Lessons to learn? Using the Inter-American Court of Human Rights’ jurisprudence on amnesties and pardons in the context of the Russo-Ukrainian War

Abstract

The standards of the Inter-American Court of Human Rights on amnesties and pardons in mass atrocity cases have been influential in Latin America and beyond. In turn, discussions about possible transitional justice mechanisms related to the Russo-Ukrainian war have involved issues of amnesty and pardon. However, the dicta of the Inter-American Court of Human Rights do not formally bind Ukraine and Russia. By connecting the two (semi-)peripheries of international law – namely, Latin America and Eastern Europe – the present article examines whether and to what extent the jurisprudence in question can shed light on legal and policy solutions for addressing the amnesty and pardon challenges posed by the Minsk agreements, domestic developments in Ukraine and Russia, and a potential future peace accord.

1. Introduction

Latin America¹ and Eastern Europe² lie on the (semi-)peripheries of international law. While it is not uncommon to compare a (semi-)periphery with the Global North³ or to compare problems within one continent,⁴ a comparative exercise between two (semi-)peripheries remote from each other is atypical.⁵ Yet, such an exercise transcends the structural biases of West-centrism and the parochialism of Global North versus Global South binaries, aiming to reinforce the often-neglected agency of Latin America and Eastern Europe, both of which have faced imperialism and impunity for mass atrocities in their past and present.⁶

To cut the Gordian knot of responsibility, Latin American and Eastern European states considered – or are contemplating – the transitional justice mechanisms of clemency or exemption measures, which prominently include amnesties and pardons. Whereas amnesties are usually rendered by the legislative power to bar or cancel the prosecution or trial of those who are allegedly responsible for crimes, pardons are discretionary acts generally granted by the executive power to benefit individuals already declared criminally responsible by a court.⁷ It is unclear whether international law prohibits amnesties/pardons owing to the absence of any treaty-based ban and the frequent use of these measures in state practice.⁸

In Eastern Europe, the latest mass atrocities are those committed during the Russo-Ukrainian war from 2014 onwards. It is very difficult to predict the outcome of this armed conflict. If Russia wins, Ukraine will probably cease to exist, as the Kremlin’s preposterous goal of ‘de-Nazification’ essentially means the destruction of the Ukrainian nation and state.⁹ If Ukraine succeeds in liberating the Russian-occupied territories, the question of amnesty/pardon will agitate the Ukrainian society: first, Ukrainian citizens who joined¹⁰ or collaborated with the Russian armed forces or worked in the occupying administrations, including the perpetrators of international crimes; second, members of the Ukrainian armed forces and other Ukrainian defenders who committed international crimes; and third, foreigners, notably Russian citizens, who committed international crimes. In any case, prisoners of war (POWs), routinely exchanged during the Russo-Ukrainian war, will have to be promptly released and repatriated after the active hostilities end.¹¹

For its part, Latin America in previous decades – mainly between the late 1960s and the early 2010s – was engulfed by serious violations of international human rights law (IHRL) and international humanitarian law (IHL), including those constitutive of international crimes. While amnesties/pardons were ubiquitous, the Inter-American Court of Human Rights (IACtHR) played a fundamental role in checking and controlling the compatibility of these measures with IHRL.

Against this background, the present article aims to contribute towards filling the above-mentioned gap in international law scholarship on justice for mass atrocities in two (semi-)peripheries. The main research question is to determine whether and to what extent the IACtHR’s amnesty/pardon jurisprudence can shed light on legal and policy solutions for addressing transitional justice issues, particularly those related to amnesties/pardons, in the context of the Russo-Ukrainian war.

This article has three main sections. Section 2 outlines justifications for and limitations of the comparative exercise between Latin America and Ukraine. Section 3 analyses the legal and political developments in Ukraine with emphasis on war-related amnesty/pardon initiatives. Section 4 deals with the IACtHR’s jurisprudential standards on amnesties/pardons and their potential applicability to Ukraine. Thus, this article seeks to provide an avant-garde approach to the object of study by examining the advisability of – or even the need for fostering – academic and practice-oriented dialogues to evaluate whether and to what extent legal innovations introduced in one region can be applied to another region, bearing in mind that the same pressing issue is present: justice for mass atrocities.

2. Justifications and limitations of the comparative exercise between Latin America and Ukraine

2.1 Justifications

There are good reasons that may justify the consideration of the IACtHR’s amnesty jurisprudence when examining whether and to what extent amnesties/pardons can – or should – be used in the context of the Russo-Ukrainian war.

First, mass atrocities in Latin America and Ukraine are, to an important extent, substantively similar. They involve serious IHL/IHRL violations, perpetrated on a large scale or in a systematic manner, affecting vast numbers of victims; and some of these violations amount to international crimes, with similar actus rei including murder, rape, torture, inhuman treatment, enforced disappearance, unlawful confinement, and persecution.¹² Several IACtHR cases have concerned amnesties/pardons for such atrocities committed during armed conflicts in Colombia,¹³ Guatemala,¹⁴ Peru,¹⁵ and El Salvador,¹⁶ as well as cases related to mass atrocities committed by Chile’s and Argentina’s military dictatorships, albeit not in armed conflict contexts.¹⁷ More specifically, some IACtHR cases related to El Salvador¹⁸ and Colombia¹⁹ addressed amnesties or similar measures, such as sentence reduction, which were discussed and adopted during (then) ongoing armed conflicts and as part of complex peace-making processes.

Second, the IACtHR’s amnesty/pardon jurisprudence has burgeoning authority beyond Latin America. This jurisprudence has contributed to shaping the practice of hybrid criminal tribunals²⁰ and the African Court on Human and Peoples’ Rights.²¹ More importantly for this article, the IACtHR’s jurisprudence on amnesties/pardons has influenced the European Court of Human Rights (ECtHR) and the International Criminal Court (ICC), both of which have jurisdiction over mass atrocities committed during the Russo-Ukrainian war within their respective mandates. Furthermore, the International Court of Justice (ICJ), which has handled two contentious cases between Ukraine and Russia,²² has increasingly invoked the IACtHR’s jurisprudence, especially that on reparations.²³

In addition, the ECtHR has growingly engaged with war-related amnesties/pardons. In some cases concerning the Yugoslav Wars and the Armenian-Azerbaijani armed conflict, the ECtHR and/or applicants have invoked the IACtHR’s jurisprudence.²⁴ It remains to be seen how the ECtHR, in a new inter-state case between Ireland and the United Kingdom, will deal with the Northern Ireland Troubles (Legacy and Reconciliation) Act 2023, which, if certain conditions are met, guarantees immunity from prosecution for Troubles-related offences.²⁵

The ICC has also referred to the IACtHR’s amnesty jurisprudence when examining admissibility issues under the principle of complementarity.²⁶ Furthermore, the ICC has relied extensively on the IACtHR’s jurisprudence when construing its own legal framework on reparations for victims of mass atrocities.²⁷

In turn, the IACtHR has increasingly invoked the ECtHR’s amnesty jurisprudence and the case law of international/hybrid criminal courts that generally ban or restrict amnesties/pardons for gross IHRL violations.²⁸ This process of judicial cross-fertilization where international institutions influence one another in interactive dialogues puts the IACtHR’s jurisprudential developments closer to Ukraine. Despite their different mandates, international courts tend to reject amnesties/pardons granted to those accused or convicted of mass atrocities. Not only has the IACtHR laid the foundations of such a trend, but it has also taken the lead on amnesty case law, since supranational and national courts outside Latin America have cited the IACtHR’s jurisprudence.²⁹ It can hence be argued that diverse supranational courts speak mutatis mutandis with the same vocabulary about amnesties/pardons in mass atrocity cases.

