England and Wales High Court (Administrative Court) Decisions

IN THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION
DIVISIONAL COURT

		Royal Courts of Justice Strand, London, WC2A 2LL
		22 September 2021

Stuart-Smith LJ:

1. This is the judgment of the Court to which we have both contributed.
2. Pursuant to the inherent jurisdiction of the Court, the Court has directed and implemented the anonymisation of the appellant referred to in this judgment as Mr A, his partner, his daughter or the police officers referred to as Mr X and Superintendent Y and directs that no information reasonably capable of identifying them shall be reported without further order of the Court. This step is necessary and proportionate to protect the safety of those persons so far as is reasonably practicable.

Introduction

3. The Appellants appeal pursuant to section 26 of the Extradition Act 2003 ("the 2003 Act") against orders for their extradition made respectively by Deputy Senior District Judge Tan Ikram on 30 July 2019 and District Judge Sam Goozée on 20 February 2020. The extradition of the Appellants is requested pursuant to European Arrest Warrants ("EAWs") issued by two French Judicial Authorities ("the Respondents"). The two appeals were joined by way of an order made by Julian Knowles J on 1 February 2021 as both cases raised the issue of prison conditions in France.
4. There is no reason to anonymise the participants in the Esmaili appeal from the order of DJ Goozée. But in the other appeal we refer to the appellant as Mr A and to various other participants by letter to protect their identity and safety so far as is possible.
5. In briefest outline, there are four grounds of appeal for determination:

i) Ground 1: Each Appellant submits that the Judges ought to have decided differently the question whether their extradition would amount to a violation of Article 3 of the European Convention on Human Rights ("ECHR") due to prison conditions in France;

ii) Ground 2: The Appellant Mr A submits that the Judge ought to have decided differently the question whether extradition was an abuse of process;

iii) Ground 3: The Appellant Mr A submits that the Judge ought to have decided differently the question whether there are substantial grounds for believing that extradition would give rise to a real risk of the execution of an arbitrary and disproportionate sentence contrary to Articles 5 and 3 of the ECHR;

iv) Ground 4: The Appellant Mr A submits that the Judge ought to have decided differently the question whether extradition pursuant to the EAW would be in breach of Article 5 as there is no mechanism under French law to challenge the proportionality of the EAW.

6. The Respondents submit that the decisions of the Judges below were correct. In summary:

i) Ground 1: the Respondents submit that the Appellants have provided no objective, reliable, specific and properly updated evidence to demonstrate that conditions in the prisons of Fresnes, Villepinte, Rennes Vezin, Lyon Corbas and Villefrance-sur-Saone are such as to breach Article 3;

ii) Ground 2: In relation to the Appellant Mr A, extradition would not amount to an abuse of process. There is no suggestion that the extradition proceedings are being usurped. The Respondent Judicial Authority has requested the Appellant's extradition for him to serve a lawfully imposed sentence of imprisonment. The circumstances as alleged by the Appellant do not amount to an abuse of process. In any event, the Appellant can provide no objective evidence to demonstrate that he was assured he would not be prosecuted for offences committed abroad. Quite the contrary. The grounds on which litigation was settled in Ireland do not disclose that such an assurance was made;

iii) Ground 3: In relation to the Appellant Mr A, extradition would not amount to a real risk of violation of Articles 3 and 5. The Appellant has been lawfully convicted. The Respondent has confirmed by a letter dated 12 October 2018 that should new evidence arise, the Appellant will have the opportunity make an exceptional application to appeal.

iv) Ground 4: In relation to the Appellant Mr A, extradition would not amount to a breach of Article 5. This issue was addressed in Parquet Général du Grand-Duché de Luxembourg and Tours C-566/19 PPU where the CJEU found that national procedural rules allow for the proportionality of the decision of the Respondent to issue an EAW to be judicially reviewed, where an individual has not been convicted. Additionally, the Updated Questionnaire and Compilation on the Requirements for Issuing and Executing Judicial Authorities in EAW Proceedings pursuant to the CJEU's Case-law identifies the mechanism in place for the proportionality of EAWs to be challenged where a sentence has been imposed.

The Background Facts – Mr Esmaili

7. The background facts in Mr Esmaili's case are unremarkable. Mr Esmaili was convicted in his absence of people smuggling after eight Iraqi nationals, including two who were underage, were found concealed in the back of a truck which he was driving to England in January 2016. He was sentenced to 15 months' imprisonment (of which 14 months and 27 days are left to serve) and permanent exclusion from French territory. He advances Ground 1 as the sole objection to his extradition.

The Background Facts – Mr A

8. The background facts in Mr A's case are complex and unusual. It is necessary to address them in detail in order to provide a framework for the discussion and application of the principles raised by Grounds 2 to 4. We do so in the knowledge that the court below declined to consider the factual basis for Mr A's submission because "the analysis of evidence and scope of the availabilityof any defence was, and is, a matter for the French courts applying their law and not one this court is able to reflect upon." In response to the submission that there was now new evidence that had not been available to the French courts, the court below took the view that the French courts provided a mechanism for revision hearings on fresh evidence.
9. Mr A's case arises from his arrest near Lyon on 7 August 2000 when he was returning to Ireland driving a lorry that carried a load of furniture in which was hidden 187 kgs of cannabis resin. From the outset he maintained that he was acting as an informer and under the supervision of the Irish Police, An Garda Síochána ("the Garda"), and that he had been assured that he would be protected from criminal process both in Ireland and abroad (including France). As a matter of fact, the Garda had not notified the French authorities about what was happening and Mr A was not immune from French criminal process. Furthermore, the information that was subsequently provided by the Garda was not sufficient to persuade the French prosecuting authorities not to prosecute or the French Courts not to convict and sentence Mr A to a substantial sentence. Mr A appealed to the Court of Appeal in Lyon and thence to the Cour de Cassation, France's highest court of appeal, which rejected his appeal on 3 October 2007. His conviction was upheld and he was sentenced to four years' imprisonment and permanent exclusion from French territory, along with forfeiture of his lorry and a fine.
10. There can be and is no criticism of the French proceedings or outcome on the basis of the information that was available to the French courts at the time. Mr A's appeal to this Court is founded on the submission that the French courts were not given full and proper information by the Irish authorities and that, if they had been, there is at least the prospect that the outcome would have been different. Specifically, had the true extent of his involvement and cooperation with the Garda been made known to the French prosecuting authorities they may have exercised a discretion not to prosecute or, if they did, the French courts may have declined to convict Mr A or may have imposed a lesser sentence that did not require him to serve a sentence of imprisonment.
11. In order to examine this factual submission, it is convenient to start with the information that was provided to the French prosecuting authorities and courts and the impact that appears to have had on those proceedings, and then to examine the evidence which Mr A relies upon to found his submission that the information provided by the Irish authorities was incomplete and inadequate.
12. The Court has extracts and core documents from the French investigation file. These documents were available to the court below but, as we have indicated, the Judge declined to make even a limited enquiry into what happened between Mr A and the Garda. We look at them solely to see whether the factual basis underlying Mr A's submissions are well founded – not with a view to criticising or commenting on the French internal procedures.
13. Mr A was interviewed by the French police on 8 and 9 August 2000. He said that he had entered into an agreement with the Garda, identifying his two contacts as Mr X and Superintendent Y. He said he had previously trafficked tobacco but contacted the police after being approached to import drugs. He said that "they proposed that I inform them during my various deliveries. They promised me that if they were able to seize drugs as well as the traffickers they would keep me out of it." He described previous runs where he had provided information and that he had been paid by the Garda for the information he had provided. When asked about the exact terms of the deal with the Garda, he said that he would continue to make money (from the traffickers) carrying out the traffic "while informing the police subject to nothing happening to me. It was agreed that for arrests and seizures of drugs, the police hide my involvement. I spoke to the police in case I had problems with the foreign authorities concerning my "load", they told me that as the drugs would always be hidden in the furniture, I only had to deny all knowledge of these drugs which I was transporting. In the event this wasn't taken into account the Irish police told me that they would take care of everything. I haven't had any reason not to believe [Superintendent Y] because he often travelled abroad in connection with other European police." Later he was asked whether making money was his sole motivation for informing the police, to which he replied that "My main motivation was money, at no time was I forced to collaborate with the police but I would never have done it otherwise as it was too risky." He identified phone numbers on his mobile as including those of Mr X and Superintendent Y and identified one call as being with Mr X so that he could meet Superintendent Y before his departure on the trip; he identified another as being to a business associate whose role was to pass information on to the Garda.
14. Immediately after his arrest, Mr A's wife went to France, where she too was interviewed on 9 August 2000. Her (hearsay) evidence was that Superintendent Y had encouraged her husband to transport drugs, while informing him, so that the Irish police could make arrests and seizures. Mr A had told her that he transported drugs in collaboration with Superintendent Y "on the understanding that if he had problems [Superintendent Y] would take care of the situation." She said he had met the two police officers on the day before he set off on the current trip; and that she had been told that Superintendent Y was going to Dublin "to try to contact someone at the Police to sort out the problem."
15. On 14 August 2000 the examining Magistrate submitted an International Rogatory Commission request to the Irish Minister of Justice and authorities in which she set out Mr A's account, including that he had explained that he was transporting the narcotics "under the surveillance of the Irish Police." It recorded Mr A's assertion that on each run he had kept his business associate informed of his movements, with that person being responsible for keeping contact with the police; and that Mr A was in possession of the private phone numbers of Mr X and Superintendent Y. The investigating magistrate asked the competent Irish authorities for information including confirmation of the phone numbers for Mr X and Superintendent Y and for them to "indicate whether Mr A has worked in collaboration with Irish police: manner; duration; results in terms of arrests and drug seizures."
16. The Irish authorities responded formally on 9 November 2000, by which time they had received a visit from a French police officer who was acting pursuant to the International Rogatory Commission. They provided a typed statement from Superintendent Y, the salient parts of which were:

"[Mr A] was unknown to Superintendent [Y] and Garda [X] prior to June 1999. He was introduced to both Gardai by …, his business partner. Through receipt of confidential information, both Gardai suspected Mr [A] was previously involved in the importation of smuggled tobacco and drugs.

After a number of meetings, [Mr A] indicated to both Gardai how he had been approached to bring a consignment of drugs to Ireland from the Continent. [Mr A] indicated he needed the money and was prepared to do the drugs run for the money. He also indicated he would be prepared to assist the Gardai in the recovery of the drugs and arrests if they wished. Both Gardai spoke to [Mr A] of the dangers involved for both himself and his family and also the risk of being arrested in the Continent in possession of drugs. [Mr A] indicated on a number of occasions he would proceed with the drug importations whether the Gardai wanted them or not as he needed the money.

The predicament facing both Gardai was as follows:

A) Did they listen to the information he was prepared to give them voluntarily with the hope of making sizeable captures of illegal drugs and the arrest of some of the top Drug Barons in Ireland?

B) Did they ignore this, refuse to have anything to do with [Mr A] and allow the drug shipments to Ireland? Without full time surveillance on [Mr A] (which was not viable) Gardai would not have been in a position to ascertain if he was bringing drugs into the country or not.

It was decided to run with option A and as a result the captures [referred to in a conversation with the French visitor to Ireland] came about.

[Mr A] was never in a position to tell either Gardai of exactly where in Europe he was to pick up the drugs until he returned to Ireland. Contact was only made with him while en route as to whether or not the drugs were available for collection and where. He sometimes went to Europe in the expectation of getting a consignment of drugs to bring back to Ireland, but for various reasons, the organisers of such importation cancelled them while he was en route.

[Mr A] did provide invaluable assistance to Gardai in the past which lead to the capture of almost £3,000,000 worth of drugs, the arrest of some of the biggest drug dealers in this country, and the recovery of guns and stolen property. He was an excellent informant.

I have little doubt, but that if he managed to return to Ireland with the drugs he was found with in Lyon in August, 2000 that Gardai would have been in a position to recover them and make arrests.

Both Superintendent Y and Garda X wish to emphasise the following.

A. There was no personal gain in this for either Garda.

B. [Mr A] was not forced or coerced in any way to carry out this work.

C. It was [Mr A] himself who offered to assist the Gardai but indicated because he needed cash he was willing to do such work whether the Gardai wanted to get involved or not.

D. He had been paid a financial reward following each of the previous two captures from Secret Funds.

E. Mr A was aware of the dangers involved in such work and of being arrested in possession of drugs in Ireland and the Continent."

