1 Ndindiliyimana et al., Judgement, ICTR-00-56-T, 17 May 2011, paras. 18–19, 22–23, 2242.

2 Ibid., paras. 2183–2242.

3 See ibid., para. 2261 (using the phrase ‘good deeds’ in describing the mitigating acts of defendant Sagahutu).

4 Although I focus only on sentencing in this article, evidence of defendants’ good deeds often appears in the guilt stage of trials as well. For example, defendants can use their good deeds to try to disprove mens rea or actus reus, and the prosecution can sometimes rely on defendants’ good deeds in proving a superior–subordinate relationship or knowledge of criminal activity. For example, Krstić, Judgement, Case No. IT-98-33-A, 19 April 2004, para. 132 (finding that efforts of Krstić to ensure safety of Bosnian Muslim civilians being transported out of the Srebrenica region cast doubt on whether he had genocidal intent and therefore reducing the mode of liability to aiding and abetting); Mpambara, Judgement, Case No. ICTR-01-65-T, 11 September 2006, para. 70 (discrediting a witness's testimony of Mpambara's orchestration of killings in light of testimony by other witnesses that Mpambara was trying to stop the killings); Čelebići, Judgement, Case No. IT-96-21-A, 20 February 2001, para. 213 (using evidence that Mucić had camp guards treat prisoners better as evidence of a superior–subordinate relationship); Kunarac et al., Judgement, Case No. T-96-23 & 23/1, 22 February 2001, para. 590 (relying on evidence that Vuković had assisted Bosnian Muslims in danger in finding that he knew of a widespread or systematic attack directed against Bosnian Muslims).

5 For purposes of this article, I need not address the lower bounds of what constitutes a ‘humanitarian step’. Suffice it to say that such a step should be genuine and at least marginally helpful to those on the other side. Thus, a commander's promise that a victim will no longer be mistreated cannot be a ‘good deed’ if it is found to be a false promise – as when the witness is then beaten daily for the next 12 days. See Naletilić and Martinović, Judgement, Case No. IT-98-34-T, 31 March 2003, para. 427. I will, however, treat humanitarian steps affirmatively taken in the course of otherwise illegal conduct as good deeds. Thus, I treat an order that no harm come to civilians as a good deed, even if given in the context of a forcible transfer operation amounting to the crime of persecution. See Krstić, Judgement, Case No. IT-98-33-T, 2 August 2001, paras. 358–359.

6 Indeed, the ICTY prosecution has made this argument. See Appeal Brief, Blagojević and Jokić, ICTY Prosecution, 9 May 2005, paras. 6.46–6.49 (arguing that Jokić's role in guiding Bosnian Muslim boys safely through a minefield should not have counted as a mitigating factor in sentencing, since ‘under international law, Jokić had a special duty to protect children from the negative effects of the conflict’); but see Blagojević and Jokić, Judgement, Case No. IT-02-60-A, 9 May 2007, para. 342 (rejecting this argument).

7 Judgement, 1 Trial of the Major War Criminals before the International Military Tribunal, Nuremberg, 14 November 1945–1 October 1946 (1947) (hereafter, ‘IMT Judgement’), at 314–15 (Dönitz), 333 (Speer), and 336 (von Neurath).

8 At least two of the US Nuremberg Military Tribunal judgements also took the good deeds of defendants into consideration in mitigation. See Schabas, W., ‘Sentencing by International Tribunals: A Human Rights Approach’, (1997) 7 Duke JCIL 461Google Scholar, at 492–3 (describing mitigation with regard to Waldemar von Radetzky in the Einsatzgruppen Trial for his having helped certain victims to escape, with regard to Ernst Dehner in the Hostage case for his conscientious efforts to apply the laws of war, and with regard to several other defendants); see also Beresford, S., ‘Unshackling the Paper Tiger: The Sentencing Practices of the Ad Hoc International Criminal Tribunals for the Former Yugoslavia and Rwanda’, (2001) 1 International Criminal Law Review 33CrossRefGoogle Scholar, at 74–5 (discussing the same cases).

9 IMT Judgement, supra note 7, at 330; see also Beresford, supra note 8, at footnote 186 (noting the differential treatment between Speer and von Neurath on the one hand and Seyss-Inquart on the other).