Third, notwithstanding state practice that includes serious IHRL/IHL violations or international crimes as part of the scope of amnesties/pardons, there are also opposing national trends.³⁰ The latter coincides with the IACtHR’s traditional position on the non-applicability of amnesties to the said atrocities. These trends are conspicuous in European states recovering from hostilities: Bosnia-Herzegovina,³¹ Croatia,³² and Georgia.³³ Although Russia’s amnesty laws concerning the Russo-Chechen wars were inapplicable to those responsible for certain serious crimes, such as genocide, terrorism, intentional homicide, rape, and kidnapping, those responsible for war crimes and other offences proscribed under Russia’s criminal legislation were granted amnesties.³⁴ Paradoxically, during the travaux préparatoires of what would later be Article 6 of Additional Protocol II to the Geneva Conventions, the Soviet Union stated that amnesties could not apply to crimes against humanity and war crimes.³⁵

Based on, inter alia, these examples of national and international practice, the International Committee of the Red Cross found the existence of the following customary IHL norm:

At the end of hostilities, the authorities in power must endeavour to grant the broadest possible amnesty to persons who have participated in a non-international armed conflict, or those deprived of their liberty for reasons related to the armed conflict, with the exception of persons suspected of, accused of or sentenced for war crimes.³⁶

Another customary international law exception, albeit mentioned only in passing, was crimes against humanity.³⁷ This exclusion of war crimes and crimes against humanity from the scope of amnesties overall coincides with the IACtHR’s traditional amnesty jurisprudence. Finally, the growing consensus in international law whereby amnesties and other exemption measures for international crimes are unacceptable is implicitly reinforced by multilateral treaties – to which Ukraine, Russia, and many Latin American states are parties – requiring prosecution of these offences³⁸ without applying statutory limitations.³⁹

2.2 Limitations

Since there are no one-size-fits-all solutions in transitional justice,⁴⁰ the potential use of the IACtHR’s jurisprudence for guidance in the context of the Russo-Ukrainian war should not be done automatically.⁴¹ This jurisprudence must be instead adapted to the pertinent factual, legal, and political circumstances. The principal limitations are as follows.

First, the IACtHR’s amnesty jurisprudence concerns non-international armed conflicts (NIACs) (Colombia, Peru, El Salvador, and Guatemala) or dictatorial regimes in ‘peace’ times seeking impunity for their own members through self-amnesties or blanket amnesties (e.g., Brazil and Chile). However, El Salvador’s NIAC occurred in the 1980s’ Central American context, which saw some level of internationalization of armed conflicts, culminating in the ICJ’s landmark Nicaragua case. El Salvador unsuccessfully tried to intervene in this case, claiming to have suffered from Nicaragua’s aggression.⁴² In any event, amnesties/pardons triggering cases at the IACtHR were normally granted outside war contexts (e.g., Chile and Uruguay) or when hostilities substantially decreased (e.g., Peru and Colombia). These contextual and factual features contrast with the Russo-Ukrainian war, an international armed conflict (IAC) that started with the Russian occupation of the Crimean Peninsula and some Donbas territories in 2014,⁴³ although the Kremlin denied its military involvement on the Crimean Peninsula (briefly) and in Donbas (until 2022).⁴⁴ As examined later, amnesties/pardons in Ukraine have been discussed or adopted by democratic authorities as part of cease-fire negotiations and in a scenario of protracted hostilities.⁴⁵

Second, in terms of the approaches to atrocity-related amnesties/pardons, the IACtHR and the ECtHR have been at odds. Under its control of conventionality doctrine,⁴⁶ the former has adopted an interventionist or demanding approach to amnesties/pardons by requiring that these measures be fully compatible with the American Convention on Human Rights (ACHR) and other inter-American human rights treaties, the IACtHR’s jurisprudence, and general IHRL.⁴⁷ Such an approach was especially rigid in the IACtHR’s early case law leading to the annulment of amnesty laws in Barrios Altos v. Peru, but it has become more flexible in recent years, as examined later.

Pursuant to its margin of appreciation doctrine,⁴⁸ the ECtHR has conversely adopted an approach to amnesties/pardons whereby it recognizes that states have a broader degree of discretion than that acknowledged by the IACtHR.⁴⁹ But the ECtHR seems to have been more demanding to guarantee that states’ actions are consistent with the Convention for the Protection of Human Rights and Fundamental Freedoms (ECHR) in amnesty/pardon cases, arguably conducting a sort of ‘implicit’ review of these national legislations in recent years.⁵⁰ In any event, unlike the IACtHR, the ECtHR has not yet directly reviewed amnesties/pardons, let alone nullifying these measures or declaring them effectless. Authors have recognized that: the majority of past amnesties outside Latin America have remained in effect; Latin American amnesties generally differ from those granted in other parts of the world, as the former were predominantly self-amnesties enacted by dictatorial regimes; Latin American societies have comparatively better transitioned into peaceful and democratic contexts than states in other regions; and Latin American dynamics are absent in other regions.⁵¹ For example, the ECtHR has been less demanding than the IACtHR when it comes to amnesties/pardons.⁵²

Third, beyond Latin America, there is no clearly conclusive evidence of a customary international law norm forbidding amnesties in cases of international crimes or gross IHRL/IHL violations.⁵³ Indeed, the IACtHR’s methodology of cherry-picking relevant domestic practices has been subject to criticism.⁵⁴ Furthermore, there is some opposite practice in Europe. For instance, Azerbaijan pardoned its soldier who had decapitated an Armenian officer during a NATO training, which was an atrocity committed out of ethnic hatred and related to the Armenian-Azerbaijani armed conflict.⁵⁵ Nevertheless, the ECtHR found Azerbaijan responsible for ECHR violations.⁵⁶ Other examples are the clauses on exemption measures for mass atrocities under the Minsk agreements, accepted by Ukraine due to Russian pressure.⁵⁷ In this regard, the UN Special Rapporteur on extrajudicial, summary, or arbitrary executions expressed concern that a blanket amnesty under the Minsk agreements could foster impunity for international crimes and serious human rights violations.⁵⁸ These political instruments are discussed in Sections 3.2 and 4.2.

Regarding national and international practice (including the examples just presented), the above-quoted customary IHL norm on the exclusion of war crimes from amnesties is, according to the International Committee of the Red Cross, applicable to NIACs but it does not extend to international(ized) armed conflicts. Indeed, the IACtHR has yet to decide on amnesty laws related to cases stemming from IACs.

3. War-related pardon and amnesty initiatives in Ukraine

In states experiencing armed conflicts, such as Ukraine, the demand for retribution plays a prominent role.⁵⁹ In Ukraine, the wheels of criminal justice do not stop even during the (more) active phase of hostilities. Yet, the society would return to normal life sooner or later and it is imperative to discuss various models of transitional justice before the Russo-Ukrainian war ends.⁶⁰ A key component of such models is the question of amnesty/pardon.

Amnesty, unlike pardon, is defined in Ukraine’s legislation. Amnesty means ‘a full or partial release from serving a sentence for persons’ found guilty of a crime, or for those whose criminal cases have been adjudicated but whose corresponding judgments have not yet entered into force.⁶¹ While amnesty law is adopted by the Ukrainian Parliament for a specific category of persons or a specific person,⁶² pardon is granted by the Ukrainian President for a specific person.⁶³ Pardon can only be granted to convicted persons.⁶⁴

3.1 Pardon

In Ukraine’s legal framework, pardon for perpetrators of grave and particularly grave crimes, i.e., including war crimes, genocide, and the crime of aggression,⁶⁵ may be granted ‘in exceptional cases and subject to extraordinary circumstances’,⁶⁶ a legal provision criticized for its vagueness by the ECtHR.⁶⁷ If a person is sentenced to life imprisonment, pardon replaces it with a minimum of 25 years’ imprisonment after the convicted person has served at least 15 years (or 20 years, before the amendment of 22 December 2023) in prison.⁶⁸

For Pysmenskyi, the need to exchange POWs falls within the ambit of the above-mentioned exceptionality.⁶⁹ Is this interpretation correct?

