This guest post was written by Carole Lyons, Law School, RGU, Scotland
On 27 May 2014, a Grand Chamber of the ECtHR, in Margus v Croatia, pronounced upon the contentious issue of the use of amnesties in post-conflict settings. The case concerned a Croatian army commander who had been convicted of several murders of civilians in 1991. He had benefited from an amnesty in relation to the murders in 1997 but in 2007 was convicted of war crimes. Just two months before Croatia became a signatory to the European Convention on Human Rights (ECHR) in November 1996, the Croatian Parliament had passed a Law on General Amnesty. Under the provisions of the latter, immunity from prosecution was granted in relation to crimes committed during the war which took place between 1991 and 1995 after Croatia’s declaration of independence from the former Socialist Federal Republic of Yugoslavia.
The resort to institutionalized forgetfulness and amnesty in conflicted entities and troubled communities is neither new (for example, all of the army personnel involved in the wrongful conviction of Alfred Dreyfus in France in 1894 were granted amnesty from prosecution) nor exceptional in Europe (see the continuing enforcement of Spain’s 1977 Amnesty Law) and such resort was not uncommon in the Western Balkans after the 1990s conflicts. However, the recourse to amnesty is a very divisive issue in itself (see the recent revelations about the so-called “on-the-run letters” in Northern Ireland), pitting those who see the need for reconciliation and the need to prevent the past from endlessly haunting communities against victims and survivors seeking justice and reparation (for example, the family of Luis Dorado Luque seeking compensation for his death in 1936 at the ECtHR).
On an institutional level, the European Parliament issued a Resolution in 1993 stating that there “should be no question of impunity for those responsible for war crimes in the former Yugoslavia”. In 2012 the EU’s External Action Service warns about a “culture of impunity” in transitional justice contexts but in April 2014 appears to support amnesty in Ukraine. Also in 2014, the UN Special Rapporteur on transitional justice recommends withdrawal of the 1977 Amnesty Law in Spain. In Strasbourg, there is conflicting case law on amnesty; in Ould Dah (2009) the Court stated that “an amnesty law was generally incompatible with the States’ duty to investigate acts of torture or barbarity.” In Tarbuk (2012) (which related to the Croatian General Amnesty Act) the Court stated that “even in such fundamental areas of the protection of human rights as the right to life, the State is justified in enacting any amnesty laws it might consider necessary” (§ 50).
Facts and National Proceedings
The case originates in December 1991 – a turning point for Europe. On 10 December, two civilians are murdered in the Vrbik forest by Fred Marguš, a Croatian army commander, near his hometown of Osijek in eastern Croatia. On the same day, in Maastricht, the European Council is meeting to finalise the Treaty which will instigate closer and political union, the high point of post WWII peaceful cooperation in Europe. There is no mention of Fred Marguš, of Croatia or of contiguous war crimes in the Council’s optimistic Conclusions. But EU closer union and the Vrbik forest murders will coalesce, 23 years later, in Strasbourg (one year after Croatia joins the EU).
Criminal proceedings against Marguš (inter alia for the Vrbik forest murders and the murders of other civilians including the SB family) are instigated in 1991 based on the Croatian Criminal Code but are terminated on 24 June 1997, pursuant to the 1996 Amnesty Act, by the Osijek County Court, without the accused ever having been tried for any of the offences alleged. The County Court does not consider whether or not Marguš’s offences might have constituted war crimes – which would have precluded the application of amnesty (1996 Act 3 (1)). Instead, it categorises his actions as “being closely connected to the armed conflict” and therefore covered by General Amnesty (1996 Act 3 (2)). Furthermore, the County Court judges mention the exceptional courage Marguš showed during late 1991 when the murders took place.
The initial post-conflict reluctance in Croatia towards indicting its own citizens in relation to the homeland war alters and in, 2007, Marguš is convicted of war crimes against civilians under the Geneva Convention 1949 (a conviction upheld by both Supreme Court (2007) and Constitutional Court (2009). The Supreme Court’s approach adds to the confusion in this convoluted case; on 19 November 2007 it reverses the June 1997 amnesty decision because there was found to be no link between the alleged crimes and the 1991-1995 armed conflict. On the same day the court upholds Marguš’s conviction for war crimes. But the link between Marguš’s amnesty and war crimes was not directly considered by any of the three levels of the Croatian judicial system. Yet, it is precisely this issue which dominates the two Strasbourg judgments in this case.
