I then turn my attention to the proper role of quantitative factors in the case selection and prioritization calculus. Though I again find an emphasis on these factors is advisable, I also judge the risk of suggesting de facto immunity for certain perpetrators more problematic. I suggest that the OTP compensate by periodically prosecuting cases which quantitative metrics would assess to be of relatively low gravity, and which are under-prosecuted considering some additional parameter as well.

In Part VI, I turn to the emphasis on persons most responsible. After offering some reasons why the policy might diminish net deterrence, I argue that, given the lower mens rea requirement for high-ranking perpetrators, directing prosecutions toward persons most responsible is, in fact, a model deterrence policy. I suggest that the OTP should still prosecute lower-ranking perpetrators when convenient, with an emphasis on crimes resulting in death or otherwise involve harming witnesses for the purpose of reducing evidence.

In Part VII, I consider how a resource surplus might change the picture. I tentatively conclude that, assuming the Chambers remain legalistic, the OTP will usually improve deterrence when it initiates an investigation—even if a political bias on the part of the OTP is sometimes perceptible.

In Part VIII, I discuss the way the complementarity requirement necessarily empowers political facts to control the ICC’s behavior, and that maximizing deterrence in light of the requirement will require that the ICC let political considerations inform its behavior. Finally, in Part IX, I offer some notes on the justice dilemma.

I. Theories of ICC Deterrence

The optimist’s view is straightforward: The ICC offers a credible threat of punishment; that threat is added as a consideration which weighs against the commission of crime; therefore, the ICC deters crime.8 Some read the statistical data and statements of leaders to suggest that the ICC’s effect has already been significant, while others are doubtful that many could fear the ICC considering its slim record of convictions and lack of a police force.9 On this theory, the strength of the ICC’s deterrent effect increases as perpetrators perceive a higher likelihood of punishment.10 However, some hold the opposite view, arguing that the threat of conviction sabotages peace negotiations and encourages perpetrators to remain committed to conflicts they might have otherwise abandoned.11

Daniel Krcmaric sees both effects reflected in the empirical data—ex ante deterrence on the one hand, and ex post “bargaining for resurrection” on the other.12 Krcmaric suggests that, since the birth of the Rome Statute and the capture of Pinochet in the United Kingdom, culpable leaders no longer find offers for safe harbor credible, predicting their prospective hosts would succumb to international pressure to surrender them.13 The result is that culpable leaders are now six times less likely to flee than their innocent counterparts, whereas culpable and innocent leaders were equally likely to flee prior to 1998.14 This has caused conflicts to last longer on average, as certain culpable leaders—who previously would have taken up exile abroad—prefer to “bargain for resurrection” rather than be punished for their crimes.15 However, the fact that leaders no longer perceive that an effective exile option will be available has also resulted in leaders’ being five times less likely to initiate atrocities in the first place.16 A justice dilemma therefore arises, as the same conditions that increase ex ante deterrence also increase ex post bargains for resurrection.17

II. Refining the Theoretical Framework

Scholarship on the use of statistical evidence in the courtroom has produced an elegant theoretical framework that synthesizes the many different angles from which scholarship has approached the issue of ICC deterrence into a single all-encompassing rule.18 The puzzle out of which this theory arose is traditionally illustrated with the following thought experiment: Plaintiff, having been struck by a bus, sues Blue Bus Company for negligence. In one scenario, the court finds Blue Bus Company liable on the sole basis that 70% of the buses operating in the area belong to Blue Bus Company, with the remainder operated by Red Bus Company. In a different scenario, the court bases the same finding on the testimony of a witness known to be 70% reliable, who identifies Blue Bus Company as the culprit. Most sense an important distinction, though vindicating it—or even identifying it—has proven a challenge. David Enoch notes the following: When the court relies on witness testimony, a finding of liability is the probable outcome whenever the Blue Bus Company is actually at fault, while a finding of no liability is the probable outcome whenever Red Bus Company is at fault. In other words, if Blue Bus Company were not at fault, the witness, being that they are 70% reliable, probably would not have testified that Blue Bus Company was at fault, and so the court probably would not have found Blue Bus Company liable. By contrast, the court that uses statistical evidence will produce the same finding no matter the defendant’s actual fault. Drawing on the work of Chris Sanchirico on character evidence,19 Enoch concludes that the relevance of this distinction lies primarily in its deterrence implications.20 If an actor knows that the court’s findings will be based on statistical evidence, they dispense with any incentive to omit crime, as their decision has no effect on their expected outcome.21 Actors for whom acquittal is guaranteed will be undeterred. Actors for whom conviction is guaranteed are at best undeterred; such actors may even be encouraged to reap what benefits they can from crime, knowing they will face punishment anyway. The rule that emerges is that deterrence requires that the courts base findings of fault on evidence that renders the behavior to be deterred biconditionally related to the likelihood of sanctions, as the perceived relationship between omission of crime and low certainty of punishment is also important.