First, it wrongly implies that POWs can be convicted of crimes that are not deemed international. This mistake used to be ubiquitous in the Ukrainian criminal justice system. Ukraine treated members of the Russian army and Russian proxies in Donbas as ordinary criminals, notably terrorists.⁷⁰ Even at the beginning of Russia’s full-scale invasion of Ukraine, Russian combatants were arrested and charged with illegal crossing of the border.⁷¹ This practice was declared to be abandoned,⁷² and rightly so, because POWs enjoy combatant immunity and their internment is not a punitive measure.⁷³ In a recent verdict, the Novozavodskyi District Court of Chernihiv held that the accused, a Russian serviceman, ‘cannot bear individual criminal responsibility for the fact of participation in an armed conflict’.⁷⁴

The second implication of Pysmenskyi’s interpretation of Ukraine’s criminal law is that pardon can be granted to POWs convicted of international crimes. Indeed, according to the IHL principle of assimilation, POWs enjoy the same right to apply for pardon as members of the detaining power’s army.⁷⁵ Yet, pardon to all such convicts, whether from the detaining power’s army or adversary army, is questionable due to the gravity of international crimes.⁷⁶

Finally, pardon has been claimed to be a useful political tool to achieve a national or inter-state compromise and cessation of armed conflict.⁷⁷ However, the practicality of pardon as an individual act at the discretion of one person – the president – is debatable when it comes to mass atrocities. Moreover, should we agree that the concept of pardon must include elements of remorse and compensation for the damage inflicted by the convicted person,⁷⁸ other mechanisms, such as amnesty, appear to be more suitable during the war and thereafter.

3.2 Amnesty

3.2.1 The draft law of 16 September 2014

As a measure to implement the first Minsk agreement, Ukraine had to ‘[e]nact a law prohibiting the prosecution and punishment of persons in connection with the events that took place in certain areas of [Donbas]’.⁷⁹ On 16 September 2014, 287 parliamentarians voted in favour of such a draft law.⁸⁰ Two categories of persons were to be exempted from criminal responsibility if, within specified time limits, they informed the relevant pre-trial investigation body that they did not hold hostages, keep weapons, occupy the premises, or block the work, of state authorities. These categories comprised ‘members of armed groups or persons involved in the activities of such groups’ and/or persons who ‘participated in the activities of self-proclaimed bodies in the Donetsk and Luhansk provinces or opposed the anti-terrorist operation’.⁸¹ The bill was expressis verbis inapplicable to perpetrators of certain crimes, particularly terrorist acts, genocide, murder, human trafficking, rape, and sexual violence, and those involved in the downing of MH17 and/or obstructing its investigation.⁸² The bill did not come into force, as the President did not sign it. The main reason was that the law on the local self-government in Donbas, which ruled out prosecution and criminal punishment of ‘participants of events in the Donetsk and Luhansk provinces’ and set the conditions for local elections,⁸³ was not complied with by the Russian proxies there.⁸⁴

The stillborn bill contained a legion of procedural, conceptual, and other deficiencies. First, Ukraine could not control its enforcement: Russian armed forces and their proxies held sway over prosecutors and judges in Donbas,⁸⁵ whereas officials from other regions of Ukraine, if seconded to Donbas, would work in danger.⁸⁶ Second, anyone could submit the required declaration to the relevant pre-trial investigation body, which would then be obliged to exempt him/her from criminal responsibility, without any procedure to verify the declarant’s honesty.⁸⁷ Third, the draft law gave rise to legal uncertainty as to its scope ratione loci due to the constantly fluctuating boundaries of the anti-terrorist operation.⁸⁸ Fourthly, the draft law failed to conceptually distinguish between national and international crimes and enshrined amnesty for some grave offences, notably war crimes.⁸⁹ Fifthly, there was no distinction between war-related and other crimes.⁹⁰ Sixthly, the draft law did not distinguish between Ukrainian citizens, who could participate in a national reconciliation process, and foreigners.⁹¹ Finally, the draft law failed to amend – and hence contradicted – the Ukrainian Criminal Code, which shall be the only law determining ‘[t]he criminal illegality of an act, as well as its … criminal law consequences’.⁹² In light of the foregoing, what was called amnesty essentially amounted to pseudo-amnesty.⁹³

3.2.2 The draft law of 19 March 2015

As a measure to implement the second Minsk agreement, unanimously endorsed by the UN Security Council,⁹⁴ Ukraine had to ‘[e]nsure pardon and amnesty by enacting the law prohibiting the prosecution and punishment of persons in connection with the events that took place in certain areas of [Donbas]’.⁹⁵ This unconditional amnesty/pardon clause, which might be seen as a necessary evil to achieve peace, could negatively affect the functioning of Ukraine as a state.⁹⁶ Indeed, the Kremlin viewed the second Minsk agreement as a means to erode Ukraine’s sovereignty.⁹⁷ Moreover, the formulation was problematic, because pardon, unlike amnesty, shall be granted by a presidential decree rather than a law. Yet, a new amnesty bill was registered for consideration at the Parliament on 19 March 2015.⁹⁸

This bill applied exclusively to Ukrainian citizens who (allegedly) committed an act that (potentially) fell within the definition of a crime from the specified list, provided that the act did not result in the victim’s death, medium bodily injury, or grievous bodily injury.⁹⁹ The list in question did not include certain grave and particularly grave crimes, such as terrorist acts, genocide, war crimes, and the crime of aggression.¹⁰⁰ Moreover, the new requirement for individuals to declare their ‘participation in the events’ in Donbas¹⁰¹ lacked a procedure to verify the declarant’s honesty. The bill was eventually withdrawn, since it failed to address all the deficiencies of the earlier legislative attempt to introduce amnesties. Further efforts to convert a similar bill to legislation were unsuccessful.¹⁰²

3.2.3 The draft law of 25 April 2016

On 25 April 2016, a draft amnesty law specifically targeting members of the Ukrainian army and other lawfully created armed groups participating in the anti-terrorist operation was registered for consideration at the Ukrainian Parliament.¹⁰³ The need for amnesty was justified primarily by the belief that hostilities placed a tremendous psychological burden on persons committing crimes.¹⁰⁴ The scope of the bill’s applicability did not cover certain crimes, including terrorist acts, genocide, intentional homicide, torture, rape, violence against civilians, and the use of weapons of mass destruction.¹⁰⁵ Yet, a person maltreating POWs could, as a rule, receive amnesty. The bill was criticized for such loopholes¹⁰⁶ and ultimately withdrawn.

