The United Nations ad hoc tribunals in the former Yugoslavia and Rwanda had primacy over national judicial agents for crimes committed in these countries during the most notorious civil wars and genocide of the 1990s. The UN Charter granted the Security Council the right to establish a tribunal for Yugoslavia in the context of ongoing civil war and against the will of recalcitrant national agents. The Council used that same right to punish individuals responsible for a genocide that it failed earlier to prevent in Rwanda. In both cases the Council delegated a portion of its coercive title to independent tribunal agents, thereby overriding the default locus of punishment in the world order: sovereign states.

By contrast, the 1998 Rome Statute gives the International Criminal Court (ICC) a coercive title that is “complementary to national criminal jurisdictions.”Footnote ¹ The ICC may prosecute and punish individuals for committing genocide, crimes against humanity, war crimes, and, eventually, aggression, but only when states are genuinely unable or unwilling to do so.Footnote ² Indeed, the complementarity principle limits the ICC's titular powers even relative to nonparty states sanctioned by the Security Council. If the Security Council refers a situation to the ICC (in a resolution based on Chapter VII of the UN Charter), as with Sudan (2005) and Libya (2011), the ICC does not thereby suddenly enjoy a primacy akin to that of the Yugoslav and Rwandan tribunals.

This distinction between primacy and complementarity may seem like a technical issue with little practical difference. For instance, the Yugoslav tribunal's title to investigate and prosecute did not alter state behavior for years. Yugoslavia, its successor states, and intervening Western states variously resisted, obstructed, or ignored the tribunal at different stages despite a Security Council resolution compelling them to do otherwise.Footnote ³ The ICC is similarly dependent on states' physical and material capacities and willingness to cooperate. Yet the political power of international courts depends also on the rules and principles that entitle supranational agents relative to sovereign state authorities. The authority to coerce is distinct from an ability to punish. With primacy, clear and semi-hierarchical lines of authority are set out for criminal tribunals. With complementarity, these lines are more political: they need to be asserted and defined in practice, and may be subject to intense conflict. As the other contributors to this roundtable show, states commonly obstruct or behave at cross purposes with the ICC. Under the terms of the Rome Statute, states have considerable latitude to discharge their duties as they see fit, and the ICC is authorized to counter perceived state resistance and noncompliance. Authority may be settled by legal principle, but political practice is unsettling, particularly when the legal principle is as open-ended as complementarity.

Drawing on Immanuel Kant's political theory, I outline two reasons the ICC's coercive authority matters for its practical and political success. First, for Kant, coercive authority rests on a noninstrumental criterion of success: the duty to establish and maintain a general system of rights. If this criterion is supreme, the ICC's coercive authority has moral legitimacy if and when it effectively supports (or substitutes for) the default role of sovereign states in systematic rights vindication. Moreover, whether the ICC succeeds in a particular trial or in effectively wielding its authority to withstand and counter political resistance by states is a secondary, instrumental concern.

Second, uncertainty and ambiguity about who possesses a final coercive authority can slow the pursuit of international justice, but it is not fatal to the noninstrumental quest for (re)establishing a sovereign state capable of systematic rights vindication. The ICC's complementarity regime is messy in terms of separating clear lines of authority. But as long as the ICC aims to help reconstruct national systems of rights vindication, it possesses what Kant calls a “provisional right” to coerce alongside national authorities. Among the ICC's first cases are Uganda and Kenya, and both illustrate the tensions and conflicts that the complementarity regime creates. While it may be tempting for anti-impunity advocates to argue that the ICC should behave as though it had a conclusive, final authority to bypass sovereign states in punishing atrocity crimes, I argue that the ICC's complementary regime is appropriate to its provisional moral authority to support the reconstruction of state sovereignty in the aftermath of atrocity.

Complementarity Conflicts: The Uganda and Kenya Examples

The ICC's involvement in Uganda and Kenya illustrates the political conflicts that inevitably arise when both sovereign states and the court exercise simultaneous coercive authority. These examples are snapshots of how the complementarity principle generates a distinct set of hurdles the ICC must navigate to succeed.

