The European Court of Human Rights (ECtHR) vis-à-vis amnesties and pardons: factors concerning or affecting the degree of ECtHR’s deference to states

ABSTRACT States have adopted amnesties/pardons concerning serious human rights violations to transition from crises, dictatorships, or conflicts worldwide, including Europe. Although the ECtHR has yet to review amnesties/pardons directly, it has increasingly decided on the effects of amnesties/pardons on the rights of individuals. Thus, the main research question herein is to identify which factors may determine whether and to what extent the ECtHR defers to states regarding amnesties/pardons in cases of serious human rights violations, namely, factors concerning or affecting the degree of ECtHR’s deference to states in these cases. Based on ECtHR’s jurisprudence on amnesties/pardons, this article argues and finds that these factors generally are: the national process of adoption, application, and/or validation of amnesties/pardons; consideration of the margin of appreciation or discretion given to states; state compliance with international obligations on human rights; and potential impact on transitions to peace, reconciliation, democracy, and/or the rule of law. This article aims to fill a gap in scholarship by proposing an explicit, detailed, and analytical systematization of factors that, in light of ECtHR’s jurisprudence, may explain whether and to what extent the ECtHR defers to states in cases of amnesties/pardons concerning serious rights violations, including identification of ECtHR’s (emerging) jurisprudential trends.


Introduction
In societies that transitioned from crises, dictatorships, or armed conflicts to (more) peaceful or democratic scenarios, states have exercised their sovereignty and discretion by adopting exemption measures such as amnesty laws and presidential pardons concerning serious human rights violations that may constitute international crimes. 1 However, these measures have been directly or indirectly challenged at international courts and, therefore, they have been subject to international judicial scrutiny under international law obligations. Accordingly, the main research question of this article is to identify which factors may determine whether and to what extent the ECtHR defers to states concerning amnesties/pardons in cases of serious human rights violations, namely, factors concerning or affecting the degree of ECtHR's deference to states.
The ECtHR has not yet directly, explicitly, or stricto sensu conducted a judicial review of amnesties/pardons, contrasting with the Inter-American Court of Human Rights (IACtHR). 2 This may explain that there are only very few academic materials on amnesties/pardons at the ECtHR, 3 which contrasts with the robust scholarship on these measures at the IACtHR. 4 Nevertheless, the ECtHR has increasingly issued judgments that concern the effects of amnesties/pardons on human rights and the compatibility of these effects with state obligations under the European Convention on Human Rights (ECHR). As Hamilton and Buyse remark, although the ECHR 'does not deal explicitly with prosecution or amnesty', 5 it 'prohibit[s] the underlying violations, and provide [s] a right to a remedy […] and to a hearing before a competent tribunal for violations of rights'. 6 Furthermore, several ECtHR cases 'evince the human dimension at the heart of every transitional claimindividual quests for justice, exoneration, amnesty, truth, inclusion, equality, representation, protection, restitution or compensation'. 7 Hence, this article endeavours to cover the said gap in academic literature. Such inquiry is also important because amnesties/pardons involve state's ius puniendi: a core element of state sovereignty. Accordingly, it is key to identify which criteria underlie the ECtHR's scrutiny of that state power.
Generally, the ECtHR has developed its margin of appreciation doctrine, which comparatively implies a flexible approach to deference to states. 8 Thus, even if the ECtHR may decide to directly review amnesties/pardons in the future, it is not expected to adopt a strict control over these measures, namely, nullify or revoke them. As this article analyses, however, the ECtHR has increasingly dealt with and examined the effects of amnesties/pardons on individual rights in cases of serious human rights violations and it has been in principle sceptical about the compatibility of these effects with the ECHR.
Based on an analytical survey of ECtHR's judgments on amnesties/pardons, this article proposes that it is possible to identify four main factors or criteria that have seemingly underlain, affected, determined, or calibrated the degree of the ECtHR's deference to states in cases of amnesties/pardons for serious human rights violations. These factors are arguably: (i) the national process of adoption, application, and/or validation of amnesties/pardons; (ii) ECtHR's consideration of the margin of appreciation or discretion given to states; (iii) state compliance with international obligations on human rights; and (iv) potential impact on transitions to peace, reconciliation, democracy, and/or the rule of law.
This article acknowledges that some of these factors may partially overlap and that other authors may suggest different criteria in the future. However, no academic publication seemingly has yet proposed an explicit, detailed, and analytical systematization of the factors that, in light of ECtHR's jurisprudence, may explain whether and to what extent the ECtHR defers to states in cases of amnesties/pardons concerning serious human rights violations. Thus, it is important to identify and individualise specific criteria that determine ECtHR's deference to states in amnesty/pardon cases of serious human rights violations, including identification of emerging trends and developments in ECtHR's jurisprudence. This also helps to analytically specify and systematize the centripetal and centrifugal factors that determine expansions or contractions of ECtHR's deference to states in matters of those measures. This article has eight sections. Section 2 examines the influential IACtHR's amnesty/ pardon jurisprudence, particularly related scholarship, including some general references to domestic case-law. Section 3 discusses the selection criteria regarding ECtHR's cases used herein. Section 4 provides the general framework on the above-mentioned factors concerning or affecting the degree of ECtHR's deference to states in amnesty/ pardon cases. Sections 5-8 respectively analyse one factor per section, including specific conclusions. Section 9 provides general conclusions.

Amnesties/pardons at the IACtHR and related scholarship
This section primarily considers IACtHR's amnesty/pardon jurisprudence and related scholarship for these reasons. First, both the IACtHR's amnesty/pardon case-law and related scholarship provide important criteria which are arguably mutatis mutandis or to some extent similar or connected to the factors concerning or affecting the degree of ECtHR's deference to states in amnesty/pardon cases. Second, the IACtHR's amnesty/pardon jurisprudence merits consideration herein because it is the most robust worldwide in the field and it has been to a greater or lesser extent influential at national and supranational courts, including the ECtHR. Finally, and comparatively, the IACtHR and the ECtHR are supranational courts with mandates over human rights.
In Barrios Altos v. Peru (2001), the IACtHR was the first international court that found domestic self-amnesty laws on serious human rights violations to be null. 9 These amnesties favoured death squad members during President Fujimori's term. In La Cantuta v. Peru (2006), the IACtHR applied the above-mentioned finding regarding the same amnesty. 10 In Almonacid-Arellano v. Chile (2006), it found that a self-amnesty law concerning torture adopted during Pinochet's regime did not have legal effects. 11 In Gomes-Lund v. Brazil (2010), the IACtHR determined that a Brazilian military dictatorship's amnesty law lacked legal effects. 12 In Gelman v. Uruguay (2011), it found an amnesty law incompatible with state obligations, dismissing that such law was ratified in a democratic regime via two referendums. 13 In these cases, the IACtHR established that amnesty laws for atrocities were incompatible with the American Convention on Human Rights (ACHR) and states were found responsible for breaching their obligations to inter alia investigate/prosecute atrocities.
