1. Introduction
What are we to make of Augustin Ndindiliyimana? The chief of staff of Rwanda's Gendarmerie nationale during much of the genocide, he commanded and left unpunished gendarmes who participated in appalling massacres. In May 2011, an International Criminal Tribunal for Rwanda (ICTR) Trial Chamber found him guilty and recognized the ‘gravity of [his] crimes’.Footnote 1 Yet, it is hard to read this Trial Chamber's assessment of Ndindiliyimana for sentencing purposes without feeling that Ndindiliyimana was in some sense a decent man – one who disapproved of the genocide and took personal risks to try to limit its magnitude. Prior to being forced out by his colleagues, Ndindiliyimana opposed the extremist positions of Theoneste Bagosora, co-operated with UNAMIR forces, maintained a Tutsi as his personal secretary, posted gendarmes to protect a hotel where Tutsis were hiding, and sheltered numerous Tutsis in his own home.Footnote 2 He was not a saint, but nor was he much of a villain.
While Ndindiliyimana presents an extreme example, many international criminal defendants do make at least occasional efforts to behave responsibly towards soldiers or civilians on the other side of the conflict. They are likely to have at least a few moments of decency towards the other side mixed in with whatever else they do – perhaps due to flashes of humanity, to international pressure, or to personal ties with particular members of the other group. Evidence of these efforts, which I will call ‘good deeds’,Footnote 3 frequently comes up at their sentencings. This raises the question of how international judges should treat these good deeds, particularly where (as is often the case) these good deeds benefited not the particular victims against whom the defendant committed crimes, but rather other individuals on the same side of the conflict as these victims. Should trial judges find such good deeds to be mitigating? And, if so, how should they evaluate the level of mitigation?
This article looks at these questions in light of the practice at international criminal tribunals to date, with a particular focus on ICTR and International Criminal Tribunal for the former Yugoslavia (ICTY) jurisprudence.Footnote 4 In section 2, I will show that evidence of good deeds comes up frequently at international criminal sentencing and indeed can be found as early as the judgment of the International Military Tribunal (IMT) at Nuremberg. But, as I also show, the tribunals have not developed a uniform or coherent theoretical framework for dealing with such evidence. Instead, judgements both within and across tribunals contain wide disparities on questions like when such mitigating evidence should matter. In section 3, I seek to develop a theoretical framework for when and how evidence of good deeds should matter for sentencing mitigation – one that rests not simply on domestic-law approaches to mitigating evidence, but rather specifically on the particular nature and goals of international criminal law. More precisely, I propose that international criminal judges should view good deeds as mitigating the retribution due to international criminal defendants (with the degree of mitigation to depend on the magnitudes of the defendants’ crimes and good deeds) and that, depending on the specifics of the good deeds, these judges should sometimes also treat these deeds as evidence of rehabilitatable character.
Throughout, I use the phrase ‘good deeds’ to refer to humanitarian steps taken by a defendant that are aimed at those on the other side of the ongoing conflict and that are conscientious relative to the culture in which the defendant is operating.Footnote 5 This use is intentionally broad, encompassing both behaviour that a defendant has a duty to undertake pursuant to international law and behaviour that is separate from what duty requires. Thus, I will consider that a commander who actively prevents his unruly troops from killing civilians has done a good deed, even though he is obligated to do so under international law. While, in theory, one could argue against assigning value to what is already obligatory,Footnote 6 in practice, compliance with international law in the midst of a culture of noncompliance is beneficial indeed. Moreover, such duty-bound good deeds will often prove more significant than free-standing good deeds, since the obligations imposed by international law tend to be associated with the exercise of power and thus have a potentially greater scope for influence.
2. Good deeds as mitigating factors in ICTR and ICTY sentencing
Defendants’ good deeds come up frequently in the sentencing portions of international criminal judgments. This section surveys the treatment of such mitigating evidence briefly with regard to the IMT and, at more length, with regard to the ICTR and ICTY. I show that, within and across tribunals, there is significant variation in the way judgments treat good deeds in mitigation.