3.2.4 Amnesty as a part of transitional justice initiatives

In 2019–2020, a group of Ukrainian officials, together with the civil society and academics, designed a non-public transitional justice roadmap. It is reported to have stipulated that amnesties/pardons could not be granted to persons, including members of the Ukrainian armed forces, who committed the gravest crimes, particularly war crimes and crimes against humanity.¹⁰⁷ It also allegedly included provisions for accountability, notably lustration, targeting leaders and aiders of the occupying administrations, as their activities infringed on Ukraine’s sovereignty and territorial integrity.¹⁰⁸ Although a detailed list of professions was deliberately omitted (as it is more important what each person actually did), the roadmap’s drafters reassured that 95 per cent of persons working in the Russian-occupied areas of Ukraine would not be adversely affected.¹⁰⁹ Indeed, some Ukrainian experts believed that entrepreneurs and people who got paid from public funds, such as teachers, medical doctors, utility workers, and even the so-called ‘officials’, should be amnestied.¹¹⁰ The Ukrainian President was expected to approve the roadmap via a decree and request the government to prepare an implementation plan, but neither transpired.¹¹¹

Meanwhile, the legislators began to discuss the bill ‘On the Principles of State Policy of the Transition Period’ in August 2021.¹¹² Article 9 of this bill stipulated, inter alia, that neither special exemption from criminal responsibility, nor amnesty shall be granted to persons in the Russian occupying forces and administrations who committed certain crimes ‘in connection with’ the occupation of Ukraine.¹¹³ The list of these crimes contained war crimes, crimes against humanity, genocide, torture, national security crimes, and, potentially, the crime of aggression, etc.¹¹⁴

Further details were supposed to be established in a separate law,¹¹⁵ which made some Ukrainian civil society organizations express concern that the bill undermined the principle of legal certainty and resembled more of a political declaration.¹¹⁶ This situation also precluded the Venice Commission from assessing the compatibility of the entire legal framework with international legal standards.¹¹⁷ In addition, the phrases ‘in connection with’ were criticized for their lack of clarity,¹¹⁸ although such wording is pervasive in transitional justice, as exemplified by the Colombian peace agreement.¹¹⁹ The Venice Commission was also puzzled by the non-applicability of Article 9 to crimes committed outside the occupied territories of Ukraine.¹²⁰ By inserting the catalogue of perpetrators of national security crimes, the bill not only controversially narrowed the requirements of the law ‘On the Application of Amnesty in Ukraine’¹²¹ but also breached the principle of non-discrimination.¹²² Furthermore, the differentiated treatment of offenders, which was not envisaged by the Minsk agreements, сould be detrimental to the reconciliation process.¹²³ Finally, the Kremlin threatened that the adoption of the bill would amount to Ukraine’s withdrawal from the Minsk agreements.¹²⁴ Instead, Russia insisted on blanket amnesties.¹²⁵ On 25 January 2022, the draft law was withdrawn.

3.3 Legal developments after 24 February 2022

In response to Russia’s full-scale invasion of Ukraine, the Ukrainian Parliament promptly criminalized collaboration with and aiding the aggressor state, both broadly defined.¹²⁶ Thus, the legislators moved away from the more reconciliatory settlement options.¹²⁷

Thousands of international crimes were allegedly committed during the initial months of the full-scale invasion.¹²⁸ At the same time, the number of POWs grew exponentially. The necessity to exchange these persons without undermining criminal justice processes loomed large.

According to the newly introduced Article 84¹ of the Ukrainian Criminal Code, the convicted person shall be released from serving a sentence upon his/her written consent, if the competent state body has decided to exchange him/her as a POW.¹²⁹ Although Article 84¹ is placed immediately before Articles 85–87 on amnesty/pardon, the new provision has a military rationale. For example, the Shevchenkivskyi District Court of Kyiv gave the green light to the exchange of a Russian artillerist, who had been convicted of a war crime and sentenced to 11 years and six months in prison.¹³⁰

A similar procedure was introduced for suspects and accused persons.¹³¹ After the exchange of POWs, the criminal trial in Ukraine can continue even if the person accused of certain crimes, including the crime of aggression, war crimes, and genocide, is absent (in absentia).¹³²

Yurii Bielousov, Head of the Department for Countering Crimes Committed During Armed Conflict of the Prosecutor General’s Office of Ukraine, reassured that POWs who committed a ‘grave war crime’, such as murder or rape, were not planned to be transferred to Russia before they serve their sentences.¹³³ To a certain extent, this viewpoint reflects some popular opinions on amnesty for war crimes without fatal consequences, as demonstrated in Section 3.4. As a matter of law, however, Ukraine’s legislation provides no gravity-based typology of war crimes. As a matter of fact, the assurance in question can be trumped by a military desideratum.¹³⁴ For example, a Russian pilot, who had been exchanged for a few Ukrainian pilots,¹³⁵ was convicted in absentia of the war crime of murder.¹³⁶

3.4 Public opinion polls

Public opinion research in Ukraine has often focused on the topic of war-related amnesty over the last decade. One of the latest examples is the computer-assisted telephone interviewing survey conducted by the Sociological Group ‘Rating’ in October 2023. The respondents’ support of amnesty varied depending on the category of collaborators in the Russian-occupied territories of Ukraine: from 58 per cent for medical doctors, teachers, and social workers to 1 per cent for members of local political parties (see Table 1).¹³⁷ No conditions for such amnesties were put forward, which is a drawback.

Table 1. Amnesty for various categories of collaborators

As regards accountability for war crimes, the respondents supported amnesty if certain conditions were met (see Table 2). For example, 42 per cent would agree to amnesty if war crimes had no fatal consequences, and 40 per cent would favour it if the perpetrator engaged in socially useful work to restore Ukraine.¹³⁸ It is worthy of note that the survey did not differentiate between Ukrainian citizens and foreigners, which is a commendable approach.

Table 2. Amnesty for war criminals

Nevertheless, sociological data about an ongoing armed conflict, such as the Russo-Ukrainian war, should be taken with a grain of salt, as they arguably lack reliability.¹³⁹ In terms of representativeness, millions of Ukrainian refugees across the world and millions of Ukrainians residing in the Russian-occupied territories are not heard in public opinion research,¹⁴⁰ including that discussed above. Furthermore, some people may be fearful of telling the truth, responding to a politically sensitive question, or taking part in a wartime survey in the first place.¹⁴¹ After all, the rights to freedom of expression and respect for one’s correspondence are restricted in Ukraine, as stated in its derogations from certain obligations under the ECHR and the International Covenant on Civil and Political Rights; notably, it is prohibited to produce and disseminate information ‘that may destabilize the situation’.¹⁴²

4. The IACtHR’s jurisprudential standards on amnesties/pardons and potential applicability thereof to Ukraine

4.1 The IACtHR’s jurisprudential standards on amnesties/pardons

This subsection provides an analytical summary of the main trends in the IACtHR’s amnesty/pardon jurisprudence.¹⁴³ In Barrios Altos v. Peru, the IACtHR became the first supranational court that nullified national legislation, namely self-amnesty laws on gross human rights violations that benefited members of the death squad group Colina during the term of Peru’s ex-President Fujimori.¹⁴⁴ According to the IACtHR, this finding had general effects and, thus, was applicable to not only the case at hand but also other proceedings or cases affected by those Peruvian self-amnesty laws.¹⁴⁵ Hence, the IACtHR in La Cantuta v. Peru applied the Barrios Altos jurisprudential standard concerning the same self-amnesty laws.¹⁴⁶

This initial trend of the IACtHR’s zero tolerance for amnesty laws, which concerned blanket amnesties or self-amnesties or similar measures that sought to grant impunity by precluding criminal proceedings against or freeing those accused or convicted of gross human rights violations constitutive of international crimes, continued for several years. Thus, the IACtHR in Almonacid-Arellano et al. v. Chile determined that a self-amnesty law concerning systemic state torture implemented during the Pinochet dictatorship had no legal effect.¹⁴⁷ This judgment is also pivotal because it was the first time the IACtHR defined its control of conventionality doctrine, which contrasts with the ECtHR’s margin of appreciation doctrine.

In Gomes-Lund et al. v. Brazil, the IACtHR declared that an amnesty law issued during the Brazilian military dictatorship had no legal effect.¹⁴⁸ In Gelman v. Uruguay, the IACtHR established that national legislation on amnesties and statutes of limitations adopted during Uruguay’s dictatorial regime breached the state’s IHRL obligations, even if this legislation was ratified in a democratic regime through referenda and upheld by the judiciary during democracy.¹⁴⁹ Scholars such as Gargarella have criticized the IACtHR’s complete disregard for the democratic validation of the said legislation in the Uruguayan case.¹⁵⁰ In these cases, which arguably correspond to a first jurisprudential period (from 2001 to approximately 2011), the IACtHR adopted a rigid position focused solely on assessing the compatibility of national legislation on amnesties and similar measures with the state’s obligations under IHRL, particularly the ACHR, without (explicitly) taking into account or (substantially) examining other considerations. However, even during this jurisprudential period, some degree of progressive flexibility can be observed: while the IACtHR in Barrios Altos v. Peru declared national legislation null and void, its focus in subsequent cases shifted primarily to the lack of effect or inapplicability of such legislation.