17. A hand-written statement from Superintendent Y dated 8 November 2000 stated that on the trip that led to Mr A's arrest the Garda "were aware the night before he left that he was going to Europe." He said that, with another officer, "we spoke with [Mr A] and explained the dangers to him. He always understood this but because of his greed he was willing to take the chance himself despite our warning. [Mr A] knew he was taking serious risks. He was always warned as to the dangers."
18. The main points of difference between what Mr A was saying and the information provided by the Irish authorities were:

i) Mr A admitted to having trafficked tobacco; but the Irish authorities said that they had suspected him of trafficking drugs before he came to them to offer his services as an informant;

ii) Mr A said he had been assured by the Garda that "they would keep him out of it" and that if anything happened abroad "they would take care of everything"; but the Irish authorities said that it had been made clear to him that he travelled on the continent at his own substantial risk;

iii) Mr A said that, although money was his main motivation, he would not have undertaken the trafficking journeys without the protection of his collaboration with the Garda because it was too risky; but the Irish authorities said that his greed was such that he was going to go ahead with the trafficking in any event.

19. The materiality of these differences is reflected in the Judgment of the Lyon Court of Appeal in its judgment on 9 January 2007, to which we turn in a moment. In the meantime, however, receipt of the information from the Irish authorities was sufficient to persuade the French authorities to release Mr A. He maintains that he understood that his "unconditional" release meant that there would be no further proceedings against him; and there is evidence that others in Ireland shared his mistaken assumption. The Court below, having heard his evidence on this point, did not accept that he genuinely thought that the case was over on his release in November 2000. That said, any misunderstanding became clear when the French authorities mounted a prosecution against him. He did not attend in person. After a re-trial, at which he was represented in his absence, on 12 September 2005 he was convicted by the criminal court in Bourg-en-Bresse of multiple charges relating to the illegal trafficking of narcotics. He was sentenced to 3 years' imprisonment, a fine was imposed, his lorry was confiscated, and he was banned from French national territory for 10 years.
20. Mr A appealed against both conviction and sentence to the Lyon Court of Appeal. In its judgment, the Court of Appeal recorded amongst the facts that had arisen from the proceedings and debated the following:

i) Mr A's account, broadly following the terms of his initial interviews, that he had been carrying out the transport of the drugs under the surveillance of the Irish police; that it had been agreed between Mr A, Mr X and Superintendent Y that Mr A would carry out deliveries of narcotic substances about which he was to inform the police officer, who afterwards would take people in for questioning; and that there had been two trips after which Mr A had received a financial reward;

ii) Mr A's associate's account was that he would be notified of the journeys by regular telephone calls from Mr A, and that he would inform the police of the moves of the lorry;

iii) "Superintendent [Y] confirmed the bulk of these declarations, as well as those expressed by [Mr A] before the investigating judge. [Mr A] had been introduced to him by [his associate] during the month of July 1999. Without any duress, [Mr A], warned of the penal risks which he could incur abroad, had carried out three journeys, …. Finally, the import from Spain of the bars of cannabis resin was to give rise, logically to new questionings."

21. In response to Mr A's assertion that he was acting in agreement with the Irish police, the judgment stated:

"Whereas [Mr A] has acknowledged having imported these narcotic drugs into France and has claimed to be acting in agreement with the Irish police who have confirmed his allegations; whereas, nevertheless, this agreement, of which it is not even contended that the French authorities had been advised, does not constitute a case of irresponsibility^[1];

Whereas in effect, the French authorities have never been asked, either by Superintendent [Y], or by any other Irish authority, and contrary to all the principles of international mutual repressive^[2] assistance, to implement the French legislative provisions concerning surveillance of the transport of substances being used to commit crimes or offences of narcotic drugs trafficking, measures provided for by the United Nations Convention against Illicit Traffic in Narcotic Drugs and Psychotropic Substances adopted in Vienna on 19th December 1988, to which Ireland and France are both party;

Whereas likewise, only the officers of the criminal investigation department can be authorised to carry out undercover operations subject to strict arrangements concerning their execution, and not suspect individuals, habitually indulging in dubious illegal activities and prey to financial difficulties; whereas [Mr A] does not thus assume the status of undercover agent, which, moreover, has never been awarded him by the Irish police authorities;

Whereas in this regard, Superintendent [Y] and the police officer, [Mr X], have affirmed that [Mr A] was already suspected of importing into Ireland smuggled tobacco, which he has acknowledged, and narcotic drugs when he offered his services to them, explaining that he needed money and that he was ready, in any case, to indulge in drugs trafficking whether that interested the police services or not; whereas they have stated that [Mr A] was perfectly informed of the dangers inherent in this activity and the risks of being arrested in possession of narcotic drugs in Ireland and on the continent; whereas [Mr A] cannot describe as inoperative the declarations of those he considered as his "protectors", without producing the slightest probative argument in support of his allegations;

Whereas in this way, whatever is true of the agreements made by the narcotic drugs importers and the Irish police, it has been perfectly established that [Mr A] has knowingly committed the offences of importing, acquiring, holding and transporting without authorisation 187 kilograms of cannabis resin, offences committed in France, of which he must be, by confirmation of the appealed judgment, declared guilty; whereas the customs crime of the irregular circulation of prohibited goods reputed to have been smuggled, in this instance cannabis resin, is equally settled;"

22. Later in its judgment, the Court of Appeal stated, in relation to sentence, that:

"Whereas there is scarcely the need to underline the seriousness of the acts of drugs trafficking committed by the defendant, who remained remarkably circumspect on the conditions in which he was supplied and being very careful not to supply the slightest indication which would make it possible to catch the importers who will be able to pursue their activity as soon as they have found a replacement for him; whereas his conscious participation in such acts which are as lucrative for the perpetrators as they are harmful to public health, justifies sentencing him to 4 years imprisonment;

…[T]he interested party had only come to France in order there to commit the offences being prosecuted."

23. The Court of Appeal confirmed the sentence imposed by the court below, save that it increased the term of imprisonment from 3 years to 4.
24. Mr A appealed to the Cour de Cassation. In its judgment on 3 October 2007, the Cour de Cassation repeated the passage from the Lyon Court of Appeal set out at paragraph 21 above. It appears that the submission on the first ground of cassation was that, notwithstanding the failure to notify the French authorities in advance and that Mr A was not an officer or agent of the French judicial police he should have been treated as having the status of an infiltrated agent, which would confer immunity upon him. The second ground of cassation was that the sentence was not justified by general considerations about the lucrative and harmful nature of drug trafficking given that "the operation of infiltration in which Mr A took part was precisely to put an end to such intrigues". The appeal was rejected.
25. The judgments of the French courts have in common that, while it was accepted that Mr A had acted as an informer, the Irish police maintained that he crossed the continent at his own risk; that the Irish police had not informed the French authorities in advance of what was going on; that Mr A would have undertaken the trafficking irrespective of his position as an informant; and that the claim to be treated as an infiltrated agent, which could provide immunity, was rejected because of the lack of notice and the fact that Mr A was a private individual and not an agent of the judicial police. They also have as a central feature of their reasoning that the conviction and sentence were to be imposed "whatever is true of the agreements made" between Mr A and the Garda i.e. even if Mr A's account were to be accepted in full.
26. Mr A submits that the Irish police failed to inform the French prosecuting authorities that they had "authorised the drugs run and promised him immunity both in Ireland and overseas". It is plain from the passages that we have set out above that the Irish authorities confirmed that Mr A had been an informant in relation to the trafficking of illegal drugs, that they had known in advance that he was going on the journey that led to his arrest, and that they would have taken advantage of information provided to them by Mr A about that journey. Whether they "authorised" it is more equivocal. Their evidence to the French authorities was that, knowing of the journey in advance, they advised Mr A of the risks and advised against it but he said he was going anyway. Mr A's case is that there were no such reservations or contrary advice and that he would not have gone but for his role as an informer. What the Irish authorities flatly denied was that Mr A was promised immunity from prosecution when abroad. To the contrary, their evidence was that they emphasised and reminded Mr A that he was taking substantial risks of arrest and prosecution when outside Ireland.
27. Whether Mr A was promised by the Irish authorities that he would have immunity from prosecution when abroad remains in issue between Mr A and the Respondent in the present proceedings; and it will be necessary to examine the evidence, including evidence that has come to light since the hearing in the court below.
28. Before considering the material that Mr A says compels the conclusion that he was promised immunity while in France, two preliminary points should be noted. First, this is an appeal pursuant to s. 26 of the 2003 Act, which limits the circumstances in which this Court may intervene. Second, in relation to Mr A's abuse of process argument, the court below declined to make any findings of fact despite having heard evidence from Mr A and three others, including Mr A's Irish solicitor.
29. As appears from the expert evidence to which I refer later, the issue about Mr A's arrangements is important for issues 2, 3 and 4 because it is Mr A's case that the French courts convicted him on a false basis, that the outcome would have been different if the French courts had known what he says was the true position, and that there is no longer any means of overturning his conviction or adjusting his sentence. He submits that it would be an abuse of the extradition process to surrender him to serve a sentence which he cannot appeal or reopen (Ground 2); that he has no judicial protection in respect of the decision to issue the EAW or the proportionality of that measure, and that the lack of such protection would render his detention arbitrary and in breach of Articles 5 and 13 ECHR (Ground 3); and that the French criminal system does not provide him with an effective means of seeking a review of his sentence with the result and disproportionate sentence contrary to Articles 5 and 3 of the Convention (Ground 4).
30. The information upon which Mr A says that the Court should rely in order to conclude that he was convicted on a false basis comes in part from his own witness evidence and that of his partner and in part from the file of the French investigation file to which we have referred. In addition, in 2006 Mr A commenced High Court proceedings in Ireland against the Commissioner of the Garda and the Attorney General of Ireland ["the Irish proceedings"]. In the course of the Irish proceedings, disclosure was given which has been made available for use in the English courts and which provides information about the relationship and dealings between Mr A and the Garda. The Irish proceedings were listed for trial on the issue of liability in November 2019, after the hearing of the present extradition proceedings in the court below, and was heard in the High Court over five days from 20 to 27 November 2019. Mr A gave evidence on 20 and 21 November 2019. On 27 November 2019, before Mr A's case had been closed, the parties entered into settlement discussions which led to the case settling that day on terms which included an admission of breach of a duty of care owed by the Garda to Mr A.
31. We do not have the pleadings from the Irish proceedings, but we have a partial description in an interim judgment of Baker J on a claim for privilege, which was delivered in November 2016. The description is:

"The plaintiff claims damages against the defendants for negligence, breach of duty, breach of contract, in quasi-contract, misrepresentation and for breach of constitutional rights, and under the European Convention on Human Rights. The claim arises from the role engaged by the plaintiff at all material times as a garda informant. The plaintiff claims that he entered into an agreement with the [Garda] in July, 1999 that he would so act as informant in relation to the activities of certain persons thought to be engaged in drug trafficking activities believed to be criminal in nature. The plaintiff claims that it was an express term of the agreement and/or that it was represented to him that the first defendant would take all necessary steps to safeguard his safety, keep his identity confidential and protect him from criminal prosecution in the State, or abroad.

…

It is not denied by the [Garda] that the plaintiff did perform some surveillance role or role as a person providing information, nor is it denied that payments were made to him in consideration of information that he provided to the first defendant."

