Royal Courts of Justice
Strand, London, WC2A 2LL
Before :
MR JUSTICE JULIAN KNOWLES
Between :
VASILE MIHAILA | Appellant |
- and – | |
JUDECATORIA PIATRA NEAMT (ROMANIA) | Respondent |
Jonathan Swain (instructed by Lawrence & Co) for the Appellant
David Ball (instructed by CPS) for the Respondent
Hearing dates: 16 May 2023
Approved Judgment
This judgment was handed down remotely at 10.30am on 23 January 2024 by circulation to the parties or their representatives by e-mail and by release to the National Archives.
.............................
Mr Justice Julian Knowles:
Introduction
This is an appeal under Part 1 of the Extradition Act 2003 (EA 2003) with the permission of Linden J. The Appellant appeals the decision of the district judge dated 16 November 2020 ordering his extradition to Romania. The Appellant was arrested before 11pm on 31 December 2020 and so the ‘old’ European arrest warrant (EAW) arrangements continue to apply.
The Appellant is represented by Mr Swain. The Respondent is represented by Mr Ball. I am grateful to them both.
The Appellant is sought pursuant to a conviction EAW. This is based on a sentence imposed on 10 April 2019, which remained final through the decision of the Court of Appeals, Bacau, on 1 April 2020.
The EAW was issued by Gabriela Bratu, Judge of the Judecatoria Patra Neam, Romania, on 8 April 2020 and issued for service by the NCA on 20 April 2020. The warrant is based on a final and enforceable judicial order dated 10 April 2019. This was passed by the Piatra Neamt Trial Court. The decision became final by a decision dated 1 April 2020 by the Court of Appeal Bacau.
The EAW relates to three offences from 2006-2007. It is said the Appellant began a ‘sexually deviant attitude’ in relation to two children (aged seven and eight) ‘of his concubine’.
The behaviour complained of is described in Box (e) of the EAW. Around 2002 the Appellant began cohabiting with a woman whom I will call PA. She had two daughters who, when she and the Appellant met, were toddlers. In around 2006-2007, when the daughters were aged approximately seven and eight respectively, he began to sexually abuse them. The warrant sets out how he ‘periodically secured the satisfaction of his sexual needs by touching the minors' private parts and by contact of his penis with the bodies of the victims, ejaculating subsequently.’ He would ‘emotionally blackmail’ and mentally coerce his partner into thinking this is normal behaviour. The conduct is said to have, ‘severely jeopardised the minors’ moral development.’
The Appellant was present at his trial but not his sentence. A sentence of 10 years and four months’ imprisonment was imposed, of which nine years nine months and 22 days remains to be served (less any time spent on remand).
Procedural history
The Appellant was arrested on 24 April 2020 and produced before the lower court the same day. The following grounds under the EA 2003 were raised at the extradition hearing under Part 1 in relation to some or all of the Romanian offences: s 2; s 10; s 17; and Articles 3 and 8 of the ECHR read with s 21 of the EA 2003. The Article 3 argument related to prison conditions in Romania and alleged risks from other prisoners to the Appellant as a sex offender. The Appellant relied on a report from the European Committee for the Prevention of Torture and Inhuman or Degrading Treatment or Punishment (CPT) from 2019 (based on prison visits in Romania in 2018).
The Appellant filed and served Perfected Grounds of Appeal dated 15 January 2021. On 22 December 2021, he applied to vary his grounds of appeal. In an order of 17 February 2022 Griffiths J granted that application, directing that the Appellant must file and serve Amended Perfected Grounds of Appeal within 28 days of the date of the order.
Further, Griffiths J directed that that the Appellant should, within 14 days of the date on which the decision of the Divisional Court in Marinescu (CO/4264/2020) being handed down (the lead case on Romanian prison conditions), inform the Respondent as to whether the ground is maintained and, if so, on what basis, together with appropriate submissions.
The Appellant filed and served Amended Perfected Grounds of Appeal dated 18 March 2022 in line with those directions. On 12 September 2022, the Court handed down its decision in Marinescu and others v Romanian Judicial Authority [2022] EWHC 2317 (Admin). The Appellant then filed and served updated Amended Perfected Grounds of Appeal in light of that decision, incorporating his updated submissions on Article 3. An application to adduce a further assurance was lodged by the Respondent in October 2022.