Judgment and Comment
The devil is in the detail here and that persists through to Marguš’s application to the ECtHR in 2009 where his claim is based inter alia on the right not to be tried or punished twice (non bis in idem) enshrined in Article 4 of Protocol No. 7 to the Convention. Paragraph 2 of Article 4 contains an exception to the right not to be tried twice “if there has been a fundamental defect in the previous proceedings, which could affect the outcome of the case” (emphasis added). Formally, this exception forms the kernel of the November 2012 Chamber judgment in Marguš with the finding (§ 75) that there was a ‘fundamental defect’ by the Osijek County Court in June 1997 when it terminated the trial proceedings under the 1996 Amnesty Act. In short, the Chamber judges assess the two cases (1997, 2007) as concerning the same offences but apply the Article 4 (2) ‘fundamental defect’ exception to find against Marguš.
But there is a distortion here; according to the Chamber, the ‘fundamental defect’ for Article 4 (2) purposes resides in the fact that the amnesty was granted in respect of war crimes (§§ 72 – 76, citing Ould Dah). However, Marguš’s amnesty was not in fact granted on this basis and his offences came to be categorised as war crimes only 10 years later. This judgment effectively chastises the County Court retrospectively for not having considered the possibility that war crimes were relevant in 1997. But this means that the ‘fundamental defect’ is incorrectly identified by the Chamber as the latter lay not in the act of granting an amnesty but rather in the omission of consideration of war crimes relevancy.
Moreover, the overt attempt by the Strasbourg judiciary to generate an amnesty law precedent out of this case (against the background of cases such as Tarbuk (amnesty supported) and Gutierrez Dorado (amnesty not considered)) is highlighted by the statement that “Granting amnesty in respect of “international crimes” – which include crimes against humanity, war crimes and genocide – is increasingly considered to be prohibited by international law.” (§ 74) A considerable amount of both the Chamber and Grand Chamber judgments is taken up with detailed comparative analysis of the current law on amnesty. This transforms the case from one merely limited to Article 4 Protocol 7 (the main basis of Marguš’s claim) to being, specifically, a covert criticism of Croatian post-conflict amnesty policy and practice and, generally, a Strasbourg position on amnesty law for future post-conflict settings.
Leaving aside a predictable judge made policy argument here, the amnesty law position might be more persuasive if there were not quite so many cracks in the whole Marguš edifice. The Grand Chamber judgment of 27 May 2014 is notable in several respects, both formal and substantive. There were three changes of judge during the course of the case; the Croatian judge was replaced by the Latvian judge after the hearing; the Bulgarian and French judges both withdrew from the case and were replaced by substitutes. In addition to the majority judgment there are five separate opinions, four (nominally) concurring involving altogether ten judges and one partly dissenting (one judge). (This rather calls into question the maths of majority in the court on the Ill.) In a judgment of 44 hard copy pages, 33 of those are taken up with a description of the Croatian facts and proceedings and analysis of amnesty law. There are two issues ancillary to Marguš’s main Article 4 Protocol 7 claim which are quickly dismissed, namely Article 6 based arguments (the same judge had been on the 1997 amnesty court and the 2007 war crimes conviction; Marguš had been removed from court during the 2007 war crimes trial) and a question of temporal jurisdiction (the ECHR entered into force in Croatia after the initial June 1997 amnesty decision) answered by reference to the date (2007) of the ‘second’ set of proceedings.