32. It can therefore be inferred that Mr A's central contention in the Irish proceedings was that it was an express term of his agreement with the Garda that they would take all necessary steps to safeguard his safety, keep his identity confidential and protect him from criminal prosecution with Ireland or, critically, abroad. Even without access to the pleadings and in the absence of evidence suggesting that English and Irish law differ in this respect, it is easy to see how such an agreement, if proved, could give rise to a duty of care owed to Mr A by the Garda. It is also possible that, if such a representation were to have been made, the maker of the representation would have owed a duty to Mr A in making it: but that does not materially add to the argument as the making of the representation would have provided the basis for the duty to take reasonable care to protect him from arrest or prosecution.
33. All that is known about the contents of the Defence is that an affidavit filed on behalf of the Garda in March 2018 by Superintendent Y said of it that "At paragraph 7 it is denied that there was ever an agreement between [Mr A] and [Superintendent Y]." That denial is also implicit in the judgment of Baker J to which we have just referred.
34. We have reviewed all of the documentation disclosed in the Irish proceedings and now before the Court. It demonstrates that Mr A and his partner asserted the existence of such an agreement immediately upon his arrest in France and consistently thereafter. Amongst other information, the direct numbers for Mr X and Superintendent Y were stored on Mr A's phone, which also showed that he had received a call from Mr X the evening before he set off on the fateful trip. The Garda's internal documents also show a marked change of attitude towards Mr A after his arrest. Documents until then show enthusiasm for Mr A as an extremely valuable ("immensely reliable") informant who was to be cultivated in the hope and expectation that he would provide further valuable information. The change of attitude is exemplified by the report of Superintendent Y on 10 August 2000 which asserted that neither he nor Mr X "would have been aware of [Mr A] being in possession of drugs outside the country at any time. The only information passed on [by Mr A to Mr X and Superintendent Y] related to drugs in Ireland and when it was in Ireland."
35. The documentation taken as a whole lends strong support to a submission that the Garda underplayed Mr A's role both internally and in their dealings with the French authorities in the aftermath of the arrest. It also appears that the Irish authorities may well have thought that, when Mr A was released from custody in November 2000, that was and would be the end of the matter. Unfortunately for Mr A, both he and the Irish authorities were wrong about that and their misunderstanding could not and did not influence the subsequent course of events. However, what the contemporaneous documentation does not do is to prove the existence of an agreement that the Garda would keep Mr A safe from arrest or prosecution when abroad.
36. The existence of an agreement to protect Mr A safe from prosecution in Ireland was eventually acknowledged by Superintendent Y in his affidavit sworn for the Irish proceedings in March 2018. However, Superintendent Y maintained the denial of any agreement relating to prosecution abroad: "It is not and could not be within the power of the Garda Siochana to interfere with criminal proceedings of a Court in another jurisdiction. … It is denied that [Mr A] was given any assurance as to his safety from prosecution should he be discovered in possession of drugs outside the jurisdiction." This remained the position until the trial of the Irish proceedings.
37. We are not privy to the thinking of the Garda in deciding to enter into settlement negotiations and, in any event, their thinking would not be admissible as an aid to construction of the settlement agreement or as evidence for this Court about what happened in 2000. However, the settlement agreement represents the formal position of the Defendants in the Irish proceedings and is in the following terms:

"The Plaintiff and Defendants have agreed, in full and final settlement of the above proceedings and all claims howsoever arising between the parties (to include all claims arising out of the European Arrest Warrant proceedings in the United Kingdom and France) to compromise same on the following terms:

1. The Defendants shall pay to the Plaintiff the sum of €300,000 …, comprising €250,000 by way of damages and €50,000 by way of contribution towards the Plaintiff's costs in defending the European Arrest Warrant proceedings in the United Kingdom, … .

2. It is hereby acknowledged by An Garda Siochána that [Mr A] was a participating informant between 1999 and his arrest in France in August 2000, that An Garda Siochána owed a duty of care in that regard, and that An Garda Siochána breached their duty of care to [Mr A].

3. The Defendants shall pay to the Plaintiff the costs of these proceedings, … ."

38. The Respondent submitted that we should speculate about the reasons that may have underlain the decision of the Defendants in the Irish proceedings to settle. We disagree: by definition any speculation would lack the certainty that could make it reliable or useful for the Court, quite apart from its inadmissibility as an aid to construction. In our judgment the only sure guidance is to be derived from the terms of the settlement agreement itself. There are three features of the agreement that stand out. First, the admitted duty subsisted between 1999 "and [Mr A's] arrest in France in August 2000." It follows that the admitted breach of that duty occurred on or before his arrest. Second, the fact that a separate sum is allocated as a contribution to Mr A's costs in these proceedings demonstrates that the settlement is directed to the consequences of Mr A's arrest in France, including these proceedings. Third, for the admitted breach of the Garda's duty of care to be relevant to his arrest and its consequences, the duty must have involved (at its lowest) a duty to exercise reasonable care to protect Mr A from that arrest. It follows, in our judgment, that the settlement can and should be taken as establishing acceptance by the Garda that it owed Mr A a duty to exercise reasonable care to protect him from arrest while in France and that the Garda (a) breached that duty and (b) by that breach of duty caused Mr A to suffer relevant damage.
39. The settlement agreement does not establish the existence of an express agreement as alleged by Mr A; but that does not matter. Even if there were no express agreement, the settlement agreement acknowledges that the Garda's dealings with Mr A gave rise to the relevant duty and that they breached it. On this basis, when the Garda informed the French authorities in 2000 that it had been made clear to Mr A that he travelled on the continent at his own risk, they misrepresented the true position. To be accurate they should have told the French authorities that he travelled to France in circumstances where the Garda had assumed a duty to protect him from arrest or prosecution (whether by express agreement or otherwise).
40. The settlement agreement is not specific about how the Garda should or could have complied with their acknowledged duty. Since the operations necessarily involved the transporting of drugs through foreign jurisdictions, it seems implausible that the duty could have been discharged by warning Mr A that he was at risk of arrest abroad and that the Garda could do nothing about it; and, in any event, if that had been a way in which the duty could have been discharged, the admission of breach demonstrates (for present purposes at least) that it was not.
41. At this point, the difficulty identified by the decisions of the French courts comes into focus. The Garda did not alert the French authorities to Mr A's passage through France in advance. But even if they had done so, it is the evidence of the experts in French law and is reflected in the judgments of the French courts that legal immunity is not available to participating informants who are private citizens. French legislation only provides immunity where the participating person is a police agent or a customs agent. They, and only they, are granted legal immunity for any illegal act that they might have to commit, including acquiring, selling, or transporting drugs in order to pose as an accomplice. Such agents are also allowed to acquire drugs in order to reveal the underlying offence; but again this immunity does not extend to private individuals such as Mr A. Hence the references (a) in the first paragraph cited at paragraph 21 above to immunity not being available even if the Garda had notified the French authorities and (b) in the third paragraph to authorisation to carry out undercover operations being available only to officers of the criminal investigation department.
42. In the absence of any legal immunity, it is common ground that the French authorities were fully entitled to act as they did. That leaves the question whether it would have made any practical difference if the Garda had disclosed to the French authorities the full extent of their involvement with and duty to Mr A. This question is the subject of evidence from suitably qualified experts in French law.
43. The expert evidence is contained in two reports served on behalf of Mr A, dated 6 July 2018 and 14 May 2021 provided by experienced French criminal lawyers, Aurelia Grignon and Thiibault Kempf of Soulez Larivière, whose expertise has not been questioned. Their first report was available to the court below; their second was not. The Respondent, while not formally agreeing their evidence, has not submitted any expert evidence to contradict or qualify it.
44. In their first report, the experts were asked what would have been the outcome of the French proceedings if the Garda had confirmed to the French authorities that Mr A had been led to believe that he would be protected from prosecution overseas. Specifically, they were asked (1) whether, as a matter of practice, the French authorities would have proceeded to prosecute; (2) whether Mr A may have been acquitted, whether by arguing that his prosecution was akin to an abuse of process or for some other reason of law or fact; (3) whether his conviction would have been quashed on appeal; and (4) whether some other form of clemency would have been available to him. In addition, the experts were asked whether, if his claim in the Irish proceedings were to be successful he could apply to reopen his conviction and, if so, what would be the prospects of success of such an application.
45. In answer to the first question, the experts point to the existence of a discretion vested in prosecutors not to initiate a prosecution where it would be "inopportune" to do so and to it being, as a matter of practice, common for French prosecutors to guarantee offenders a protection from prosecution in return for valuable information. Their opinion, for reasons given, is that "had the Gardai confirmed to the French authorities early on in the investigation that [Mr A] was in fact an extremely valuable informant benefiting from protection in Ireland and that they had told him he would be protected from prosecution overseas, … based on our experience, we think there would have been a fair chance that the Prosecutor decide not to prosecute him. … By "fair chance" we mean that there would have been sufficient factual grounds for the Prosecutor in charge to recourse to its discretionary power … and not to prosecute even though [Mr A] was involved in the drug trafficking." They express this opinion despite the fact that the Irish authorities had not given prior notice or initiated the mutual cooperation that would have been necessary for Mr A to be rendered immune from prosecution as a matter of law rather than of discretion if that immunity had been available to him. In the absence of prior notification, had the Gardai notified the prosecutor subsequently of the position, it would not have bound the prosecutor and "he or she would not have been required to drop the criminal proceedings against [Mr A]. It is therefore impossible to predict conclusively that a prosecution would not have been launched." In the course of answering this question, the experts refer to the fact that different states within the EU have different rules about who may or may not be granted immunity and in what circumstances and to the difficulties that such different rules create for international cooperation in relation to cross-border drug trafficking. This, in our judgment, is a salutary reminder that the Court is operating in an area of mutual trust where different approaches are adopted in different jurisdictions for reasons that are properly within the ambit of those jurisdictions' margin of appreciation.
46. In answer to the second question, because Mr A was not a member of the police or customs services, he could not benefit from legal immunity from prosecution – as was noted by the Lyon Court of Appeal and the Cour de Cassation. However, it is the experts' evidence that, "it is possible that the confirmation of the Gardai that [Mr A] was promised immunity would have led the Court to find the components of the offence did not exist in this case." However, the experts identify significant difficulties in the way of such an approach. Specifically, there is the Lyon Court of Appeal's ruling, which we have set out above, that whatever the nature of the agreements between Mr A and the Gardai, Mr A had knowingly committed the offences. Having pointed out that the Lyon Court of Appeal was evidently influenced by the information provided by Superintendent Y, their (suitably guarded) opinion is that "it is possible that [the Lyon Court of Appeal], or another, presented with information which substantiated Mr [A's] good faith, would have had a different appreciation of the criminal intent. However, it is also possible that even if there was an admission by the Irish police that they had promised him immunity overseas (as well as in Ireland) [that] would be insufficient on its own to obtain an acquittal." In considering whether there were any possible grounds for an "excuse from liability" the experts emphasise that "French judges very rarely accept that an error of law has been made and often oppose that "no one is supposed to ignore the law". This could prove even more difficult to demonstrate here, as it would require for the French court to accept that the error of law could legitimately result from the false information given by a foreign authority."
47. In answer to the third question, the position would be the same on appeal as before the first instance court, because they rely on the same laws and case law, though the appeal court must make its own assessment. Accordingly, an appeal court could in theory reverse either an acquittal or a conviction.
48. In answer to the fourth question, it is the experts' view that, if convicted, it would have been open to Mr A to argue for a reduced sentence if the Garda had acknowledged that they were following the journey closely and coordinating with Mr A to make arrests. The experts, while identifying possible arguments and outcomes, express no opinion on what would have been the outcome had Mr A's case been fully confirmed by the Irish authorities.
49. In summary, the experts in their first report are unable to say that the outcome would have been different if the Irish authorities had accepted that they had given Mr A an assurance that he would not be subject to criminal proceedings if stopped on the continent; the most that they are able to say is that there is a "fair chance" (as explained above) that the prosecutor would not have instituted criminal proceedings and, if criminal proceedings had been instituted, that it is "possible" that a court would have acquitted Mr A or, if convicted, would have imposed a lesser sentence (including a sentence that would not involve serving a sentence of imprisonment). The French prosecuting authorities and courts would have been under no obligation to act in this way.
50. In answer to the questions about the possibility of reopening his conviction if the Irish proceedings were to be successful, the experts identify the existence in French law of revision proceedings (a "demande en revision"), which they describe as an extraordinary remedy. Pursuant to Article 622 of the French Procedural Code as modified in 2014 "the review of a final criminal decision may be requested for the benefit of any person convicted of a crime or offence when, after a conviction, a new fact or an information, unknown to the jurisdiction on the day of the trial, likely to establish the innocence of the convicted person or to cast a doubt regarding his guilt, is revealed." While accepting the theoretical possibility of making such an application if the Irish proceedings were successful, they identify that the prospects of success of such an application depend upon the ability to cast a doubt upon Mr A's guilt. As a matter of fact, applications "do not often prosper" and those that do are those "where the new information … clearly demonstrated, above all doubts, that the person was innocent." The problem for Mr A, as identified by the citations from the judgments of the French courts to which we have referred, is that he did not have immunity for the actions he knowingly carried out. The experts' assessment is that the prospects of a revision would be low because the Court would probably focus on the lack of mutual assistance or legal immunity rather than Mr A's understanding and because revision decisions, which are quite unusual, tend to rely upon fresh evidence "that leave no margin of interpretation regarding the person's guilt."
51. As a matter of chronology, we record that, in the light of the settlement of the Irish proceedings, the French prosecutor was asked whether she intends to withdraw her application for Mr A's extradition. She has confirmed that she does not intend to do so.
52. The experts' second report was prepared in May 2021 in the light of the settlement. The experts were asked six questions: (1) whether there is an effective remedy for Mr A to contest the issue or maintenance of the EAW or the underlying arrest warrant other than by seeking to reopen the case of his conviction or sentence; (2) whether Mr A could bring an action for a declaration of invalidity under Article 170 of the French Code of Criminal Procedure and, if so, whether that would give him an effective remedy to contest the maintenance of the EAW or underlying national arrest warrant; (3) whether the French prosecutor has any discretion whether to withdraw an EAW in a conviction case; (4) whether, if the prosecutor has a discretion to withdraw an EAW in a conviction case, a refusal to do so can be challenged; (5) whether the admission of liability in the Irish proceedings alters the opinions expressed in their first report; and (6) whether there is any effective mechanism by which Mr A could obtain a review of his sentence (as opposed to his conviction).
53. In answer to the first two questions, it is the experts' view that there is no effective remedy available under French law to contest the issue or maintenance of the EAW or the underlying national arrest warrant other than by trying to reopen his conviction. Article 170 only applies during the investigation phase and not after conviction.
54. In answer to the third and fourth questions, it is the experts' view that the prosecutor has no discretion to withdraw the EAW, because of express provisions of the French Code of Criminal Procedure.
55. In answer to the fifth question, the experts' view as expressed in their first report is unchanged. Given the "laconic nature of the admissions", it is their opinion that "the Cour de Revision would find them, at the very least, subject to interpretation." They re-emphasise a point made in their first report, that the timing of the admission is important because it comes at a time when all other available judicial avenues are exhausted: accordingly they assess the chances of success as "extremely limited."
56. Finally, in answer to the sixth question, it is the experts' view that conviction and sentence are inextricably linked and that sentence cannot be reopened without a reopening of the whole case. That said, after serving 2/3 of the sentence or when there is less than 2 years left to serve, application could be made for the balance of the sentence to be served outside prison.

