A Response to Kevin Jon Heller by Carsten Stahn

[Carsten Stahn responds to Kevin Jon Heller, A Sentence-Based Theory of Complementarity. This post is part of the Third Harvard International Law Journal/Opinio Juris Symposium. Carsten Stahn is Professor of International Criminal Law and Global Justice at Leiden University and Programme Director of the Grotius Centre for International Legal Studies. He is inter alia editor of ‘The International Criminal Court and Complementarity: From Theory to Practice’, Cambridge University Press, 2011. ]

One Step Forward, Two Steps Back? – Second thoughts on a ‘sentence-based’ theory of complementarity

I am delighted to comment on Kevin Heller’s thought-provoking essay. I have been following his scholarship for a number of years. It is a great honor to engage with this challenging topic on ‘Opinio Juris’. His scholarship stands out for his obvious originality and critical engagement with policy dilemmas of ICC practice. This piece is no exception. I have no doubt that it will give rise to significant debate and might even divide scholars and practitioners engaged in the complementarity debate. I share certain hesitations regarding the central claim of the article. In my view, the argument that the ICC should focus ‘exclusively on sentencing’ when determining whether ‘ordinary’ crime prosecution is admissible is neither desirable nor manageable in all cases. I will focus on three aspects: The assumptions underlying the central claim, the desirability of a new methodology, and its manageability.

1. Underlying assumptions

In my view, some of the core assumptions underlying Heller’s claim merit critical reflection.

(i) Impact of the ICC

Heller challenges existing theories (i.e. ‘hard’ and ‘soft mirror’ thesis), based on the premise that the complementarity regime ‘pressures’ states to prosecute international crimes under an international label. This starting point is partly misleading. It overstretches the role of the ICC. The Statute does not per se oblige States to investigate and prosecute under an ‘international crimes’ label. Even under international law, the duty to penalize crimes does not necessarily coincide with a duty to prosecute a specific crime under the respective label. An ‘ordinary’ crime label might not be ideal since domestic law might not reflect all specific aspects of international crimes, capture their specific ‘gravity’ or provide ‘effective penalties’ that take into account their contextual elements. But the obligation to use a specific ‘international crime’ label results mostly from domestic provisions, such as prosecutorial duties to pursue the most ‘serious’ charges or charges that reflect the context and characteristics of the crime most suitably. This principle is largely independent of Western or non-Western cultures. It is thus somewhat artificial to ‘blame’ the ICC admissibility system for introducing a pull towards an ‘equivalence’ rule that is hard to comply with, or to associate this trend with a ‘Western’ bias.

(ii) Cost-benefit of ‘ordinary’ crime prosecution

Heller supports his criticism of the ‘equivalence’ principle by a generalized preference for ‘ordinary’ crime prosecutions. He argues that ‘national prosecutions of ordinary crimes are far more likely to succeed than national prosecutions of international crimes’. This claim is based on a slippery dichotomy. To present these two options as two diametrically opposed schemes is artificial. While ‘international crimes’ may be more difficult to investigate or prosecute, they also offer certain advantages which must be brought into the equation. An ‘international crime’ label might offer a broader basis for jurisdiction (i.e. prosecution of extraterritorial acts), curtail the applicability of statutes of limitation or extend the prospects for cooperation and judicial assistance. Moreover, in practice, domestic and international crime labels are de facto often interrelated in a domestic setting. Many jurisdictions rely on a mix of ‘international’ and ‘ordinary crime’ definitions in order to try offences, or adjust modes of liability to capture the conduct in question. These factors are not taken into account in the Heller’s ‘cost-benefit’ analysis. Paradoxically, in existing practice, ‘ordinary crime’ prosecutions (e.g. war crimes) are often criticized since they result in lower sentences. To reduce incentives for international crimes prosecution is thus partially counter-productive.

(iii) Flexibility v. checks and balances

Heller’s justification for the turn to a ‘sentencing heuristic’ is founded on the overall assumption that (even) wider flexibility for States to investigate and prosecute might create better compliance. This logic is open to challenge. It seems to overstretch the reach of the law. More often than not, the causes of action or inaction are influenced by contextual factors, such as the nature of the conflict, internal political factors (i.e. regime change) or general attention to atrocities.