2.1. IMT
At Nuremberg, the IMT sometimes took the good deeds of defendants into account in mitigation. With regard to Karl Dönitz, Albert Speer, and Constantin von Neurath, the IMT explicitly found certain good deeds to be mitigating – citing, in the case of Dönitz, his responsible treatment of British naval prisoners of war; in the case of Speer, his opposition to Hitler's scorched-earth policy towards the end of the war; and, in the case of von Neurath, his efforts to arrange for the release of certain Czechoslovaks and students.Footnote 7 These three defendants all received sentences for terms of years, whereas all but one of the other convicted defendants received sentences of either life imprisonment or death by hanging.Footnote 8 But, with regard to the good deeds of Arthur Seyss-Inquart, the IMT was more opaque. The judgment recognized that:
[i]t is . . . true that in certain cases Seyss-Inquart opposed the extreme measures used by . . . other agencies, as when he was largely successful in preventing the Army from carrying out a scorched earth policy, and urged the Higher SS and Police Leaders to reduce the number of hostages to be shot.Footnote 9
Yet, the IMT did not explicitly describe these actions as mitigating and, in light of the death sentence given to Seyss-Inquart, it is also clear that the IMT did not deem them to be materially mitigating.
The IMT did not explain how it distinguished between Dönitz, Speer, and von Neurath on the one hand and Seyss-Inquart on the other. It could have drawn a principled distinction by finding that Seyss-Inquart's crimes – including the deportation of Dutch Jews to Auschwitz – were so great as to make any mitigating power of his good deeds immaterial. But it did not draw this distinction, and instead left any reasoning unstated. This lack of clarity foreshadows the approaches found in today's ad hoc tribunals.
2.2. ICTR
ICTR defendants frequently assert good deeds in seeking sentencing mitigation. Indeed, a recent study has found that ‘assistance to the victims’ is cited in 42.7 per cent of ICTR cases – more than any other mitigating factor.Footnote 10 These good deeds are usually of a similar cloth: namely direct personal assistance to certain Tutsis, often relatives or friends of the defendants. With rare exceptions, ICTR Trial Chambers spend very little time discussing these good deeds or their implications for the sentence, usually devoting no more than three or four sentences in a trial judgement to these issues. This brevity is perhaps not surprising given the discretion available to judges in sentencing decisions and the many factors they may consider.Footnote 11 What is surprising is that even these brief discussions reveal multiple and sometimes inconsistent approaches to the overarching questions of when and how much good deeds should count in mitigation.
Most notably, ICTR Trial Chambers have taken different approaches to whether direct personal assistance to certain Tutsis during the genocide is a mitigating factor. Broadly speaking, Trial Chambers take one of three approaches: (i) not counting such assistance as a mitigating factor; (ii) not clearly resolving whether or not such assistance is a mitigating factor, but in any event finding it to have no weight; or (iii) counting such assistance as a mitigating factor (even if the trial court then goes on to give limited or no weight to it).
An example of the first approach occurs in the case of Yussuf Munyakazi, a farmer and Interahamwe leader convicted of leading genocidal attacks. In that case, the Trial Chamber stated that it:
acknowledges the assistance [Munyakazi] provided to a number of Tutsi friends during the genocide. However, as the Appeals Chamber of the Tribunal has explicitly affirmed, it is well within the province of a Trial Chamber to disregard such ‘selective assistance’ to Tutsis as a mitigation factor. . . . Therefore, the Chamber does not regard these as mitigating factors.Footnote 12
The Trial Chamber thus treated Munyakazi's good deeds as irrelevant for mitigation purposes. The Trial Chamber's rejection seems to turn here on the ‘selective’ nature of Munyakazi's good deeds, although it does not clarify why this selectivity makes these deeds non-mitigating. The Trial Chamber's concern could be with the extent of Munyakazi's efforts: that he did not do as much as he could have done to help Tutsis. If so, this would make it difficult for good-deed evidence ever to count in mitigation, since convicted international criminal defendants have obviously not done their utmost to prevent international crimes. From the limited language above, however, the most likely reason seems to be that the Trial Chamber's concern was with Munyakazi's motivations. In other words, the Trial Chamber did not treat Munyakazi's actions as mitigating – regardless of whether they saved Tutsi lives – because it thought Munyakazi acted not from disinterested motives, but rather from personal ties to the Tutsis at issue.
