role will be determined when the supranational Secretariat's and Council's powers are determined--and the ECHR, which some observers see as so completely changed from its initial enforcement design as to now fit in this category.

In sum, Table 2 presents a stark paper trend of creating ICs with both private access and compulsory jurisdiction. Deeper investigation of this trend reveals that wide access is mostly for other-binding roles--administrative review and constitutional review of IO outputs. Diagram 2 below maps actual delegations of IC authority onto the categories in Diagram 1 to capture which actual delegations are sovereignty-compromising. Remember that ICs can be delegated more than one role. The most extensive delegations of authority--in terms of the different roles ICs are given, and the sovereignty-compromising nature of the design of ICs--appear in economic agreements. It is interesting to note that the most active ICs are those where the sovereignty costs of delegation are highest. So one cannot conclude that sovereignty-compromising delegation is merely symbolic. It is also true, however, that not all rulings emerging from active courts compromise sovereignty, which may be why the sovereignty costs are more politically palatable.

[TABLE OMITTED]

There is strong evidence that states have tried to limit the sovereignty-compromising nature of delegation to ICs. But the fundamental risk of delegation to ICs remains. The original delegations of authority to the ECJ and ECHR were not so sovereignty-compromising. But the ECHR and the ECJ have ended up exercising their powers in ways that are deeply compromising of national sovereignty. (74) It was not so much the original grant of authority that created this outcome, but rather the bold assertiveness of the ICs as they exercised their authority. Indeed, the ATJ has the same structure as the ECJ--with the same delegated roles and even wider private access to trigger litigation--yet it has not exercised its authority in as sovereignty-compromising a way. (75) Meanwhile, the WTO's appellate body was granted only a dispute-adjudication role, and it lacks private access, but it has ended up ruling in ways that do compromise national sovereignty whereas similarly designed bodies (ECCIS, EFTAC, BCJ, CACJ, and COMESA) have not. (76) All of this suggests that the substance of the legal suits--itself an artifact of the court's role--matters more in determining the extent to which sovereignty is compromised than does the fact of delegation or the design of the IC.

IV CONCLUSION: THE SOVEREIGNTY COSTS OF DELEGATION TO INTERNATIONAL COURTS

This article has aimed to correct the impression of what delegation to ICs is about. Many scholars and practitioners assume that ICs primarily play an interstate dispute-adjudication role, along the lines of the ICJ. (77) Eric Posner and John Yoo go so far as to suggest that given the heterogeneity of state preferences, noncompulsory dispute adjudication is the only role ICs can play effectively. (78) Although dispute adjudication is a prevalent role delegated to ICs, it is not the only role, and dispute adjudication is increasingly combined with compulsory jurisdiction, making ICs more about enforcement--precisely what Posner and Yoo dislike. (79) In terms of IC dockets, interstate dispute adjudication clearly is not the most prevalent role ICs play.

This analysis raises theoretical challenges for existing theories of IC independence as it relates to compromise of national sovereignty. In Principal-Agent (P-A) literature, ICs are presumed to be agents of the states that create them, and independence is assessed in terms of the rules that shape the principal's ability to change the delegation contract. (80) Principal-Agent theory expects that the harder it is to sanction an agent through recontracting, the more independent the agent, and the greater the agent slippage. (81) The problem with this theory, as applied to ICs, is immediately apparent. The rules for sanctioning ICs through recontracting are largely uniform: international judges are appointed for short terms (four to eight years); changing international rules, the original delegation contract to punish judges, or a combination of both tends to be hard (requiring unanimity or super-majorities); and cutting budgets slows the administration of justice but does not affect judicial autonomy. Given the vast similarity of the sanctioning mechanisms for ICs, it is clear that recontracting rules cannot explain variation in the extent of charges of IC slippage. (82) The failure of P-A theory to explain the variation in IC slippage highlights a problem with the assumptions of P-A theory; it is not true that the more independent ICs, the more likely judges are to deviate from the wishes of the principal. Rather, judicial roles significantly define how judges approach their interpretive task. In some IC roles, courts are not really agents of the states but are mechanisms states use to oversee the behavior of others (IOs, or signatories to the agreement). In other roles, ICs are designed to challenge illegal state practices. ICs would lose their legitimacy as legal actors if they shied from their given role because of political pressure. If IC judges must choose between adopting a transparently political decision or accepting a political sanction, ICs may actually find the political sanction to be the more appealing option.

Legal literature does not focus on state ability to "sanction" judges through recontracting, but shares the assumption that independence is associated with slippage. Posner, Yoo, and to some extent Bradley and Kelley expect IC independence and thus the sovereignty costs of delegation to ICs to be shaped by whether international courts have compulsory jurisdiction and whether private actors are allowed to initiate disputes. (83) Posner and Yoo expect compliance with IC rulings to be less likely when a state is an unwilling litigant, and thus they expect ICs with compulsory jurisdiction to be less effective overall in inducing compliance with the law. (84) Table 2 shows a trend of creating ICs with compulsory jurisdiction and private access, and thus highly independent ICs. There is more controversy surrounding ICs today than in the past, but given that there are more ICs, and given the seventy-five percent rise in IC activity, the increase in controversial rulings is not surprising. It is hard to say that the design trend itself has led to an increase in ICs being charged with "running amok." Indeed, a number of ICs with compulsory jurisdiction and private access do not seem to generate controversy (for example, the ATJ, COMESA, the CACJ, and OHADA)--and not simply because these systems are not used. Meanwhile, ICs without these design features do seem to engender controversy--such as the ICJ, which lacks compulsory jurisdiction and private access, and the WTO and ICC, which lack private access. The reason is the same as above--certain roles are inherently other-binding (and thus not sovereignty-compromising) and in certain roles judges are more likely to be deferential to legislative intent, thereby avoiding compromising sovereignty or engendering controversy.