On 19 December 2022, Lane J refused permission on all grounds. The Appellant applied to renew permission to appeal. On 20 December 2022 he made an application to rely by way of fresh evidence upon a report from 2022 from the CPT based on visits in 2021. On 26 January 2023, Linden J granted permission to appeal.
The single ground of appeal now advanced by Mr Swain on behalf of the Appellant is that he would be at a real risk of a breach of Article 3 if surrendered to Romania due to the risk of harm he faces from other prisoners because of the nature of his offences (sexual offences against children) and that he will not be adequately protected in prison. Mr Swain relied on the principle that although Article 3 is aimed primarily at ill-treatment, etc, by state agents, the state will also be liable under Article 3 where the relevant risk emanates from non-state actors (such as other prisoners) and the state is unable or unwilling to act to provide ‘reasonable protection’ to the defendant: see R (Bagdanavicius) v Secretary of State for the Home Department [2005] 2 AC 668, [24].
Submissions
On behalf of the Appellant, Mr Swain submitted that the nature of the Appellant’s offending would render him vulnerable to violence at the hands of prisoners. In addition to the Appellant’s own past treatment in prison, there is evidence from the CPT that individuals in custody for similar offending are particularly at risk, and that the Romanian authorities fail to adequately protect them (and so will not protect the Appellant).
The Appellant had given evidence in his proof (which he adopted orally as his evidence in chief), in particular at [37], that he had been ill-treated in prison in Romania because of his offending. The judge had therefore been wrong to say there was no evidence of risk to the Appellant arising from his offences when he said at [42]:
“I have heard no evidence about the risks that are faced specifically by the RP should his extradition be ordered. In his proof of evidence, the RP describes his troubling experiences while on remand in Romania, however those experiences are not as a result of the offences for which the RP was on remand.”
The Appellant’s evidence had not been challenged by the Respondent, and should have been accepted. At [44] the judge mentioned the assurance from Romania about prison conditions, but this did little to alleviate risks from other prisoners. The 2019 CPT Report before the judge referred to inter-prisoner violence and made criticisms, eg, of failures to conduct proper risk assessments in relation to cell-sharing. The judge had not referred to this.
Mr Swain said that the judge had wrongly not engaged with that CPT Report and that had he done so, he would have been bound to discharge the Appellant.
Before me, Mr Swain also sought to rely on the CPT Report from 2022 which I referred to earlier. Plainly this post-dates the district judge’s decision. He said it was decisive and should be admitted under the well-known principles in Szombathely City Court v Fenyvesi [2009] 4 All ER 324 and Zabolotnyi v Mateszalka District Court, Hungary [2021] 1 WLR 2569.
Mr Swain said the 2022 Report includes pertinent information, which demonstrates that deficiencies inherent in the Romanian prison system would place those accused of sexual offences at risk: see, for example, at [72]:
“72. The CPT’s delegation found that instances of inter-prisoner violence appeared to be lower in the prisons visited in 2021 by comparison with those visited in 2018. Nevertheless, many persons stated that tensions were exacerbated by the overcrowding, lack of activities and limited access to hot water/showers which did, at times, spill over into violence between prisoners. In each prison, the CPT’s delegation received allegations of fights happening mostly in cells and occasionally in the exercise yards. In this context, persons of Roma origin, as well as persons accused or convicted of sexual offences, appeared to be particularly at risk. The situation appeared especially problematic at Galaţi and Giurgiu Prisons.”
Mr Swain said there was nothing in the assurances offered by Romania (including the one from October 2022) which met these concerns. Overall, he said that the Aranyosi threshold had been crossed (Criminal proceedings against Aranyosi [2016] QB 921) so as to require further specific information about the Appellant’s position. I will return to Aranyosi later.
Absent an adequate response from the Respondent which engaged directly with the issue of risks to the Appellant, Mr Swain submitted that there is a real risk that Romania would fail to adequately protect the Appellant from the risk of harm.
