The ILC Draft: Empowerment of International Prosecution

The 1994 the ILC draft was the accepted basis for negotiations. Attempting to reflect the perceived political consensus of the day regarding an ICC, it envisaged the ICC as a part-time institution to be activated when needed to deal with serious breaches of humanitarian law (war crimes), crimes against humanity, genocide, and aggression. Proposed as subordinate to national courts, the ICC would thus be empowered to intervene only if national systems were unavailable. Its jurisdiction depended on the consent of the state on whose territory the crimes had been committed and the state that had custody over the suspect (state-consent mechanism). With the exception of genocide, accepting the court's jurisdiction was to be optional. After ratification, states were to choose for which crimes (and for what period of time) they were willing to accept the court's jurisdiction (opt-in mechanism). The proposal was for a court closely affiliated with and readily available to the Security Council. The council had the power to refer cases to it irrespective of whether the states in question consented. Only if the council identified an act of aggression could the court instigate prosecution. Finally, the ICC could take up only cases the council was not seized with under its Chapter VII responsibilities. These structures severely hamstrung a court that largely preserved national sovereignty. The ILC's commentary reveals that its considerations were based on both “practicality” and achieving “the widest possible adherence of states” to the statute. ILC members believed that states would be unwilling to surrender significant elements of their national sovereignty to a supranational judicial body and that the support of the major powers would be ensured only if the court was readily available to them.Footnote ¹⁵

However, negotiations in the PrepCom (1996–98) indicated that the ILC draft was far from uncontroversial. Before long, two key negotiation blocks emerged. The first was composed of small and middle powers known as the Like-Minded (LM) group that was supported by a burgeoning nongovernmental organization (NGO) coalition, the Coalition for an International Criminal Court (CICC). This group called for a strong, independent court with an independent prosecutor, limitation of the Security Council's power of referring cases to the court, and automatic jurisdiction over the core crimes. The second group, under the leadership of the Permanent Members of the Security Council (P5 group), favored a kind of permanent ad hoc court along the lines of the ILC draft. Thus, in the run-up to the Rome Treaty Conference in June 1998, the draft statute contained ninety-nine articles and about 1,300 square-bracketed dissents on specific issues. Remarkably, however, during the night of July 17–18, the Rome Statute was not only adopted by 120 states in favor, with twenty-one abstentions and seven nations against (namely the United States, Israel, China, Yemen, Qatar, Libya, and Iraq), but it also contained many features that went considerably beyond the ILC draft.

The Rome Statute: Duty of International Prosecution

As established by the Rome statute, the ICC has jurisdiction over the crimes of genocide, the most serious war crimes, and crimes against humanity. In line with the principle of complementarity, the ICC steps in only if a state is unwilling or unable to prosecute matters on its own. The court has automatic jurisdiction over such crimes if either the territorial state (within whose borders the alleged crime has been committed) or the suspect state (nationality of the suspect) consent to or are state parties to the statute.Footnote ¹⁶ Three mechanisms were established to launch an investigation. First, any state party can file petitions concerning so-called “situations” with the prosecutor. Secondly, the Security Council can refer situations under Chapter VII of the UN Charter to the ICC, in which case the ICC automatically has jurisdiction without any state consent needed. The third trigger mechanism grants the prosecutor the right ex officio to start an investigation. Additionally, the Security Council has the authority to halt investigations (by a qualified majority vote, whereby the vote of the P5 must be unanimous) regarding situations within the scope of its responsibilities under Chapter VII of the UN Charter.

It is clear from the outlined features of the statute that the conservative model envisioned by the ILC was rejected by a large majority. On all matters regarding the independence of the court, the Rome Statute differs radically from the ILC draft. The statute's adoption was not only “a gift of hope for succeeding generations” as UN Secretary General Kofi Annan put it,Footnote ¹⁷ it also signaled a profound normative and institutional change. Although the statute formally founds an institution designed to enforce compliance with already existing norms, it also establishes a new norm: the duty to ensure international prosecution of serious violations of humanitarian law on the assumption that such crimes threaten international peace and security.Footnote ¹⁸ This change could be accounted for by different theoretical approaches.

Power-Based Approaches

In power-based approaches, such as the theory of hegemonic stability and broader neorealist thinking, (legal) norms and institutions are perceived as epiphenomena, reflecting the power relations underlying the international system. In essence, they are instruments hegemonic powers use in pursuit of their interests.Footnote ¹⁹ This implies that norms and institutions that strongly encroach on the national sovereignty of said powers are unlikely, especially where security issues are concerned.Footnote ²⁰ The ICC clearly addresses security policy issues, especially in military terms, as certain conduct is not only prohibited but even criminalized.Footnote ²¹ Additionally, the ICC interferes with the core of national sovereignty, namely “the ultimate application of the power of the state to persons within its territory.”Footnote ²² From a power-based perspective, such norms and institutions can be established only by the will of a hegemon or major power coalition that can control their application,Footnote ²³ so one would expect the creation and design of the ICC to reflect the preferences of the hegemon or major powers more broadly. The powerful states would have to retain control of the institution or, at least ensure the institution is so weak as not to interfere with their interests.

