EGMR 39692/09, 40713/09, 41008/09 - Austin et al. v. The United Kingdom
 
Judgment
 by the Grand Chamber on 15 March 2012
-- Applications nos. 39692/09, 40713/09 and 41008/09 --
In the case of
Austin and Others v. the United Kingdom
The European Court of Human Rights, sitting as a Grand Chamber composed of: Françoise Tulkens, President, Nicolas Bratza, Jean-Paul Costa, Josep Casadevall, Nina Vajic, Dean Spielmann, Lech Garlicki, Ineta Ziemele, Päivi Hirvelä, Giorgio Malinverni, Luis Lopez Guerra, Ledi Bianku, Kristina Pardalos, Ganna Yudkivska, Vincent A. De Gaetano, Angelika Nußberger, Erik Mose, judges, and Michael O'Boyle, Deputy Registrar,
Having deliberated in private on 14 September 2011 and 15 February 2012,
Delivers the following judgment, which was adopted on the last-mentioned date:
 
Procedure
1. The case originated in one application (no. 39692/09) lodged with the Court under Article 34 of the Convention for the Protection of Human Rights and Fundamental Freedoms ("the Convention") by a British national, Ms Lois Amelia Austin ("the first applicant"), on 17 July 2009; one application (no. 40713/09) lodged by a dual Greek/Australian national, Mr George Black ("the second applicant") on 27 July 2009; and one application (no. 41008/09) lodged by a dual British/Australian national, Ms Bronwyn Lowenthal ("the third applicant") and a British national, Mr Peter O'Shea ("the fourth applicant"), on 27 July 2009. All three applications were brought against the United Kingdom of Great Britain and Northern Ireland.
2. The first applicant was represented before the Court by Ms L. Christian, Ms K. Craig, Ms H. Williams QC and Ms P. Kaufmann. The second applicant was represented by Mr J. Welch. The third and fourth applicants were represented by Mr B. Emmerson QC, Mr M. Fordham QC, Mr A. Bailin and Mr J. Halford. The United Kingdom Government ("the Government") were represented by their Agent, Mr J. Grainger, Foreign and Commonwealth Office.
3. The applicants complained that their restriction within a police cordon (a measure known as "kettling") for up to seven hours during the course of a demonstration in central London amounted to a deprivation of their liberty in breach of Article 5 § 1 of the Convention.
4. The application was allocated to the Fourth Section of the Court (Rule 52 § 1 of the Rules of Court). On 21 September 2010 the Court decided to join the applications and communicate them to the Government. It also decided to rule on the admissibility and merits of the applications at the same time (Article 29 § 1). On 12 April 2011 the Chamber decided to relinquish jurisdiction to the Grand Chamber.
5. The composition of the Grand Chamber was determined according to the provisions of Article 27 §§ 2 and 3 of the Convention and Rule 24 of the Rules of Court.
6. The applicants and the Government each filed a memorial on the admissibility and merits.
7. A hearing took place in public in the Human Rights Building, Strasbourg, on 14 September 2011 (Rule 59 § 3).
There appeared before the Court:
(a) for the Government: Mr J. Grainger (Agent), Mr D. Pannick QC, Mr J. Segan (Counsel), Mr C. Papaleontiou, Ms M. Purdasy (Advisers);
(b) for the applicants: Mr B. Emmerson QC, Ms P. Kaufmann QC, Ms A. Macdonald, Mr I. Steele (Counsel), Ms K. Craig, Mr J. Halford, Mr J. Welch (Advisers), Ms L.A. Austin, Mr G. Black, Ms B. Lowenthal (Applicants).
The Court heard addresses by Mr Pannick, Ms Kaufmann and Mr Emmerson.
 
The Facts
 
I. The Circumstances of the Case
8. The first applicant was born in 1969 and lives in Basildon; the second applicant was born in 1949 and lives in London; the third applicant was born in 1972 and lives in London; and the fourth applicant was born in 1963 and lives in Wembley.
9. The facts of the case may be summarised as follows.
A. The applicants' accounts of what happened to them on 1 May 2001
10. On 1 May 2001, during a demonstration in central London, the applicants were contained within a police cordon at Oxford Circus (the junction between Regent Street and Oxford Street).
11. The first applicant, Ms Lois Austin, is a member of the Socialist Party and had been on many demonstrations, including previous May Day demonstrations. On 1 May 2001 she left her 11-month-old daughter at a crèche, planning to collect her at 4.30 p.m., and travelled from Essex to central London with her partner. They attended a protest against globalisation outside the World Bank before walking with other protesters to Oxford Circus, arriving at about 2 p.m. Around 3.45 p.m. Ms Austin needed to leave the demonstration to collect her daughter from the crèche. She explained her situation to two police officers maintaining the cordon but was told that she could not leave and that it was not known how long it would be before she would be able to leave the area. Ms Austin therefore arranged for a friend to collect the child from the crèche. She was finally allowed to leave at about 9.30 p.m.
12. Between 2 and 2.30 p.m. on 1 May 2001, the second applicant attempted to cross Oxford Circus to go to a bookshop on Oxford Street. He was told by a police officer that he could not walk down Oxford Street because of the approaching crowd of demonstrators and was advised to take Margaret Street, a parallel road to the north. The applicant followed this advice but between Margaret Street and Regent Street he was met by a wall of riot police with shields and helmets moving south. The applicant was forced into Oxford Circus at about 2.30 p.m. He immediately asked to be allowed out of the cordon and was informed that there was an exit for non-protesters at the Bond Street side of Oxford Circus; when he went there he was told that there was no exit. The applicant was not able to exit the cordon until 9.20 p.m.
13. The third applicant had no connection with the demonstration. She worked in the Oxford Circus area and was on her lunch break at 2.10 p.m. when she was prevented from returning to her workplace by a line of police officers blocking the road. She turned and tried to pass in another direction but found that that exit was also now blocked by police officers, who began to advance towards her. She was held within the cordon at Oxford Circus until 9.35 p.m. She and others repeatedly requested to be allowed to leave the cordoned area but was told by the policemen she approached that they were under orders to allow no one to pass.
14. The fourth applicant also worked in the Oxford Circus area and was also caught up in the cordon while walking through Oxford Circus on his lunch break. He was able to leave at approximately 8 p.m.
B. The domestic proceedings
1. The High Court
15. Following the events on 1 May 2001, approximately 150 people who had been confined in Oxford Circus contacted various firms of solicitors with the intention of commencing proceedings. The various potential applicants, their legal representatives and the representatives of the Metropolitan Police entered into correspondence with a view to progressing the claims in an efficient manner. It was agreed that the first applicant and Mr Geoffrey Saxby, who was a passer-by caught up in the cordon, would act as "test" claimants. They commenced proceedings in the High Court, claiming damages for false imprisonment and under the Human Rights Act 1998 for breach of the right to liberty guaranteed by Article 5 of the Convention. The first applicant also initially advanced a claim for interference with her rights of freedom of expression and assembly contrary to Articles 10 and 11 of the Convention, but she subsequently abandoned these claims. The Metropolitan Police provided undertakings to the legal representatives of the other claimants (including the second, third and fourth applicants) that they would not raise any limitation argument if they brought claims in the domestic courts after the test case had been determined.
(a) The facts as found by Mr Justice Tugendhat
16. The hearing in the High Court before Tugendhat J lasted three weeks, including six days of oral evidence. He considered live evidence from 18 lay witnesses and two experts, statements from a further 138 witnesses, thousands of pages of documentary evidence and video footage from hand-held and security cameras and police helicopters. In his judgment, delivered on 23 March 2005 ([2005] EWHC 480 (QB)), Tugendhat J devoted 500 paragraphs to his assessment of the evidence and findings of fact. His factual findings can be summarised as follows.
17. Tugendhat J found that on 18 June and 30 November 1999 and May Day 2000 there had occurred very serious breakdowns in public order in London, which the police feared would be repeated in 2001. The theme of all three demonstrations was protest against capitalism and globalisation. The organisers of the event on 18 June 1999 had failed to cooperate with the police and had distributed publicity material similar to that distributed by the organisers of the May Day 2001 demonstration. During the afternoon of 18 June 1999 a crowd of some 3,000 to 5,000 people, wearing masks, caused approximately 2 million pounds sterling worth of damage and injury to members of the public and police officers, eleven of whom required hospital treatment. Demonstrations on these themes had also resulted in serious breakdowns in public order in other countries at about this time, including in Seattle on 30 November 1999 (the World Trade Organisation meeting), in Washington DC on 16 April 2000 (the International Monetary Fund meeting), in Melbourne on 11 to 13 September 2000 (World Economic Forum Asia Pacific summit), in Prague on 26 September 2000 (another International Monetary Fund meeting), and on 22 April 2001 in Quebec (a Summit of the Americas meeting). The planning for May Day 2001 reflected experience at these and earlier demonstrations, particularly in London, and recommendations made in the light of them.
