Aurélie Van Baelen, researcher at the Human Rights Centre (University of Ghent)
On 16 February 2021, the European Court of Human Rights (ECtHR) delivered its long-awaited judgment in Hanan v. Germany (application no. 4871/16). The ruling presents another episode in the saga of cases regarding States’ conduct, and more specifically their compliance with international human rights law, during military operations abroad under the auspices of an international organisation. Facing no fewer than eight intervening governments, showing the importance of the case at hand for troop-contributing States, the ruling is – unsurprisingly – one of give and take.
As I will argue in this blog post, the judgment is remarkable for finding extraterritorial jurisdiction for the purposes of Article 1 of the European Convention of Human Rights (ECHR) and for explicitly reconfirming a State’s duty to investigate civilian casualties in correspondence to human standards. The Court’s finding of a non-violation of the procedural limb of Article 2 ECHR however is unfortunate, but expected. The Court delivered a strategic judgment, in which it attempts to not dissuade States to engage in multilateral military operations, or to institute domestic investigations into deaths occurring during military operations abroad, on the one hand, and to protect fundamental rights on the other. A carefully constructed balance which, regrettably, has tilted in favour of the former, which for airstrike victims, is the umpteenth disappointment on their quest to justice.
Background & Facts
In the aftermath of the 9/11 attacks, the United States together with the United Kingdom launched military operations in Afghanistan under the umbrella of ‘Operation Enduring Freedom’. As part of this operation, Germany agreed to the deployment of up to 3,900 German soldiers. A couple of months following the launch of this operation, in the beginning of December 2001, prominent Afghan leaders met in Bonn under the auspices of the United Nations, where they set up an Afghan Interim Authority. The so-called Bonn Agreement also foresaw the establishment of the International Security Assistance Force (ISAF) to maintain the security in Kabul and surrounding areas. In 2003, NATO assumed command of ISAF and the latter’s mission, to maintain security, was gradually expanded until ISAF was responsible for all of Afghanistan. At the relevant time, a unit of approximately 1,500 German soldiers (named ‘PRT Kunduz’) was stationed in the Kunduz province under the command of Colonel K.
On 3 September 2009, Taliban fighters hijacked two fuel tankers. Later that day, the tanks got stuck on a sand bank in the Kunduz river. The Taliban insurgents tried to remobilise the tanks by asking nearby villagers to siphon off some of the fuel. An informant tipped of the German troops about this incident around 8 p.m. Upon receiving this information, Colonel K. ordered a surveillance aircraft to locate the tankers. At around 12:30 a.m. the informant reported the partial emptying of the tankers as well as the absence of civilians at the sand bank.
After confirming with the informant that only insurgents and no civilians were present in the area, Colonel K. ordered two F-15 aircrafts to bomb the fuel tankers. At 1.49 a.m. two 500 pound bombs destroyed the two fuel tankers and killed a number of civilians. The exact number of victims has never been established. Official reports deem the number of deceased insurgents between 13 to 132. The estimation of deceased civilians ranges from 14 to 142. Among the civilian casualties were Adbul Bayan Hanan and Nesarullah Hanan, the applicant’s sons who were twelve and eight years old, respectively.
Investigations into the airstrike
Later that same morning, an investigation team of the German military police was sent from Masar-i-Sharif to Kunduz to support PRT Kunduz with the investigation into the airstrike. Before the team from Masar-i-Sharif arrived, a team from PRT Kunduz already paid a visit to the sand bank and came across an extensively altered site: no bodies were found and only a few destroyed cars remained. During the course of the next few days, members of PRT Kunduz together with the military police interviewed some villagers regarding the airstrike. Colonel K. was involved in the interviews and was kept up-to-date regarding the investigation.