The history of international criminal tribunals would confirm this hypothesis. The Nuremberg and Tokyo tribunals as well as the ICTY and ICTR were all established at the will of the powerful and used against the weak or the defeated. Prosecution was clearly limited to crimes committed by the latter, while those who established the tribunals exempted themselves from scrutiny. The power-based explanation is further reinforced by the selectivity of ad hoc tribunals: tribunals were established for Yugoslavia and Rwanda, yet not for Uganda or Burundi. As political observers note, these tribunals were established not out of the conviction that prosecution was necessary but in order to avoid costly and risky military interventions that none of the major powers wanted.Footnote ²⁴ In this regard, one would have to understand the establishment of the ICC as an attempt by major powers to share the costs of setting up tribunals with the international community as a whole.Footnote ²⁵ Indeed, in the early 1990s major powers, and especially the United States, were skeptical of a permanent court.Footnote ²⁶ Only when a certain tribunal fatigue materialized in the face of the enormous administrative and financial costs associated with ad hoc tribunalsFootnote ²⁷ did major powers become cautiously supportive of the idea of a permanent court.

This support began to dwindle, however, the more the ILC draft was challenged during negotiations. The draft originally granted the Security Council clear control of the court and envisaged not a duty but an empowerment of international prosecution. Once negotiations started to favor enshrining a duty of international prosecution, major powers (China, Russia, United States, and, for a long time, Great Britain and France) began to oppose the ICC. However, their opposition did not prevent the norm being established within the new institution.Footnote ²⁸ This is puzzling, considering the ICC would have direct relevance for major powers as well. The jurisdictional rules allow for the prosecution of nonstate party nationals when they are suspected of having committed crimes on the territory of a state party. Thus, the faster the ratification of the statute proceeds, the more the national autonomy of all states is infringed. Such reasoning led the U.S. government to execute a policy of active opposition to the ICC, but it did not prevent the court from coming into existence.

Power-based approaches can explain the selective establishment of ad hoc tribunals and the initial interest of major powers in the ILC draft, but they cannot account for the legalization of a duty of international prosecution, that is, the ICC's institutional design.

Rational Choice-Based Approaches

Rational choice-based approaches such as neoliberal institutionalism view the creation of international institutions as a strategic choice by self-interested actors to resolve recurrent cooperation problems relating to the transaction costs of renegotiations and monitoring and enforcement problems. Recent research has shifted focus away from the creation of international institutions toward explaining their institutional design. The legalization project perceives legalization as a “particular set of characteristics that institutions may … possess”Footnote ²⁹ in terms of the obligations their rules create, the precision of those rules, and the degree to which the monitoring, application, and enforcement of rules are delegated to judicial bodies. The higher institutions place on all these levels, the higher the degree of legalization they exhibit. The ICC constitutes a case of high legalization of the “atrocities regime.” Its rules are obligatory and precise, and the ICC represents a delegation for the interpretation and application of rules.Footnote ³⁰

High legalization is particularly likely in situations where actors expect great benefits from cooperation but simultaneously face enforcement and monitoring problems.Footnote ³¹ High legalization reduces the transaction costs for subsequent interactions, strengthens the credibility of commitments, curbs cheating, and resolves contracting problems.Footnote ³² Similarly, the rational design project stipulates that the centralization of decision making within institutions increases as the severity of enforcement problems increases. But high legalization or centralization also imposes significant sovereignty costs on actors. An increase in the legal obligations imposed by rules and the delegation of their application to judicial bodies means that actors lose significant control over the application and interpretation of rules. The higher the sovereignty costs, the more limited the prospect for centralization.Footnote ³³ Sovereignty costs fall when greater control of the institution is facilitated by means of voting rules or veto powers, for example, or if flexibility is factored into institutions, for example, through escape clauses or opt-out provisions.Footnote ³⁴ Finally, rational design theorists expect control of an institution to be asymmetrical in line with the capabilities of different actors and their importance for the viability of an institution.Footnote ³⁵

Regarding the institutional design of the ICC, neoliberals stipulate that the prosecution of atrocities produces a collective action problem that can only be resolved by a “hard-law regime”—an institutional design involving high degrees of centralized decision making or delegation. However, as sovereignty costs can be expected to be rather high, one should observe high levels of asymmetric control of the institution by special voting rights or the inclusion of flexibility mechanisms such as opt-out provisions.

Before the ICC came into existence, it was the task of states to prosecute or extradite perpetrators. However, states were rarely willing to do so. The prosecution of crimes committed elsewhere incurred high costs and little return. Furthermore, smaller states especially feared that the prosecution of other states' nationals harbored the risk of impairing their relations to those states or even fomenting political crises. From a cost-benefit perspective, the national prosecution of international crimes is not a rational strategy,Footnote ³⁶ a fact that generated an incentive to centralize prosecution.

The general preference for a permanent court can hence be illuminated by a rational choice-based approach, but what about its specific design? The centralization of prosecution by a permanent court entails enormous sovereignty costs for actors, because it “impinges on the relations between a state and its citizens or territory, the traditional hallmarks of (Westphalian) sovereignty.”Footnote ³⁷ Logically, such sovereignty costs would obstruct centralization if they were not offset by high levels of individual control and/or flexibility provisions. The ILC draft provided extensively for both by the inclusion of opt-in and state-consent procedures for exercising jurisdiction and, finally, massive supervision by the Security Council. However, during negotiations these provisions were all abandoned or minimized in favor of a design with minimal controls and flexibility provisions despite a high degree of centralization and delegation. This choice then must be related to enormous attendant benefits.