In more recent jurisprudence on amnesties/pardons, the IACtHR has arguably nuanced its approach by increasingly balancing competing interests in post-conflict contexts and/or post-dictatorships. The IACtHR in Tiu-Tojin v. Guatemala and Garcia-Lucero v. Chile did not annul amnesty laws but, to determine state responsibility, examined whether they were actually applied. 14 In La Rochela v. Colombia, it did not decide whether the Justice and Peace Law was an amnesty due to the early stage thereof: it instead indicated guiding implementation principles. 15 In El Mozote v. El Salvador, IACtHR's former President García-Sayán differentiated between amnesties adopted to end internal conflicts and those rendered by dictatorships. 16 In monitoring implementation of its judgments and as for a pardon of former Peruvian President Fujimori, the IACtHR deferred the matter to Peru's jurisdiction but subject to requirements and guidelines, 17 which Peru's Supreme Court followed. 18 As scholars remark, IACtHR's jurisprudence has had a 'spill-over effect' over Latin-American national courts, 19 which have adopted its anti-amnesty/anti-pardon doctrine. 20 Courts in Argentina, 21 Colombia, 22 Chile, 23 Peru, 24 etc. evidence so. Brazil has traditionally been the major exception; 25 however, certain recent Brazilian jurisprudence aligns with IACtHR's jurisprudence. 26 Some academic analysis of IACtHR's jurisprudence on amnesties/pardons may mutatis mutandis be to some extent related to the factor called herein national process of adoption, application, and/or validation of amnesties/pardons. Mallinder finds that contextual matters such as amnesties adopted by democratically elected governments or upheld in national referendums did not prevent the IACtHR from finding these measures incompatible with the ACHR when amnesties/pardons included gross abuses. 27 She has identified these trends: the IACtHR prohibits unconditional amnesties/pardons for atrocities while it permits limited amnesties excluding atrocities if the respective limitations are applied in practice. 28 She also concludes that the IACtHR normally scrutinises the amnesty text and its implementation. 29 Gargarella considers that the IACtHR should further argumentatively differentiate amnesty laws adopted/ratified via democratic proceedings from amnesties adopted by autocracies/dictatorships. 30 According to him, IACtHR's further consideration of the different levels of democratic pedigree of diverse national amnesties would be needed. 31 The factor labelled herein as ECtHR's consideration of the margin of appreciation or discretion given to states corresponds to the ECtHR's unique approach to the principle of subsidiarity. Nevertheless, regarding other approaches to the same principle, scholarship examining IACtHR's jurisprudence on amnesties/pardons remarks the importance of the IACtHR's control of conventionality doctrine. Under IACtHR's jurisprudence, 32 scholars thus note that this doctrine requires domestic organs to examine the compatibility of national laws or decisions with the ACHR/inter-American treaties, IACtHR's jurisprudence, and international law. 33 As authors remark, Latin-American courts have increasingly applied such doctrine. 34 This doctrine contrasts with the ECtHR's margin of appreciation doctrine that gives (much) more deference to states. While scholars such as IACtHR's former President Cançado-Trindade have encouraged the non-application of the margin of appreciation doctrine due to deficient local judiciaries and impunity/autocratic regimes in Latin America, 35 others criticise the IACtHR's control of conventionality as intrusive or interventionist. 36 Contesse suggests that the IACtHR and the ECtHR substantially differ in their approaches to the subsidiarity principle because the IACtHR has traditionally avoided granting margin of appreciation to states, instead asserting its supranational authority even when a democratic regime and through democratic proceedings adopted amnesties. 37 Binder remarks that IACtHR's jurisprudence on amnesties/pardons substantially limits national courts and legislators in how to give effect to the ACHR domestically. 38 Nevertheless, as Contesse points out, IACtHR's more recent jurisprudence on amnesties/pardons may indicate some emerging trend of 'constrained deference' whereby it defers to national authorities to decide on merits but this deference is subject to constraints on what national courts can do, which is however less intrusive than the IACtHR's traditional position. 39 The IACtHR may be becoming more aware of its subsidiary role. 40 Regarding the factor indicated herein as state compliance with international obligations on human rights, it can be mutatis mutandis identified some similar academic analyses of IACtHR's jurisprudence on amnesties/pardons. As Contesse comments, the IACtHR has continuously determined that amnesties for gross abuses are incompatible with inter-American human rights law, especially state's international obligations to investigate, prosecute, and punish. 41 Mallinder finds that the IACtHR has developed progressive interpretations of state duties to guarantee rights, harmonise domestic law with international obligations, and ensure victims' right to justice, going beyond other supranational approaches. 42 As she concludes, the IACtHR has increasingly attempted to balance state's international obligations with other considerations by not finding that a state breached its obligations provided that amnesties/pardons were not applied in practice to prevent investigation/prosecution of gross abuses. 43 Gargarella critically remarks that the IACtHR has adopted an inflexible view of the protection of rights and related state obligations by downplaying other factors. 44 Binder comments that IACtHR's jurisprudence aims that state's human rights obligations are implemented, 45 and such jurisprudence 'considerably extended the standard of review […] when examining whether a violation of the respective state's human rights obligations had occurred'. 46 As Contesse remarks, e.g. Argentina's Supreme Court fully embraced IACtHR's amnesty jurisprudence because this concerns state obligations involving anti-impunity doctrines. 47 Additionally, as certain authors note, 48 some European courts have found amnesties involving international crimes committed in Latin America 49 and elsewhere 50 not to bind these courts under universal jurisdiction and international law obligations. One of them concerned a Mauritanian amnesty, triggering the ECtHR's Ould Dah v. France case examined later.
Regarding the factor identified herein as potential impact on transitions to peace, reconciliation, democracy, and/or the rule of law, the following can be found mutatis mutandis and/or partially in scholarship examining IACtHR's amnesty/pardon jurisprudence. Mallinder identifies that the IACtHR has increasingly considered the impact of amnesties/pardons on societies in transition, particularly in La Rochela v. Colombia and Judge García-Sayán's Concurring Opinion in El Mozote v. El Salvador which, she remarks, 'indicate that the Court may be willing to distinguish between amnesties enacted during or after dictatorship, and amnesties […] to end violent conflict'. 51 As she explains, this evidences an IACtHR's flexible approach when examining state measures to balance state obligations to prevent further atrocities by finalising an armed conflict with state obligations concerning victims and it means a partial/ nuanced move towards conditional approaches to 'different roles that amnesty can play within different types of transitions'. 52 Binder finds that the IACtHR's focus on accountability and effective human rights protection 'facilitates the efforts of domestic institutions […] to implement human rights and the rule of law', and 'supports democratic transition and consolidation […] and, ultimately, to domestic self-determination' in Latin-American states. 53 As for domestic case-law vis-à-vis IACtHR's jurisprudence, Mallinder qualified Brazil as a regional 'outlier'. 54 She also found that, unlike most Latin-American courts, Brazil's Supreme Federal Court remarked the importance of the Brazilian amnesty law for a peaceful and democratic transition. 55 Comparatively, mutatis mutandis, South Africa's Constitutional Court also considered that the South African amnesty was pivotal in the South African negotiated settlement, assisting in the reconciliation and reconstruction process; 56 however, authors find that the Court did not answer whether such amnesty was compatible with international law. 57 It is thus possible to identify certain criteria in the IACtHR's amnesty/pardon jurisprudence and related scholarship that mutatis mutandis and to a greater or lesser extent are similar to the factors found in the ECtHR's jurisprudence when the ECtHR has decided on its degree of deference to states in amnesty/pardon cases. Precisely, the rest of the article examines these factors in the ECtHR's amnesty/pardon jurisprudence.

Remarks on the selection of ECtHR's cases
Concerning case selection in this article, all ECtHR's cases (a total of eight cases) and respective judgments that met cumulatively the following four selection criteria as of 1 October 2021 were considered: (i) the effects of amnesties/pardons on rights of individuals were the central issue or one of the main issues; (ii) effects of amnesties/pardons on human rights were discussed in some detail and as part of the ratio decidenci and merits rather than merely some generic or obiter dicta reference and/or discussion confined to preliminary objections; (iii) amnesties/pardons concerned or were related to serious human rights violations, specifically, gross violations of the rights to: life, freedom from torture/ill-treatments, or liberty/security when linked to one or the other two rights; and (iv) amnesties/pardons were connected to transitional processes and/or abusive state policies. Other cases that only met some of these selection criteria were excluded. 58 Thus, the analysis is arguably representative of the cases on the matter examined before the ECtHR.
The cases were identified by considering the scarce literature on ECtHR's jurisprudence on amnesties/pardons 59 and using the ECtHR-HUDOC database via these keywords: 'amnesties', 'pardons', 'serious violations', 'torture', 'inhuman or degrading treatment or punishment', 'murder', 'killing', 'extrajudicial execution', 'international crimes', 'crimes against humanity', 'war crimes', 'forced disappearance', 'Article 2', and 'Article 3'. 60 This article identified eight cases meeting the above-mentioned four cumulative selection criteria, which include thirteen case-law sources: eight judgments, one Grand Chamber's judgment, two joint concurring opinions, one concurring opinion, and one partly dissenting opinion. Not all cases/judgments receive the same level of discussion herein due to their different levels of relevance concerning this research. While all the eight cases identified are discussed, some cases are discussed in (much) further detail and/or in all below sections of the four factors on deference to states depending on: depth of ECtHR's examination that is relevant to this article in each case/judgment and usefulness thereof to better illustrate the respective factor. Contextual case information is provided. Cases/decisions are presented chronologically to identify certain jurisprudential trends better.
The cases/decisions identified and examined are presented in the below Table 1 ('ECtHR's amnesty/pardon jurisprudence').
Since the ECtHR is not static over time as its judges change, the ECtHR's jurisprudence on amnesties/pardons has increased since 2011 (as the below Table 1 shows), particularly in terms of breadth and depth of discussion. Moreover, there is arguably a trend of ECtHR's increasing scrutiny of the effects of amnesties/pardons on human rights, which has seemingly impacted its degree of deference to states as seen later. What should be mentioned now is that ECtHR's jurisprudence on amnesties/pardons has contained deeper judicial analyses in recent years, including concurring opinions (Margus v. Croatia) as well as a partly dissenting opinion and some indirect or implicit review of a pardon (Makuchyan and Minasyan v. Azerbaijan and Hungary). This may arguably be attributed to inter alia the composition of the respective judicial bench and the increasing importance that matters related to amnesties/pardons have gained at diverse supranational courts, particularly human rights courts and international/ hybrid criminal tribunals.

4.
Factors concerning or affecting the degree of ECtHR's deference to states in its amnesty/pardon jurisprudence: general framework This section starts the main part of the article, namely, discussion of the factors concerning or affecting the degree of ECtHR's deference to states in its jurisprudence on amnesties/pardons. Each of the four (overlapping) factors identified is discussed in its respective section (sections 5-8) by analysing ECtHR's jurisprudence (complemented with scholarship) that illustrates or evidences the factor in question, including conclusions about each factor in each section. As the next sections explain, illustrate, and demonstrate in detail, the following general remarks can be made.