As the IACtHR increasingly appraised national legislation on amnesties and similar measures related to (then) ongoing NIAC (Colombia) and periods postdating NIACs (Guatemala and El Salvador), along with the respective peace-making processes, the IACtHR started looking at other factors when assessing the compatibility of these national measures with the ACHR and international law as a whole. This contrasts with the judicial approach in the above-mentioned Peruvian cases which, unlike other cases of the first jurisprudential period unrelated to NIACs, occurred within the broader context of a NIAC: the Peruvian one. In the Peruvian cases, the IACtHR mostly zoomed in on the self-amnesty laws and similar measures but without further contextual consideration. Thus, it is possible to identify a second jurisprudential period, which became noticeable around 2012 but arguably began in 2007. This period can be characterized by the IACtHR’s attempts to nuance its strict approach and, thus, balance various competing interests in ongoing wars, post-war contexts, and/or post-dictatorial scenarios.

In Rochela Massacre v. Colombia (2007), the IACtHR did not decide whether, in the context of the (then) ongoing NIAC between the Colombian government and the Revolutionary Armed Forces of Colombia–People’s Army (FARC), the Justice and Peace Law was an amnesty because such a legislative piece was still at an early stage.¹⁵¹ The judges rather indicated guiding principles to implement the said law.¹⁵² In turn, the ICC seemingly tolerated certain amnesty-related measures under the Justice and Peace Law during the ICC’s preliminary examination concerning Colombia.¹⁵³ In Tiu Tojín v. Guatemala and García Lucero et al. v. Chile, the IACtHR did not annul amnesty laws but, to determine state responsibility, instead examined whether these laws were actually applied.¹⁵⁴

In El Mozote v. El Salvador (2012), the IACtHR’s then President García-Sayán differentiated between amnesties adopted to end NIACs and those issued by dictatorial or autocratic regimes, reflecting certain evolution in the IACtHR’s jurisprudence. It was thus recognized that amnesties in such contexts can lead to diverse outcomes in which two opposing forces interact: criminal accountability for gross human rights violations vis-à-vis a negotiated solution to an NIAC and reconciliation.¹⁵⁵ In turn, solutions to the dilemmas brought about by such interactions are context-specific rather than universal, but some guidelines must be borne in mind.¹⁵⁶ President García-Sayán’s additional remarks highlight several influential standards: (i) NIACs, characterized by large numbers of offenders and victims, constitute exceptional situations that typically necessitate ‘exceptional mechanisms of response’; (ii) methods for assessing tensions between opposing forces are essential; (iii) consideration of both judicial and non-judicial elements is crucial in pursuit of justice, truth, and reparation; and (iv) addressing the demands arising from mass atrocities, responding to conflict consequences, and seeking long-term peace requires concurrent measures from both states and society as a whole to ensure the simultaneous fulfilment of victims’ rights to justice, truth, and reparation.¹⁵⁷ In conclusion, President García-Sayán noted that there is a need to:

devise ways to process those accused of committing serious crimes … for example … routes towards alternative or suspended sentences … but, without losing sight of the fact that this may vary substantially according to both the degree of responsibility for serious crimes and the extent to which responsibility is acknowledged and information is provided about what happened … Reduction of sentences, alternative punishments, direct reparation from the perpetrator to the victim, and public acknowledgment of responsibility are other ways that can be considered.¹⁵⁸

In 2018, while monitoring the implementation of its judgments and concerning the 2017 pardon of ex-President Fujimori, the IACtHR deferred the matter to Peru’s jurisdiction, subject to specific requirements and guidelines.¹⁵⁹ The Peruvian Supreme Court followed them.¹⁶⁰ Yet, the Peruvian Constitutional Tribunal decided to execute the pardon of Fujimori and ordered his release in December 2023 despite the IACtHR’s request not to do so.¹⁶¹ Finally, in Aldea Chichupac/Molina Theissen v. Guatemala, the IACtHR requested that the respondent state refrain from passing Bill 5377, which aimed to reform the 1996 National Reconciliation Law by granting amnesty for all serious IHRL/IHL violations committed during Guatemala’s NIAC.¹⁶² Although the IACtHR invoked the need to protect the right of victims to access justice, it did not nullify Bill 5377. The IACtHR applied its control of conventionality doctrine in a nuanced manner, acknowledging the importance of initiatives to address the effects of armed conflicts or violent situations, which involve complex processes.¹⁶³ Moreover, Judge Vio Grossi’s partial dissent only urged Guatemala to consider the existence of potential violations of international obligations if and/or when deciding to pass Bill 5377.¹⁶⁴

In light of the jurisprudence summarized above, two potentially conflicting factors chewed over by the IACtHR can be identified. The first factor is the state’s compliance with its IHRL obligations, particularly those concerning the rights of victims of mass atrocities. Contesse rightly points out that the IACtHR has continuously held that amnesties (and similar measures) for gross abuses are inconsistent with inter-American human rights law, especially the obligations of states to investigate, prosecute, and punish.¹⁶⁵ In turn, Mallinder refers to the IACtHR’s ‘progressive’ interpretations of the state duties made in order to guarantee rights, harmonize national law with international law, and protect the right of victims to justice, going beyond the approaches adopted by other supranational courts.¹⁶⁶ For Binder, the IACtHR strives for an effective implementation of the IHRL obligations of states¹⁶⁷ and, hence, its jurisprudence ‘considerably extended the standard of review … when examining whether a violation of the respective state’s human rights obligations had occurred’.¹⁶⁸ Yet, as Mallinder observed, the IACtHR has growingly attempted to strike a balance between the IHRL obligations of states and other considerations by not finding that a state violated its obligations if amnesties/pardons were actually not applied and thus did not prevent the investigation, prosecution, or sanction of mass atrocities.¹⁶⁹

The second factor is a rising recognition of the potential impact of amnesties/pardons on transitions to peace, reconciliation, democracy, and/or the rule of law. The IACtHR has growingly considered the impact of amnesties/pardons on societies in flux. The IACtHR may, according to Mallinder, ‘be willing to distinguish between amnesties enacted during or after dictatorship, and amnesties … to end violent conflict’.¹⁷⁰ This suggests the emergence of a flexible judicial approach to evaluating state measures adopted to strike a balance between, on the one hand, the need to prevent further atrocities and bring armed conflicts to an end, and, on the other hand, state obligations concerning victims of mass atrocities. Depending on the conditions of transition, amnesty can therefore play different roles.¹⁷¹ In turn, Binder argues that the IACtHR’s focus on accountability and effective human rights protection clears the way for domestic bodies ‘to implement human rights and the rule of law’ and bolsters democratic transition, consolidation, and, eventually, domestic self-determination in Latin American states.¹⁷² In a more critical stand, Gargarella points out that the IACtHR’s earlier case law (labelled herein as the first jurisprudential period) usually adopted a rigid view of human rights protection and related state obligations by neglecting or undermining the consideration of other factors.¹⁷³

In conclusion, the IACtHR’s approach to amnesty during the first jurisprudential period, which emphasized the inadmissibility of amnesties/pardons (as well as the scholarship supporting this approach), strictly prioritized full respect for IHRL, notably the rights of victims. In Latin America, this took place when impunity through amnesties/pardons, particularly self-amnesties or blanket amnesties, threatened justice in mass atrocity scenarios. Thus, supranational bodies like the IACtHR serve as mechanisms of last resort to oversee states that, under the guise of transitional justice or peace-making considerations, may either intentionally pursue impunity and breaches of IHRL obligations or inadvertently end up doing so at the expense of victims’ rights.