An example of the second approach occurs in the judgement on retrial in the case of Tharcisse Muvunyi, a lieutenant colonel in the Rwandan Army found guilty of incitement to genocide for giving a virulently anti-Tutsi speech at a public meeting. The Trial Chamber stated that:
The Defence presented evidence of four character witnesses whom the Chamber found credible. These witnesses testified that Muvunyi: (1) sent soldiers to protect a bishop and some Tutsi refugees in Butare prefecture [and] (2) placed several Tutsi orphans in orphanages. . . . Exercising its discretion, the Chamber does not consider that the assistance Muvunyi provided to a handful of Tutsis during the genocide warrants mitigation because it was limited and selective. . . . The Chamber therefore concludes that there are no mitigating circumstances that should be taken into account in the determination of his sentence.Footnote 13
This last sentence is open to two possible interpretations. The first is that, as in the Munyakazi trial judgement discussed above, the Trial Chamber did not view Muvunyi's actions as mitigating.Footnote 14 The second is that, regardless of whether or not these actions were mitigating, the Trial Chamber simply did not view them as rising to a level that ‘should be taken into account’. In any event, the ultimate outcome is clear: Muvunyi's assistance to certain Tutsis did not count in his favour in mitigation. The Trial Chamber reached this finding even though some of Muvunyi's good deeds seem to have been done towards Tutsis who were not his friends or relatives.Footnote 15
The third approach, which is the most common, is for Trial Chambers to describe personal assistance to Tutsis during the genocide as a mitigating factor. For example, Simba described the defendant's ‘rather selective assistance’ to Tutsi friends and family as mitigating; Serugendo referred to the defendant's saving a Tutsi from a Hutu mob as mitigating; and Rutaganda described the defendant's aid to a handful of Tutsis (some or all of whom may have been friends or family) as mitigating.Footnote 16
Within this third approach, however, there is considerable variation in the language that Trial Chambers use to describe the extent to which the defendant's good deeds are mitigating. Some judgements simply describe such assistance as mitigating without trying to quantify the degree of mitigation. Serugendo and Rutaganda do this, for example, and, more recently, Ntawukililyayo simply noted that the defendant's provision of supplies to Tutsi survivors was a factor that, in conjunction with other factors, ‘require[d] mitigation in Ntawukulilyayo's sentence’.Footnote 17 Other judgements have language aimed at quantifying the weight accorded to evidence of the defendant's personal assistance to Tutsis – which is usually quite low. As examples, Simba described the good-deed evidence as of ‘limited weight’ and, more recently, Rukundo observed that, even if proven, evidence of the defendant's assistance to various Tutsis would carry ‘limited, if any, weight’.Footnote 18 Still other judgements use language aimed at quantifying the weight accorded to the defendant's good-deed evidence relative to the gravity of his crimes. For example, Niyitegeka described the defendant's intervention to protect some Tutsis from a roving band of Hutus as of ‘limited weight’ given his ‘crimes of a heinous nature against civilians prior to and after this episode’, and Kajelijeli found that the defendant's assistance in evacuating ‘one Tutsi man and his family is insufficient to mitigate Kajelijeli's sentence, in light of the number of Tutsis whom Kajelijeli not only failed to protect, but whose deaths he actively brought about’.Footnote 19 Different Trial Chambers thus frame their mitigation analysis with different levels of specificity. It is hard to tell, however, whether these differences stem from different choices of wording, from variations in factual circumstances, or from more fundamental differences in approaches to weighing mitigating evidence.
In sum, different Trial Chambers describe their treatment of good deeds for sentencing purposes in quite different terms. The Appeals Chamber has accepted this variation among Trial Chambers by affirming the Trial Chambers’ treatment of good-deed evidence regardless of which approach was taken below.Footnote 20 In doing so, it tolerates the direct doctrinal inconsistencies as to whether good-deed evidence is a mitigating factor in the first place, as well as the softer variation among Trial Chambers in terms of whether or not they expressly assess the weight to be given to good-deed evidence in mitigation.
