Detention conditions & release mechanisms: the approach to countries outside the Council of Europe

Stefan Hyman, 9BR Chambers


If there is one certainty in extradition, it is that conditions of detention are never far off the agenda. UK courts, like their counterparts across the continent, receive an ever-increasing number of extradition requests from countries outside the Council of Europe (‘CoE’) or ‘third states’. These are not bound by rights, and the interpretation thereof, enshrined in the European Convention on Human Rights (‘ECHR’).

Over the last couple of months, the European Court of Human Rights (‘ECtHR’) handed down three decisions concerning third state extradition requests. Two of these concerned US extradition requests: one case originating from Italy (McCallum v Italy (20863/21)), the other from England & Wales (Sánchez-Sánchez v UK (22854/20)). The third involved extradition from Poland to China (Liu v Poland (37610/18)). After exhausting their challenges under Article 3 ECHR before national courts, the requested persons each applied for, and received, an ‘interim measure’ under Article 39 ECHR staying, in effect, their removal.

There is a particular nexus between McCallum and Sánchez-Sánchez since these cases considered the prospect of a requested person receiving a sentence of life imprisonment without the possibility of parole (‘LWOP’), also known as a ‘irreducible life sentence’. The original sections to which the cases were allocated relinquished jurisdiction to the Grand Chamber so that Strasbourg could consider the diverging lines of authority in Trabelsi v Belgium (140/10) and Harkins & Edwards v UK (2012) 55 EHRR 19, of which more below.

Before drawing conclusions from these decisions, here is a summary of the facts and the reasoning behind the ECtHR’s decisions.

Liu v Poland

On 6 August 2017, Mr. Liu, a Taiwanese citizen, was arrested on foot of an INTERPOL Red Notice in Poland. This sought his detention for purposes of extradition to face charges in respect of a vast fraud allegedly committed on a telecommunications syndicate. The Chinese Government’s extradition request came before the Warsaw Regional Court. During first instance proceedings, the Polish Prosecutor sought a guarantee that Mr. Liu would be afforded a fair trial in China and details of where he would be detained. The Ministry of Public Security responded that Mr. Liu would be held at the Boluo Detention Centre in Guangdong Province and provided information about the human-rights protection system in China together with a guarantee that Mr. Liu’s fundamental rights would be protected. The Warsaw Regional Court approved Mr. Liu’s extradition finding that the request complied with the technical requirements of Polish law, there was prima facie evidence of guilt, and that the assurances received were sufficient. The ECtHR pointedly observed that reports from reputable international bodies adduced before the Warsaw Regional Court were swiftly dismissed, with the national court finding that they ‘always highlight irregularities, whilst ignoring developments in the State concerned’ (§10).

The Warsaw Court of Appeal upheld the Regional Court’s decision and reasoning. The mere fact that Mr. Liu was a Taiwanese citizen was insufficient to give rise to real risk of ill-treatment whilst in detention. Moreover, it noted that the crimes of which he was accused were not political in nature and the reports relied on were general and not specific (§11).

Upon the ECtHR’s request, the Government of Poland confirmed that it had not sought diplomatic assurances from China. Whilst the case was awaiting determination before the ECtHR, the Polish Commissioner for Human Rights intervened and appealed the Court of Appeal’s decision to the Supreme Court. The Supreme Court dismissed the Commissioner’s appeal (§§15-16).

In its judgment (§§34-44), the ECtHR summarised the conclusions of various governmental and non-governmental bodies about the treatment of suspects in custody and at the Boluo Detention Centre. This included the UN Committee against Torture, the UN Special Rapporteur on Torture, the US State Department, Amnesty International and Human Rights Watch. Unlike many Article 3 cases, moreover, the evidence did not focus on inter-prisoner violence (see HLR v France (24573/94), for instance), but on ill-treatment at the hands of state agents (i.e., police officers, guards, etc.)

Finding a violation of Article 3, the ECtHR criticised the domestic courts’ analysis of Mr. Liu’s evidence as ‘superficial’ holding that ‘the domestic courts [should have] perform[ed] a more in-depth analysis of available sources. It heralded ‘[w]here there are serious allegations of widespread ill-treatment in the country of destination, the domestic authorities have a special obligation to verify whether the person concerned would be exposed to a real risk of treatment contrary to Article 3’ (§69).