The benefit of prosecution is its deterrent effect, which should help foster international peace and security.Footnote ³⁸ However, there is evidence that the prosecution of such crimes neither deters similar future crimes nor fosters peace and stability.Footnote ³⁹ Therefore, the ICC becomes puzzling for the rational-choice account for legalization: high sovereignty costs are weighed against uncertain benefits.

Still, this unattractive cost-benefit ratio does not apply to all states alike. In the 1990s, the majority of Western states did not face severe sovereignty costs. They had functioning national judicial systems and a low probability of involvement in massive violations of humanitarian law. Their costs were comparatively low because they had only a negligible risk of having to appear before the ICC. However, the constellation of interests during the negotiations was at odds with such an expectation. While a majority of Western liberal states was in favor of the ICC, others opposed it. More puzzling, many developing countries ended up vehemently supporting the court, despite having initially been somewhat reluctant. Plagued by military uprisings and failing state structures, these countries could realistically expect their heads of state to be summoned before the court, so they had to expect the highest costs. This constellation of interests cannot be explained by side payments, public pressure, or issue linkages.Footnote ⁴⁰

Finally, the fact that the United States—the state most crucial to the institution's success—was left behind is most puzzling of all.Footnote ⁴¹ Although the United States vociferously expressed its concerns, it failed to achieve any major design concessions. This puzzle also dogs another potential explanation for the conditions under which a change in the normative system of international law leads to a change in its operating system, that is, the creation of the ICC. Specifically, while operating system change necessarily hinges on a perception of necessity as well as political shocks to overcome the inertia within the system,Footnote ⁴² change is still highly unlikely when it threatens the self-interests of the dominant states or is actively opposed by these. If change nevertheless occurs, it should be so shallow as not to challenge the major powers' interests.Footnote ⁴³ In the case of the ICC, neither expectation materialized. The operating system was changed despite opposition of dominant states, and so profoundly that the United States deemed it necessary to challenge the new institution with an armada of bilateral and multilateral instruments, threatening to block UN missions if peacekeepers were not granted immunity from the ICC.

The adoption of the statute by the overwhelming majority of states as well as the swift ratification worldwide despite the active U.S. opposition indicates that neither the pre-existing power distribution nor the initial interests of states can sufficiently explain the emergence of a duty of international prosecution as reflected in the ICC's institutional design. Indeed, they suggest that those interests changed during negotiations.

Persuasion and Discourse

Such endogenous change is more inclined to a constructivist explanation of institutional change that focuses on persuasion and discourse. Constructivists argue that law cannot be reduced to a functional, cost-benefit equation; equally important is the inherent normativity of law, based on norms of due process, shared legal principles and broader normative understandings.Footnote ⁴⁴

However, constructivists have not yet developed a compelling conceptual understanding of the process of changes of interests, that is, persuasion,Footnote ⁴⁵ such as that offered by the broader debate on a discursive or deliberative understanding of law and democracy.Footnote ⁴⁶ Discourse theory views the creation of legal norms as the result of rational discourses. Discourses are expected to transform actors' preferences by the ‘noncoercive coercion’ of the better argument.Footnote ⁴⁷ This view presupposes that actors share a common frame of reference, which implies common knowledge and normative understandings, that is, that their life worlds sufficiently overlap.Footnote ⁴⁸ Furthermore, to ensure that agreement depends solely on the strength of arguments, a discourse has to ensure inclusiveness, equal communicative rights, sincerity, and freedom from repression and manipulation.Footnote ⁴⁹ Discourses need to be safeguarded against the asymmetric power resources of participants and everyone affected must be able to participate to ensure that the public interest prevails.

Because this type of lawmaking in a rational discourse is a (democratic) ideal, not a regular empirical occurrence,Footnote ⁵⁰ the question arises under what conditions persuasion and discourse affect real-world collective decisions. Even in national settings, its approximation is necessarily based on a legal institutionalization of discourses. Law enables and stabilizes discourses by minimizing the risk of exploitation by strategically oriented actors and guaranteeing free and equal access and fair compromises.Footnote ⁵¹ Institutionalization is the central means by which “practical reason is implanted in the very forms of communication and institutionalized procedures, [so] it need not be embodied exclusively or even predominantly in the heads of collective or individual actors.”Footnote ⁵²

While this internal linkage of law and politics is already fragile within national societies, it does not exist at all in the international political arena. Although fundamental institutions, such as sovereignty, pacta sunt servanda, and, more specifically, the culture of modern multilateral diplomacy and international public lawFootnote ⁵³ constitute a thin veneer of a shared life world, these features are meager compared to national societies. These institutions cannot substitute for the shadow of hierarchy within the state that is capable of minimizing or even offsetting the risk of dissension and exploitation innate to discourse.Footnote ⁵⁴ This situation poses doubts about whether discourse theory can indeed be applied to international politics.

However, institutions do not generate discourse quasi-automatically but should instead be understood as opportunity structures that foster trust, stabilize expectations among participants, and offer shared underlying normative understandings and procedural mechanisms that resemble some of the formal conditions of discourse.Footnote ⁵⁵ Institutions should be capable of leveling the power asymmetries between actors and guaranteeing equal participation and principles of public argumentation.Footnote ⁵⁶ That said, even such thin normative and institutional features are not an empirical regularity in multilateral negotiations, which are determined by power and information asymmetries as well as normative fragmentation.Footnote ⁵⁷ Actors engaged in negotiations may themselves attempt to alter the normative and institutional setting of negotiations to foster persuasion and discourse.Footnote ⁵⁸ These attempts can focus on (1) conditions of life world overlap (shared normative and factual background understandings) and (2) the procedures of rational discourse (inclusiveness, publicity, equality, and fairness).

