Strasbourg Observers

Romeo Castaño: “meticulously elaborated interpretations” for the sake of prosecution

September 10, 2019

By Mattia Pinto, PhD Candidate at the London School of Economics, Department of Law

 On 9 July 2019, the Second Section of the European Court of Human Rights (ECtHR or the Court) delivered its judgement in Romeo Castaño v. Belgium, concerning Belgium’s failure to execute multiple European Arrest Warrants (EAWs) issued by Spanish authorities in relation to a suspected ETA terrorist. Extradition cases involving the ECtHR usually concern complaints of ill-treatment likely to occur if an individual is extradited to a country where human rights appear not adequately protected (see, e.g., Soering, Trabelsi, Othman (Abu Qatada) and Pirozzi). In the case here at issue, the situation is reversed: the applicants complained that Belgian authorities’ refusal to surrender would amount to a breach of their right to an effective investigation into their father’s murder. The Second Section accepted this complaint and ruled unanimously that Belgium had breached its procedural obligation to cooperate under Article 2 ECHR. It is the first time the Court has found a violation of the Convention because a State refuses to surrender an individual sought by an extradition request. The decision is interesting but also controversial in its attempt to engage with multiple and complex issues, involving the relation between the ECHR and EU law, positive obligations to prosecute human rights violations and the principle of non-refoulement in EAW requests. In my opinion, the Court tries but eventually fails to properly deal with these issues.


In 1981, Colonel Ramón Romeo was murdered by an ETA commando. One of the suspects, N.J.E., fled to Mexico and, subsequently, settled in Belgium. In 2013, Spanish authorities issued two EAWs to have N.J.E. surrendered to Spain so that she could be prosecuted (trials in absentia are not possible in Spain). However, Belgium refused to execute the request due to a concern that the suspect would have been exposed to a risk of human rights violations. The decision was justified on the basis of “Spain’s contemporary political history” (Romeo Castaño § 12. All the translations are mine) and a Report issued in 2011 by the European Committee for the Prevention of Torture. The latter, in particular, criticised the Spanish legal system for issues of incommunicado detention and various forms of ill-treatment for terrorist suspects. In 2016, upon a second EAW, Belgium again refused to surrender N.J.E. to Spain on similar grounds. The Belgian authorities cited the UN Human Rights Committee’s Sixth Periodic Report of Spain (2015), which reiterated concerns at incommunicado detention for terrorist suspects. The children of Colonel Romeo then filed an application with the ECtHR alleging violations of Article 2 (right to life) and Article 6 (Right to a fair trial). The Second Section decided to focus only on the complaints under Article 2, namely that the Belgian authorities’ refusal to execute the EAWs “made in fact impossible” the prosecution of their father’s alleged murderer (Romeo Castaño § 66).


In the judgment, the Court found a violation of Belgium’s procedural obligation to cooperate arising from the procedural aspect of Article 2 ECHR. The Second Section dismissed Belgium’s argument that N.J.E.’s surrender to Spain would be contrary to Article 3 ECHR (Prohibition of torture). On the Court’s account, the Belgian authorities did not provide a sufficient factual basis to substantiate their refusal to surrender the alleged murderer.

Let’s unpack the decision a little more. Spain requested Belgium’s cooperation in the context of the EAW. Relying on Güzelyurtlu and others (at §§ 232-235), the Court considered that it should examine (1) whether the Belgian authorities responded properly to the request for cooperation; and (2) whether the refusal to cooperate was based on legitimate grounds (Romeo Castaño § 82).

As to point (1), the Court observed that Belgium had provided Spain with a properly reasoned response. In the opinion of the Second Section, the approach taken by the Belgian courts was compatible with the ECtHR’s decision in Avotiņš and Pirozzi, namely that, upon execution of an EAW, the EU “principle of mutual trust” should not be automatically applied to the detriment of fundamental rights (Romeo Castaño §§ 83-84).

As to point (2), the Court held that a risk of ill-treatment for the person whose surrender is requested “may constitute a legitimate ground for refusing to execute the EAW”; yet, the assessment of such a risk should have sufficient factual basis (ibid § 85). In this regard, the Court noted that the Belgian authorities had used outdated reports and had neither conducted an individual assessment of the case at hand nor identified any structural shortcomings regarding the conditions of detention in Spain (ibid § 86). The Second Section also highlighted the fact that Belgium had not requested from Spain any additional information concerning the application of the prison regime in N.J.E.’s case, as it would have been possible under Belgian law (Romeo Castaño § 89). For these reasons, the Court was able to find that the assessment carried out by the Belgian courts was not “sufficiently complete” for it to conclude that the risk of ill-treatment for N.J.E., in case of surrender, had a “sufficient factual basis” (ibid § 90). Therefore, Belgium was in breach of its duty to cooperate.