18. For 1 May 2001, two events had been notified to the police, namely a trade-union May Day March and Rally and a Young Socialist Students March, taking place in different parts of London. In addition, intelligence indicated that activists representing a broad coalition of environmentalist, anarchist and left-wing protest groups intended to stage various protests at twenty-four locations across London based on the squares of "Monopoly", a board-game. The final event was to be a rally at Oxford Circus at 4 p.m. The organisers of the "May Day Monopoly" protest did not make any contact with the police, nor attempt to seek authorisation for the demonstrations, and attempted to maintain secrecy about the locations and nature of the protest. Protesters were directly and indirectly encouraged to wear masks and to engage in looting and violence (see Tugendhat J's judgment, §§ 206-25). The intelligence available to the Police Special Branch was that there would be "500 to 1,000 hard-core demonstrators looking for confrontation, violence and to cause public disorder". The Special Branch assessment was that the protest would involve one of the most serious threats to public order ever seen in London, with a real risk of serious injury and even death, as well as damage to property, if they did not effectively control the crowd. Those at risk included members of the public, police officers and other demonstrators. On 24 April 2001 the Mayor of London wrote an article in the principal London evening newspaper, expressing the view that the organisers of the May Day protest deliberately sought to create destruction and urging Londoners to stay away. Similar warnings were reported in a number of other newspapers in March and April 2001.
19. The police plan for the day involved the deployment of nearly 6,000 officers on foot wearing high-visibility jackets, in addition to mounted police officers. At that time, this was probably the largest number of police officers that had ever been deployed in London. The policemen and women responsible for policing on May Day 2001 were the most experienced public-order officers in England. Since it was foreseen that the final event would be a gathering at Oxford Circus at 4 p.m., a speaker system was installed there. The strategic intentions of the police operation were stated as being to provide public reassurance and ensure public safety; facilitate and police all legitimate protest; prevent public disorder and protect key buildings such as Buckingham Palace and the Houses of Parliament; prevent crime and take all reasonable steps to apprehend offenders if crime was committed; and generally minimise disruption. However, the police had little idea of what to expect or how they would react to it if and when it happened.
20. During the morning of 1 May 2001 there were a number of fairly small demonstrations across London. At around 1 p.m. demonstrators started gathering outside the offices of the World Bank in the Haymarket area. They walked towards Piccadilly Circus then up Regent Street to Oxford Circus. By 2 p.m. it was estimated that there were over 1,500 people in Oxford Circus and that more were steadily joining them. A number of people moving up Regent Street were wearing face masks. The police intelligence had indicated that the gathering at Oxford Circus was to take place at 4 p.m. and the size of the crowd there at this time took them by surprise, with insufficient officers in the area to prevent more people entering it.
21. At approximately 2 p.m. the police decided to put up a cordon to contain the crowd. The decision was based on the available intelligence, which had estimated that 500 to 1,000 individuals intent on violence would take part in the May Day protest and on what had occurred at previous similar demonstrations, rather than on the behaviour of the crowd up until that point. The decision was made in conscious exercise of common-law powers to prevent a breach of the peace. Once the decision to put in a cordon was made, it took five to ten minutes to put in a loose cordon and, as more police officers arrived, twenty to twenty-five minutes to put in a full cordon. There was sufficient space within the cordon for people to walk around and there was no crushing. Nonetheless, as the afternoon progressed, conditions became uncomfortable. The weather was cold and wet. No food or water was provided and there was no access to toilet facilities or shelter.
22. No announcement was made to the crowd when the cordon was first put in place because of police concern that it would not be strong enough to resist a concerted effort by the crowd to break through. The first announcement through the public-address system was made at 4 p.m., informing the crowd that they were being contained to prevent a breach of the peace. It was accepted by the Police Commander in evidence before the domestic courts that the announcement to the crowd could have been made earlier, perhaps at 3.15 or 3.30 p.m.
23. At 2.25 p.m., five minutes after the full cordon was put in place, the Chief Superintendent commanding the operation planned to commence a controlled release from within the cordon up Regent Street north. However, release had to be delayed when members of the crowd both within and outside the cordon started to throw missiles and use violence against the police and when there was an attempt by the crowd to break through the cordon into Regent Street. At 2.55 p.m. a dispersal north was again planned but suspended because of violence by protesters on both sides of the cordon. At around this time more people started moving towards Oxford Circus to take part in the event which had been planned to take place there at 4 p.m. By 3.40 p.m. the police situation report was that officers were sandwiched between crowds, with pushing and shoving and bottles being thrown. At 4.30 p.m. there was a crowd of between 400 and 500 people outside the cordon following a samba band, which made dispersal into Oxford Street difficult. The situation was reviewed at 4.55 p.m. but collective release was ruled out because of the risk of violence and disorder. At 5.15 p.m. there was serious disorder in Oxford Street caused by a core of about twenty-five masked protesters which attracted a crowd of several hundred. At 5.20 p.m. the crowd within the cordon was calm but the police were not willing to risk collective release because of the presence of other large, disorderly crowds nearby.
24. At 5.55 p.m. the decision was taken to disperse the crowd within the cordon. However, the crowd became violent again and at 6.15 p.m. the decision to release was reversed. At 7 p.m. release was commenced, with small groups and individuals being escorted away from the containment. However, at 7.20 p.m. the release was put on hold because of the difficulty in policing the crowds outside the cordon, some of whom were throwing large bits of masonry and burning missiles at the police, and because demonstrators released from the cordon were remaining in the vicinity. At 7.30 p.m. the collective dispersal was resumed, with additional police resources to escort those released from the cordon. However, the dispersal was soon halted again because it became apparent that those being released from Oxford Circus were joining another large crowd, which had previously been violent, to the north in Great Portland Street. By 8 p.m. Portland Place was clear and the collective release of the Oxford Circus crowd recommenced, in groups of ten. By 9.45 p.m. the dispersal was almost complete. Over 100 people were arrested as a result of the disorder in Oxford Circus and the surrounding area. It was part of the collective-release policy to search some or all of those released, take a record of their names and addresses and photograph them.
25. The police estimated that there were a maximum of 2,000 people within the cordon at the peak time and 1,000 in the crowd outside it. Analysis of the documentary and video evidence admitted at trial indicated that, over the course of the afternoon, some 392 people were released individually. It was accepted that this figure was unlikely to be accurate, but Tugendhat J found that the number of individual releases was likely to be "nearer 400 than 200". Of these, most were from the north and south sides of Oxford Circus, with very few people being released to the east and west. Most of the recorded releases were before 4 p.m.; 12 were between 4 p.m. and 5 p.m.; 89 between 5 p.m. and 6 p.m.; 59 between 6 p.m. and 7 p.m.; and 12 individual releases after 7 p.m. It was difficult for police officers to identify members of the crowd for individual release on the basis that they posed no threat of violence. Some of those released were identified in police records as being bystanders caught up in the demonstration. Others were described as being in physical distress, pregnant women, elderly people or children.
(b) Tugendhat J's conclusions
26. The judge concluded that, in the light of the violence which had occurred at previous demonstrations, the intelligence available to the police, the lack of cooperation from the organisers and the condsuct of certain sections of the crowd, the police had reasonable grounds to believe that there was a real risk of damage to property, serious physical injury and even death. The main risks were from crushing and trampling, but there were also risks from missile throwing. Given the situation at Oxford Circus, if they were to prevent violence and injury, the police had no alternative at 2 p.m. but to impose an absolute cordon and this was, therefore, a proportionate response by the police to the presence of the crowd. The principal purpose of the containment was to ensure the safety of persons (including those within the cordon), the preservation of property in Oxford Street and the protection of other rights of third parties. The police also intended to segregate some members of the crowd from others, if appropriate by asking them questions or by searching them.
27. From 2.20 p.m., no one in the crowd was free to leave without permission. The measure was a close confinement, with minimal liberty in Oxford Circus, and its effects were severe, increasingly so as time went by. However, it was never expected that the containment would last so long and the possibility of safely releasing the crowd was kept under review at all times.