Such attempts can be strategically motivated, as a shift from bargaining to discursive interaction also changes the underlying power structure of negotiations, weakening classical bargaining resources while strengthening the force of arguments. Such attempts might therefore be expected from those actors who will profit the most from discourse—the small and middle powers who, at least compared to major powers, lack sufficient bargaining weight—and even more so from NGOs. Interactions based on threats and promises essentially diminish their influence. NGOs in particular have only discursive resources: expertise, arguments, and publicity.

Constructivist studies on normative change have repeatedly pointed out how NGOs frame issues within negotiations to make them resonate with prior knowledge, agreed-upon principles, and norms,Footnote ⁵⁹ a strategy that is especially apt for lawmaking processes in which NGOs can rely on a body of well-known, broadly shared normative understandings and principles. These actors will try to adjust or enrich the life world conditions of negotiations to increase the likelihood of persuasion and discourse. By dispersing information and pointing to agreed-upon principles, they aim to establish a common frame of reference to which actors within negotiations can allude in evaluating arguments. Specifically, when brokering rights and duties, as with the ICC negotiations, only those norms that are in principle universally acceptable can lay claim to validity.Footnote ⁶⁰

Discourse theory would thus lead one to expect actors to seek to link negotiations to principles and norms that lend themselves to universality. Such reframing activities are closely connected to attempts to alter the institutional frame of reference. NGOs often try to change the institutional setting of negotiations if they are marginalized within the established setting. A well-known example is the so-called “Ottawa-Process.” Initially the landmine issue was dealt with in a security context, which is traditionally dominated by major powers, at the Geneva Conference on Disarmament (CD). Only after NGOs, in cooperation with a number of middle powers, initiated an alternative negotiation process outside the CD and reframed the landmine problem as a humanitarian issue did a comprehensive landmine ban become realistic.Footnote ⁶¹

Such alterations are most likely to increase the chance of persuasion if they approximate the conditions of discourse—if they increase norm density, transparency, inclusiveness, and equality; rely on consensus-based principles; and increase a sense of fairness and trust among participants. This example also indicates that NGOs frequently need to identify partners among progressive, often small, middle powers if they want to achieve major change in the institutional setting.Footnote ⁶²

Given the ideal character of rational discourse, the expectation is to discover “islands” of persuasion—approximations of discourse in certain phases and between certain actors during negotiations. While this rules out any notion of a rational consensus as the outcome of negotiations, such “islands” might fundamentally alter the course and result of negotiations. The case study on the ICC illuminates these dynamics. It shows how NGOs and a group of small and middle powers intentionally altered the institutional and normative setting of negotiations in order to encourage persuasion and discourse, and how this affected the negotiations, shifting the range of legitimate arguments within negotiations and the scope of possible outcomes.

Research Design, Methods, and Data

In order to apply this discourse approach, which is based on idealizations, to an empirical study, the research methodology, traces change back from the outcome (persuasion) to its potential causes.Footnote ⁶³ The first step is hence to detect possible candidates for persuasion (potential instances of an endogenous change of interests within discourse). The negotiations are then analyzed to figure out whether the theoretically inferred normative and institutional conditions conducive to persuasion and discourse correspond positively with these candidates (see Table 1).

1. 1. To identify possible candidates for persuasion, negotiations were screened for turning points, that is, phases in which crucial changes in state positions could be observed. Such turning points can indicate persuasion and discourse unless they can be explained by alternative factors, such as power, public pressure, side payments, issue linkages, or domestic factors such as elections or protests. To this effect, the positions of all participating states were traced over time.

2. 2. The next step was to evaluate whether the normative setting of negotiations changed and whether the direction of change shifted in favor of more generalizable principles and thus corresponded with the conditions of persuasion. The analysis traced the development of normative frames used by states to justify their positions over time.Footnote ⁶⁴ Similarly, changes in the institutional setting were traced to evaluate whether they were more or less conducive to persuasion and discourse (in terms of norm density, inclusiveness, equality, consensus, and fairness principles).

3. 3. The final step was to assess which kind of actors were responsible for changes in normative and institutional settings, and what effects these processes had on the negotiations.

TABLE 1. Retrospective methodology: Chain of causation and observable implications for the plausibility of the discourse approach

Obviously, this kind of analysis hinges on several caveats. Because the theoretical approach rests on idealizations and claims that approximations of these idealizations affect collective decision making, one cannot simply test for the existence of certain factors and contrast them with hypothesized effects. Instead, one has to carefully trace relative changes within these factors and determine how, if at all, they affect the course and outcome of negotiations. The analysis cannot completely prove that interests have changed. Since one cannot look into peoples' heads, it is nearly impossible to provide direct, reliable, empirical proof of an interest change.Footnote ⁶⁵ Instead, the analysis has to focus on indirect evidence to render such a change at least plausible. Therefore, in the absence of alternative factors to account for such changes, one infers an interest change from a change in position plus a change in the normative frame that actors employ to justify their position.

The analysis does not trace individual state-level changes. If one stipulates that certain normative and institutional setting changes increase the chance of discourse within which actors may change their interests in response to normatively superior arguments, it would be futile to focus on individual changes. Instead I focus on strong correlations between the theorized variables by simultaneously excluding potential alternative explanations for observed changes. Furthermore, the analysis uses a triangulation of different methods and data to improve the reliability of its assessments.