The difference in deterrence between the ex ante and ex post stages fits neatly into this framework and can be used for illustration. Deterrence obtains at the ex ante stage because actors perceive that their decisions will affect their likely fate, with omission guaranteeing freedom and commission raising the chances of conviction. At the ex post stage, the decision to omit additional crime has the capacity to reduce the likelihood of sanctions. In fact, some perpetrators judge that omitting crime will effectuate a lower likelihood of freedom, choosing to bargain for resurrection on the slim chance of regaining power and thereby staying out of the ICC’s reach.

Accordingly, maximizing deterrence of Rome Statute violations writ large requires behavior on the part of the ICC that most strengthens the perceived biconditional relationship between commission of crime and sanctions for the most possible contexts, with consideration also given to the differing frequency with which respective contexts arise. In other words, the ICC maximizes deterrence when it commits actions that indicate both that commission of crime will precipitate an increase in the likelihood of ICC intervention, and that omission of crime will preserve a low likelihood of the same; the stronger the perceived connections, the greater the deterrence.

This rule also neatly identifies the mechanism by which politically motivated behavior can harm deterrence. When the ICC enters political facts into the calculus, it becomes importantly like the court that bases its verdicts on statistical or character evidence. As a general rule, a politically motivated ICC action sends a message that the political facts on which it is based (or perhaps political facts more generally), rather than the commission of crime, dictate the likelihood of sanctions. I call this the “background” or “general” harm to deterrence that results from politically motivated behavior. Politically motivated behavior also results in a special harm which reflects the specific criterion or criteria discounted in the political action, as well as the nature of the political considerations which overwhelmed the legal ones. Some legal criteria involve considerations that track the commission of crime rather closely—like, as I argue in this comment, the scale factor of the gravity assessment, for example. Therefore, it is important that the ICC not discount legal criteria that closely reflect the commission of crime. Equipped with this theory, I can begin my analysis of ICC deterrence.

III. Prescription for the Trial Chamber and Appeals Chamber

The permissibility of engaging in political behavior will vary according to the specific ICC organ and imposition under discussion. The conclusion is most straightforward concerning convictions and acquittals: The TC and AC should maintain a commitment to acting legalistically so as to ensure that guilt is a necessary and sufficient condition for conviction. Deterrence is harmed by politically motivated convictions and acquittals alike. If the Chambers frequently return convictions that lack a strong legal basis, actors will perceive little reward to remaining innocent, while acquittals that contradict the evidence diminish the perceived risks associated with commission of crime. Conversely, if conviction follows in every case in which the evidence is strong, actors will perceive a high cost to crime, and if acquittals follow every time the evidence is weak, actors perceive value in remaining innocent.

The severity of imposition which a conviction represents is also an important factor in this heightened duty of legalism. Prosecutorial deterrence likely overwhelmingly reflects the fear of apprehension and conviction, rather than the fear of whatever costs independently derive from investigations. The same is likely true of bargains for resurrection, as it is hard to fathom how a fight to the (probable) death could appear to offer the better expected outcome if otherwise. As is explored in Part VI, another consequence of this disparity in severity is that a legalistic TC and AC will reduce the risk of harm resulting from perception of OTP bias, as actors who believe themselves victims of political bias can still count on the facts to determine ICC decisions of real consequence.

Finally, the ICC’s credibility is likely staked particularly strongly on perceptions about the fairness of verdicts—and not only because its decisions are of most consequence. The very function of a judicial body is to apply law to facts. If even the judicial organ of the ICC does not observe a commitment to legalism, it becomes hard to regard the ICC as more than a political tool. Therefore, the TC and AC must convict the guilty and acquit the innocent.

IV. Introduction to Admissibility, Selection, and Prioritization

The analysis is more complicated for other ICC behavior. One thing to note is that the OTP and Pre-Trial Chamber (PTC) ultimately lack the power to convict, while, as suggested above, the impositions associated with mere investigation are likely meaningless to deterrence.22 Thinking first in very general terms, this would seem to suggest that the OTP should err on the side of overinvolvement. In Part VII, I consider the hypothetical situation where the ICC is experiencing a resource surplus and therefore can decide whether to add additional investigations. First, however, I will consider the current situation, where an overwhelmed OTP must choose among the many situations that call for its intervention, deciding in the process what message it sends to the world about its priorities and the factors that determine its behavior, and the scope of the ICC’s deterrent effect by extension.