28. It was not practicable for the police to release the crowd collectively earlier than they did. There were periods when the dispersal route was blocked by other crowds attempting to get into Oxford Circus. It would not have been reasonable or safe to allow these crowds to join each other without controlling their movement. In addition, there were long periods during which the police did not have the resources to provide for safe dispersal, and it was not suggested that the Commissioner of the Metropolitan Police could or should have made available more police officers than he had on the streets that day. One reason the resources were inadequate was that a substantial proportion of the crowd was not cooperating with the police on the cordons around them. The judge estimated that about 40% were actively hostile at any given time, pushing and throwing missiles, and otherwise showing a lack of cooperation. Those not pushing or throwing missiles were not disassociating themselves from the smaller minority who were. As a result the cordon had to be manned by enough officers to resist a concerted push by these people to break out. A cooperative crowd could have been contained by fewer officers, leaving others available to control the dispersal. The other reason why there were insufficient resources was that other crowds outside the containment were refusing to accept control by or directions from the police. The police were doing the best that they could in the most difficult circumstances. The resources they needed to control the dispersal of the crowd in Oxford Circus were necessarily and properly deployed elsewhere. That did not mean that the police had to allow the crowd in Oxford Circus to disperse without control. According to the judge, that would have been an abnegation of their duty to prevent a breach of the peace, and of their duty of care and the positive obligation incumbent upon them to protect the members of the crowd and third parties, including police officers, from the risk of serious injury, as well as to protect third parties from risk of damage to property.
29. In the circumstances that existed in Oxford Circus, and in particular the difficulty for the police to distinguish between peaceful and violent or potentially violent individuals within the cordon, the judge was unable to find any individual release policy, other than that applied, which would have been workable. Once the cordon was in place, any measure of controlled release was bound to have taken a considerable time before all of the crowd was released. It was impossible to say how long it would have taken if there had been no searches or evidence gathering, but it would have been more than a matter of twenty minutes. If a release was to be combined with searches and evidence gathering, it was bound to take as long as this one took from the time it restarted at 7.30 p.m., that is about one to two hours at least.
30. As regards the false-imprisonment claim, Tugendhat J held that the defence of necessity was available and that the police action was necessary.
31. In connection with Article 5, the judge found that containment within the cordon amounted to deprivation of liberty within the meaning of Article 5 § 1. Although he found that there was never an intention on the part of the police to bring everyone contained at Oxford Circus before a judge, the purpose was to contain the crowd so that the police could arrest and bring before a judge all those they reasonably considered had committed offences and those whose arrest was necessary to prevent them committing offences, and that this was sufficient for the requirements of Article 5 § 1 (c).
32. In addition, Tugendhat J found that, on the unusual facts of the case, there had been no interference with the rights of freedom of speech and assembly. None of the witnesses was able to explain what the purpose of the procession to Oxford Circus was or what it was proposed should have happened either there, or anywhere else, if the police cordon had not been imposed. He found that the literature distributed in advance by the organisers was intended to encourage at least a substantial minority of those present to engage in some form of disorderly and criminal activity, probably including public-order offences such as affray, criminal damage and theft. If there had been no cordon, it would in practice have been impossible in this environment of disorder for anyone intending to carry out lawful acts of protest to do so. Moreover, there was no evidence that anyone at Oxford Circus intended to exercise any rights of speech that they did not in fact exercise. Tugendhat J therefore concluded that the case was about public order and the right to liberty, and not about freedom of speech or freedom of assembly, and he dismissed all the plaintiffs' claims.
(c) The Court of Appeal
33. Tugendhat J granted Ms Austin and Mr Saxby leave to appeal against his findings as regards the claims for false imprisonment and under Article 5 of the Convention. In a judgment delivered on 15 October 2007 ([2007] EWCA Civ 989), the Court of Appeal dismissed the appeal.
34. In connection with the claim of false imprisonment, the Court of Appeal held that, in order to prevent a breach of the peace threatened by others, the police could lawfully take action which interfered with or curtailed the lawful exercise of the rights of innocent third parties, but only where all other possible steps had been taken to avoid a breach of the peace and to protect the rights of third parties, and where the action taken was reasonably necessary and proportionate. Applying this test, in the circumstances of the demonstration at Oxford Circus, Ms Austin's containment was lawful because it was necessary to prevent an imminent breach of the peace by others.
35. In connection with the claim under Article 5, the Court of Appeal concluded that the detention did not amount to a deprivation of liberty. Sir Anthony Clarke MR, giving the judgment of the court, held as follows.
    "102. (...) [T]he first question is whether the appellants were deprived of their liberty from the outset. In our opinion they plainly were not. The position at that time was not markedly different in terms of detention from a number of different types of confinement or detention to which the judge referred which would not be regarded as a deprivation of liberty within Article 5 § 1. A good example is perhaps a football crowd. It is commonplace for such a crowd to be contained for what may turn out to be quite long periods, partly for the protection of individuals in the crowd and partly (in some cases) to avoid crowd violence, perhaps as between groups of opposing supporters (...) Other examples would be (...) for example where motorists are unable to leave a motorway, perhaps for many hours, because of police action following an accident. In such cases it may be necessary for police to confine individuals in particular areas for what may be much longer than originally intended.
    103. In our opinion this was plainly such a case. On the judge's findings of fact, the police had no alternative but to impose the cordon which they did. They anticipated orderly release over two or three hours in order to avoid violence. The judge identified their various purposes, which included safety and the prevention of crime by individuals in the crowd many of whom could not be identified. In these circumstances the original imposition of the cordon could not, in our judgment, properly be regarded as the kind of arbitrary detention which the Strasbourg authorities would describe as deprivation of liberty within the meaning of Article 5 § 1. For these reasons we hold that the judge erred in principle in concluding that the appellants were unlawfully detained as from 2.20 p.m.
    104. On that basis, it is for us to consider afresh the remaining question, namely whether they were unlawfully detained thereafter. In our judgment the answer to that question is 'No'. So for example, (...) on a number of occasions during the afternoon the police gave the order to commence controlled release, only to find that they could not safely carry it through (...) On three occasions the decision to commence controlled dispersal north had to be reviewed or suspended because of the conduct of protesters either inside or outside the contained area, with the result that the final release phase did not begin until 8.02 p.m. (...) During the whole period there was very considerable violence, although not it must be stressed by the appellants (...) As the judge concluded (...) this was not simply a static crowd of protesters in Oxford Circus surrounded by police and held in place for seven hours. It was a dynamic, chaotic, and confusing situation in which there were also a large number of other protesters in the immediate vicinity outside the cordon who were threatening serious disorder and posing a threat to the officers both on the cordon and within it.
    105. In these circumstances it could not sensibly be held that there came a time in which what was originally something less than a deprivation of liberty subsequently became a deprivation of liberty within the meaning of Article 5 § 1 of the Convention. We therefore hold, contrary to the conclusion of the judge, that, if all the relevant circumstances are taken into account, there was not here the kind of arbitrary deprivation of liberty contemplated by the Convention."
(d) The House of Lords
36. Ms Austin, like Mr Saxby, was granted leave to appeal by the House of Lords in connection with the issues under Article 5 § 1 of the Convention. The case was heard on 24 and 25 November 2008, and on 28 January 2009 a unanimous judgment was delivered, dismissing the appeal on the ground that Article 5 § 1 did not apply since the applicant had not been deprived of her liberty ([2009] UKHL 5).
37. Lord Hope of Craighead, with whom all the other Law Lords agreed, explained his approach to the interpretation of "deprivation of liberty" as follows.
    "23. The application of Article 5 § 1 to measures of crowd control is an issue which does not appear so far to have been brought to the attention of the court in Strasbourg. So there is no direct guidance as to whether Article 5 § 1 is engaged where the police impose restrictions on movement for the sole purpose of protecting people from injury or avoiding serious damage to property. The need for measures of crowd control to be adopted in the public interest is not new, however. It is frequently necessary, for example, for such measures to be imposed at football matches to ensure that rival fans do not confront each other in situations that may lead to violence. Restrictions on movement may also be imposed by the police on motorists in the interests of road safety after an accident on a motorway, or to prevent local residents from coming too close to a fire or a terrorist incident. It is not without interest that it has not so far been suggested that restrictions of that kind will breach Article 5 § 1 so long as they are proportionate and not arbitrary.
    24. The restrictions that were imposed by the police cordon in this case may be thought, as compared with the examples that I have just mentioned, to have been greater in degree and intensity. But Lord Pannick QC for the respondent submitted that one could not sensibly ignore the purpose of the restriction or the circumstances. Detention in the paradigm sense was not in the minds of anyone. There would have been no question of there being a deprivation of liberty if the cordon had remained in place for only twenty minutes. The fact that it remained in place for much longer ought to make no difference, as the fact that it was not possible to release everyone from the cordon earlier was due to circumstances that were beyond the control of the police. This was a case, he said, where the answer to the question whether what was done was within the scope of Article 5 § 1 was to be determined by striking a fair balance between the rights of the individual and the interests of society. It was, of course, necessary to give full effect to the fact that Article 5 was a fundamental right whose importance was paramount. But the fact that infringement was not open to justification except in the cases listed in sub-paragraphs (a) to (f) pointed to the need for care to be taken to identify the limits of its application.