10 Holá, B. et al. , ‘International Sentencing Facts and Figures’, (2011) 9 JICJ 411Google Scholar, at 432.

11 The ICTR statute and its Rules of Procedure and Evidence give very little guidance on sentencing generally (let alone on the narrower issue of how to treat good deeds in mitigation). See Art. 23(1)–(2) ICTRSt; Rule 101(B) ICTR RPE; see also Art. 24(1)–(2) ICTYSt; Rule 101(B) ICTY RPE (containing provisions similar to those in the ICTR). The lack of precision in the foundational documents and the resulting broad discretion available to Trial Chambers have led to a wide range of sentencing practices in the ad hoc tribunals. This has been discussed in general terms by a number of commentators. See, generally, e.g., Beresford, supra note 8; Harmon, M. B. and Gaynor, F., ‘Ordinary Sentences for Extraordinary Crimes’, (2007) 5 JICJ 683Google Scholar; Henham, R., ‘The Philosophical Foundations of International Sentencing’, (2003) 1 JICJ 64Google Scholar; Nemitz, J. C., ‘The Law of Sentencing in International Criminal Law: The Purposes of Sentencing and the Applicable Method for the Determination of the Sentence’, (2001) 4 YIHL 87CrossRefGoogle Scholar, at 120.

12 Munyakazi, Judgement, Case No. ICTR-97-36A-T, 5 July 2010, para. 520. For another example, see Muvunyi, First Judgement, Case No. ICTR-2000-55A-T, 12 September 2006, para. 540 (a judgement reversed and remanded for other reasons by Muvunyi, Judgement, Case No. ICTR-00–55-A-A, 29 August 2008).

13 Muvunyi, Second Judgement, ICTR Trial Chamber, 11 February 2010, paras. 147, 150–151. For other examples of this approach, see Bikindi, Judgement, Case No. ICTR-01-72-T, 2 December 2008, para. 457; Zigiranyirazo, Judgement, Case No. ICTR-01-73-T, 18 December 2008, para. 465; cf. Nahimana et al., Judgement, Case No. ICTR-99-52-T, 3 December 2003, paras. 850, 1101 (noting incidents in which Ngeze assisted individual Tutsis without making clear whether it was counting these incidents in mitigation).

14 See Holá et al., supra note 10, at 434 (reading similar language in Bikindi this way and briefly noting the conflicting trends in the jurisprudence).

15 See Muvunyi, Second Judgement, ICTR Trial Chamber, 11 February 2010, para. 130 (describing his good deeds as ‘limited and selective, or offered to Tutsis who were close to either his friends or family’).

16 Simba, Judgement, Case No. ICTR-01-76-T, 13 December 2005, para. 442; Serugendo, Judgement, ICTR-2005-84-T, 12 June 2006, paras. 68–69; Rutaganira, Judgement, ICTR Trial Chamber, 14 March 2005, paras. 153–155.

17 Ntawukulilyayo, Judgement, Case No. ICTR-05-82-T, 3 August 2010, para. 475. For additional examples, see Ntakirutimana, Judgement, Case No. ICTR-96-10 & ICTR-96-17-T, 21 February 2003, para. 909; Sentencing Nzabirinda, Judgement, Case No. ICTR 2001-77-T, 23 February 2007, paras. 75–77; Ruggiu, Judgement, Case No. ICTR-97-32-1, 1 June 2000, paras. 73–74; Serushago, Judgement, Case No. ICTR-98-39-S, 5 February 1999, para. 38.

18 Simba, supra note 16, para. 442; Rukundo, Judgement, Case No. ICTR-2001-70-T, 27 February 2009, para. 602. For another example, see Bagosora et al., Judgement, Case No. ICTR-98-41-T, 18 December 2008, para. 2273.

19 Niyitegeka, Judgement, Case No. ICTR-96-14-T, 16 May 2003, paras. 494–495; Kajelijeli, Judgement, Case No. ICTR-98-44A-T, 1 December 2003, paras. 949, 951. For another example, see Nchamihigo, Judgement, Case No. ICTR-01-63-T, 24 September 2008, paras. 393–394.

20 E.g., Munyakazi, Judgement, ICTR Appeals Chamber, 28 September 2011, paras. 172–175; Bikindi, Judgement, Case No. ICTR-01-72-A, 18 March 2010, paras. 162–163; Niyitegeka, Judgement, Case No. ICTR-96-14-A, 9 July 2004, para. 266.