The analysis in this study is based on a qualitative content analysis of the ILC report of 1994 (to grasp the normative environment in the starting point of negotiations) and of all summaries of state speeches within negotiations as recorded by UN press releases and summary minutes from 1994–98Footnote ⁶⁶ (to establish changes in positions and normative frames).Footnote ⁶⁷ I also conducted a series of twelve interviews with delegations, NGOs, and UN officials to take account of alternative factors and institutional changes and to cross-check whether the interviewees supported the findings of the content analysis.

The content analysis focused on the opinions of the delegations concerning four issues, which were most closely linked to an obligation to pursue international prosecution, namely: whether delegations advocated or opposed automatic jurisdiction; whether they were for or against an independent prosecutor; whether the Security Council should have little or extensive control over procedures, and, finally, how they viewed the necessity and desirability of the ICC in general. The analysis traced the positions of all states over time with regard to the first three issues to detect potential turning points. The unit of analysis was the summaries of state speeches in negotiations. Since negotiations were soon dominated by two negotiation groups, coding followed this binary structure. Taking the inclusion of an independent prosecutor, for example, states could either support an independent prosecutor, which was coded as support for the LM group (LM), oppose the independent prosecutor, coded as support for the P5 group (P5), or they could hold a different position, coded as undecided/others (U). Each state and year was coded to arrive at estimates of the relative strength of the LM and P5 positions in negotiations over time.Footnote ⁶⁸

Apart from the analysis of state positions, the normative frames used by states to justify their positions were also relevant to how and whether the normative setting of the negotiations had changed. For this purpose, the fourth issue, focusing on the necessity and desirability of the ICC, was added to the content analysis.Footnote ⁶⁹ In this step, coding focused on only the justifications states gave in defense of their position (sample position: “The court should have an independent prosecutor”; sample justification: “Only an independent prosecutor can ensure the impartiality of the court”). This also meant that the size of the sample decreased since not all states justified their positions at all times.

These justifications were coded inductively. In a first round, state speeches were screened for the prominence of normative principles, which were then used to develop a scheme of categories for coding (including sovereignty, noninterference, universal adherence, independence, equality and impartiality, and universal applicability).Footnote ⁷⁰ Because the categories still overlapped, they were subsequently condensed into normative frames (political reality and public interest).Footnote ⁷¹ State speeches were then coded again in several rounds using the frames. Finally, the frames were weighted in their strengths in relation to all framings employed in the justifications in order to evaluate whether one frame gained dominance during negotiations.

Turning Points

The negotiations underwent two turning points. The first can be located in the move from the ad hoc committee in 1995 to the PrepCom in 1996. In contrast to the ad hoc committee, where especially major powers argued against an ICC,Footnote ⁷² no state publicly objected to the ICC as such in the PrepCom,Footnote ⁷³ though the type of court envisaged remained controversial, partly due to differing legal systems and cultural contexts. At its heart lay a conflict over the normative basis of the court, that is, over a potential obligation to pursue international prosecution including the question of how to instigate proceedings (trigger mechanisms), the role of the Security Council, and the type of jurisdiction.

A second turning point lies between 1997 and 1998, after the PrepCom sessions and before the Rome conference. During this period, positions changed dramatically. Until the end of 1996, a majority of states pushed for a very conservative ICC model. Led by the P5, this group called for a court under the control of the Security Council. This meant that the council would have to approve each case upon investigation (veto of P5). They also rejected the initial proposals for an ex-officio prosecutor, and they disliked automatic jurisdiction but favored several opt-in/opt-out mechanisms or a state consent regime to establish the court's jurisdiction.Footnote ⁷⁴ In essence, they wanted a kind of permanent ad hoc tribunal.Footnote ⁷⁵ The P5 defended this conservative position by arguing that their particular responsibility for peace and security should justify a special protection from the ICC.Footnote ⁷⁶

This approach was countered by a dozen states that then coordinated their positions, emerging as the LM group, strongly supported by the NGO network (CICC). The LM group, which grew to forty-two members by the end of 1997, soon established broad principles to which all its members could subscribe despite minor differences on specific issues.Footnote ⁷⁷ The group lobbied for automatic jurisdiction, an ex-officio prosecutor, and called for the Security Council's role to be limited to the referral of cases. A veto for the P5 was perceived as a politicization of the court and an undue interference in the workings of a judicial institution.

The majority of states, however, lay somewhere in between. While some, such as Syria or China, did not want any court, most had no coherent position, supporting LM positions on one issue and the P5 on another. By the beginning of the treaty conference in Rome in June 1998, however, majority opinion had changed drastically; the majority of states now backed LM positions for a strong and independent court, and this trend gained strength in the course of the conference. This increased support came especially from those regions that counted as the silent majority during the PrepCom meetings and were assumed to be reluctant to back an ICC.Footnote ⁷⁸ They included the Southern African Development Community, West African states (usually thought to be Francophile), and parts of the RIO group, comprising eighteen Latin American countries (see Table 2).

TABLE 2. Strength of support for LM and P5 positions in negotiations on three selected issues

Notes: Table is based on a content analysis of all summaries of state speeches during negotiations. LM = support for like-minded position. P5 = support for P5 position. U = undecided/others. N = total number of coded positions per issue and year.