On behalf of the Respondent, Mr Ball said that the height of the Appellant’s case was that he will be at real risk of a breach of Article 3 inter-prisoner violence because he is a child sex offender. (Although he also denies that he is a child sex offender, and that the allegations are made up.) However, he said that the evidence relied upon by the Appellant falls short of what would be required to make good this proposition. Mr Ball divided up the Appellant’s challenge into two: (a) supposed errors by the district judge in relation to the material before him; (b) the 2022 CPT Report.
As to the first limb, Mr Ball said the judge’s analysis had been sound, in particular [43] where he said that:
“43. The JA will have a wealth of experience in the incarceration of prisoners convicted of sexual offences. Faced with no evidence to lead me to a contrary view, I have trust and confidence in the management of prisons by the JA. There is no reason to believe that the RP will not be properly and safely kept within the prisons. I am satisfied that the RP will be properly accommodated and that necessary steps will be taken to ensure his safety.”
As to the Appellant’s own evidence, this amounted to at most an allegation that a prisoner had climbed into his bed and tried to sexually assault him, whilst other prisoners laughed. Mr Ball said that even taken at its height this did not come close to satisfying the Article 3 test arising from ill-treatment by non-state actors (discussed later). There was an alarm button and the Appellant was later moved.
As to the second limb, Mr Ball accepted criticisms had been made by the CPT in its 2019 and 2022 Reports but said they did not establish the Appellant’s case under Article 3. He said at [45]-[46] of his Skeleton Argument:
“45. The [2019] CPT report simply does not come close to establishing that there are general conditions for child sex offenders as a class in Romania that mean they are at real risk of ill treatment. There are aspects about the Romanian prison estate that can be improved of course. But the material does not support the proposition that child sex offenders are generally unsafe in Romanian prisons. On the contrary, as set out above, the evidence which was before the District Judge of the Appellant’s specific experiences highlighted that there were safety features in place and which were ultimately effective. No prison unfortunately, given that they frequently house some of the most dangerous people in society, can ever guarantee the absolute safety of all of its prisoners, all the time.
46. The District Judge did not err in his treatment of the evidence. He certainly was not wrong to find that extradition was compliant with Article 3. The evidence of the Appellant and the CPT in 2019 fall far short of establishing any systemic deficiencies as they relate to child sex offenders. There are no reports or judgments relied upon that make this out. There is no expert report which highlights specific difficulties faced by child sex offenders as a class.”
So far as the 2022 CPT Report is concerned, Mr Ball said this showed and that inter-prisoner violence had fallen in recent years and that Romania had taken steps to remedy concerns highlighted in the earlier 2019 report. Further, he pointed out that the 2022 Report had been considered in Marinescu and found not to call for any further Aranyosi enquiry.
Discussion
Legal principles
Article 3 of the ECHR provides that ‘No one shall be subjected to torture or to inhuman or degrading treatment or punishment.’
The principles in relation to Article 3 are not materially in dispute. A defendant wishing to defeat extradition under Article 3 must show strong grounds for believing that, if returned, s/he will face a real risk of being subjected to torture or to inhuman or degrading treatment or punishment contrary to Article 3: see R (Ullah) v Special Adjudicator [2004] AC 323, [24]. As regards the test for a ‘real risk’, ‘the burden of proof is less than proof 'on the balance of probabilities', but the risk must be more than fanciful’: see Badre v Court of Florence [2014] EWHC 614 (Admin), [40].
In Rae v Government of the United States of America [2022] EWHC 3095 (Admin). [64] and [86] the Court said:
‘ 64. Thus, the position can be summarised as follows:
The prohibition of Article 3 ill-treatment is absolute. There is no distinction to be drawn between the minimum level of severity required to meet the Article 3 threshold in the domestic context and the minimum level required in the extra-territorial context. The extradition of a person by a contracting state will raise problems under Article 3 where there are serious grounds to believe that he would run a real risk of being subject to treatment contrary to Article 3 in the requesting state: see, most recently, Sanchez-Sanchez, at [99]. "Serious grounds" in this context means "strong grounds": Ullah, [24].
Article 3 is not "relativist" in the sense suggested by Lord Hoffmann in Wellington. In an individual case, the question whether treatment in the requesting state will reach the Article 3 level of severity does not admit of a balancing exercise between the treatment on the one hand and the seriousness of the offence for which extradition is sought or the importance of the public interests in favour of extradition: Harkins & Edwards, [124]-[128]; Ahmad, [172]-[175]; Sanchez-Sanchez, [99].