Lurking behind these doctrinal inconsistencies is a much harder question of practical significance – namely do these inconsistencies matter in practice? In other words, are similarly situated defendants getting treated differently because of these inconsistencies? This is a hard question to answer because of the overall lack of transparency in sentencing decisions. Trial chambers do not precisely quantify how many months or years they write off the sentence in mitigation, let alone break this down based on individual mitigating factors.Footnote 21 It may be that good-deed evidence will rarely have more than a marginal effect and often will have none at all, since the crimes of the defendants usually dwarf their good deeds to a dramatic extent. So far, the Ndindiliyimana case discussed at the beginning of this article is the only ICTR case in which the ICTR Trial Chamber seems to have viewed good-deed evidence as making an enormous difference at sentencing – a difference so enormous that it led the Trial Chamber to sentence Ndindiliyimana only to time served.Footnote 22
Yet, even if the doctrinal inconsistencies in how ICTR Trial Chambers approach good deeds for mitigation purposes do not matter much in practice, they are still worthy of concern. Doctrinal clarity has value for its own sake. In addition, even small effects arising from doctrinal inconsistencies can matter to defendants and victims. Accordingly, I will close this survey of how ICTR Trial Chambers treat good deeds in mitigation by comparing two cases in which, in my view, the different approaches taken by the Trial Chambers may be best explained by differences in the doctrinal approaches rather than in the underlying facts.
These cases are Muvunyi and Serugendo. Both cases involved the crime of incitement to genocide (and Serugendo also involved persecution as a crime against humanity). Tharcisse Muvunyi's conviction rested on a speech he gave at a public meeting in Gikore in which he called for the murder of Tutsis, including of Tutsi women married to Hutu men.Footnote 23 Joseph Serugendo pled guilty to incitement stemming from his role as an Interahamwe leader and as a board member/adviser of Radio Télévision libre des Mille Collines (RTLM). While he was not found to have personally made a speech calling for genocide, he played an ongoing and important role across many months in a radio station that he and others had ‘planned . . . to disseminate an anti-Tutsi message, intended to foment racial hatred and ultimately to destroy the Tutsi ethnic group’.Footnote 24 In terms of good deeds, Muvunyi was found to have helped save multiple Tutsi refugees and orphans (of whom at least some appear not to have been friends or family), while Serugendo was found to have saved one Tutsi.Footnote 25
Based on these facts, one might expect that good deeds would be more likely to mitigate Muvunyi's sentence than Serugendo's. Muvunyi's good deeds were more significant than Serugendo's, and his criminal acts were far less sustained and probably also less widespread in their impact. Yet exactly the opposite happened. The Trial Chamber in Muvunyi entirely disregarded his good deeds in mitigation on the grounds that they were ‘limited and selective’, while the Trial Chamber in Serugendo took his far more modest good deed into account in mitigation. We cannot tell the extent to which these different doctrinal approaches meaningfully affected the sentences ultimately given – in Serugendo's case, his guilty plea and terminal illness were doubtless the main factors behind his extraordinarily low six-year sentence – but, at the very least, they created the possibility for disparate treatment.
2.3. ICTY
As in the ICTR, defendants frequently raise good deeds in seeking mitigation in ICTY sentencing procedures. A recent study found that ‘assistance to victims’ is cited in 38 per cent of ICTY cases, second only to family circumstances.Footnote 26 As with the ICTR, the ICTY has inconsistent strands of jurisprudence with regard to how to take account of good-deed evidence. These strands overlap partially but not completely with those of the ICTR. Where the ICTR strands focus on whether defendants’ assistance to Tutsis merits mitigation when it is selectively targeted to friends or family or at least limited in scope, the ICTY strands have focused on the question of whether good deeds should count at all when they are not directed to the specific individual victims against whom the defendant is found to have committed crimes. In addition, ICTY cases have an unacknowledged tension in how they count good deeds in mitigation. Some count these good deeds directly, while others seem to count them only as evidence of good character.