I can address only a limited number of legal criteria and policies in this comment. In addition to prioritizing policies of particular importance, my selections are intended to illustrate the range of ways in which the deterrence rule I endorse in this comment can be applied in the deterrence analysis. I begin my analysis with the current role of gravity assessments in informing OTP action, the manner in which the OTP performs those gravity assessments, and the degree to which political considerations should have any influence on either.

V. Gravity and Deterrence

Gravity considerations feature prominently at multiple stages of OTP action. First, gravity is implicated as a threshold requirement for case admissibility, with Article 17(1) of the Rome Statute providing that a case is inadmissible when “not of sufficient gravity to justify further action by the Court.” Article 53 of the Rome Statute implicates the gravity requirement under Article 17(1) in the decision to initiate an investigation, as the OTP is instructed to consider whether potential cases arising out of a situation would be admissible. Assessments of relative gravity inform the OTP’s selection and prioritization of cases.23

The precise function of the threshold gravity requirement for admissibility can seem mysterious. The AC in Hassan stated that each of the four crimes within the ICC’s jurisdiction is in principle of sufficient gravity to satisfy the requirement,24 and the PTC in Comoros interpreted the design of the requirement to place a limit on the rejection of cases.25

The four gravity factors outlined in the Policy Paper—scale, nature, manner of commission, and impact26—appear to have successfully captured the colloquial notion of gravity.27 The many subfactors combine to form an expansive view of gravity that reduces the likelihood for perpetrators to regard their crimes as essentially unlike the grave atrocities they believe the ICC has in mind (admittedly, some conclude the exact opposite, finding that the factors’ vagueness risks encouraging actors to rely on their own idiosyncratic assessments of gravity)28.

The scale factor is often thought to carry the most weight,29 and it is assessed according primarily to quantitative metrics, with the number of victims a “key consideration.”30 In light of these quantitative metrics, assigning the scale factor particular weight is likely good for deterrence. A quantitative focus strengthens and universalizes the connection between crime and ICC sanctions by ensuring that each additional atrocity brings the perpetrator closer to crossing the threshold. By taking both intensity and duration into account, perpetrators are placed in a double bind, from which abandonment of the conflict is the only escape. Deference to quantitative factors has the additional virtue of presenting as apolitical: While an assessment of gravity that references the crime’s nature and manner of commission is liable to reflecting a particular cultural, political, or personal perspective, quantitative factors can generally be assessed more objectively, enhancing the ICC’s credibility.

However, some might be concerned that an emphasis on scale will embolden certain perpetrators to commit low-scale atrocities on reliance that the gravity requirement will shield them from accountability. Though the lack of clarity about the way in which gravity is assessed is often thought to interfere with the ICC’s ability to deter,31 my intuition is that the information available strikes the right balance for maximizing deterrence. With no definite quantifications available, perpetrators cannot achieve confidence in their ability to game the requirement; meanwhile, the scale factor assures perpetrators that each additional crime and day of continued conflict raises the likelihood that their crimes cross the admissibility threshold. The quantitative component renders it easy for actors to identify a range of actions which would increase the gravity of their cases, while the absence of a brightline rule encourages them to err on the side of caution. Further, the nature of the concept necessitates that any such rule would be arbitrary.

Additionally, the ICC has revealed that a side door to admissibility is available for certain small-scale cases.32 For example, in Abu Garda, an attack on U.N. peacekeepers, though isolated and resulting in relatively few direct victims, impacted the safety of millions and risked setting in motion events which would result in an immense number of atrocities, therefore exceeding the gravity threshold despite its small scale.33 High-impact cases like Abu Garda are good candidates for atypical findings of sufficient gravity, as their admissibility is still based on quantitative factors. This expansion of the sufficient gravity notion does empower consideration of political dynamics. Still, it is an advisable one. The claim that deterring such crimes will serve the goal of preventing atrocities has an incredibly straightforward logic—in fact, if certain crimes are in fact likely to trigger the large-scale commission of atrocities, it seems there can be little argument about whether preventing those crimes will serve the goal of minimizing the commission of atrocities. I suppose the only countervailing concern is that such actions could undermine the credibility of the ICC such that deterrence is harmed on balance. This seems unlikely to me. First, as suggested before, quantitative considerations are objective. Further, just as the deterrence rationale is clear, so is the more conceptual claim that such crimes are, indeed, grave. This is again owing to the quantitative considerations on which that assessment is based. Additionally, here the effect is to increase the number of guilty perpetrators who are eligible for facing accountability, not to let political considerations spare perpetrators from consequences. However, predictions which appear too speculative could indeed carry a risk of undermining the ICC’s credibility. Therefore, the claimed consequences should be proximate and fairly certain to follow—or else compensate with exceptional gravity.

