 |
 |
Extraordinary Evil, Ordinary Crime: | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
2. Communication/Condemnation/Social Solidarity in Transitional Justice |
||
The international debate about the use of prosecutions in transitional justice has focused on the conditions that permit prosecuting those who commit human rights violations. Some critics suggest that international law imposes a duty to prosecute a former regime’s atrocious crimes, and contend that states have overstated claims that prosecutions are impossible.[1] These critics contend that while governments should not press prosecutions when there is “a genuine and serious threat to national life,”[2] states should assume “reasonable risks” associated with prosecution, including military discontent.[3] Other scholars, however, criticize the proponents of prosecutions for assuming that prosecutions will be possible in the wake of human rights disasters. Not only is an amnesty for human rights abusers often a precondition for securing a smooth political transition,[4] they argue, but “[m]any fledgling democracies have simply not had the power, popular support, legal tools, or conditions necessary to prosecute effectively.”[5]
While participants in this debate disagree as to when trials are possible in practice, they generally share a basic assumption: prosecuting perpetrators of injustice is the optimal method for dealing with past atrocities.[6] The assumption that prosecutions are preferable, while perhaps not always feasible, has fostered a belief that alternative approaches, such as truth commissions, are an inferior substitute for prosecution. Predictably, strong proponents of prosecution argue that “[w]hatever salutary effects it can produce, an official truthtelling process is no substitute for enforcement of criminal law through prosecutions.”[7] But even those who advocate non-prosecution alternatives generally concede the desirability of prosecution.[8] Such critics tend to justify non-prosecution alternatives by referring to the difficulty of bringing perpetrators to trial due to inadequate legal systems, the staggering number of potential defendants, or the political consequences of trials.[9]
What accounts for the widespread assumption that the best way to deal with perpetrators of serious human rights atrocities is to prosecute them?
Scholars ascribe a wide range of values to such trials. Douglass Cassel, for example, argues that the International Criminal Court[10] would contribute to justice through the “identification, exposure, condemnation and proportionate punishment of individuals who violated fundamental norms recognized internationally as crimes, and . . . reparations to their victims, by means of fair investigations and fair trials by an authorized judicial body.”[11] Stephan Landsman suggests that prosecution “makes possible the sort of retribution seen by most societies as an appropriate communal response to criminal conduct.”[12] Moreover, he argues, prosecution can educate and deter, provide a predicate for compensating victims, enhance the rule of law, and help to heal a society’s wounds.[13] Similarly, David Crocker concludes that “[e]thi-cally defensible treatment of past wrongs requires that those individuals and groups responsible for past crimes be held accountable and receive appropriate sanctions or punishment.”[14]
Underlying such justifications is the assumption that atrocious human rights violations are in fact crimes. The instruments of systemic savagery can take many forms, from industrialized gas chambers to machete-wielding mobs, from sophisticated torture chambers to steel-toed boots. But the actions of individual human rights violators, such as murder, rape, assault, and torture, are prohibited by almost every domestic criminal justice system.[15] Since the Western conceptual framework for dealing with ordinary crime revolves around prosecutors, judges, and trials, it is easy to assume that the same mechanisms should be used to deal with genocide and similar horrors.
Is ordinary crime really an appropriate analogy for massive human rights atrocities, what Kant has called “radical evil?”[16] Ordinary crime—individual conduct that violates domestic criminal law and is undertaken for non-political purposes—concerns individual criminals. Extraordinary evil—massive or systematic human rights violations prohibited by international
law[17]—involves individuals committing many of the same actions, such as the unjustified intentional taking of human life, that constitute ordinary crimes. The ordinary crime analogy presumes, as Martha Minow puts it, that “even . . . massive horrors can and should be treated as punishable criminal offenses perpetrated by identifiable individuals.”[18] Yet are genocide and ethnic cleansing really just more egregious versions of premeditated murder? Are they merely much larger conspiracies with many more victims? Or are such atrocities qualitatively different from ordinary crime because of the numbers of victims involved and because they are typically undertaken or at least countenanced by state or quasi-state actors for political reasons? Moreover, would such qualitative differences require different responses and remedies?
Assuming that the analogy of extraordinary evil to ordinary crime is an appropriate framework within which to examine transitional justice, one must ask, why do societies prosecute ordinary crimes? The short answer is that different societies have different goals for criminal justice. Even within societies there are often fundamental disagreements about the purposes of domestic criminal justice. For example, in the United States, there is a de-
bate about whether offenders should be punished in prisons or rehabilitated in half-way houses; whether or not harsh sanctions have an appreciable deterrent effect; and whether or not reconciliation programs can be attached effectively to prosecutions. Despite such disagreements, however, in the West the determination of individual culpability through prosecution is commonly regarded as necessary to redress criminal actions. The widely held faith in prosecution—whether the goal is punishment or deterrence, condemnation or rehabilitation—disguises disputes about the underlying purposes of criminal justice.
This Article, while discussing and employing the ordinary crime analogy, ultimately questions the utility of this analogy. The aim here is to examine the reasons why societies prosecute and punish “common criminals,” and to question whether prosecutions are the best method of redressing criminal actions in the context of transitional justice. Specifically, this Article will challenge the assumption that prosecutions are always the best way to pursue justice in societies in transition by arguing that the choice between prosecution and non-prosecution alternatives should depend on what one is seeking to achieve.
The best known alternative to prosecution is the truth commission. However, truth commissions are not the only alternative; other options include reports by international delegations, lustration, civil liability, reparations, and historical inquiry.[19] Additionally, each of these alternatives may take many different forms.[20] The purpose of this Article is not to champion any specific alternative, although contrasts will be drawn between prosecution and alternative mechanisms—particularly truth commissions—in order to highlight the distinctive features of prosecutions. Rather, this Article will challenge the primacy of prosecution and will argue for goal- and culture-specific responses to mass atrocity. As the term “non-prosecution alternative” suggests, prosecution[21] and forgetting are not the only options available in the context of transitional justice.
The social and political realities of a particular transitional context will affect the kind of justice that can be pursued. Before one determines whether or not prosecution is feasible, however, one must ask if it is even desirable. Prosecutions are better designed to achieve some goals than others. Non-prosecution alternatives are indeed a second-rate option when prosecution, though politically difficult, would best serve the goals of transitional justice. But one cannot presume the inferiority of non-prosecution alternatives without first articulating the desired goals of transitional justice. Some goals can be best achieved through non-prosecution alternatives, regardless of whether prosecutions are politically feasible. As Argentinian philosopher and human rights activist Carlos Nino has suggested, “the extent of the duty of a government . . . to prosecute past human rights abuses depends . . . on the theory that underlies the justification of punishment.”[22] In other words, we must decide what we want from transitional justice before we can decide if prosecutions are the best way to achieve it.
In the context of domestic crime, penologists have provided a variety of theoretical frameworks for justifying punishment and for dealing with offenders and the crimes they commit.[23] These approaches may be classified generally as: desert/retribution/vengeance, deterrence, rehabilitation, restorative justice, and communication/condemnation/social solidarity.[24] This
Article will argue that although each of these themes is relevant to transitional justice, none can blindly be transposed from the domestic context. Each of the five main Sections below corresponds to one approach, and will begin with a brief outline of the theoretical underpinnings of that approach as applied in the domestic context. Each Section then considers the implications of that approach for transitional justice, highlighting the ways in which the scale and nature of “radical evil” complicate not only the pursuit of the five possible goals of punishment, but also the very analogy of extraordinary evil to ordinary crime.
Prosecution can certainly satisfy a society’s demand for retribution in reaction to massive human rights violations. However, if in reaction to massive human rights violations a society seeks to achieve another of the possible goals of punishment, a nuanced inquiry is required to determine which mechanism would be best suited to that goal. In order to decide whether prosecution or non-prosecution alternatives are more likely to promote deterrence, restorative justice, rehabilitation, or condemnation, we need a clearer understanding of what these concepts mean in the context of transitional justice, as well as how they might apply in a country undergoing transition.
If the appropriate mechanism for confronting grave human rights violations depends upon specified goals, who has the authority to set these goals? If transitional societies themselves have the right to decide, then we must recognize that different societies will have differing goals.[25] Some societies emerging from mass trauma will demand retribution, while others will focus on compensation. Still others may concentrate on rebuilding a shattered economy or on strengthening democratic institutions. If different societies want different things, and if prosecution is a more effective tool for achieving some goals than others, we cannot presuppose that all societies in transition should choose prosecution.
If, on the other hand, the international community has the right to set goals for transitional justice, we must ask whether or not the international community even agrees about what the most important goals should be.[26] An indiscriminate duty to prosecute[27] assumes that the international community shares a fixed hierarchy of goals, agrees that these goals are best served by prosecution, and feels comfortable imposing such a vision on the society in question. Certainly there is at least superficial agreement that human rights atrocities should be prevented and condemned. But this limited consensus provides little aid in deciding how to prioritize these two goals as against others, nor does it signify complete agreement that prosecution is the best way to prevent future atrocities or communicate shared outrage. Moreover, even if the international community could make a clear choice as to its goals and the means of achieving them, it is not obvious that such an international strategy should trump the wishes of the local society.[28] Those who have not suffered cannot presume to determine for those who have what should be attempted through transitional justice.[29] Domestic criminal justice systems have sometimes been criticized for ignoring the needs and desires of victims, for “stealing conflicts” from those involved.[30] Similar problems will arise if transitional justice reflects only the priorities of the international community (or of powerful states within it), and not those of the affected country.