A comparison of three cases – Kunarec et al., Krajišnik, and Blagoje Simić et al. – illustrates the tension in ICTY jurisprudence over whether the sentence for crimes committed against certain individuals on the other side of the conflict can be mitigated by good deeds towards different individuals on the other side of the conflict. In Kunarec et al., defendant Zoran Vuković was found guilty of twice raping a 16-year-old Bosnian Muslim girl, who was given the pseudonym FWS-50. Although recognizing aid provided by Vuković to certain other Bosnian Muslims,Footnote 27 the Trial Chamber did not discuss this fact in mitigation. On appeal, Vuković claimed that the Trial Chamber erred in not treating his aid to other Bosnian Muslims in mitigation. The Appeals Chamber rejected this argument. Specifically:
[t]he Appeals Chamber holds that the Appellant's help to other Muslims in the conflict does not change the fact that he committed serious crimes against FWS-50. If he is to be punished for his acts against FWS-50, it is to these acts that any possible mitigating factors should be linked.Footnote 28
This approach ties any mitigating effects of good deeds very closely to the particular crimes committed by the defendant – and indeed to the particular victims harmed by his crimes.Footnote 29 This restrictive approach is likely to sharply limit the role of good deeds in mitigation.
A somewhat less restrictive approach is found in Krajišnik. In what is perhaps the longest abstract discussion in ICTY jurisprudence on how good deeds should be treated for sentencing purposes, the Trial Chamber stated the following:
Good conduct contemporaneous to the crimes may serve as a mitigating factor when the convicted person had taken steps to save lives or alleviate the suffering of the victims. The Chamber may mitigate a sentence where the convicted person provided selected assistance to the victims or persons of the same ethnicity as the victims, even though his or her acts had little practical effect. The mitigating effect is less, however, where the convicted person is shown to have been in a position to take steps to control or prevent all acts of violence. In such case, sporadic benevolent acts or ineffective assistance may be disregarded.Footnote 30
This framework holds considerably broader potential for the role of good deeds in mitigation, as it recognizes that good deeds can count even if aimed not at the specific victims, but rather at ‘persons of the same ethnicity as the victims’. But it also includes one notable limit, namely that these good deeds be ‘contemporaneous to the crimes’. The Trial Chamber does not explain why this temporal element is important and thus does not justify why good deeds towards those on the other side of the conflict should count if they occur at the same time as the crimes but not at other times during the conflict.
Where Krajišnik expressly inserts a temporal element, other Trial Chambers have simply treated good deeds towards others of the victims’ ethnicity as mitigating without a contemporaneity requirement. In most of these cases, the good deeds taken into account do turn out to be roughly contemporaneous with the crimes (especially if the crimes occurred across a long period of time), although the cases do not note this as a formal requirement.Footnote 31 This is not always true, however. In Blagoje Simić et al., the Trial Chamber's brief discussion of good-deed evidence in mitigation in relation to defendant Miroslav Tadić expressly suggests that good deeds throughout the war (rather than narrowly around the times of the crimes) can be relevant. Specifically, the Trial Chamber ‘accept[ed] evidence showing that Miroslav Tadić helped some Bosnian Muslims during the war’.Footnote 32
As in the ICTR, then, we see inconsistencies within the ICTY as to the conditions under which a good deed should be found to be a mitigating factor.Footnote 33 In the ICTY, we further find a tension in the manner in which Trial Chambers use good deeds in mitigation. Sometimes, ICTY cases treat good deeds as mitigating in and of themselves and sometimes they treat these deeds as mitigating because they shed positive light on the character of the defendant. The cases discussed in the prior paragraphs take the former approach and treat the good deeds as their own mitigating factor. The Blaškić case is an example of the latter approach. There, the Trial Chamber observed that ‘[a]nother indication that the accused's character is reformable is evident in his lending assistance to some of the victims’.Footnote 34 This language suggests that the defendant's good deeds are relevant in mitigation not for their own sake per se, but rather for what they reveal about the defendant's character.Footnote 35
The interplay between good deeds as directly mitigating and good deeds as indirect evidence of character comes across in Popović et al. in relation to defendant Vinko Pandurević. A lieutenant colonel in the Bosnian Serb forces at Srebrenica, he was found guilty of certain crimes against humanity and war crimes (but not genocide) that the Trial Chamber termed ‘serious and grave’ and ‘normally associated with heavy sentences’.Footnote 36 As in the ICTR case of Ndindiliyimana, however, the Trial Chamber viewed Pandurević's good deeds to be so important that they merited ‘significant weight’.Footnote 37 It sentenced him only to 13 years. Most importantly, the Trial Chamber emphasized that Pandurević opened a corridor that allowed safe passage for many Bosnian Muslim men at Srebrenica ‘in contravention of the orders from his superiors and with knowledge that it would potentially put him in jeopardy’.Footnote 38 In words that have a startlingly approving ring given that they are directed at a man whom it had just convicted, the Trial Chamber stated that:
Pandurević's action in this regard stands out as an instance of courage and humanity in a period typified by human weakness, cruelty, and depravity. . . . [E]ven if Pandurević's motivations in opening the corridor included military considerations and protecting Serb lives, this does not detract from the fact that objectively he saved thousands of lives. The Trial Chamber is overall convinced that Pandurević's action in opening the corridor was a clear and compelling instance of assistance to potential victims. . . . [t]he Trial Chamber therefore gives significant weight to these acts [in conjunction with other ‘brave acts’] by Pandurević as mitigating factors in the determination of his sentence.Footnote 39
This language suggests two possible reasons why the good deed was mitigating – first, that it showed courage and humanity on the part of Pandurević (i.e., was indirect evidence of good character) and, second, that, regardless of how mixed his motivations may have been, it ‘objectively’ saved lives (i.e., was directly mitigating). In other words, although the Trial Chamber does not say so expressly, it seems to recognize that good deeds can be both directly mitigating and evidence of good character rather than simply one or the other.