Discerning the optimal deference to quantitative considerations in the case selection and prioritization context is more difficult. On the one hand, centering quantitative metrics within the assessment appears by far the best system for ensuring that the casual relationship between the commission of crimes and likelihood of ICC intervention is maximized over the most contexts. The policy preserves a disincentive to commit additional crimes for perpetrators who judge their cases to have already crossed the threshold requirement, including even those who perceive acquittal to be impossible should their case be tried, as each additional crime continues to harm the perpetrator’s position by raising the likelihood that their case will be selected or prioritized. On the other hand, too strong a deference to quantitative factors could risk suggesting that perpetrators can safely hide behind the graver crimes of others, and so the risk that perpetrators might attempt to game the policy is more appreciable. Lack of clarity will not help matters here, as disparities between crimes can be obvious in many cases. For example, former Prosecutor Luis Moreno Ocampo stated that the atrocities of the Lord’s Resistance Army were “much” graver than those of the Ugandan government.34 Again, noting this policy, perpetrators could perceive that a certain allowance for atrocities accrues to them whenever obviously larger-scale or higher-impact atrocities are identifiable elsewhere, even if occurring in the very same situation.

The ICC could address this risk by periodically selecting or prioritizing cases which quantitative assessments would deem of low relative gravity, thereby setting a precedent that prosecuting such cases is within the realm of possibility. If this discounting of quantitative metrics is infrequent, there should be little change in the perceived predictability for escalation to increase the likelihood of prosecution. At the same time, the ICC sets a precedent that could dispel assumptions of de facto immunity. Even if potential perpetrators understand that such cases are anomalous, the transition from a negligible to non-negligible perception of risk is a valuable increase.

The gravity assessment also considers qualitative factors, and the Policy Paper provides for a few additional considerations to inform the OTP’s case selection and prioritization. In my view, the most important of these is the emphasis on charging crimes which have historically been under-prosecuted.35 My reasoning here is similar to the expressivists’, though motivated by a different goal. Expressivists emphasize the conceptual connection between expressions of condemnation and justice, and some suggest that the OTP should prioritize under-prosecuted cases for educational and norm-bolstering purposes.36 While deterrence is not the primary rationale for the expressivist, some do identify enhanced deterrence as a collateral benefit.37 Ideally, this is where the OTP would spend its small allowance for low-gravity cases (as characterized according to quantitative factors), as this would empower the OTP to set multiple precedents in one action, while the infrequency with which such cases would be pursued would ensure that quantitative factors continue to be perceived as most important. Further, deterrence considerations are explicitly mentioned as a factor in the Policy Paper,38 so the OTP will not appear unprincipled. In any event, I find it doubtful that such an infrequent deviation from stated OTP policy could significantly undermine the ICC’s credibility, especially given that the cases at issue would still be characterized by the degree of gravity necessary for admissibility.

VI. Pursuing Those Most Responsible

Next, I turn to the OTP’s policy of prioritizing those most responsible for the crimes. Though this policy is based on a deterrence rationale,39 the risk that many low-ranking perpetrators will expect de facto immunity seems glaring. In fact, the AC noted this concern in clarifying that this policy is not a legal requirement, questioning how a rule which would limit prosecutions to a small subset of perpetrators to the exclusion of low-ranking perpetrators per se could enhance the ICC’s deterrence capacity.40

However, concerns could persist even in light of this softer policy. For example, thinking in terms of this comment’s preferred theoretical framework, it is not immediately obvious how this policy strengthens the perceived causal connection between the commission of atrocities and adverse consequences for the perpetrator. The policy appears instead to shift the power from the decision to commit crime to ascension in rank, therefore making ascension in rank the most likely object of its deterrence. While there might be some benefit associated with a chilling of ambition amongst the lower ranks, directing prosecution toward those most responsible could cause high-ranking positions to more strongly select for risk tolerance, placing the most undeterrable perpetrators in the positions of most influence.

One could also argue that, because low-ranking perpetrators can be apprehended at a lower average cost to the ICC, a lesser emphasis on high-ranking perpetrators would enable the ICC to quickly establish the sizable conviction record it needs to establish in order to be taken seriously. One might even suggest that directing more attention to low-ranking perpetrators offers a way out of the justice dilemma, as the risk of sabotaging peace negotiations is diminished in the case of the low-ranking perpetrator considering their decisions have a limited effect on the trajectory of the greater conflict in which they are involved.