However, the question whether treatment reaches the minimum level of severity required to engage Article 3 is intensely fact-sensitive and contextual. In a domestic case, the court is looking backwards at a concrete factual situation. In an extra-territorial case, the court is looking forward and attempting to gauge whether there is a real risk of Article 3 ill-treatment. Given the highly contextual nature of the assessment required, this may make it more difficult to establish a real risk of a breach: Harkins & Edwards, [130]; Ahmad, [178].
This is particularly so where the requesting state is one with a long history of respect of democracy, human rights and the rule of law, such as the USA: Harkins & Edwards, [131]; Ahmad, [179].
…
The question whether treatment reaches the high level of severity necessary to engage Article 3 depends on a holistic assessment of the conditions of detention. As to personal space, unusually, Muršic creates a bright line rule giving rise to a strong presumption of breach. As to other conditions of detention, it will be rare that one element taken on its own will be sufficient to trigger the application of Article 3 in the domestic context and, a fortiori, in an extradition case: see para. 64(c) and (d) above.”
Where matters have ‘moved on’ evidentially since the date of the district judge’s decision it is appropriate to adopt the approach taken by the Divisional Court (Stuart-Smith LJ and Jay J) in Modi v Government of India [2022] EWHC 2829 (Admin), [102]-[104], which is to concentrate more on the up-to-date position rather than on whether the judge's decision is legally flawed, and whether the further evidence is ‘decisive’ (Szombathely City Court v Fenyvesi [2009] 4 All ER 324).
A failure to protect a prisoner from violence may give rise to a real risk of inhuman or degrading treatment: see Jane v Lithuania [2018] EWHC 1122 (Admin), [16]. The test when the risk emanates from non-state actors (such as other prisoners) is whether the state is unable or unwilling to act to provide ‘reasonable protection’ to the requested person. It is incumbent on the requested person to establish not merely that he faces a real risk of suffering serious harm from non-state agents but that the receiving country does not provide for those within its territory a reasonable level of protection. The burden of proof does not therefore shift to the requesting state once a real risk of harm is demonstrated. In Bagdanavicius Lord Brown said at [24]:
“24. The plain fact is that the argument throughout has been bedevilled by a failure to grasp the distinction in non-state agent cases between on the one hand the risk of serious harm and on the other hand the risk of treatment contrary to article 3. In cases where the risk "emanates from intentionally inflicted acts of the public authorities in the receiving country" (the language of para 49 of D v United Kingdom 24 EHRR 423, 447) one can use those terms interchangeably: the intentionally inflicted acts would without more constitute the proscribed treatment. Where, however, the risk emanates from non-state bodies, that is not so: any harm inflicted by non-state agents will not constitute article 3 ill-treatment unless in addition the state has failed to provide reasonable protection. If someone is beaten up and seriously injured by a criminal gang, the member state will not be in breach of article 3 unless it has failed in its positive duty to provide reasonable protection against such criminal acts. This provides the answer to Mr Nicol's reliance on the UK's obligation under article 3 being a negative obligation and thus absolute. The argument begs the vital question as to what particular risk engages the obligation. Is it the risk merely of harm or is it the risk of proscribed treatment? In my judgment it is the latter. The very identification of the issue for determination by the House in the agreed statement of facts and issues illustrates the confusion:
‘If, on removal to another country, there is a real risk that a person would suffer torture or inhuman or degrading treatment or punishment from non-state agents, will removal violate article 3 ECHR, or must the person concerned also show that there is in that country an insufficiency of state protection against such ill-treatment?’
Non-state agents do not subject people to torture or the other proscribed forms of ill-treatment, however violently they treat them: what, however, would transform such violent treatment into article 3 ill-treatment would be the state's failure to provide reasonable protection against it.”
As to Aranyosi, in Mohammed v Portuguese Judicial Authority [2017] EWHC 3237 (Admin), the Divisional Court (Beatson LJ and Sir Wyn Williams) said:
“15. In Aranyosi, the CJEU decided that the consequence of the execution of an EAW must not be that the requested person will, if returned, suffer inhuman or degrading treatment. At [88] – [89], [91] – [92], [95] and [98] the CJEU set out the procedure that must be followed where the judicial authority of a member state is in possession of evidence of a real risk of inhuman or degrading treatment of individuals detained in the state that has issued the EAW.