Overall, it may be argued that the four factors identified in ECtHR's jurisprudence on amnesties/pardons have been considered explicitly/implicitly and to a greater or lesser extent when the Court has decided on its degree of deference to states in cases involving the effects of amnesties/pardons on rights of individuals. As mentioned, these factors present some overlapping elements, but they are arguably autonomous enough to consider them separately in light of ECtHR's jurisprudence on amnesties/pardons. The interaction of these factors has expanded or constrained the said degree of ECtHR's deference to states.
The specific national process of adoption, application, or validation of amnesties/ pardons 61 may be regarded as (usually) the starting point considered in the ECtHR's assessment to calibrate its degree of deference to a state in cases related to amnesties/ pardons. Depending on inter alia how the said process develops throughout different stages in time and the scope of amnesties/pardons (whether atrocities are included), the ECtHR's deference to states can be expanded or constrained.
By applying its doctrine of margin of state appreciation or discretion, 62 the ECtHR in turn has usually maintained its deference to states in matters related to amnesties/ pardons as showed later. This is arguably the most important factor for the ECtHR to expand its degree of deference to states based on ECtHR's overall/prima facie respect for state's local expertise, state's legitimate interests, state's discretionary policies, etc. Traditionally, the ECtHR's above-mentioned flexible doctrine has sharply contrasted with the IACtHR's conventionality control doctrine.
As an important counter-balancing factor vis-à-vis the margin of state appreciation in ECtHR's jurisprudence on amnesties/pardons, it can be identified the level of state compliance with international human rights obligations 63 to investigate, prosecute, try, and punish atrocities. The more a state does not comply with these obligations, the less deference will likely be accorded by the ECtHR. This arguably occurs especially (but not only) when state officers commit atrocities, and not limited to torture/ill-treatment but also including murders. The ECtHR's mandate involves supervising state compliance with those obligations.
Furthermore, the ECtHR has increasingly considered the potential impact of amnesties/pardons on transitional processes towards peace, democracy, and/or the rule of law 64 to expand or reduce the degree of deference to states. Should an amnesty/ pardon seek to achieve one or more of these legitimate aims in a specific context rather than just merely impunity, the ECtHR's jurisprudential trend seemingly indicates that it is more likely that the ECtHR can expand the degree of deference recognised to a state. This is also showed/discussed later.
As the next sections also demonstrate, the interactions among these four factors and the relative/comparative weight or attention given by the ECtHR to each factor identified have been dynamic and have experienced some evolution in recent years. In ECtHR's more recent jurisprudence on amnesties/pardons for serious abuses, it may be seemingly identified a potentially lower degree of deference to states or deference to states subject to some guidelines/conditions in amnesty/pardon cases.
As an additional or related (emerging) trend, it is generally sustained that the ECtHR has seemingly begun moving towards some indirect or implicit review of amnesties/ pardons in its latest jurisprudence. Nevertheless, unlike the IACtHR, the ECtHR has not yet proceeded with a direct, explicit or stricto sensu review of amnesties/pardons. Thus, the ECtHR's jurisprudence on amnesties/pardons (exclusively) concerns the effects of these measures on the rights of individuals.

The national process of adoption, application, and/or validation of amnesties/pardons
An absolute approach to the duty to prosecute in the ECHR would impact the scope of deference that the ECtHR should give to political decision-makers about amnesties/ pardons. 65 Based on certain ECtHR's jurisprudence, 66 Jackson correctly highlights that the specific historical-political context 'may widen the margin of appreciation'. 67 This general remark can be extrapolated to amnesties/pardons. Due to the nature and goals of these measures, the national process of their adoption, application, and/or validation corresponds to specific political-historical circumstances. Examining the process of each amnesty/pardon hence makes sense.
As Mallinder and others jointly identify concerning amnesties and the ECHR, amnesties are measures and processes related to the past and reconciliation and they may accompany other processes linked to investigations, reparations, etc. 68 As Jackson notices, amnesties/pardons may not result from balancing different goals but they may emerge from state's bad faith to exonerate its agents from liability. 69 Yet, distinguishing different types of amnesties/pardons is complex and these measures do not necessarily pursue legitimate ends. 70 Lexa v. Slovakia (2008) provides insights into the ECtHR's consideration of the national process of adoption, application, and validation of amnesties/pardons. The applicant received a presidential amnesty regarding his alleged involvement as the Slovakian Secret Service's Director in the abduction and intoxication of the then Slovak President's son amidst political controversies; however, a new Prime Minister revoked the amnesty and the applicant faced criminal charges. 71 Unlike other ECtHR cases that involved amnesties/pardons adopted by law, this case concerned Slovakia's Constitution (Article 102) that, as the ECtHR noted, entitles Slovakia's President 'to pardon […] sentences imposed by courts […] and to expunge convictions by means of an individual pardon or an amnesty'. 72 Nevertheless, this power is not entirely automatic. As the ECtHR remarked, Slovakia's Constitution subjects the validity of pardons/amnesties to the Prime Minister's signature. 73 Moreover, the ECtHR paid attention to interpretations of the Slovakian Constitutional Court and Slovakian criminal courts, a commentary by authors headed by the Slovakian Constitutional Court's President, and the prevailing legal opinion in other States Parties to the ECHR. 74 The ECtHR concluded that: presidential amnesties cannot be quashed under the Slovakian Constitution; and neither national practice nor legal theory allows revoking presidential amnesties/pardons. 75 Concerning Slovakian courts, the ECtHR added that they 'based their conclusion in discontinuing the proceedings on their own assessment of the facts of the case and interpretation of the relevant law'. 76 Indeed, the ECtHR found no basis to question Slovakian courts' interpretation, which determined that the Presidential amnesty covered the offences for which the applicant had been prosecuted. 77 The ECtHR explicitly considered the factual and legal scenario surrounding the case, related political controversies, diverging legal views, and public debates. 78 Yet, it ultimately found Slovakia responsible for violating Article 5(1) ('Right to liberty and security') of the ECHR. 79 This was because the applicant who had benefited from the Presidential amnesty was subsequently prosecuted following the revocation of the amnesty by the new Prime Minister as Slovakia's Acting President. 80 Such outcome illustrates that an important later deficit in the long process of application or validation of amnesties/pardons may cause effects that breach human rights under the ECHR although the adoption and initial application of these measures were legally valid. Thus, States Parties to the ECHR should be aware of the ECtHR's potential scrutiny of not only a particular instance of the adoption or application process of an amnesty/pardon but the whole process related to the amnesty/pardon concerning its effects on human rights recognised in the ECHR.
In Association 21 December 1989 v. Romania (2011), the ECtHR examined the effects of the 2008 Draft Amnesty Law on human rights violated by the use of lethal force by Romanian armed forces against protesters when the Romanian dictator Nicolae Ceauşescu was overthrown. 81 This bill (Article 2) included an amnesty: 'military officers and service personnel who have been tried and convicted, or against whom judicial proceedings have been brought on account of their participation in the events of December 1989 shall qualify for the amnesty'. 82 Hence, the scope of the Romanian draft amnesty law arguably included all crimes without excluding serious human rights violations or international crimes: it seemingly aimed at granting a blanket amnesty. Additionally, unlike Lexa v. Slovakia, the present case concerned a bill.
The ECtHR drew attention to the legal, political, and historical process surrounding this bill. This means the alleged use of the bill to influence prosecutors investigating the protests against Ceauşescu and ensuing violent crackdown resulting in numerous victims (December 1989-June 1990). 83 As the ECtHR noted, the applicant association claimed that the head of the military prosecutor's office disseminated the bill to prosecutors, which the association considered 'an attempt to influence the prosecutors and to suppress definitively the investigations'. 84 The ECtHR noted that the applicant association regretted that an NGO-backed bill to speed up investigations was not 'disseminated to prosecutors, as the draft amnesty law had been'. 85 Moreover, the ECtHR observed that the Ministry of Defence's Legal Directorate transmitted the draft amnesty law to military prosecuting authorities at the High Court of Cassation for consultation. 86 The ECtHR referred to the rights of victims and their families to know the truth about massive violations and that this case concerns 'widespread use of lethal force against the civilian population during anti-Government demonstrations preceding the transition from a totalitarian regime to a more democratic system'. 87 Thus, the ECtHR concluded that it 'cannot accept that an investigation has been effective where it is terminated as a result of the statutory limitation of criminal liability, when it is the authorities themselves who have remained inactive of fundamental rights'. 88 Importantly, the ECtHR invoked its jurisprudence to remark that 'an amnesty is generally incompatible with the duty incumbent on the States to investigate acts of torture […] and to combat impunity for international crimes. This is also true in respect of pardon'. 89 Unsurprisingly, the Court found Romania's responsibility for violations of the 'procedural limb' of Article 2 ('Right to life') of the ECHR. 90 The excessively broad scope of the amnesty bill and the attempted use thereof to stop investigations seriously affected the validity and legitimacy of the national process of adoption of the Romanian amnesty. These features arguably underlay the ECtHR's finding of Romania's state responsibility for procedural violations of Article 2 of the ECHR. Such outcome illustrates how important is that States Parties to the ECHR avoid failures in any step in the national process of adoption, application, and/or validation of amnesties/pardons because noticeable deficits in this path can later lead to their being found internationally responsible by the ECtHR for violations of rights recognised in the ECHR.