21 The only empirical study to try to answer this question focuses on the ICTY and finds that ‘[g]iven all other factors, a sentence is on average reduced by 0.6 years (7 months) for each mitigating factor’. Holá, B. et al. , ‘Is ICTY Sentencing Predictable? An Empirical Analysis of ICTY Sentencing Practice’, (2009) 22 LJIL 79CrossRefGoogle Scholar, at 94.

22 See Judgement, Ndindiliyimana et al., supra note 1, paras. 2188–2242, 2266 (noting in a very long discussion of mitigation that the mitigating factors – many of which involved good deeds — carried ‘considerable weight’ and amounted to ‘circumstances [that] are unique and distinguish Ndindiliyimana from . . . other Accused who have come before this Tribunal’).

23 Muvunyi, Second Judgement, ICTR Trial Chamber, 11 February 2010, paras. 94–98.

24 Serugendo, supra note 16, paras. 24–27.

25 Muvunyi, Second Judgement, ICTR Trial Chamber, 11 February 2010, paras. 130, 147.

26 Holá, supra note 10, at 433.

27 Kunarac et al., Judgement, Case No. IT-96-23-T & IT-96-23/1-T, 22 February 2001, para. 434.

28 Kunarac et al., Judgement, Case No. IT-96-23 & IT-96-23/1-A, 12 June 2002, para. 408.

29 An interesting comparison to this case is Perišić, where the Trial Chamber found that the defendant had in fact done good deeds towards his victims – certain captured Bosnian Muslim soldiers – but declined to give any weight to this factor on the grounds that he ‘had himself helped precipitate this situation’. Perišić, Judgement, Case No. IT-04-81-T, 6 September 2011, para. 1831.

30 Krajišnik, Judgement, Case No. IT-00-39-T, 27 September 2006, para. 1162 (footnotes omitted). Applying this framework to the facts, the Trial Chamber then recognized that the defendant had ‘made some efforts during the indictment period to provide help to non-Serb individuals’ but deemed this assistance ‘sporadic’; ibid., para. 1163. On appeal, the Appeals Chamber similarly remarked that the defendant's ‘sporadic assistance . . . can only have a limited impact on the sentence’; Krajišnik, Judgement, Case No. IT-00-39-A, 17 March 2009, para. 817.

31 E.g., in Kupreskić et al., the Trial Chamber found as a mitigating factor that defendant Josipović assisted two Bosnian Muslims in escaping. From the Judgement, it is clear that Josipović aided one of these individuals at around the same time as he committed his crimes, but it is not spelled out when he aided the other individual. Kupreskić et al., Judgement, Case No. IT-95-16-T, 14 January 2000, para. 860. For other examples in which the good deeds noted in mitigation seem to be roughly contemporaneous with the crimes but where no contemporaneity requirement is expressly noted, see Sentencing Nikolić, Judgement, ICTY Trial Chamber, 18 December 2003, para. 266; Br đanin, Judgement, Case No. IT-99-36-T, 1 September 2004, para. 1121; Čelebići, Judgement, Case No. IT-96-21-T, 16 November 1998, paras. 1248, 1270; Čelebići, supra note 4, para. 776; Aleksovski, Judgement, ICTY Trial Chamber, 25 June 1999, para. 238; Krstić, supra note 4, paras. 272–273; Kvočka et al., Judgement, Case No. IT-98-30/1-T, 20 November 2001, paras. 715, 730, 739.

32 Simić et al., Judgement, Case No. IT-95-9-T, 17 October 2003, para. 1096. For other examples, see Lukić & Lukić, Judgement, ICTY Trial Chamber, 20 July 2009, para. 1092; Delić, Judgement, ICTY Trial Chamber, September 15 2008, paras. 581–585.

33 Also, as in the ICTR, once an ICTY Trial Chamber determines a good deed to be mitigating, there is significant variation in the extent to which the Trial Chamber spells out how much or how little weight it attaches to this factor. It is thus hard to assess how consistent the Trial Chambers are in how they weigh the impact of these mitigating factors.