Ultimately, there are compelling reasons for thinking that this policy will reduce the commission of atrocities on aggregate. Most obvious among them is that high-ranking perpetrators’ decisions have a heightened impact. The policy’s deterrence capacity is also empowered by its surrounding legal framework. The key is the lower mens rea requirement for high-ranking perpetrators, which is met if the defendant knew or should have known about the atrocities committed by their subordinates41 (convicting a low-level perpetrator instead requires a finding of intent).42 While linking any particular low-level perpetrator to any particular result may often present an evidentiary challenge, the collective results of atrocities can be imputed to high-ranking perpetrators. Consequently, high-ranking perpetrators have good reason to expect being held accountable if they do not take preventative measures. Because prevention efforts need not be successful, high-ranking actors will not regard conviction as unavoidable. In this way, the emphasis on persons most responsible, operating in concert with the relevant legal framework, places the high-ranking actor’s fate squarely in their own hands, and is therefore a model deterrence policy.

However, my intuition is that, here too, some flexibility would be most efficient for deterrence. At the very least the OTP should prosecute low-level perpetrators in cases of exceptional convenience or efficiency. Additionally, the OTP could turn its attention to low-ranking perpetrators in situations where it is resolved not to pursue a culpable high-ranking perpetrator for fear of provoking a bargain for resurrection.

As a final note, though the ICC has not assembled a hierarchy for the crimes falling under its jurisdiction,43 perhaps it would be advisable to focus prosecutions of low-level perpetrators on crimes resulting in death or that otherwise involve victimization intended to minimize available evidence. While bargains for resurrection might less frequently appear the better option where low-ranking perpetrators are concerned, such perpetrators might also be more incentivized to harm witnesses should the threat of prosecution become more credible. If all crimes are otherwise equally important, this consideration could serve to identify these crimes as the best candidates.

VII. Direct Consequences of OTP Bias

Given the ICC’s limited resources, the OTP communicates its priorities through its choices of situations and cases. Because deterrence maximization depends on actors’ believing that each additional crime raises the odds of apprehension, the OTP will often be required to defer to quantitative considerations in order to foster and maintain that perception.

However, it might be worth exploring what allowance for politically motivated OTP behavior might be tolerable if this particular concern did not apply. For instance, we could imagine that ICC supply has come to exceed demand—maybe the ICC is already engaging with every admissible case and yet still has resources to spare.44 Assuming confidence in the integrity of the TC and AC is widespread, my sense is that the harm that would result from perceptions of OTP bias would be diminished such that even OTP actions perceived to amount to politically biased audits, or “fishing expeditions,” would in most cases enhance the ICC’s deterrence effect.

I have supposed that the hardships that independently derive from investigations are too unsubstantial to trigger bargains for resurrection. With the Chambers remaining legalistic, innocent players whom the OTP is thought to be, or perhaps actually is, targeting for political reasons can continue to expect omission of crime to guarantee freedom. Meanwhile, such players will judge the heightened scrutiny to increase the risk associated with the initiation of atrocities. Empirical data on the ICC suggest that the initiation of investigations into cases results in a large spike in deterrence for rebel groups, greatly exceeding that which follows the initiation of an investigation into the situation.45 For example, following the initiation of the OTP’s investigation into the Lord’s Resistance Army, attacks on civilians decreased and many rank-and-file members of the Lord’s Resistance Army defected.46 Similarly, empirical data on domestic crime suggest that a visible police presence deters street crime.47 While the traditional account connects the increased deterrence to perceptions about certainty of punishment, some suggest that it is also partially mediated by a decrease in the perceived rewards of crime, which is mediated partially by an increase in the perceived celerity of apprehension.48 Of course, the situation is quite different in the OTP context, as investigators cannot apprehend perpetrators on the spot. On the other hand, perceptions of increased celerity may implicate considerations in the international crime context not applicable in the domestic context. For example, the initiation of an investigation might signal to perpetrators involved in the early stages of a struggle for power that they will be unable to prevail in time.

Of course, there are certainly risks associated with perceptions of OTP bias. Disproportionate OTP attention on Africa preceded accusations of anti-African bias and threats by African States to withdraw from the Rome Statute.49 An ICC that becomes loose with the complementarity requirement could greatly strain its relationship with the international community. Initiating investigations without sufficient evidence could cause involved players to become emboldened at the conclusion of the investigation, satisfied to have had their names cleared for life. The conclusions in this section are intended to be mostly illustrative. I offer them to suggest that perceptions of fairness on the part of the TC and AC would significantly reduce the harm associated with perceptions of OTP bias; with the initiation of investigations offering significant deterrence boosts, this reduction in harm could make it such that even biased investigations would frequently improve the ICC’s deterrent effect on balance. In any event, the principles in this analysis can easily be transferred to less extreme situations.

VIII. Complementarity and Deterrence

Complementarity gives the ICC no option but for political facts to control its behavior. The complementarity requirement limits admissibility to those cases which territorial judicial systems are unwilling or unable to prosecute themselves, with the sufficiency of the efforts assessed according to their genuineness.50 While this limit on ICC action may appear to hamper the ICC’s deterrence power, the complementarity requirement generates some indirect benefits to deterrence. For instance, it helps ensure that the ICC’s limited resources are directed at precisely those perpetrators who would otherwise enjoy impunity.