Parallel with the investigation in Afghanistan, a preliminary investigation was conducted in Germany by the Dresden Public Prosecutor General in cooperation with the Federal Ministry of Defence. The domestic investigation was later taken over by the Federal Prosecutor General who opened a criminal investigation against Colonel K. in light of his possible liability under the Code of Crimes against International Law. Eventually, the criminal investigation was discontinued due to a lack of sufficient grounds for suspicion. The Federal Prosecutor General concluded that Colonel K. did not have the intent to kill or harm civilians and had assumed that only Taliban fighters were present on the site, who were a legitimate target, as were the fuel tankers. Therefore, he was not liable.
The applicant, Mr. Hanan, as the next-of-kin of the two victims was not involved in the investigation. Neither in Afghanistan, nor in Germany. He also only got (limited) access to the investigation file, after multiple requests and after the investigation had already been closed. This was the start of a legal battle between Mr. Hanan and the German authorities. Mr. Hanan’s numerous attempts to continue to reopen the investigation into Colonel K.’s criminal liability fell foul.
Together with the intervening governments of Denmark, France, Norway, Sweden and the United Kingdom, Germany contested its extraterritorial jurisdiction in Afghanistan, and thus the Court’s competence ratione loci. The governments argued that Germany exercised neither ‘effective control over an area’ nor ‘State agent authority and control’ (§104, see Al-Skeini and Others v the UK or Jaloud v the Netherlands). Relying on the Court’s judgment in Banković and Others v Belgium and Others, they also stated that the airstrike did not establish a jurisdictional link between the respondent State and the victims as it was an ‘instantaneous extraterritorial act’ and Article 1 of the Convention does not admit to a ‘cause and effect’ notion of jurisdiction (§107). The intervening governments further deemed the mere institution of investigative measures by the German authorities not sufficient to establish a jurisdictional link, and consequently to make complaints under the procedural limb of Article 2 ECHR (§108).
The Court confirms that instituting a domestic criminal investigation into any death which has occurred anywhere in the world is in itself insufficient to establish a jurisdictional link. It also recognises that deciding the contrary could create a ‘chilling effect’ on instituting investigations at the domestic level, which would result to an inconsistent application of the Convention (§135). The Court then relies on the ‘special features’-test as established in Güzelyurtlu and Others v Cyprus and Turkey to find jurisdiction in the present case. This reliance is rather surprising as the Court stresses, throughout its reasoning, the different circumstances in Güzelyurtlu compared to the present case. Nevertheless, the Court mentions the possible existence of ‘special features’ with regards to Hanan which could establish a jurisdictional link bringing into effect the procedural obligation imposed by Article 2 ECHR, even in the absence of an investigation or proceeding having been instituted by a State into a death which has occurred outside its jurisdiction (§136). The Court further confirms that this ‘special features’-assessment also applies in extra-territorial situations outside the legal space of the Convention as well as in respect of events occurring during the active hostilities phase of an armed conflict.
The Court then continues by identifying three ‘special features’ with regards to Germany which, cumulatively, establish a jurisdictional link. Firstly, Germany was obliged under customary international humanitarian law to investigate the airstrike as it concerned the individual criminal responsibility for a potential war crime of German armed forces (§137). Secondly, as a result of an ‘exclusive jurisdiction’ provision in ISAF’s Status of Forces Agreement, Afghan authorities were prevented from instituting a criminal investigation (§138). Thirdly, German prosecution authorities also had an obligation under their domestic law to institute a criminal investigation, a direct consequence of Germany’s ratification of the Rome Statute of the International Criminal Court (§138). Accordingly, the combination of these ‘special features’ trigger the existence of a jurisdictional link for the purposes of Article 1 of the Convention in relation to the procedural obligation to investigate under Article 2 (§142).
After establishing the jurisdictional link and confirming its competence ratione loci, the Court proceeds to examine the alleged ineffectiveness of the investigation into the airstrike that killed inter alios the applicant’s two sons. The Court reiterates that the procedural duty under Article 2, in cases such as the present one, must be applied realistically. Therefore, while the investigation should be in accordance with the principles of effectiveness (§202), thoroughness (§203), promptness and independence (§207) and with a sufficient element of public scrutiny (§208), the challenges and constraints for the investigation authorities due to the context of an armed conflict must be taken into consideration (§200).