Given the latent insufficiencies at national level, the ECtHR decided to analyse the material itself (§70). The Fifth Section recalled that where a court considers the risk of treatment contrary to Article 3, be it at first instance or appellate level, it must do so with the most-up-to-date information. This is because “risk assessment” is forward-looking (§73).

The ECtHR concluded that there was a strong “international consensus” concerning the routine ill treatment of suspects in detention, both in China (generally) and the Boluo Detention Centre (specifically). It observed that the conclusions of the reputable bodies were consistent (§74 & §79). Moreover, given that the Chinese Government had not signed many international human rights instruments, there was limited scope for an independent assessment of compliance with international standards (§§75-78).

As to evidential presumptions when assessing Article 3, the Fifth Section held that ‘where there are many significant shortcomings in the domestic legislation in the country of destination and allegations of serious abuses identified in independent reports, coming from numerous sources, the benefit of the doubt should be granted to an individual seeking protection’ (§80).

The ECtHR criticised the Government of Poland for solely seeking ‘informal declarations’ that Mr. Liu’s human rights would be respected. The documents received were not ‘diplomatic’ and thereby would did not permit the Strasbourg Court itself to decide whether they were sufficient. In support of this proposition, it cited Othman (Abu Qatada) v UK (2012) 55 EHRR 1 (§§187-189).

Lastly, the ECtHR held that it ‘considers that the extent to which torture and other forms of ill-treatment are credibly and consistently reported to be used in Chinese detention facilities and penitentiaries may be equated to the existence of a general situation of violence. Thereby the applicant is relieved from showing specific personal grounds of fear, it being enough that it is established that, upon extradition, he will be placed in a detention centre or penitentiary’ (§83).

This will no doubt be of assistance to any practitioner litigating an extradition request to China.

Irreducible sentences amongst the COE states

Over the last twenty years, the ECtHR has considered whether an LWOP sentence gives rise to a violation of Article 3 and, if so, at what point. The core principle is that a penalty like ‘a whole life order’ in England & Wales (see section 321 of the Sentencing Act 2020), may amount to ill treatment where the sentence is not de jureand de facto irreducible. At the point at which it is imposed, a prisoner should know what he or she must do to receive a review which affords a prospect of release on parole. Such a review must consider whether the prisoner’s progress in detention is so exceptional that continued incarceration remains justified on legitimate penological grounds. This prospect is meant to empower a prisoner, if so minded, to strive for release (see Hutchinson v UK [2017] ECHR 65 at §§42-43). Whilst some CoE states permit an automatic review mechanism after a fixed period (normally after twenty or twenty-five years – see Hutchinson at §44), which is broadly akin to a ‘minimum term’ in England & Wales (see section 28 of the Crime (Sentences) Act 1997), such a fixed mechanism is not required to comply with Article 3. However, as is apparent from Article 5(2) of the EAW Framework Decision, some EU Member States have established this benchmark in domestic law.

Although the Hutchinson case law is well-known in the UK, it is worth noting that several other CoE states have had their de facto and de jure release schemes analysed: see inter alia Karkaris v Cyprus (2009) 49 EHRR 35, Murray v The Netherlands (10511/10), TP & AP v Hungary (37871/14 & 73986/14), Harakchiev & Tolumov v Bulgaria (15018/11 & 61199/12), Čačko v Slovakia (49905/08), Bodein v France (40014/10), Matiošaitis and Others v. Lithuania (22662/13, 51059/13, 58823/13, 59692/13, 59700/13, 60115/13, 69425/13 & 72824/13), Petukhov v. Ukraine (no 2) (41216/13), Marcella Viola v Italie (No 2) (77633/16). Two points merit attention. First, it is not only the UK which finds the question of the possible release of prisoners convicted of the most heinous crimes a sensitive (and, no doubt, politicised) question. Second, the ECtHR’s ultimate determination on the facts (i.e., whether a review mechanism is sufficient in fact and in law) is not always internally consistent.