An important element that indicates the goals underlying the national process of adoption, application, and/or validation of amnesties/pardons is the scope of these measures, particularly, the kind of crimes covered by these exemption measures. In Yeter v. Turkey (2009), which involved torture to death followed by the termination of disciplinary proceedings against police officers due to an amnesty law, the ECtHR determined that 'the granting of an amnesty or pardon should not be permissible' concerning crimes violating Article 3 ('Prohibition of torture') of the ECHR. 91 As part of the applicable or relevant law in both Margus v. Croatia and Tarbuk v. Croatia, the ECtHR paid attention to the 1996 Croatia's General Amnesty Act, which concerned criminal offences committed during the armed conflicts in Croatia in the nineties. 92 This act (Section 1) 'grants general amnesty from criminal prosecution and proceedings to the perpetrators of criminal offences committed during […] armed conflicts in […] Croatia'. However, its Section 3 crucially establishes that: 'No amnesty […] shall be granted to perpetrators of the gravest breaches of humanitarian law, which have the character of war crimes, […] genocide […]'. Unlike the bill in Association 21 December 1989 v. Romania, the Croatian cases concerned an act. Another key difference is that the Croatian amnesty legislation did not provide a blanket amnesty for all offences. As quoted, this amnesty act explicitly excluded military personnel suspects of international crimes or serious human rights violations.
The ECtHR in Margus v. Croatia (2012) actually paid close attention to the contents of the above-mentioned Croatian General Amnesty Act and, especially, how different branches of the Croatian state interpreted and applied it. The ECtHR agreed with the Croatian Supreme Court's conclusions 'to the effect that the Amnesty Act was erroneously applied in the applicant's case' since the offences committed by the applicant constituted war crimes, namely, the amnesty could not benefit the applicant. 93 By invoking practices of international bodies related to the prevention or prohibition of amnesties for war crimes, the ECtHR indeed 'accepts the Government's view that the granting of amnesty to the applicant in respect of acts which were characterised as war crimes against the civilian population amounted to a fundamental defect in the proceedings'. 94 Hence, the ECtHR found no Croatia's state responsibility for the reopening of criminal proceedings against the applicant. As the ECtHR established, Croatia met the conditions under Article 4(2) ('Right not to be tried or punished twice') of Protocol 7 to the ECHR for reopening proceedings. 95 All of this sharply contrasts with the above-examined Romanian situation.
Finally, in Makuchyan and Minasyan v. Azerbaijan and Hungary (2020), Hungarian courts had sentenced an Azerbaijan soldier (referred to as 'R.S.') to life imprisonment for a hate crime that involved the killing of an Armenian soldier during a NATO training in Budapest. 96 However, Azerbaijan pardoned, released, and reinstated him in the army upon his transfer to Azerbaijan where he would serve the rest of his sentence. 97 Regarding the process of adoption, application, and validation of this pardon, the following is noteworthy. First, the ECtHR invoked findings of the Council of Europe's Parliamentary Assembly: the pardon violated the good faith principle and the rule of law. 98 As the ECtHR noted, this was because Azerbaijan in his request for the transfer of R.S. promised Hungary to continue R.S.'s prison sentence but it freed him immediately upon his return to Azerbaijan. 99 Second, the ECtHR observed that 'there is nothing in the case file to indicate that a formal request to this end [pardon] was ever made, and nor is there any indication that there ensued any kind of reflection process or legal procedure for the pardon'. 100 Third, the ECtHR described Azerbaijani officials' statements 'glorifying R.S., his deeds and his pardon' as disturbing. 101 Therefore, the ECtHR's cases examined in this section arguably evidence that the process through which an amnesty/pardon was adopted, applied, and/or validated by domestic authorities and/or national courts has constituted a factor underlying, concerning, or affecting the degree of ECtHR's deference to states when examining the effects of amnesties/pardons on human rights. Regarding the national process of adoption of these measures and their effects on human rights, the ECtHR has to a greater or lesser extent considered which organ granted these measures, the normative instrument through which these measures were adopted, the scope (subject matter) of these measures, and/or how the measure was adopted.
As for the national process of application and validation of amnesties/pardons and effects on human rights, the ECtHR has considered not only the initial implementation or execution of these measures but also subsequent or later actions adopted by (later) governments, authorities, or (judicial) organs of the same state. Thus, a later national decision related to an amnesty/pardon may lead to effects on human rights that trigger state responsibility for violations of the ECHR at the ECtHR even if the initial adoption, application, or validation of the measure violated no human right recognised in the ECHR. In turn, this will contribute to expand or constrain ECtHR's deference to states in cases involving amnesties/pardons for serious human rights violations.

ECtHR's consideration of the margin of appreciation or discretion given to states
The ECtHR has developed its well-known doctrine of 'margin of appreciation'. As scholars have discussed, margin of appreciation involves 'room for manoeuvre', 102 an 'ambit of discretion', 103 or a 'latitude of deference or error' 104 that the ECtHR is prepared to 'accord national authorities in fulfilling their obligations under the European Convention on Human Rights' 105 or that it is willing to give to 'national authorities in assessing appropriate standards of the Convention rights, taking into account particular values and other distinct factors woven into the fabric of local laws and practice'. 106 The margin of appreciation or discretion doctrine relates to the supplementary or subsidiary character of the ECHR and the ECtHR vis-à-vis sovereign national systems concerning human rights protection. 107 Thus, the margin of appreciation or discretion enables the ECtHR to balance the sovereignty of the States Parties to the ECHR and their human rights obligations under the ECHR. 108 In examining ECtHR's amnesty-related jurisprudence, Jackson observes that, if the duty to prosecute is absolute, this could inhibit how the margin of appreciation should operate in amnesty cases. 109 Regarding deference on the international-local axis, such a strict approach would unduly limit ECtHR's deference to national decisionmakers. 110 The ECtHR has not directly reviewed amnesties/pardons. Yet, ECtHR's amnesty-related jurisprudence seemingly promises a degree of flexibility in how the ECtHR may address the permissibility of amnesties/pardons. 111 However, Jackson notes that the ECtHR's approach might potentially keep out considerations related to 'margin of appreciation and the scope of judicial deference' which may 'push decision making from the national level to the international level, and from political decision makers to the Court. As to the former, these are exactly the cases in which a wider margin of appreciation ought to be granted to national authorities'. 112 Pinto interestingly remarks that although the ECtHR's practice suggests a growing trend that requires criminal law enforcement as part of state obligations to investigate, prosecute, and sanction: 'this has been affirmed only with regard to serious violations, namely the right to life, acts of torture and ill-treatment […] the ECtHR has yet to go as far as the IACtHR to require punishment […] Nevertheless, there are signs of a potential development in this direction'. 113 The level to which the ECtHR may calibrate the margin of appreciation or discretion given to states involves competing elements. In particular, positive duties stemming from absolute rights vis-à-vis political scenarios suggesting deference from the ECtHR to the national level. 114 As the ECtHR established, due to 'their direct knowledge of their society and its needs, the national authorities are in principle better placed than the international judge to appreciate what is in the public interest'. 115 Yet, Jackson soundly suggests confining amnesties 'to exceptional situations, where the state is pursuing a compelling public interest' and also maintaining 'the Court's overarching supervisory jurisdiction […] demanding justification from the state for its decision. Nonetheless, recognition of the potential conflict opens up the possibility of the state defending its amnesty […]'. 116 Regarding ECtHR's practice on amnesties/pardons, Mallinder and others jointly identified criteria that may arguably relate to ECtHR's consideration of the margin of state appreciation and limits thereof. 117 Thus, States Parties to the ECHR can adopt amnesties/pardons if these measures: (i) are necessary to fulfil state's legitimate aims, including peaceful resolution of armed conflicts; (ii) are exceptional and applicable to particular events and/or offenders rather than a general impunity practice that undermines the rule of law; and (iii) co-exist with or support investigations. 118 Based on certain ECtHR's jurisprudence, 119 Mallinder and others jointly suggest that amnesties may be applicable with a larger degree of flexibility when criminal proceedings concern private individuals rather than state officers. 120 In Lexa v. Slovakia (2008), the ECtHR provided key principles mainly tailored to presidential pardons. It acknowledged the important amount of state appreciation or discretion underlying these acts, remarking the following. First, the head of state's powers concerning clemency measures and potential review thereof depend on each national constitutional model. 121 Second, pardons granted by the executive power are usually 'atypical discretionary acts […] in the framework of relations between the branches of power […] they cannot be construed as normal administrative measures subject to ordinary judicial review'. 122 Third, pardons are overall delimited by rule of law principles such as public order, need for sanctions, legal certainty, power separation, equality, and/or constitutional norms; however, the judicial review or overturn of a pardon is seemingly very limited in States Parties to the ECHR. 123 Finally, the ECtHR established that 'the discretionary character of these measures does not, in principle, allow for their revocation'. 124 Ould Dah v. France (2009) concerned France's exercise of universal jurisdiction against the applicant who was then a Mauritanian army captain. 125 The applicant had benefited from a Mauritanian Amnesty Law for armed/security forces who committed crimes connected with a Mauritanian armed conflict and violent acts (1989)(1990)(1991)(1992). 126 However, while in France, he faced criminal charges and trial and the French judiciary in absentia convicted him of torture committed in Mauritania. 127 By relying on its previous jurisprudence, 128 the ECtHR determined that the States Parties to the ECHR '[are free] to determine their own criminal policy, which is not in principle a matter for it to comment on' and that 'a State's choice of a particular criminal justice system is in principle outside the scope of the supervision it carries out at European level, provided that the system chosen does not contravene the principles set forth in the Convention'. 129 Furthermore, the ECtHR recognised that States Parties to the ECHR via judicial interpretation of their criminal law provisions can respond to the ever-existing 'need for elucidation of doubtful points and for adaptation to changing circumstances […]' and that in those states 'the progressive development of the criminal law through judicial law-making is a well-entrenched and necessary part of legal tradition'. 130 It remarked that Article 7 ('No punishment without law') of the ECHR 'cannot be read as outlawing the gradual clarification of the rules of criminal liability through judicial interpretation from case to case' provided that this is consistent with the essence of the crime and could be foreseen reasonably. 131 The ECtHR importantly pointed out the principle of gradual development of law in any state that is subject to the rule of law and democratic regime 'factors which constitute the cornerstones of the Convention'. 132 Based on inter alia these considerations, the ECtHR acknowledged that French courts in application of France's legislation can exercise universal jurisdiction over torture. 133 Thus, the ECtHR declared the application inadmissible. 134 This meant that the ECtHR did not consider that the Mauritanian amnesty law benefiting the Mauritanian applicant was a decisive element to preclude the margin of appreciation or discretion given to France to arrest him and in absentia convict him of torture.