Of course, the international community does have both a role and an interest in transitional justice. Just as ordinary crime is not simply an offense against the individual victim but against the entire society, so extraordinary evil is not merely an assault on the particular traumatized society but on humanity as a whole. As a result, the choice of retribution or deterrence, reconciliation or condemnation cannot be left solely to either the international community or the local society. Transitional justice must reflect the needs, desires, and political realities of the victimized society, while at the same time recognizing the international community’s right and responsibil-
ity to intervene. In practice, the balance between domestic and international control over transitional justice will likely be determined in large part by the political realities of a particular transition.[31] Nonetheless, we must think more carefully about the extent to which transitional justice should reflect local rather than international choices.
This Article does not suggest that any one of the goals discussed below should be given priority, either in general or with regard to any particular society considering how best to come to grips with its past. Indeed, since the choice of goals will affect the choice of mechanisms, and since the choice of goals and effectiveness of mechanisms will depend at least in part on cultural factors, it would be far from realistic to suggest that any one means of dealing with the past will always be superior. The main thrust of this Article, then, is that greater honesty in explaining why we punish those who commit horrific abuses will clarify what transitional justice can achieve, and increase the likelihood of achieving it.
Before moving to a discussion of each of the different approaches, it is necessary to identify three assumptions implicit in this Article. First, forgetting is unacceptable. In part this is because victims of horrible atrocities are simply unable to forget. Without some form of accounting, past atrocities inevitably fuel future ones.[32] A nation’s unity, explains Chilean human rights advocate José Zalaquett, “depends on a shared identity, which in turn depends largely on a shared memory. The truth also brings a measure of healthy social catharsis and helps to prevent the past from reoccurring.”[33] Truth and accountability are essential if traumatized societies are to begin resolving their political, ethnic, racial, and religious conflicts through democratic processes, rather than through torture, rape, and genocide. The recent horrors in Rwanda and the former Yugoslavia can be attributed in part to the failure to confront earlier ones. By contrast, denazification contributed to the stability of Germany’s post-war democracy.[34] As Richard Goldstone, the former chief prosecutor for the International Criminal Tribunals for the former Yugoslavia and Rwanda, has noted, “[t]he only hope of breaking cycles of violence is by public acknowledgement of such violence and the exposure
of those responsible for it.”[35] Victims have a deep need for acknowledgment, he continues, and they “cannot get on with building the future until their calls for justice have been answered.”[36] Amnesia or blanket amnesties without any recognition of the past are simply not an option.[37]
Second, if prosecutions are undertaken, they must comport with accepted standards of due process. While this may seem obvious, in fact procedural abnormalities are common in trials undertaken in transitional contexts.[38] Jon Elster has cataloged the use in such prosecutions of exceptional measures, including illegal internments, the presumption of guilt, biased selection of judges and jurors, lack of appeals mechanisms, collective guilt, and retroactive legislation.[39] Moreover, it is not self-evident that punishment undertaken in the name of transitional justice necessarily requires trials. For example, as World War II drew to a close, Britain and Russia argued for the summary execution of their enemies: to them, the crimes were so apparent that no trial was necessary.[40] Although a number of German war criminals were eventually tried at Nuremberg, elsewhere summary executions did take place; in post-World War II France, between 20,000 and 50,000 alleged collaborators were assassinated.[41] When trials are eschewed, the criminal justice analogy is typically replaced by an alternative conceptual framework. For example, political leaders who have insisted that evidence against those who orchestrated the recent World Trade Center and Pentagon attacks need not be sufficient for an indictment in a court of law have done so within the framework of a “war on terrorism.”[42] Within this framework, terrorists may be punished without trials or due process. However, if the conceptual framework were to shift back to criminal justice, and suspected terrorists were to be brought to trial, due process safeguards should still apply, just as they have in other terrorism-related trials.[43]
While certain procedural deviations may be acceptable—a relaxed application of the concept of nulla poena sine lege, for example—basic due process rights should be respected, even when those on trial are accused of genocide.[44] There are ethical as well as practical reasons to respect defendants’ rights. By upholding standards of fairness, one can, in Vaclav Havel’s famous words, show that “we are not like them.”[45] Due process is designed in part to protect the innocent from punishment and prevent excessive punishment of the less guilty. But due process is also what gives legitimacy to trials and convictions. Much of what prosecutions can achieve, from communicating reprobation to legitimating societal demands for revenge, depends for its success on a public belief in the fairness of trials. The failure to adhere scrupulously to fair trial standards and apply laws equally may lead to the perception that trials are merely exercises in partisan politics or victor’s justice.[46]
Of course, due process requirements must also apply to non-prosecution alternatives. For example, the ongoing debate about whether truth commissions should make public the names of offenders reflects the concern that those individuals have not had the chance adequately to defend themselves.[47] Arguably, however, there is a critical distinction between due process in the non-prosecution and prosecution contexts. When applied to non-prosecution alternatives, due process is an inherently flexible concept.[48] For example, due process might require more procedural safeguards if a commission can strip offenders of their jobs than if it can merely publish reports about their conduct. In criminal prosecutions, by contrast, there is a fairly fixed idea of what due process means, reflecting in part the severe consequences of a
criminal conviction. Despite considerable variation among legal systems, “fair trials” are generally understood to require a set of basic rights, such as the right to an impartial tribunal, the right to counsel, the right to cross-examine one’s accuser, and the right to present evidence in one’s own defense.[49]
Respecting due process requirements certainly makes it harder to convict human rights violators, just as it makes it more difficult to convict common criminals. While we must be willing to pay this price, it can be dear indeed. In the case of Chilean ex-dictator Augusto Pinochet,[50] a Chilean appeals court suspended his prosecution on the grounds that he was not mentally fit to stand trial.[51] Human rights activists argued that Pinochet’s illness was exaggerated and that he “could [have] be[en] tried without a violation of due process.”[52] But given the court’s finding of incompetence, due process under Chilean law prohibited Pinochet’s prosecution.[53] Less well-known is the case of Jean-Bosco Barayagwiza, a Rwandan media leader who incited Hutus to take up arms against Tutsis and who has been called the “number
one” culprit in the Rwandan genocide.[54] In November 1999, the International Criminal Tribunal for Rwanda (“ICTR”) dismissed the case against him on the grounds that Mr. Barayagwiza’s fundamental rights had been violated by prolonged detention without trial.[55] Respecting the speedy trial rights of a perpetrator responsible for the deaths of thousands is one cost of using a prosecution-based approach premised on the rule of law.[56]
Due process requirements also affect who is prosecuted. For example, in deciding how to focus the limited resources of the Office of the Prosecutor of the International Criminal Tribunal for the Former Yugoslavia (“ICTY”), Richard Goldstone and his colleagues had hoped to target those who planned and ordered the commission of crimes, rather than those who simply undertook them. But because the Tribunal requires sufficient evidence for guilt to be proven beyond a reasonable doubt, and because there was often not enough evidence to establish who ordered the commission of the atrocities, the Office of the Prosecutor was forced to focus on lower-level cases where a longer paper trail and more witness testimony were available.[57] Thus, the evidentiary requirements of due process may make it not only generally more difficult to convict, but also particularly difficult to convict those who are most culpable.
The third assumption implicit in this Article is that prosecutions are necessarily selective. Often entire societies are implicated in atrocities.[58] As José Alvarez notes, massive human rights violations “usually involve massive complicity by large numbers of perpetrators, at all levels of domestic and international society, and not merely by a select group of government elites.”[59] Most commentators recognize that in the wake of such widespread guilt, only a small number of even the worst perpetrators will ever stand trial. Because mass atrocities are generally perpetrated by a large number of
people, “[p]rosecution of every single participant in the planning, ordering, or implementation of the atrocities in question—not to mention all those who collaborated with them—would be politically destabilizing, socially divisive, and logistically and economically untenable.”[60] Indeed, even the strongest advocates of prosecution do not believe that it is possible to prosecute every perpetrator.[61] The fact that the ICTY has only two trial chambers demonstrates that, in Goldstone’s words, it only has the capacity to try “a relatively small number of cases.”[62] As of October 2001, sixty-one accused had appeared in proceedings before the ICTY. Only twenty-three defendants had been tried, of whom two had been acquitted, and twenty-one convicted.[63] As of September 2001, there had been fifty-two ICTR detainees, of whom one had been acquitted, three had pled guilty, and five had been convicted.[64]
Prosecutions, in part because of the due process requirements involved, are also expensive.[65] Costs escalate if trials are held outside of the transitional country, as witnesses and evidence must be transported to the distant forum. Criminal justice resources—not only funds, but staff and expertise as well—may be sufficient to prosecute “only a small fraction of those responsible for gross violations of human rights.”[66] In its efforts to prosecute widely, the Rwandan government has confronted the problem of inadequate resources. While at least 1420 people have been tried, Amnesty International has reported that some trials have been unfair, that tens of thousands of genocide suspects have been detained for years under horrible conditions, and that due to the backlog the government plans to introduce a system for handling lower-level offenders which “do[es] not conform to basic international standards for fair trials.”[67] As Paul van Zyl notes, “[c]riminal justice systems . . . are designed for societies in which the violation of law is the exception and
not the rule. Once the violation of the law becomes the rule, criminal justice systems simply cannot cope.”[68] The result, he concludes, is a de facto amnesty for many.[69]
While selective prosecution is used in virtually every legal system[70] and prosecutors everywhere are faced with the need to concentrate their limited resources on the most culpable offenders, selective prosecution after genocide or massive human rights abuses presents a qualitatively different challenge. In a functioning criminal justice system, a prosecutor’s decision not to proceed typically reflects a lack of evidence (so that a conviction is unlikely), an understanding that the defendant’s actions were justifiable (so that a defense to the crime exists), or a recognition that the crime was relatively insigni-ficant (so that obtaining a conviction is not an efficient use of resources). Selective prosecution in the transitional justice context, on the other hand, takes place despite compelling evidence that the perpetrators have committed the most heinous of crimes, and have done so without justification.