As to the main claim of a breach of Article 4 Protocol 7, the Grand Chamber comes to a decision contrary to that of the Chamber; the ultimate finding is that this claim is inadmissible in respect of two of Marguš’s 1991 murder offences (Vrbik forest) and inapplicable in relation to three other 1991 offences (the SB family – two murders and the wounding of a child). The complex and messy history of indictments, withdrawals, amnesties, re charging… in the Croatian judicial system thus feeds directly into this fragmented assessment with the Grand Chamber distinguishing the Vrbik offences (withdrawn) from the SB family offences (amnestied). But although Marguš clearly loses his case there is a shift in the Grand Chamber’s reasoning which renders this more of judgment against and about Croatia than a mere assessment of the rights of an accused. The judges introduce new arguments, not raised by the parties, and rely on Articles 2 and 3 of the Convention to state that, in general, granting amnesty in relation to the killing of civilians constitutes a breach of a State’s fundamental human rights obligations. However, the 2007 war crimes conviction of Marguš demonstrated Croatia’s compliance with these provisions (§§ 127, 139 and 140) and for that reason Protocol 7 is deemed not applicable. This interjection is justified by the need for the Convention and Protocols to be read as a whole (§ 128).
Is the foot being reshaped to fit the shoe here? It might be argued that the Court, finally, felt obligated – in the face of an army commander amnestied (and praised) for civilian murders – to make a statement on amnesty law against a background where “so far no international treaty explicitly prohibits the granting of amnesty in respect of grave breaches of fundamental human rights.” (§131). The Court ends, what has all the signs of a rather tortuous decision making process, by stating that amnesties which are accompanied by reconciliation and compensation processes may be possible despite the growing tendency against amnesty in international law (§ 139).
This statement in itself is a significant contribution to the emerging thinking on transitional justice in Europe with the Court essentially stating that Convention jurisprudence could endorse amnesty laws as long as they are part of a wider transitional justice set of processes. But there are many questions left unanswered by the judgment such as the use of Articles 2 and 3 instead of Protocol 7, the lack of detailed discussion of the relevance of Article 4 (2) Protocol 7 and the issue of what might constitute a ‘fundamental defect’ under that provision. Significantly, the precise circumstances in which an amnesty would be acceptable under Convention law even though it might be unacceptable in international law are not clarified in this judgment.
On the face of it, the Marguš case has the semblance of another appeal to the ECtHR to resolve rights abuses dating from post 1989 conflicts. The Strasbourg judiciary are increasingly (see the work of Eva Brems and the work of Antoine Buyse and Michael Hamilton) being faced with high profile European historical crimes and rights abuses (Katyn, Kononov). Even though the Court has explicitly stated “that it is not its task to settle possible points of debate among historians” it has become, de facto, a forum for the processing of many reparations originating in the past. This arguably points to institutional and theoretical deficits in terms of transitional justice processes. The Strasbourg tribunal and its Convention are not equipped – structurally, philosophically, institutionally – to deal with the complex intermingling of law, politics, rights abuse, bitter memories, divisions, desperation, discriminations, quotidian concerns alongside crimes against humanity… which constitute post conflict situations. (European) human rights law was not envisaged as a “super glue” solution for the many, varied and long lasting vestiges of war and conflict scenarios. (See further, Kieran McEvoy on the general need to revisual the relationship between traditional legal approaches and emerging transitional justice methods and thinking.) However, the dearth of alternative forums and comprehensive transitional justice principles in the European legal space has seen the evolution of a culture and practice whereby the legalistic road to Strasbourg is taken by many “victims” (both actual victims and perpetrators) of past conflicts. The Strasbourg court appears to have accepted this mantle of default transitional justice forum but the confused messages emanating from Marguš testify to Strasbourg’s laudable but restricted ability to handle the wide variety of post-conflict disputes.
For all the complexities and lack of clarity in this judgment, it is no doubt destined to become a much debated Strasbourg precedent on amnesty law (perhaps until (inevitably) the Ukraine conflict is processed). But the Marguš case, like all high profile cases involving the abusers rather than the abused, ultimately eclipses those who were the real victims in this case. There were many victims from many communities in the Croatian conflict but one of the abiding images of that time is of the Croat Serb exodus away from Operation Storm in 1995 with the roads clogged with tractors. Small people, making significant law…
*Now as news come in of each neighbourly murder, we pine for ceremony, customary rhythms: the temperate footsteps of a cortege, winding past each blinded home. S Heaney, Funeral Rites, North (1975, Faber)
Section 3 of the General Amnesty Act has two exclusions; (1) no amnesty in the case of serious violations of humanitarian law, war crimes and genocide; (2) no amnesty in the case of criminal offences not connected with the armed conflict in Croatia.