The Grounds of Appeal

57. We now turn to consider the various grounds of appeal. We consider Ground 2 first while recognising that abuse of the process is a residual jurisdiction.

Ground 2: Abuse of the Process

58. The twin pillars of the Appellant's case on this ground are (i) the failure of the Garda to disclose properly to the French investigating magistrate the nature of their agreement (or arrangement) with Mr A, and (ii) that Mr A now has no realistic prospect of appealing his conviction or reopening the case against him and, in any event, the French court has no jurisdiction to protect against unfairness caused by a third state. He submits it would be manifestly unfair for him to be extradited when he committed the offence as a participating informant for the Garda, having been assured by Garda officers of protection and immunity both in Ireland and overseas.
59. English law has developed an abuse of process doctrine by which the English court protects its involvement in the extradition process against manipulation or usurpation. Throughout the relevant authorities one recurring theme is that the mutual trust between states that are party to the Framework Decision informs the entire scheme. A second recurring theme is that the doctrine focuses on abuse by the requesting judicial authority. The history of the development of the doctrine of abuse as it applies in extradition proceedings was reviewed by a different constitution of this court in Wawrzyczek v The District Court in Bielko-Biala, Poland [2021] EWHC 64 (Admin) and it is not necessary to repeat that exercise in full. However, the following citations illustrate the importance attached to each of the recurring themes.
60. In R (Bermingham) v Director of the Serious Fraud Office [2006] EWHC 200 (Admin), [2007] QB 727 the court held that the appropriate judge conducting an extradition hearing under Part 2 of the Extradition Act 2003 has a discretion to stay proceedings as an abuse of process in order to ensure that 'the [extradition] regime's integrity' is not usurped by abuse of process. At prargraph 100 Laws LJ, giving the judgment of the court, said:

"The prosecutor must act in good faith. Thus if he knew he had no real case, but was pressing the extradition request for some collateral motive and accordingly tailored the choice of documents accompanying the request, there might be a good submission of abuse of process. …"

61. In R (Government of the United States of America) v Bow Street Magistrates Court [2006] EWHC 2256 (Admin), [2007] I WLR 1157 (commonly referred to as "Tollman") at paragraph 82 the Court applied to extradition proceedings the statement made by Bingham LJ in R v Liverpool Stipendiary Magistrate ex parte Ellison [1990] RTR 220 in relation to conventional criminal proceedings:

"If any criminal court at any time has cause to suspect that a prosecutor may be manipulating or using the procedures of the court in order to oppress or unfairly to prejudice a defendant before the court, I have no doubt that it is the duty of the court to inquire into the situation and ensure that its procedure is not being abused. Usually no doubt such inquiry will be prompted by a complaint on the part of the defendant. But the duty of the court in my view exists even in the absence of a complaint." 

Again, although the citation derived from a purely domestic case, the focus was on the role of the prosecutor.

62. In Symeou v Public Prosecutor's Office at the Court of Appeals, Patras, Greece [2009] EWIHC 897 (Admin) [2009] 1 WLR 2384 Ouseley J, giving the judgment of the Court, said in a passage that bears extensive repetition:

"33 … The focus of this implied jurisdiction is the abuse of the requested state's duty to extradite those who are properly requested, and who are unable to raise any of the statutory bars to extradition. The residual abuse jurisdiction identified in R (Bermingham) v Director of the Serious Fraud Office … and the Tollman case … concerns abuse of the extradition process by the prosecuting authority. We emphasise those latter two words. That is the language of those cases. It is the good faith of the requesting authorities which is at issue because it is their request coupled with their perverted intent and purpose which constitutes the abuse. If the authorities of the requesting state seek the extradition of someone for a collateral purpose, or when they know that the trial cannot succeed, they abuse the extradition processes of the requested state.

34 The abuse jurisdiction of the requested state does not extend to considering misconduct or bad faith by the police of the requesting state in the investigation of the case or the preparation of evidence for trial.

35 The reason for the distinction lies in the respective functions of the courts of the requested and requesting state in the European arrest warrant framework. The former are entitled to ensure that their duties and the functions under the Extradition Act 2003 Part I are not being abused. It is the exclusive function of the latter to try the issues relevant to the guilt or otherwise of the individual. This necessarily includes deciding what evidence is admissible, and what weight should be given to particular pieces of evidence having regard to the way in which an investigation was carried out. It is for the trial court in the requesting state to find the facts about how statements were obtained, which may go to admissibility or weight, both of which are matters for the court conducting the trial. It is the function of that court to decide whether evidence was improperly obtained and if so what the consequences for the trial are. It is for the trial court to decide whether its own procedures have been breached.

36 As those issues are for decision by the trial court in the requesting state, it cannot be an abuse of the extradition process of the requested state for such an issue to be shown to exist and for its resolution to be available only in the courts of the requesting state. The courts of the requested state cannot decide, let alone do so on partial and incomplete evidence, what it is for courts of the requesting state within the European arrest warrant framework … to decide about such issues and with what effect on the trial."

63. To similar effect, in Zakrzewski v Regional Court in Lodz, Poland [2013] UKSC 2, [2013] 1 WLR 324 at paragraph 11, Lord Sumption (with whom the other Justices of the Supreme Court agreed) considered the safeguards that existed to prevent an unjustified extradition if the prescribed particulars in a warrant were incorrect. The first does not concern us in the present case. Lord Sumption continued:

"The second safeguard lies in the inherent right of an English court, as the executing court, to ensure that its process is not abused. One form of abuse of process is the fortunately rare case in which the prosecutor has manipulated the process of the executing court for a collateral and improper purpose: …"

64. In Belbin v Regional Court of Lille, France [2015[ EWHC 149 (Admin) Aikens LJ, giving the judgment of the court, said at prargraphs 43-44:

"43. It is clear from statements of this court in R (Bermingham) v Director of the Serious Fraud Office … ('Bermingham'), R (Government of the USA) v Bow Street Magistrates' Court … ('Tollman') and Symeou v Public Prosecutor's Office at the Court of Appeals, Patras, Greece … ('Symeou') that both the Magistrates' Court and the High Court on appeal retain an implied jurisdiction to refuse to extradite a requested person under Part 1 of the EA on the basis that there has been an abuse of the process of requesting extradition by the prosecuting authority or other emanation of the judicial authority seeking extradition. In Tollman (which involved extradition proceedings under both Parts 1 and 2 of the EA) and in Symeou (Part 1 extradition) the court emphasised that the abuse of the process has to be that of the prosecuting authority. But, given that, under the Framework Decision of 2002 on which Part 1 of the EA is based, all extradition requests must be made by a Judicial Authority, it seems to us that the court has an implied jurisdiction to consider whether there has been an abuse of the extradition process under Part 1 of the EA by a requesting judicial authority. We note, of course, the point made by Sir John Thomas, then President of the Queen's Bench Division, at [49]-[50] of Swedish Prosecution Authority v Assange [2011] EWHC 2849 (Admin) that the acts of a prosecutor, in contradistinction to those of a judge, must be subjected to "rigorous scrutiny" because a prosecutor is (unlike a judge) a party to the criminal proceedings in the requesting state. That "rigorous scrutiny" must be applied when considering whether a prosecuting authority, acting as a Judicial Authority for the purposes of the extradition request, has conducted itself in a way that is an abuse of the extradition process. It is important to note that the abuse of process jurisdiction does not extend to considering misconduct or bad faith by the police of the requested state in the investigation of the case nor in the preparation of evidence for the trial in the requesting state: see [34] of Symeou.

44. However, whether it is the prosecuting authority's behaviour or that of another entity that constitutes the Judicial Authority of the requesting state that is being criticised, it will only amount to an abuse of the extradition process if the statutory regime in the EA is being 'usurped' (see [97] of Bermingham). It would, for example, be 'usurped' by bad faith on the part of the Judicial Authority in the extradition proceedings or a deliberate manipulation of the extradition process. But any issues relating to the internal procedure of the requesting state are outside the implied abuse of process jurisdiction concerning extradition proceedings: see [36] of Symeou. Moreover, as is clear from the decision of this court in Federal Public Prosecutor, Brussels, Belgium v Bartlett [2012] EWHC 2480 (Admin), this 'usurpation' of the statutory extradition regime has to result in the extradition being 'unfair' and 'unjust' to the requested person. In this regard it has also to be shown that, as a result of the 'usurpation' of the statutory regime, the requested person will be unfairly prejudiced in his subsequent challenge to extradition in this country or unfairly prejudiced in the proceedings in the requesting country if surrendered there."

And at prargraph 59:

"Secondly, the court will only exercise the jurisdiction if it is satisfied, on cogent evidence, that the Judicial Authority concerned has acted in such a way as to "usurp" the statutory regime of the EA or its integrity has been impugned. We say "cogent evidence" because, in the context of the European Arrest Warrant, the UK courts will start from the premise, as set out in the Framework Decision of 2002, that there must be mutual trust between Judicial Authorities, although we accept that when the emanation of the Judicial Authority concerned is a prosecuting authority, the UK court is entitled to examine its actions with "rigorous scrutiny."

