Punishment and policy in international criminal sentencing: an empirical study.

Author:Doherty, Joseph W.

The International Criminal Tribunal for the Former Yugoslavia (ICTY) and the International Criminal Tribunal for Rwanda (ICTR) have sentenced over 130 perpetrators for genocide, crimes against humanity, grave breaches of the Geneva Conventions, or war crimes. Sentencing judgments discuss the factors considered by the judges and impose a term of imprisonment. We regressed the sentence length meted out for each perpetrator on the doctrinal factors said to explain the term of imprisonment and on other factors rumored to affect sentencing. We find that the gravity of the crime and aggravating factors declared by the Tribunals as sentencing factors are significantly related to sentence length but that the mitigating factors proclaimed by the Tribunals--all but one of which are associated with diplomatic and policy objectives--are not significantly related to the term of imprisonment. We conclude that international criminal sentences prioritize punishment of the individual based on the seriousness of the crime over the other diplomatic and policy goals that the judges claim to be pursuing. We conjecture that this discrepancy is based on functional differences: the sentencing judgment discussion seeks to advance the many policy objectives of the Tribunals, while the declared term of imprisonment is largely an expressive act of retributive justice, which might also facilitate deterrence and reconciliation.


    In the early 1990s, as genocide and other mass atrocities unfolded in Yugoslavia and Rwanda, the ICTY and the ICTR were founded in 1993 and 1994, respectively, as vehicles for retribution against perpetrators, (1) preventing future atrocities, (2) substituting and reflecting local customs, (3) establishing a historical record of the events, (4) building international law and jurisprudence on the topic, (5) expressing humanity's stake in redressing horrendous crimes, (6) and achieving justice. (7) Some commentators and most diplomats were adamant, however, that a key objective of establishing the Tribunals should also be the restoration of peace and reconciliation, (8) which some worried were incompatible with establishment of a justice mechanism. (9) Concern arose that the prospect of postconflict criminal accountability could deter leaders from agreeing to end a conflict; a justice mechanism could destabilize the peace due to the arrest and incarceration of powerful figures to whom allegiance was owed; and the process would inevitably create perceptions of bias. (10) The diplomatic result was to establish tribunals that would embrace all of the foregoing policy objectives, including those aimed at advancing traditional criminal justice objectives, such as retribution and deterrence, as well as peace and reconciliation.

    From the outset, diplomats, lawyers, judges, and commentators recognized that the Tribunals' capacity to advance these goals would turn on a few key elements apart from the capacity of the Tribunals to prosecute and convict perpetrators while respecting the rights of the accused, namely the sentencing rationales and practices that would be employed. (11) Hence, the Tribunals expressly adopted all of the objectives identified above and began building their sentencing doctrine around them. (12) Meanwhile, United Nations diplomats, political figures in the two regions, prosecutors, defense counsel, and commentators championed a broad range of considerations that they argued should affect the sentences of those convicted. The Tribunals considered these factors in their sentencing jurisprudence, expressly accepting approximately fourteen factors and rejecting others. The resulting sentencing jurisprudence is far more elaborate than that employed by the International Military Tribunal at Nuremberg, (13) and this sentencing jurisprudence has now been applied to more than 130 perpetrators who have been convicted and sentenced by the ICTY or ICTR.

    This article assesses the extent to which sentences issued by the Tribunals correspond with their sentencing doctrine: (14) Do the sentencing factors that the Tribunals proclaim to be considering in fact affect sentence length? Do the sentencing factors rejected by the Tribunals as inappropriate in fact bear upon sentence length? Do factors not discussed at all by the Tribunals affect sentences? Using empirical methods increasingly employed in international legal scholarship, (15) we find that most factors embraced in doctrine that bear upon the seriousness of the criminal act (those related to the gravity of the crime and aggravating factors) in fact affect sentence length. In contrast, the pronounced mitigating factors that would be expected to reduce sentence length, most of which are intended to advance policies other than retribution and deterrence--including those factors primarily intended to facilitate peace, reconciliation, humanitarian assistance during conflict, and judicial efficiency--bear no significant relationship (16) to the actual sentence issued. (17)

    These findings raise important questions that we address in the conclusion: Why do the Tribunals' decreed sentences not correspond more fully with sentencing jurisprudence? And why do Tribunal sentences privilege criminality factors over mitigating policy factors?