Saying no to impunity in the shape of amnesties/pardons may work well in post-war or post-dictatorial scenarios. Such a principled or normative-oriented approach might, however, be deemed excessively unyielding and even counter-productive in the context of ongoing armed conflicts, where there is a pressing need to balance divergent interests, such as negotiating peace, transitioning from scenarios of mass atrocities, and complying with IHRL obligations. States may also become alienated from co-operating with overly demanding supranational human rights bodies or, more radically, withdraw from them, to the detriment of the affected communities. Even if there is no withdrawal, for example, states may decide not to establish or fund reparation programmes out of frustration with the approach of a supranational body like the IACtHR that strictly opposes amnesties/pardons. These potential problems highlight the advantages of pragmatism in the IACtHR’s more recent approach to amnesty, unlike its early jurisprudence.

Under no circumstances, however, should victims’ legitimate claims for justice be ignored. A fundamental aspect of the evolution of the IACtHR’s amnesty/pardon jurisprudence is the extent to which retributive justice – namely, criminal prosecution and punishment, and the complete inadmissibility of amnesties/pardons – should give precedence to restorative justice measures, such as reparations for victims through national reparation programmes. Indeed, justice is a complex phenomenon comprising not only punitive but also other dimensions.

4.2 Potential applicability to Ukraine

This subsection examines the potential applicability of the IACtHR’s amnesty/pardon case law to the context of the Russo-Ukrainian war. In doing so, it stresses the importance of direct communication between Ukraine and Latin America, explores the challenges of paying heed to the views of Ukrainian victims, engages with the expert discussion of amnesties/pardons while considering the fundamental factor of Russia’s aggression, and hypothesizes about amnesty/pardon as part of a potential future peace accord.

4.2.1 Importance of direct communication

Ukrainian experts are open to incorporating the best amnesty practices from abroad, including Latin America, in drafting the necessary legislation.¹⁷⁴ Neither the ACHR, nor the IACtHR’s jurisprudence is a binding or directly applicable source of law in Ukraine or Russia. Yet, this formality can in no way preclude Ukrainian national authorities from considering the IACtHR’s judicial practice in the law-making process.

One may wonder whether the European system of human rights protection should be consulted first, because Ukraine is a member state of the Council of Europe. The Venice Commission’s opinions are deemed authoritative, albeit not legally binding, in Ukraine.¹⁷⁵ More importantly, Ukrainian courts are obliged to apply the ECHR and the ECtHR’s judgments as sources of law.¹⁷⁶ Nevertheless, the main problem is that the ECtHR, unlike the IACtHR, has yet to directly conduct a judicial review of amnesties/pardons in atrocity cases. This holds true despite the ECtHR’s quasi-review of a presidential pardon in Makuchyan and Minasyan v. Azerbaijan and Hungary.¹⁷⁷

The relevance of the IACtHR’s amnesty/pardon jurisprudence for Ukraine is reinforced by the fact that judicial developments in Strasbourg regarding amnesties/pardons for atrocities related to IACs or dictatorial regimes¹⁷⁸ are somewhat analogous to those in San José during the previously analysed second jurisprudential period. In their examination of the effects of amnesties/pardons, both courts have considered seemingly conflicting criteria, notably state compliance with IHRL obligations vis-à-vis the potential impact on the transition to peace, the rule of law, and (even) reconciliation.¹⁷⁹

As discussed in Section 2.1, the ECtHR has increasingly invoked the IACtHR’s amnesty/pardon case law. Thus, the latter’s jurisprudential standards can be indirectly or tacitly used in the Ukrainian criminal justice system through the former’s judicial pronouncements. Furthermore, the Venice Commission has recommended Ukraine to adhere to the IACtHR’s first jurisprudential standard, despite the absence of explicit references to the inter-American system of human rights protection.¹⁸⁰ That being said, direct communication between Ukraine and Latin America would be a better strategy to enhance Ukraine’s visibility in the Global South (and vice versa), build intellectual bridges, and eschew distortions by Western, Russian, or other intermediaries.¹⁸¹

4.2.2 Voices of Ukrainian victims

To avoid social and political destabilization, an amnesty law ought to be perceived as legitimate and just by the population.¹⁸² However, the Ukrainian Constitution does not allow referenda on a draft amnesty law.¹⁸³ To some extent, this resembles the IACtHR’s jurisprudence of not subjecting draft amnesty laws to referenda provided that the exemption measure excludes those suspected or accused of mass atrocities from the universe of amnesty/pardon beneficiaries. In Gelman v. Uruguay, the IACtHR held that a national referendum does not automatically legitimize amnesties under international law.¹⁸⁴ This implies that even the public backing of amnesties may be neglected.

At the same time, an amnesty law should take into account the voices of victims, notably their readiness to forgive and their desire to receive compensation and other reparation forms.¹⁸⁵ In this regard, the IACtHR’s jurisprudence on the rights of victims may provide a lot of food for thought. A consistent jurisprudential line in the IACtHR’s amnesty/pardon case law throughout decades has been precisely the importance given to the rights of victims to access to justice, truth, and reparations.¹⁸⁶ Even in the more recent IACtHR jurisprudence, victims’ rights are still a key factor that weighs heavily with the judges.¹⁸⁷

The process of lending an ear to victims or even the entire population of Ukraine¹⁸⁸ may be hampered by social disagreements, including diametrically opposed views on amnesty. Opinions may also change over time. The latest instructive lesson from Latin America comes from Colombia. The 2016 Colombian peace agreement setting forth a detailed legal framework on amnesty was rejected by a narrow margin in a referendum, which, however, did not prevent Colombia’s Congress from approving a revised version of the accord.¹⁸⁹ While the Inter-American Commission on Human Rights welcomed this peace agreement,¹⁹⁰ the IACtHR has not yet had the opportunity to assess the compatibility of the agreement’s amnesty provisions with IHRL.

Thus, the IACtHR in Gelman v. Uruguay and Colombia’s Congress ignored, respectively, the will of the Uruguayan and Colombian populations. In Colombia, this disregard led to upholding the amnesty clause, unlike the outcome in Uruguay. These contrasting scenarios underscore the absence of one-size-fits-all solutions, notably for Ukraine, and the difficulties in considering the sentiments of the affected communities.

4.2.3 Opinions of experts

Currently, there is no full agreement on amnesties within Ukrainian society and, more specifically, political elites.¹⁹¹ Yet, the dominant approach among academics and practitioners is reminiscent of the IACtHR’s original jurisprudential standard that amnesty shall not be admissible for perpetrators of international crimes.

Ukrainian experts have emphasized that there shall be no amnesty for perpetrators of violent crimes (e.g., torture, murder, and rape),¹⁹² grave crimes,¹⁹³ or grave and particularly grave crimes,¹⁹⁴ whereas amnesty for other offences can only be possible after thorough criminal investigations.¹⁹⁵ By citing the IACtHR’s cases of Velásquez-Rodríguez v. Honduras and Barrios Altos v. Peru in his scholarship, Gnatovskyy, the ECtHR’s judge in respect of Ukraine, stresses that international law prohibits amnesties for perpetrators of international crimes.¹⁹⁶ According to Gnatovskyy, this conclusion holds true for the Council of Europe system of human rights protection as well; however, he admits that the case of Marguš v. Croatia (concerning amnesties related to the Yugoslav Wars) suggests a more lenient approach to amnesties if they are necessary to achieve social reconciliation and provide reparations to victims.¹⁹⁷ For Kotelva, a judge in the Halytskyi District Court of Lviv, no amnesty should be granted at least to perpetrators of international crimes, as these offences pose a threat not only to Ukraine but also humankind.¹⁹⁸ No regional case law is cited in her op-ed, though.