65. The only case of which we are aware that supports the submission that abusive behaviour by a third party may form the basis for an abuse of process argument is Ristea v Prosecutor's Office, Court of Bologna, Italy [2018] EWHC 1876 (Admin). On its facts it was a case about double jeopardy: it was said that the requested person had already served the sentence in respect of which Italy wished to extradite him because it had been merged with another sentence in Romania and served there. Accordingly extradition would expose the requested person to double jeopardy. In the alternative it was submitted that the extradition of the requested person in these circumstances would be an abuse of process because, from the requested person's perspective (and that of the court in Romania), he had already served his sentence. The Respondent accepted that if the Judge were to find (as she duly did) that the requested person had served his sentence, it would be an abuse of the process for him to be surrendered so as to serve that sentence for a second time. This concession appears to have been based upon Newman v Poland [2012] EWHC 2931 (Admin), where Pitchford J stated at prargraph 19 that "it would be an abuse of the process of this court and the court below to continue to seek the extradition of a person who has, in effect, served his custodial sentence in full." The judgment in Ristea does not indicate that the finding of abuse of process that the Judge made was reached on any wider basis than that. Since the double jeopardy rule provided a bar to extradition, we question whether the abuse of process doctrine could have been applicable in any event.
66. The Appellant made extensive reference to purely domestic cases in support of its argument. The existence of an abuse of process doctrine in purely domestic cases is not in doubt; but that doctrine cannot simply be transposed into the sphere of extradition proceedings, nor do the authorities considering the domestic doctrine provide helpful analogies when considering the scope of the abuse of process doctrine applicable to extradition proceedings, for two main reasons. First, the domestic doctrine is concerned with the control of purely domestic proceedings and focuses on the need to ensure that a trial is fair, whereas most issues of abuse that arise in extradition proceedings relate to the integrity of the extradition system as such. Second, there is no equivalent in the purely domestic cases to the mutual trust that underpins the extradition process. Specifically, the English court had repeatedly said that the internal procedure of the requesting state's criminal system are outside the implied abuse of process jurisdiction concerning extradition proceedings: see Symeou at prargraph 36, Belbin at prargraph 44. The only exception to this may be found in cases such as R v Horseferry Road Magistrates' Court, ex parte Bennett [1994] 1 AC 42 where extradition was secured by abusive conduct by the police and prosecuting authority in this jurisdiction in such an egregious fashion that all proper legal processes were sidestepped. The explanation for these cases, and for Lord Lowry's broad statement of principle on which particular reliance was placed by the Appellant, is that but for the misconduct of the executive the defendant would not have been tried in this country at all. This exceptional principle cannot avail Mr A.
67. With these principles in mind, we turn to consider Ground 2 as advanced by the Appellant.
68. For the purposes of the abuse argument, we accept the factual position most favourable to Mr A, namely (a) he was acting as a participating informer for the Garda at the time of his arrest, (b) he had been assured by the Garda that they would protect him from arrest or prosecution both in Ireland and abroad, (c) the Garda failed to keep their assurance, (d) the Garda failed to provide a full and accurate account of the Mr A's role and their involvement (including their assurance that the would protect him from arrest or prosecution in Ireland and abroad), (e) the Garda's failures were in breach of a duty of care that they owed to Mr A, (f) had either the prosecutor or the court been fully informed about (a) and (b) above, the outcome could have been different because the prosecutor could under French law have decided not to prosecute or the French courts could have taken a different and more lenient view in relation to either conviction or sentence (though they would have been under no obligation to do so), and (g) there is now no realistic prospect of overturning Mr A's conviction or sentence.
69. We have come to the conclusion that the doctrine of abuse of process in extradition proceedings has no applicability to the facts as we have outlined them earlier in this judgment even approaching those facts on the basis most favourable to Mr A. There are two reasons for this conclusion. First, the doctrine applies to abuse by the "prosecuting authority or other emanation of the judicial authority seeking extradition": see Belbin at [43]. Save in the context of cases such as Bennett, it does not even apply to conduct of the police of the requesting state. A fortiori, it cannot apply in respect of conduct of the police of a third party state. In the present case, there is no question of "usurpation" or "manipulation" of the process by the Respondent and, as we have said, there is no criticism that is or could be made of the Respondent prosecutor's conduct. On the contrary, the requesting authority is acting on the basis of a proper conviction after all due process, including Mr A's successive appeals up to the Cour de Cassation. Mr A's real complaint is not and cannot be directed at the integrity of the extradition process or the good faith of the requesting authority. His complaint is squarely directed at the conduct of the Garda; and his complaint is vindicated by the terms of the settlement agreement which acknowledge that the Garda acted in breach of the duty of care that it owed him. Although we do not regard this as determinative, it is a feature of the factual background to the present case that he has secured substantial compensation from the Garda for that breach – compensation which evidently relates to the consequences of his arrest including being subjected to these extradition proceedings and the risk of extradition. Though it is clear that Mr A would still wish to avoid extradition to France, that compensation goes some way to tempering what he sees as the unfairness of his predicament that the Garda's breach of duty (on his case) has caused.
70. We consider that it is established on the authorities that, had it been the French police who had acted in the way alleged against the Garda, that would not have brought the doctrine of abuse of process into play. The reason for that is that the doctrine is directed solely and exclusively to the conduct of the requesting authority itself. Here, as we have said, the prosecuting authorities have acted properly throughout the extradition process. If similar behaviour by the French police would not have triggered the doctrine, we can see no grounds for reaching a different conclusion in relation to the behaviour of the Garda.
71. Second, Mr A's submission that the end result is manifestly unfair is a direct challenge to the internal procedure of the French criminal process. Although we have recorded the expert evidence and have accepted, for the purposes of argument, that the outcome might have been different if the Garda had provided full information to the French prosecuting authorities and the French courts, we regard speculation about what would or should have been the outcome in different circumstances as inadmissible and irrelevant to the application of the doctrine of abuse of process in extradition proceedings. This is because of the mutual trust that informs the working of the extradition scheme. Similarly, there being no realistic prospect of overturning Mr A's conviction or sentence is itself a consequence of the internal procedure of the French criminal process. While it is easy to assert that this state of affairs is "unfair", it may be remembered that Mr A had, and took advantage of, the provisions of French criminal process that enabled him to overturn his original conviction and then, after re-conviction on a second trial (where he was represented but chose not to attend in person) to appeal successively to the Court of Appeal in Lyon and thence to the Cour de Cassation.
72. There is a further pragmatic reason why we would be disinclined to stigmatise the position as "unfair" even if it were open to this Court to investigate the internal workings of the French system. We have recorded the experts' view that the outcome might have been different if the prosecutor or the French court had been fully informed. However, the perils of going down this route are shown by the judgments of the French courts, cited at paragraphs 21-22 above. They were premised on Mr A not having had legal immunity in any circumstances and the seriousness of the offences, and were expressly said to take into account "whatever is true of the agreements made [by Mr A and the Irish police]." This approach, first by the Court of Appeal in Lyon and affirmed by the Cour de Cassation is the most informative and reliable indication of what the attitude of the French internal system would have been. It is not for this Court to opine on what it considers to be any rights or wrongs of the French approach or to assert that the outcome should (as opposed to could) have been different if Mr A's account had been accepted.
73. We return to the summary of principle at paragraph 59 of Belbin. In the present case there is no evidence (cogent or otherwise) that the Respondent has acted in such a way as to usurp the statutory regime of the Extradition Act or to impugn its integrity. It would in our judgment be an unwarranted and unprincipled extension of the doctrine of abuse to apply it in circumstances where a perceived unfairness has arisen in such circumstances. We therefore dismiss the appeal on Ground 2.

Ground 3: arbitrary and disproportionate sentence contrary to Articles 5 and 3 of the Convention

74. Article 5, so far as relevant, provides:

"1.  Everyone has the right to liberty and security of person. No one shall be deprived of his liberty save in the following cases and in accordance with a procedure prescribed by law:

(a) the lawful detention of a person after conviction by a competent court;

…

2. …

3. …

4. Everyone who is deprived of his liberty by arrest or detention shall be entitled to take proceedings by which the lawfulness of his detention shall be decided speedily by a court and his release ordered if the detention is not lawful."

75. The Appellant accepts that detention following conviction is lawful under Article 5(1)(a) where it is sufficiently causally connected to the conviction and the objectives of the sentence and that in principle matters of appropriate sentencing fall outside the scope of the Convention. However, the Appellant points to cases where changing circumstances may render continued detention unlawful in the absence of further judicial supervision, e.g Stafford v United Kingdom (2002) 35 EHRR 32, Kafkaris v Cyprus (2011) 53 EHRR SE 14, Todorov v Bulgaria (App 71545, ECtHR, 19 January 2017). In the absence of ECHR authority specifically in point, the Appellant submits that the settlement of the Irish proceedings and the terms of settlement mean, in the absence of a mechanism for review of the sentence imposed by the French courts, that his sentence is or may now be arbitrary or disproportionate.
76. Article 5 does not automatically mandate a continuing right of recourse to the court. This is made clear by De Wilde, Ooms and Versyp v Belgium No 1, (1971) 1 EHRR 373, a case that concerned the inability to challenge a magistrate's administrative decision to detain vagrants. At paragraph 76 the ECtHR said:

"… it is clear that the purpose of Article 5 (4) is to assure to persons who are arrested and detained the right to a judicial supervision of the lawfulness of the measure to which they are thereby subjected: … . Where the decision depriving a person of his liberty is one taken by an administrative body, there is no doubt that Article 5 (4) obliges the Contracting States to make available to the person detained a right of recourse to a court; but there is nothing to indicate that the same applies when the decision is made by a court at the close of judicial proceedings. In the latter case, the supervision required by Article 5 (4) is incorporated in the decision; this is so, for example, where a sentence of imprisonment is pronounced after "conviction by a competent court" (Art. 5 (I) (a) of the Convention)."

77. The Respondent submits that the cases cited by the Appellant are distinguishable on their facts. Stafford v United Kingdom concerned the lawfulness of detention following the completion of the tariff in relation to a life sentence. Kafkaris v Cyprus concerned a review of a whole life term. Todorov v Bulgaria concerned the execution of a sentence that had become statute barred. In each case it could be said that the necessary causal connection between the conviction and the objectives of the sentence had been broken. That, submits the Respondent, is not the case here, quite apart from the existence of a judicial remedy in appropriate cases. We agree.
78. Here, in our judgment, the sentence is neither arbitrary nor disproportionate to the gravity of the crime, having been imposed by a competent court of law after all due process including the exercised right of successive appeals. Not only has the Appellant had the advantage of the conventional appeals process under French law, he has available the potential remedy of a "demande en revision" even if the prospects of success in his case are slim. It cannot therefore be said that there is no remedy available under French law. The existence of the "demande en revision" process creates something of a dilemma. If the fact and terms of settlement are sufficient to lead to his conviction being overturned, Mr A's sentence will fall. If they are not, there is no reason to think that the reasons justifying the original sentence cease to be appropriate or that there ceases to be a sufficient causal connection between the conviction and the stated objectives of the sentence. We therefore reject the submission, which underpins this ground, that there is a real risk that he will serve a sentence that is arbitrary and disproportionate.
79. We therefore dismiss the appeal on Ground 3.

Ground 4: Proportionality and Articles 5 and 13 ECHR

80. The Appellant accepts that he cannot succeed on this ground if, as we have held, he fails on Ground 3. That is because he accepts that, unless it is arguable that the sentence imposed by the French court is arbitrary, there is no basis for attacking the proportionality of the EAW. That concession is correctly made and determines the outcome of this Ground of Appeal.
81. Accordingly, we dismiss the appeal on Ground 4.

Ground 1: Article 3 and prison conditions

82. Both Appellants contend (by Ground 1 in the case of Mr A, and by the sole Ground in the case of Mr Esmaili) that the District Judges ought to have reached different decisions on the issue whether their extradition would amount to a breach of their rights under Article 3 of the Convention in view of prison conditions in France. Given that both parties to the appeal are relying in part on fresh evidence, the reasoning and conclusions of the District Judges in both cases need not be closely examined. Ultimately, we will be deciding the Article 3 issue for ourselves, also acknowledging that these appeals are being treated as test cases. Choudhary v France [2020] EWHC 1966 (Admin) enjoyed that status last year, with other French prison appeals being stayed behind it; but the appeal was conceded by the IJA on a separate issue before the Article 3 question could be addressed.
83. Although the Article 3 ground is common to the two appeals, there are some factual differences which we should explore. Mr Esmaili has 14 months and 27 days remaining to be served under a judgment of the Court of Appeal of Rouen dated 20^th March 2017 imposed for facilitating immigration offences. Following an initial stay at Villepinte or Fresnes maisons d'arrêts ("MAs"), he would be transferred to Rennes-Vezin to serve the remainder of his sentence. Mr A's initial incarceration would also be at Villepinte or Fresnes MAs, although thereafter he would be transferred to Lyon-Corbas or Villefranche-sur-Saône. In Mr A's case, the CPS sought further information about the location of custody and how the French authorities proposed to accommodate the medical conditions from which he suffers. Somewhat laconically, the IJA stated that Mr A's health conditions would be taken into account and he would be sent to a "prison specially equipped". No further details were provided.

The Governing Law

84. The parties were in general agreement as to the law applicable to these appeals, and we have noted the various summaries provided in similar cases decided in this jurisdiction in the past. In those circumstances, it is unnecessary for us to reinvent the wheel, but we will seek to highlight key areas requiring particular emphasis.
85. In Shumba v France [2018] EWHC 1762 (Admin) ("Shumba No 1"), this Court (Singh LJ and Carr J, as she was then) summarised the principles in an Article 3 prisons case as follows:

"34. Article 3 can in principle apply where a Contracting State proposes to extradite a person to another state, whether or not that other state is itself a party to the ECHR. As it happens France is, like the United Kingdom, a party to the ECHR.