Sánchez-Sánchez v UK

Based on an arrest warrant that a US District Court issued, the US Government sought Mr. Sánchez-Sánchez, a Mexican national, for purposes of prosecution in respect of various drugs supply offences. He was said to be the joint head of a drug trafficking organisation who supervised and managed the distribution of large quantities of drugs in Georgia. Of relevance, a co-conspirator is said to have died because of fentanyl overdose in circumstances related to supply (§§9-10). At first instance, Mr. Sánchez-Sánchez submitted that there was a real-risk that he would receive an LWOP sentence. He adduced evidence of the US Guidelines Manual, together an opinion of an experienced US lawyer regarding actual sentencing practice, including circumstances where a person had died. Based on the evidence, District Judge Crane found that there was a possibility that Mr. Sánchez-Sánchez would receive a whole life order, but that this was discretionary. Relying on Shaw v USA [2014] EWHC 4654 (Admin), she concluded that there were mechanisms at federal level, namely executive clemency or compassionate release, which meant that the sentence was neither de facto nor de jure irreducible (§§12-16).

The argument was not pursued on appeal (Sanchez v USA [2020] EWHC 508(Admin) at §4) as the Divisional Court had then recently disposed of the same point in another case (Hafeez v USA [2020] 1 WLR 1296). In spite of the ECtHR’s judgment in Trabelsi, whichheld that the principles in Hutchinson applied to third states,the Divisional Court followed the House of Lords’ judgment in R (Wellington) v SSHD [2009] 1 AC 335 instead, the ratio of which was approved in Harkins. Whilst all their Lordships considered that the two routes for release described above meant that the sentence which Mr. Wellington may face if convicted was neither de facto nor de jure irreducible, the Committee was divided over whether the approach within CoE states applied to third states (§§28-34). It is exactly this point which characterised the Harkins v Trabelsi dispute

The Grand Chamber turned to this discordance and how it might be resolved. First, it rehearsed the lines of authority and case law in some detail (§§33-56). It then considered the sentencing principles set out in Title 18 of the United States Code (§57) and the pathways for release (§§58-62) before considering statistical information from the US Sentencing Commission (§63) and a letter from a federal prosecutor concerning the likely disposal in Mr. Sánchez-Sánchez’s case (in summary, if a plea bargain were agreed, he would receive a determinate sentence measured in years; he had a right of appeal; none of his co-conspirators, including one in a broadly comparable position to him, had received an LWOP sentence; if convicted, he would benefit from a pre-sentence report; all federal courts had discretion and a statistical analysis suggested it was properly used; just 1.1% of prisoners in federal facilities were serving a de facto life sentence) (§64).

In the dispositive part of its decision, the Grand Chamber emphasised the absolute prohibition of an Article 3 violation (§78). Referring to Kafkaris, it rehearsed the LWOP sentence principles amongst CoE states: (a) the imposition of a life sentence of an adult offender was not, within itself, incompatible with Article 3; (b) however, in certain circumstances it ‘might raise an issue under Article 3’; (c) ‘the principal question […] was whether a life prisoner could be said to have any prospect of release’; and (d) this, thus, engaged a de facto and de jure analysis of a particular mechanism (§79).

Drawing from its judgment in Vinter & Ors v UK [2014] Crim LR 81, the ECtHR emphasised the margin of appreciation conferred on COE states: it would not prescribe the nature (i.e., executive or judicial) or timing of the review mechanism (although it gave a strong “steer” on the latter). (§§80-81). Nevertheless, referring to Murray, it held that, to be effective, the review must entail ‘an actual assessment of the relevant information and it must be surrounded by sufficient procedural guarantees’ so that the prisoner ‘know[s] what he or she what he or she must do to be considered for release and under what conditions, reasons might have to be provided, and this should be safeguarded by access to judicial review’ (§82).