The inevitable selectivity of prosecution does not render it an unacceptable approach. Despite this limitation, trials can achieve important criminal justice goals. Nevertheless, the narrow scope of prosecution—due in part to the requirements of due process—is central to an assessment of how well prosecution can fulfill each of the five possible goals of criminal justice.
Desert theory is premised on the relationship between punishment and culpability. “Retributivism is a very straightforward theory of punishment,” writes penologist Michael Moore. “[W]e are justified in punishing because and only because offenders deserve it.”[71] Unlike other theories of criminal justice, “just deserts” is thus explicitly backward-looking. For a pure retributivist, there is a Kantian categorical imperative to punish, whether or not punishment will prevent future crime.[72] While desert theory is distinctive in its concern for the past rather than the future, it is like most other approaches to criminal justice in its focus on the offender. Victims may re-
ceive satisfaction from the knowledge that the perpetrator is punished.[73] But because crime is an offense against society as a whole, it is society, and not the victim, that determines the perpetrator’s guilt and the appropriate level of punishment. Moreover, decisions about the seriousness of the offense are based on the offender’s moral culpability, not on the degree of harm suffered by the victim.
Three basic questions usually arise in considering approaches to punishment: who should be punished, how much should they be punished, and why should they be punished? Desert theorists generally agree on the answers to the first two questions.[74] Regarding the first question, desert theorists believe that punishment should be restricted to those who have committed crimes.[75] As for how much to punish, desert theorists rely on the principle of proportionality, the notion that the severity of a sentence should be proportional to the seriousness of the criminal conduct.[76] The related requirement of ordinal proportionality determines the degree to which a crime should be punished relative to other crimes.[77] The real debate in desert theory concerns the third question: why punish? The tension among desert theorists—like the skepticism of critics of retribution—is grounded in a deep discomfort with vengeance. Revenge, some fear, will lead to a downward spiral of violence and recrimination.[78] Moreover, explains scholar Jeffrie Murphy, one tends to think “that it is only primitives who would actually hate criminals and want them to suffer to appease an anger or outrage
that is felt toward them. Good people are above such passions or at least they try to be.”[79]
Desert theorists have responded to the “vengeance problem” in several ways. Some have sought to prove that retribution and vengeance are not the same thing, or that retribution can be justified without reliance on the concept of revenge.[80] Other theorists, notably Susan Jacoby, have sought to rehabilitate the concept of revenge itself.[81] Jacoby argues that we should recognize that criminal justice is based on revenge[82] and stop pretending that “justice and vengeance have nothing, perish the uncivilized thought, to do with each other.”[83] Law serves to channel vengeance, thereby both discouraging less controlled forms of victims’ justice, such as vigilantism,[84] and restoring the moral and social equilibrium that was violently disturbed by the offender.[85]
Some desert theorists, while avoiding the term “vengeance,” rely on the similar concept that “[m]ost people react to [atrocious crimes] with an intuitive judgment that punishment (at least of some kind and to some degree) is warranted.”[86] Under this theory, qualms about the legitimacy of retribution blind one to the fact that the belief in punishment is actually moti-
vated by the desire for retribution.[87] The problem with such intuition-based arguments for retribution is that not everyone shares the desire to punish;[88] in fact, some victims plead for clemency for their tormentors. Nor does sympathy for the retributive victim’s desire that her wrongdoer suffer necessarily mean that such suffering is justified, or even morally right. As Nino argues, retributivism “presupposes that it is sometimes appropriate to redress one evil with another evil. However, when I add the evil of the crime to the evil of the punishment . . . my moral arithmetic leads me consistently believe that we have ‘two evils’ rather than ‘one good.’”[89]
How effective is prosecution in achieving the goal of retribution against those who commit massive human rights abuses? Because it provides a legitimate way in which to impose severe punishment, prosecution is better suited to retribution than other forms of transitional justice. In the face of atrocity, both individuals and societies have a powerful need to call those who caused the suffering to account. It is difficult to accept that the worst perpetrators of genocide and war crimes should escape responsibility.
In fact, advocates of prosecution emphasize its retributive qualities. Some openly use the image of “getting even.” Aryeh Neier, for example, expresses the wish that in establishing the ICTY, the international community had simply stated, “[f]rom now on, those who commit great crimes will pay.”[90] Generally, however, retribution is more politely described in terms of combating impunity or bringing perpetrators to justice. For example, Diane Orentlicher writes that the world community “has resolved emphatically that it will not countenance impunity for massive atrocities against persecuted groups.”[91] Similarly, the argument for prosecutions made in a recent report to the United Nations by a group of experts assigned to assess possibilities for transitional justice in Cambodia is that “crimes such as those of the Khmer Rouge deserve punishment as a matter of morality and fundamental considerations of justice.”[92] The UN Security Council resolution that established the ICTY likewise speaks of the world’s “determin[ation] . . . to bring to justice” those responsible for the atrocities.[93] Although not framed
in the language of vengeance or retribution, the underlying assumption of such statements is that such horrible crimes should not go unpunished.
Prosecution is not, of course, the only way to exact retribution. Lustration can strip perpetrators of their jobs or remove them from elective office.[94] Through civil suits, victims and their heirs can exact financial penalties from those who committed abuses.[95] National truth commissions can “generate social opprobrium,”[96] turning perpetrators into social outcasts and forcing them to face victims on television. However, the sanctions imposed through alternative mechanisms—sanctions such as social opprobrium, ostracism, money judgments, or loss of jobs and privileges—simply are not proportional to the crimes committed by human rights violators.[97] As we have seen, because courts afford defendants far more due process protections than truth commissions or lustration committees, prosecutions can legitimately sanction behavior more severely. As a result, “[t]he greater the felt need for punishment, the more seriously the prosecution option must be considered.”[98] In other words, if one adopts a retributive theory of transitional justice, prosecutions have significant advantages over other accountability mechanisms.
While prosecutions may be more effective than other approaches in achieving retributive goals, true retributive justice is almost always unachievable in the wake of radical evil. This is true for several reasons. First, it is often impossible even in prosecutions to impose a punishment that is proportional to the crime. As Minow argues, massive human rights atrocities “call for more severe responses than would any ordinary criminal conduct, even the murder of an individual . . . . And yet, there is no punishment that could express the proper scale of outrage.”[99] What punishment would be proportional to such heinous offenses? Theoretically, one could torture torturers and rape rapists. But societies committed to human rights and individual dignity must not take this path. Killing killers remains a popular form of retribution in many domestic criminal justice systems. However, while certain courts of transitional justice, such as that established at Nuremberg after World War II, have condemned human rights abusers to death, the death penalty now clearly violates international human rights norms.[100] Even if capital punishment is used, it cannot satisfy the requirements of or-
dinal proportionality because it fails to distinguish the punishment of radical evil from the price paid by an ordinary murderer.[101] Moreover, where a perpetrator’s actions may have caused the deaths of thousands, the loss of one life is not a punishment that adequately reflects the gravity of the crime.
Adequate retribution is impossible unless those inflicting punishment violate the rights of human rights violators, and even then the punishment often fails to approach the horror of the crime. “[J]ustice,” explains theologian Donald Shriver, “falls limp before monster-sized evil.”[102] In the final analysis, however, the ultimate futility of retribution should not be a reason to discount the utility of prosecution. Even if human rights violators can never be punished enough, they can still be punished severely. And if the desired goal for transitional justice is retribution, inadequate penal sanctions imposed after trial are still preferable to grossly inadequate civil liability or public shaming.