As for the investigation’s adequacy, the Court notes that the German domestic authorities did not have the competence to undertake investigative measures in Afghanistan. Despite this incompetence, the Federal Prosecutor General had a considerable amount of material from different sources concerning the circumstances and impact of the airstrike to its disposal. The Prosecutor also conducted interviews with Colonel K. and other soldiers present at the command centre at the relevant time (§215). The Court therefore sees no reason to doubt the Federal Prosecutor General’s assessment regarding Colonel K.’s criminal liability (§216). As for the fact that to date, the precise number of victims remains unknown, the Court acknowledges that this constitutes an essential element under normal circumstances but accepts that the circumstances of active hostilities, as well as the social and religious mores of the population prevented a more accurate assessment (§218).
Regarding the standards of promptness and independence, the applicant complained about a delay and a lack of independence in relation to the on-site reconnaissance (§223). While the Court concurs with the applicant with regards to the possibility of performing an additional reconnaissance with a drone prior to the on-site visit, it does not see how this could have resulted in gleaning information beyond what could have been established through the inspection by an unmanned aircraft in the early morning following the airstrike. As for the element of independence, the Court agrees that the initial on-site would preferably not have done exclusively by members of PRT Kunduz, and that Colonel K. should not have been involved in the investigative steps in Afghanistan, given that the investigation concerned his own responsibility in connection with ordering the airstrike. Nevertheless, the Court cannot conclude that this rendered the investigation ineffective. The Court based this finding primarily on the fact that the Federal Prosecutor General relied mainly on Colonel K.’s mens rea at the time of the airstrike for determining that the Colonel was not criminally liable for the attack. Since Colonel K.’s mens rea was corroborated by evidence which could not be tampered with (e.g. audio recording of the radio traffic between the command centre and the pilots of the F-15 aircrafts), there was no risk of contamination or unreliability of the evidence that was used for the decision of criminal liability (§227).
Lastly, Mr. Hanan complains about the fact that he was not granted access to the file until after the investigation had been closed. The Court acknowledges that at first sight, this might seem problematic since the applicant, as the father of two persons killed by the airstrike, could have possessed relevant information (§230). However, in the circumstances of the present case, the Court found that this did not render the investigation deficient (§231). Referring to Colonel K.’s mens rea, the Court does not doubt that the applicant would not have been able to provide additional insights that would have been relevant to the decision of the Colonel’s criminal liability.
In view of the foregoing, and with regards to the specific circumstances of the case taking place in the context of an armed conflict, the Court finds that there has been no violation of the procedural limb of Article 2 of the Convention.
Separate opinion of Judges Grozev, Ranzoni and Eicke
While the Court unanimously found that there had been no violation of the Convention, Judges Grozev, Ranzoni and Eicke dissented on account of the case’s admissibility. They disagree with the majority’s extended application of ‘special features’ to events which have taken place outside the legal space of the Convention. In the present case, the applicant only complained about an alleged violation of the procedural limb of Article 2 ECHR, not about a substantive breach. Nevertheless, to date, the Court had always first assessed whether the incident itself fell under a State’s jurisdiction within the meaning of Article 1 of the Convention, and if so, whether it was attributable to the State concerned. The Court did this consistently, also when the applicant did not complain about a substantive breach of Article 2 (see Ukraine v Russia (re Crimea), Al-Skeini, Jaloud). When applying the Court’s established principles in relation to the exercise of jurisdiction outside the territory of the Contracting States, namely ‘State agent authority and control’ and ‘effective control over an area’, the dissenting judges conclude that the death of the applicant’s sons did not occur within the ‘jurisdiction’ of Germany. Thus, by applying the ‘special features’ in the present case, the dissenting judges fear that the Court has created a procedural duty to investigate a loss of life, which is expressly not ‘attributable’ to the Contracting State, in relation to which that State has no substantive Article 2 obligation nor even jurisdiction.