It is agreed herein with the differentiated approach to amnesty in Ukraine. First, amnesty should not be admissible for perpetrators of international crimes, regardless of their nationality, due to the gravity of these offences and the increasing recognition of this prohibition under international law. Second, other offenders may be amnestied under certain terms. What conditions should be considered?

Again, some practices from Latin America are instructive. One of the examples given by Ukrainian experts is Colombia with its experience of the restoration of schools and hospitals by amnestied persons.¹⁹⁹ Given that around 3,800 educational facilities,²⁰⁰ 1,800 healthcare facilities,²⁰¹ and 250,000 residential buildings²⁰² have been destroyed or damaged in Ukraine, public works as elements of restorative justice would indeed be beneficial.

At the same time, Ukrainian experts have turned a blind eye to the more controversial aspects of Colombia’s amnesty formula. According to the Colombian peace agreement, all crimes listed in the ICC Statute are ineligible for amnesty/pardon, but perpetrators who acknowledge the truth and commit to non-repetition at the outset of special judicial proceedings in Colombia may receive alternative sentences, including reparative labour and some restrictions on liberty, rather than imprisonment.²⁰³ Although this creative framework may be deemed a form of conditional amnesty/pardon, alleged perpetrators are tried by a court and these trials may lead to a criminal punishment and imposition of a requirement to redress the harm inflicted, get involved with the affected communities, and work towards socio-economic development.²⁰⁴ When designing its transitional justice framework, Colombia seemingly drew inspiration from, but did not copy, that of South Africa,²⁰⁵ where amnesties were granted by the Truth and Reconciliation Commission in exchange for truth-telling about offences associated with a political objective, including grave crimes.²⁰⁶ For its part, South Africa took into account the previous amnesty experiences of Argentina, Chile, and El Salvador.²⁰⁷

It is submitted that neither Colombia’s alternative sentences for international crimes, nor South Africa’s conditional amnesty for grave crimes would be appropriate in Ukraine for at least two reasons. First, South Africa adopted its amnesty law before the establishment of the ICC; and the Colombian framework was, arguably, made possible due to the ICC’s focus on positive complementarity and refrain from opening an investigation.²⁰⁸ By way of comparison, the ICC has already opened an investigation within the Situation in Ukraine and issued six arrest warrants for Russia’s high-ranking officials, including President Putin. These arrest warrants cannot easily sink into oblivion.²⁰⁹

Second, in contrast to the Colombian context,²¹⁰ it is naïve to think that Ukraine’s promise of amnesty would serve as an incentive for members of the Russian army and Russian proxies to put down their arms and demobilize. After all, this objective has not been achieved by the Ukrainian project ‘I want to live’, which entices Russian combatants to surrender, become POWs, and receive treatment in compliance with IHL.²¹¹ Crucially, the Constitutional Court of South Africa differentiated between amnesties related to IACs and those related to violent political conflicts within a state, such as apartheid in South Africa.²¹² This contextual distinction calls for a discussion of Russia’s involvement in the Russo-Ukrainian war.

4.2.4 The Factor of Aggression

Transitional justice responses should be adapted to local contexts.²¹³ Russia’s aggression against Ukraine since 2014²¹⁴ is an important – if not the most important – external factor affecting the elaboration of such responses.²¹⁵

After Russia and its proxies defeated the Ukrainian armed forces in the battles of Ilovaisk and Debaltseve, Ukraine halted Russia’s further advance by signing the first and second Minsk agreements, respectively.²¹⁶ Both instruments essentially treated the war in Donbas as a NIAC, since Russia was not explicitly labelled a warring party. By calling, inter alia, for a cease-fire, blanket amnesties, national dialogue, and the socio-economic recovery of Donbas, the Minsk agreements seem closer to the recent developments in the IACtHR’s jurisprudence.

As scrutinized above, the IACtHR’s case law has become more flexible in terms of considering amnesties and similar measures as potential options in the context of (ongoing) NIACs, subject to two conditions. First, these amnesties and similar measures can contribute to peace-making, the consolidation of democracy and the rule of law, and/or reconciliation in societies transitioning after wars and/or dictatorial regimes. Second, these processes should be conducted with due respect for the rights of victims of mass atrocities, including reparations to redress the harm suffered.²¹⁷ As for the latter condition, the Minsk agreements were silent about victims’ rights. As regards the former condition, it is unfathomable how these agreements could have set the scene for peace in Ukraine if Russia’s participation in hostilities was not even acknowledged. Moreover, the objective of achieving peace through amnesties²¹⁸ can be realized when periods of mass atrocities, such as the ongoing Russo-Ukrainian war, are coming to an end.²¹⁹ Yet, the Kremlin aimed to preserve the status quo and cost-effectively impede Ukraine’s progress instead of bringing peace to the neighbouring state.²²⁰ Busol aptly observed that the Minsk agreements ‘were perceived as “peace settlements” anywhere but in Ukraine’.²²¹

The IACtHR’s recent jurisprudence, unlike the Minsk agreements, has not validated blanket amnesties expressis verbis. This can be explained by the fact that amnesty laws and similar measures in Latin America corresponded for many years to the so-called self-amnesty laws adopted by dictatorial regimes. Although amnesties related to armed conflict and democratic transition pursue different goals, the boundaries between these two categories are somewhat porous.²²² It is submitted that the topic of self-amnesty should be brought to the forefront of legal research into the Russo-Ukrainian war. The Kremlin’s insistence on blanket amnesties via the Minsk agreements aimed to shield members of the Russian army and Russian proxies in Donbas from criminal responsibility. Since Moscow officialdom has denied its military involvement in Donbas from 2014 until 24 February 2022,²²³ any criminal trials against Russians or pro-Russian separatists could establish the truth, undermine the Kremlin’s persistent disinformation narrative, and weaken the Russian dictatorship.²²⁴ Behind the smokescreen of Russian propaganda about a ‘NIAC’ in Ukraine, the Minsk agreements thus paved the way for self-amnesties bringing benefits to the Russian autocratic regime.

Following Russia’s full-scale invasion of Ukraine, the Kremlin no longer needed to hide the presence of its troops in Donbas, but the self-amnesty agenda remained. In July 2023, the Russian Parliament adopted a law imposing Russian criminal law and procedure upon the Donetsk, Kherson, Luhansk, and Zaporizhzhia provinces of Ukraine, effectively granting blanket amnesty for virtually all crimes committed in the interests of Russia and its proxies until 30 September 2022,²²⁵ the date when Russia attempted to illegally annex these territories. The new law has nothing to do with transitional justice, contravenes Russia’s national legislation and international obligations, and deprives victims of their human rights, particularly the right to an effective remedy.²²⁶ Emboldened by impunity, Klishas, a co-drafter of the law, envisaged an extension of its applicability to other Ukrainian regions after their conquest by Russia.²²⁷

Admittedly, there is a risk that Ukrainian military men and women may also enjoy impunity for the alleged commission of international crimes. The above-mentioned draft law of 25 April 2016 is a good example to the point. However, this draft law was not adopted, whereas Russia succeeded in foisting the Minsk agreements on Ukraine in 2014–2015 and enacted its self-amnesty legislation in 2023.

Russia’s full-scale invasion of Ukraine made the Minsk agreements obsolete. As a result, Ukraine’s new frame of reference is based on the ideas of state survival, deterrence, and retribution rather than national reconciliation. In the sense that Ukraine is less open to exemption measures, there is an increasing predilection for the IACtHR’s original strictness regarding the admissibility of amnesties/pardons for mass atrocities.