From the ECtHR's judgment in Yeter v. Turkey (2009), it can be inferred that the ECtHR may respect the margin of state appreciation in amnesty/pardon cases involving serious abuses provided that the respective national criminal-law system is applied rigorously and possesses dissuasive effects to prevent atrocities effectively. 135 In turn, the ECtHR in Enukidze and Girgvliani v. Georgia (2011) was 'struck' by the presidential pardon granted to state agents convicted of murder and it commented that although Georgia's government expected that the Georgian society would accept a lenient sentence, amnesties/pardons 'can scarcely serve the purpose of an adequate punishment' concerning murders. 136 In Tarbuk v. Croatia (2012), criminal proceedings were discontinued against the applicant, who had been detained, because he benefited from Croatian amnesty legislation. 137 Nevertheless, he unsuccessfully claimed compensation for his detention before Croatian courts. 138 The ECtHR dismissed the applicant's claim of Croatia's violation of Article 6 ('Right to a fair trial') of the ECHR. 139 The ECtHR inter alia closely followed and respected the Croatian Constitutional Court's finding. Under the Croatian Constitution (Article 25 (4)), which recognises compensation solely for persons unlawfully detained/convicted, Croatia's Constitutional Court established that there is no compensation when the proceedings are not unlawful or flawed. 140 Importantly, the ECtHR highlighted that 'In the light of the principle that the domestic authorities are best placed to interpret the domestic law […] the Court accepts such an interpretation of the relevant domestic rules, particularly since they appear to be in compliance with the relevant provisions of the Convention'. 141 The ECtHR also referred to the Croatian General Amnesty Act 'as a sovereign act resulting in the applicant's immunity from criminal prosecution'. 142 By invoking the extinct European Commission of Human Rights' case-law, 143 the ECtHR remarked that: […] even in such fundamental areas of the protection of human rights as the right to life, the State is justified in enacting, in the context of its criminal policy, any amnesty laws it might consider necessary, with the proviso, however, that a balance is maintained between the legitimate interests of the State and the interests of individual members of the public. 144 In Margus v. Croatia (2014), the ECtHR's Grand Chamber respected the decision of Croatian authorities who brought a fresh indictment against the applicant and convicted him of war crimes against civilians although he had benefited from the previously examined Croatian amnesty legislation. 145 As discussed later, such finding mainly corresponded to that the new criminal proceedings against the applicant were consistent with ECHR/international law obligations. 146 Yet, the outcome also seemingly reveals an ECtHR's respectful approach to the margin of appreciation or discretion given to Croatian authorities. The ECtHR's Grand Chamber respected Croatia's discretion to re-start proceedings, which had been discontinued against the applicant under Croatian amnesty legislation. Consequently, the Chamber inter alia declared inadmissible the complaint under Article 4 of Protocol 7 to the ECHR regarding the applicant's right not to be tried or punished twice concerning charges that involved killing. 147 The Grand Chamber's Judgement in Margus v. Croatia contain relevant points related to the ECtHR Judges' consideration of the margin of appreciation or discretion given to a state. While the joint concurring opinion of Judges Šikuta, Wojtyczek, and Vehabović remarked that human rights violations cannot go unpunished and that amnesty laws may potentially have 'perverse effects' when provide impunity to the perpetrators of those violations, they stated: 'Different countries have devised widely varying approaches enabling them to put grave human rights violations behind them and restore democracy and the rule of law'. 148 They added that 'international rules imposing a blanket ban on amnesties in cases of grave violations of human rights [are] liable, in some circumstances, to reduce the effectiveness of human rights protection' and, in certain cases, 'there may be practical arguments in favour of an amnesty that encompasses some grave human rights violations'. 149 Thus, they were 'in favour of allowing the States concerned a certain margin of manoeuvre […] to allow the different parties to conflicts engendering grave human rights violations to find the most appropriate solutions'. 150 Finally, Judge Vučinić's concurring opinion implicitly invoked the importance of the ECtHR's respect for the margin of appreciation or discretion afforded to a state: […] retrial and final conviction of the applicant have to be understood as a legal and legitimate effort on the part of the Croatian authorities to correct the previously mentioned defects in the domestic proceedings. This, I believe, is fully in accordance with the letter and spirit of Article 4 of Protocol No. 7. 151 In Makuchyan and Minasyan v. Azerbaijan and Hungary (2020), the ECtHR remarked that 'pardons and amnesties are primarily matters of member States' domestic law'. 152 It added that they are generally consistent with international law unless they concern gross atrocities. 153 The ECtHR then examined Azerbaijan's reasons for releasing the pardoned Azerbaijani soldier, namely humanitarian concerns for his mental condition and alleged unfair criminal proceedings against him in Hungary, and the ECtHR found these reasons unconvincing. 154 First, the ECtHR determined that: 'it is difficult to seriously question the fairness of criminal proceedings conducted in another Council of Europe Member State'; 155 Hungarian criminal proceedings against the pardoned soldier were fair; and he did not file an application under Article 6 of the ECHR when the proceedings finished. 156 Second, the ECtHR concluded that: the soldier's mental difficulties alone could hardly justify Azerbaijan's failure to enforce punishment for a serious hate crime; he was mentally fit when committing the crime as medical experts found during his trial; and Azerbaijan promoted him, which indicates (if any) a non-serious mental condition. 157 In commenting Makuchyan and Minasyan v. Azerbaijan and Hungary, Ryngaert and Istrefi remark that the ECtHR's assessment of the reasons for releasing 'practically amounted to reviewing the reasons for pardoning R.S. […] Without pronouncing it, the Court appears to have conducted a form of judicial review of the presidential pardon'. 158 However, as they also recognise, the ECtHR did not determine that the pardon itself in the R.S.'s case is incompatible with Article 2 of the ECHR. 159 Milanovic and Papic note that the ECtHR implicitly reviewed the justifiability of the said presidential pardon. 160 Yet, they reckon that the ECtHR focused on whether releasing rather than pardoning the convicted soldier violated Article 2. 161 Although the pardon was a formal condition for releasing the convicted soldier, the ECtHR did not review the pardon directly or explicitly. Indeed, Judge Pinto-de-Albuquerque in his partly dissenting opinion sustained that the ECtHR should have decided on the pardon directly. 162 Even though the applicants suggested that the ECtHR could order revoking the pardon, 163 the ECtHR actually dismissed this by remarking that the state in principle is 'free to choose the means by which it will discharge its legal obligation' to execute an ECtHR's judgment provided that such means are compatible with the judgment conclusions. 164 As the ECtHR highlighted, 'this discretion as to the manner of execution of a judgment reflects the freedom of choice attached to the primary obligation of the Contracting States' to secure rights. 165 Thus, the ECtHR concluded that it 'does not consider it appropriate to indicate the need for any general or individual measures in respect of Azerbaijan'. 166 Therefore, ECtHR's jurisprudence on amnesties/pardons examined in this section has pointed out that the ECtHR has considered and overall or in principle respected the margin of appreciation or discretion given to states. Such criterion is evidenced by and based on the ECtHR's explicit acknowledgement of inter alia: the (inherently) discretionary nature of amnesties/pardons, the fitness and authority of domestic authorities to interpret national rules better than the ECtHR can do, free determination of the national criminal policy, state's legitimate interests, and state's discretion to choose the means to execute an ECtHR's judgment. Thus, these elements are closely related to the ECtHR's consideration of and (expected) overall or prima facie respect for the margin of appreciation or discretion given to states in amnesties/pardons, which corresponds to the ECtHR's general doctrine of margin of appreciation.