The second and more fundamental critique of a retributive approach to transitional justice is that it depends on the concept of blame, which requires character evaluation.[103] As Nino has argued, the viability of character evaluation in the context of radical evil is unclear.[104] Noting Hannah Arendt’s puzzlement that Adolf Eichmann organized mass murder in order to advance his career rather than to cause harm to others,[105] Nino concludes,
Of course, banal evil is still evil. But are we prepared to blame a character which we evaluate as banal rather than full of burning hatred, sadistic inclinations, and cruelty? Even if retributive punishment were justified in general . . . it may be unsuitable for radical evil.[106]
Yet perhaps it is easier to blame the banal than Arendt and Nino recognize: we might find a technocrat who believes that managing the details of geno-
cide will further his career more reprehensible than a hate-filled murderer. Moreover, societies regularly punish crimes that are undertaken in pursuit of commonplace goals. While the burglar’s incentive may be a bigger paycheck and the drug pusher’s motive may be a promotion in the drug ring, retribution may still be appropriate. Nevertheless, Arendt’s and Nino’s insights are powerful. Radical evil involves horrific acts that even ordinary criminals would find appalling.[107] It is often committed by average people who would never commit ordinary crime. How can one understand intent in such circumstances? Perhaps, as Nino suggests, the proper response is to suspend reactive judgments (much as one might do with the insane), because the perpetrators of mass atrocity “have gone beyond the pale of humanity by rejecting the framework of interactions that blame presupposes.”[108]
Character evaluation in the aftermath of mass human rights violations is further complicated by the difficulty of assigning individual responsibility, especially in the case of lower-level participants. Minow argues that
[t]he central premise of individual responsibility portrays defendants as separate people capable of autonomous choice—when the phenomena of mass atrocities render that assumption at best problematic. Those who make the propaganda but wield no physical weapons influence those with the weapons who in turn claim to have been swept up, threatened, fearful, mobilized.[109]
Of course, individuals always have choices. As the German philosopher Karl Jaspers wrote shortly after World War II, “I, who cannot act otherwise than as an individual, am morally responsible for all my deeds, including the execution of political and military orders. It is never simply true that ‘orders are orders.’”[110] But individual autonomy in the context of dictatorship or mass violence may not be the same in the context of ordinary crime. Unlike ordinary criminals, who violate social norms by committing crimes, individuals who are swept up in mass violence do not step outside the prevailing moral framework. Rather, they succumb to intense social pressure. While this does not relieve such individuals of moral agency or responsibility, it does make them more difficult to judge. Former German President Richard von Weiz-säcker has noted a widespread and unfortunate tendency among young Germans “to believe that then people were evil but today they are good.”[111] As Reinhold Neibuhr has quipped, the universality of sin is the only concept in
the Judeo-Christian tradition that is empirically verifiable.[112] Those who have never faced such choices should not be too quick to assume that they would have acted differently.
In addition to the conceptual obstacles to a retributive understanding of transitional justice, prosecutions themselves complicate the retributive framework. The first difficulty lies in the relationship between retribution and vengeance. Ethnic conflicts around the world, in which each side justi-fies the atrocities it inflicts by referring to the wrongs it has suffered, demonstrate the tendency of vengeance to lead to a downward spiral of violence. Prosecutions are designed to channel these demands for vengeance and to break the cycle of personal revenge. As Minow explains, in order
to avoid such escalating violence . . . [one must] transfer the responsibilities for apportioning blame and punishment from victims to public bodies acting according to the rule of law. This is an attempt to remove personal animus, though not necessarily to excise vengeance. Tame it, balance it, recast it as the retributive dimension of public punishment.[113]
Whether or not we equate retribution and vengeance,[114] it is clear that in the context of transitional justice prosecutions founded on a desire for retribution will have many of the drawbacks of vengeance. Where prosecutions are publicly perceived as a form of victor’s justice, they will be unlikely to break the cycle of violence. As José Alvarez notes, “[t]he majority of the thousands detained in Rwanda’s jails today report, and perhaps genuinely feel, that ‘they have done nothing wrong’ and are being victimized merely because they were on the ‘wrong side of the war.’”[115] Even if prosecutions satisfy demands for revenge from Tutsis, they may stoke the desire for vengeance by Hutus. Nino describes a related problem in Argentina, where human rights groups, by adopting an “all-out retributive” approach and demanding that all the guilty be punished, ended up undermining their own credibility, fanning a backlash by military and government forces against the trials, and ultimately weakening the impact of those trials that did take place.[116] Nino suggests that although
many people approach the issue of human rights violations with a strong retributive impulse, almost all who think momentarily about the issue are not prepared to defend a policy of punishing
*** Top of Page 61 ***
those abuses once it becomes clear that such a policy would probably provoke, by a causal chain, similar or even worse abuses.[117]
The second difficulty is that prosecutions, as we have seen, are necessarily limited and selective. Retribution theory, by contrast, is predicated on the notion that everyone should get their just deserts. In the wake of genocide or other mass violence, many people in society will “deserve” to be punished. Yet prosecutions, even at their most extensive, will only reach a few of the culpable. Many offenders who raped, tortured and killed will never be tried. As will be argued below, a deterrence-based rationale for prosecutions can account for exemplary prosecutions of genocidaires. A retributive one, by contrast, cannot.[118]
Selective prosecution further undermines retributive goals because prosecutors rarely succeed in targeting only the most culpable. “[T]he actual set of individuals who face prosecution,” notes Minow, “is likely to reflect factors far removed from considered judgments about who deserves prosecution and punishment.”[119] The failure to prosecute all equally culpable individuals, however, violates the principle of proportionality, which dictates that like crimes should be treated alike.[120] Proportionality is further undermined when prosecutions target lower-level offenders while ignoring more blameworthy ones, since such decisions do not reflect the relative level of social disapproval accorded to different crimes.[121] Yet accepted legal principles may make it extremely difficult to convict those who orchestrated abuses rather than simply carrying them out. For example, reunified Germany found it difficult to hold East German leaders adequately accountable for ordering that fleeing citizens be shot, and decided instead to prosecute several young East German border guards.[122] Such selective, limited prosecu-
tions—the only kind possible in transitional justice—fail to meet the basic requirements of retributive justice.
The third complication is the fact that transitional societies must continue to deal with ordinary crime even as they confront the extraordinary evil of the past. In many societies transitional periods have been marked by soaring crime rates and accompanying efforts to prosecute and punish offenders.[123] In light of this reality, a retributivist must ask who, for example, is more guilty: the young, uneducated, unemployed black carjacker from a Johannesburg township, or the white police officer who killed the boy’s father during a political rally several years before? Due to the inevitable limitations on prosecutions in the transitional context, even those prosecutors who believe that the youth is less culpable will concede that they will be more likely to prosecute him than his father’s murderer.[124] It may well be true, as Paul van Zyl has argued, that “[i]f the police and prosecuting authorities were to devote a significant share of their resources to dealing with human rights violations, many of which occurred a decade or more ago, the country would almost certainly lose the current battle against ongoing crime.”[125] But the need to control crime is no answer to the basic retributivist objection: the greater the culpability, the more severe the punishment should be. Retributive theory cannot support a form of prosecutorial selectivity that excuses more serious past crimes in order to address less serious current ones.
Furthermore, amnesties for past offenses typically do not apply to ordinary crime. South Africans, for example, could only seek amnesty if their offenses were “associated with political objectives and committed in the course of conflicts of the past.”[126] From a retributive point of view, it is not immediately clear why a murderer who kills for political reasons should be entitled to amnesty in return for the truth, while one who kills out of passion or greed should not. Society’s willingness to forgo punishment for some political offenses (whether through an express amnesty or a de facto one resulting from selective prosecution), but not for common crimes, suggests
that individual culpability is understood very differently in the contexts of extraordinary evil and ordinary crime.
To summarize, in the face of horrific crimes, many people, both in the affected society and the international community, share a powerful sense that such atrocities should not go unpunished. Prosecution, which provides a legitimate way to impose severe punishment, is the most effective method of pursuing retributive goals in transitional societies. But it has limits: one cannot punish fairly, and one cannot punish enough. As an Indonesian victim and prosecution proponent explains, “[it] will have to be a very selective justice and, of course, selective justice is not exactly justice.”[127] An advocate for retributive prosecution might answer these criticisms by simply arguing for more prosecution. But unless everyone can be prosecuted, which is impossible, the objections remain. Although the desire for retribution is a justifiable response to extraordinary evil, retribution itself is only partially achievable.