Nor does the extent to which sovereignty is compromised correspond entirely with the legal effect or sanctioning power of a court. All ICs considered in this analysis can issue binding rulings. (85) There are variations in enforcement mechanisms for IC rulings, but these variations do not seem to account either for variation in sovereignty risks or in respect for IC rulings. For two reasons, sanctioning power is not key: First, in all cases, courts rely primarily on voluntary compliance by the parties--indeed Martin Shapiro argues that all courts, from weak to strong, seek the consent of their parties, crafting rulings that offer each side the chance to claim partial victory. (86) Indeed, most actors follow IC rulings simply because the IC is the authoritative body charged with interpreting the law. Second, the stronger the enforcement mechanism, the less likely it is to actually be used. For example, ICJ rulings can be backed up by the use of force, but the United Nations Security Council has never authorized such a backup because doing so would be a drastic step of great political significance. Indeed, international legal systems with sanctioning mechanisms--like the systems of the WTO and ECJ--rarely invoke the sanctioning mechanisms, nor is it clear that the mere possibility of appealing to sanctions systematically increases compliance with legal rulings. (87) These reasons are why international lawyers like Louis Henkin, Abraham Chayes, Harold Koh, and Thomas Franck emphasize the legitimacy of legal rulings over the strong, direct sanctions such as the use of force or criminal punishment. (88) Chayes goes so far as stating that efforts to improve compliance by adding sanctions are a "waste of time." (89)

Rather than focus on contract design, this article examined the roles delegated to courts showing how contract design largely follows from the judicial roles delegated to courts. International Courts with explicit enforcement and constitutional roles require compulsory jurisdiction for these roles. International Courts with explicit administrative roles also require private access. Indeed, ICs granted constitutional, administrative, and enforcement roles were also granted compulsory jurisdiction (with the exception of the IACHR), and, in the case of administrative and constitutional roles, they were granted private access. Because the administrative and constitutional roles are primarily other-binding, little national sovereignty is being compromised through delegation--which is why compulsory jurisdiction and private access do not per se translate into a compromise of national sovereignty.

The assumptions about IC independence create misperceptions, which are then fed by a bias in American scholarship on ICs. Most scholars follow the political controversy--writing about rulings because the decision upset expectations or the desire of powerful actors, especially the United States or European states. The assumption is that controversial rulings are the most significant and sovereignty-compromising IC rulings. But really, the preferences of state actors, rather than the legal or policy significance of a ruling or the extent to which sovereignty is compromised, determine whether an IC ruling is controversial. American politicians have reacted strongly to WTO rulings even when the WTO rulings represented reasonable interpretations of the law, and the cost of the ruling was fairly insignificant in dollar and political terms. Meanwhile, when the ECJ extended the reach of European gender-equality provisions, ruling that the German constitutional ban on women in combat support roles violates European law, there was relatively little political controversy. The ECJ's ruling was deeply sovereignty compromising--requiring Germany to change its constitution and fundamentally change the German military--an institution extremely close to the heart of national sovereignty. (90) But, neither the ECJ ruling nor the constitutional change was controversial because many Germans supported increasing the role of women in the military. (91)

If, instead of following controversy, scholars followed the litigants, they would be writing more about ICs' involvement in private-public dispute adjudication, enforcement, and administrative review, and about how most of these rulings are exactly what states hoped for when they delegated authority to ICs. If scholars focused more on IC jurisprudence in its various judicial roles and on the political impact of the jurisprudence, we would have a greater sense of when and how ICs facilitate state compliance with international rules--which really is the only way to ascertain how effective an international legal system actually is. We may also find that whether or not a ruling is sovereignty compromising is epiphenomenal--that IC rulings that do not compromise sovereignty nonetheless generate controversy whereas highly sovereignty-compromising IC rulings elicit compliance without complaints.

Finally, this analysis reveals the limits of focusing on design features to explain IC behavior. This study can tell us what roles were explicitly delegated to ICs, but not what roles ICs come to play. It can tell us which roles, and thus which courts, are more likely to end up compromising national sovereignty, but not which ICs actually do end up compromising national sovereignty. In the end, the cases that are raised, and the audacity of judges in exercising their authority (or extending their authority) will ultimately determine when and to what extent national sovereignty becomes compromised by delegation to ICs.

Fearing that sovereignty will be compromised, conservative commentators condemn nearly all delegations to international courts. This analysis reveals the extent to which fearful critics like Robert Bork, Claude Barfield, and Jeremy Rabkin are offering as examples just a small sliver of what ICs actually do. (92) The point is not to eliminate self-binding delegation while retaining all other-binding forms of delegation--indeed, eliminating all risk would be impossible. Before we discard the baby with the bathwater, we would be better off considering the benefits and costs of delegation to ICs as a package deal. Delegation to ICs provides many benefits--far more than ICs as "simple problem solving devices" that can provide information that helps states resolve disagreements. (93) In the vast majority of cases, ICs are providing benefits by doing exactly what member states asked them to do--reviewing administrative decisionmaking, ensuring international institutions do not exceed their power, and enforcing international agreements so that states can capture the benefits of the treaties. In a small minority of cases, national sovereignty is compromised--often by design, but also in surprising ways. People will come out differently in weighing this balance. For some, simply the idea that an international actor can tell a national actor what to do is intolerable. But, it is worth pointing out the direction of the trend. The empirical record shows an increasing willingness to create and use ICs, suggesting that most states are quite comfortable with the balance of costs to benefits as it stands.

This Article is also available at http://www.law.duke.edu/journals/lcp.

(1.) This article adopts the definition of an IC created by the Project on International Courts and Tribunals (PICT): ICs are (1) permanent institutions, (2) composed of independent judges (3) that adjudicate disputes between two or more entities, one of which is a state or international organization. They (4) work on the basis of predetermined rules of procedure and (5) render decisions that are binding. For a discussion of this definition, see PICT's synoptic chart. The Project on Int'l Courts and Tribunals, Research Matrix (2000), http://www.pict-pcti.org/matrix/matrixhome.html (last visited Feb. 7, 2008). This discussion does not include certain African courts that exist on paper but do not yet exist in practice. See discussion infra note 35. Thus, if anything, this article underreports the trend of delegation to international legal bodies.

(2.) See Cesare Romano, The Proliferation of International Judicial Bodies: The Pieces of the Puzzle, 31 N.Y.U.J. INT'L L. & POL. 709 (1999). See also Jose Alvarez, The New Dispute Settlers: (Half) Truths and Consequences, 38 TEX. INT'L L.J. 405 (2003); Benedict Kingsbury, Is the Proliferation of International Courts and Tribunals a Systemic Problem?, 31 N.Y.U.J. INT'L L. & POE. 679, 680 (1999); Benedict Kingsbury, Nico Krisch & Richard B. Stewart, The Emergence of Global Administrative Law, 68 LAW & COMTEMP. PROBS. 15, 16 (2005); Chester Brown, The Proliferation of International Courts and Tribunals: Finding Your Way Through the Maze, 3 MELBOURNE J. INT'L L. 453, 454-57 (2002) (reviewing MANUAL ON INTERNATIONAL COURTS AND TRIBUNALS (Philippe Sands, Ruth Mackenzie & Yuval Shany eds., 1999)).