3. Towards a more coherent sentencing jurisprudence
The prior section has shown that evidence of good deeds comes up frequently in ICTY and ICTR sentencing decisions, yet the tribunals have not developed consistent theories for whether, when, and how they should consider this evidence to be mitigating. Different domestic legal systems answer this question in different ways,Footnote 40 and the question the tribunals should consider is what approach would make the most sense in the context of international criminal law. This section seeks to articulate such an approach. I begin with Jan Nemitz's observation that ‘[t]he starting point for every consideration of a mitigating factor has to be the following question: does the granting of mitigation for certain conduct of the accused serve the sentencing purposes of the court?’Footnote 41 For international criminal tribunals, two of the most important sentencing purposes are undoubtedly deterrence and retribution.Footnote 42 It is thus appropriate to focus on whether the use of defendants’ good deeds in mitigation would advance these purposes.
As regards deterrence, there is an argument that the good deeds of defendants should be considered mitigating. To the extent that potential international criminals are sensitive to the practices of international tribunals, then, by suggesting that good deeds can lead to lighter sentences, the tribunals theoretically provide incentives for these individuals to do good deeds. I would not place weight on this factor, however. Its effect is reduced for at least two reasons. First, to the extent that potential international war criminals can expect mitigation, their incentives to act lawfully are also lessened. Given a choice between focusing on deterring bad conduct and encouraging good conduct, international criminal tribunals should focus on the former. Second, and more practically, it seems unlikely that potential international criminals are sensitive to the fine points of international sentencing practices.
In considering whether it serves the purposes of retribution to treat the good deeds of defendants as mitigating, it is necessary to first address whom retribution is for. Is retribution for the specific victim alone, as Kunarec et al. implies in rejecting the relevance of Vuković's good deeds towards other Bosnian Muslims, or is it also for a broader group, such as those on the same side of the conflict as the victim or even society at large?
In the domestic criminal context, retribution is largely for the specific victim and perhaps to some extent for society at large for having its peace disturbed. I would suggest that, in the international criminal context, however, retribution may also be due to the collective group to which the victim belonged. International crimes are not simply crimes aimed at individuals; rather, they are crimes that take place in the context of a larger struggle between groups. Doctrinally, this is shown by the elements of international crimes. War crimes require an armed conflict and genocide requires an intent to destroy a protected group. Both types of crime thus plainly involve a struggle between groups. In theory, crimes against humanity require no such group conflict, as any widespread or systematic attack directed against civilians can give rise to crimes against humanity. In practice, however, crimes against humanity prosecuted in international criminal tribunals have typically involved the same kinds of group conflict that are found in war crimes and genocide.