The term “deterrence” is often used interchangeably with “prevention.” In fact, deterrence is only one way to prevent crime. Under deterrence theory, potential offenders may still be capable of committing crimes (since they are not incapacitated) and may still desire to commit crimes (since they are not rehabilitated). But despite their capacity and desire, potential offenders are inhibited by the “intimidation or terror of the law.”[128] There are two main types of deterrence: individual deterrence and general deterrence. Individual deterrence seeks to prevent future crime by setting sentences that are strict enough to ensure that a particular offender will not reoffend.[129] General deterrence, on the other hand, attempts to prevent crime by “induc[ing] other citizens who might be tempted to commit crime to desist out of fear of the penalty.”[130] Notably, deterrence theory does not allow, much less require, the punishment of all who are guilty. Moreover, general deterrence does not even require punishment of all those who might be deterrable as individuals. If
exemplary punishments adequately prevent future crime, they are suffi-cient.[131]
Deterrence occurs, writes Nigel Walker, when people refrain from certain actions because they fear the possible consequences of those actions.[132] In other words, the potential benefit of committing crime is outweighed by the risk of sanctions. Deterrence thus assumes that, were it not for the possibility of adverse consequences, people would engage in crime. “A person is not deterred,” writes Walker, “if he refrains because he is not tempted, or is tempted but restrained by his code of manners or morals.”[133] Deterrence also assumes that the potential offender will undertake a two-part calculation, assessing both the gravity of the consequences and the likelihood of getting caught. This calculation is based not on the objective severity of sanctions or the real risk of apprehension, but on the potential offender’s subjective assessment of these factors. Thus, the effectiveness of any deterrent depends on the potential offender’s perception of possible sanctions, and on her assessment of her ability to evade law enforcement.[134] The actual severity or certainty of punishment is less important than its perceived severity or certainty.[135]
While the logic of deterrence is intuitively appealing, the available empirical evidence regarding the effectiveness of deterrence in domestic criminal justice systems is inconclusive.[136] There are several possible reasons for this.[137] First, it is difficult to prove that threats of legal sanctions, rather than other motivations, have prevented people from offending.[138] “Marginal deterrence,” the amount by which deterrence increases or decreases based on changes in the severity of sanctions, is particularly difficult to demonstrate.[139] The problem is not in showing that deterrence can occur when punishment is certain, swift, and severe, but in determining when and to what extent it occurs under real-world conditions, under which punishment is never certain, rarely swift, and only sometimes severe.[140] Second, effec-
tiveness depends on context.[141] The risk of detection and punishment must not be so low as to be readily discounted, and the penalty must be adequately publicized.[142] Moreover, the offender and the crime must be deterrable; that is, they must be rational.[143] Heightened enforcement and increased sanctions may reduce crimes that reflect rational choices, but they are unlikely to have an impact on irrational offenders. Deterrence, then, only works in relation to some crimes and some offenders.[144]
Scholars and human rights activists have trumpeted deterrence as perhaps the most important justification for prosecution in transitional justice. Orentlicher, for example, writes that “[t]he fulcrum of the case for criminal punishment is that it is the most effective insurance against future repression.”[145] Neil Kritz shares Orentlicher’s confidence in the effectiveness of trials as a deterrent, noting that the failure to prosecute at least key figures “can be expected not only to encourage new rounds of mass abuses in the country in question but also to embolden the instigators of crimes against humanity elsewhere.”[146] M. Cherif Bassiouni adds that “[t]he relevance of prosecution and other accountability measures to the pursuit of peace is that through their effective application they serve as deterrence, and thus prevent future victimization.”[147]
Given the manner in which deterrence functions in the domestic criminal justice context, how tenable is the widespread deterrence justification for prosecution in the wake of massive human rights violations, war crimes, or genocide?[148] Deterrence theory is quite useful as a justification for selective
prosecution.[149] In particular, because deterrence does not require punishment of all the guilty, it does a better job than desert theory of explaining why exemplary punishments of a few offenders may be acceptable. Although the failure to prosecute heinous abuses is impermissible on retributive grounds, deterrence theory can accommodate selective prosecution. The explicitly utilitarian goals of deterrence also provide a justification for abandoning post-transition prosecutions when “the mischief [they] would produce would be greater than what [they] prevented.”[150] When ousted leaders retain significant power in a country in transition, harsh punishment could lead them to seize power again. Thus, criminal punishment may have a positive or negative overall effect on human rights.[151] A deterrence rationale allows for a cost/benefit analysis in which one assesses whether the advantages of preventing crime through prosecution outweigh the costs to democracy and human rights that might result if trials lead to political instability.[152] Deterrence permits a decision not to prosecute where trials would cause more problems than they would solve.
If deterrence is the justification for prosecution, one must determine if prosecutions actually prevent human rights abuses. It is even more difficult to show the effectiveness of prosecution as a method of deterring mass atrocity than it is to demonstrate its effectiveness in the context of ordinary crime. In fact, it is virtually impossible to assess whether or not the threat of prosecution has ever prevented genocide and war crimes. As Minow notes, “[n]o one really knows how to deter those individuals who become potential dictators or leaders of mass destruction . . . . One hopes that current-day prosecutions would make a future Hitler, or Pol Pot, or [Bosnian Serb leader] Radovan Karadzic; change course, but we have no evidence of this.”[153] Given the unyielding stream of atrocities the world has witnessed since Nuremberg, it is difficult to argue that these trials had any discernable effect.[154] Similarly, many of the worst atrocities in the former Yugoslavia took place after the ICTY was established.[155] It could be that in the absence
of these prosecutions, many more such atrocities would have taken place. But those who point to the deterrent effect of prosecutions bear a heavy burden of proof indeed.
Any deterrent effect that prosecutions might have will depend on context, including the risk of “getting caught,” the severity of penalties, the extent of public knowledge about such sanctions, and the degree to which the crime and the offender are deterrable. In domestic criminal justice, the risk of “getting caught” typically concerns the risk of detection, since in a functioning criminal justice system, offenders whose guilt can be proven will generally be punished if they can be apprehended.[156] In the transitional justice context, on the other hand, “getting caught” usually has little to do with the risk of detection; indeed, many atrocities are committed in plain view.[157] Rather, “getting caught” primarily concerns the chance of being punished. Thus, the fact that in the wake of mass atrocities only a small number of those implicated will ever be prosecuted undermines the logic of the deterrence argument.[158] Those who “merely” kill, rape, and plunder, but do not mastermind the carnage, have little to fear from prosecution. Even for those who orchestrate human rights abuses, the risk of “getting caught” is low. “[I]t is not irrational,” writes Minow, “to ignore the improbable prospect of punishment given the track record of international law thus far.”[159] Of course, in the rare cases where perpetrators do “get caught,” the sanctions may be considerable. However, it seems doubtful that severe penalties for massive human rights violations will have much deterrent value when they are so heavily discounted by the negligible likelihood of prosecution.[160] Nor is it clear that the foot soldiers of atrocity will even be aware of the heavy sanctions imposed on a few high-level perpetrators in some far-off land.[161]
In response, one might argue that the international community should substantially increase the likelihood that human rights abusers will face criminal prosecution. But it is not clear how much of a deterrent effect increased prosecutions would have on genocidaires. Potential war criminals may underestimate the actual risks. Robert Jackson, the lead prosecutor at Nuremberg, questioned the degree to which that tribunal could serve as a deterrent, given that wars are almost always started in anticipation that they will be won. “Personal punishment, to be suffered only in the event the war is lost,” he argued, “is probably not [enough] to be a sufficient deterrent to prevent a war where the war-makers feel the chances of defeat to be negligible.”[162] Moreover, as Jon Elster notes, “even if violations are harshly punished now, how can future would-be violators know that they, if overthrown, will be treated in the same way? Incentive effects presuppose stable institutions, which almost by assumption do not exist.”[163]
Scholars have also debated whether or not massive human rights violations involve crimes or criminals that are deterrable. Are such crimes subject to a rational assessment of costs and benefits? One former prosecutor of the ICTY has claimed that “deterrence has a better chance of working with these kinds of crimes [war crimes, genocide, crimes against humanity] than it does with ordinary domestic crimes because the people who commit these acts are not hardened criminals; they’re politicians or leaders of the community that have up until now been law abiding people.”[164] Such analysis seems fundamentally misguided. When “ordinary” people commit horrible crimes, it suggests that the normal restraints of law and deterrence are not working, or that these people are no longer functioning rationally.[165] At the same time, some individuals do make rational choices when committing horrific crimes, and would therefore potentially be deterrable. As Douglass Cassel notes, certain dictators like former Yugoslav President Slobodan Miloševic; are manipulators, not fanatics, and might be restrained by credible threats.[166] Other perpetrators, such as Hitler, however, are probably undeterrable. Moreover, some atrocities are carefully planned and staged, while others are “crimes of passion” or crimes of hate. Just as in the domestic context, deterrence, if it works at all, will only work against some offenders and some crimes.[167]
It is worth remembering that deterrence does not work when perpetrators, while recognizing and fearing the possibility of punishment, nevertheless decide to engage in crime because the potential rewards outweigh the risks. Dictators who knowingly accept the risk of prosecution, but nevertheless commit atrocities in order to retain political and economic control, may rationally believe that crime does pay. In the domestic context, a deterrence theorist’s response to this problem would be to continually increase both the penalties and the likelihood of apprehension. In the transitional context, however, penalties could not be much more severe,[168] and the chances of punishment, while they could be increased, will likely never be extremely high.