For obvious reasons, the Rome Statute prevents simultaneous legal proceedings at the supranational and national level for crimes committed by the same individual. Again, however, because the ICC lacks primacy, if a state has the will and ability to exercise its default authority, the court is prevented from asserting a supranational coercive title. But the process of adjudicating state and ICC assertions and evaluations about state ability and willingness is inherently conflictual. As William Schabas argues, “As originally conceived, the term ‘complementarity’ may be somewhat of a misnomer, because what is established is a relationship between international justice and national justice that is far from ‘complementary.’ Rather, the two systems function in opposition and to some extent with hostility vis-à-vis each other.”Footnote ⁴ In theory, the admissibility criteria set out in Article 17 of the Rome Statute should provide a clear-cut mechanism for settling the issue of when a particular state is genuinely unable and unwilling to act. This article gives the court the title to decide. But states retain a general coercive authority on their own territory, even when they have been declared unwilling or unable to deal with certain crimes. Moreover, even if states give the court authority by declaring themselves incapable and by initiating a so-called self-referral—something not specifically permitted or excluded by the Rome Statute—states maintain a legal obligation to prosecute the core crimes. Indeed, some criticize the ICC for accepting self-referrals where there is not a manifest incapability of states to hold trials (as in the Uganda case), and thus weakening the core notion of state obligation in the statute's preamble.Footnote ⁵ As Schabas writes, “If the prosecutor is sincere about his desire to stimulate national systems [to act], he might be better to send the case back, and give the State in question a lecture about its responsibilities in addressing impunity.”Footnote ⁶ Nevertheless, with both states and the ICC having a coercive title over a political situation, conflict is a predictable outcome and potential obstacle to success.

More needs to be said about a state's authority over a general political situation and the legal domain over which the ICC has authority. The Rome Statute creates a legal concept of the “situation” in which the crimes occurred.Footnote ⁷ However, even if the crimes are no longer being perpetrated, the national political situation often remains fluid and dynamic. In Uganda, for instance, ICC Prosecutor Luis Moreno-Ocampo accepted Uganda's self-referral in the midst of a long civil war. The ICC then indicted the leaders of the rebel Lord's Resistance Army (LRA) while the conflict was still ongoing. This decision was highly controversial, as many believe Uganda instrumentalized the court by legally cornering the rebel LRA and garnering international legitimacy for its own strategic advantage. It was also controversial because Prosecutor Moreno-Ocampo opted to delay consideration of investigating the government for its own human rights abuses and possible crimes.Footnote ⁸

Although Uganda never formally requested a return of the coercive titles it chose to relinquish through a self-referral, President Yoweri Museveni exercised the state's default coercive authority in ways that hindered the ICC practically. Thus, distinct from explicit noncooperation with the court, which can be masked in a number of ways, state decisions to manage or end civil conflict have effects that thwart pure legalism or a strict anti-impunity agenda. Politically, states are required to make many coercive decisions to uphold and/or rebuild a civil order that prevents future crimes. Thus, for instance, after having referred its situation to the ICC, Uganda entered into peace negotiations with the LRA and held out the possibility of amnesty in exchange for peace.Footnote ⁹ As both Kenneth Rodman and Benjamin Schiff note in their contributions to this roundtable, this can be justified on consequentialist grounds—that is, the need to stop the war and its harm to civilians. Indeed, many Ugandans have expressed a preference for ending conflict over punishment for LRA leaders, such as Joseph Kony.Footnote ¹⁰ But the same idea may be justified by a non-consequentialist reason: the need to reestablish a constitutional order or an effective general system of rights vindication. Whether this is best achieved with or without prosecution or criminal punishment for LRA leaders is a source of genuine conflict among Ugandans and between the state and the ICC.

International law typically assumes that states have an exclusive title to coerce. But politically states are neither monolithic nor unitary. This has implications not just for (non)cooperation with the ICC but also for contestation of the court's coercive title. If states are ensembles of institutions, with inter alia executives, legislatures, military, and police, then the ICC's coercive title within a state may encounter multiple points of resistance. In Kenya, for instance, police forces, worried about being blamed for the postelection violence, refused at one point to provide statements requested by the ICC prosecutor, despite an order to comply from Kenya's attorney general and ICC assurances that it was those at the top who were being sought out.Footnote ¹¹ Kenya's legislature also passed a motion to strip the court of its title by having Kenya exit the Rome Statute. Although the executive branch rejected Parliament's extreme option, leaders in that branch took clear steps to hinder the ICC when Moreno-Ocampo announced his intent to indict cabinet members and other political elites for crimes against humanity for the 2007–2008 postelection violence. Despite polls that indicated that as many as 73 percent of Kenyans supported the ICC's role and saw international prosecution as the only way to end a national culture of impunity, the government successfully lobbied the African Union to demand the ICC to stop its actions and make way for a national criminal justice mechanism. Kenya's government also requested the UN Security Council to defer ICC action for one year.Footnote ¹² These efforts have failed to hinder the ICC's ability to act on a coercive title, but they are nonetheless significant.Footnote ¹³