Additionally, the political dynamics which result from the complementarity requirement may end up maximizing the prosecution of atrocities. The requirement may incentivize states to prosecute perpetrators themselves, as states tend to regard the intervention of international courts as an imposition on their sovereignty.51 In this way, the requirement pays a passive income to the ICC that expands the range within which prosecution is threatened.

The ICC can also take a more active role. There is some evidence that preliminary investigations intensify the pressure on states, with some identifying ICC pressure as the catalyst for domestic prosecutions in the DRC, for example.52 The ICC might therefore be tempted to try a more coercive strategy: initiating preliminary investigations with the purpose of prompting action from territorial courts. Executed skillfully, this strategy offers a cheap method for expanding the ICC’s reach. At the same time, unskillful execution could result in significant harm. For example, the ICC will need to proceed as normal in the case that the preliminary investigation fails to produce the desired result; otherwise, preliminary investigations lose the capacity to credibly threaten ICC intervention—to both territorial courts and perpetrators. The ICC could find itself overextended in ways unplanned, impairing its ability to function effectively and forcing it to pass on cases that more strongly call for its involvement.

Also, while the ability to leverage territorial courts has its benefits, there may be situations in which the ICC should hope for continued inaction on the part of the territorial court. For instance, domestic prosecutions might in some cases be associated with a higher risk of provoking bargains for resurrection, as territorial justice systems often threaten longer sentences, to be served in harsher prisons facilities, than threatened by the ICC. Similarly, while evidence sharing could be an effective way of empowering territorial courts,53 the ICC should consider whether a given territorial court’s use of evidence provided by the ICC would send a favorable message to potential perpetrators. To illustrate, a territorial court that applies a substantially lower evidentiary standard or less discriminating rules of evidence might harm deterrence if it convicts on the sole basis of sparse evidence it received from the ICC. Conversely, a court which applies more onerous standards and which consequently acquits a defendant despite having been provided strong evidence could contribute to perceptions of impunity. In this way, the ICC must engage with political considerations so as to ensure that actors receive the message which is best for the purposes of deterrence.

These examples illustrate just a few of the many ways in which the complementarity requirement can reward skillful political maneuvering, punish unsound political judgment, and thereby render political behavior on the part of the ICC a necessary component of deterrence maximization.

IX. Notes on the Justice Dilemma

Admittedly, the theoretical framework I apply in this comment does little to suggest a solution to the justice dilemma. While I am tempted to simply accept the tradeoff, a “net welfare” analysis is plainly in order, as Krcmaric suggests.54 Such an analysis could also indirectly shed some light on how best to improve the balance. Until then, I can offer some notes.

One suggestion sees the ICC offering perpetrators immunity in exchange for an end to the conflict.55 This strategy seems to me to be too heavy-handed. Widespread availability of this exchange would effectively eliminate the ICC’s capacity to effectuate prosecutorial deterrence—but not before incentivizing perpetrators to demonstrate their commitment to continued horrors. Even a single concession of this sort would appear to me to carry an unacceptably large risk of inspiring subsequent attempts to obtain the same offer.

On the other hand, the ICC would retain at least some capacity to effectuate social deterrence, which is already likely responsible for a significant portion of the current deterrence for which the ICC can take credit.56 If not for the concern that perpetrators would escalate their conflicts as a negotiation tactic, this policy would appear a safe way of ensuring that the ICC’s effect on deterrence is at least positive.

On the less extreme side, and though admittedly an uncreative suggestion, one might try simply reducing the maximum sentence length. Deterrence more strongly reflects perceived certainty of apprehension than perceived severity of punishment.57 Additionally, increases in punishment severity yield diminishing deterrence returns.58 The severity of sentence expected must be an important factor in the decision to bargain for resurrection, as the odds of prevailing are so slim. Considering also that the price of failure is often death, we should maintain hope that there is some reduction in maximum sentence which would significantly reduce the frequency with which bargaining for resurrection appears the better option while minimally diminishing ex ante deterrence such that the balance of deterrence (or utility) is increased and positive.

Conclusion

The deterrence question is an empirical one, and so would benefit most from more empirical data and more statistical analysis. However, my hope is that this comment was successful at least in showcasing the virtues of David Enoch’s deterrence rule. This rule does little more than combine the rules already noted and then distill the combination, so there should be little reason to fault it.