Of central importance, the Strasbourg Court distinguished between an extradition and a non-extradition context and acknowledged the “pull and push” between the dual international obligations of criminal cooperation and compliance with the ECHR (§83). Drawing on FG v Sweden (43611/11) and Saadi v Italy (27201/06), the Grand Chamber observed that extradition cases involve a forward-looking assessment as opposed to a review of past acts/omissions. It held that ‘the distinction is important’ as, in the latter, a person has been convicted and sentenced and the requirements of review are well known to domestic courts and can be re- assessed by the Court (§91). On the other hand, the former call for ‘complex risk assessment’ involving ‘a tentative prognosis’ that will ‘inevitably be characterised by a very different level of uncertainty’ (§92).  Referring both to ‘principle’ and ‘practical concerns’, Strasbourg distinguished between the ‘substantive’ obligation to ensure that the penalty imposed is not incompatible with Article 3 and ‘procedural safeguards’ which ‘appear to be better suited to a purely domestic context’. The Grand Chamber concluded, ‘[c]onsequently [these considerations] do not arise in relation to an individual whose extradition has been requested by a third State, as this would be an over-extensive interpretation of the responsibility of a Contracting State in such a context’. (§93). Accordingly, CoE states are not expected to scrutinise a third state’s ‘relevant law and practice’, nor ‘its degree of compliance with these procedural safeguards’ and, as such, the Contracting Parties cannot be held responsible for ‘deficiencies in the system of a third state’ (§93).

The Court then drew the distinction between a domestic case, in which, following a finding that Article 3 had been violated, the Contracting Party would be obliged to introduce a Convention-complaint review mechanism as a remedy (see, for instance, Harkins following Vinter supra) and the extradition context in which there would be impunity unless the requested person could be prosecuted in the requested state, or the requesting state could supply sufficient assurances. This was particularly grave as that those who may face an LWOP sentence would be accused of the most serious crimes (§94).

In terms of process, at an extradition hearing, a requested person must, first, prove that there is a real risk that he will receive an LWOP sentence. A violation of Article 3 ‘will more readily be established if the applicant faces a mandatory sentence of life imprisonment’ (§95). If that hurdle is surmounted, then the executing state must establish whether, in the requesting state, there is a ‘mechanism of sentence review which allows the competent authorities there to consider whether any changes in the life prisoner are so significant, and such progress towards rehabilitation has been made in the course of the sentence, as to mean that continued detention can no longer be justified on legitimate penological grounds’.

Nevertheless, this need not include an examination of the sufficiency of ‘procedural safeguards’ (§96) so long as, ‘from the moment of sentencing, there is a review mechanism in place which allows the domestic authorities to consider the prisoner’s progress towards rehabilitation or any other ground for release based on his or her behaviour or other relevant personal circumstances’ (§97). The Grand Chamber expressly overruled Trabelsi (§98). On the facts of Mr. Sánchez-Sánchez’s case, it found that he had failed to meet the first stage of the test. Based on all the evidence, the ECtHR did not believe that there was a real risk that he would receive a sentence of life without parole (§109).

McCallum v Italy

The US sought Ms. McCallum’s extradition for purposes of prosecution in respect of an allegation of murder. The deceased was her husband who disappeared in May 2002. Although his mortal remains was discovered soon after, as they were so badly burned, they were not identified as his until thirteen years later (§14). In the interim, Ms. McCallum had left the US (§16). Her daughter and a friend of her daughter were charged with murder (§17). The former was convicted after trial of first-degree murder and received an LWOP sentence; the latter pleaded to second-degree murder and was sentenced to a period of determinate detention (§18).

Having been arrested in Italy, the Court of Appeal in Rome examined the extradition request. It relied on information that US authorities provided concerning the possibility of an appeal against an LWOP sentence and Ms. McCallum’s right to seek a pardon or commutation of sentence from the Governor of Michigan. Further, it considered domestic jurisprudence concerning extradition to Michigan in which the Court of Cassation held that an LWOP sentence was not grossly proportionate and there were mechanisms which meant that sentences were reducible (§21). On appeal, the Court of Cassation held that the US system was generally comparable with the Italian system and thus, ostensibly, compatible with the ECHR. It held that Ms. McCallum’s arguments were a ‘generic bundling of arguments and precedents without relevance to the present case’ (§23).