The Ugandan and Kenyan examples show that complementarity generates conflict between national and supranational titles to coerce in the same territory. In theory, this should be settled legally by the ICC judging whether, under the statute's admissibility criteria, there is a genuine inability and unwillingness of national authorities to prosecute the core crimes. In practice, such judgments will be challenged and resisted by national political actors with a plausible claim that they have the first duty to reconstitute a constitutional order of rights vindication. Given the challenges of this reality, I argue that the ICC should embrace its role as a provisional authority seeking to help reconstitute a state's general obligations as a sovereign authority. Next, I outline why—from a Kantian point of view—this is a difficult and yet vital role.

Kant on Coercion and Politics: Implications for the ICC

Kant's concept of coercion provides insight into the ICC's dilemmas and its role in world politics. His political thought is not an exercise in moralistic wishful thinking—that people ought to live in harmony. Rather, Kant's core ambition is to distinguish ethical from unethical coercive acts and actors. This is a difficult undertaking because, at one level, coercion is neither bad nor good but an intrinsic part of political life. But Kant thinks that because humans have a unique and moral capacity to choose, some forms of coercion are morally distinguishable from others based on the ideal of human freedom. Thus, while most see coercion in terms of instrumental threats of violence, harm, sanction, and punishment, Kant thinks of the concept in terms of noninstrumental political and legal relationships that give such threats a moral meaning and force.

As Arthur Ripstein writes, Kant sees coercion as “the limitation of freedom.”Footnote ¹⁴ Moreover, “An act is coercive if it subjects one person to the choice of another.”Footnote ¹⁵ A morally unjustified act is inherently coercive because it hinders one person's free choice and subordinates it to the choice of another, a violation of a person's humanity. Thus, morally justified coercion is a counterforce to another's first coercive choice: “Coercion is objectionable where it is a hindrance to a person's freedom, but legitimate when it takes the form of hindering a hindrance to freedom.”Footnote ¹⁶ Both aggressive and defensive acts are coercive, but only the latter are morally justified.

Kant defends the sovereign state because of its unique institutional capacity, at least as an ideal, to serve as an authorized coercive actor that systematically counters immoral coercion. Without the sovereign, each actor is free to do as he sees fit: one may act aggressively by hindering others' choices or one may act defensively to counter hindrances on one's own (or others') freedom. However, Kant thinks the aggressive/defensive dichotomy is unsustainable and meaningless in a “state of nature” context because every actor is free to make one-sided or partial judgments of the facts.Footnote ¹⁷ In supplanting the state of nature, the sovereign state is given primacy or conclusive title to coerce in a specific constitutional order.

Kant rejects the idea of a world state for a number of complex reasons.Footnote ¹⁸ But he argues that states have a moral duty to depart from a pure state of nature toward a peace federation and a semi-juridical legal order. In this order, a “provisional” title to coerce is possible among states even in the absence of an overarching world sovereign. States may resist hindrances to their own or others' freedom when they act collectively in a peace federation under limited conditions.Footnote ¹⁹ This has important implications for understanding the political ethics of the ICC.

On the one hand, based on his ideal theory, Kant rejects the notion that states, as authorized coercive actors in their own constitutional order, may be coerced by any other state or international agent except for defensive purposes.Footnote ²⁰ He rejects an imperial or hegemonic model where more powerful states intervene and project their own authority structures in the domestic affairs of an independent people. Multiple claims to authorized coercion in any state's constitutional order produce a contradictory situation and a backward step—a degeneration to the state of nature. On the other hand, based on a more pragmatic reading of political change, Kant acknowledges that no state is perfect and that both sovereign states and international law rely on a “provisional” title to coerce by common authorities in a transitional context, both internationally and domestically.Footnote ²¹ Kant's idea of a provisional title to coerce relaxes his ideal notions of politics while keeping sight of an overarching objective: establishing a unified coercive order based on freedom at the national, international, and supranational level.