    25. Ms Williams QC for the appellant, on the other hand, said that the purpose for which the measure was employed was irrelevant. The fact that it was a necessary response and was proportionate was a pre-condition for establishing the measure's legality for the purpose of sub-paragraphs (a) to (f) of Article 5 § 1. But it went no further than that. There was no balance to be struck when consideration was being given to the initial question whether Article 5 § 1 applied to the measures adopted by the police. Questions of purpose and balance only arose when consideration was being given to the cases listed in sub-paragraphs (a) to (f).
    Is purpose relevant?
    26. The decision whether there was deprivation of liberty is, of course, highly sensitive to the facts of each case. Little value can be derived therefore from decisions on the application of Article 5 that depend entirely on their own facts. But they are of value where they can be said to illustrate issues of principle. In the present context some assistance is to be derived from the cases as to the extent to which regard can be had to the aim or purpose of the measure in question when consideration is being given as to whether it is within the ambit of Article 5 § 1 at all.
    27. If purpose is relevant, it must be to enable a balance to be struck between what the restriction seeks to achieve and the interests of the individual. The proposition that there is a balance to be struck at the initial stage when the scope of the Article is being considered was not mentioned in Engel v The Netherlands (No 1) (1976) 1 EHRR 647 or Guzzardi v Italy (1980) 3 EHRR 333. Nor can it be said to be based on anything that is to be found in the wording of the Article. But I think that there are sufficient indications elsewhere in the court's case-law that the question of balance is inherent in the concepts that are enshrined in the Convention and that they have a part to play when consideration is being given to the scope of the first rank of fundamental rights that protect the physical security of the individual."
Lord Hope then reviewed a number of judgments and decisions of the Court and Commission, including X. v. the Federal Republic of Germany, no. 8819/79, Commission decision of 19 March 1981, Decisions and Reports (DR) 24, p. 158; Guenat v. Switzerland, no. 24722/94, Commission decision of 10 April 1995, DR 81-B, p. 130; H.M. v. Switzerland, no. 39187/98, ECHR 2002-II; Nielsen v. Denmark, 28 November 1988, Series A no. 144; Soering v. the United Kingdom, 7 July 1989, Series A no. 161; O'Halloran and Francis v. the United Kingdom [GC], nos. 15809/02 and 25624/02, ECHR 2007-III; and N. v. the United Kingdom [GC], no. 26565/05, ECHR 2008. He concluded:
    "34. I would hold therefore that there is room, even in the case of fundamental rights as to whose application no restriction or limitation is permitted by the Convention, for a pragmatic approach to be taken which takes full account of all the circumstances. No reference is made in Article 5 to the interests of public safety or the protection of public order as one of the cases in which a person may be deprived of his liberty. This is in sharp contrast to Article 10 § 2, which expressly qualifies the right to freedom of expression in these respects. But the importance that must be attached in the context of Article 5 to measures taken in the interests of public safety is indicated by Article 2 of the Convention, as the lives of persons affected by mob violence may be at risk if measures of crowd control cannot be adopted by the police. This is a situation where a search for a fair balance is necessary if these competing fundamental rights are to be reconciled with each other. The ambit that is given to Article 5 as to measures of crowd control must, of course, take account of the rights of the individual as well as the interests of the community. So any steps that are taken must be resorted to in good faith and must be proportionate to the situation which has made the measures necessary. This is essential to preserve the fundamental principle that anything that is done which affects a person's right to liberty must not be arbitrary. If these requirements are met however it will be proper to conclude that measures of crowd control that are undertaken in the interests of the community will not infringe the Article 5 rights of individual members of the crowd whose freedom of movement is restricted by them."
Lord Neuberger of Abbotsbury agreed that there had been no deprivation of liberty, and observed as follows.
    "58. The police are under a duty to keep the peace when a riot is threatened, and to take reasonable steps to prevent serious public disorder, especially if it involves violence to individuals and property. Any sensible person living in a modern democracy would reasonably expect to be confined, or at least accept that it was proper that she could be confined, within a limited space by the police, in some circumstances. Thus, if a deranged or drunk person was on the loose with a gun in a building, the police would be entitled, indeed expected, to ensure that, possibly for many hours, members of the public were confined to where they were, even if it was in a pretty small room with a number of other people. Equally, where there are groups of supporters of opposing teams at a football match, the police routinely, and obviously properly, ensure that, in order to avoid violence and mayhem, the two groups are kept apart; this often involves confining one or both of the groups within a relatively small space for a not insignificant period. Or if there is an accident on a motorway, it is common, and again proper, for the police to require drivers and passengers to remain in their stationary motor vehicles, often for more than an hour or two. In all such cases, the police would be confining individuals for their own protection and to prevent violence to people or property.
    59. So, too, as I see it, where there is a demonstration, particularly one attended by a justified expectation of substantial disorder and violence, the police must be expected, indeed sometimes required, to take steps to ensure that such disorder and violence do not occur, or, at least, are confined to a minimum. Such steps must often involve restraining the movement of the demonstrators, and sometimes of those members of the public unintentionally caught up in the demonstration. In some instances, that must involve people being confined to a relatively small space for some time.
    60. In such cases, it seems to me unrealistic to contend that Article 5 can come into play at all, provided, and it is a very important proviso, that the actions of the police are proportionate and reasonable, and any confinement is restricted to a reasonable minimum, as to discomfort and as to time, as is necessary for the relevant purpose, namely the prevention of serious public disorder and violence.
    61. It was suggested on behalf of the appellant that, at any rate in some of the examples I have given, consent to being confined could be imputed to the people concerned. I am not sure that that is a satisfactory analysis, not least because, unless the consent is to be treated as being involuntary or irrebuttably deemed to be given, it would not deal with the case of a person who informed the police that he objected to being confined. However, if imputed consent is an appropriate basis for justifying confinement for Article 5 purposes, then it seems to me that the confinement in the present case could be justified on the basis that anyone on the streets, particularly on a demonstration with a well-known risk of serious violence, must be taken to be consenting to the possibility of being confined by the police, if it is a reasonable and proportionate way of preventing serious public disorder and violence.
    62. So, in agreement with the Court of Appeal, I would hold that, in the light of the findings of the Judge, as summarised in para [57] above, the actions of the police in the present case did not give rise to any infringement of the appellant's Article 5 rights. The feature of the present case which gives particular cause for concern is the length of the period of confinement, nearly seven hours. However, having reached the conclusion that reasonable and proportionate constraint, which is requisite to prevent serious public disorder and violence, does not infringe Article 5, it seems to me hard to contend that the mere fact that the period of constraint was unusually long can, of itself, convert a situation which would otherwise not be within the ambit of Article 5 into one which is. I think that some support for that view can be found in cases where it has been held that detention in prison is not taken out of Article 5 because it was only for a short time -- see e.g. Novotka v Slovakia (Application No 47244/99) 4 November 2003.
    63. As already indicated, it appears to me that the intention of the police is relevant, particularly in a non-paradigm case, such as this, and where the intention is manifest from the external circumstances. If it transpired, for instance, that the police had maintained the cordon, beyond the time necessary for crowd control, in order to punish, or 'to teach a lesson' to, the demonstrators within the cordon, then it seems to me that very different considerations would arise. In such circumstances, I would have thought that there would have been a powerful argument for saying that the maintenance of the cordon did amount to a detention within the meaning of Article 5. However, as is apparent from the clear and careful findings made by the Judge, which have quite rightly not been challenged on appeal, there could be no question of such a contention being raised in the present case.
    64. Furthermore, it is worth bearing in mind that, at least as I see it, if the restraint in the present case did amount to detention within Article 5, it would not be possible for the police to justify the detention under the exceptions in paras (b) or (c), not least because of the reasoning of the European Court in Lawless v Ireland (No 3) (1961) 1 EHRR 15. I consider that the fact that the restraint in the present case could not be justified under any of the exceptions in paras (a) to (f) supports the contention that the constraint did not amount to detention within Article 5 at all. It would appear to me to be very odd if it was not open to the police to act as they did in the instant circumstances, without infringing the Article 5 rights of those who were constrained."
Lord Carswell agreed with Lord Hope, and Lord Scott of Foscote agreed with Lords Hope and Neuberger, emphasising that "the purpose of the confinement or restriction and the intentions of the persons responsible for imposing it rank very highly in the circumstances to be taken into account in reaching the decision" whether there has been a deprivation of liberty.
Lord Walker of Gestingthorpe agreed with Lord Hope, but added as a "footnote":
    "43. In paras 26ff of his opinion Lord Hope poses the question 'Is purpose relevant?' His conclusion is a very guarded one, that is (para 34) that there is room, even in the case of fundamental rights, for a pragmatic approach which takes full account of all the circumstances. I respectfully agree that it is right to be cautious on this point. The Strasbourg Court has frequently made clear that all the surrounding circumstances may be relevant in determining whether there is a deprivation of liberty: see for instance HM v Switzerland (2004) 38 EHRR 314, para 42:
    (...)
