Non-State Actors and Non-Refoulement: The Supreme Court’s Decision in Zain Taj Dean

Lord Advocate v. Zain Taj Dean [2017] UKSC 44 concerned an extradition request, made by the Republic of China in Taiwan (‘ROC’). Dean, a British national, had lived in Taiwan for many years. In 2011, he was convicted for manslaughter, drunk-driving and leaving the scene of an accident by an ROC court. While on bail, pending an appeal, he fled to Scotland. His convictions and four-year sentence were upheld, in absentia, in 2012. The request was made pursuant to an ad hoc ROC/UK MOU, and in accordance with section 194 of the Extradition Act 2003. The Edinburgh District Court ruled that Dean could be extradited but the Scottish Appeal Court disagreed. The Supreme Court had to decide whether Dean’s extradition, to serve out the remainder of his sentence in Taipei prison, would violate Article 3 of the ECHR.

As the greatest risk of harm emanated from other prisoners – rather than from public officials or the prison conditions themselves – the Supreme Court decided that the correct test was whether the requesting ‘State’ had offered to put in place reasonable protective measures to obviate this risk. To this end, it drew a distinction between State agents and non-State actors for this purpose despite the fact that the prison would be under the public authorities’ direct authority and control at all times. This post argues that this approach amounts to a misapplication of the Strasbourg jurisprudence, invoked by the Supreme Court, with potentially serious consequences for the interpretation of the non-refoulement principle in detention cases.  

The Article 3 Challenge

In the Appeal Court, counsel had agreed that the applicable test was “whether substantial grounds have been shown for believing that there is a real risk of treatment incompatible with Article 3” (Saadi v Italy (2009) 49 EHRR 30, para 128). According to the Lord Advocate’s expert, Taipei prison failed to meet a number of CPT Standards. It was grossly overcrowded and significantly understaffed. It had unsanitary conditions and serious problems with bullying. Dean claimed that media reports in Taiwan had generated considerable public hostility which put him in danger of violent attack by other prisoners.

Due to growing concerns about the compatibility of Dean’s detention with Article 3, the ROC authorities gave a series of undertakings regarding his treatment. Specifically, they promised that he would be detained in an Article 3 compliant two-man cell; he would be separated from other prisoners; and that any alleged breach of the assurances, made by British consular officials, would be remedied. The key question was whether these special arrangements were sufficient to remove the risk of ill-treatment.

The Appeal Court’s Judgment

In its judgment [2016] HCJAC 83, the Appeal Court decided, by a majority, that Dean’s extradition would violate Article 3. It held that there was a real risk of ill-treatment which the diplomatic assurances did not remove. Lady Paton was disturbed by the likelihood that Dean would have to remain in his cell due to his vulnerability to attack. This danger would be heightened, in the majority’s view, by the privileged conditions in which Dean would be kept. Lady Paton pointed out that, according to the CPT Standards, separation for a prisoner’s own protection may amount to ill-treatment (para 53).

The ECtHR has accepted that gaps in the protection of human rights may be bridged by using diplomatic assurances where the requesting State is not a Contracting State. In Othman v UK (2012) 55 EHRR 1, it indicated that – in essence – as long as relations of mutual trust and confidence exist between the two governments; the undertakings are comprehensive, specific and cogent; they are given on behalf of the central government; and the requesting State has a good record of honouring similar assurances then a court can accept them as a practical guarantee against ill-treatment (para 189).

In Dean, the majority decided that the Othman criteria were not satisfied. In addition, to concerns about the non-existence of prisoners’ rights in the ROC’s legal system; they were anxious about the lack of international monitoring of prison conditions in Taiwan; the absence of diplomatic ties; and the general deficiencies in the system of diplomatic protection [see here]. The Lord Advocate appealed to the Supreme Court.

The Supreme Court’s Judgment

The Supreme Court decided that the wrong test had applied to the facts. In its view, the correct test required a court to satisfy itself: (1) that the ROC’s undertakings would provide Dean with a reasonable level of protection against violent attack by other prisoners in Taipei prison; and (2) whether the protective measures themselves would infringe Article 3.

Lord Hodge relied upon Bagdanavicius v. Secretary of State for the Home Department [2005] 2 AC 668, which followed the approach adopted by the ECtHR in D v. UK (1997) 24 EHRR 423 and HLR v. France (1997) 26 EHRR 29. In Bagdanavicius, the House of Lords distinguished between cases where the risk of ill-treatment “emanates from the intentionally inflicted acts of the public authorities in the receiving country” (D v. UK, para 49) and those where the risk of harm arises from non-State actors. In that case, Lord Brown observed that the latter situation cannot trigger an Article 3 violation per se because ill-treatment cannot be committed by private individuals, no matter how severely the victim may have been treated (or how grave the risk of harm). However, he acknowledged that violence, or its threat, may be transformed into an Article 3 violation, if the receiving State fails to afford reasonable protection against it (para 24).