Given that prosecution is not a particularly effective deterrent against gross human rights violations, are alternative mechanisms any better? In comparing prosecutions and other sanctions—such as civil liability, lustration, or public shaming—we must examine their relative effectiveness in terms of marginal deterrence, which as we have seen is difficult to assess. If deterrence depends on “intimidation or terror of the law,” it is likely that a potential human rights violator will fear incarceration more than a money judgment, the loss of his job, or the shame of a public confession. Moreover, even dictators who never expect to stand trial may fear the consequences of an indictment. Once an international warrant of arrest has been issued against a suspected war criminal, that person is liable to arrest in virtually every country in the world, making it difficult for her to hold high public office in her own country and virtually impossible for her to participate in international negotiations.[169] But even though trials threaten more severe punishment than alternative mechanisms, it is not clear what effect this increased potential sanction has on deterrence. Deterrence depends not only on the severity of the sanction but also the certainty of punishment. While the likelihood that human rights violators will face any sort of sanction remains small, some accountability mechanisms are better equipped than others to handle large numbers of offenders. Prosecutions may yield severe punishments, but they are rare. Will a highly uncertain but severe punishment have a greater deterrent effect than a lesser but more likely sanction? Perhaps the deterrent effect differs depending on the person being deterred; would-be dictators might be dissuaded by the fear of prosecution, while
low-level functionaries, well-aware that they are unlikely ever to be tried, might regard lustration or truth commissions as a more credible threat.
Prosecutions may deter some future human rights abusers, and prosecutions may even have a greater deterrent value than alternative post-transition mechanisms. However, it is unlikely that post-atrocity prosecution is the most effective way to prevent future atrocities. As penologists have noted in the context of ordinary crime, the severity and likelihood of a legally imposed penalty are not “the only or necessarily the most powerful influence[s] on a person’s behavior.”[170] An individual’s actions are often affected to a greater degree by moral norms than by fear of punishment. The reason most people do not murder is not because they are afraid of getting caught, but because they believe that murder is wrong. Similarly, the constraints a society imposes on itself may have more to do with its political culture and form of government than with concern about the possible consequences of misbehavior.[171] Even where a person does refrain from taking a desired action out of fear of the possible consequences, legal sanctions may play only a minimal role. Frequently, non-legal deterrents, a pervasive fact of human life[172], are much more powerful than legal ones. An individual’s decision not to assault someone after an insult in a bar may have more to do with a fear of being beaten up than with any worry about what the police might do. Similarly, while the threat of prosecution may deter some leaders contemplating atrocities, such persons are probably more likely to hold back out of fear of vigorous public criticism, political pressure, diplomatic isolation, economic sanctions, or even military intervention.
To summarize, if deterrence is our goal, our underlying concern will be the prevention of future crimes. It is by no means clear that prosecution is the most effective mechanism for preventing atrocities. If the international community sits by and watches while atrocities occur, demanding prosecution only after the violence has stopped, arguments about the deterrent effect of such trials will ring hollow.[173] Alvarez is right to warn that the international community will lose its credibility unless “[i]nternational efforts to prevent the continuation of genocidal acts and other acts of violence . . . pre
cede attempts at criminal accountability.”[174] This is not to say that prosecutions, or other transitional justice mechanisms, have no deterrent value. But if the goal is to deter future human rights abuses by making potential abusers afraid to act, the international community has bigger sticks to shake than the threat of trial.
Rehabilitation seeks to prevent the future commission of crime by curing previous offenders of their criminal tendencies.[175] Thus, the success of rehabilitation is measured by recidivism rates, rather than by changes in the aggregate incidence of crime.[176] There are two possible reasons for wanting to “cure” the offender. First, one might argue that society will be safer once the offender is rehabilitated and is no longer committing crimes.[177] Second, one could believe that offenders should be given the opportunity to have productive lives for their own sake.[178] Rehabilitation de-emphasizes the link between the gravity of the crime and the severity of the sentence. Appropriate rehabilitative sentences reflect the measures necessary to reintegrate a particular offender into the community. Once an offender has been rehabilitated, further punishment is unnecessary. Strict proportionality is not required; among offenders who commit the same crime, some will take longer than others to rehabilitate.[179]
The primary objection raised by critics of rehabilitation is that it simply does not work.[180] Even researchers who favor rehabilitation believe that treatment programs are effective only with certain types of offenders.[181] Other critics argue that rehabilitation denigrates human dignity.[182] Still
others attack the allegedly benevolent purpose of rehabilitation, since rehabilitative regimes may in fact inflict a greater deprivation of liberty on their subjects than avowedly punitive programs.[183]
Advocates for prosecutions as the optimal form of transitional justice frequently use the language of rehabilitation, but their focus is on societal, not individual, rehabilitation.[184] Few indeed would think of prosecution and punishment as a way to redeem despots like Pol Pot[185] and Pinochet. By contrast, support for societal rehabilitation—the idea that prosecutions can change a society’s moral values by “foster[ing] respect for democratic institutions and thereby deepen[ing] a society’s democratic culture”[186]—is widespread.
Prosecutions are believed to have at least three curative powers. First, prosecutions help to establish the truth. Most scholars of transitional justice agree that exposure and acknowledgement of the past is a prerequisite for future social stability.[187] Prosecutions educate the public about the nature and extent of prior wrongdoing[188] and contribute to a shared historical understanding. Through this educational process, writes Stephan Landsman, prosecutions “may serve both to inoculate the populace against lapses into oppressive behavior and as a means of establishing an accurate account of what actually transpired before the democratic regime came to power.”[189] Second, prosecutions help to establish the rule of law. “Holding violators accountable for their misdeeds,” explains Landsman, demonstrates to “all members of society that the law’s authority is superior to that of individuals.”[190] By contrast, failure to enforce the law undermines its authority.[191]
Third, prosecutions reinforce moral norms and contribute to a shared understanding that certain behavior is wrong. In order to prevent future atrocities, one must establish not only the truth about past abuses, but also a national and international consensus that such acts are unacceptable.[192] One important way to communicate such a moral consensus is through the criminal law.[193] Punishment thus not only reflects but also shapes moral values.
The fact that prosecutions can promote societal rehabilitation does not necessarily mean, however, that trials are more effective than non-prosecution alternatives at curing societies of their evil tendencies. The superiority of prosecution as a means of accomplishing the first goal—establishing the truth, educating the public, and forming a shared historical understanding—is dubious. Some scholars have suggested that trials provide a “higher quality” truth than alternative mechanisms because they are more narrative and dramatic.[194] While trials may have moments of high drama, their formalism and rigidity can also make them excruciatingly boring.[195] This is particularly true if due process standards are respected, since “[w]hat makes for a good ‘morality play’ tends not to make for a fair trial.”[196] Other forms of dramatic truth-telling, such as the televised confrontations between victims and perpetrators before the South African Truth and Reconciliation Commission, may be as good or better at capturing the public imagination.
Some scholars have suggested that, because of the higher evidentiary standards imposed on trials, the truth produced through prosecution is more accurate than that established through alternative mechanisms.[197] In fact,
prosecution is at best an imperfect means to develop a complete record of the past.[198] First, rules of evidence typically reflect not only a desire to ascertain the truth, but also competing public policy or constitutional concerns. For example, in the United States relevant and probative evidence is routinely excluded from criminal prosecutions when it is considered excessively prejudicial, denies the defendant her right of confrontation, or forces a defendant to incriminate herself. Yet in the wake of gross human rights violations, gruesome photographs, flagrant hearsay and perpetrator confessions are essential to developing an accurate picture of the past.[199] Moreover, while the threat of prosecution can be an important tool in forcing perpetrators to participate in other truth-seeking mechanisms,[200] prosecutions themselves are ill-suited to eliciting testimony from perpetrators, the very people who know the most about the atrocities. “The primary sources of information concerning those infamies, the perpetrators themselves,” notes the late South African Constitutional Court Justice John Didcott, “would hardly be willing to divulge it voluntarily, honestly, and candidly without the protection of exemptions from liability.”[201]
The most serious deficiency of prosecutions when it comes to truth-seeking is that trials focus on select individuals and thus do not account, in Minow’s words, for “the complex connections among people that make massacres and genocides possible.”[202] The history produced by judges is “the by-product of particular moments of examining and cross-examining witnesses and reviewing evidence about the responsibility of particular individuals.”[203] Moreover, international tribunals are often located far from the affected society in transition, making the process less accessible to victims, witnesses, and the public. In sum, “if the goal to be served is establishing consensus and memorializing controversial, complex events, trials are not ideal.”[204]
Alternative mechanisms may be better suited than prosecution to developing full records of the past. Some have argued that due to the lower standard of proof and evidentiary and discovery advantages of civil proceedings,
civil liability is better than criminal prosecution at establishing “a definitive, historically accurate account of the atrocities.”[205] This may be true, but civil trials, just like criminal trials, focus on individual responsibility and shift public focus away from systemic, shared culpability. By contrast, truth commissions are not limited to the facts of individual cases, but highlight the vast scope of and widespread complicity in the human rights violations. In addition, the history produced through truth commissions focuses not just on perpetrators but also “on victims, including forgotten victims in forgotten places.”[206]
The merits of prosecution in achieving the second goal of rehabilitation, establishing the rule of law, are also unclear. Trials conducted before impartial courts that scrupulously observe due process requirements may showcase the benefits of the rule of law and contrast favorably with the lawless behavior of the defendants.[207] Yet when those same due process protections free those who are perceived to be guilty, fair trials may well inspire contempt for the rule of law.[208] On the other hand, trials which are conducted unfairly or which offend legal principles will undermine the “spirit of legality”[209] that such trials are supposed to inculcate. For example, human rights activists initially cheered the prosecution in Senegal of former Chadian dictator Hissène Habré[210] as demonstrating that even such an “ultimate untouchable”[211] may not violate the law. However, after a change of government in Senegal, the case against Habré was dismissed,[212] reinforcing the notion that the powerful are untouchable.[213] A similar problem arises when the “small fry” are prosecuted instead of high level officials, fostering a perception that prosecutions are a form of scapegoating, not a means of achieving justice.