In my view, these group elements should inform the nature of retribution for international crimes.Footnote 43 Some retribution is on behalf of the specific victims (and perhaps some on behalf of society at large), but some retribution is also on behalf of the group. With regard to this group-based retribution, less is due to a defendant who committed crimes against some members of the group but also helped other members of the group than is due to a defendant who committed the same crimes but did not help other members of the group.Footnote 44 This is not to say that individual-level retribution should be forgotten. A defendant who has committed an international crime against one member of the other side but saved the lives of many members of the other side would justly serve time, as retribution is partially due to his specific victim. Nonetheless, it seems fair to take the good deeds of defendants towards members of the other side into account in mitigation in a way that might not be recognized in the domestic criminal context. International crimes are international crimes in part because they involve wider conflicts between groups and, accordingly, in measuring the sentence due to a defendant, it seems appropriate to consider the overall impact he has had on members of the opposing group.
Under this approach, a defendant who has done good deeds towards those on the other side of the conflict merits less retribution, from a collective perspective, than a comparable defendant without such good deeds. Accordingly, the approach taken in Kunarac et al. is less desirable than those taken in Krajišnik and Blagoje Simić et al. As between Krajišnik and Blagoje Simić et al., the latter has the better approach. There is no reason to limit this collective perspective to the time frame contemporaneous with the defendant's crimes, as Krajišnik would do, since the benefits of good deeds can accrue to the other group at any time during the conflict.Footnote 45 There is also no reason to limit the treatment of good deeds as mitigating factors to instances where the defendant's motives are strictly pure and disinterested. Whether a defendant saves a friend from the other group or a stranger from it, he has still done something beneficial to this other group, and thus the collective retribution due to him is less than with regard to a defendant who committed the same crimes but did no good deeds at all. In terms of ICTR jurisprudence, then, the approaches taken in Munyakazi and Muvunyi, which decline to treat limited or selective good deeds as mitigating factors, should be disavowed. Good deeds, even selective ones, should be considered mitigating factors, as cases like Simba and Serugendo recognize.
In terms of how much mitigation is warranted, I would suggest a rough comparison between the defendant's crimes and good deeds. A good deed would mitigate a crime of equal magnitude partly (by substantially reducing the collective component of retribution), but only partly (because, unless directed at the specific victim, it would not reduce the individual component of retribution). The rough comparison between crimes and good deeds need not be spelled out – indeed, it would be tasteless to do so – but the Trial Chamber should give an indication of how much mitigation it ultimately thinks appropriate, such as ‘no material weight’, ‘limited weight’, or ‘moderate weight’. In most cases, as existing trial judgements suggest, good deeds will likely have at most limited weight. The crimes committed by most international criminal defendants are typically so heinous that their good deeds are unlikely to merit significant mitigating effect.Footnote 46 Nonetheless, international criminal tribunals will occasionally encounter the rare defendant like Ndindiliyimana and Pandurević whose good deeds make appropriate a significant reduction in the sentence.
So far, I have suggested that good deeds should be treated as an independent mitigating factor and that the focus should be on the objective value that these good deeds brought to the other side of the conflict rather than on the defendant's subjective motivations. But, as the discussion in Popović et al. with regard to defendant Pandurević suggests, a defendant's subjective motivations may also be relevant as evidence of his character. His character may in turn inform yet another possible sentencing purpose, namely rehabilitation.Footnote 47 For good deeds to be evidence of good character, however, they must stem from motivations that we would consider signs of good character. Selective assistance to family or friends on the other side of the conflict is less likely to meet this requirement than is disinterested assistance to strangers on the other side of the conflict.
Good deeds thus have two potential ways in which they can influence sentencing mitigation. One is directly, as good deeds, and the other is indirectly, as evidence of good character. These two ways are complementary, not in conflict, although they have different focuses. As a direct mitigating factor, good-deed evidence should emphasize the magnitude of the benefit brought to the other group, while, as evidence of good character, it should emphasize the subjective motivations that led to this good deed.
4. Conclusion
This article has addressed what I have termed the ‘good deeds’ of international criminal defendants for purposes of sentencing mitigation. I have argued that international criminal tribunals lack a consistent framework for when and how they treat these good deeds as mitigating factors. Such a framework is needed because international criminal defendants raise good deeds in mitigation in a large number of cases. Their good deeds usually are minor in comparison with the atrocities for which they are convicted, meaning that they should receive little if any sentence reduction due to their good deeds. On rare occasions, however, this is not the case. Sometimes, defendants who are high up in the organizational structure of what becomes the abusive regime push back against it – not completely by resigning or avoiding the commission of any crimes, but strongly enough to merit substantial reductions in their sentences.