(3.) Karen J. Alter, Private Litigants and the New International Courts, 39 COMP. POL. STUD. 22, 25-27 (2006).

(4.) Compulsory jurisdiction and private access limit the ability of states to block a case from proceeding to court. These features are emphasized in Curtis Bradley and Judith Kelley's introduction to this volume as shaping the extent to which delegation to ICs is sovereignty-compromising. These are the critical features of ICs discussed in the debate over IC independence between Eric Posner, John Yoo, Laurence Heifer, and Anne-Marie Slaughter. See Laurence Heifer & Anne-Marie Slaughter, Why States Create International Tribunals: A Response to Professors Posner and Yoo, 93 CAL. L. REV 899 (2005); Eric A. Posner & John C. Yoo, Judicial Independence in International Tribunals, 93 CAL. L. REV. 1 (2005).

(5.) See discussion infra Part III.C.

(6.) Curtis A. Bradley & Judith Kelley, The Concept of International Delegation, 71 LAW & CONTEMP. PROBS. 1 (Winter 2008).

(7.) As many have shown, the voting thresholds required to reverse legal interpretation are particularly challenging to surmount because reversing legal rulings means disempowering actors who prefer the legally created status quo, and if reversal is perceived as political interference in a legal domain, defenders of the rule of law will rally to the side of judges. See Karen J. Alter, Who Are the Masters of the Treaty?: European Governments and the European Court of Justice, 52 INT'L ORG. 121, 136-40 (1998); Brian A. Marks, A Model of Judicial Influence on Congressional Policy Making: Grove City College v. Bell (1984) (1989) (unpublished Ph.D. dissertation, Wash. Univ., St. Louis) (on file with author).

(8.) of course this binding is somewhat fictitious, since the self-binding could be undone through a new constitutional act. The metaphor Jon Elster uses is that of Ulysses, who ties himself to the mast to avoid the temptation of the Sirens. See JON ELSTER, ULYSSES UNBOUND: STUDIES IN RATIONALITY, PRECOMMITMENT, AND CONSTRAINTS (2000).

(9.) The concept of an "ideal type" was introduced by Max Weber. "Ideal types" are intellectual constructs representing definitions that are logically controlled and conceptually pure. By definition, ideal types are not meant to represent reality. MAX WEBER, FROM MAX WEBER: ESSAYS IN SOCIOLOGY 59-60 (H.H. Gerth & C. Wright Mills eds., trans., 1946).

(10.) The categories underemphasize variation because within a single role (for example, administrative review, constitutional review, et cetera), different national designs and differences in the powers given to courts will be important in shaping how a court plays its role. For example, the political role of constitutional courts will vary based on whether constitutional courts have abstract judicial-review authority, concrete judicial-review power, or both. In addition, variation in how judges and legal cultures employ notions like standing, burden of proof, the standard of review, et cetera, will lead to meaningful cross-national variation despite the similarity in role across systems. The categories overemphasize variation across roles because, in practice, cases can involve multiple issues, leading a court to assume multiple roles within a single case. For an example of these differences, see ALEC STONE SWEET, GOVERNING WITH JUDGES: CONSTITUTIONAL POLITICS IN EUROPE (2000).

(11.) On the logic of appropriateness, see James March & Johan Olsen, The Institutional Dynamics of International Political Orders, in EXPLORATION AND CONTESTATION IN THE STUDY OF WORLD POLITICS 303, 311-12 (Katzenstein, Keohane & Krasner eds., 1999).

(12.) MARTIN SHAPIRO, COURTS: A COMPARATIVE POLITICAL ANALYSIS 17-20 (1981).

(13.) This desire to hold state actors accountable in similar contractual terms as private actors is so compelling that over time foreign sovereign immunity, a fundamental diplomatic courtesy, has been compromised. Policies like that expressed in the Foreign Sovereign Immunity Act of 1976, Pub. L. No. 94-583, 90 Stat. 2891 (codified as amended in scattered sections of 28 U.S.C.), have been passed to revoke sovereign immunity with respect to commercial interactions. This Act has reverberated through the international system, creating new doctrines that limit sovereign immunity. See MALCOLM N. SHAW, INTERNATIONAL LAW 628-38 (2003).

(14.) WERNER LEVI, LAW AND POLITICS IN THE INTERNATIONAL SOCIETY 70-71 (1976).

(15.) The World Trade Organization (WTO) system, for example, explicitly blurs the line into an enforcement role. The case starts as dispute resolution--both state parties pick panelists they prefer in the hopes of finding a middle-ground resolution. But the case can end as enforcement, with the WTO's Dispute Settlement Body determining the extent of the damage caused by the violating country's behavior and authorizing the victim state to do what would be otherwise illegal--to construct a purposely discriminatory and trade-diminishing barrier. For more on the WTO dispute-resolution system, see JOHN H. BARTON ET AL., THE EVOLUTION OF THE TRADE REGIME 67-74 (2006).

(16.) Although compliance is more problematic, it is not necessarily true that ICs are therefore less effective. Compliance is a poor indicator of effectiveness. See Kal Raustiala, Compliance and Effectiveness in International Regulatory Cooperation, 32 CASE W. RES. J. INT'L L. 387 (2000).

(17.) For infringement proceedings, commissions are highly susceptible to political pressure. The UNSC can block the ICC's prosecutor from raising a case, but even without this formal block, it is unlikely that a prosecutor will pursue a case when there is significant transnational political opposition to doing so.

(18.) This conceptualization of administrative review is consistent with the argument made by Barry Weingast and Mark Moran, and by Christopher Edley who gives as a subtitle to his book "rethinking judicial control of bureaucracy." See CHRISTOPHER EDLEY, ADMINISTRATIVE LAW: RETHINKING JUDICIAL CONTROL OF BUREAUCRACY (1990); Barry R. Weingast & Mark J. Moran, Bureaucratic Discretion or Congressional Control? Regulatory Policymaking by the Federal Trade Commission, 91 J. POE. ECON. 765 (1983).