Applying this rule, I quickly determine that convictions and acquittals should never give a voice to political considerations. I refer to the deterrence rule when I suggest that an emphasis on quantitative considerations creates the strongest link between the decision to commit crime and ICC action. Though quantitative assessments are more objective, performing them properly will often involve consideration of political dynamics, such as in impact assessments.

The infrequent exception is also a theme in this comment. Deriving as aways from the deterrence rule, it also centers around the notion that some changes in risk perception are more valuable—or more harmful—than others. The OTP should consider whether it has the option of increasing a perceived risk such that it becomes non-negligible. Another important and related idea is that the OTP can efficiently improve deterrence by looking for actions that send more than one message. Finally, I suggest that a commitment to legalism from the TC and AC would significantly reduce the harm resulting from perceptions of OTP bias.

Endnotes — (click the footnote reference number, or ↩ symbol, to return to location in text).

1. 1.

   Rome Statute of the International Criminal Court, Adopted by the United Nations Diplomatic Conference of Plenipotentiaries on the Establishment of an International Criminal Court, Jul. 17, 1998, U.N. Doc. A/CONF.183/9, as amended [hereinafter Rome Statute], Preamble, available online. ↩

2. 2.

   Mark Kersten, Lubanga and the Trouble with ICC Deterrence, Just. in Conflict (Mar. 20, 2012), available online. ↩

3. 3.

   See Hyeran Jo & Beth A. Simmons, Can the International Criminal Court Deter Atrocity?, 70 Int’l Org. 443 (Mar. 8, 2016), paywall, earlier version available online, doi. ↩

4. 4.

   See Jack Goldsmith & Stephen D. Krasner, The Limits of Idealism, 132 Daedalus 47 (2003), available online. ↩

5. 5.

   Daniel Krcmaric, The Justice Dilemma: Leaders and Exile in an Era of Accountability (Sep. 15, 2020), paywall. ↩

6. 6.

   David Enoch, Levi Spectre & Talia Fisher, Statistical Evidence, Sensitivity, and the Legal Value of Knowledge, 40 Phil. & Pub. Aff. 197 (Dec. 7, 2012), paywall, archived, doi. ↩

7. 7.

   Chris William Sanchirico, Character Evidence and the Object of Trial, 101 Colum. L. Rev. 1227 (Oct. 2001), paywall, earlier version available online, doi. ↩

8. 8.

   Jo & Simmons, supra note 3. ↩

9. 9.

   Christopher W. Mullins & Dawn L. Rothe, The Ability of the International Criminal Court to Deter Violations of International Criminal Law: A Theoretical Assessment, 10 Int’l Crim. L. Rev. 771, 781 (Feb. 10, 2010), paywall, doi. ↩

10. 10.

    Id. ↩

11. 11.

    Goldsmith & Krasner, supra note 4. ↩

12. 12.

    Krcmaric, supra note 5. ↩

13. 13.

    Id. ↩

14. 14.

    Id. ↩

15. 15.

    Id. ↩

16. 16.

    Id. ↩

17. 17.

    Id. ↩

18. 18.

    Enoch, Spectre & Fisher, supra note 6. ↩

19. 19.

    Sanchirico, supra note 7. ↩

20. 20.

    Enoch, Spectre & Fisher, supra note 6. ↩

21. 21.

    Id. at 22. ↩

22. 22.

    I do not include an analysis of arrest in this comment. My conclusion would depend on whether such an imposition is, itself, severe enough to trigger bargains for resurrection or deter crime. While the OTP and PTC can render the TC and AC powerless to deter, it is also easy to imagine that the costs associated with being forced to appear at trial, especially for high-ranking actors, could be significant enough to trigger bargains for resurrection, even if eventual acquittal is expected. ↩

23. 23.

    Office of the Prosecutor, ICC, Policy Paper on Case Selection and Prioritisation ¶ 36 (Sep. 15, 2016) [hereinafter Policy Paper], available online, archived. ↩

24. 24.

    The Prosecutor v. Al Hassan Ag Abdoul Aziz Ag Mohamed Ag Mahmoud, ICC-01/12-01/18 OA, Judgment on the appeal of Mr Al Hassan against the decision of Pre-Trial Chamber I entitled “Decision relative a l’exception d’irrecevabilite pour insuffisance de gravite de l’affaire soulevee par la defense”, ¶¶ 55–56 (ICC AC, Feb. 19, 2020), available online. ↩

25. 25.

    Situation on the Registered Vessels of the Union of the Comoros, the Hellenic Republic and the Kingdom of Cambodia, ICC-01/13, Decision on the “Application for Judicial Review by the Government of the Comoros”, ¶¶ 96–97 (ICC PTC I, Sep. 16, 2020), available online. ↩

26. 26.

    Policy Paper, supra note 23. ↩

27. 27.