Finally, the Court was clear that the finding of a violation did not necessarily imply that Belgium has to surrender N.J.E. to Spain (ibid § 92). According to the Second Section, its judgment should not be read as diminishing the States’ obligation to verify that a person is not extradited to a country where he or she may run the risk of a treatment contrary to Article 3 ECHR (ibid).

Judge Spano attached a concurring opinion, joined by Judge Pavli, which highlights the importance of “symmetry” between the ECHR and EU law.


As others have already argued (here and here), Romeo Castaño is certainly an important decision for the law of extradition, the relation between the ECHR and EU law and the doctrine of positive obligations under the Convention. However, its very concise reasoning hides some assumptions and unspoken premises that make Romeo Castaño problematic in several ways. I will focus here on three of them, namely i) the relation between the ECHR and EU law; ii) the duty to cooperate in connection with the state’s duty to investigate and prosecute serious human rights violations; and iii) the relation between procedural obligations and absolute substantive rights.

First, Romeo Castaño shows a very deferential attitude towards EU law and, in particular, towards the highly controversial EAW. Judge Spano explicitly accepts it, when, in his concurring opinion, he highlights the need to reconcile “the minimum human rights guarantees set out in the Convention with the requirements of uniformity and harmonization of standards in the framework of EU law” (Romeo Castaño, concurring opinion of Judge Spano, joined by Judge Pavli, § 1). The “symmetry” between the ECHR and EU law has been a well-observed phenomenon since the decision in Bosphorus. In this judgment the ECtHR held that the protection of fundamental rights afforded by the EU legal system was in principle equivalent to the one provided by the Convention (presumption of equivalent protection) (§ 72). In Michaud, the Court stressed that this finding applied a fortiori with the entry into force of Article 6 (amended) of the Treaty on European Union, which gave Convention rights the status of general principles of EU law (§ 106). In Romeo Castaño, the ECtHR accepts as a given that the EU “principle of mutual trust” should regulate European States’ cooperation in criminal matters (§§ 82-86). On the basis of this principle, EU Member States are required to accept and enforce each other’s decisions as if they were their own. Since such a “blind trust” may be detrimental to fundamental rights, some limits are necessary (Romeo Castaño § 84). However, even in the context of the ECHR, the limits are arguably not elaborated by the ECtHR, but by the European Court of Justice (CJEU) (see Melloni, Aranyosi & Căldăraru and LM). Although the Second Section does not mention any CJEU’s decision in the merits of its judgment, by referring to Avotiņš and Pirozzi (Romeo Castaño § 84), it implicitly accepts the CJEU’s case-law concerning the grounds for refusal of an EAW (see also ibid § 86). However, by failing to disclose the origins of the rules applied, the Court does not provide adequate clarity. When assessing the risk of ill-treatment of a person whose extradition is requested, should States always apply the principles elaborated in Melloni, Aranyosi & Căldăraru and LM? Or, in specific circumstances (perhaps those not involving EU law), can they move away from the CJEU’s jurisprudence by relying on the absolute nature of Article 3 ECHR? Furthermore, one could argue that the presumption of equivalent protection falls short in the context of the EAW, since its application is not primarily designed to protect the suspects’ fundamental rights but rather to expedite EU Member States cooperation in criminal justice. One may even speculate that the ECtHR wanted to assure the CJEU after its Opinion 2/13.[i] While there (see §§ 191-194) the CJEU was worried that EU accession to the ECHR could upset the principle of “mutual trust”, here the ECtHR responds that it won’t, and the presumption of equivalent protection stands strong even in the EU area of freedom, security and justice. Yet the ECtHR does not explain any of this. A longer elaboration on the relation between the ECHR and the EAW would have certainly provided more clarity in this respect.