Latin American armed conflicts that have given rise to exemption measures have been non-international, in contrast to the Russian war against Ukraine, which constitutes an IAC. The differentiation between types of armed conflicts matters for the ratione personae scope of exemption measures. The release of Russian POWs in Ukraine under IHL applicable to IACs interplays with the idea of granting amnesties/pardons, but such a factual and legal factor has not been prominent in Latin America, including the IACtHR’s jurisprudence. Nevertheless, in amnesty/pardon-related cases, notably Barrios Altos v. Peru and La Cantuta v. Peru, where the IACtHR has engaged with the consideration of persons who to a greater or lesser extent participated in hostilities and were then detained as suspects, accused, or convicted of mass atrocities in NIACs, no weight has been given to the said status. Thus, the release of Russian POWs who are then tried in absentia in Ukraine is, arguably, consistent with the IACtHR’s jurisprudence.

The differentiation between types of armed conflicts also matters for the ratione materiae scope of exemption measures. Ukrainian experts have – in line with the IACtHR’s jurisprudential standards – excluded war crimes, crimes against humanity, and the crime of genocide. While the IACtHR’s jurisprudence offers limited assistance in addressing the crime of aggression, some domestic practices from Latin America may still provide valuable insights. For example, Ecuador’s Constitution prohibits amnesty for all four international crimes, including the crime of aggression.²²⁸

Moreover, as pointed out in Section 2.2, El Salvador’s NIAC was arguably internationalized to some extent and during a certain period. In the Nicaragua case, in which El Salvador tried to intervene, the ICJ found that the United States of America violated the customary international law prohibition on the use of force against Nicaragua.²²⁹ Overall, these factors help to explain why peacemaking efforts and transitional justice mechanisms, including amnesties, in the Salvadorian context involved the UN and some Central American presidents.²³⁰ If the IACtHR’s amnesty jurisprudence related to El Salvador is considered through the lens of internationalized armed conflict, this jurisprudence arguably becomes (even) more relevant to the context of the Russo-Ukrainian war.

4.2.5 Amnesty as part of a potential future peace accord?

The inter-state nature of the Russo-Ukrainian war suggests that amnesty/pardon issues may be components of a potential future peace agreement between Russia and Ukraine. As demonstrated by El Salvador²³¹ and Colombia,²³² peace negotiations can benefit from the good offices of third parties, such as other states or the UN Secretary-General. Thus, the 2024 Summit on Peace in Ukraine, held in Switzerland, marked the first step in the right direction, although Russia was not invited. One of the upcoming conferences is anticipated to address the issue of justice.

From the standpoint of Ukraine, accountability for ‘the most serious crimes under international law’ is a sine qua non for a future peace.²³³ Globally, this understanding is shared by 141 states, including the overwhelming majority of Latin American states that are parties to the ACHR and recognize the IACtHR’s contentious jurisdiction, as evidenced by their votes at the UN General Assembly in February 2023.²³⁴ Although the ban on amnesties/pardons is not explicitly mentioned in this resolution, the emphasis on ‘justice for all victims’ and ‘accountability … through investigations and prosecutions’²³⁵ leaves little-to-no room for the inclusion of exemption measures. Indeed, as observed by the IACtHR in Gelman v. Uruguay, amnesties for mass atrocities granted to secure peace often fail to achieve their intended goal.²³⁶

At this point, it should be remarked that Additional Protocol II to the Geneva Conventions urges the contracting parties ‘to grant the broadest possible amnesty’ to NIAC participants.²³⁷ In El Mozote v. El Salvador, the IACtHR held that this provision is not absolute, as IHL requires states ‘to investigate and prosecute war crimes’ allegedly committed in such conflicts.²³⁸ At the same time, IHL treaties concerning IACs do not address the amnesty issue while obligating the investigation and prosecution of war crimes.²³⁹ In the context of IACs, it is therefore reasonable to assume that the IACtHR would handle amnesty for mass atrocities as a peace-making tool in the same manner.

Nevertheless, the feasibility of Ukraine’s Peace Formula faces numerous practical obstacles, hinging on the military and political power of Russia, which is a far more influential actor on the world stage than non-state armed groups or even states involved in Latin American NIACs. Driven by its idée fixe to destroy the Ukrainian nation and state and bolstered by its lack of accountability, the Kremlin is likely to enhance the role of amnesties/pardons in potential future peace negotiations, as it did in the Minsk agreements. Would such a solution follow the IACtHR’s flexible approach to amnesties/pardons during the second jurisprudential period? It is doubtful that a balance could be struck between, on the one hand, the need to prevent further atrocities and bring the Russo-Ukrainian war to an end, and, on the other hand, state obligations concerning the rights of victims to justice, truth, and reparation. Russia’s understanding of international law questions the sovereignty of smaller states and is antagonistic to anthropocentrism.²⁴⁰ Moreover, impunity is entrenched in Russia’s approaches to IHL, as Moscow officialdom has consistently denied not only the facts of its wrongdoings in Ukraine and elsewhere but also the very existence of the Russo-Ukrainian war.²⁴¹ Indeed, the Kremlin’s imperialistic narrative of the ‘special military operation’ reduces Ukraine to the status of a fake state and treats it as a constituent part of Russia.²⁴² Since Russia’s aspiration to be recognized as a great power, along with its claims over Ukraine, predates President Putin’s dictatorial regime and is widely supported across the Russian populace,²⁴³ it is submitted that Russia’s de-imperialization is a vital condition for bringing lasting peace to Eastern Europe.

5. Conclusions

This article has augmented the incipient legal research on transitional justice connecting the two (semi-)peripheries of international law, namely Latin America and Eastern Europe. Amidst the active phase of hostilities in the Russo-Ukrainian war, it is difficult to reach definitive conclusions. In such circumstances, the demand of the Ukrainian government and Ukrainians for retributive justice naturally eclipses the capacity for reconciliation. Furthermore, the decision-making processes are largely motivated by military considerations. It is very complicated to predict how and when this long-lasting armed conflict will end and how the Ukrainian society will look thereafter. Yet, the avenues of (possible) amnesties/pardons need to be explored in advance.

More attention should be paid to the IACtHR’s case law on amnesties/pardons, as it helps to explain the legal and political intricacies of the Minsk agreements and the ensuing domestic developments in Ukraine and Russia. The Kremlin’s insistence on blanket amnesties via these controversial agreements aimed to shield members of the Russian army and Russian proxies in Donbas from criminal responsibility, contrary to the IACtHR’s jurisprudence on self-amnesties adopted by dictatorial regimes. At the same time, the Minsk agreements could hardly contribute to peace-making and reconciliation in Ukraine, as they failed to acknowledge Russia’s military involvement in the IAC and prioritize the rights of victims of mass atrocities. Therefore, the IACtHR’s recent jurisprudence, which is relatively flexible in terms of considering amnesties and similar measures as potential options in NIACs, is of limited relevance in this context.

Although various transitional justice initiatives have been put on the table in Ukraine since the beginning of the Russo-Ukrainian war, none have been validated. It remains to be seen what legal framework will be embraced in post-war Ukraine. In potential future peace negotiations, the Kremlin is likely to use its military and political power to enhance the role of amnesties/pardons, as it did in the Minsk agreements. Meanwhile, the prevailing approach among Ukrainian experts bears a resemblance to the IACtHR’s original jurisprudential standard: amnesties/pardons should be inadmissible for perpetrators of international crimes, regardless of their nationality. As for other offences, it appears reasonable to employ restorative justice. This outline should be the starting point for further debates. Without doubt, the Latin American fons et origo of the norms in question should always be publicly acknowledged, given that Ukraine strives to win hearts and minds in the Global South.