(19.) Bradley & Kelley, supra note 6, at 22. Kal Rausiala makes a similar argument. See Kal Raustiala, Police Patrols & Fire Alarms in the NAAEC, 3 LOY. L.A. INT'L & COMP. L. REV. 389 (2004).

(20.) In the United States, Italy, and to some extent Germany, all courts feel empowered to reject laws that judges deem unconstitutional. In these systems, supreme constitutional courts serve primarily as appellate bodies. But many national systems maintain a strong distinction between administrative and constitutional review. These systems, as in France, only allow constitutional courts to conduct constitutional review.

(21.) Christopher Edley discusses how standards of review can vary. See Edley, supra note 18, at 96-129.

(22.) Indeed the ECJ's innovative preliminary-review mechanism was created for this purpose--to allow challenges to the implementation of European rules that were raised in domestic courts to be channeled to the ECJ for review. See Pierre Pescatore, Les travaux du "Groupe Juridique" dans la negociation des Traites de Rome, 34 STUDIA DIPLOMATICA 172 (1981).

(23.) See THOMAS HOBBES, LEVIATHAN; OR, THE MATTER, FORME AND POWER OF A COMMON WEALTH, ECCLESIASTICALL AND CIVIL (1962); JEAN-JACQUES ROUSSEAU & HENRY JOHN TOZER, THE SOCIAL CONTRACT, OR, PRINCIPLES OF POLITICAL RIGHT (Henry John Tozer, trans., 1905).

(24.) See John Locke, THE SECOND TREATISE OF GOVERNMENT AND A LETTER CONCERNING TOLERATION (1956).

(25.) ELSTER, supra note 8, at 115-18.

(26.) For more on the variety of constitutional delegations, see SWEET, supra note 10, at 47.

(27.) Treaty Establishing a Constitution for Europe, Dec. 16, 2004, 2004 O.J. (C310) 1; Treaty Creating the Court of Justice of the Cartagena Agreement ch. III, May 28, 1979, 18 I.L.M 1203; Treaty Establishing the Common Market for Eastern and Southern Africa art. 26, Nov. 5, 1993, 33 I.L.M. 1067.

(28.) Germany, for example, challenged the EU's Banana Protocol, which was passed despite its objections. Karen J. Alter & Sophie Meunier, Banana Splits: Nested and Competing Regimes in the Transatlantic Banana Trade Dispute, 13 J. EUR. PUB. POL'Y 362,367 (2007).

(29.) Joseph Weiler, The Transformation of Europe, 100 YALE L.J. 2403, 2416-19 (1991). See also Eric Stein, Lawyers, Judges and the Making of a Transnational Constitution, 75 AM. J. INT'L L. 1 (1981). The equivalent ATJ ruling is Andean Tribunal of Justice Interpretaciones Prejudiciales 1 IP 87, available at http://www.comunidadandina.org/canprocedimientosinternet/ interpretacion_prejudicial.htm (last visited Feb. 7, 2008).

(30.) Petersmann believes the AB's jurisprudence has constitutionalized the WTO, while John Jackson and Jeffrey Dunoff disagree. See Jeffrey L. Dunoff, Constitutional Conceits: The WTO's 'Constitution' and the Discipline of International Law, 17 EUR. J. INT'L L. 647 (2006); JOHN H. JACKSON, THE WORLD TRADE ORGANIZATION: CONSTITUTION AND JURISPRUDENCE (1998); ErnstUlrich Petersmann, The Transformation of the Worm Trading System Through the 1994 Agreement Establishing the World Trade Organization, 6 EUR. J. INT'L L. 161 (1995).

(31.) Laurence R. Helfer, Redesigning the European Court of Human Rights: From International Tribunal to Constitutional Court, Presentation at the NYU Law School Colloquium on Compliance with International Human Rights (Jan. 29, 2007).

(32.) The differences in logic is explained further in Karen J. Alter, Agent or Trustee: International Courts in Their Political Context, 14 EUR. J. INT'L REL. 33 (2008).

(33.) Bradley & Kelley, supra note 6, at 21, fig. 3.

(34.) This Part draws on material previously published in Alter, supra note 3.

(35.) See supra note 1 for PICT's definition. The year the treaty was signed is the year the court was established. Often courts were not created until a threshold number of states ratified the court treaty, thus there is a gap between the date of establishment and the date of creation.

(36.) Not included because of a lack of information are the Instance Judiciare of the Arab Maghreb Union, the Court of Justice of the East African Community, the Court of Justice of the Central African Economic and Monetary Community, the African Court on Human and Peoples' Rights, the African Court of Justice, the Court of Justice of the Economic Community of West African States, and the Tribunal of the Southern African Development Community. For more on these courts, see the website for the African International Courts and Tribunals, http://www.aict-ctia.org (last visited Feb. 7, 2008).

(37.) Courts that lack general compulsory jurisdiction usually have optional protocols which states can sign to commit to compulsory jurisdiction among signatory states.

(38.) E.g., Alvarez, supra note 2, at 425-27; Laurence Heifer & Anne-Marie Slaughter, Toward a Theory of Effective Supranational Adjudication, 107 YALE L.J. 273,288-91 (1997).

(39.) Robert Keohane et al., Legalized Dispute Resolution: Interstate and Transnational, 54 INT'L ORG. 457,474 (2000).

(40.) John E. Noyes, The International Tribunal for the Law of the Seas, 32 CORNELL INT'L L.J. 109, 118-26, 133-35,138 (1998).

(41.) Based on the author's coding. Please contact the author for more information about this coding.

(42.) Sixty-two states (out of 191 United Nations members) have agreed to the ICJ's compulsory jurisdiction, though thirteen limit their assent to issues other than cases arising from belligerent action.

(43.) This is how the United States came to withdraw twice from the ICJ's compulsory jurisdiction-first from the ICJ's general compulsory jurisdiction, and second with respect to the Vienna Convention on Consular Relations.

(44.) The World Trade Organization's Appellate Body, the Andean Community, the European Union, the European Free Trade Area, the East and South African Common Market, the Caribbean Community, and the Benelux Community.

(45.) The Economic Court of the Commonwealth of Independent States and the Court for the Harmonization of Corporate Law in Africa.

(46.) James McCall Smith, The Politics of Dispute Settlement Design, 54 INT'L ORG. 137, 145-50 (2000).