Second, in Romeo Castaño, through the doctrine of positive obligations, the ECtHR de facto behaves as a court of third or fourth instance (Practical Guide on Admissibility Criteria §§ 261-268). Here, the Second Section does not limit itself to verifying Belgium’s compliance with the rights defined in the Convention, but it deals with errors of fact and law (rectius EU law) allegedly committed by Belgian courts in assessing N.J.E.’s risk of ill-treatment in case of surrender. This is a common issue with procedural violations and the “horizontal applicability of human rights”, but in this case the problem is enhanced. In decisions such as Güzelyurtlu and others and Rantsev, the obligation to cooperate is connected to the duty to investigate. It could admittedly be said that there is a certain indirect linkage between the duty to investigate and a violation of Article 2. As the argument usually goes, investigation is required to secure retrospectively the substantive right of the victim. Recent case-law has also linked the obligation to investigate to the substantive obligation to prevent interferences with a person’s life (see, e.g., Nicolae Virgiliu Tănase § 193). But in Romeo Castaño, as recognised by the Second Section (§§ 38-39), Belgium does not have an obligation to investigate Colonel Romeo’s death. To find a (procedural) human rights violation, the Court has to assume two points that, in fact, should not be taken for granted. Firstly, the Court has implicitly to recognise an individual’s right to have alleged perpetrators prosecuted. Even though the Second Section refers only to “effective official investigation by Spain” (Romeo Castaño § 84), the real concern is with prosecution. Arguably, effective investigation can be carried out even without the surrender of one of the suspected persons. Yet, through the recognition of an individual’s right to secure the prosecution of another, the Court shows a sort of “criminal fetishism”, whereby absence of criminal accountability is regarded as a human rights violation per se. It also opens the door to the risk of coercive overreach, which may be highly problematic, especially if no serious assessment about the role of criminal law is carried out. Secondly, the Court has to find a breach of the Convention even though, in this case, the violation is only future and potential, arising if the ETA suspect is not surrendered (but – see § 92 – the Belgian authorities have no absolute duty to do so). Belgium’s failure to cooperate at worst prevented Spain from exercising its duty to investigate and prosecute, but it is hardly arguable that it itself breaches the right to life of Colonel Romeo’s children. Neither has Belgium a duty to secure retrospectively the substantive right of the applicants (through investigation) nor is its obligation to cooperate linked to a substantive obligation to prevent interferences with the right to life. Conceiving the duty to cooperate as a self-standing obligation under Article 2 stretches the application of the right to life well-beyond the protection of an individual’s life and towards ensuring efficient cooperation in criminal matters.

Third, by finding Belgium in breach of Article 2, the ECtHR hides a very problematic assumption: procedural obligations under Article 2 may lessen substantial protection under Article 3. The Court focuses on Belgium’s failure to provide “sufficient factual basis” for invoking Article 3 (N.J.E.’s risk of ill-treatment in case of surrender) rather than focusing on Belgium’s “duty” to surrender the suspect (§§ 85-90). Yet, in §§ 86-89 the Court seems quite sceptical that Belgium would be able to provide such justification. The Court rejects the UN 2015 Report as outdated and points out that Belgium executed other Spanish EAWs without delay. The duty to cooperate is here transformed into a duty to assess the risk of ill-treatment in a way that suspects should in principle be surrendered. Only in exceptional circumstance, adequately justified, the principle of non-refoulement under Article 3 would be able to stay an EAW request. Therefore, the procedural obligation to conduct an adequate assessment merely qualifies the substantial obligation to surrender. What can domestic authorities draw from this? States may be encouraged to reduce the non-refoulement guarantees to comply with this new (and ambiguous) obligation to cooperate (rectius to extradite).[ii] The European judges seem to appreciate this issue in § 92 (“this judgment can not be interpreted as reducing the obligation of States not to extradite a person to a country requesting extradition where there are substantial grounds for believing that the person concerned, if extradited to that country, will run a real risk of being subjected to treatment contrary to Article 3”). However, the finding of a violation of Article 2 has in fact the effect of making this provision completely immaterial. The message is: if you don’t want to be found in breach of Article 2, you should go easy on Article 3 when examining EAW requests.


With a very concise and seemingly straightforward decision, the ECtHR goes a long way towards redefining the relationship between Convention rights and EU law; it recognises an individual’s right to have alleged perpetrators prosecuted; and it expands the application of the duty to cooperate to extradition requests or EAWs. It does so in a way that shows itself to be driven mainly by the concern to preserve “the ‘symmetry’ between Convention law and EU law’ (Romeo Castaño, concurring opinion of Judge Spano, joined by Judge Pavli, § 8). However, under implicit premises and “meticulously elaborated interpretations” (ibid), the Court fosters a “crime control model” which prioritises maximal efficiency over a better protection of fundamental rights.


[i]       Thanks to Laurens Lavrysen for pointing this out.

[ii]      Thanks to Laurens Lavrysen for suggesting this problem and for the email exchange that followed on this aspect of the case.

Print Friendly, PDF & Email

Related posts

Leave a Reply

Your email address will not be published. Required fields are marked *