35. There must be substantial grounds for believing that, if extradited, the Appellant faces a real risk of being subjected to inhuman or degrading treatment.

36. Once such evidence has been adduced by the Appellant it is for the requesting state to dispel any doubts about it: see Saadi v Italy [2009] 49 EHRR 30, at prargraphs. 129 and 140.

37. There is a presumption that parties to the ECHR, such as France, are willing and able to fulfil their obligations, in the absence of "clear, cogent and compelling" evidence to the contrary. However, that presumption can be rebutted where that evidence comes from an internationally recognised source or is specific to an individual.

38. There may also be a duty on the Court in this jurisdiction to request further information from the state concerned where this is necessary to dispel any doubts.

39. In the context of prison overcrowding, there will be a strong presumption of a breach of Article 3 if any of the following criteria are absent: (1) a private sleeping place within a prison cell; (2) at least 3m² of floorspace per prisoner; and (3) an overall surface area of the cell which is such as to allow the detainees to move freely between the furniture items.

40. Where a detainee is allocated between 3 and 4m² of personal space, a violation of Article 3 will be found if there are other aspects of inappropriate physical conditions: in particular, regard will be had to access to outdoor exercise; natural light or air; availability of ventilation; adequacy of room temperature; access to private toilet facilities; and compliance with basic sanitary and hygiene requirements."

86. In the light of the parties' submissions on the law, the following specific matters arise.
87. First, the reference to "an internationally recognised source" has been variously interpreted. In Krolik v Poland [2012] EWHC 2357, at paragraphs 4, 6 and 7, this Court referred to the need for "clear, cogent and compelling evidence" and "something approaching an international consensus", namely "a significant volume of reports from the Council of Europe, the UNHCR and NGOs". In Criminal Proceedings against Aranyosi [2016] QB 921, the ECtHR expanded this category to include its own judgments and judgments of courts of the issuing state. Further, in JMB v France [2020] ECHR 91, promulgated on 30^th January 2020), the ECtHR appeared to draw inferences from France's failure to rebut what was described as the "credible and reasonably detailed" information provided by the various applicants in that case. Ms Malcolm QC for the Respondent pointed out that JMB was not a "pilot judgment"; Mr Keith QC for the Appellants characterised it as a "quasi-pilot judgment". In our view, this terminological debate does not really assist (c.f. the decision of this Court in Brazuks v Latvia [2014] EWHC 1021 (Admin)). JMB is undoubtedly an important decision from these Appellants' perspective because of the obvious saliency of the evidence (or perhaps the lack of it) under scrutiny, and we will return to it below.
88. Secondly, at the first stage of the analysis the real question is whether the Appellants, on whom the burden resides, have adduced objective, reliable, specific and properly updated evidence that indicates a real risk of a violation of Article 3: see Aranyosi. If, at the second stage, we were to assess that there are substantial grounds to believe that the Appellant concerned would be exposed to that risk, it would be incumbent on us in the first instance to request supplementary information of the IJA under Article 15(2) of the Framework Decision. Mr Keith did not submit that we should determine the Article 3 issue in his clients' favour, at least at this juncture. That would arise at Stage 3 after the supplementary information had been furnished.
89. Thirdly, we must acknowledge that the maximum period that Mr A and Mr Esmaili would be held in either Villipente or Fresnes MAs would be 10 days. M. Dominique Tricaud in affidavit evidence dated 12^th November 2019 stated that the period would be as short as "a couple of days", although in his oral evidence he said – without giving chapter and verse – that the average period was in fact 10 days. In Shumba No. 2 the supplementary information from the IJA was that the maximum period would be 6 days. In our view, we could not reasonably go beyond 10 days, and the better view must be that the period is in fact 6 days. In Mursic v Croatia [2017] 65 EHRR 1, the Grand Chamber of the ECtHR stated that the strong presumption of a violation of Article 3 in a situation where a detainee's personal space falls below 3m² will normally be capable of being rebutted if three factors are cumulatively met: (1) the reductions in the required minimum space are "short, occasional and minor"; (2) there is sufficient freedom of movement outside the cell; and (3) there are no other exacerbating factors. In Szalai and Zabolotnyi v Hungary [2019] EWHC 934 (Admin), this Court concluded that a period of one to two weeks could fall within (1) above. Given that, as we will explain further below, items (2) and (3) have not been put seriously in issue in relation to Villipente and Fresnes MAs, we accept Ms Malcolm's submission that even if Mr Keith were correct in his contention that the current overcrowding is such that we should conclude in the light of that and all the available evidence that the strong presumption is engaged, these Appellants would not in fact be held in these institutions for sufficient time to raise an issue under Article 3.
90. Although Mr Keith had the forensic point that in Shumba No. 2 the Article 3 claim did not fail for reasons of brevity of duration, we understand that the point was not argued. Be that as it may, and in the light of the abiding expectation that this judgment should be providing general guidance on Article 3 in relation to a number of French prisons, we will examine the state of the evidence bearing on Villipente and Fresnes MAs.

Key Jurisprudence on French Prison Conditions

91. As we have said, Mr Keith placed particular reliance on the decision of the ECtHR in JMB. In order to provide greater context, we can borrow freely from the summary that this Court (Popplewell LJ and Jay J) provided in Choudhary, handed down in July 2020, expanding it to reflect additional evidence to which we have been referred.
92. In February 2018 the French prisons inspector, the CGLPL, reported on conditions in a number of establishments including Fresnes MA. Its final recommendations of general application were as follows:

"Recommendation 1

The right to individual imprisonment must be effective for all detained persons. This right implies that the cells of a place, because of their area less than 11 m², are occupied by only one person. People who express the wish to be or who, in fact, are in a group must be in a room suitable in terms of living space and equipment. An action plan to reduce the use of extra mattresses must be implemented without delay, having regard to the unacceptable worsening of the conditions of detention which results for people and the consequences which seriously jeopardize their prospects for reintegration.

Recommendation 2

The calculation of places and capacity of penal establishments must be reviewed and updated in a standard of a regulatory nature. This standard must take into account the recommendations of Council of Europe bodies. In addition, no other data than operational capacity should be taken into account when calculating the occupancy rate of an establishment. In addition to the number of people detained, the number of operational places and the occupancy rate per establishment, it is necessary for the prison administration to acquire more precise statistical tools for measuring prison overcrowding and individual cell occupancy. The rate of individual cell occupancy and the number of additional mattresses must be produced each day by establishment, with regard to the specific characteristics of each of these, in particular the number and type of cells (individual, double or multiple). The notion of density should be further developed in remand centres, in order to know the area allocated to each person detained and to measure overcrowding. The monthly statistics should show, by establishment, the number of vacant places and calculate the difference between the operational capacity, reduced vacant places, and the number of persons detained.

Recommendation 3

The implementation of a policy to reduce the prison population cannot be seriously envisaged for lack of precise knowledge of the state of overcrowding and the execution of sentences. The management of the prison administration must once again be able to produce, via the GENESIS software, statistics relating to the composition of the penal population of each establishment.

Recommendation 4

The lack of staff and the resulting "degraded mode" management have detrimental effects on the conditions of detention which prison overcrowding aggravates, when it is not one of the causes. Failing to fill the posts provided for in the staff organizational charts within the establishments, the prison administration must define criteria for the abolition of posts and prohibit some, in particular those having the consequence of reducing access to visiting rooms, to healthcare, medical and other activities.

Recommendation 5

Overcrowding must cease to be understood as an essentially prison problem. The fight against prison overcrowding must become real public policy, to which clear and lasting resources must be allocated.

Recommendation 6

Judges who pronounce prison sentences must be attentive to the conditions of detention in remand prisons under their jurisdiction and ensure that this incarceration can make sense. It is the responsibility of the magistrates to know the places of detention and the context specific to the establishments under their jurisdiction. To do this, they must in particular effectively monitor places of detention and rely on sentence enforcement commissions to put in place real policies to combat overcrowding, by intensifying the exchange of information on data available and by creating suitable management tools.

Recommendation 7

It is time to take the necessary steps to end the excessive use of the prison sentence; to readjust the perimeter of the prison sentence in application of the principle of the need for punishment, in particular by replacing prison terms for certain offenses by other penalties, as well as by decriminalizing measures.

Recommendation 8

The public authorities must question the effect of short prison sentences which most often have the effect of causing real disruption in the life of a convicted person without being able to benefit from any assistance in prison because of the brevity of the stay.

Recommendation 9

The way in which our criminal jurisdictions operate and the whole process of the execution and enforcement of sentences needs to be thought through, in conjunction with the goal of prison deflation. Quantified targets must be set and subject to reinforced monitoring.

Recommendation 10

A national prison regulation mechanism must be set up by legislation and be accompanied by binding local protocols, associating the various actors under the responsibility of the judicial authorities. Its purpose is to prevent any establishment from exceeding an occupancy rate of 100%."

93. As at 1 January 2019, which was the end-point of the ECtHR's consideration in JMB, overall prison density in France was 116.5% and the figure for MA establishments was 140%. The evidence before us demonstrates that in October 2019 and January 2020 the matching figures were 116% and 138%; in April 2021 they were 107.2% and 125.4% respectively. Further, it was not until December 2020 that Recommendation 3 of the CGLPL and the recommendation of the ECtHR itself was implemented in the form of an action plan.
94. The CGLPL report dated February 2018 also covered Fresnes MA following a visit in October 2016. Her report contains four different statements on the issue of cell sizes:

"1. From one division to another, the cells are identical to a few details: 3.94m deep by 2.46m wide and 2.99m high, i.e. an area of 9.69m² and a volume of 28.93m²"

"2 – the current situation, with three persons in a cell of 9.8 m², constitutes an assault on the dignity and is not acceptable"

"3 – The cells are in principle individual, approximately 10 m². In these cells, once the size of the beds (three bunk beds), toilets and table has been deducted, three people must then live in a space of barely 6 m²"

"4. These cells are only 10 m², once you deduct the bed space, the toilets and the table…."

95. According to the CGLPL's report, only 13% of the prison population benefited from an individual cell; 31% had to share a cell with one other remand prisoner; 56% with two others. All the above dimensions, which vary slightly in the report although its author also states that the cells are "à peu près identiques", take account of the toilet facilities. As for the cell furniture, the CGLPL report supported the opinion of the European Committee for the Prevention of Torture that the floor space it covered should be deducted, although the ECtHR did not go that far.
96. Further, there was evidence in JMB in the form of a letter from the Parlement Français dated January 2018. At that point, the overcrowding rate in Fresnes prison was 203%; there were wet and dirty cells, as well as hygiene problems. Inmates had only three showers a week and bedbugs proliferated.
97. The CGLPL's report of February 2018 was considered by this Court in Shumba 1: see paragraphs 56-63 of its judgment. It was only one piece of the evidential jigsaw. Given the uncertainty surrounding the floor dimensions, this Court decided to ask a number of questions of the relevant French authorities. In Shumba 2 [2018] EWHC 3130 (Admin), handed down on 16 November 2018, the responses of the French Ministry of Justice were closely analysed. The position in relation to Fresnes prison was as follows:

"… smaller single cells (8 to 9m²) (excluding sanitary facilities) have space for one or two detainees, leaving a minimum of 4m² per person. As of 23 July 2018, the 386 larger single cells (10m²) accommodated three detainees, i.e. 3m² per person (excluding sanitary facilities)."

As for other "non-spatial" factors:

"In Fresnes permanent eradication of bed bugs is a priority action, with regular 'disinsectisation' operations two or three times a week and whenever a specific report is made …"

98. At paragraphs 13-22 of its judgment in Shumba 2, this Court concluded that a violation of Article 3 had not been demonstrated. It was held that the further, up-to-date information from the IJA showed that each Appellant would have at least 3m² of floorspace despite the occupancy rate in October 2018 being 192%.
99. As we have said, in JMB the ECtHR was considering evidence in relation to a number of French prisons including Fresnes MA up to 1 January 2019. The occupancy rate at Fresnes MA was 214% as at 18 April 2017 and 195.6% as at 1 January 2019. The applicants RM and AT each claimed to be sharing, or to have shared, a cell of 9m² including furniture and toilet with two other detainees. The ECtHR concluded that each had in fact been detained with two others in cells of 9.5m² before deduction of the sanitary facility, resulting in less than 3m² per detainee. In relation to applicant ABA, the ECtHR found that he had been detained in the same cell size as the others with one other detainee, and therefore had a personal space of 4m².
100. The ECtHR was clearly hampered by the French government's failure to produce documentary evidence confirming amongst other things the precise area of the cells. It accepted that the applicants' account was credible and reasonably detailed, and that the burden of proof was therefore transferred to the government. The ECtHR concluded as follows:

"259. In the cases examined, the Court noted that the Government had produced information on the end of the applicants' detention or on the date of their end of sentence. On the other hand, it notes that the accuracy of the information communicated by the Government on the applicants' personal space is limited. These are sometimes non-existent, as is the case for detainees from Faa'a-Nuutania, Baie-Mahault and Nice. For others, they are incomplete because they do not always specify the area of the cells and do not indicate whether the sanitary annexes are included in these areas. Finally, the information is not always supported by a written document such as a co-ownership history. The Court noted these evidentiary shortcomings in the applications concerning the prisons of Ducos and Fresnes.