(47.) NAFTA has similar provisions for investor disputes. See, e.g., North American Free Trade Agreement, U.S.--Can.--Mex., Dec. 17, 1992, 32 I.L.M. 605, 639. NAFTA is not on Table 1 because its legal bodies are not permanent.

(48.) The owner of the boat may bring the suit, but the plaintiff's government must first consent for the case to go forward. Noyes, supra note 40, at 123.

(49.) For more information on the CACJ, see Project on International Courts and Tribunals, Central American Court of Justice, http://www.pict-pcti.org/courts/CACJ.html (last visited Feb. 7, 2008).

(50.) The lack of large dispute-adjudication case loads--with the notable exception of the WTO-may reflect a bias in the data. If most international disputes are dealt with through alternative-dispute-resolution bodies--via arbitration, diplomacy, or legalized dispute resolution undertaken by courts that are not permanent or that deal with private-actor disputes only--my reliance on PICT's categorization may actually hide where the action occurs. This bias may exist, but the data nonetheless suggests that most international adjudication by ICs does not involve interstate dispute adjudication.

(51.) See text accompanying note 13.

(52.) See discussion supra note 15.

(53.) The ECCIS enforcement role is specific; Article 32 of the Charter of CIS allows the Economic Court to help "ensure the observation of economic obligations." Commonwealth of Independent States Charter art. 32, June 22, 1993, 34 I.L.M. 1279. This has been interpreted by the ECCIS court as an enforcement role that pertains to any rule that gives rise to "tangible benefits" for a party. For more information on the ECCIS court, see Worldcourts.com, Economic Court of the Commonwealth of Independent States Jurisdiction, http://www.worldcourts.com/eccis/eng/jurisdiction.htm (last visited Feb. 7, 2008).

(54.) Federico Mancini & David Keeling, Democracy and the European Court of Justice, 57 MOD. L. REV. 175, 183 (1994).

(55.) JONAS TALLBERG, EUROPEAN GOVERNANCE AND SUPRANATIONAL INSTITUTIONS: MAKING STATES COMPLY 76-84 (2003).

(56.) See Tanja Borzel, Non-Compliance in the European Union: Pathology or Statistical Artifact, 8 J. EUR. PUB. POL'Y 803 (1994).

(57.) ECHR can award compensation, but not punitive damages, to victims. Thus the fine is rarely sufficient to serve as a deterrent. On ECHR fines, see DINAH SHELTON, REMEDIES IN INTERNATIONAL HUMAN RIGHTS LAW 14740 (1999).

(58.) The ECHR's early case load was as follows: from 1959 through 1969, 10 decisions; from 1970 through 1979, 26 decisions; from 1980 through 1984, 58 decisions; from 1985 through 1989, 111 decisions. Data from A.H. ROBERTSON & J. G. MERRILLS, HUMAN RIGHTS IN EUROPE 310 (1994).

(59.) See Andrew Moravcsik, The Origins of Human Rights Regimes: Democratic Delegation in Postwar Europe, 54 INT'L ORG. 217 (1994).

(60.) E. Barlow Keener, The Andean Common Market Court of Justice: Its Purpose, Structure, and Future, 2 EMORY J. INT'L DISPUTE RESOLUTION 39, 50 (1987).

(61.) These changes were adopted by the Protocol Amending the Treaty Creating the Court of Justice of the Cartagena Agreement, concluded on May 28, 1996, available at http://www.comunidadandina.org/normativa/tratprot/cochabamba.htm (last visited Feb. 7, 2008) [hereinafter Cochabamba Protocol] (amending the Treaty Creating the Court of Justice of the Cartagena Agreement, concluded on May 28, 1979, 18 ELM. 1203 (1979) [hereinafter ATJ Treaty[).

(62.) These represent the cases not settled out of court; there have been 201 reasoned opinions during the same time period (not all of which find infringements). Retaliatory sanctions in the Andean context are similar to the WTO--a state is allowed to suspend concessions against another state up to the authorized amount. Thus, noncompliance can be sanctioned only when states are interested in retaliating. See SECRETARY GENERAL, ANDEAN COMMUNITY, INFORME DE LA SECRETARIA GENERAL DE LA COMUNIDAD ANDINA 2006-2007 (June 14, 2007) (on file with author).

(63.) GARY JONATHAN BASS, STAY THE HAND OF VENGEANCE: THE POLITICS OF WAR CRIMES TRIBUNALS 8-16 (2000).

(64.) The International Tribunal for the former Yugoslavia is perhaps the clearest case of this model, though the principle that all sides are accountable certainly holds for the ICTY's Rwandan and Sierra Leone counterparts.

(65.) GATT's dispute-adjudication authority started as noncompulsory. After the U.S. started unilaterally "enforcing" GATT rules, GATT states decided a more usable enforcement mechanism would be preferable. When the WTO was created in 1994, its dispute-resolution mechanism was made compulsory. On the WTO system, see BARTON ET AL., supra note 15, at 70-73.

(66.) Karen J. Alter, Resolving or Exacerbating Disputes? The WTO's New Dispute Resolution System, 79 INT'L AFF. 783, 786 (2003). See also Joost Pauwelyn, Enforcement and Countermeasures in the WTO: Rules Are Rules--Toward a More Collective Approach, 94 AM. J. INT'L L. 335 (2000).

(67.) Note that an IC can end up engaged in administrative review in other ways. Dispute-resolution cases, for example, can end up asking essentially administrative-review questions. ICs can exercise this review, but their rulings will not per se nullify the questionable administrative decision, nor will they be reviewing "failures to act."

(68.) A number of ICs have authority to adjudicate disputes between IOs and their employees.

(69.) The exception to this is the ITLOS body. Access is wide to the Seabed authority in this role, but the types of challenges are circumscribed:
   Without prejudice to article 191, in exercising its jurisdiction
   pursuant to article 187, the Seabed Disputes Chamber shall not
   pronounce itself on the question of whether any rules, regulations
   and procedures of the Authority are in conformity with this
   Convention, nor declare invalid any such rules, regulations and
   procedures. Its jurisdiction in this regard shall be confined to
   deciding claims that the application of any rules, regulations and
   procedures of the Authority in individual cases would be in
   conflict with the contractual obligations of the parties to the
   dispute or their obligations under this Convention, claims
   concerning excess of jurisdiction or misuse of power, and to claims
   for damages to be paid or other remedy to be given to the party
   concerned for the failure of the other party to comply with its
   contractual obligations or its obligations under this Convention.