Nevertheless, the ECtHR's jurisprudence on amnesties/pardons also establishes that such margin of state discretion or appreciation needs to be balanced against other considerations, particularly human rights of individuals and ECHR core principles. It may be additionally sustained that the ECtHR has accorded a wider margin of state discretion or appreciation when the respective state has handled the effects of amnesties/pardons in manners that take into account these other (potentially) competing elements.
Consideration of the margin of appreciation or discretion given to states is arguably the most decisive criterion for the ECtHR to expand its deference to states in cases of amnesties/pardons for gross abuses. Conversely, the next section examines state compliance with international human rights obligations as, arguably, the strongest factor for the ECtHR's contraction of such deference to states in cases involving amnesties/pardons for atrocities.

State compliance with international human rights obligations
In examining ECtHR's amnesty jurisprudence, Jackson remarks that although absolute/ non-derogable rights give rise to state duties such as prosecution, this obligation has been construed problematically due to the 'underlying assumption that each of the duties to which an absolute right gives rise carries that absolute status'. 167 As he comments, treating the state obligation to prosecute as absolute would render it non-derogable, which 'closes a valuable means of political flexibility […]'. 168 The ECtHR should explain: how the absolute/non-derogable nature of the right translates into each related obligation; and to what extent the absolute status of certain rights includes obligations beyond prohibition of atrocities. 169 Yet, Mallinder and Jackson acknowledge that amnesties/pardons at the ECtHR and beyond concern victims' rights as challenges from victims and their families against amnesties/pardons evidence. 170 Thus, ECtHR's developments of state procedural obligations under the ECHR are justified in cases of serious violations of rights. 171 As scholars have commented, the ECtHR has been particularly vigilant of state compliance with international obligations when state officers committed gross human rights violations. 172 As Pinto remarks, the ECtHR has adopted a censorious approach to state measures that impede criminal justice, including amnesties/pardons. 173 When these measures involve the right to life, the said ECtHR's approach reflects a perspective whereby those measures are permissible only if they are exceptional and necessary for a legitimate aim, differing from the IACtHR's absolute prohibition. 174 As Pinto recognises, however, the ECtHR has traditionally adopted a stricter approach when amnesties/pardons involve torture (ECHR, Article 3). 175 Jackson also notes this ECtHR's traditionally different degree of flexibility concerning the state obligation to prosecute in amnesty cases related to breaches of Article 2 ('Right to life') vis-à-vis amnesties concerning violations of Article 3. 176 Likewise, Mallinder and others jointly sustain that the ECtHR has seemingly followed a more flexible approach to amnesties for violations of Article 2 compared to amnesties for violations of Article 3. 177 Nevertheless, Mallinder and others jointly also state that similar to 'Article 2 violations, Article 3 clearly requires prompt, effective, transparent, and independent investigations and where an amnesty blocks such investigations it is likely to be viewed as impermissible'. 178 Based on ECtHR's jurisprudence, Mallinder and others jointly indicate that although international law has not yet evolved to ban amnesties for serious human rights violations and international crimes, amnesties should be accompanied 'by compliant investigative processes, reparations for victims, and other reconciliation measures […] amnesties that are limited by the exclusion of particular offences may not offer permanent protection to perpetrators of these crimes'. 179 Pinto summarises the ECtHR's general standing about amnesties/pardons: 'according to the ECtHR, measures that would prevent criminal justice are generally incompatible with the ECHR since they foster impunity and hinder accountability'. 180 In Lexa v. Slovakia (2008), the ECtHR invoked its jurisprudence to remark that 'it is of the utmost importance' that amnesties/pardons 'should not be permissible' when state agents are charged with serious crimes, including torture or ill-treatment. 181 Additionally, it stated that a similar conclusion can be obtained from international law and practice, 182 invoking the UN-General Assembly's Enforced Disappearance Declaration, IACtHR's jurisprudence, and Special Court for Sierra Leone's case-law. 183 In Yeter v. Turkey (2009), the ECtHR determined the inadmissibility of amnesties/ pardons, namely, when a state agent is accused of crimes that violate Article 3 ('Prohibition of torture') of the ECHR 'an amnesty or pardon should not be permissible'. 184 The Court explained that, in this case, the manner in which national law was applied 'rendered the disciplinary proceedings ineffective', and it expressed its concern about the Turkish government's repeated failure to address the said problem. 185 Since this case involved a killing as a result of torture and the ECtHR found a substantive and procedural violation of Article 2 ('Right to life') of the ECHR, the Court concluded about the alleged violation of Article 3: 'although this complaint is admissible, there is no need to make a separate examination of its merits'. 186 In Ould Dah v. France (2009), the ECtHR found that, in cases of serious violations of human rights such as torture, the state obligation to prosecute offenders 'should not therefore be undermined by granting impunity to the perpetrator in the form of an amnesty law that may be considered contrary to international law'. 187 It also derived an important procedural consequence concerning amnesties: international law does not preclude an individual who benefited from an amnesty before trial in his/her own state from being tried by another state. 188 The ECtHR invoked the International Criminal Court (ICC) Statute (Article 17), which does not include the said situation among inadmissibility grounds. 189 Despite a Mauritanian amnesty benefiting the applicant, the ECtHR welcomed the Nîmes Assize Court's conviction of the applicant as this corresponded to France's obligations to criminalise torture in its legislation and exercise jurisdiction over the applicant under the UN-Convention against Torture and the UN-Committee against Torture's recommendations to France. 190 As for state agents, especially law-enforcement officers, convicted of crimes that violated Article 2 ('Right to life') of the ECHR, the ECtHR in Enukidze and Girgvliani v. Georgia (2011) found that 'an amnesty or pardon can scarcely serve the purpose of an adequate punishment'. 191 It importantly added that 'the Court expects States to be all the more stringent when punishing their own law-enforcement officers for the commission of such serious life-endangering crimes than they are with ordinary offenders '. 192 Regarding state agents, the ECtHR emphasised that what is at stake is not only individual criminal law liability but also the state obligation to fight impunity and the respect for the law-enforcement system. 193 In Association 21 December 1989 v. Romania (2011), the ECtHR remarked the importance of the right of victims and their families to know the truth about events underlying serious human rights violations and that involves the right to effective judicial proceedings and the right to compensation. 194 It added that, in contexts such as widespread lethal force against civilians, it cannot accept ineffective investigations and 'an amnesty is generally incompatible with the duty incumbent on the States to investigate acts of torture […] and to combat impunity for international crimes. This is also true in respect of pardon '. 195 In Margus v. Croatia (2012), the ECtHR remarked that amnesties/pardons for international crimes such as genocide, crimes against humanity, and war crimes are 'increasingly considered to be prohibited by international law'. 196 The ECtHR invoked customary international humanitarian law rules, human rights treaties, decisions of international and regional courts, developing state practice, and 'a growing tendency for international, regional and national courts to overturn general amnesties enacted by Governments '. 197 In Margus v. Croatia (2014), the ECtHR's Grand Chamber established that 'an amnesty or pardon should not be permissible' when state agents have been charged with serious crimes such as torture. 198 It highlighted that national authorities should not give the impression that they willingly permit that serious violations of human rights remain unpunished. 199 Besides ECtHR's jurisprudence, the ECtHR's Grand Chamber invoked practices of the UN-Human Rights Committee and the International Criminal Tribunal for the Former Yugoslavia to find that amnesties are generally incompatible with the state obligation to investigate and prosecute individuals responsible for serious violations of human rights and international crimes, namely, such state duty 'should not therefore be undermined by granting impunity to the perpetrator in the form of an amnesty law that might be considered contrary to international law'. 200 The Chamber also relied on a teleological interpretation. Thus, it affirmed that amnesties/pardons for gross abuses such as intentional killings and torture/ill-treatment would be inconsistent with the ECHR's object and purpose as an instrument to protect persons and which requires an interpretation and application of the ECHR provisions to have practical and effective safeguards. 