Reconstituting Sovereign Authority: The ICC's Role

Since 2002 the ICC has tread cautiously to build legitimacy and support. Some think its first prosecutor, Moreno-Ocampo, has been too hasty and careless in his approach to cases in Africa. Again, however, conflicts inevitably arise when the ICC asserts its coercive authority in states' jurisdiction. Any prosecutor, no matter how prudent, will attract criticism for actions and nonactions conducted under the Rome Statute's complementarity principle. Kant's political theory encourages an evaluation of the court in broader, systemic terms. The challenges posed by states to the ICC under a complementarity regime are simply part of the political context in which the ICC can play its moral role. As Michael Struett argues in these pages, the ICC may pretend it is above politics to tactically cope with this reality. But managing a political context is not the chief moral role for the ICC, nor is it how its success should be measured.

Kant's notion of a provisional title to coerce provides an important explanation for the ICC's role. The complementarity regime permits the ICC to bridge—but not supersede—the tension between the ideal of coercive agents having a presumptive, conclusive authority over a singular constitutional or civil order and messier political realities of transitional justice.Footnote ²² These realities include war crimes, crimes against humanity, and genocide. States have a default primacy in terms of preventing and punishing these crimes within their own constitutional ambit; but the complementarity principle assumes that, because states are imperfect, they often have a title without capacity or have a capacity unworthy of the title. Such tensions are not bridged by misinterpreting complementarity as the foundation of a post-sovereign world of multilevel governance, where no actor has sovereign authority. Even though the ICC is rightly designed to be a permanent institution, the idea of a dispersion of the sovereign state's title to coerce to supranational authorities is problematic. Instead, complementarity should be viewed appropriate to a situation where common authorities—both states in transition and the ICC—have claim to a provisional title to coerce as the political situation evolves toward an effective system of rights vindication. The aim of both national and supranational authorities ought to be that of constituting (or reconstituting) the state to a situation of primacy, with the ICC eventually fading into the cosmopolitan background.

In playing a bridging role, the ICC would subordinate instrumental calculations surrounding a particular case or specific situation to the systematic and noninstrumental objective of establishing or restoring that state's best possible constitutional order. And, where feasible, it would be open to working with other provisional or common authorities within the state. But of course in some cases this goal would mean forcefully asserting a legal title to coerce against disingenuous states, as with Kenya. In other words, it would mean imposing international justice. But it could also mean, as Schabas argues, rejecting disingenuous claims about an inability to prosecute, such as Uganda's. In each instance, the judgments ought to be about the noninstrumental goal of (re)establishing a general system of rights vindication, no matter how difficult or distant a prospect this may be in some countries.

To some, the vision of the ICC working with other, possibly flawed, state authorities may conflict with or dilute the ideal of an absolute anti-impunity regime. Certainly, the ICC is obligated to pursue justice robustly and impartially and despite certain political realities, as Struett argues here. But even in playing a supportive, transitional justice role, the ICC is a significant advance compared to the reactive ad hoc tribunals of the 1990s. Ultimately, the ICC's provisional title to coerce may be insufficient to prevent and deter atrocities systematically. It may not serve as an especially effective tool for international conflict management or coercive diplomacy, as Rodman and Schiff observe. It may fail to eliminate impunity on a global scale, or even in “problem spots,” such as Africa. For these reasons, it is tempting to pretend that the ICC has the type of primacy or finality akin to the ideal of sovereign coercive actors. But there are problems and risks with misrepresenting the relationship between the ICC and states as essentially hierarchical. These problems and risks have rightly been criticized by those who worry that the ICC has acted as an imperial power over African states and societies that are coping with complex problems of political transition in nonideal conditions.Footnote ²³

While the ICC's authority is more uncertain and ambiguous than that of traditional sovereign actors or compared with the primacy given to the ad hoc war crimes tribunals of the 1990s, it has an important moral role as an agent of global transitional justice. There are bound to be political conflicts between states and the ICC under the complementarity regime. Kant's political theory, however, suggests that these conflicts are only morally relevant when they prevent the (re)constitution of a state's coercive order of systematic rights vindication.