    It is noteworthy that the listed factors, wide as they are, do not include purpose.
    44. The purpose of confinement which may arguably amount to deprivation of liberty is in general relevant, not to whether the threshold is crossed, but to whether that confinement can be justified under Article 5 § 1 (a) to (f): see for instance (in relation to Article 5 § 1 (e)) Nielsen v Denmark (1988) 11 EHRR 175; Litwa v Poland (2001) 33 EHRR 1267; Wall v Sweden, (10 December 2002) admissibility decision 41403/98; HM v Switzerland (above); HL v United Kingdom (2005) 40 EHRR 32; Enhorn v Sweden (2005) 41 EHRR 633; and Storck v Germany (2006) 43 EHRR 96. If confinement amounting to deprivation of liberty and personal security is established, good intentions cannot make up for any deficiencies in justification of the confinement under one of the exceptions listed in Article 5 § 1 (a) to (f), which are to be strictly construed.
    45. Many of these Article 5 § 1 (e) cases also raise issues as to express or implied consent (to admission to a psychiatric ward or old people's home). Some of the earlier cases seem questionable today insofar as they relied on 'parental rights' (especially Nielsen, which was a nine-seven decision that the admission to a psychiatric ward of a 12-year-old boy was not a deprivation of liberty, because of his mother's 'parental rights'). Storck has, I think, sent out a clear message indicating a different approach to the personal autonomy of young people (although the unfortunate claimant in that case was 18 years of age at the time of her compulsory medication in a locked ward in the clinic at Bremen, for which she was made an exceptionally large award for non-pecuniary loss).
    46. I also feel some unease about the decision in X v Germany (19 March 1981) admissibility decision 8819/79; police stations can be intimidating places for anyone, particularly children, and it seems rather disingenuous to reason that 'in the present case the police action was not aimed at depriving the children of their liberty but simply to obtain information from them about how they obtained possession of the objects found on them and about thefts which had occurred in the school previously'.
    47. Having said all that, however, I conclude that it is essential, in the present case, to pose the simple question: what were the police doing at Oxford Circus on 1 May 2001? What were they about? The answer is, as Lord Hope has explained in his full summary of the judge's unchallenged findings, that they were engaged in an unusually difficult exercise in crowd control, in order to avoid personal injuries and damage to property. The senior officers conducting the operations were determined to avoid a fatality such as occurred in Red Lion Square on 15 June 1974. The aim of the police was to disperse the crowd, and the fact that the achievement of that aim took much longer than they expected was due to circumstances beyond their control."
 
The Law
 
I. Alleged Violation of Article 5 § 1 of the Convention
38. The applicants complained that they were deprived of their liberty in breach of Article 5 § 1 of the Convention, which reads as follows:
    "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;
    (b) the lawful arrest or detention of a person for non-compliance with the lawful order of a court or in order to secure the fulfilment of any obligation prescribed by law;
    (c) the lawful arrest or detention of a person effected for the purpose of bringing him before the competent legal authority on reasonable suspicion of having committed an offence or when it is reasonably considered necessary to prevent his committing an offence or fleeing after having done so;
    (d) the detention of a minor by lawful order for the purpose of educational supervision or his lawful detention for the purpose of bringing him before the competent legal authority;
    (e) the lawful detention of persons for the prevention of the spreading of infectious diseases, of persons of unsound mind, alcoholics or drug addicts or vagrants;
    (f) the lawful arrest or detention of a person to prevent his effecting an unauthorised entry into the country or of a person against whom action is being taken with a view to deportation or extradition."
39. The Government argued that Article 5 § 1 did not apply because there had been no deprivation of liberty. In the alternative, they contended that any deprivation of liberty had been in conformity with Article 5 § 1 (b) and/or (c).
A. The parties' submissions
1. The Government
40. The Government submitted that the police had not deprived the applicants of their liberty, within the meaning of Article 5 § 1. They emphasised that it was one of the most basic principles inherent in the Convention that a fair balance should be struck between the interests of the community and the protection of the rights of the individual. The need to strike such a balance was a theme running through the Court's case-law and it should also be taken into account when determining the existence of a deprivation of liberty. Unlike Articles 8 to 11 of the Convention, Article 5 § 1 was not subject to a general justification provision. It was therefore important not to adopt too broad a concept of "deprivation of liberty".
41. The relevant principles for identifying a deprivation of liberty were first set out by the Court in Engel and Others v. the Netherlands, 8 June 1976, §§ 58-59, Series A no. 22, followed in Guzzardi v. Italy, 6 November 1980, §§ 92-93, Series A no. 39 and numerous subsequent cases. This case-law made it clear that the question whether there was a deprivation of liberty had to be determined with reference to the specific facts. The duration of the measure was one factor to be taken into account, but the fact that controls were imposed for a significant period was not in itself sufficient to trigger a deprivation of liberty, as was clear from the cases on night curfews: see Raimondo v. Italy, 22 February 1994, Series A no. 281-A and Trijonis v. Lithuania, no. 2333/02, 15 December 2005. The purpose for which a measure was imposed was a relevant factor and could weigh against the Court finding a deprivation of liberty, even where there was physical confinement in a specific place for a lengthy period: see Engel and Others, cited above, § 59, where the Court observed that the context of military discipline had to be taken into account; see also Nielsen v. Denmark (28 November 1988, Series A no. 144) and H.M. v. Switzerland (no. 39187/98, ECHR 2002-II), where the Court, in holding that Article 5 did not apply, took into account the humanitarian purpose behind the confinement. This approach was correct in principle since the objective of Article 5 § 1 was to prevent arbitrary and unjustified detention.
42. None of the Court's case-law to date addressed circumstances such as those in issue here, where it had been necessary for the police to take proportionate action to confine persons for a limited time to prevent serious public disorder involving a substantial risk of death or serious injury. If Article 5 were to be interpreted as preventing the imposition of a cordon at Oxford Circus on May Day 2001, the police in Contracting States would be obliged to prepare alternative methods of dealing with violent demonstrations, which would not raise Article 5 issues but which might be far more dangerous for all concerned, such as the use of tear gas or rubber bullets.
43. The House of Lords and Court of Appeal correctly applied principles derived from the Court's case-law to find that there was no deprivation of liberty when the cordon was initially imposed. A temporary restraint on freedom of movement along a public highway, even if absolute, did not amount to a deprivation of liberty, as was clear from the examples involving crowds at football matches or traffic on motorways referred to by the domestic courts. The question then was whether the length of time for which the cordon was imposed made it into a deprivation of liberty. That had to depend on all the circumstances, in particular the purpose of the police to protect the safety of those within and outside the cordon and the necessity of the measure since there were no other steps which the police could have taken to prevent serious public disorder.
44. In the alternative, if there had been a deprivation of liberty, it was justified under Article 5 § 1 (b) to secure the fulfilment of the "obligation prescribed by law" to assist a constable in dealing with a breach of the peace. In the further alternative, any deprivation of liberty also fell within the exception of Article 5 § 1 (c), in that the confinement of each applicant was necessary in order to enable the police to prevent the apprehended breach of the peace.
2. The applicants
45. In the applicants' submission, in order to establish whether there had been a deprivation of liberty it was necessary objectively to evaluate the individual's concrete situation, in particular whether there had been "confinement to a certain limited place for a not negligible length of time", and whether he or she "validly consented" to this happening (see Storck v. Germany, no. 61603/00, § 74, ECHR 2005-V). Where the measure employed did not involve detention in the classic sense of imprisonment, it should be evaluated by reference to the nature and extent of the confinement, the manner of its implementation, its duration and its effect upon the applicant. Thus, for example, the greater the extent of the confinement and the greater the degree of coercion by the authorities, the shorter the duration required before a deprivation of liberty would be found.
46. The applicants reasoned that the Government's observations were based on the novel and controversial proposition that containment required for a benevolent or public-interest purpose would not amount to a deprivation of liberty, at least outside the classic imprisonment situation. The applicants disagreed and submitted that if a measure was employed in circumstances that would otherwise amount to a deprivation of liberty, the intention or purpose with which it was undertaken was irrelevant to the assessment of whether a deprivation of liberty had occurred. The purpose for which the measure was undertaken was relevant only to determine whether an established deprivation of liberty was justified by reference to the six purposes specified at Article 5 § 1 (a) to (f), which were, in any case, to be narrowly construed. It was not possible to interpret the concept of deprivation of liberty differently in relation to restrictions imposed on grounds of public order, as opposed to measures imposed for any other benevolent or public-interest purpose.