Perhaps the most distinctive contribution prosecution can make to societal rehabilitation is in establishing the wrongfulness of past atrocities. The prosecution and punishment of atrocities is a forceful way to disavow such conduct. A process of censure not only expresses disapproval of those who violate international human rights norms, but also serves to define and strengthen the norms themselves. While prosecutions help to establish
moral norms, one cannot assume that in every case “once accusations are leveled and indictments and judgments are issued, all will come to acknowledge the barbarous evils committed by genocidaires.”[214] Rather, one must identify those situations in which prosecutions will help to forge common moral values.[215] While prosecutions can produce moral consensus, they can also create scapegoats and feelings of bitterness, particularly where the selection of defendants appears to be politically motivated or where there is a perception that the trial represents victor’s justice. One must be wary of a vicious circle: where prosecution is not grounded in moral consensus it will be seen as victor’s justice. And if it is seen as victor’s justice, it will lose its effectiveness as a tool for creating moral consensus. A society cannot be cured of a condition it does not regard as a disease.
Finally, is prosecution the most effective way to foster societal rehabilitation? Certainly prosecutions shape norms by condemning undesirable conduct. But trials also express the belief that “specific individuals—not entire ethnic or religious or political groups—committed atrocities.”[216] While individual crime is a component of mass atrocity, “radical evil,” as Carlos Nino notes, also
requires an evil political and legal framework in which to flourish. Without that framework, it is unlikely that massive, state-sponsored human rights violations will ensue regardless of whether punishment for previous violations takes place. With that framework in place and given certain antecedent circumstances, however, violations are highly likely even with previous convictions and punishment for human rights violations.[217]
Perhaps funds spent on prosecutions would have a stronger rehabilitative effect if spent on reforming a society’s political and legal framework. Prosecutions may well be less successful in rehabilitating a society than a concerted effort to reduce inequalities in wealth, provide basic public education, create functioning courts, establish civilian control over the military, ensure the independence of the press, or hold free and fair elections.
Two major paradigms fall within the rubric of restorative justice.[218] The first focuses on compensating victims[219] and views crime as a harm that criminal justice should seek to undo.[220] Under this view, the purpose of punishment is to repair injuries to victims,[221] and thus the goal of criminal justice is for the offender to provide restitution to the victim. The second paradigm envisions crime as conflict and criminal justice as a form of conflict resolution.[222] The basic assumptions of this approach are:
(1) Crime is primarily a conflict among individuals resulting in injuries to victims, communities and the offenders themselves; only secondarily is it lawbreaking.
(2) The overarching aim of the criminal justice process should be to reconcile parties while repairing the injuries caused by crime.
(3) The criminal justice process should facilitate active participation by victims, offenders and their communities. It should not be dominated by the government to the exclusion of others.[223]
Under this view, the goal of criminal justice is the reconciliation of the offender, victim, and community.[224]
These two paradigms, compensation and conflict resolution, are often linked. An apology without restitution may mean little; if a friend apologizes for taking a pen but does not return it, her statement is worthless.[225] “Apologies set the record straight; restitution sets out to make a new rec-
ord,” explains theologian Donald Shriver.[226] Yet conflict resolution cannot rest on compensation alone, particularly since it is often impossible to restore to the victim what she has lost or repair the harm that she has suffered. While a stolen pen can be replaced, a murdered child cannot.
Restorative justice differs from retribution, deterrence and rehabilitation in its focus on the victim.[227] Restorative justice raises questions about the identities of the parties to the conflict. In other words, it asks whose interests are relevant to the case.[228] In the West, crimes are defined as offenses against the state; in this way, as Nils Christie argues, the state has stolen conflicts from victims,[229] communities and offenders. “Virtually every facet of the criminal justice system works to reduce victims, offenders and communities to passive participants,” notes Daniel Van Ness.[230] Restorative justice aims to return conflicts to the parties to the conflict. Yet the idea that crimes are offenses against society at large should not be ignored, for offenders harm not only specific victims but also entire societies. Thus, restorative justice must consider the interests of both the individual victim and the wider community.
Theoretically, the compensatory paradigm of restorative justice is somewhat disconnected from culpability, since the degree of harm caused may not reflect the blameworthiness of the offender.[231] Proportionality in this context is based on the amount of harm inflicted on the victim, not the maliciousness of the offender’s intent.[232] The conflict resolution paradigm, on the other hand, is more likely adequately to address culpability; a victim will be angrier at someone who hits her intentionally than at someone who does so accidentally, and will require more of the offender in order to resolve the conflict.[233]
In the context of transitional justice, reparations can be seen as a form of compensation, while approaches seeking to heal society’s wounds can be un-
derstood within a conflict resolution paradigm. In both cases, as in the domestic context, victims claim a central role.[234] This Section will first look at reparations as a form of transitional justice, and then assess the capacity of prosecutions to reconcile victims, offenders, and society as a whole.
Reparations in the wake of massive human rights violations are designed to provide at least partial restitution to victims. While compensation typically involves monetary payments to individuals, other types of restitution, like building memorials or naming streets for victims, are less tangible and less directly focused on individuals. The truth itself can also be understood as a form of reparation. When the silence is broken, and families learn where the bodies are buried or victims discover the identities of their torturers, the injury caused by past abuse may begin to be repaired.
Reparations should not only acknowledge the survivor’s loss, but also repair the harm caused. While this is a valuable goal, it is often difficult to achieve in practice. Where former regimes have harmed both individuals and entire communities, should scarce resources be used to pay compensation to individual victims, or to rebuild a society victimized by poverty, appalling health care and lack of education? Moreover, monetary measures cannot remedy non-monetary harms, like the loss of a child or the agony of remembered torture.[235] “[N]o market measures exist,” writes Minow, “for the value of living an ordinary life, without nightmares or survivor guilt.”[236] Reparations are likely to be grossly disproportionate to the damage caused, and many thus trivialize suffering.[237] In fact, victims frequently express only modest demands for reparations, such as a tombstone or death certificate for their loved ones or the removal of bullets from their own bodies.[238] Often, their most significant demand is for the truth. Thus while restitution can never fully compensate victims, it can serve an important symbolic function.
Proponents of prosecution sometimes argue that trials are useful as a basis for reparations.[239] This justification seems dubious, at least insofar as it is used to support the primacy of prosecutions. Since prosecution focuses on establishing the guilt of a few individual perpetrators, it is not an ideal tool for identifying victims who deserve compensation. Moreover, because prosecutions are adversarial, defendants may never reveal those facts about the past which victims most want to know. Arguably, alternative mechanisms are better able to identify victims, as well as to provide them with a more
comprehensive, and more personal, account of the past.[240] Truth commissions and civil suits offer at least the possibility that victims will be compensated.[241] Reparations may even take the form of governmental or international aid programs providing medical treatment, scholarships for victims’ children, or preferential access to government services such as public housing or transportation. Any of these approaches is more likely to provide real restitution to victims than prosecutions. Thus the reparative paradigm of restorative justice offers little justification for prosecution.
How do prosecutions hold up under the conflict resolution paradigm of restorative justice? First, we must ask if reconciliation or forgiveness are even possible after massive human rights violations. In the domestic context, restorative justice programs are usually restricted to minor offenses. Yet conflict resolution in the context of transitional justice requires interaction between those who carried out and those who suffered from horrific atrocities. “Healing,” worries Minow, may be “an absurd or even obscene notion for those who have died,” as well as for the survivors, who often feel as if they “have died or live among the dead.”[242] Susan Dwyer expresses a similar concern that “[r]econciliation is being urged upon people who have been bitter and murderous enemies, upon victims and perpetrators of terrible human rights abuses, upon groups of individuals whose very self-concep-tions have been structured in terms of historical and often state-sanctioned relations of dominance and submission.”[243] Absent a clear explanation of reconciliation and what it requires, she continues, “proposing reconciliation will seem like a political sop aimed at masking moral defeat.”[244]
So, what is reconciliation? Scholars of transitional justice distinguish between the repair of relationships that will suffice for a society to move forward and unrealistic expectations of transformative interactions between victims and perpetrators. Dwyer, for example, believes apologies and forgiveness are not absolutely required for future interaction. She contrasts personal reconciliation between victims and perpetrators, which may be too much to ask, with national reconciliation, which she argues is more possible.[245] What is necessary if perpetrators and victims are to live together in
the future, she claims, is the development of a common national historical narrative, based on agreed-upon facts and a shared interpretation of them.[246] Shriver, in contrast, prefers the concept of “forgiveness.”[247] Believing victims must have more courage than Dwyer requires, Shriver asks victims “to face still-rankling past evils with first regard for the truth of what actually happened; with resistance to the lures of revenge; with empathy—and no excusing—for all the agents and sufferers of the evil; and with real intent on the part of the sufferers to resume life alongside the evildoers or their political successors.”[248]
Whatever words we use to describe it, the concept of restorative justice is certainly relevant to transitional justice. The horrors of the past can be seen as a form of conflict, and the goal of transitional justice as conflict resolution. Transitional justice can be organized so as to give victims a central role and repair relationships between them, perpetrators, and society at large. In this way, transitional justice can strive for at least enough forgiveness, reconciliation, or healing to make coexistence possible.