The existing architecture of the complementarity regime foresees a nuanced system of checks and balances: compliance by threat and consensual burden-sharing. In practice, Heller’s test leads to a more determinate ‘all-or nothing’ choice. According to the application of the ‘sentencing’ logic, either the ICC or a domestic jurisdiction assume ‘ownership’ over the case. This leaves little space for burden-sharing or parallel action which might sometimes be an asset from a compliance point of view..

(iv) ‘Higher sentence’ – ‘better justice’

Finally, Heller’s theory operates on the critical assumption that a justice system based on ‘higher sentences’ provides better and more efficient justice than a system with potentially lower sentences. This vision appears to go against the very rationales of sentencing which typically preserves a great degree of flexibility in order to pay adequate tribute to individual interests. It is further ill-suited to provide an appropriate logic for forum choices in situations in which sentence and penalties may be of lesser importance, such as transitional justice scenarios.

2. Is a change of methodology desirable?

Given the existing status quo, it is questionable whether there is a need to introduce a new ‘heuristic’.

(i) Relevance of a ‘sentencing heuristic’

Due to the large number of crimes committed under ICC jurisdiction, conflicts over the prioritization of ‘ordinary’ v. ‘international crime’ prosecution are a relatively rare exception. In many existing ICC situations, states have entrusted the Court with the mandate to investigate and prosecute. In these circumstances, it would be largely unfeasible to force the ICC to encourage the exercise of domestic jurisdiction, based on the consideration that the domestic system might potentially impose a higher sentence. The Court cannot compel states to exercise jurisdiction, nor should it required to suspend its own proceedings, based on the speculation that domestic authorities might act – be it under an ‘ordinary’ or an ‘international’ crime label.

Potential conflicts might arise in cases of competing proceedings involving the ‘same conduct’, and specifically in cases where domestic investigations or prosecutions cover the same ‘incident’. Such conflicts can be avoided through prosecutorial selection practice and appropriate charging strategy. A ‘sentencing’ heuristic is not strictly necessary to save ICC resources. In many instances, the same result may be achieved through other constraints (e.g. budgetary (self-) restriction) or the proper exercise of prosecutorial discretion. In circumstances, where there is overlap, there is a range of different options. The Prosecutor might on his/her motion, decide not to proceed further, e.g. based on ‘new facts or information’ under Article 53 (4). The respective State might seek to convince the Prosecutor to ‘withdraw the charges’ (Article 61 (9). Moreover, ICC may defer to domestic jurisdictions if investigations or prosecutions are genuine. In the first two cases, the application of the ‘sentencing’ logic is not required. In the last case, it is doubtful whether the estimated gravity of the ‘sentence’ should be the ‘exclusive criterion’ to make this admissibility judgment.

(ii) ‘Hard’ cases

Heller mentions two examples where the existing regime might produce unsatisfactory results. The first is the charging of an ‘inadequate mode of participation’, and the second is an ‘overprotective’ use of domestic defences. Both examples are less clear cut than suggested, and might not necessarily require exclusive application of the ‘sentencing heuristic’. Even Rule 11bis jurisprudence before the ad hoc international criminal tribunals has left domestic authorities some flexibility in relation to the adjudication of modes of liability (Prosecutor v. Ademi and Norac, Referral Decision, 14 September, para, 46). The second scenario might often be resolved through application of the traditional admissibility criteria under Article 17 (2) (i.e. ‘purpose of shielding’, ‘intent to bring the person to justice’), or 17 (3), which already allows the Court to take sentencing considerations into account.

The most difficult scenarios are ‘borderline’ cases that relate to context. The resolution of such cases might deserve a more nuanced consideration than ‘sentencing’. For instance, there might be a legitimate interest for the ICC to consider the broader context of charges if a perpetrator is charged as an isolated actor, i.e. irrespective of a link to a State or organizational policy, to an armed conflict or the widespread or systematic commission of crimes. This finding might arguably provide a reason not to defer, if such a strategy ultimately limits responsibility to low-level perpetrators and prevents proceedings at a higher echelon.