47. The applicants further contended that the Government's attempt to support their argument by reference to the search for a fair balance between the demands of the public interest and the need to protect the rights of individuals was misconceived. This fair balance had already been struck by the very formulation of the rights protected by the Convention. The Court was not free to weigh competing public-interest considerations to narrow down the scope of protection. Any assessment of fair balance undertaken by the Court was informed by the structure of the particular Article and only occurred within the spaces left by that structure, for example in delimiting the scope of positive obligations arising from certain Articles. If the Government's central contention were correct, States would be able to circumvent the protections of Article 5, detaining people for a wide range of reasons beyond the scope of Article 5 § 1 (a) to (f), provided that necessity was shown and without those individuals enjoying the safeguards, procedural and substantive, afforded by Article 5.
48. The applicants did not contend that there was a deprivation of liberty from the moment the cordon was imposed. However, in view of the nature of the confinement, its coercive enforcement, long duration and effect upon them, the containment within the police cordon was clearly a deprivation of their liberty. The fact that the domestic Courts found it was a necessary public-order measure was irrelevant to this issue.
49. The deprivation of liberty was not justified under any of the sub-paragraphs of Article 5 § 1. In particular, as regards Article 5 § 1 (b), the applicants were not detained to "secure the fulfilment of any obligation prescribed by law", since this justification could only arise where a specific and concrete obligation upon the individual had arisen and the deprivation of liberty was incidental to its fulfilment. The "obligation" could not be a requirement to submit to the deprivation of liberty itself. Furthermore, detention pursuant to Article 5 § 1 (c), "when it is reasonably considered necessary to prevent his committing an offence", required both that the deprivation occurred to prevent the individual in question from committing a particular offence and that it was the intention of the authorities at the time of the deprivation to bring the individual before the competent legal authority in the course of criminal proceedings. Neither of these requirements could be shown in their case.
B. The Court's assessment
1. Admissibility
50. The Court considers that the question whether the applicants were deprived of their liberty, and therefore whether Article 5 § 1 applies, is closely linked to the merits of the applicants' complaints. It therefore joins this preliminary issue to the merits.
51. It finds that the case is not inadmissible on any other grounds and it therefore declares it admissible.
2. Merits
(a) General principles
52. It is true, as the parties point out, that this is the first time that the Court has considered the application of Article 5 § 1 of the Convention in respect of the "kettling" or containment of a group of people carried out by the police on public-order grounds. In interpreting Article 5 § 1 in these circumstances, and in particular in determining whether there has been a deprivation of liberty, the Court draws guidance from the following general principles.
53. Firstly, as the Court has underlined on many occasions, the Convention is a living instrument which must be interpreted in the light of present-day conditions and of the ideas prevailing in democratic States today (see, among other authorities, Tyrer v. the United Kingdom, 25 April 1978, § 31, Series A no. 26; Kress v. France [GC], no. 39594/98, § 70, ECHR 2001-VI; Christine Goodwin v. the United Kingdom [GC], no. 28957/95, § 75, ECHR 2002-VI; and, most recently, Bayatyan v. Armenia [GC], no. 23459/03, § 102, ECHR 2011). This does not, however, mean that to respond to present-day needs, conditions, views or standards the Court can create a new right apart from those recognised by the Convention (see Johnston and Others v. Ireland, 18 December 1986, §§ 51-54, Series A no. 112) or that it can whittle down an existing right or create a new "exception" or "justification" which is not expressly recognised in the Convention (see, for example, Engel and Others, cited above, § 57, and Ciulla v. Italy, 22 February 1989, § 41, Series A no. 148).
54. Secondly, the Convention must be read as a whole, and interpreted in such a way as to promote internal consistency and harmony between its various provisions (see Stec and Others v. the United Kingdom (dec.) [GC], nos. 65731/01 and 65900/01, § 48, ECHR 2005-X).
55. Given the context in which this containment measure took place in the instant case, the Court considers it appropriate to note, for the sake of completeness, that Article 2 of Protocol No. 4 to the Convention guarantees the right to liberty of movement. It is true that the applicants did not rely upon this provision, since the United Kingdom has not ratified Protocol No. 4 and is thus not bound by it. In the Court's view, however, taking into account the importance and purport of the distinct provisions of Article 5 and of Article 2 of Protocol No. 4, it is helpful to make the following reflections. Firstly, Article 5 should not, in principle, be interpreted in such a way as to incorporate the requirements of Protocol No. 4 in respect of States which have not ratified it, including the United Kingdom. At the same time, Article 2 § 3 of the said Protocol permits restrictions to be placed on the right to liberty of movement where necessary, inter alia, for the maintenance of public order, the prevention of crime or the protection of the rights and freedoms of others. In connection with Article 11 of the Convention, the Court has held that interferences with the right of freedom of assembly are in principle justified for the prevention of disorder or crime and for the protection of the rights and freedoms of others where demonstrators engage in acts of violence (see Giuliani and Gaggio v. Italy [GC], no. 23458/02, § 251, ECHR 2011). It has also held that, in certain well-defined circumstances, Articles 2 and 3 may imply positive obligations on the authorities to take preventive operational measures to protect individuals at risk of serious harm from the criminal acts of other individuals (see Giuliani and Gaggio, cited above, § 244, and P.F. and E.F. v. the United Kingdom (dec.), no. 28326/09, § 36, 23 November 2010). When considering whether the domestic authorities have complied with such positive obligations, the Court has held that account must be taken of the difficulties involved in policing modern societies, the unpredictability of human conduct and the operational choices which must be made in terms of priorities and resources (see Giuliani and Gaggio, cited above, § 245, and P.F. and E.F. v. the United Kingdom, cited above, § 40).
56. As the Court has previously stated, the police must be afforded a degree of discretion in taking operational decisions. Such decisions are almost always complicated and the police, who have access to information and intelligence not available to the general public, will usually be in the best position to make them (see P.F. and E.F. v. the United Kingdom, cited above, § 41). Moreover, even by 2001, advances in communications technology had made it possible to mobilise protesters rapidly and covertly on a hitherto unknown scale. Police forces in the Contracting States face new challenges, perhaps unforeseen when the Convention was drafted, and have developed new policing techniques to deal with them, including containment or "kettling". Article 5 cannot be interpreted in such a way as to make it impracticable for the police to fulfil their duties of maintaining order and protecting the public, provided that they comply with the underlying principle of Article 5, which is to protect the individual from arbitrariness (see Saadi v. the United Kingdom [GC], no. 13229/03, §§ 67-74, ECHR 2008).
57. As mentioned above, Article 5 § 1 is not concerned with mere restrictions on liberty of movement, which are governed by Article 2 of Protocol No. 4. In order to determine whether someone has been "deprived of his liberty" within the meaning of Article 5 § 1, the starting-point must be his concrete situation and account must be taken of a whole range of criteria such as the type, duration, effects and manner of implementation of the measure in question. The difference between deprivation of and restriction upon liberty is one of degree or intensity, and not of nature or substance (see Engel and Others, § 59; Guzzardi, §§ 92-93; Storck, § 71, all cited above; and also, more recently, Medvedyev and Others v. France [GC], no. 3394/03, § 73, ECHR 2010).
58. As Lord Walker pointed out (see paragraph 37 above), the purpose behind the measure in question is not mentioned in the above judgments as a factor to be taken into account when deciding whether there has been a deprivation of liberty. Indeed, it is clear from the Court's case-law that an underlying public-interest motive, for example to protect the community against a perceived threat emanating from an individual, has no bearing on the question whether that person has been deprived of his liberty, although it might be relevant to the subsequent inquiry whether the deprivation of liberty was justified under one of the sub-paragraphs of Article 5 § 1 (see, among many examples, A. and Others v. the United Kingdom [GC], no. 3455/05, § 166, ECHR 2009; Enhorn v. Sweden, no. 56529/00, § 33, ECHR 2005-I; and M. v. Germany, no. 19359/04, ECHR 2009). The same is true where the object is to protect, treat or care in some way for the person taken into confinement, unless that person has validly consented to what would otherwise be a deprivation of liberty (see Storck, cited above, §§ 74-78, and the cases cited therein and, most recently, Stanev v. Bulgaria [GC], no. 36760/06, § 117, ECHR 2012; see also, as regards validity of consent, Amuur v. France, 25 June 1996, § 48, Reports of Judgments and Decisions 1996-III).