United Nations Convention on the Law of the Sea art. 189, Dec. 10, 1982, 1833 U.N.T.S. 477.

(70.) There is no need to require heavy-handed remedies--administrative-review rulings primarily remand an action to the administrative actor, nullifying the existing decision and requiring them to issue a new one.

(71.) Andean Tribunal cases seem to be mostly about intellectual property--well over ninety percent of the cases. Laurence Helfer, Karen Alter, and Maria Flo Guerzovich have a project underway examining this activity. Laurence Helfer, Karen J. Alter & Maria Flo Guerzovich, Constructing an Intellectual Property Rule of Law in the Andean Community (unpublished manuscript, on file with author).

(72.) Rachel Cichowski, Courts, Rights and Democratic Participation, 39 COMP. POE. STUD. 50, 65 (2006).

(73.) Bradley & Kelley, supra note 6, at 22.

(74.) KAREN J. ALTER, ESTABLISHING THE SUPREMACY OF EUROPEAN LAW: THE MAKING OF AN INTERNATIONAL RULE OF LAW IN EUROPE 182-208 (2001): TALLBERG, supra note 55, at 92-127. See also Helfer, supra note 31.

(75.) Karen J. Alter, Exporting the European Court of Justice Model: The Experience of the Andean Common Market Court of Justice (2007) (unpublished manuscript, on file with author).

(76.) CLAUDE BARFIELD, FREE TRADE, SOVEREIGNTY, DEMOCRACY 37-69 (2001). See also Alter, supra note 66.

(77.) See Richard H. McAdams, Adjudicating in Anarchy: An Expressive Theory of International Dispute Resolution, 45 WM. & MARY L. REV. 1229 (2004); Richard H. McAdams, The Expressive Power of Adjudication, 2005 U. ILL. L. REV. 1045 (2005).

(78.) Posner & Yoo, supra note 4.

(79.) See id. at 66.

(80.) See Geoffrey Garrett & Barry Weingast, Ideas, Interests and Institutions: Constructing the EC's Internal Market, in IDEAS AND FOREIGN POLICY 173 (Judith Goldstein & Robert O. Keohane eds., 1993); Paul B. Stephan, Courts, Tribunals and Legal Unification - the Agency Problem, CHI. J. INT'L L. 333 (2002).

(81.) Darren G. Hawkins et al., Delegation under Anarchy: States, International Organizations, and Principal-Agent Theory, in DELEGATION AND AGENCY IN INTERNATIONAL ORGANIZATIONS 3, 27-31 (Darren G. Hawkins et al. eds., 2006).

(82.) See Karen J. Alter, Delegation to International Courts and the Limits of Re-Contracting Political Power, in DELEGATION AND AGENCY IN INTERNATIONAL ORGANIZATIONS 312 (Darren G. Hawkins et al. eds., 2006).

(83.) See Posner & Yoo, supra note 4, at 25; Bradley & Kelley, supra note 6, at 30.

(84.) Posner & Yoo's argument, supra note 4, has been ably critiqued. See Helfer & Slaughter, supra note 38.

(85.) The ability to issue binding rulings is inherent to PICT's definition of an IC. Supra note 1.

(86.) MARTIN SHAPIRO, COURTS: A COMPARATIVE POLITICAL ANALYSIS 8-17 (1981).

(87.) Eric Reinhardt and Marc Busch have found that states are most likely to make concessions before a WTO ruling is issued, so that it is in fact the hardest cases--those where compliance is least likely--that end up in court. Marc L. Busch & Eric Reinhardt, Bargaining in the Shadow of the Law: Early Settlement in GATT/WTO Disputes, 24 FORDHAM INT'L L.J. 158, 158-59 (2000-2001). There are certainly examples in which countries continued noncompliance up until the day that retaliatory sanctions would kick in, but in systems where enforcement mechanisms were added (such as the ECJ), there is little evidence that general compliance improved once sanctions for noncompliance became possible. See TALLBERG, supra note 55, at 57-68, 135-38; Borzel, supra note 56.

(88.) See THOMAS M. FRANCK, THE POWER OF LEGITIMACY AMONG NATIONS 193 (1990); LOUIS HENKIN, HOW NATIONS BEHAVE 9-10, 45-46 (1968); ABRAM CHAYES & ANTONIA HANDLER CHAYES, THE NEW SOVEREIGNTY (1995); Harold Hong Koh, Why Do Nations Obey International Law?, 106 YALE L.J. 2599 (1997).

(89.) Chayes & Chayes, supra note 88.

(90.) GRUNDGESETZ [GG] [Constitution] art. 12 (F.R.G.) (as amended through Dec. 12, 2000).

(91.) Case C-285/98, Tanja Kreil v. Bundesrepublik Deutschland, 2000 E.C.R. 1-69; Gerhard Kuemmel, Changing State Institutions: The German Military and the Integration of Women 7 (2003).

(92.) See ROBERT H. BORK, COERCING VIRTUE: THE WORLDWIDE RULE OF JUDGES (2003); JEREMY A. RABKIN, LAW WITHOUT NATIONS?: WHY CONSTITUTIONAL GOVERNMENT REQUIRES SOVEREIGN STATES (2005); JACK L. GOLDSMITH & ERIC A. POSNER, THE LIMITS OF INTERNATIONAL LAW (2005); Posner & Yoo, supra note 4.

(93.) Posner & Yoo, supra note 4, at 6, 19, 22.

KAREN J. ALTER, Associate Professor of Political Science, Northwestern University. The author would like to thank Curtis Bradley, Sean Gailmard, Oona Hathaway, Larry Helfer, Judith Kelley, Jennifer Landsidle, and Richard Steinberg for helpful comments on this paper.
Table 1: The Four Judicial Roles Compared

Judicial Role       Dispute                Criminal
                    Adjudication           Enforcement

                    When combined          Compulsory
                    with compulsory        jurisdiction,
                    jurisdiction,          prosecutor
                    dispute                initiates case.
                    adjudication
                    becomes
                    contract
                    enforcement.