201 Importantly, the ECtHR's Grand Chamber considered that, in amnesty-related matters, it 'should take into account developments in international law in this area. The Convention and its Protocols cannot be interpreted in a vacuum but must be interpreted in harmony with the general principles of international law of which they form part'. 202 Generally, scholars have highlighted ECtHR's use of 'external' international law sources. 203 The Chamber acknowledged that no international treaty explicitly prohibits amnesties for serious human rights violations. 204 However, it remarked that diverse international bodies/courts held that amnesties: should not be granted to those who seriously breached international human rights law/humanitarian law; and are inadmissible when they seek to prevent investigation and punishment of those responsible for gross atrocities. 205 The Chamber noted the IACtHR's firm stance on the inadmissibility of amnesties regarding serious human rights violations as they are incompatible with state duties to investigate and punish offenders. 206 The Chamber weighed in different goals pursued by amnesties such as peace and reconciliation but also it mainly acknowledged the need for appropriate punishment of those responsible for atrocities under state obligations stemming from the ECHR and international law: A growing tendency in international law is to see such amnesties as unacceptable because they are incompatible with the unanimously recognised obligation of States to prosecute and punish grave breaches of fundamental human rights […] The Court considers that by bringing a fresh indictment against the applicant and convicting him of war crimes […] the Croatian authorities acted in compliance with the requirements of Articles 2 and 3 of the Convention and in a manner consistent with the requirements and recommendations of the above-mentioned international mechanisms and instruments. 207 Finally, in Makuchyan and Minasyan v. Azerbaijan and Hungary (2020), the ECtHR stated that pardons/amnesties 'are in principle not contrary to international law, save when relating to acts amounting to grave breaches of fundamental human rights'. 208 Regarding Azerbaijan's pardon, release, and military promotion of an Azerbaijani soldier convicted of an ethnically motivated murder, the ECtHR determined that Azerbaijan's acts 'in effect granted R.S. impunity for the crimes committed against his Armenian victims', which is incompatible with Azerbaijan's obligation under Article 2 of the ECHR to effectively deter crimes against human lives. 209 Thus, the ECtHR found that Azerbaijan violated Article 2 'under its procedural limb'. 210 Moreover, it determined that the measures were ethnically/racially motivated. 211 Accordingly, it found that Azerbaijan violated Article 14 ('Prohibition of discrimination') together with Article 2. 212 This section has hence shown that, whether and to what extent the state complies with its international human rights obligations to investigate, prosecute, try, and punish serious abuses when granting amnesties/pardons is a powerful factor considered by the ECtHR to decide on the degree of deference to accord to a state. Indeed, state compliance with the above-mentioned international human rights obligations arguably works as a counter-balancing factor against the ECtHR's consideration and overall or prima facie respect for the margin of appreciation or discretion given to states. This means that the ECtHR has relied on the said state compliance to constrain its quota of deference to states in cases of amnesties/pardons for serious abuses. International human rights obligations that bind States Parties to the ECHR when adopting amnesties/pardons stem from the ECHR as developed in ECtHR's jurisprudence. Moreover, the ECtHR has increasingly relied on diverse 'external' international law sources such as the practice of UN-human rights bodies, the IACtHR, and international/hybrid criminal tribunals, which have generally banned or rejected the admissibility of amnesties/ pardons in cases of serious human rights violations or international crimes. Thus, the more a state does not comply with international human rights obligations to investigate, prosecute, try, and punish serious abuses, the less likely is that the ECtHR will defer to this state in cases of amnesties/pardons for atrocities. This is particularly (but not only) the case when state officers committed these atrocities.
The ECtHR has traditionally been very strict in monitoring the compliance of state obligations in cases of amnesties/pardons for torture offenders (ECHR, Article 3), which corresponds to the absolute prohibition of torture, its impact on procedural obligations, and frequent torture/ill-treatment episodes in ECtHR cases. By inter alia invoking the above-mentioned 'external' international law sources, the ECtHR arguably has also expanded or began to expand a similar strict approach to violations of the right to life (Article 2). This is a welcome development. In contexts of armed conflicts, dictatorships, or serious political unrest, atrocities such as torture, arbitrary killing, and enforced disappearance are all serious human rights violations that may likely constitute international crimes, particularly, crimes against humanity when committed in a widespread or systematic manner against civilians, or war crimes when linked to armed conflicts. 213 Hence, establishing or implying some hierarchy among these atrocities may potentially send the wrong message to States Parties to the ECHR about more 'permissible' gross abuses regarding amnesties/pardons. Indeed, this ultimately could distort the level of deference that the ECtHR should or may accord to states. 8. Potential impact on transitions to peace, reconciliation, democracy, and/or the rule of law As Jackson remarks, a reason why the amnesty debate is relevant to the ECHR's space is that 'transition from conflict to peace is relatively recent in a number of Council of Europe states, some of which are currently debating how to deal with abuses committed in the past'. 214 Peace negotiations or transitional processes may involve amnesties/ pardons. If so, these measures may be challenged under the ECHR and actors may invoke the ECHR/ECtHR's practice in negotiations. 215 In ECtHR's cases involving amnesties, Mallinder and others jointly find that the ECtHR has provided very little guidance on what would be an appropriate reconciliation process and to what extent law would need to regulate it. 216 Nevertheless, they crucially point out that the ECtHR 'may look more favourably on amnesties that are designed to bring a conflict to an end and encourage a society to reconcile, particularly where they are accompanied by appropriate investigative processes and reparations for victims'. 217 As they remark, ECtHR's amnesty jurisprudence has seemingly experienced a change throughout time: from certain reluctance to considering the impact of amnesties on the promotion of conflict resolution or reconciliation to leaving the possibility that amnesties even for very serious crimes may be in certain circumstances acceptable, including a reconciliation process. 218 Pinto highlights that the ECtHR has not discarded that amnesties for war crimes could be acceptable under certain circumstances, including reconciliation. 219 If the ECtHR considers criminal prosecutions as the only manner to address gross rights violations, other accountability forms are excluded. 220 Nevertheless, Pinto appropriately remarks that other mechanisms such as truth and reconciliation commissions and civil proceedings as well as non-legal measures including commemoration, public inquiries, or legislative reparations should also be adopted if 'finding those responsible for abuses, preventing future crimes, restoring the rule of law and giving redress are the goals to be pursued with a view to human rights protection'. 221 Thus, the ECtHR should consider the specificities of each case to determine to what extent and how criminal law works alongside other measures in a particular socio-political context. 222 In Ould Dah v. France (2009), some ECtHR's considerations arguably involved the potential impact of amnesties/pardons on transitions to post-crisis or post-conflict societies. The ECtHR found that a Mauritanian amnesty law was not adopted after the applicant had been tried and convicted but it was given to prevent prosecution against him. 223 The ECtHR stated that a conflict between the need to prosecute offenders and a country's determination to promote reconciliation in society may emerge. 224 Nevertheless, it contextualised this general statement in light of Mauritania's situation: 'no reconciliation process of this type has been put in place in Mauritania'. 225 Concerning French courts' universal jurisdiction over a Mauritanian army captain accused of torture and who benefited from a Mauritanian amnesty, the ECtHR remarked that judicial interpretation constitutes part of the progressive development of national criminal law and is subject to a democratic regime and the rule of law 'which constitute the cornerstones of the Convention'. 226 In Association 21 December 1989 v. Romania (2011), the ECtHR invoked the right to the truth held by victims of serious abuses and their relatives. 227 Importantly, it contextualised these serious human rights violations committed by state agents in highlighting that: they were perpetrated as part of the widespread use of lethal force against civilians; and this excessive and lethal force was deployed during anti-government protests 'preceding the transition from a totalitarian regime to a more democratic system'. 228 This contextualisation of the Romanian society's transition to democracy is crucial to understand the applicants' opposition to the 2008 draft amnesty law related to crimes imputed to armed forces.