59. However, the Court is of the view that the requirement to take account of the "type" and "manner of implementation" of the measure in question (see Engel and Others, § 59 and Guzzardi, § 92, both cited above) enables it to have regard to the specific context and circumstances surrounding types of restriction other than the paradigm of confinement in a cell (see, for example, Engel and Others, cited above, § 59, and Amuur, cited above, § 43). Indeed, the context in which action is taken is an important factor to be taken into account, since situations commonly occur in modern society where the public may be called upon to endure restrictions on freedom of movement or liberty in the interests of the common good. As the judges in the Court of Appeal and House of Lords observed, members of the public generally accept that temporary restrictions may be placed on their freedom of movement in certain contexts, such as travel by public transport or on the motorway, or attendance at a football match (see paragraphs 35 and 37 above). The Court does not consider that such commonly occurring restrictions on movement, so long as they are rendered unavoidable as a result of circumstances beyond the control of the authorities and are necessary to avert a real risk of serious injury or damage, and are kept to the minimum required for that purpose, can properly be described as "deprivations of liberty" within the meaning of Article 5 § 1.
60. Article 5 enshrines a fundamental human right, namely the protection of the individual against arbitrary interference by the State with his or her right to liberty. Sub-paragraphs (a) to (f) of Article 5 § 1 contain an exhaustive list of permissible grounds on which persons may be deprived of their liberty and no deprivation of liberty will be compatible with Article 5 § 1 unless it falls within one of those grounds (see, amongst many other authorities, Al-Jedda v. the United Kingdom [GC], no. 27021/08, § 99, ECHR 2011). It cannot be excluded that the use of containment and crowd-control techniques could, in particular circumstances, give rise to an unjustified deprivation of liberty in breach of Article 5 § 1. In each case, Article 5 § 1 must be interpreted in a manner which takes into account the specific context in which the techniques are deployed, as well as the responsibilities of the police to fulfil their duties of maintaining order and protecting the public, as they are required to do under both national and Convention law.
(b) Application of these principles to the facts of the case
61. The question whether there has been a deprivation of liberty is, therefore, based on the particular facts of the case. In this connection, the Court observes that within the scheme of the Convention it is intended to be subsidiary to the national systems safeguarding human rights (see A. and Others v. the United Kingdom, cited above, § 154). Subsidiarity is at the very basis of the Convention, stemming as it does from a joint reading of Articles 1 and 19. The Court must be cautious in taking on the role of a first-instance tribunal of fact, where this is not rendered unavoidable by the circumstances of a particular case. As a general rule, where domestic proceedings have taken place, it is not the Court's task to substitute its own assessment of the facts for that of the domestic courts and it is for the latter to establish the facts on the basis of the evidence before them. Though the Court is not bound by the findings of domestic courts and remains free to make its own appreciation in the light of all the material before it, in normal circumstances it requires cogent elements to lead it to depart from the findings of fact reached by the domestic courts (see Giuliani and Gaggio, cited above, § 180). Nonetheless, since pursuant to Articles 19 and 32 of the Convention it is the Court's role definitively to interpret and apply the Convention, while it must have reference to the domestic court's findings of fact, it is not constrained by their legal conclusions as to whether or not there has been a deprivation of liberty within the meaning of Article 5 § 1 (see, for example, Storck, cited above, § 72).
62. Tugendhat J's judgment at first instance followed a three-week trial, during which he considered a substantial body of evidence about the events at Oxford Circus on May Day 2001, including oral testimony and documentary, video and photographic evidence (see paragraph 16 above). He found, inter alia, that the information available in advance to the police indicated that the demonstration would attract a "hard core" of 500 to 1,000 violent demonstrators and that there was a real risk of serious injury, even death, and damage to property if the crowds were not effectively controlled. The police were expecting a crowd to form at Oxford Circus at around 4 p.m. and they were taken by surprise when over 1,500 people gathered there two hours earlier. In the light of the intelligence they had received and the behaviour of crowds at earlier demonstrations on similar themes, the police decided that, if they were to prevent violence and the risk of injury to persons and damage to property, an absolute cordon had to be imposed at 2 p.m. From 2.20 p.m., when a full cordon was in place, no one in the crowd was free to leave the area without permission. There was space within the cordon for people to walk around and there was no crushing, but conditions were uncomfortable, with no shelter, food, water or toilet facilities. Throughout the afternoon and evening, attempts were made by the police to commence collective release but the violent and uncooperative behaviour of a significant minority both within the cordon and in the surrounding area outside led the police repeatedly to suspend dispersal. In consequence, full dispersal could not be completed until 9.30 p.m. However, the police permitted approximately 400 individuals, who could clearly be identified as not being involved in the demonstration or who were seriously affected by being confined, to leave (see paragraphs 17-25 above). These findings were not disputed by the parties to the present proceedings and the Court sees no ground to depart from them. The first, second and third applicants were confined within the police cordon for approximately seven hours and the fourth applicant for five and a half hours.
63. The Court must analyse the applicants' concrete situation with reference to the criteria set out in Engel and Others and the subsequent case-law (see paragraph 57 above). Although there were differences between the applicants, in that the first applicant was present in Oxford Circus as a demonstrator whereas the other applicants were passers-by, the Court does not consider that this difference is relevant to the question whether there was a deprivation of liberty.
64. In accordance with the Engel and Others criteria, the Court considers that the coercive nature of the containment within the cordon, its duration and its effect on the applicants, in terms of physical discomfort and inability to leave Oxford Circus, point towards a deprivation of liberty.
65. However, the Court must also take into account the "type" and "manner of implementation" of the measure in question. As indicated above, the context in which the measure was imposed is significant.
66. It is important to note, therefore, that the measure was imposed to isolate and contain a large crowd, in volatile and dangerous conditions. As the Government pointed out (see paragraph 42 above), the police decided to make use of a measure of containment to control the crowd rather than having resort to more robust methods, which might have given rise to a greater risk of injury to people within the crowd. The trial judge concluded that, given the situation in Oxford Circus, the police had had no alternative but to impose an absolute cordon if they were to avert a real risk of serious injury or damage (see paragraph 26 above). The Court finds no reason to depart from the judge's conclusion that in the circumstances the imposition of an absolute cordon was the least intrusive and most effective means to be applied. Indeed, the applicants did not contend that, when the cordon was first imposed, those within it were immediately deprived of their liberty (see paragraph 48 above).
67. Moreover, again on the basis of the facts found by the trial judge, the Court is unable to identify a moment when the measure changed from what was, at most, a restriction on freedom of movement to a deprivation of liberty. It is striking that, some five minutes after the absolute cordon was imposed, the police were planning to commence a controlled release towards the north. Thirty minutes later, a second attempt by the police to begin release was begun but suspended because of the violent behaviour of those within and outside the cordon. Between about 3 p.m. and 6 p.m., the police kept the situation under review but the arrival of a new group of protesters and the dangerous conditions within the crowds led them to consider that it would not be safe to attempt to release those within the cordon. Controlled release was recommenced at 5.55 p.m. but stopped at 6.15 p.m.; resumed at 7 p.m. and suspended at 7.20 p.m.; begun again at 7.30 p.m. and again abandoned; then carried out continuously, in groups of ten, until the entire crowd had been released at 9.45 p.m. (see paragraph 24 above). Thus, the trial judge found the same conditions which required the police to contain the crowd at 2 p.m. persisted until about 8 p.m., when the collective release was finally able to proceed without interruption (see paragraph 24 above). In these circumstances, where the police kept the situation constantly under close review, but where substantially the same dangerous conditions which necessitated the imposition of the cordon at 2 p.m. continued to exist throughout the afternoon and early evening, the Court does not consider that those within the cordon can be said to have been deprived of their liberty within the meaning of Article 5 § 1. Since there was no deprivation of liberty, it is unnecessary for the Court to examine whether the measure in question was justified in accordance with sub-paragraphs (b) or (c) of Article 5 § 1.
68. The Court emphasises that the above conclusion, that there was no deprivation of liberty, is based on the specific and exceptional facts of this case. Furthermore, this application did not include any complaint under Articles 10 or 11 of the Convention and the Court notes the first-instance judge's finding that there had been no interference with the Article 10 and 11 rights of freedom of expression and assembly of those contained within the cordon (see paragraph 32 above). It must be underlined that measures of crowd control should not be used by the national authorities directly or indirectly to stifle or discourage protest, given the fundamental importance of freedom of expression and assembly in all democratic societies. Had it not remained necessary for the police to impose and maintain the cordon in order to prevent serious injury or damage, the "type" of the measure would have been different, and its coercive and restrictive nature might have been sufficient to bring it within Article 5.
69. In conclusion, since Article 5 is inapplicable, there has been no violation of that provision in this case.
For these reasons, the court
1. Declares the applications admissible unanimously;
2. Holds by fourteen votes to three that there has been no violation of Article 5 of the Convention.
Done in English and in French, and delivered at a public hearing in the Human Rights Building, Strasbourg, on 15 March 2012.
 Michael O'Boyle (Registrar), Françoise Tulkens (President)
In accordance with Article 45 § 2 of the Convention and Rule 74 § 2 of the Rules of Court, the following separate opinion of Judges Tulkens, Spielmann and Garlicki is annexed to this judgment.