34 Blaškić, Judgement, Case No. IT-95-14-T, 3 March 2000, para. 781.

35 For other examples in which good deeds are deemed to shed positive light on a defendant's character, see Hadžihasanović and Kubara, Judgement, ICTY Trial Chamber, 15 March 2006, para. 2080 (considering Hadžihasanović to have ‘a character which can be rehabilitated’ based in parts on his efforts to train his troops to comply with international humanitarian law); Blagojević and Jokić, Judgement, Case No. IT-02-60-T, 17 January 2005, para. 854 (finding that Jokić's aid to Bosnian Muslim boys in a minefield established ‘both [his] character and the fact that [he] did not discriminate against Bosnian Muslims’); see also Gotovina et al., Judgement, Case No. IT-06-90-T, 15 April 2011, para. 2610 (finding that evidence that defendant Markač has assisted several victims was ‘not sufficient for the Trial Chamber to assess whether [he] had a good character which it could consider as a mitigating factor’).

36 Popović et al., Judgement, Case No. IT-05-88-T, 10 June 2010, para. 2210.

37 Ibid.x, para. 2219.

38 Ibid.

39 Ibid., paras. 2219–2222. In the next paragraph, the Trial Chamber then observes that evidence of certain other good deeds by Pandurević during the course of the war constituted ‘other evidence of Pandurević^’s good character’ but stated further, without explanation, that it would give ‘only limited weight to [this evidence] as a mitigating factor’; ibid., para. 2223.

40 Some systems create no explicit space for taking beneficial acts into account; others provide explicit space for them only where they are aimed at the specific victims; and still other systems will take into account beneficial acts unconnected to the victims. See United States Sentencing Commission, Guidelines Manual, § 5H1.11 (November 2006) (‘prior good works are not ordinarily relevant in determining whether a departure [from the guidelines] is warranted’); German Criminal Code, §§ 46(2) and 46a (stating that steps taken to make restitution towards the victim can be deemed mitigating); D. Thomas QC, Current Sentencing Practice (2007), § C2-2H (noting that, in the United Kingdom, even worthy acts unconnected to the particular victims may serve as mitigating circumstances).

41 Nemitz, supra note 11, at 120.

42 There is debate about what all the proper sentencing purposes of international criminal tribunals are, but deterrence and retribution are broadly accepted as proper purposes – although concerns about the efficacy of deterrence may make it a less important purpose than retribution. See, e.g., Harmon and Gaynor, supra note 11, at 691–6.

43 See, e.g., Sloane, R. D., ‘The Expressive Capacity of International Punishment: The Limits of the National Law Analogy and the Potential of International Criminal Law’, (2007) 43 Stanford JIL 39Google Scholar, at 41 (noting ‘the collective nature of the victim of international crimes’); Henham, R., ‘Developing Contextualized Rationales for Sentencing in International Criminal Trials’, (2007) 5 JICJ 757Google Scholar (noting the need to consider collective claims of justice in the course of international criminal-law sentencing).

44 This argument may also hold true with regard to society at large. Indeed, I think that is why we see different approaches in domestic jurisdictions to whether a defendant's beneficial acts to those other than the victim can count in his favour in mitigation. See supra note 40. Some jurisdictions choose to factor in these social benefits, while others focus more strictly on retribution for the specific victim. In the international criminal context, however, we can bypass this domestic debate entirely by relying on the group-based component to retribution that is present in the international context but not in the domestic one.

45 A related question is whether post-conflict assistance to the other side should be mitigating. E.g., Plavšić, Sentencing Judgement, Case No. IT-00-39&40/1-S, 27 February 2003, paras. 91–94; Judgement on Sentencing Appeal, Milan Babić, ICTY Appeals Chamber, 18 July 2005, paras. 53–62; cf. Judgement on Sentencing Appeal, Miodrag Jokić, ICTY Appeals Chamber, 30 August 2005, para. 53. This issue raises distinct questions of its own, however, and is beyond the scope of this article.

46 This is particularly true as international criminal tribunals move away from the ICTY's early approach of trying any suspects it could get custody over, whether low-level or high-level, and more towards the approach taken by the ICTR, the SCSL, and the ICC of focusing on high-level perpetrators believed responsible for numerous atrocities.

47 A debate exists about whether and to what extent rehabilitation – which often turns upon a defendant's character – should be a relevant sentencing purpose. See, e.g., Harmon and Gaynor, supra note 11, at 693 (suggesting that rehabilitation should be at most a lesser sentencing purpose). For purposes of this article, I assume that it is at least a modest sentencing purpose.