In addition, the Court was unable to know precisely the area of the sanitary part of the cells, with the exception of those of the MA of Nîmes, which made it difficult to calculate the applicants' personal space when it had information on the total area of the cell. It then assumed that such a space was between 1 and 2m².

…

260. In these circumstances, and while admitting the overcrowding situation in all the prisons concerned, the Court considers that the Government have not convincingly refuted the allegations made by the applicants of the CP of Ducos, Faa'a-Nuutania, Baie-Malhaut, Nice and Fresnes (as regards RM and AT for this last establishment) according to which they would have had less than 3 m² of personal space during their entire detention (paragraphs 29, 49, 59, 92 and 113 above). These allegations are further corroborated by relevant information from national authorities such as the CGLPL or from international bodies such as the CPT.

…

299. Having regard to what it said in paragraph 260 above, and as regards the detention of RM and AT, the Court concludes that there is a strong presumption of violation of Article 3 of the Convention. This presumption cannot be called into question in the absence, in this case, of the first of the three cumulative factors challenging this rebuttal, namely periods of "short, occasional and minor" reduction in the applicants' personal space compared to at the minimum required. It follows that there is no need to examine the other factors (mutatis mutandis, Nikitin and others, cited above, § 184).

300. As regards ABA, the Court held that he had a personal space of approximately 4 m² (see paragraph 113 above). In their observations, the Government indicate that the personal space of ABA is not less than 3 m² (see paragraph 250 above). In these circumstances, the Court considers that it must be held that the applicant had personal space of between 3 and 4 m² throughout his detention.

The Government also indicate that the recreation and the possibility of playing sports are sufficient to consider that the threshold of gravity required by Article 3 of the Convention has not been reached. In view of the findings of the judge hearing the application for interim measures, the CGLPL and the CPT, which observe and describe the very degraded conditions of detention within the MA of Fresnes (see paragraphs 6, 108, 151 and 152 above), the Court does not share this point of view. It notes that it appears from their decisions and reports that the MA of Fresnes, obsolete because of its age and lack of renovation, is repeatedly confronted with the presence of pests, and in particular bedbugs in the beds of detainees, and that the latter suffer from the lack of light and humidity in the cells (idem). It also notes that if the length of the exercise walks in the prison courtyards is not disputed by the applicants, it is the condition of these places which is in question: in its emergency recommendations published in December 2016, the CGLPL indicated that the areas were cramped (twenty-five people in 45 m²) and lacked shelters and toilets and that the rats were moving around there en masse (see paragraph 106 above). The Court does have no information on the current state of these facilities but the description made by the applicants detained in Fresnes at the time of the introduction of their request in 2017 corresponds to that which was made by the CGLPL in 2016 (paragraph 106 above) and the finding of the domestic judge in 2018 who considered that the conditions under which the exercise walks take place are detrimental to the dignity of detained persons (see paragraph 109 above).

In view of the above, the Court considers that the conditions of detention of A.B.A. amounted to degrading treatment within the meaning of Article 3 of the Convention."

101. In the light of JMB, this Court in Choudhary sought supplementary information from the IJA, relating in particular to Fresnes MA, which was provided in August 2020. The Respondent seeks to rely on inter alia this information in defence of the present appeals, a course that is not opposed by Mr Keith. This material has not yet been subject to judicial scrutiny and we will address it below.
102. Villipente MA was also considered by this Court in Shumba. As at March 2017 the occupancy rate was 201% and there was evidence of inmates enduring less than 3m² personal space. The IJA provided supplementary information which was considered by this Court in Shumba No. 2. By the autumn of 2018, the occupancy rate was down to 184%. Prisoners were able to move freely between furniture in the cells. They were on average confined to their cells for 15-16 hours a day, and were offered a minimum of 5 hours' activity a day. As for the cell-space:

"there are 480 single cells measuring 9 to 10m² including sanitary facilities, meaning that there is 8 to 9m² of space, i.e. a minimum of 4m² for the one to two detainees. The double cells measure from 11m² to 12m² including sanitary facilities i.e. 5m² per person. The 40 cells (as of 27 August 2018) with three detainees have at least 3m² per prisoner. Less than 10% of Villepinte's prisoners have less than 4m² floorspace, the other 90% have more than 4m²"

The Action Plan

103. The Covid-19 pandemic has seen a reduction in the prison population from 71,679 on 16 March 2020 to 58,720 on 11 May 2020. However, by 19 November 2020 it had risen to 62,502. Apart from the building programme which we address below, there have been significant changes in the overall penal regime to reduce the number of people serving prison sentences.

Further Evidence Bearing on Fresnes MA

104. Fresnes prison, including the MA, was completed in 1898.
105. In a petition by M. Bou Magassa and eight others to the ECtHR published on 14 December 2020, various complaints were made about inter alia Fresnes MA, although these have not yet been subject to formal judicial consideration. These complaints included the following: two or three inmates occupying cells of 9m²; a general lack of hygiene; cell toilets not fully partitioned; lack of ventilation and of light; no heating; cockroaches, bedbugs and rats; poor food; inmates held 22 hours a day in their cells; etc.
106. The cell occupancy rates were 166.6% as at October 2019; 163.6% as at January 2020; 130.3% as at 1 April 2021; and 127.4% as at 1 May 2021.
107. The further information supplied by the IJA in Choudhary in August 2020, which on 14 June 2021 the French authorities said remained valid, stated the following. It was said that the duration of night confinement is limited to 12 hours, and that "legal and regulatory provisions" allow inmates regularly to be released from their cells for the purpose of exercise, training and working in concession workshops. In terms of cell occupancy:

"The Penitentiary Centre of Fresnes (men and women …) includes 1,391 cells:

- 1,376 cells of 9 to 10 m²

- 3 cells of 10 to 11 m²

- 6 cells of 11 to 12 m²

- 4 cells of 13 to 14 m²

- 2 cells from 19 to 24 m²

- the size of the sanitary facilities being a little less than 2 m²

- showers not in the cells

In the men's area, 1,063 cells can accommodate 3 prison inmates.

As of 29 May 2020, the prison of Fresnes allows 1,551 inmates … Apart from one cell, none accommodates more than two inmates. Six hundred and ninety three inmates are individually locked up, representing 44% of the total number of inmates …"

108. It is also said that the pressure on Fresnes MA will ease in the light of additional places that will be created in the Île-de-France region. Even so, the "first deliveries" will not be until 2022, with a second phase being implemented in 2028. The evidence before us was that 1,927 prison places in the Paris region (presumably Île-de-France) had been brought into service since May 2017. By 19 November 2020 the capacity of Fresnes MA had risen to 1,852.

Further Evidence Bearing on Villipente MA

109. Villipente MA was built in 1990 and started taking prisoners in 1991.
110. The occupancy rates for October 2019, January 2020, April 2021 and May 2021 were 184%, 177.9%, 164.8% and 159.6% respectively.
111. Our attention was drawn to the witness statement of M. Robert Bechian dated 27 August 2020. This was not considered by the District Judge. He was imprisoned at Villipente MA between 21 December 2019 and 8 April 2020. His statement refers to gross overcrowding, filthy conditions, and to being locked up in his cell for 22 hours a day. If true, and free from exaggeration, it paints a troubling picture. It appears that owing to oversight on the part of the Respondent's legal team M. Bechian's statement was provided to the IJA too late for them to comment.
112. However, we accept Ms Malcolm's submission that, despite M. Bechian's solicitor's evidence, this statement could have been obtained in time to be subject to forensic scrutiny before the District Judge. As such, it does not constitute the sort of objective and reliable evidence that may properly weigh with us.

Lyon-Corbas MA

113. Lyon-Corbas MA started taking prisoners in May 2009. Plainly, it is of modern construction.
114. The occupancy rates for October 2019, January 2020, April 2021 and May 2021 were 138.8%, 139.1%, 132.7% and 131.8% respectively.
115. Mr Keith relied on the statement of M. Dominque Tricaud dated 16 May 2018 which was considered by the District Judge. He is a French lawyer with particular knowledge of French prison conditions.
116. Apart from his evidence dealing with conditions generally, which we have taken into account, M. Tricaud drew attention to the CGLPL report dated October 2017 following a visit in December 2014. The delay in publishing these reports has not been explained. There were no mattresses on the floor at the time of the visit, but greater overcrowding since then has led to this. M. Tricaud refers to evidence from a Lyonnais lawyer to that effect. Moreover:

"… it has been observed that cells were supposed to be individual, and adding extra beds implies continuous promiscuity [sic], being 2 or 3 per cells. The surface of cells is generally around 10.5m² …"

117. M. Tricaud's evidence was also to the effect that the fabric of the building is deteriorating and that there are frequent floods; that it is difficult to obtain medical appointments; and, that there is a "climate of violence".
118. We have not ignored other aspects of the CGLPL report but this material is somewhat stale and in our view these other matters, taken both individually and cumulatively, are not of sufficient gravity to raise an issue under Article 3.
119. In an email dated 11 June 2018 from Mr Amid Khallouf of the International Observatory of Prisons:

"We don't have a lot of information about Corbas.

…

The CGLPL reports … are quite old, but the infrastructure and the furniture didn't changed [sic] since then."

120. Deputy District Judge Ikram's analysis of Mr A's Article 3 case was as follows:

"I have considered the affidavits of Dominique Tricaud. The witness states that the RP is likely to be placed at either Lyon-Corbas or Villefranche Sur Saone prisons. I note the Visit Reports at the prisons date back to 2009/2012/2014.

The RP bears the burden to satisfy a real risk that detention there would create a real risk of a breach of his Art 3 rights. The witness has not suggested that the RP would be accommodated in space less than 3m².

As regards mattresses have been added to the floor, their Lordships in Shumba rejected the suggestion that the same would breach Article 3 and stated that:

'There is no proper basis for asserting that there would be additional furniture such as to impede free movement, nor would the presence of mattresses interfere with it.'

…

I am not persuaded that the RP has discharged the burden upon him that there would be a breach of his Article 3 rights."

121. We observe at this stage that the District Judge has not correctly characterised M. Tricaud's evidence on the space issue.

Villefranche-sur-Saône MA

122. Villefranche-sur-Saône MA started taking prisoners in 1990.
123. The occupancy rates for October 2019, January 2020, April 2021 and May 2021 were 116.2%, 114.4%, 88.3% and 90.2% respectively.
124. The CGLPL report of 2016 following a visit in November 2012 does not represent up-to-date evidence. Even so, unless the position has improved materially, this prison is dilapidated and dirty, there is a "high climate of violence", and medical facilities are inadequate.
125. Mr A relies on the evidence of M. Tricaud and of Mr Khallouf. The former cannot materially add to the CGLPL report, save to point out that prisoners from Lyon-Corbas are generally allocated to Villefranche-sur-Saône MA owing to overcrowding. Mr Khallouf has provided a number of hyperlinks relating to the behaviour of the prison guards and poor visiting facilities. One of the hyperlinks references a visit by a French senator, Mme. Esther Benbassa, who made an unannounced visit to this institution on 7 December 2017. We have examined this material as best as we can (there is no translation) and it presents a worrying picture of violence and suicide. Photographs, which have been included in the bundle, show cramped conditions, unpartitioned sanitary facilities and filthy showers.