In Tarbuk v. Croatia (2012), the ECtHR to some extent indirectly considered Croatia's post-conflict transitional context when examining Croatian amnesty's effects on the applicant's rights. Thus, it examined this amnesty and Croatian jurisprudence, both of which explicitly refer to the armed conflicts in Croatia during the nineties. 229 Compared to Tarbuk v. Croatia, however, Margus v. Croatia provided by far much more detailed analysis of the impact of amnesties/pardons to achieve certain (legitimate) goals in transitional contexts. In Margus v. Croatia (2014), the ECtHR's Grand Chamber paid attention to arguments from academic experts who were third-party interveners in this case. As the Chamber summarised, these experts claimed that: (i) there were difficulties in negotiating international treaty clauses on amnesties (ICC-Statute and UN-Convention on Enforced Disappearance) and the 2012 UN-General Assembly's Declaration on the Rule of Law at the National and International Levels; (ii) amnesties have been the most frequently used transitional justice form; (iii) there has been relatively stable reliance on amnesties within peace accords worldwide ; (iv) several supreme courts worldwide upheld amnesties laws due to contributions thereof to achieving peace, democracy, and reconciliation; etc. 230 Furthermore, the ECtHR's Grand Chamber noted that although the interveners accepted that amnesties might cause impunity for those responsible for serious human rights violations and undermine the protection of such rights, they also remarked that 'strong policy reasons supported acknowledging the possibility of the granting of amnesties where they represented the only way out of violent dictatorships and interminable conflict'. 231 Thus, the interveners pleaded for a nuanced approach to amnesties rather than a total ban on them. 232 The Chamber remarked that: the interveners invoked the lack of state agreement concerning a total ban on amnesties for serious human rights violations, including those covered by the ECHR (particularly Articles 2 and 3); and they sustained that amnesties can be used as a tool in ending protracted armed conflicts that may lead to positive outcomes. 233 Nevertheless, the Chamber concluded that: […] the applicant was granted amnesty for acts which amounted to grave breaches of fundamental human rights such as the intentional killing of civilians and inflicting grave bodily injury on a child […] Even if it were to be accepted that amnesties are possible where there are some particular circumstances, such as a reconciliation process and/or a form of compensation to the victims, the amnesty granted to the applicant in the instant case would still not be acceptable since there is nothing to indicate that there were any such circumstances. 234 As examined, the ECtHR's Grand Chamber thus found that the decision of the Croatian authorities to convict the applicant of war crimes was consistent with state obligations under the ECHR and international law. 235 In this case involving international crimes, the Chamber seemingly gave more weight to the need that amnesties/pardons do not impede national criminal prosecution of gross atrocities. In their joint concurring opinion, Judges Ziemele, Berro-Lefèvre, and Karakaş remarked that serious human rights/humanitarian law violations should not end in amnesties/pardons. 236 Yet, the ECtHR's Grand Chamber considered the importance of amnesties/pardons to reach peace, reconciliation, etc. In the joint concurring opinion of Judges Šikuta, Wojtyczek, and Vehabović, there were also significant remarks about the potential impact of amnesties/pardons on processes towards post-conflict scenarios. These judges contextualised the scenarios where amnesties/pardons normally occur: 'Different countries have devised widely varying approaches enabling them to put grave human rights violations behind them and restore democracy and the rule of law'. 237 They reckoned that 'in certain circumstances there may be practical arguments in favour of an amnesty that encompasses some grave human rights violations' and indicated that they 'cannot rule out the possibility that such an amnesty might in some instances serve as a tool enabling an armed conflict or a political regime that violates human rights to be brought to an end more swiftly, thereby preventing further violations in the future'. 238 Moreover, they established that concerns related to guaranteeing effective human rights protection point 'in favour of allowing the States concerned a certain margin of manoeuvre […] to allow the different parties to conflicts engendering grave human rights violations to find the most appropriate solutions'. 239 Finally, in Makuchyan and Minasyan v. Azerbaijan and Hungary (2020), the ECtHR found that Azerbaijan did not adequately respond to a very serious ethnically-biased murder against an Armenian victim committed by an Azerbaijani soldier. 240 It established that 'the soldier had been pardoned for overtly political reasons relating to the ongoing Nagorno-Karabakh conflict with Armenia'. 241 It added that 'in view of the extremely tense political situation between the two countries, the authorities should have been all the more cautious, given that the victims of the crimes in the present case were of Armenian origin'. 242 The Court highlighted that the applicants claimed that the Armenian ethnic origin of the victims was the main reason for the crimes and 'subsequent actions of the Azerbaijani authorities, including the pardoning and glorification of the perpetrator'. 243 It also deplored statements of Azerbaijani officials supporting and congratulating the pardoned soldier for his crimes against Armenians. 244 Thus, the pardon, release, and promotion of the soldier did not promote peace or reconciliation between Azerbaijan and Armenia but, instead, worsened the relationship between these states and their respective societies.
Therefore, this section finds that the ECtHR has increasingly considered and discussed arguments on the (potential) impact of amnesties/pardons on helping to achieve or restore peace, reconciliation, democracy, and/or the rule of law in states and societies which transitioned or were still transitioning from situations of armed conflict, dictatorial regimes, or serious political turmoil. The ECtHR may be more prone to accept amnesties/pardons as tools helping a state to transition to peace, reconciliation, or democracy provided that: (i) the respective state demonstrates that these measures are genuinely adopted to achieve those legitimate aims rather than just serve as impunity mechanisms; and/or (ii) when those measures are accompanied with other transitional justice mechanisms such as some level of (criminal) accountability or reparations for victims.
To expand its level of deference to states, the ECtHR seemingly requires the existence of genuine and legitimate aims and/or complementary transitional mechanisms when examining the use of amnesties/pardons as peace-making, democracy building, or reconciliatory instruments in specific contexts. This jurisprudential approach arguably relates to that justice for victims of serious human rights violations/international crimes and related state obligations are at stake in societies transitioning from dark chapters of their recent history.

Conclusions
In the examined case-law, it is possible to identify ECtHR's several considerations that determine whether and to what extent the ECtHR has deferred to states in cases of amnesties/pardons for atrocities, namely, factors concerning or affecting the degree of ECtHR's deference to states. These jurisprudential considerations can arguably be grouped under four main factors that correspond to the examined cases and judgments, namely: (i) the national process of adoption, application, and/or validation of amnesties/ pardons; (ii) consideration of the margin of appreciation or discretion given to states; (iii) state compliance with international obligations on human rights; and (iv) potential impact on transitions to peace, reconciliation, democracy, and/or the rule of law.
Although the ECtHR has yet to conduct a direct, explicit, or stricto sensu judicial review of amnesties/pardons concerning serious human rights violations, the ECtHR has increasingly scrutinised and decided on the effects of amnesties/pardons on the ECHR rights of individuals who filed applications against states at the ECtHR. The said factors can and should be considered in future ECtHR's jurisprudence related to the effects of amnesties/pardons on the rights of individuals. Moreover, the ECtHR has indeed conducted its first indirect or implicit review of a pardon (Makuchyan and Minasyan v. Azerbaijan and Hungary). Importantly, these factors may and should apply if and when the ECtHR decides to review amnesties/pardons directly, explicitly, or stricto sensu.
The national process through which amnesties/pardons have been adopted, applied, and/or validated may be regarded as (usually) the starting factor considered by the ECtHR to determine its degree of deference to a state. This factor includes which organ granted these measures and through which normative instrument, whether these measures apply to serious abuses in national proceedings, and whether subsequent actions adopted by (later) national authorities negatively affected or changed originally admissible amnesties/pardons.
ECtHR's consideration of the margin of state discretion or appreciation has generally been the factor that has most expanded the degree of ECtHR's deference to a state in amnesty/pardon cases. This criterion is illustrated by the ECtHR's overall or prima facie respect for or acknowledgement of the: (inherently) discretionary nature of amnesties/pardons, fitness and authority of domestic authorities to interpret national rules better than the ECtHR can do, state's free determination of national criminal policies, state's legitimate interests, state's discretion on how to execute an ECtHR's judgment, etc.
Conversely, poor levels of state compliance with international obligations on human rights to investigate, prosecute, try, and punish atrocities when granting amnesties/ pardons generally require the ECtHR to constrain its deference to states. This occurs particularly (but not exclusively) when state officers committed atrocities, and concerning not only torture/ill-treatment but also (increasingly) murders. Besides the ECHR and its jurisprudence, the ECtHR has growingly invoked 'external' sources (e.g. IACtHR's practice), which generally reject amnesties/pardons. Overall, the ECtHR in several atrocity cases found that the effects of amnesties/pardons on human rights breached state obligations under the ECHR.
In turn, the impact of amnesties/pardons on achieving or pursuing legitimate goals of peace, reconciliation, or democratic transition when states adopt amnesties/pardons may increase the degree of ECtHR's deference to states as opposed to exemption measures seeking impunity, which constrain such deference. Thus, the ECtHR may accept amnesties/pardons in a specific context provided that: the state demonstrates that these measures are genuinely adopted to achieve the said legitimate aim(s) rather than impunity; and/or when these measures are accompanied with other transitional justice mechanisms such as reparations.
As noticed in the examined ECtHR's jurisprudence, these four factors can interact with each other in the same or opposite directions, which may in turn determine the degree of deference that the ECtHR is willing and able to accord to states. As a result of this interaction, it may be overall identified an emerging trend of potentially less deference to states and/or more conditioned deference to states in ECtHR's jurisprudence on amnesties/pardons related to serious abuses in recent years.
Therefore, this paper found factors that, to a greater or lesser extent, have expanded or contracted the degree of ECtHR's deference to states in cases involving amnesties/pardons for serious human rights violations, which in turn enables or helps to better identify jurisprudential developments and trends. By doing so, this article aims to contribute to the scarce existing literature on the said topic so that other researchers can consider these or similar factors (even different ones) as starting points or framework criteria to analyse ECtHR's existing and future amnesty/pardon jurisprudence more systematically and more comprehensively. The present article may mutatis mutandis and partially be also useful for researchers examining practices of other supranational bodies (and potentially domestic courts) regarding amnesties/pardons for gross abuses and related matters.