These turning points cannot be explained by alternative factors: first, positions changed contrary to the existing material power structure. After all, it was the position of the major powers that lost support. Second, there is no evidence of side payments or issue linkages in negotiation protocols to account for this change, nor were they mentioned in interviews; nor did public pressure play a role, as negotiations were largely ignored by the public up until the Rome conference.Footnote ⁷⁹ Although some specific changes can be credited to domestic factors (for the second turning point the change of positions of the U.K. delegation and partly that of France followed changes in government in 1997), these factors cannot account for the overall swing, especially among developing countries. However, an analysis of the development of normative framings underlying positions in negotiations strengthens the hypothesis that the change may indicate persuasion and discourse, that is, a change in interests.

Normative Frames

The analysis of all available summaries of state speeches reveals that the different positions toward the ICC also reflect different normative framings. From the outset, there was unanimity over the criterion on which to base the ICC: it had to be effective, capable of performing its task. “Effectiveness” had already been granted center stage in all statements in the ILC, though its meaning was highly disputed. The analysis identified two frames, one emphasizing politics (and in particular security politics) and the other explicitly focusing on fundamental principles of law and morality.

For the political reality frame, effectiveness is equated with universal adherence to the statute. Only a court acceptable to all states would be able to duly discharge its duties: “Universal participation [is] essential for the authority and effectiveness of the court.”Footnote ⁸⁰ However, universal adherence can be achieved only if the court follows the political realities and not some principled ideals: “If we approach the court from an academically pure perspective, without regard for political realities and what states are willing to participate in and fund, we will have wasted our time.”Footnote ⁸¹ Only a step-by-step approach,Footnote ⁸² which accepts that sovereignty and noninterference are sacrosanct,Footnote ⁸³ could ensure that the majority of states would accept the statute: “The only durable basis for the development of such international cooperation is scrupulous regard for the fundamental principles of the UN Charter, notably the sovereign equality of the states, non-discrimination and non-interference in internal affairs.”Footnote ⁸⁴ But political reality is not only a principle of political wisdom. The frame posits political realities versus principled ideals. It refers to the power asymmetries within the international system that the court has to reckon with and which it should not amend but anticipate: “Realism dictates … that account must be taken of the diversity of regional interests, different stages of development and social and cultural traditions, and the position of major powers.”Footnote ⁸⁵ Additionally, political reality as framed by the major powers implies that international prosecution is a potential threat to efforts at securing peace and security: “The ICC can be really effective only if it … is inscribed into the existing system of maintenance of international peace and security and takes into account the prerogatives of the Security Council.”Footnote ⁸⁶ A judicial intervention into conflicts by the ICC is perceived as a threat as it might question the legality of decisions made by military leaders of members of the Security Council with regard to military interventions: “Soldiers deployed far from home need to do their jobs without exposure to politicized proceedings.”Footnote ⁸⁷ Political reality is a conservative frame that perceives the ICC as an instrument of security politics—an area in which traditionally major powers and power politics prevail. It highlights the political over the principled. Universal adherence does not refer to a moral argument (what everyone can expect from one another) but instead propagates a power-preserving approach that safeguards the particularistic interests of states and subsumes law under politics. This framing was the dominant reference of the P5.

The alternative frame interpreted effectiveness as resulting from universal applicability. This public interest frame was at the heart of the CICC's statements but was also taken up by the LM group. In this frame, the ICC should not bow to the existing power asymmetries but be given the power to override them. Only a court able to interfere “regardless of national, political or other identity of the perpetrators”Footnote ⁸⁸ could end impunity,Footnote ⁸⁹ and restore peace:Footnote ⁹⁰ “The rule of law precludes selective justice and means that victims should be able to seek redress for crimes of concern to the international community as a whole when a domestic system cannot provide it.”Footnote ⁹¹ Independence from and impartiality toward the political interests of states is a leitmotiv. The court is presented as a project of the public interest vis-à-vis the particular interests of states.Footnote ⁹² This presupposes, as its adherents formulate, “that we accept abandoning, to a certain degree, the classical concept of sovereignty of states.”Footnote ⁹³ Consequently, public interest clearly perceives the court as a condition for peace and security: “Without justice there can be no reconciliation and without reconciliation no peace.”Footnote ⁹⁴ In normative terms, while the political reality frame highlights politics as the guiding reference, the former public interest frame demands that the ICC “must not be clouded by politics' formulation of what the law is.”Footnote ⁹⁵ Public interest is based on the potentially universal principles of equality and impartiality, shared by all major law systems.

The ILC draft was clearly dominated by a political reality frame. Ensuring universal adherence was the predominant justification within the commentary.Footnote ⁹⁶ This frame remained dominant in deliberations in the General Assembly, where it could be identified in 57 percent of all justifications in 1995 and in 46 percent in 1996. However, in 1997 both frames were nearly equally represented (public interest with 31 percent; political reality with 28 percent). Indeed, at the beginning of the Rome conference public interest finally gained predominance, figuring in 47 percent of all statements while political reality stagnated at 28 percent (see Figure 1).

FIGURE 1. Strength of frames “political reality” and “public interest” in negotiations on “necessity and desirability of an ICC” in relation to all normative references

The frame analysis supports the hypothesis that the negotiations leading to the creation of the ICC were affected by persuasion and discourse. At least with regard to the second turning point the framing correlates with the shift in positions toward the ICC (see Figure 2).Footnote ⁹⁷ More importantly, public interest, a clearly principled frame relating to universal values and norms, central to all major law systems, gains center stage in the negotiations. This kind of argumentative reference corresponds to the kind of arguments expected from a rational (legal) discourse. As I will show, this successful reframing resulted from a combined effort of the coalition of LM states and the CICC to alter institutional settings to gain support for their frame.