Rennes-Vezin MA

126. Rennes-Vezin MA started to accept prisoners in 2010.
127. The occupancy rates for October 2019, January 2020, April 2021 and May 2021 were 137.5%, 150.5%, 130.4% and 133.8% respectively.
128. According to the CGLPL's report dated February 2012 following a visit in December 2010, this prison comprises two separate "wards" known as "MA1" and "MA2". As for MA1, the surface area is 14m² for dual occupancy and 10.5m² for single occupancy. The sanitary facilities occupy 1.6m². As for MA2, there are 73 individual cells of 9.5m², 11 "double" cells of the same surface area, and 47 cells with two places of 13.6m². These dimensions, assuming no more than double occupancy, do not raise an issue under Article 3.
129. There was a second CGLPL visit in January 2017. The reporters noted a general deterioration in conditions with an occupancy rate of 152%, 37% of the MA1 cells and 31% of the MA2 cells being subject to triple occupancy (i.e. 174 prisoners in all), and 10% of inmates living on mattresses. Further, 41% of the MA1 cells and 47% of the MA2 cells were in double occupancy (i.e. 222 prisoners in all). At the very least, this raised an issue as to overcrowding and insufficient personal space. There was also noted to be a disturbing level of violence.
130. CGLPL's recommendation was as follows:

"Bed deprivation, the obligation to sleep on a mattress placed on the floor, the lack of available space and the ensuing promiscuity in the cell, constitute serious attacks on the dignity of persons. This situation must be ended."

131. District Judge Goozee's judgment contained no analysis of the overcrowding issue in relation to Rennes-Vezin MA.

The Rival Contentions Summarised

132. In view of this material, Mr Keith reminded us that Article 3 is a fact-sensitive jurisdiction. He made a number of overarching submissions. First, he submitted that both Appellants had fulfilled "stage 1" of the Aranyosi procedure, that the onus shifted to the IJA to provide sufficiently detailed information as to the conditions in which they would be held, that the IJA had failed to do so, and that this Court should therefore activate the Article 15(2) procedure and seek supplementary information as soon as possible. Secondly, he submitted that although some of the hard evidence in these cases could be described as somewhat stale, the inferences to be drawn from more recent material were that matters had not improved. Thirdly, he submitted that the IJA could derive nothing from the "action plan" because it did not address specific prisons and in any case the timescales were lengthy. Fourthly, he maintained that the two District Judges had failed to grapple with the evidence that was available to them, and that in any case the situation has worsened.
133. As for the specific prisons, Mr Keith submitted that the latest information regarding Fresnes MA simply did not add up. The evidence that 44% of inmates were individually locked up could not be squared with the evidence of considerable overcrowding. Mr Keith submitted that the evidence relating to Lyon-Corbas is concerning, and he further submitted that the arithmetic shows that in relation to Rennes-Vezins there would be a plain breach of Article 3. In this regard, we may draw directly from paragraph 47 of Mr Keith's skeleton argument in Esmaili's appeal:

"There should have been a more detailed consideration [by the District Judge] of the material within the 2017 CGLPL report, in which the ratios of cell space to occupancy were still accurate. A cell area of 9sqm, taking out the area for sanitary facilities of 1.6sqm, equates to 7.4sqm. Therefore, for the 112 detainees who are in two-man cells, the space per person is between 3-4sqm (3.7sqm). However, for the 102 detainees who share three to a cell, there would be under 3sqm of space (2.47sqm). Without any further information or assurances to the contrary, this therefore represents a plain breach of Article 3."

134. Ms Malcolm's overarching submission was that both Appellants have failed to adduce sufficient reliable and objective evidence to surmount Aranyosi stage 1. Much of the evidence on which the Appellants relied was hearsay and reportage; greater weight should be given to the information provided by the IJA in Choudhary. Overcrowding per se did not demonstrate that there was a real risk that any prisoner would have to endure less than 3m² of personal space: more was required, and such evidence was not available. Ms Malcolm submitted that the situation has improved recently. She also contested Mr Keith's arithmetic in relation to Rennes-Vezin.

Analysis and Conclusions

135. The principal issue for us to resolve in respect these five establishments is whether the Appellants have adduced sufficient evidence to satisfy Aranyosi stage 1 in relation to personal space, in particular the question of there being a real risk of having less than 3m².
136. Overcrowding can be no more than a general indicator of individual personal space. In Shumba No. 2, the occupancy rate at Fresnes MA was as high as 192% but this did not establish that individual personal space was less than 3m². Assuming that inmates tend to be evenly distributed across the available cells, that would tend to suggest that the occupancy rate would need to be in excess of 200% before the relevant threshold was threatened. We know that it was as high as 203% in January 2018, but since then it has fallen considerably.
137. There is an apparent inconsistency between the conclusions of the ECtHR in JMB and those of this Court in Shumba No. 2 regarding Fresnes. In JMB, the ECtHR held in relation to two applicants that their personal space was less than 3m², and in relation to a third his personal space was between 3-4m² but the violation was made out by appalling physical conditions. However, we emphasise that the inconsistency is apparent for this reason: the French authorities provided no assistance to the ECtHR in JMB, and the vast majority of the complaints related to the period before October/November 2018, which was when the supplementary information was provided to this Court in Shumba.
138. Another facet of Shumba was that the IJA provided supplementary information of "disinsectisation" operations at Fresnes. Whether these included steps to eradicate vermin is unclear, but what is apparent is that this Court was given assistance by the French authorities on an issue which was an important aspect of the adverse conclusion in JMB in relation to the applicant who had established that his personal space would be between 3-4m².
139. In any event, the evidence has moved on from both JMB and Shumba. The IJA has confirmed that the supplementary information provided in Choudhary in August 2020 remains valid. It demonstrates, in our view, that there is no real risk that any inmate would have less than 4m² of personal space let alone less than 3m², because all bar one cell is occupied by no more than two prisoners. We cannot accept Mr Keith's submission that the IJA's figures cannot be squared with the statistic that the cell occupancy level is considerably above 100%. For Mr Keith's submission to be correct, he would need to show exactly how the occupancy rate has been calculated. That evidence has not been provided, and has not been sought. Given that we know that a 192% occupancy rate does not result in a less than 3m² area for personal space, it is reasonable to infer that a 100% occupancy rate would mean that each inmate had well in excess of 4m² of individual space. We would not be minded to seek supplementary information from the IJA to clarify this issue.
140. It follows, in our judgment, that Aranyosi stage 1 has not been fulfilled in relation to Fresnes MA.
141. The evidence in relation to Villipente MA is sparser, but the upshot is the same. Without M. Bechian's evidence, which as we have said should have been made available at a time when it could be tested before the District Judge, the Appellants have not adduced sufficient evidence to satisfy Aranyosi stage 1.
142. We should make it clear that these conclusions are free-standing. We have already concluded that the short duration of stay at these establishments is fatal to the Appellants' case that an Article 3 violation could be made out. Had we concluded otherwise, we would still have found that the Appellants' case fell at the hurdle of Aranyosi stage 1.
143. We consider that the position in relation to Lyon-Corbas and Villefranche-sur-Saône is different, for this straightforward reason. Thus far, we have not touched on Mr A's physical and mental health. The evidence demonstrates that Mr A suffers from a range of conditions which may be summarised as follows. He has a number of orthopaedic problems and is likely to require a right knee replacement within the next three years, and a replacement of the left hip and knee thereafter. He has one functioning kidney, suffers from hypertension and proteinuria, and requires active medical treatment and supervision. He also suffers from a moderately severe depressive illness which would likely deteriorate were he to be extradited. Mr A also suffers from other medical problems but it is unnecessary to mention these specifically. We would need to be satisfied that the range of conditions from which he suffers could be properly treated in the establishment where he would be spending his sentence. The IJA has provided only very general information in this respect. Question marks exist as regards the medical facilities at both of these institutions. In our judgment it is appropriate to seek supplemental information from the IJA as to the expertise of medical practitioners and the availability of treatments for Mr A's specific conditions within these two establishments; and, should it be deemed necessary to remove Mr A to a different prison, like information in that regard. Furthermore, in view of Mr A's medical conditions it is also appropriate to seek further information as to personal floor space etc. as outlined in prargraph 108 of his skeleton argument.
144. We invite the parties to formulate and agree the text of the questions that should be asked and an appropriate timetable for the IJA's replies. In our view, the IJA should also be sent copies of the key medical evidence in addition to this judgment.
145. This brings us to Mr Esmaili's appeal and Rennes-Vezin. The issue here is extremely narrow. In order to arrive at 2.47m² of personal space in relation to cells which are occupied by three inmates, Mr Keith has taken an overall cell area of 7.4m² by subtracting 1.6m² for the sanitary facilities from 9m². Ms Malcolm's competing arithmetic is based on gross figures of 10.5 and 14m². 10.5m² divided by 3 is in fact just under 3m², but it is not altogether clear to us which gross figures should be employed for this purpose.
146. The point remains a narrow one but (1) we cannot say, at least at this stage, that Mr Keith is incorrect, and (2) even on Ms Malcolm's calculations there is, albeit only just, a presumptive violation of Article 3. In the circumstances, there is every merit in seeking to resolve this issue by seeking supplemental information from the IJA. As in Mr A's case, we invite the parties to formulate and agree the text of the question or questions that should be asked and an appropriate timetable for their replies.
147. Annex 1 to this judgment contains the Article 15(2) requests for Supplemental Information in the case of Mr A, and Annex 2 contains the equivalent requests in the case of Mr Esmaili.
ANNEX 1: ARTICLE 15(2) REQUESTS FOR SUPPLEMENTAL INFORMATION IN THE CASE OF MR A
148. In relation to the prisons at Lyon-Corbas and Villefranche-sur-Sâone, and any other institutions in which the Appellant may be detained, please provide the following information:

i) Will the Appellant be accommodated in a cell which provides him with at least 3m² of space (excluding any in-cell sanitary facility) at all times throughout his detention? If the answer is Yes, will he have between 3m² and 4m²? Please give full details of how the surface areas have been calculated including details of the surface area of the sanitary facilities.

ii) Will the overall surface of the cell allow him to move freely between the furniture items in the cell at all times throughout her detention, bearing in mind his mobility difficulties? Please include details of the dimensions of the furniture in the various cells in which the Appellant may be held (i.e. one, two and three man cells, as the case may be.

iii) What will the other detention conditions be for the Appellant throughout his detention, including:

a. Whether he will be accommodated in a cell where he or someone he is sharing with is sleeping on a mattress on the floor;

b. What the sanitary facilities there will be;

c. Whether the toilet will be fully partitioned from the rest of the cell;

d. How many hours a day he will be allowed out of her cell;

e. What meals he will receive;

f. How the cells are heated and to what temperature;

g. The arrangements that will be made to accommodate the Appellant's mobility difficulties and his medical problems, in terms of accessing shower facilities, medical treatment and areas outside his prison cell for work, exercise and recreation;

g. Whether there remains a serious problem with rats and bed bugs at the prison?

149. The Appellant suffers from the following medical conditions:

i) Widespread degenerative pathology affecting both his neck and lower back. Osteoarthritis and increasing pain on his right hip (he is likely to require a right knee replacement within the next three years, and a replacement of the left hip and knee thereafter;

ii) Only one functioning kidney;

iii) Hypertension;

iv) Proteinuria;

v) High blood pressure;

vi) Gout;

vii) Perforation to his right eardrum which is prone to infection;

viii) Asthma;

ix) Chronic plaque cirrhosis;

x) Moderately severe depressive illness.

Relevant medical evidence is attached to these questions.

In relation to Lyon-Corbas and Villefranche-sur-Sâone prisons, and any other institutions in which the Appellant may be detained, please can you specify in detail in relation to each of the medical conditions:

xi) The expertise of medical practitioners in treating these conditions;

xii) The availability of treatments for each of these conditions;

xiii) The arrangements that will be made for the Appellant, with his mobility and other medical difficulties, to access the treatment facilities at the prison and his daily medication.

xiv) The arrangements that will be made to ensure that the Appellant will have ready access to the treatment he may need including interpreters within the medical facility of the prison in question.

ANNEX 2: ARTICLE 15(2) REQUEST FOR SUPPLEMENTAL INFORMATION IN THE CASE OF MOHAMMED ESMAILI

In relation to Rennes Vezin prison: if the Appellant is accommodated in a cell for three inmates, will he be provided with at least 3m² of space (excluding any in-cell sanitary facility) at all times throughout his detention? Please explain this calculation, giving details of the overall surface area of the cell and the surface area of the sanitary facilities.

Note 1   Translated in the equivalent extract from the subsequent judgment of the Cour de Cassation as “a cause of absence of liability”.    [Back]

Note 2   Translated in the equivalent extract from the subsequent judgment of the Cour de Cassation as “international mutual criminal assistance”.    [Back]