A reconciliation-based argument for prosecution is premised on the notion that retribution is a precondition for societal healing. There has been considerable debate about the tension between “justice,” the requirement that those who violate human rights be punished, and “peace,” the desire for both social cohesion and an end to human rights violations.[249] Historically, the concern has been that pushing for justice would jeopardize peace. Increasingly, however, diplomats involved in settling violent disputes and politicians seeking to move their countries forward recognize that a focus on past atrocities is not an obstacle to stability and conflict resolution, but is, in Kritz’s words, “an integral and unavoidable element of the peace process.”[250]
Proponents of prosecution have equated this new framework, in which justice is a precondition for peace, to an argument that retributive punishment must precede social healing.[251] Accordingly, before there can be reconciliation in places like Bosnia and Rwanda, there must be retribution, at least as far as the most serious offenders are concerned.[252] To achieve closure, “individuals emerging from massive abuse and trauma [must] develop ap-
propriate mechanisms to confront and process that past experience.”[253] But the need to face one’s past does not adequately explain why such a confrontation must be retributive. Forgiving, after all, is not actually the same as forgetting.[254]
Perhaps, as is frequently asserted, one cannot forgive what one cannot punish.[255] While this argument has strong intuitive appeal, it is problematic. First, it is a conclusion, not a verifiable fact. It seems equally plausible that those most likely to demand punishment are those least likely to forgive. Second, the extent to which forgiveness is contingent on punishment likely depends on context. Some offenses are easier to forgive in the absence of punishment than others, and some cultures and religions promote forgiveness more than others. Third, if one cannot punish all, how can one possibly forgive all? If only the ringleaders are punished, can one really expect victims to forgive when they must pass their tormentors in the street on a daily basis? If punishment is a prerequisite, reconciliation between the perpetrators and their victims is impossible. Fourth, and most importantly, the argument that one cannot forgive what one cannot punish suggests that where one cannot adequately punish extraordinary evil, one cannot forgive. If reconciliation depends on forgiveness, and forgiveness depends on punishment, does the impossibility of proportionate retribution in the context of extraordinary evil render reconciliation impossible?
Truth commissions might be better suited to a restorative model of transitional justice than prosecutions. Truth commissions can focus on the victims, craft a shared narrative about the past as the basis for a shared future, and facilitate the active involvement of victims, perpetrators, and the larger community. In essence, truth commissions return the conflict to those who participated in it. The more open, discursive nature of truth commissions better addresses the problem of widespread complicity in massive human rights violations. Moreover, the informality of the process brings other benefits such as the ability of officials to grieve publicly with victims. “[Truth and Reconciliation Commission Chairperson Archbishop Desmond] Tutu cries,” Justice Albie Sachs of the South African Constitutional Court has said. “A judge does not cry.”[256] Furthermore, restorative justice is multi-
directional. Prosecutions may be appropriate where individuals can be clearly classified as either victims or perpetrators. Truth commissions, on the other hand, recognize that this distinction is not always clear. During periods of mass atrocity or repression, individuals usually assume different roles over the course of the conflict. As a result, apologies must be both given and received.
Of course, truth commissions will not always work as a form of restorative justice. Repairing relationships through discussion and confrontation requires not only that all parties be engaged in the process, but also that perpetrators recognize their blameworthiness and accept responsibility for past actions. Amnesty can be understood as an annulment of the appropriate retributive penalty in return for truthtelling.[257] But offenders may be unwilling to tell the truth or to express contrition.[258] Moreover, to be meaningful, apologies must be linked to restitution. Reconciliation in the wake of atrocity requires “the credibility that can be established only by implementation of social and economic programs that concretely address the substantive injustices” of the past.[259] Even those who acknowledge their own culpability or complicity through a national truth-telling process may recoil at the loss of privileges and power that true accountability demands. How much reconciliation can be achieved if in post-apartheid South Africa, for example, whites admit that their economic, social, and political status was based on a morally bankrupt system, but then refuse to accept sharply redistributive taxation?
The fundamental difficulty with restorative justice is that it cannot adequately address demands for retribution.[260] This is not really a problem with restorative justice itself, but rather with the difficulty of reconciling two competing goals of transitional justice. This issue nevertheless deserves comment here, as it lies at the heart of many critiques of truth commissions.
Retribution requires the isolation of perpetrators and the definition of their behavior as outside the realm of human interaction. Restorative justice, in contrast, requires that we recognize the offenders’ humanity and the pos-
sibility of a continuing relationship with them. Forgiveness, argues Shriver, is about repairing fractured human relationships, and therefore “insists on the humanity of enemies even in their commission of dehumanizing deeds, and . . . values the justice that restores political community above the justice that destroys it.”[261] In the context of domestic crime, criminality is sometimes understood as in part the product of social and economic circumstances. In the context of massive atrocities, however, it is far easier to be appalled and to demand vengeance than it is to ask ourselves if, in similar circumstances, we might have done the same thing. Among the most important lessons to be learned from confronting radical evil is just how banal and widespread it really is. “Affirming common humanity,” as Minow notes, “does not mean turning the other cheek or forgetting what happened.”[262] But it does require that accountability mechanisms not dehumanize perpetrators. In essence, restorative justice seeks to affirm the humanity of those who have behaved inhumanely. It offers accountability, not vengeance.[263]
Both retributive and restorative justice envision reconciliation as a product of full accountability for wrongdoing.[264] But while retributive accountability involves proportionate punishment, restorative accountability demands an acknowledgment by offenders of their culpability and a willingness to make good. While retributive justice allows society to punish an offender as a means of achieving reconciliation, restorative justice requires society to include the offender in the process of reconciliation. Neither approach is perfect; society cannot create adequate punishments for all offenders, nor can it force all offenders to be contrite. Whether either strategy will actually bring about reconciliation will depend on the extent to which punishment, apology, and reparation can form the basis for a common future. “Simple justice is elusive,” concludes Shriver. In fact, “forgiveness thrives upon the tension between justice-as-punishment and justice-as-restora-tion.”[265] Retribution can play a restorative role, while “forgiveness can make . . . room for punishment while making wider room yet for the repair of damages and the renewal of relations among enemies.”[266]
For Émile Durkheim, punishment is a form of moral communication used to express condemnation and strengthen social solidarity.[267] Durkheim argues that the “true function [of punishment] is to maintain social cohesion intact.”[268] Crimes, according to Durkheim, are acts that violate a society’s fundamental moral code of sacred norms, thereby weakening those norms. Punishment plays a critical role in preventing the collapse of the moral order by limiting the “demoralizing” effects of crime.[269] Punishment also functions as a collective response that demonstrates and reaffirms the real force of the common moral order.[270] By punishing, a society expresses its shared moral outrage, strengthening and reinforcing the norms of social life.[271] Punitive rituals, by articulating shared sentiments, help to reflect and sustain a society’s moral values, thereby strengthening the bonds of community.[272] For Durkheim, sanctioning offenders is a way to communicate the continuing validity of the law, a “language” which “expresses the feeling inspired by the disapproved behaviour.”[273] Thus the two central premises of the “Durkheimian school” of thought are that punishment communicates social condemnation, and that it therefore plays a role in reaffirming, or even creating, social identity and/or social solidarity.[274]
In examining the Durkheimian school, three points are worth noting. First, to argue that punishment is a form of communication sheds no light on the message being conveyed. Different forms of punishment communicate different messages and produce different forms of social solidarity. Moreover, because other approaches to criminal justice rely on the communication of particular messages, like those of deterrence or moral reformation, the communicative paradigm is intertwined with, though conceptually distinct from, the theories discussed above.
Second, in describing the impact of punishment on the construction of social identity, Durkheimian theorists focus on why a society punishes, not why a society says it punishes. After all, in adopting purpose provisions for penal codes, legislatures do not express a desire to shore up the social order, but rather use the language of retribution, deterrence, rehabilitation, or restorative justice. Moreover, descriptions of the consequences of punishment for social control or identity do not explain why one should punish. David Garland may well be right that punishment is sometimes used to express and construct the dominant moral order.[275] But the idea that ruling elites employ punishment as a way to establish the legitimacy of their social vision is not an affirmative reason to punish.
Third, if punishment is imposed in order to express societal reprobation, then condemnation, not punitive treatment, is the central goal. Under this view, while a prisoner