(iii) Side-effects

Even in its ‘moderate’ form, Heller’s ‘sentencing heuristic’ produces some side-effects that are not necessarily desirable. It pays little attention to the interests of defendants. It might require the Prosecutor or Judges to make a ‘sentencing’ hypothesis even prior to the ‘confirmation of charges’, or the closure of the investigation. Moreover, it would implicitly encourage a ‘race to top’ in terms of penalties, both at the ICC and at the domestic level. Such a prioritization of ‘higher’ penalties through ICC procedure is not necessarily in the spirit of the Statute. The Statute is neutral in this respect as confirmed by the wording of Article 80.

The current debate on admissibility in relation to Libya illustrates some of the difficulties of Heller’s position. Taken to the extreme, Heller’s argument might be understood as an incentive for States to extend the death penalty for conduct underlying ‘core crimes’. This poses problems on several levels. The ICC is not a forum or appellate body mandated to remedy general human rights violations occurring in domestic criminal proceedings (i.e., akin to a human rights court). But in some cases, considerations related to the ‘proportionality’ of the domestic penalty might have to be assessed. Moreover, Heller’s logic would ultimately force the Prosecutor or Judges to justify a forum choice and deferral to domestic jurisdiction because that domestic system recognizes capital punishment as the main sentence. This positive endorsement of the potential application of the death penalty might be difficult to reconcile with Article 21 (3) of the Statute and Rule 11bis practice. Ultimately, the ICC might not be prohibited from deferring a case to a State applying capital punishment, if one takes the view that admissibility assessment should not involve consideration of potential violations to the detriment of the defendant. But even if one adopts that view, Heller’s proposal is not appealing from a ‘policy’ perspective. The problem with Heller’s ‘sentencing heuristic’ is that it places the ICC in the uncomfortable position of making that choice in the first place. By natural instinct, Court officials will be inclined to avoid entering into such determinations. This makes this test very hard to apply in practice.

Finally, Heller’s theory would ultimately treat states who reject the ‘death’ penalty less ‘favorably’ in terms of deference than States who apply it. This inequality will be hard to defend.

3. Is a change of methodology manageable?

Heller gives in-depth consideration to the question of whether the ‘sentencing heuristic’ would be manageable in light of the absence of internationally agreed sentencing standards. The problem may not only lie in the determination of an anticipated international penalty, but in the prediction of a domestic sentence by the ICC. This determination exercise involves a great degree of uncertainty and certain risks.

(i) ‘Domestic’ sentence assessment

Heller’s proposal might require not only Judges, but also the Prosecutor to determine an estimated ‘average’ sentence for respective crimes. Such a determination may be difficult at the early stages of the proceedings (i.e. at pre-trial), where investigations continue and the scope of the ‘case’ and charges is still undefined.

Heller’s test would require the ICC to carry out in-depth analysis of domestic law in order to determine an estimated ‘average’ sentence. This would pose significant problems. Sentencing criteria typically contain a large number of individualized discretionary elements, e.g., related to the discrepancy between minimum and maximum penalty. To properly understand the specificities of a domestic system, ICC Judges or the Prosecutor might have engage deeply with applicable domestic law and jurisprudence. It is questionable whether the ICC would be properly equipped to carry out such assessment. Any calculation might entail a great degree of uncertainty and risk of misinterpretation. Further complications might arise if the ICC might have to address factors such as cumulative charging, plea bargaining or applicable mitigating circumstances under domestic law. If a party (e.g. the defendant) would appeal the initial admissibility decision under Article 82 (1) (a), the ICC might be involved more with the correct interpretation of national law, rather than its core business: to investigate and prosecute cases.

(ii) The ‘lower’ sentence dilemma

Finally, there is an inherent flaw in the design of the model. In some cases, it may be desirable to give preference to proceedings at the domestic level, although such proceedings might actually lead to the imposition of a lower sentence. Such cases might in particular arise in contexts where mitigated or alternative sentences are part of a ‘transitional justice’ strategy. Heller’s model might have to be adjusted in order to take into account such specificities.

4. Conclusion

As it stands, Heller’s approach is thus still more a creative ‘thought’ experiment than a fully ‘manageable’ model. A case-by-case assessment, which makes best use of the existing flexibility under the Statute and takes into account ‘sentencing’ criteria as part of the admissibility criteria under Article 17, might in the end present a more nuanced and suitable approach.