The dissenting judges are furthermore unable to discern what makes the three ‘special features’ so ‘special’ in this case. The first so-called ‘special feature’ consists of Germany’s obligation under international law to investigate the airstrike. This obligation is largely based on customary international law and mainly flows from the State’s practice and obligations in light of the Rome Statute, which is common to the majority of Contracting Parties to the Convention. The second ‘special feature’ concerns the exclusive jurisdiction provision in ISAF’s States of Forces Agreement, which is a ‘common practice for troop-contributing States in United Nations-authorised military missions’ (§138). The third and last ‘special feature’ consists of the fact that the German prosecution authorities were obligated under domestic law to institute a criminal investigation, which is a direct consequence of the State’s ratification of the Rome Statute. Again, this is likely to be a common feature rather than a ‘special one’ amongst the majority of Contracting Parties to the Convention. In sum, to the dissenting judges it remains unclear what makes the combination of these ‘special features’ so ‘special’ so as to justify creating ‘jurisdiction’ where it would not otherwise exist.
First things first. Allow me to quickly guide you through the first issue the Court had to tackle: the never-ending story of ‘jurisdiction’ when it comes to situations of deployment of military troops abroad under the auspices of an international organisation. The ‘jurisdiction’ concept, as well as the exceptions that have been formulated by the Court, seem quite straightforward. However, when dealing with grey areas, the jurisprudence demonstrates how rapidly the straightforwardness of jurisdiction (or not) disappears.
It all started in 1999 with the well-known judgment in Banković. Here, the Grand Chamber declared the case inadmissible on the grounds that no jurisdiction is exercised through extraterritorial airstrikes. There were no ‘boots on the ground’ in the area, and the Court explicitly rejected a ‘cause and effect’ jurisdiction or ‘effective airspace control’ that would originate from the ‘mere’ launch of airstrikes (§75). In 2007, the Grand Chamber was confronted anew with victims due to military conduct abroad in Behrami Behrami v France and Saramati v France, Germany and Norway. Again, the Court ruled the case inadmissible. This time around, the Grand Chamber attributed the impugned acts to the UN. Hence, there was a lack of jurisdiction ratione personae (§152). This judgment was widely criticized and allegedly nudged to Court to turn the tide in Al-Jedda v UK and Al-Skeini v UK. At this time, the Court attributed the deaths of Iraqi citizens to the UK and established jurisdiction. Admittedly, the military presence of the UK in Iraq, as an occupying power, made it a tad easier for the Court to hold that the UK had ‘effective control over an area’. The personal nexus was based on detention and respective exchange of fire during a military patrol. However, the case of Jaloud presented the Court with another conundrum. An Iraqi citizen was shot by a Dutch soldier when passing through a Dutch military checkpoint, however the UK exercised full operational command in the area, rendering the territorial control far weaker compared to Al-Skeini. Nevertheless, by referring to the ‘full command’ the Dutch officers retained over their military personnel, the Court found territorial control (§143). The passing through a Dutch checkpoint provided the personal nexus.
And then, there was Hanan. The facts presented to the Court contained a myriad of elements which to some extent resembled the cases it had previously dealt with, and that in certain aspects also resembled those cases where the Court had not accepted jurisdiction. As in Banković, the impugned act was an airstrike. However, contrary to Banković and like Al-Skeini and Jaloud, there was a military presence of German troops on the ground. More so, the deployed German soldiers outnumbered the Taliban fighters in the area. Furthermore, similar to Behrami and Al-Jedda, ISAF was established under a UN Resolution, on the basis of which it provided security in the area. In this regard, Hanan more resembled Al-Jedda since the troops remained under the command of the troop contributing State (i.e. Germany) and not the Secretary-General as was the case in Behrami. Lastly, similar to Jaloud, ISAF had checkpoints in place in the area and conducted daily military patrols. However, unlike Jaloud, the deaths were not caused by soldiers when the victims were passing through one of the checkpoints, bringing us back to the airstrike, consequently the issue of ‘cause and effect’ jurisdiction.