FIGURE 2. Timeline of events

The Institutional Setting

Originally, the ICC issue had been delegated to the ILC, which is renowned for neither its progressive stand toward international law nor its speed. The ILC is formally an expert forum whose members are all law authorities in their own right. This setting seems particularly conducive to rational discourse because it decouples political decision making (with its potential distributive implications) from the creative part of finding new solutions to the problem at hand, (problem-solving) which it delegates to experts.Footnote ⁹⁸ However, as insiders to the UN claim, its discussions have always been deeply influenced by the political cleavages in the GA and especially the 6th (legal) Committee.Footnote ⁹⁹

The move to the PrepCom altered the institutional setting enormously. Basically, it placed the item on the public agenda of the UN. Nearly a hundred governments participated in the PrepCom, observed by an NGO coalition. However, owing to the nature of the issue, the PrepCom shared some of its features with the ILC. The establishment of an ICC required enormous expertise. It entailed not only questions of institutional design but also drafting an international criminal code. The majority of delegates were therefore legal advisors while often higher ranking political advisors were in the minority. Decisions were based on consensus. However, the first year hardly saw any consensuses reached. Instead, delegates produced a nearly 300-page compilation of divergent and often contrary proposals. While it prompted highly expert-oriented negotiations, the sheer number and complexity of issues meant that expertise varied enormously among delegations. Wealthy states could afford to send delegations of ten or more members who were able to focus on all issues, but developing countries especially could send only one or two delegates.Footnote ¹⁰⁰ Another problem was caused by the cultural implications of encounters between experts on common law, civil law, and the Sharia.Footnote ¹⁰¹

Overall, the degree of shared knowledge and normative understandings among delegations was rather low. Although the institutional setting was formally conducive to rational discourse (as it granted equal participation rights, was based on consensual decisions, and established the principle of public argumentation), this was offset by meager normative and factual underpinnings, as well as daunting inequality in terms of resources and expertise. The fundamental problem was the lack of effective participation by developing and transitional countries from Africa, Latin America, and Central and Eastern Europe. Given the complex nature of the issues, their delegations were hardly able to cover the entire gamut of negotiations, which were increasingly split up into working groups and formal, informal, and even informal-informal meetings. The lack of participation was also caused by a general mistrust that the ICC was being set up mainly to reinforce the dominance of major powers.Footnote ¹⁰² These fears were especially widespread among African countries, many of which had recently experienced violent upheavals and could not realistically expect to be spared from them in the future.Footnote ¹⁰³ As an NGO representative from Senegal put it: “Apart from its Southern part, the continent remains unmotivated by an ICC because of a lack of information, but also because many Africans feel the ICC is created in order to put their heads of state on trial.”Footnote ¹⁰⁴

This mistrust could not be resolved amid the tense negotiations in New York. In 1997, therefore, LM delegations, in close cooperation with the CICC, began to convene regional conferences to increase knowledge on the issue and respond to the special concerns of the respective regions. They organized conferences in Latin America, Africa, and Central and Eastern Europe to lay the groundwork for the Rome conference. These regional conferences were to allow open and free discussion—not necessarily to establish clear-cut positions on the specific issues—but to generate knowledge and an understanding of the main issues at stake.Footnote ¹⁰⁵

These conferences were quite similar in structure. Designed as learning forums, they strove to approximate the ideal of a rational discourse. They were not formally bound to direct political decisions (participants were not forced to settle on official negotiation positions), but were forums for deliberation and the free exchange of views among colleagues, with input from experts. The delegations, not to mention the majority of experts and NGOs, all came from the same region. They often knew each other already, and shared similar perceptions of the problems, a common history, and cultural values. The life world certainties were much richer than in New York: they constituted an almost ideal-typical discourse setting.Footnote ¹⁰⁶ It is striking that overall, from Pretoria via Dakar to Budapest, these regional conferences resulted in unexpectedly progressive positions toward the ICC that were almost identical to those of the LM group. In interviews, LM delegations saw these regional processes explicitly as persuasion:

The analysis of the institutional setting supports the hypothesis that legalization can be attributed to persuasion and discourse. The change in positions, at least at the second turning point, corresponds to changes in the normative frames observed between 1997 and 1998, which in turn correspond to altered institutional settings resembling the formal conditions for rational discourse to a remarkable extent (see Figure 2). The analysis also highlights the crucial role of the CICC and LM group.

CICC and LM States Fostering Persuasion and Discourse

The NGOs championed a public interest frame as early as the start of the 1990s and attacked the political reality frame from the beginning. For NGOs, the ad hoc tribunals set up under Chapter VII responsibilities were a means to achieve and restore peace: “Governments had made a lot of statements in the context of the Yugoslavian and Rwanda tribunals about international justice … and it was very easy to use these statements against them to push them further.”Footnote ¹⁰⁸

The political reality frame explicitly subordinated justice to peace and security. NGOs reversed the order so that justice was the prerequisite for security and peace, as impunity “only leads to renewed cycles of violence.”Footnote ¹⁰⁹ NGOs used the discrepancies between the political reality frame and the widely supported ad hoc tribunals to shame those statesFootnote ¹¹⁰—it became simply inappropriate to publicly oppose the ICC.Footnote ¹¹¹ So it was not persuasion and discourse, but shaming and blaming that brought about the first turning point. States simply abstained from publicly voicing their opposition without in fact changing their minds—a mechanism often described as rhetorical entrapment.Footnote ¹¹²

NGOs were also involved in altering institutional settings to promote their alternative frame—often in cooperation with the LM group, with which they started to forge bonds. For the LM group, this coalition had several advantages. First, security issues regularly disadvantage small and medium states in negotiations because security is traditionally seen as a central domain of great-power politics. The LM group could refer to the support of “civil society” for their position when seeking to legitimize it in the negotiations. Secondly, the LM group was able to draw on the strong expertise of NGOs to strengthen their arguments.