    Part II of this paper presents relevant literature and factors that we hypothesize relate to sentence length, based largely on the Tribunals' Statutes and Rules of Procedure and Evidence, sentencing doctrine, commentary, and interviews with ICTY trial attorneys, in four broad categories: gravity and aggravation; mitigation; sentencing practices in the former Yugoslavia and Rwanda; and factors rumored to affect sentence length. Part III presents our methodology and information on data. Part IV presents our findings. Part V concludes, conjecturing that sentencing doctrine performs a display function intended to advance all the Tribunals' objectives, while the sentence actually decreed is fundamentally an act of retributive justice proportionate to the seriousness of the criminal act--which might also facilitate deterrence and reconciliation.


    Some commentators have distilled factors that have been proclaimed by the Tribunals as the bases for sentencing but have not statistically tested whether those factors are in fact related significantly to sentence length. (18) Several years ago, when far fewer ICTY cases had been completed, a few other commentators tested empirically whether a range of factors might be significantly related to ICTY sentence length. (19) These sophisticated studies are pioneering, but, with the benefit of hindsight as we note below, (20) those analyses could be challenged as being undertheorized, containing methodological errors, testing an incomplete set of hypothesized factors, or considering only ICTY (but not ICTR) sentence length.

    We hypothesize and test the extent to which twenty factors may be empirically related to the length of sentences issued by the ICTY and ICTR. The Tribunals have asserted that most of those factors affect sentence length but that some should not affect sentence length. Other factors, identified in author interviews of ICTY trial attorneys from the Office of the Prosecutor and defense counsel, are also examined.

    The Tribunals' respective mandates allow them to charge the following three crimes: genocide, crimes against humanity, and particular breaches of the Geneva Conventions of 1949. While the Rwandan conflict was an internal conflict, the conflict in the former Yugoslavia has been deemed international in nature, so the ICTY also has jurisdiction over violations of the laws or customs of war. (21) Article 23 of the Statute of the ICTR, and Article 24 of the Statute of the ICTY, are nearly identical and provide primary guidance concerning penalties:

    (1.) The penalty imposed by the Trial Chamber shall be limited to imprisonment. In determining the terms of imprisonment, the Trial Chambers shall have recourse to the general practice regarding prison sentences in the courts of [the former Yugoslavia] [Rwanda].

    (2.) In imposing the sentences, the Trial Chambers should take into account such factors as the gravity of the offence and the individual circumstances of the convicted person. (22) Furthermore, the penalty provisions of the respective Rules of Procedure and Evidence (RPEs) are identical to each other, except for an internal reference, and state in relevant part that the Trial Chamber shall take into account "such factors as ... (i) any aggravating circumstances; [and] (ii) any mitigating circumstances including the substantial cooperation with the Prosecutor by the convicted person before or after conviction...." (23)

    No sentencing guidelines in the Statutes or RPEs go beyond these basic provisions. The Tribunals favor wide discretion and relatively unbounded sentencing directives, declaring that each crime is unique and deserves an individual assessment, free from strict sentencing restrictions. (24) Many commentators, on the contrary, believe that stricter guidelines should have been imposed, finding that the current system is too ambiguous, resulting in widely varying sentences and discontent in the regions where the crimes took place. (25)

    Given the absence of more detailed sentencing guidelines, the sentencing factors considered by the Tribunals must be distilled from their decisions. Largely mirroring the factors suggested in the Statutes and RPEs, the Tribunals' stated rationales for sentencing address four sets of factors. The Tribunals have opined that " [b] y far the most important consideration, which may be regarded as the litmus test for the appropriate sentence, is the gravity of the offense," (26) but "gravity" subsumes many factors. The Tribunals also recite consideration of a host of aggravating and mitigating...

To continue reading