The Supreme Court concluded that the Othman criteria had been satisfied. Lord Hodge noted that the specific assurances, which had been given by senior officials in the ROC government, addressed the particular conditions of Dean’s detention – including concerns about his safety vis-à-vis other prisoners – and they had given British consular officials a credible monitoring role (para 38). Lord Hodge paid particular attention to the claim that the proposed protective regime would breach Article 3. He accepted that solitary confinement may qualify as ill-treatment, where it involved the ordering of complete sensory or prolonged social isolation (Ahmad v UK (2012) 56 EHRR 1, paras 207-210, cited para 41). However, he noted that any isolation would be at Dean’s election. Consequently, he distinguished between instances of total isolation and the “relative isolation” that was foreseeable in this case. In his view, such a protective measure would not infringe Article 3. The Supreme Court concluded that the ROC authorities had ensured that Dean would be afforded reasonable protection against the threat of harm from other prisoners; and the proposed special regime did not contravene Article 3.

Analysis

The Strasbourg jurisprudence, relied upon by the Supreme Court, holds that ill-treatment involves “the intentionally inflicted acts of the public authorities”. The observation that private individuals cannot commit ill-treatment is correct as it is closely associated with the conduct of public officials (Article 1, UN Convention Against Torture (1984) 1465 UNTS 85. However, the divide between State agents and non-State actors, especially in extradition cases involving detention, is much more contentious. It is well known that poor prison conditions can amount to ill-treatment. The public authorities are responsible for those conditions (and they will have full knowledge of them) but it would be going too far to say that State officials have intentionally inflicted ill-treatment on a prisoner by detaining him or her in such conditions. Clearly, this requirement is highly doubtful in prison cases.

Further, the cases invoked by the Supreme Court in support of its State agents/non-State actors divide are qualitatively different from prison cases. Bagdanavicius, an expulsion case, involved a claim that the receiving State would not protect the Applicants against the threat of violence arising from family members; in HLR v. France, another expulsion case, the Applicant claimed he would be at risk of harm from drugs traffickers in the receiving State; and D v. UK concerned responsibility for the transmission of AIDS. In sharp contrast, public authorities exercise continuous, direct authority and control over prisoners and are responsible for jail conditions. Accordingly, although the risk of violent behaviour from prisoners cannot be directly attributable to the State, the high level of control exercised in such situations sets them apart from cases where the receiving State is obligated to provide reasonable protection against the risk of harm arising from non-State actors throughout its jurisdiction. This difference should be enough to cast doubt on the reliability of the State agent/non-State actor divide in prison cases. In any event, it should ensure that the courts undertake a stringent review of whether protective measures are compatible with Article 3 in such cases – a test based on reasonableness is inadequate for this a task.

Cases involving the application of the non-refoulement principle are hard because they often require judges to make fine-grained assessments about the foreseeability of harm. Accordingly, judges may disagree about whether an individual should be extradited or not. Perhaps the debate about whether Dean’s de facto solitary confinement would qualify as ill-treatment provides a good illustration of this point. However, to take the view that Dean is such a finely-balanced case would be to overlook a number of background factors which should have influenced the Supreme Court’s decision.

The fact that the extradition arrangements were, necessarily, made by way of an MOU should have increased concerns about the enforceability of the ROC’s commitments. The usual argument [see here] that MOU have significant political effects, is much less convincing given the UK’s non-recognition of the ROC. Moreover, the general charge that diplomatic assurances amount to unenforceable promises has been well-documented [see here]. In any event, the undertakings given in Dean can be compared unfavourably with other cases. In Othman, an MOU was used to conclude extradition arrangements between the UK and Jordan. Nevertheless, the two governments had strong diplomatic ties; the assurances provided for recourse to judicial remedies; and an independent human rights NGO was appointed to monitor compliance with them.

Once the complexities of the ROC’s ambiguous international standing are taken into account and the Othman criteria reassessed it is suggested that the Supreme Court made a mistake in allowing the appeal. More broadly, its distinction between State agents and non-State actors threatens to undermine the safeguards that the non-refoulement principle offers to individuals in extradition cases where detention is involved.

Conclusion

If Dean ends up before the Strasbourg Court, the outcome on the question of whether the danger posed qualifies as ill-treatment would be far from certain. In Ahmad, the Court warned that it is not for Contracting States to impose Convention standards upon third States (para 177). Moreover, it observed that “it has been very cautious in finding that removal from the territory of a Contracting State would be contrary to Article 3” especially where the receiving State has, “a long history of respect of democracy, human rights, and the rule of law” (para 179). It is unclear whether the ROC would be treated as though it fitted within this bracket of receiving States, notwithstanding its established democratic credentials and avowed commitment to the rule of law (see [2015] HCJAC 52, paras 17-19). Nevertheless, such a petition would provide the ECtHR with an opportunity to correct the Supreme Court’s misapplication of its jurisprudence and to reaffirm the general applicability of the Saadi test in extradition cases involving detention.