NGOs, on the other hand, used the coalition to gain access to even the informal meetings from which they were often barred, although the PrepCom had granted them access: “We of course had a lot of things leaked to us and we would comment on them and people started showing us drafts before they submitted them. Governments started to learn how to operate within this new diplomacy. They went to NGOs to leak the issues.”Footnote ¹¹³ This alliance became even more pronounced with regular consultations between NGOs and the LM group and cooperative strategies, culminating in the creation of “islands” of persuasion and discourse through the regional conferences.

Effects on the Negotiations

The approximations or “islands” of persuasion and discourse outlined here had far-reaching effects. They changed the structure of the negotiations, the range of legitimate arguments, and finally even the range of possible outcomes of the negotiations. Concerning the structure of the negotiations, the coordinated strategies of NGOs and LM states, especially in the regional settings, diminished the traditional lines of confrontation within the negotiations. In multilateral settings, negotiations are often characterized by an almost rigid confrontation and mistrust between different groupings such as the Western group, the nonaligned movement, or the Arab group. Outcomes in such settings usually reflect the lowest common denominator. In the Rome process, however, regional allegiances played a negligible role owing to the emergence of the LM group as a coalition of states from all regions, thus cutting across traditional cleavages. The regional conferences obviously supported this development:

The public interest frame that the LM delegations and NGOs promoted was at least potentially based on universal principles, which attracted support regardless of regional or cultural differences. The regional settings combined institutional features and argumentative structures: LM delegations and NGOs held positions that were universally appealing as they were common to all major law systems, and they developed institutional settings in which those positions could prevail.

With regard to the range of legitimate arguments, these islands of persuasion also had an impact on the kind of arguments perceived as appropriate within negotiations: the more dominant the public interest frame became, the less support alternative frames attracted. This effect can be demonstrated for the U.S. delegation, which was seen by many delegations as a constructive and well-prepared negotiation partner. The arguments over the special responsibility of the United States were partly considered legitimate, and both the LM countries and NGOs tried hard to find solutions to accommodate these concerns. They started to desist from these attempts, however, when the credibility of the U.S. delegation was increasingly questioned. Although they always argued for an effective court, their perception of what that meant was fundamentally different from the LM group:

The reframing efforts now bore fruit and public interest became the dominant frame. Effectiveness did not entail special interests or responsibilities, but universal applicability based on equality and impartiality. From this new perspective, the tack taken by the U.S. delegation, though it constantly based its arguments on a political reality frame, was perceived as hypocritical. Realizing that its argumentative efforts were not translating into sufficient support, the delegation switched to classical bargaining resources such as threats and promises. Smaller states were threatened with a withdrawal of military aid if they did not support the U.S. position.Footnote ¹¹⁶ Additionally, Republican voices within the U.S. Senate proclaimed that no court that could convict a U.S. citizen would ever be passed in Senate. While the fierce rhetoric from the U.S. Senate can reasonably be said to have belonged to the shared life world of multilateral negotiations in the 1990s, the U.S. government's threats provoked outrage on the part of the other delegations once they became public.Footnote ¹¹⁷ Thus the frame was set: arguments were taken into consideration only if they related to the effectiveness of the court, and that now spelled equality and impartiality.

Notwithstanding this impact, the most decisive effect was on the range of possible outcomes. Until 1997, the prospect was either to have no court at all or a kind of permanent ad hoc court. When the treaty conference began, a court involving a duty of international prosecution became a realistic alternative. But although majorities shifted, the process fell far short of consensus. While the LM group and the CICC succeeded in raising support for their position in the regions, the confrontation with the P5 remained. Faced with the stalemate between the two camps, the conference chairman, Philippe Kirsch, actively started to work toward a compromise. At the beginning of the fourth week, his bureau circulated discussion papers on the controversial issues, leaving out the extreme positions and listing several moderate options. Kirsch then invited all delegations to express their views on these options.

Following these debates the CICC disseminated public lists reporting the country positions on the respective topics. These lists made abundantly clear that the P5 position was increasingly a minority position.Footnote ¹¹⁸ Given these numbers and delegations' resistance to compromise, Kirsch's bureau announced that it would draft a comprehensive package as a final text for the ICC's statute that would be put to an up-or-down vote. Last amendments were then negotiated. Not only for Kirsch, but also for the majority of the LM states, the support of the P5 and especially the United States remained essential. Kirsch was willing to compromise on several issues, but the only way to get the United States on board was with a passage guaranteeing that U.S. citizens could never be prosecuted before the ICC that the other delegations rejected outright. After unsuccessful attempts by the Indian and the U.S. delegation to press for amendments, the Committee of the Whole finally adopted the statute in an unrecorded vote.