Stage 1 of the procedure involves determining whether there is such a risk by assessing objective, reliable, specific, and properly updated evidence. I deal further with the type of evidence and what assessment is required at [50] – [51] below. A finding of such a risk cannot lead, in itself, to a refusal to execute the EAW. Where such a risk is identified, the court is required to proceed to stage 2.
Stage 2 requires the executing judicial authority to make a specific assessment of whether there are substantial grounds to believe that the individual concerned will be exposed to that risk. To that end it must request the issuing authority to provide as a matter of urgency all necessary supplementary information on the conditions in which it is envisaged that the individual concerned will be detained.
Stage 3 deals with the position after the information is provided. If in the light of that, and of any other available information, the executing authority finds that, for the individual concerned, there is a real risk of inhuman or degrading treatment, execution of the warrant must be postponed but cannot be abandoned.”
At [50] the Court said of Stage 1:
“50. In Aranyosi at [89] the Grand Chamber of the CJEU stated that:
‘the executing judicial authority must, initially, rely on information that is objective, reliable, specific and properly updated on the detention conditions prevailing in the issuing member state and that demonstrates that there are deficiencies which may be systemic or generalised, or which may affect certain groups of people, or which may affect certain places of detention.’
The CJEU stated that the information may be obtained from inter alia judgments of international courts, courts of the issuing member state, and also decisions, reports and other documents produced by bodies of the Council of Europe or under the aegis of the United Nations …”
Analysis
I accept that reports from the CPT have to be given very considerable significance and weight given the status and role of the CPT and its expertise and experience: see Urbonas v The Prosecutor General’s Office of the Republic of Lithuania [2024] EWHC 33 (Admin), [31]. The CPT is a committee operating under the terms of the European Convention for the Prevention of Torture and Inhuman or Degrading Treatment. It is a highly respected body which exists to prevent the ill-treatment of prisoners through dialogue with states: see Bartulis v Panevezys Regional Court (Lithuania) [2019] EWHC 3504 (Admin), [26]. These are factors essentially similar to those regarded by the Supreme Court in assessing the evidence of the United Nations Commissioner for Human Rights in R (AAA) v Secretary of State for the Home Department [2023] 1 WLR 4433, [64]-[70].
That said, I remain unpersuaded that the Appellant has made good his case even as to Stage 1 of Aranyosi and so I do not consider that any further Aranyosi enquiry is called for and I have concluded that the Appellant’s Article 3 challenge must fail. That is for the following reasons.
The starting point is the judge’s conclusion at [43] that the Romanian authorities have a ‘wealth of experience’ in the incarceration of prisoners convicted of sexual offences and that there was no evidence to the contrary. The judge was entitled to reach that conclusion.
Turning to the most up-to-date position in the 2022 CPT Report (which I have considered de bene esse), I do not consider this establishes that there is any systemic or generalised risk specifically to prisoners such as the Appellant who have been convicted of sex offences.
Although [72] the CPT referred to violence against sex offenders, and violence in general, it also noted that incidents had fallen since its last visit in 2018. This would tend to suggest that measures had been taken to reduce the problem. Further, although the CPT highlighted at [72]-[74] problems at particular prisons, none of these, according to the assurance from Romania, are prisons at which the Appellant is going to be detained (see the Respondent’s Skeleton Argument at [37]). There are pockets of violence (as unfortunately there will be in most prisons in most countries), but the picture which emerges is not one of systemic violence against sex offenders which the authorities cannot or will not stop.
The CPT at [73]-[74] made reference to shared cell risk assessments but again the context was not the targeting of sex offenders.