Rather than staying true to assessing ‘effective control over an area’ and ‘State agent authority and control’, the Court – surprisingly – turns to identifying ‘special features’ in order to determine Germany’s jurisdiction. Here the Court goes into unchartered waters. Admittedly, the dissenting judges certainly make a relevant point when arguing that they do not discern what makes the identified ‘special features’ in this case so ‘special’. Perhaps it would have been more desirable if the Court would have followed its own case-law and, accordingly, its established principles. Contrary to what the dissenting judges hold, it would not have been impossible to establish a jurisdictional link when doing so. The Court could very well have argued that the combination of the airstrike and Germany’s permanent military presence (incl. checkpoints, daily military patrols), was sufficient to conclude that Germany had ‘effective [territorial] control over [the Kunduz] area’. The personal nexus could have been based on Colonel K.’s order to launch the airstrikes. It would have been a less surprising move, and it would not have raised too many eyebrows. Especially since the Court has previously decided that jurisdiction was established whenever some elements of both the personal model and the territorial model were present in a certain case. Now, the finding of jurisdiction might seem more fabricated.
Another surprising move of the Court was not discussing whether the deaths were attributable to Germany. Admittedly, the applicant did not raise the substantive limb of Article 2 ECHR in his complaint, nevertheless the Court rigorously made such an assessment (regardless of the claims) in previous cases. A calculated move since attributability of the act would have been difficult to establish? Possibly. Be that as it may, the applicant’s strategic bet to only complain about Article 2’s procedural limb has paid off. Whereas the Court’s manoeuvring might have come as a surprise, it is also remarkably positive. It demonstrates the Court’s cautious willingness to expand the Convention’s application to situations which were, until now, at the limits of its reach. Worth mentioning in this regard, is also the fact that of the 17 judges, only three dissented with respect to the Court’s jurisdiction. This in itself should be perceived as a (small) achievement.
Equally positive is the Court’s reconfirmation of a State’s duty to comply with the human rights investigation standards (effective, thorough, independent and prompt), even in hostile environments or during active hostilities. It follows that (hopefully) in the future, it will be rather difficult for States to hide behind the fog of war when failing to at least conduct an investigation compliant with the aforementioned human rights principles (albeit interpreted in a flexible manner taking into account challenges and constraints stemming from the fact that the deaths and subsequent investigations occur in a hostile environment).
In the same way that a positive light is cast by the ruling regarding jurisdiction and the investigation standards, so too is its shadow by way of the disappointing conclusion of a non-violation of Article 2 ECHR. Admittedly, the circumstances on the ground during active hostilities are not to be underestimated. That does not change however, that some very basic principles of effective investigation were not complied with by Germany. The Court touches upon all of them: the unknown number of victims, the tardiness of the Masar-i-Sharif investigation team, the involvement of Colonel K. in the investigation process, the hierarchy between Germany’s Federal Prosecutor General and the Ministry of Justice, the political stakes involved, not involving the applicant – a possible key witness – in the investigation process, and not interviewing key witnesses altogether. But the Court presents itself remarkably lenient towards Germany and is ‘ready to accept’ that these elements do not amount to an ineffective investigation. A lenience which is particularly remarkable when recalling how strict the Court was towards the Netherlands in Jaloud.
The Hanan-judgment presents a good example of the Court’s strategic manoeuvring in a delicate matter. It is unsurprising that this judgment is one of give and take. Indeed, while the judgment certainly contains remarkably positive elements, it also demonstrates how the Court remains cautious. Despite the Court’s cautiousness, this ruling has resulted in a win regarding the Convention’s extraterritorial applicability. Taking this into account, it seems as if the Court is ready to accept that Member States cannot do abroad, what they cannot do at home. However, we are not entirely there yet, because simultaneously, the judgment constitutes another disappointment for airstrike victims. The Court has come a long way since Banković and Behrami Behrami, but for those seeking justice the road ahead is, in all probability, still a lengthy one.