Functional          Judge applies          Judge ensures
Role                state's law to         that public
                    resolve a dispute      authorities have
                    between private        reasonable
                    actors, or             evidence and
                    between public         grounds for
                    and private            punishing those
                    actors, radiating      who violate the
                    state social           law, granting a
                    control into           legal
                    private-law            imprimatur to
                    disputes.              state exercise of
                                           coercive
                                           authority.

How we know         Jurisdiction to        Jurisdiction in
it when we see      interpret the law      cases brought
it                  in concrete cases      by public
                    raised before it.      prosecutors or
                    No explicit            commission
                    authority to           regarding an
                    review the             enumerated list
                    validity of the        of crimes or a
                    law, or of public      set of rules.
                    acts. Cases are
                    raised by
                    disputants, not
                    by public,
                    prosecutor-type
                    actors.

Who is bound        Primarily other-       Primarily other-
by delegation       binding.               binding.
to domestic         Legislative body       Legislator is
courts              binds public and       creating
                    private actors to      oversight
                    judicial               mechanisms for
                    interpretation of      police forces.
                    rules set by the
                    legislative body.

Who is bound        Self-binding           Other-binding
by delegation       in the sense that      in the case of ad
to                  governments            hoc criminal
international       and in some            courts-the
courts              cases                  states creating
                    legislatures are       ad hoc courts
                    held                   usually do not
                    accountable to         fall under the
                    their                  court's
                    international          jurisdiction.
                    commitments.           Self-binding
                                           when all states
                                           fall under
                                           court's
                                           jurisdiction.

Ways to limit       Noncompulsory          By controlling
the                 dispute                the prosecutor,
sovereignty         adjudication           and the flow of
costs of            assures that           resources and
delegation          parties cannot         information to
                    be brought to          the prosecutor,
                    court                  governments
                    unwillingly.           can influence
                    States can also        the extent of
                    exempt certain         judicial
                    state actions          oversight.
                    from review.           States can also
                                           attach weak
                                           sanctions to
                                           adverse IC
                                           rulings, to
                                           lessen the costs
                                           of losing an
                                           enforcement
                                           case.

Judicial Role       Administrative         Constitutional
                    Review                 Review

                    Compulsory             Compulsory
                    jurisdiction;          jurisdiction;
                    private actor          access rules can
                    initiates case.        vary.

Functional          Judge oversees         Judge helps
Role                public                 check that
                    administrators to      legislative actors
                    ensure their           do not exceed
                    decisions were         their
                    made following         constitutional
                    proper                 authority,
                    procedure,             holding
                    consistent with        governments
                    the requirements       and legislatures
                    of the law, and        accountable to
                    are not arbitrary      constitutional
                    or capricious.         bargain.

How we know         Jurisdiction in        Jurisdiction to
it when we see      cases concerning       review the
it                  the legality of        validity of any
                    any                    legal rule of an
                    administrative         10, of a national
                    actor's                government, or
                    regulatory             of both.
                    decision, or
                    administrative
                    actor's "failure
                    to act."

Who is bound        Other-binding.         Primarily self-
by delegation       Legislature is         binding.
to domestic         binding                Constitution
courts              administrative         creates absolute
                    agencies to            limits on
                    follow their           legislative
                    rules.                 authority.

Who is bound        Other-binding          Primarily other-
by delegation       when states are        binding
to                  binding IOs to         whenever IC's
international       follow                 assess whether
courts              international          international
                    rules.                 acts are ultra
                    Self-binding           vires.
                    when ICs               Self-binding
                    oversee                whenever ICs
                    domestic               can assess the
                    application of         compatibility of
                    international          national rule,
                    rules.                 with
                                           international
                                           rules. Legal
                                           impact of an IC'
                                           ruling will he
                                           determined in
                                           large part by
                                           domestic
                                           system.

Ways to limit       A narrow               Limiting access
the                 standard of            for
sovereignty         review (for            constitutional
costs of            example,               challenges will
delegation          limiting review        limit the
                    to procedural          opportunities
                    issues or to the       courts have to
                    issue of whether       rule a law or
                    the decision was       practice
                    arbitrary and          unconstitutional.
                    capricious) will
                    limit judicial
                    oversight.

Table 2: International Courts, By Date Established

International        Date             Compulsory        Private-
Courts               Established/     Jurisdiction      Actor
                     Created                            Access

International        1945/1946        Optional
Court of                              Protocol (37)
Justice (ICJ)

European             1952/1952               X                X
Court of
Justice (ECJ)

European             1950/1959               X                X
Court of                                                (as of 1998)
Human Rights
(ECHR)

Benelux Court        1965/1974               X          Via national
(BCJ)                                                   court *

Inter-               1969/1979        Optional          Commission
American                              Protocol          is a
Court of                                                gatekeeper.
Human Rights
(IACHR)

Andean               1979/1984               X                X
Tribunal of
Justice (ATJ)

Judicial             1980/1980        So qualified      By optional
Tribunal for                          as to be          state
Organization                          meaningless.      consent.
of Arab
Petroleum-
Exporting
Countries
(OAPEC)

International        1982/1996        Limited           Seabed
Tribunal for                          compulsory        authority
the Law of the                        jurisdiction;     and seizing
Seas (ITLOS)                          Optional          of vessels
                                      Protocol.         only.

European             1988/1988               X                X
Court of First
Instance (CFI)

Central              1991/1992               X                X
American                                   (some
Court of                                exceptions)
Justice
(CACJ)

European Free        1992/1995               X          Via national
Trade Area                                              courts.
Court
(EFTAC)

Economic             1992/1993               X
Court of the
Common-
Wealth of
Independent
States
(ECCIS)

Court of             1993/1998               X                X
Justice for the
Common
Market of
Eastern and
Southern
Africa
(COMESA)

Common               1993/1997               X                X
Court of
Justice and
Arbitration for
the
Organization
for the
Harmonization
of Corporate
Law in Africa
(OHADA)

International        1993/1993               X
Criminal
Tribunal for
the former
Yugoslavia
(ICTY)

General              1953/1993
Agreement on
Tariffs and
Trade                                        X
(GATT);
World Trade          1994/1995
Organization
Appellate
Body (WTO)

International        1994/1995               X
Criminal
Tribunal for
Rwanda
(ICTR)

International        1998/2002               X
Criminal
Court (ICC)

Caribbean            2001/2005               X
Court of
Justice (CCJ)

International        2002/2002               X
Criminal
Tribunal for
Sierra Leone
(ICTSL)

Total
International
Judicial
Activity

Complete
Cases Only

International        Total Cases               Total Cases
Courts               (last year                Since 1990
                     included in               (primarily
                     figures)                  until 2006)

International        111                       50 cases filed,
Court of             contentions               30 judgments,
Justice (ICJ)        cases filed, 80           4 advisory
                     judgments, 20             opinions
                     advisory                  -2006.00
                     opinions
                     (2006).