Whilst the application was pending before the Grand Chamber, the US Embassy in Rome sent the Government of Italy a diplomatic note informing that the state prosecutors’ office in Michigan had undertaken to try Ms. McCallum for second-degree murder which carried a maximum sentence of life with parole. Other charges were also amended (§28). Based on this changed landscape, the Minister of Justice made a fresh order for extradition (§§29-30).

The Grand Chamber rehearsed Italian law (§§31-34), the nature of bilateral extradition arrangements (§35) and the Michigan Penal Code (§§37-41). In light of the reduced charges, the operative part of the judgment was short. Of interest, the ECtHR rejected the argument that a ‘diplomatic note’ could not be considered an assurance, citing Harkins at §85: ‘The Court also recognised that, in international relations, Diplomatic Notes carry a presumption of good faith and that, in extradition cases, it was appropriate that that presumption be applied to a requesting State which has a long history of respect for democracy, human rights and the rule of law, and which has longstanding extradition arrangements with Contracting States’. Relying on Article 26 of the Vienna Convention on the Law of Treaties 1969, the ECtHR considered non-compliance with the terms of the diplomatic note was tantamount to bad faith. The Government of Italy assured the Grand Chamber it considered the ‘diplomatic note’. a binding promise (§52).

Whilst Ms. McCallum sought to argue that it was possible that the Governor of Michigan and not the Parole Board would be responsible for her release, the Grand Chamber rejected this submission outright noting that the decision of the parole board could be appealed to a circuit court (§54). In any event, it relied on its reasoning in Sánchez-Sánchez holding that Ms. McCallum’s arguments pertained to (in)sufficiency of a procedural safeguard and not a substantive obligation (§53).


There are various take-aways from the 2022 trilogy on prison conditions.

  1. National courts are entitled to give weight to reports of governmental and non-governmental bodies. Where these speak with the same voice, they, without more, create a strong, albeit rebuttable, presumption of risk (Liu).
  • There is a ‘special obligation’ to analyse evidence carefully as to whether it is credible vis-à-vis incompliant treatment during, or conditions of, detention. This is because Article 3 is absolute and thus the national court’s scrutiny must be rigorous (Liu).
  • There is a difference between informal guarantees and diplomatic assurances since the latter can be monitored (Liu). Interestingly, in McCallum, the Court placed particular emphasis on a ‘diplomatic note’ transmitted from the US Embassy in Rome to the Italian Government. Given that assurances/guarantees are a regular feature of extradition amongst the COE states, we may question what status can be afforded to the documents we regularly receive in Bulgarian, Hungarian and Romanian cases.
  • There is a material difference to “the type” of review applied to LWOP sentences in a “domestic” and an “extradition” context where the requesting territory is a third state. In the former, whilst the ECtHR (and thereby national courts) must assess the procedural safeguards in place to ensure that there is a possibility of review of whether there remain legitimate penological grounds for detention, this is not so in the latter. As such, in appropriate cases, the requested state need only demonstrate that there is a mechanism, be it executive or judicial, which exists in law and may allow for a possibility of release. The executing court need not consider the sufficiency of that mechanism (Sánchez-Sánchez).
  • Prior to a national court considering whether such a mechanism exists de facto and de jure, the burden falls on a requested person to adduce sufficient evidence to demonstrate that there is a real risk that he will receive an LWOP sentence (Sánchez-Sánchez).

Sánchez-Sánchez is likely to close a chapter in the LWOP sentence cases. All a requesting state need show is that there is a possible mechanism for review. Presumably, unless a requested person demonstrates that it would be impossible to rely on such a mechanism (i.e., de facto it is entirely illusory), national courts will look no further. Whilst the ECtHR’s decision is pragmatic and will surely be welcomed by many COE states, as a matter of pure principle it is hard to reconcile the differing “requirements” of review as the terms of Article 3 are absolute. Moreover, if a requesting state has sufficient interest in securing the extradition of a person accused of the most serious offences, surely it would supply cogent evidence about its review mechanisms or, failing that, supply an assurance or even modify the charges?

Updated on

Related Articles

Need Support?
Can't find the answer you're looking for?
Contact Support