    Stuart Ford, The Meaning of Gravity at the International Criminal Court: A Survey of Attitudes About the Seriousness of Mass Atrocities, 24 U.C. Davis J. Int’l L. & Pol’y 209, 210 (2018), available online. ↩

28. 28.

    See Margaret M. deGuzman, The International Criminal Court’s Gravity Jurisprudence at Ten, 12 Wash. U. Global Stud. L. Rev. 475 (2013), available online. ↩

29. 29.

    Jon Kevin Heller, Situational Gravity Under the Rome Statute, in Future Directions in International Criminal Justice (Carsten Stahn & Larissa van den Herik eds., 2008), available online. ↩

30. 30.

    Letter from Luis Moreno-Ocampo, ICC Prosecutor, Regarding Situation in Iraq (Feb. 9, 2006), available online. ↩

31. 31.

    See deGuzman, supra note 28. ↩

32. 32.

    The Prosecutor v. Bahar Idriss Abu Garda, ICC-02/05-02/09, Decision on the Confirmation of Charges, ¶ 31 (ICC PTC I, Feb. 8, 2010), available online. ↩

33. 33.

    Id. ↩

34. 34.

    Luis Moreno-Ocampo, ICC Prosecutor, Statement on the Uganda Arrest Warrants (Oct. 14, 2005), available online. ↩

35. 35.

    Policy Paper, supra note 23, ¶ 46. ↩

36. 36.

    See Margaret M. deGuzman, Choosing to Prosecute: Expressive Selection at the International Criminal Court, 33 Mich. J. Int’l L. 265 (2012), available online. ↩

37. 37.

    Id. ↩

38. 38.

    Policy Paper, supra note 23. ↩

39. 39.

    Office of the Prosecutor, ICC, Paper on Some Policy Issues Before the Office of the Prosecutor 7 (Sep. 2003), available online. ↩

40. 40.

    Situation in the Democratic Republic of the Congo, ICC-01/04-169, Judgment on the Prosecutor’s appeal against the decision of Pre-Trial Chamber I entitled “Decision on the Prosecutor’s Application for Warrants of Arrest, Article 58”, ¶ 73 (ICC AC, Jul. 13, 2006), available online. ↩

41. 41.

    Rome Statute, supra note 1, at Arts. 6–8. ↩

42. 42.

    Id. Art. 28. ↩

43. 43.

    United Nations Diplomatic Conference of Plenipotentiaries on the Establishment of an International Criminal Court, U.N. Doc. A/CONF.183/13 (Vol. II), Summary records of the plenary meetings and of the Committee of the Whole (Jun. 1998), available online. ↩

44. 44.

    The specific circumstances would be highly relevant. However, for the purposes of this discussion, all that is important is that considerations about efficient messaging are disregarded. ↩

45. 45.

    Hyeran Jo & Beth A. Simmons, Running the Numbers on ICC Deterrence: When Does It Actually Work?, Open Global Rts. (Mar. 22, 2016), available online. ↩

46. 46.

    Hyeran Jo, Mitchell Radke & Beth A. Simmons, Assessing the International Criminal Court, in The Performance of International Courts and Tribunals 193 (Theresa Squatrito, Oran R. Young, Andreas Follesdal & Geir Ulfstein eds., Mar. 2018), paywall, earlier version available online, doi. ↩

47. 47.

    Daniel S. Nagin, Deterrence in the Twenty-First Century, 42 Crim. and Just. 199, 235 (Aug. 2013), paywall, archived, doi. ↩

48. 48.

    Id. at 212. ↩

49. 49.

    Erik Voeten, Populism and Backlashes Against International Courts, 18 Persp. on Pol. 407, 417 (2020), available online, doi. ↩

50. 50.

    Rome Statute, supra note 1, at Art. 17. ↩

51. 51.

    William W. Burke-White, Proactive Complementarity: The International Criminal Court and National Courts in the Rome System of Justice, 49 Harv. Int’l L.J. 53 (Dec. 2008), available online. ↩

52. 52.

    William W. Burke-White, Complementarity in Practice: The International Criminal Court as Part of a System of Multi-Level Global Governance in the Democratic Republic of Congo, 18 Leiden J. Int’l L. 557 (Oct. 2005), paywall, doi. ↩

53. 53.

    Burke-White, supra note 51. ↩

54. 54.

    Daniel Krcmaric, supra note 5. ↩

55. 55.

    Mohammed Abbas & Khaled Yacoub Oweis, Bombings Rock Damascus, Brother of Parliament Speaker Killed, Reuters, Nov. 6, 2012, available online. ↩

56. 56.

    Jo & Simmons, supra note 3, at 12. ↩

57. 57.

    Nagin, supra note 47. ↩

58. 58.

    Id. at 232. ↩