The Respondent also relies on the fact, as I have already mentioned, that the 2022 CPT Report was considered in Marinescu and found not to be decisive, at [46], [64], [66]:
“46. In material submitted after the hearing, the respondents invite the court to consider two documents published on 14 April 2022: the Report of visits to Romanian prisons, in May 2021, by the European Committee for the Prevention of Torture and Inhuman or Degrading Treatment or Punishment (‘CPT’) ("the Report"); and the Response to that Report by the Romanian Government (‘the Response’). The four prisons which were visited by the CPT did not include any of those at which the appellants are likely to be held. The Report (as the appellants point out) refers to overcrowding and generally poor conditions in the prisons visited; but the respondents submit that the Response (which is comprehensive) provides important evidence as to the steps being taken by the Romanian authorities. The Response acknowledges the continuing problems and reiterates the commitment of the authorities to ensure that prisoners are detained in conditions which fully respect their Convention rights. It refers to a decision taken by the NPA in 2022 to improve detention conditions and explains the way in which the NPA provides funds for ‘revamping’ of detention rooms and sanitary areas. It also explains the centralised arrangements for the supply of bedding; the periodic pest control measures; the rules for cleaning, sanitising and disinfection which have been intensified since the onset of Covid; and the establishment of "caloric values" for inmates' food.”
64. We do not find it necessary to seek any further information or assurances in accordance with the approach set out in Aranyosi. We conclude that the assurances which have been provided to the appellants satisfy the criteria encapsulated in the formulation adopted by the court in Sunca v Iasi Court of Law [2016] EWHC 2786 (Admin)
66. Having considered all the additional material de bene esse we are satisfied that, even taking it at its highest, it could not lead to a different conclusion. The proposed fresh evidence therefore could not satisfy the Fenyvesi test of decisiveness, and we accordingly decline to receive it.”
This passage is strongly against the Appellant’s case.
I therefore consider that the Respondent was right to submit that taking the evidence of the 2022 CPT report at its highest, it falls short of being decisive and so is not admissible. It does not show that child sex offenders are at any generalised risk in the Romanian prison estate. This is without my even having regard to the reply to the 2022 CPT Report described by the Divisional Court as being ‘comprehensive.’
Added to this is the 2022 assurance given by Romania. This is a ‘solemn promise’, binding as between the states concerned: Marinescu, [54] as to the conditions of detention. It provides:
“In order to ensure a safe environment and to prevent the occurrence of negative events that could affect the safety of the place of detention or of other persons in accordance with the provisions of Article 108, para 1, lit (e) of the Regulation for the implementation of Law No. 254/2013 on the execution of sentences and custodial measures ordered by the judicial bodies, approved by the Government Decision No. 157/2016, in the case of detainees for whom the security, order and discipline measures at the level of the penitentiary are not sufficient, there is a possibility that they will be transferred to other penitential units. We would like to point out that, in this situation, the detention conditions assumed by the National Administration of Penitentiaries will not be affected.”
From this I take that the prison authorities are aware of the importance of providing a for detainees – including sex offenders.
Of the 2019 CPT Report, Mr Swain referred me in particular to the following paragraphs: [70] (inter-prisoner violence and risk of sexual assault, especially for younger prisoners); [71] (violence and sexual assault at Bacău Prison); [72] (beatings and bullying at Bacău Prison); and [124] (lack of proper risk assessments for cell-sharing leading to risks of sexual assault for new prisoners).
These paragraphs do certainly make unhappy reading, however as far as the Appellant’s case is concerned, they are now five or six years out of date; as I have said, in its 2022 Report the CPT found the situation had improved; and, most importantly, they do not demonstrate a generalised risk to convicted sex offenders against which the authorities will not provide reasonable protection.
Overall, therefore I agree with this paragraph from the Respondent’s Skeleton Argument:
“61. The evidence required for Stage 1 of Aranyosi is not met. There is not objective, reliable, specific and properly updated evidence from the requisite sources that sex offenders are at real risk of inhuman or degrading treatment in Romanian prisons. There is no suggestion of any Romanian or Strasbourg judgments specific to this issue. There is no expert report making good the Appellant’s case. All there is is two CPT reports and the Appellant’s account. The two CPT reports fall significantly short of this. The most recent CPT report actually shows levels of inter-prisoner violence falling. At their highest they properly recount a number of isolated incidents of concern, but these incidents do not make out a systematic case affecting all sex offenders across the prison estate. Even the Appellant’s own account demonstrates that measures are in place through security alarms and movements of detainees to help ensure people’s safety.”
Conclusion
This appeal is accordingly dismissed.