F.T., M.O'B.
 
Joint dissenting Opinion of Judges Tulkens, Spielmann and Garlicki
(Translation)
1. We do not share the view of the majority that there was no deprivation of liberty in the present case, a finding which led them to conclude that, since Article 5 of the Convention was inapplicable, there had been no violation of that provision.
2. The judgment explains its position by "the specific and exceptional facts of this case", while also pointing out that "[h]ad it not remained necessary for the police to impose and maintain the cordon in order to prevent serious injury or damage, the 'type' of the measure would have been different, and its coercive and restrictive nature might have been sufficient to bring it within Article 5" (see paragraph 68).
3. In terms of the principles governing the application of Article 5 of the Convention, a provision which features in all the universal and regional human rights instruments and forms part of the European public order, the majority's position can be interpreted as implying that, if it is necessary to impose a coercive and restrictive measure for a legitimate public-interest purpose, the measure does not amount to a deprivation of liberty. This is a new proposition which is eminently questionable and objectionable for two reasons.
4. Firstly, the Court has always held that the aim or intention of a measure cannot be taken into account in assessing whether there has been a deprivation of liberty. These aspects are relevant only in assessing whether the deprivation of liberty was justified for the purposes listed in sub-paragraphs (a) to (f) of Article 5 § 1. In other words, the wording of Article 5 in itself strikes the fair balance inherent in the Convention between the public interest and the individual right to liberty by expressly limiting the purposes which a deprivation of liberty may legitimately pursue.
5. Next, regard being had to the structure and wording of Article 5 § 1 of the Convention, there can be no distinction in principle between measures taken on public-order grounds and measures imposed for any other legitimate/public-interest purpose. In other words, there is no reason to treat deprivations of liberty resulting from public-order considerations any differently from other kinds of deprivation of liberty for which this provision is invoked. Otherwise, States would be able to "circumvent" the guarantees laid down in Article 5 and detain people for a whole range of reasons going beyond the provisions of Article 5 § 1 (a) to (f), as long as they could show that the measure was necessary.
6. This was the approach underlying the Court's analysis in the A. and Others v. the United Kingdom judgment of 19 February 2009 ([GC], no. 3455/05, ECHR 2009), concerning a situation that was surely even more serious, namely a potential threat to national security: "The Court does not accept the Government's argument that Article 5 § 1 permits a balance to be struck between the individual's right to liberty and the State's interest in protecting its population from terrorist threat. This argument is inconsistent not only with the Court's jurisprudence under sub-paragraph (f) but also with the principle that paragraphs (a) to (f) amount to an exhaustive list of exceptions and that only a narrow interpretation of these exceptions is compatible with the aims of Article 5. If detention does not fit within the confines of the paragraphs as interpreted by the Court, it cannot be made to fit by an appeal to the need to balance the interests of the State against those of the detainee" (§ 171).
7. In this connection, the suggestion by the majority which unfortunately features in the part of the present judgment setting out general principles is problematic in our view: "It cannot be excluded that the use of containment and crowd-control techniques could, in particular circumstances, give rise to an unjustified deprivation of liberty in breach of Article 5 § 1. In each case, Article 5 § 1 must be interpreted in a manner which takes into account the specific context in which the techniques are deployed, as well as the responsibilities of the police to fulfil their duties of maintaining order and protecting the public, as they are required to do under both national and Convention law" (see paragraph 60 of the judgment). The wording of this statement appears dangerous to us in that it leaves the way open for carte blanche and sends out a bad message to police authorities.
8. The majority point out that "in certain well-defined circumstances, Articles 2 and 3 may imply positive obligations on the authorities to take preventive operational measures to protect individuals at risk of serious harm" (see paragraph 55 of the judgment). That may be so, but it has not been established in the present case that there was a clear and present danger to life or limb. In any event, the problem arising is not in fact a new one. The interaction between the protection afforded by Article 5 § 1 of the Convention and the positive obligations under Articles 2 and 3 has been examined on several occasions by the Court, which has consistently reiterated that positive obligations of this kind should be fully compatible with the guarantees set forth in Article 5. Only recently, in the Jendrowiak v. Germany judgment (no. 30060/04, 14 April 2011), the Court observed that "the State authorities could not, in the present case, rely on their positive obligations under the Convention in order to justify the applicant's deprivation of liberty which, as has been shown above (...), did not fall within any of the exhaustively listed permissible grounds for a deprivation of liberty under sub-paragraphs (a) to (f) of Article 5 § 1. That provision can thus be said to contain all grounds on which a person may be deprived of his liberty in the public interest, including the interest in protecting the public from crime" (§ 38).
9. In the present case, the paradox lies in the fact that, as rightly noted by Lords Hope and Neuberger, if there had been a deprivation of liberty, it would not have been possible for the police to justify it under the exceptions provided for in Article 5 § 1, sub-paragraphs (b) and (c).
10. We are well aware that maintaining order is a difficult task, although in the present case it was not disputed that the 6,000 police officers deployed were the most experienced in England. As the domestic courts' analysis indicates, it appears that the police prioritised effectiveness in their operation and opted for the most practical means of dealing with the situation by keeping everyone inside the cordon. This measure was thus applied indiscriminately and was also imposed against people taking no part in the demonstration. In this regard, the police could have been expected to apply less intrusive means. As it was, it seems that all people who happened to be at Oxford Circus at around 2 p.m. were treated like objects and were forced to remain there as long as the police had not solved other problems around the city.
11. The majority note that the applicants' circumstances differed in that the first applicant went to Oxford Circus to take part in the demonstration, whereas the other three applicants were passers-by. They consider, however, that this difference is not relevant to the question whether there was a deprivation of liberty (see paragraph 63 of the judgment). With all due respect, we do not agree. Admittedly, one can accept that active participants in a demonstration that is not entirely peaceful should be aware that their freedom of movement may be restricted because of the need for police measures, although that was not the case here. Indeed, the Court of Appeal overturned Mr Justice Tugendhat's finding that the police had had reasonable grounds to believe that the first applicant was about to commit a breach of the peace; on the contrary, it held that in containing her, the police had been exercising an exceptional common-law power whereby an innocent party could be detained in order to prevent a breach of the peace by others. Be that as it may, the situation was completely different regarding the other three applicants, who were at the scene by chance and had no intention of taking part in the demonstration. They could reasonably have expected that, by following police instructions, they would not be subjected to measures aimed at controlling a crowd of hostile demonstrators.
12. The Court considers itself unable to identify a moment when the measure of a restriction on freedom of movement changed to a deprivation of liberty (see paragraph 67 of the judgment). It is unclear what this observation means. Does it mean that there was no deprivation of liberty before 9.30 p.m. or that the situation became a deprivation of liberty between 2 and 9.30 p.m. but the precise moment cannot be pinpointed? In the latter event, the majority should not be able to conclude so categorically that those within the cordon cannot be said to have been deprived of their liberty within the meaning of Article 5 § 1 of the Convention. In a situation of uncertainty, the presumption is normally in favour of respect for individual rights.
13. Lastly, the Grand Chamber makes no reference whatsoever to the Gillan and Quinton v. the United Kingdom judgment of 12 January 2010 (no. 4158/05, ECHR 2010). Admittedly, the main focus of that case was Article 8 of the Convention, but Article 5 was also involved, precisely in the context of a demonstration. The interpretation of Article 5 in Gillan and Quinton was in fact much broader than in the present case since the Court found that a coercive restriction on freedom of movement amounted to deprivation of liberty within the meaning of Article 5 § 1: "The Court observes that although the length of time during which each applicant was stopped and searched did not in either case exceed thirty minutes, during this period the applicants were entirely deprived of any freedom of movement. They were obliged to remain where they were and submit to the search and if they had refused they would have been liable to arrest, detention at a police station and criminal charges. This element of coercion is indicative of a deprivation of liberty within the meaning of Article 5§ 1 (see, for example, Foka v. Turkey, no. 28940/95, §§ 74-79, 24 June 2008)" (§ 57, emphasis added). Yet the degree of coercion in the present case was much higher than in Gillan and Quinton.
14. In the present case, the applicants were confined within a relatively small area, together with some 3,000 other people, and their freedom of movement was greatly reduced; they were only able to stand up or sit on the ground and had no access to toilet facilities, food or water. The cordon was maintained through the presence of hundreds of riot police officers and the applicants were entirely dependent on the police officers' decisions as to when they could leave. Furthermore, the police could use force to keep the cordon in place, and refusal to comply with their instructions and restrictions was punishable by a prison sentence and could lead to arrest. All the applicants were contained in those conditions for six to seven hours.
15. In conclusion, we consider that there was a deprivation of liberty within the meaning of Article 5 of the Convention and that there has been a violation of that Article in the present case.