European             5765 cases                3769 cases
Court of             referred by               referred by
Justice (ECJ)        national                  national
                     courts, 7908              courts, 3248
                     direct                    direct actions,
                     actions, 822              822 appeals,
                     appeals, 342              81
                     applications              applications
                     for interim               for interim
                     measures,                 measures, and
                     and 2860                  1943
                     infringements             infringements
                     (2006).                   (2006).

European             12,310 cases              11,640 cases
Court of             deemed                    deemed
Human Rights         admissible;               admissible;
(ECHR)               7528                      7323
                     judgments                 judgments
                     (2006).                   (2006).

Benelux Court        137                       42 preliminary
(BCJ)                preliminary               references
                     references                (2006).
                     filed (10
                     rejected)
                     (2006).

Inter-               162                       153
American             judgments, 19             judgments, 9
Court of             advisory                  advisory
Human Rights         opinions, and             opinions, and
(IACHR)              245 orders                243 orders for
                     for                       provisional
                     provisional               measures
                     measures                  (2006).
                     (2006).

Andean               25                        22
Tribunal of          nullifications,           nullifications,
Justice (ATJ)        79                        78
                     infringement              infringement
                     cases, and                cases, and
                     1163                      1152
                     preliminary               preliminary
                     rulings                   rulings (2006).
                     (2006).

Judicial             2 cases                   **
Tribunal for         (1999).
Organization
of Arab
Petroleum-
Exporting
Countries
(OAPEC)

International        14 judgments (2006).
Tribunal for
the Law of the
Seas (ITLOS)

European             5227 cases completed from
Court of First       6256 cases filed (2006).
Instance (CFI)

Central              78 cases, 23 advisory opinions,
American             and 55 rulings (2006).
Court of
Justice
(CACJ)

European Free        90 opinions (2006).
Trade Area
Court
(EFTAC)

Economic             83 decisions and opinions
Court of the         (2006).
Common-
Wealth of
Independent
States
(ECCIS)

Court of             5 judgments, 2 orders (2006).
Justice for the
Common
Market of
Eastern and
Southern
Africa
(COMESA)

Common               6 opinions, 111 rulings (2006).
Court of
Justice and
Arbitration for
the
Organization
for the
Harmonization
of Corporate
Law in Africa
(OHADA)

International        73 public indictments, 31
Criminal             completed cases, 46 judgments
Tribunal for         by the Trial Chambers, anti 24
the former           judgments by the Appcals
Yugoslavia           Chamber (2006).
(ICTY)

General              229 cases, 98 rulings from
Agreement on         GATT era.
Tariffs and
Trade                357 disputes formally initiated,
(GATT);              79 appellate rulings, 192 panel
World Trade          reports in WTO era (2006;
Organization         2005 for panel reports).
Appellate
Body (WTO)

International        27 cases in progress, 27
Criminal             completed cases, 7 awaiting
Tribunal for         trial (2006).
Rwanda
(ICTR)

International        4 situations under
Criminal             investigation, six warrants for
Court (ICC)          arrest issued (2007).

Caribbean            Began operation April 2005.
Court of
Justice (CCJ)

International        13 indictments proceeding; 2
Criminal             withdrawn due to death (2006).
Tribunal for
Sierra Leone
(ICTSL)

Total                38,995                    30,311
International        admissible                admissible
Judicial             cases filed.              cases filed.
Activity

Complete             33,057                    24,863
Cases Only           completed                 completed
                     decisions,                decisions,
                     opinions, or              opinions, or
                     rulings.                  rulings.

(1) The data was compiled by author, based on visiting the
websites of the international courts and consulting scholarship
as available. Bibliography available from the author. This table does
not include labor cases involving disputes with employees of IOs or
contempt of court decisions. Courts are constantly changing how they
report usage; thus one can find discrepancies over time.

* = no cases

** = data not available. Figures exclude staff cases.

(2) The GATT system changed significantly, going from a quasi-legal
body to meeting PICT's definition of an IC.

Table 3: Delegation of Different Roles to ICs (41)

                                                       Percent of Total
Judicial Role               ICs With This Role         ICs Explicitly
                            (See Table 2 for           Delegated This
                            full court names.)         Role (n=20)

Dispute Adjudication        ATJ *, BCJ, CACJ *,              13/20
General jurisdiction        CCJ, COMESA, ECCIS,
to  interpret the           ECJ, EFTAC, ICJ,
meaning of the              ITLOS, OAPEC,                     65%
law" or to "ensure          OHADA *, WTO
that the law is
respected"

Enforcement                 ATJ *, CACJ *, COMESA,           11/20
Jurisdiction                ECHR *, ECJ, EFTAC,
regarding an                IACHR, ICC, ICTY,                 55%
enumerated list             ICTR, ICTSL
of crimes or
jurisdiction to
hear infringement
suits against states.
Cases generally
are raised by a
public-prosecutorial
type actor.

Administrative Review       ATJ *, BCJ *, CACJ *,            8/20
Jurisdiction in cases       CFI *, COMESA *,
concerning the              ECJ *, EFTAC *,                   40%
"legality of any            ITLOS *
action, regulation,
directive, or
decision" of a
public actor, or
the public actor's
failure to act.

Constitutional Review       ATJ *, CACJ *,                   4/20
Jurisdiction to             COMESA *, ECJ *
review the validity         CCJ *?                       20% (possibly
of any legislative          Post 1998 ECHR *?                  30%)
act, regulation, or
directive of an IO.

Courts in bold have compulsory jurisdiction associated with the role.

Courts with a * have private access associated with the role.
COPYRIGHT 2008 Duke University, School of Law
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2008 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Alter, Karen J.
Publication:Law and Contemporary Problems
Date:Mar 22, 2008
Words:17243
Previous Article:The concept of international delegation.
Next Article:Sovereignty and delegation in international organizations.
Topics:

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters