Hostname: page-component-848d4c4894-ttngx Total loading time: 0 Render date: 2024-05-21T06:45:10.833Z Has data issue: false hasContentIssue false


Published online by Cambridge University Press:  09 March 2023

Lewis Graham*
Law Society Fellow in Law, Wadham College, University of Oxford,
Rights & Permissions [Opens in a new window]


Article 5 of the European Convention on Human Rights enshrines the right to liberty, one of the oldest and most fundamental rights in the human rights tradition, and one of the core rights in the Convention. Central to the judicial understanding of Article 5 is the ‘exhaustive justification principle’: unlike with other rights, such as the right to privacy, interferences with liberty can only be justified by one of the specific reasons listed in Article 5 itself. This article shows that this rigidity has posed problems in practice: faced with modern developments unforeseeable at the time of the Convention's writing, such as the use of novel policing techniques and the COVID-19 pandemic, judges have interpreted Article 5 in an unusual and artificial way, sacrificing the exhaustive justification principle in doing so, in order to achieve sensible outcomes. The integrity of Article 5 has been threatened, with serious consequences for the future protection of the right to liberty. This trend is explained, evidenced and evaluated, and some (partial) solutions and concessions are considered.

Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (, which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright © The Author(s), 2023. Published by Cambridge University Press on behalf of the British Institute of International and Comparative Law


Article 5 of the European Convention on Human Rights (ECHR) enshrines the right to liberty. It begins with a simple guarantee:

Everyone has the right to liberty and security of person. No one shall be deprived of his liberty.

These two short sentences house one of the oldest and most fundamental rights in the civil liberties tradition, and one of the core rights in the Convention. The right to liberty has always been considered a particularly important right, even amongst those described as fundamental elsewhere in the Convention. It is, in the eyes of the Strasbourg Court, vital for the functioning of a ‘democratic society’Footnote 1 and has even been called a ‘first rank’ right alongside the right to life and freedom from torture.Footnote 2

Nonetheless, it is not an absolute right. The text of Article 5 includes six explicit exceptions—circumstances where an interference with the right to liberty can be considered justifiable in the eyes of the law. These are set out in a fairly particular and rigid framework, one which is not found elsewhere in the Convention. Crucially, these exceptions are exhaustive: an interference with the right to liberty can only be justified on the basis of one or more of the grounds found in that closed list.Footnote 3

This has its downsides. As this article will show, rigidity has led to problems in practice. In particular, strict adherence to the terms of Article 5 by judges would lead to some surprising outcomes, and preclude the deprivation of liberty in circumstances which judges are clearly very sympathetic to. It is to be contended that, when it comes to the application of Article 5 in practice, in a number of different cases, evidence can be found of judges ‘talking the talk’ but not ‘walking the walk’: that is, paying lip service to the fundamental nature of Article 5, characterised by the exhaustive nature of its exceptions, whilst adopting a more flexible approach which allows for preferred outcomes to be realised.

This article will analyse various cases which evidence this claim, taken from both the UK and Strasbourg, before evaluating this phenomenon as a whole and assessing its implications. First, however, some important background to Article 5 is provided, and the exhaustive justification principle, as well as other fundamental features key to Article 5, are set out.


Unsurprisingly, the right to liberty shares a number of features with other rights listed in the ECHR: the right must be interpreted in a ‘dynamic and evolutive’Footnote 4 manner, so as to provide effective, rather than ‘theoretical or illusory’Footnote 5 protection in practice. Any interference with the right must be properly prescribed by law,Footnote 6 which must satisfy certain conditions to meet the Convention standard of legality: it must, for example, be sufficiently precise and free from arbitrariness.Footnote 7 Further, any interference with the right by State authorities must pursue a legitimate aim, and the level of interference must be proportionate.Footnote 8 All of these are common features of qualified Convention rights across the board.

But the right to liberty is also distinctive. Article 5, as interpreted and developed by the relevant jurisprudence, features a set of fundamental principles with very limited relevance elsewhere in the Convention. Three of them are listed here. For the purposes of this article, and for the sake of convenience, they will be styled the exhaustive exception principle; the irrelevance of purpose principle, and the expansive definition principle. The first principle is the primary feature of Article 5, and the main focus of this article. The latter two overlap with, compliment, and safeguard the first. They are each intended to give practical effect to the unique character of the right to liberty.

A. The Core Principle: The Exhaustive Justification Principle

The exhaustive justification principle requires that any potential interference with liberty must be justified by one of the specific grounds listed in Article 5, and not any other. Article 5(1) confirms that no one shall be deprived of their liberty, ‘save in the following cases’, before setting out the following list of exceptions (further detail about each of these grounds is provided in subsequent sections of this article, where relevant):

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

  2. (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;

  3. (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;

  4. (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;

  5. (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;

  6. (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.

As noted above, courts have repeatedly confirmed that these grounds are exhaustive.Footnote 9 A State cannot seek to justify a deprivation of liberty by reference to a ground of justification which does not feature in that closed list, and the Strasbourg Court in particular has deprecated the efforts of State authorities to do so:

If a given instance of deprivation of liberty does not fit within the confines of one of the sub-paragraphs of that provision, 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 claimant].Footnote 10

As such, the Court has dismissed attempts by States to invoke justifications for deprivations of liberty based on the purported need to: protect the public from terrorist incidents;Footnote 11 manage national security threats;Footnote 12 effectively control national borders;Footnote 13 ensure the safety of the individual being detained;Footnote 14 or to prevent the occurrence of crime.Footnote 15 Even when the Court has been sympathetic to a State's concerns, they have traditionally refused to consider justifications that fall outside of the exhaustive list.Footnote 16

This is the core feature of Article 5, and one which distinguishes it from surrounding rights: given its importance, an interference with this right can only be justified under a very specific set of circumstances. Two more principles, or sub-principles, help to ensure that this central exhaustive justification principle is not undermined: first, the irrelevance of purpose principle; secondly, the expansive definition principle.

B. The Irrelevance of Purpose Principle

The irrelevance of purpose principle, straightforwardly enough, dictates that the purpose behind a deprivation of liberty—benevolent or otherwiseFootnote 17—is generally irrelevant when it comes to the question of whether there has been a deprivation of liberty in the first place. Whilst the court has listed considerations which can be positively considered when determining whether an act constitutes a deprivation of liberty—for example, its ‘type, duration, effects and manner of implementation’Footnote 18—it has at the same time specifically ringfenced ‘purpose’ as a separate consideration which can not be considered as part of this exercise. As the Court has stated, it ‘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’.Footnote 19 Rather, any consideration of the purpose behind imposing a deprivation is left for the court when it is considers whether that deprivation can be justified (under the exhaustive list of grounds under Article 5(1)(a)–(f) listed above).Footnote 20 To do otherwise would, in the words of Lord Kerr in a domestic case, ‘conflate the object of the restraints with their true character’Footnote 21—the court must first assess whether a deprivation of liberty has occurred (‘their true character’), then determine whether that deprivation is justified, whereby the purpose for their imposition (‘the object of the restraints’) may become relevant, in that it may convince the court that a specific justification was invoked, or that the interference in question was proportionate in nature. In this way, the principle ensures that the exhaustive list of exceptions in Article 5(1)(a)–(f) remain closed; judges must not consider any purpose which does not comport with the closed list of justifications. The irrelevance of purpose principle therefore helps safeguard the exhaustive justification principle.Footnote 22

C. The Expansive Definition Principle

The next principle is the expansive definition principle. The text of Article 5 neither defines ‘liberty’ nor what constitutes a ‘deprivation’ of it. It is therefore up to judges to delimit its boundaries. In doing so, judges should bear in mind the rights found in Article 2 of Protocol 4 of the Convention, which safeguard a distinct right to freedom of movement. The acts caught by Article 2 of Protocol 4 are much wider than those covered by Article 5 (and interferences with freedom of movement are assessed by courts in a very different way);Footnote 23 judges must therefore be careful, when interpreting the right to liberty, not to expand it too far so as effectively to trespass on actions which are most appropriately governed by the right to freedom of movement.Footnote 24 Nevertheless, courts must adopt a conception of a ‘deprivation of liberty’ under Article 5 which gives practical effect to its purpose wherever it arises. A deprivation of liberty must therefore extend beyond traditional detention in a prison cell. Although this remains the ‘paradigm’ case,Footnote 25 a ‘deprivation of liberty’ has been interpreted so as to include more general restrictions, so long as they meet a certain severity threshold. Courts consider all relevant factors cumulatively,Footnote 26 with the result that whilst one particular factor, such as the existence of physical restraint, or the presence of coercion, can point towards a finding that a deprivation of liberty has occurred, no single factor will be decisive in this regard.Footnote 27

The result is that somewhat less severe restrictions on movement and freedom, such as being placed under house arrest or a court order restricting movement,Footnote 28 may in some instances constitute a deprivation of liberty. Further, given that the Strasbourg Court has confirmed that Article 5 can ‘apply to deprivation of liberty of a very short length’,Footnote 29 somewhat more pedestrian interferences such as detention at airport security checks,Footnote 30 and (possibly) the exercise of stop-and-search powers by the police,Footnote 31 may, depending on the facts at hand, be characterised as deprivations of liberty for the purpose of Article 5. The principle ensures that non-paradigm examples of deprivations of liberty are still scrutinised by the court, and that States cannot interfere with the liberty of a person without justification, even if they do not happen to be sitting in a prison cell. It also ensures that States cannot manoeuvre around the fact that a given deprivation of liberty is not justifiable under any of the grounds on the closed list of justifications, and would otherwise result in a breach of Article 5, by artificially defining an action as that which is does not constitute a ‘deprivation of liberty’ in the first place.

Another part of the expansive definition principle, long since recognised by the courts, is that the list of exceptions in Article 5(1)(a)–(f) ought to be given a careful and narrow interpretation;Footnote 32 this is so that creative interpretation of the permitted exceptions cannot be used to expand the number of exceptions recognised in practice.

D. The Three Fundamental Principles Brought Together

Looking at the three fundamental principles above, it can be seen how Article 5 operates (or should operate) in practice.Footnote 33 First, a court must determine whether an individual has been deprived of their liberty. When doing so, a liberal definition of the term is to be applied (the expansive definition principle), and the State's purpose for enforcing that deprivation are irrelevant at this stage (the irrelevance of purpose principle). This is to ensure that justificatory reasoning does not creep into the judicial evaluation at an improper point (the exhaustive justification principle). Supposing a deprivation of liberty is made out, the court must then determine, at a second stage, whether this deprivation was justified. The only grounds under which a State can do so are listed in Article 5(1)(a)–(f), which must also be interpreted narrowly (the exhaustive justification principle). There must be a sufficient degree of proportionality between the interference and the justification for it. The purpose behind the measure may be relevant here insofar as it can help determine whether a legitimate justification was pursued, and the proportionality of that measure.

As such, the three fundamental principles fit together to form a coherent vision as to how Article 5 should be interpreted and applied in practice. As noted above, however, the claim in this article is that, over time, these fundamental principles have been stretched, if not completely eroded, at least in novel contexts. In a number of cases, set out below, it is suggested that the straightforward application of Article 5, treating the list of exceptions as truly exhaustive, would have resulted in an unpopular, unworkable or otherwise untenable result. In such cases, for pragmatic and sometimes consequentialist reasons, courts have creatively moved around the constraints inherent in Article 5 in order to achieve a different outcome.

It is suggested that as the Convention continues to be applied in novel situations, far removed from those the drafters could have envisaged, Article 5 is beginning to show a degree of inflexibility and rigidity. Therefore, what could, and should, be done about this? In addition, the more fundamental question is whether Article 5 continues to be fit for purpose.

To do this, the remainder of this article examines a number of different circumstances where judges have bent the rules, or departed from the fundamental principles, underlying Article 5. The cases are taken from the jurisprudence of the European Court of Human Rights as well as that of UK courts. It is structured as follows. Sections III to V will look at three situations involving the police: kettling and crowd control; preventative detention; protective detention. Section VI turns to the altogether different context of armed conflict and examines how Article 5 has been significantly diluted in that setting. Section VII considers the contemporary issue of State responses to the COVID-19 pandemic, and how Article 5 has been interpreted artificially so as to allow States maximum room for action. Section VIII brings these threads together and evaluates the overall picture emerging. Section IX asks what, if anything, ought to be done about it, and how the current situation in light of real-world obstacles to potential reform should be considered.


It is useful to begin with the case of Austin. The facts are simple: in 2001, a 3,000-strong protest took place in central London; many attendees protested peacefully, but some engaged in what the domestic court described as ‘a deliberate attempt to create destruction in the capital’.Footnote 34 In response, the police set up a cordon around the group, ostensibly to prevent outbursts of violence and potential breaches of the peace. The crowd, peaceful protestors amongst them, was contained there, unable to leave, for some seven hours, without access to food, water, shelter or toilet facilities. One of the protestors alleged, relying on the three features of Article 5 outlined above, that her right to liberty had been infringed.

The claimant argued that, considering the expansive definition principle, to be contained in a small space against her will, without being free to leave for any reason, constituted a deprivation of liberty in practice. Considering the exhaustive justification principle, it was argued that since none of the exceptions in Article 5(1)(a)–(f) applied, this deprivation could not be justified. Considering the irrelevance of purpose principle, it was argued that whilst in abstract the police's aims and purpose—preventing damage and ensuring public safety—may be understandable, they ought to play no part in the legal determination of whether Article 5 was breached. This set up a straightforward-enough case; indeed, as Lord Neuberger noted, ‘the notion that there has been no infringement of article 5 seems, at least on the face of it, surprising’.Footnote 35 Yet this is exactly what both the domestic courtsFootnote 36 and the European Court of Human RightsFootnote 37 did: both held that there was no deprivation of liberty under Article 5(1).

The House of Lords ruled 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’,Footnote 38 and found that the legitimate purpose pursued by the police allowed for what might have been a deprivation of liberty to be downgraded to a less severe restriction on liberty.Footnote 39

The European Court of Human Rights, by a majority, also found no breach, by treating the purpose of the cordon as a relevant factor in determining the ‘type, duration, effects and manner of implementation’ of the interference.Footnote 40 Whilst there is a slight shift in emphasis between the judgments of the respective courts, both were insistent upon the fact that the issue could be resolved at the first stage—whether a deprivation occurred—rather than at the justification stage, and that no deprivation occurred on the facts.

Clearly, both judgments were designed to achieve what the judges considered to be a fair result, and Article 5 was interpreted in a manner necessary to achieve this.Footnote 41 However, as other commentators have noted, to do so, the courts ‘smuggled’ purpose into the framework, despite alleging to do no such thing.Footnote 42 Indeed, following Austin, domestic judges have interpreted it as authority for the more general proposition that ‘the purpose and intention of the person [interfering with someone's liberty] may be relevant to whether there is a breach of Article 5’.Footnote 43 This approach is clearly out of step with orthodoxy.Footnote 44

This importation of purpose into the Article 5 framework can be framed in at least two different ways, and both run counter to the fundamental principles described above. The first is that the purpose behind the interference may in some circumstances become relevant to the initial question regarding the scope of Article 5, ie whether there was a deprivation in the first place (which would go against the irrelevance of purpose principle).Footnote 45 In fact, in a recent judgment Lady Arden suggested that Article 5 can be construed, albeit ‘exceptionally’, to accommodate considerations of proportionality within the question of whether a deprivation of liberty has occurred.Footnote 46 The second framing suggests that the purpose behind the interference can become relevant at the justification stage, even if it that purpose does not align with one of the otherwise exhaustive list of proscribed grounds listed in the Article (which would go against both the irrelevance of purpose principle and the exhaustive justification principle).Footnote 47 It is unsurprising, then, that Austin has been described as creating de facto general ‘purpose’ justification.Footnote 48

On either interpretation, Austin provides a clear example of a case where, upon an orthodox application of Article 5, judges must surely have found that an unjustified deprivation of liberty had taken place. That being intolerable, the courts interpreted Article 5 in a different way, going against its fundamental principles in order to do so. Whether this is defensible will be returned to below. However, as will be shown, Austin is not the only time that judges have bent the rules.


The next example concerns preventive detention (that is, detaining an individual for the purpose of preventing them from engaging in some future incident). There is no doubt that this kind of detention, involving being held in a prison cell, constitutes a deprivation of liberty. The issue for judges is, rather, whether the detention is justified. Article 5(1)(c) allows for ‘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’. So, can preventive detention be justified under the Convention?

Traditionally, the answer was no. The relevant jurisprudence firmly established that Article 5(1)(c) only facilitates detention for preventive purposes if it is done, at least in part, to bring the detainee before a judge, and for the purposes of determining whether they committed a concrete and specific offence. In the foundational case of Guzzardi v Italy, the Strasbourg Court said that Article 5(1)(c) was ‘not adapted to a policy of general prevention directed against an individual or a category of individuals who … present a danger on account of their continuing propensity to crime’.Footnote 49 In later cases, the Court described as ‘long … established’ the principle that Article 5 could not be used to justify preventive detention in the abstractFootnote 50 and that ‘preventive detention without charge to be incompatible with the fundamental right to liberty under Article 5(1)’.Footnote 51

Three recent cases signalled a change of approach. In 2014, a Chamber of the European Court handed down its judgment in Ostendorf v Germany.Footnote 52 In that case, the German authorities detained a number of football supporters with a history of violence and hooliganism, on the back of knowledge that further violence was planned at a particular football match. The majority of the court reiterated the orthodox position: Article 5(1)(c) could not, on its own, justify detention, at least where no offence was committed and there was no intent of bringing the detained before a judge.Footnote 53 A minority of judges, however, suggested that Article 5(1)(c) should be adapted to cover this kind of situation, even in the absence of a criminal charge.Footnote 54 They argued that its previous rulings, allowing detention only to bring an individual before a judge, should be jettisoned, and that Article 5 should now be interpreted so as to allow the detention of an individual where it is considered necessary to prevent the commission of an offence in the abstract.

The second case is Hicks v Commissioner of Police of the Metropolis,Footnote 55 a 2017 decision of the UK Supreme Court. In that case, a number of individuals considered likely to breach the peace at the wedding of a royal couple were detained, and then released once the event was over. In considering the lawfulness of that detention, the judges sided decisively with the minority in Ostendorf: Article 5 should be interpreted so as to permit detention of those who may imminently commit a crime (such as a breach of the peace), even if there is no intention, at the time of detention, to charge the detained with an offence or bring them before a judge.Footnote 56

The third, and most important, case was delivered by the Grand Chamber of the Strasbourg Court in 2019.Footnote 57 S v Denmark was another case involving the detention of individuals considered likely to participate in football hooliganism; as with Hicks, and notwithstanding the wealth of established jurisprudence to the contrary, the Grand Chamber ruled that ‘the fact that an arrested person was neither charged nor brought before a judge does not necessarily mean that the purpose of his or her detention was not in accordance with Article 5(1)(c)’.Footnote 58 The absence of any ‘concrete and specific [criminal] offence’, something integral to the finding of a breach of Article 5 in previous detention cases,Footnote 59 is, following S v Denmark, seemingly no longer requiredFootnote 60 and ‘ought not to constitute an obstacle to short-term detention’.Footnote 61 Rather ‘the second limb … can be seen as a distinct ground for deprivation of liberty, independently of the first limb’.Footnote 62 The prevention of the commission of an offence is, according to the Court, now an independent ground justifying detention in and of itself.Footnote 63

Whilst some have supported this new view on the basis of the language of Article 5,Footnote 64 it is, with respect, unconvincing that the approach adopted in S v Denmark is any more appropriate a reading of the text of that provision as the one adopted by the Court in cases like Ostendorf. As noted, the change upends a previously settled line of case law (which, when applied, frequently led to violations, and States paying compensation for causing those violations).Footnote 65

The new approach risks offending the exhaustive justification principle; the end result is that a ground which was previously not recognised as a valid reason justifying detention (prevention of harm, violence, or a breach of the peace) has now been positively recognised as a distinct ground in and of itself—ostensibly a new amendment to the otherwise exhaustive list in Article 5(1)(a)–(f). It also arguably offends the expansive definition principle; even if there was real ambiguity in terms of the meaning of Article 5(1)(c), the expansive definition principle should compel the adoption of a position which gives the greater protection to liberty rights (and, as such, the position which requires a more exacting justification).

Preventative detention of the kind seen in Hicks and S v Denmark presents another situation where being able to detain an individual seems, at least to some judges, like an eminently sensible idea, and one which, under a more open-ended proportionality exercise, would surely carry great weight. It is just not one which easily fits with the exhaustive nature of exceptions under Article 5(1).


As with detention for the purpose of preventing crime and, in effect, likely harm to others, detention solely for the detainee's own protection is not an explicit ground capable of justifying a deprivation of liberty under Article 5. For the avoidance of doubt, the Grand Chamber has repeatedly confirmed that, on its own, Article 5 cannot be used to effect so-called ‘protective detention’: that is, detaining someone for their own safety.Footnote 66

It was perhaps surprising, then, that in 2021 the Court of Appeal ruled, in Archer v Commissioner of Police of the Metropolis,Footnote 67 that detention considered necessary for the detainee's own protection did not breach Article 5.

In that case, the claimant was first (lawfully) detained pursuant to an assault charge, where it became clear that he himself had been the victim of a recent violent attack. The police—reasonably—believed that he would be at risk of a future attack if he were released from detention. His application for bail was therefore refused, on the grounds that continued detention was necessary for his own protection. He challenged this on the straightforward basis that this kind of detention could not be justified under Article 5(1); as the court explained:

The primary position for the appellant is simple: the appellant's detention was effected for the purpose of his own protection. Detention for that purpose is not permitted under Article 5(1)(c).Footnote 68

The Court of Appeal decided the case after the cases of Hicks and S v Denmark, described above, were decided; it therefore followed these authorities in recognising that Article 5(1)(c) involved three distinct grounds of justification, including a stand-alone ground relating to the prevention of imminent crime. However, even this more expansive interpretation of Article 5 could not, on the face of it, justify protective detention for the detainee's own sake.

Nevertheless, no breach was found in Archer's case. On the facts, the claimant's detention was justified for two reasons: first, on the basis that it was necessary for his own protection; secondly, because of the continued suspicion that he had committed a criminal offence. The detention was approved by the authorities in order to effect both purposes.Footnote 69 It seems that, because the latter reason was, in and of itself, a ground which could justify detention under Article 5(1)Footnote 70 (and the other requirements necessary to comply with Article 5 were satisfied), the detention was considered lawful. Seemingly, in this case, the ‘own protection’ ground was not truly relevant to the outcome, and may be something of a red herring: the court had already found another, alternative, ground under which Archer's detention could be justified.

Yet the court was not only being asked whether the specific instance of detention in this case was lawful, but also whether the provision under which that detention was authorised—Section 38(1)(a) of the Police and Criminal Evidence Act 1984—which on the face of it permits detention for a detainee's own protection as a stand-alone ground, was itself compatible with the Convention.Footnote 71 The Court of Appeal seemingly upheld these provisions as being compatible with Article 5 (or at the very least, did not find them incompatible with Article 5).Footnote 72 The fact the detention was compliant with Article 5 led naturally—in the Court's view inevitably—to the conclusion that the provisions which were invoked to authorise the detention must also be compliant with Article 5.Footnote 73 However, Archer's detention was not justified for reasons of his own protection on the facts: it was justified because of the suspicion that he had committed an offence. Indeed, if Archer's detention could not be justified on the basis of ‘own protection’ alone, the Court should not have upheld a provision which authorises detention on the basis of ‘own protection’ alone. But this is exactly what the Court did.Footnote 74

The decision once again illustrates a widening of the orthodox defences justifying a deprivation of liberty; if the S v Denmark interpretation of Article 5 is hard to reconcile with its text, the Archer interpretation is even less credible. The decision goes against the irrelevance of purpose principle; the (laudable enough) purpose of protecting an individual from harm is used to justify detention. Given that this justification clearly falls outside the list of proscribed grounds in Article 5(1), this also offends the core exhaustive justification principle, too.

It is, of course, easy to sympathise with the idea that bail might justifiably be refused, for a short while, in light of concrete and reliable information about the risk to the detainee's life. Such a decision, on the face of it, seems practical and justifiable in the abstract. However, judges are not tasked with answering abstract questions. Archer provides another example where judges, sympathetic to the reasonable, and even benevolent, aims of the authorities, have departed from the requirements of Article 5 in order to achieve a just and equitable result.


Sections III to V considered the application of Article 5 in the context of police powers and public order. This section turns to the application of Article 5 in a very different context: armed conflict. More specifically, it considers how the narrow approach to detention in the ECHR overlaps with the wider approach in humanitarian law, and the controversial question of which takes precedence in certain kinds of armed conflict. The analysis here will be somewhat brief, as the issue has been covered, in far more detail, and on the basis of far greater expertise, elsewhere.Footnote 75

The issue is as follows: when a party to the ECHR assumes personal authority over an individual in another jurisdiction,Footnote 76 or where it exercises ‘effective control’ of another territory,Footnote 77 it may (albeit only ‘exceptionally’)Footnote 78 be required to comply with the requirements of the Convention in that context. The UK was considered by the Strasbourg Court to have exercised jurisdiction, at least some of the time, over parts of Iraq during its occupation in the years following the 2003 invasion.Footnote 79 The ECHR, including Article 5, therefore applied. Yet, in the same circumstances, the Geneva Convention also applied, allowing the detention of certain individuals under circumstances beyond those listed in Article 5(1)(a)–(f).Footnote 80 Potentially, certain conduct by the UK in Iraq may have been considered unlawful under the European Convention, whilst justified and lawful under the Geneva Convention.

Initially, when faced with this prospect, both Strasbourg and UK courts were able to sidestep the potential for conflict altogether, often by emphasising that a challenge based on Article 5 of the ECHR must be determined according to the terms of Article 5 of the ECHR, regardless of the position under the Geneva Convention or any other international law instrument.Footnote 81 However, the Strasbourg Court seemed to grasp the nettle in Hassan v UK,Footnote 82 at least with regard to international humanitarian law.Footnote 83

In that case, the Strasbourg Court considered whether the UK forces’ detention of prisoners of war in Iraq on security-related grounds, something demonstrably justifiable under the terms of the Geneva Convention, ran contrary to Article 5 of the ECHR. Of course, no express ground justifying this kind of detention appears in Article 5. Nevertheless, in a landmark judgment, the Grand Chamber ruled that ‘the context and the provisions of international humanitarian law’ should be considered when interpreting Article 5 as a whole, including the list of justifications in Article 5(1).Footnote 84 Essentially, the right to liberty was to be interpreted so as expressly to allow, in addition to the express grounds for justification listed therein, the detention of prisoners of war on security grounds, in circumstances which satisfy the requirements of the Geneva Convention.Footnote 85

It is easy to see why the Grand Chamber ruled as it did. In the context of international humanitarian law, there is a strong argument that the rules of the Geneva Convention are better-suited than those of the ECHR. Certainly, those rules have applied in this context for a longer period of time. There are a whole host of practical problems which may arise if the ECHR were seen to ‘trump’ humanitarian law in practice, especially given that the Convention's extra-territorial application was, and continued to be, beset by controversy.Footnote 86 Nevertheless, the conclusion in Hassan represents a clear expansion of the apparently exhaustive grounds in Article 5(1). In practice, the Grand Chamber recognised a seventh groundFootnote 87 capable of justifying detention under Article 5(1). This offends the exclusive justification principle as well as the irrelevance of purpose principle, and by narrowing the circumstances under which a deprivation of liberty would be precluded, the expansive definition principle too.

But whilst the Hassan ruling represented a ‘striking modification’Footnote 88 of the orthodox position, the stretching of Article 5 in the context of armed conflict did not end there. Despite the Grand Chamber making it explicit that this new, wider defence was justified only on the basis of the specific importance of the Geneva Convention, and as such would be applicable ‘only in cases of international armed conflict’,Footnote 89 the UK Supreme Court subsequently ruled, in the case of Al-Waheed v Ministry of Defence,Footnote 90 that further defences to liberty interferences could be recognised in at least some further conflict-related contexts, including non-international armed conflicts where the Geneva Convention did not apply.Footnote 91 Considering the reasoning in Hassan, Lord Wilson saw ‘no reason to afford any less interpretative significance to the resolutions of the Security Council than to the Geneva Conventions’Footnote 92 which led the majority of the Court to find that the justifications in relevant United Nations Security Council Resolutions provided, in effect, further grounds for justifying detention, above and beyond those in the text of Article 5, and beyond those in the Geneva Convention as recognised in Hassan.Footnote 93 Indeed, that case was interpreted as authority for the strikingly wide principle that the previously exhaustive list of justifications in Article 5 did not apply outside of ‘peacetime conditions’;Footnote 94 the Supreme Court ruled that, during armed conflict, detention for ‘imperative reasons of security’ should also be permitted.Footnote 95 Whilst the majority's conclusion was met with a strong dissent from Lord Reed, arguing that the relevant Security Council Resolutions did not authorise the kind of detention implemented in that case,Footnote 96 each of the nine judges in the UK Supreme Court handing down the case seemed to agree that an extension of the grounds justifying detention in Article 5(1) in the context of armed conflict was, in and of itself, permissible.

Once again, judges applied Article 5 in a manner which clearly goes against the exclusive justification rule and the expansive definition rule; despite decades of case law to the contrary, the list of justifications under Article 5(1) are now, according to Lord Sumption in Al-Waheed, now only ‘ostensibly exhaustive’Footnote 97 rather than actually exhaustive, in the context of armed conflict.


Perhaps the most pertinent challenge to the continued suitability of the Article 5 framework in recent times, and the last case study in this article, concerns the impact of, and response to, the COVID-19 pandemic. This section considers the measures put in place across parts of the UK at different points during the first two years following the outbreak of the pandemic; the analysis is likely to be applicable to similar measures imposed in other parts of Europe, too. The kind of measures adopted were practically unprecedented in modern times: partial or wholesale lockdowns, stay-at-home orders and mandatory isolation periods. Whilst these measures, by and large, garnered the support of the majority of the general public,Footnote 98 they present, in theory, a significant curtailment of traditional freedoms, and may be more difficult to square with Article 5 than has perhaps been presumed.

On the face of it, there is a strong case that lockdowns (meaning here a general requirement to remain at home, sometimes with specific and usually very limited exceptions for purposes or classes of person),Footnote 99 constitute ‘deprivations of liberty’ in the sense necessary to engage Article 5.Footnote 100 Factors in the case law which have pointed towards a deprivation of liberty, which are of potential application in the COVID context, include: being required to live in a specific location or restricted area;Footnote 101 having social opportunities limited or meaningful contact with the ‘outside world’ removed;Footnote 102 needing permission or specific reasons to leave the location;Footnote 103 being unable to visit others;Footnote 104 and having visits from others limited or prohibited.Footnote 105 All of these are of potential application to some of the UK's lockdown measures, which often required the general public to remain in one place—typically their own home—for a significant period of time; civil and criminal penalties for non-compliance could be, and were, enforced. Whilst there were exceptions to even the most stringent lockdown rules, allowing, for example, for the purposes of essential shopping, or for taking one short period of exercise, these were very limited in nature and their existence does not necessarily foreclose the possibility of a deprivation of liberty. Courts have found that being ‘free to leave’ a site of detention means being free to leave for reasons of one's own choosing, rather than a specific purpose chosen by another.Footnote 106

Further, courts have been more likely to find that a measure constitutes an interference with Article 5 on the face of it where it is set out in terms which are general in nature and applicable across the board, without the possibility of sufficient adaption and alteration to suit the needs of a particular individual's circumstances.Footnote 107 Again, some of the lockdown measures in the UK could be said to fall afoul of this. There is, then, a good argument that the UK's lockdown restrictions constituted a deprivation of liberty, which needed to be justified in order to comply with Article 5.Footnote 108

Why, then, when the European Court of Human Rights had an opportunity to consider the compliance of Romania's lockdown rules—very similar in nature to those applicable in the UK—did it so blithely decide that lockdowns of this kind did not engage Article 5 at all?Footnote 109 It is suggested that in doing so, the Court fell into a similar error of reasoning as it did when it decided Austin, above: the fact that the measures seemed so obviously and compellingly justifiable in the abstract, albeit not necessarily justifiable in a sense that would fit easily with the terms of Article 5(1)(a)–(f),Footnote 110 probably fed into the prior question of whether a deprivation of liberty arose in the first place.

This is evident from the language the Court employed in its (brief) reasoning, when declaring the case inadmissible due to being manifestly ill-founded: the ‘specific context’ of the COVID pandemic was key to its decision,Footnote 111 and especially the pressing ‘health reasons’ behind the imposition of lockdown.Footnote 112 The Court spoke candidly about the laudable ‘aim’ of the measures adopted,Footnote 113 and expressed clear sympathy with the plight facing the authorities. The Court, against the irrelevance of purpose principle, seemingly allowed its agreement with the benevolent purpose behind the measures to justify a finding that the measures ought not be considered a deprivation in the first place. It is hard to see how positioning lockdowns outside the ambit of Article 5 sits comfortably with the expansive definition principle, either. The ultimate result, much like in Austin, was a de facto justification being imported into the definition stage, bypassing the need for the measure to fall within one of Article 5(1)(a)–(f). It goes without saying that this undermines the exhaustive definition principle.

The Strasbourg Court was not the only court to adopt such an artificial view. The Court of Appeal adopted a similarly reticent attitude towards a similar claim made against the UK lockdown regime in Dolan; it being an apparent ‘mischaracterisation to refer to what happened under the regulations as amounting in effect to house arrest or even a curfew’,Footnote 114 the argument that there had been a deprivation of liberty was considered ‘unarguable’.Footnote 115

It is very easy to sympathise with the approach taken here. The COVID-19 pandemic was sudden, unexpected and escalated quickly. The need for some kind of lockdown measures was considered inevitable by much of the scientific community at the time.Footnote 116 It is obvious, whatever else may be thought of lockdown measures across Europe, that they contributed to a significant mitigation against widespread loss of life, especially of the most vulnerable, during the early stages of the pandemic.Footnote 117 To declare them unlawful, on the basis that provision for responding to a global pandemic was unthinkable at the time Article 5 was drafted, would be a very severe step.

All the same, a finding that lockdowns do not deprive individuals of their liberty feels contrary to common sense. Surely, lockdowns deprive those affected by them of their liberty (at least in a legal sense, for reasons set out above); the pertinent question ought to be whether that deprivation is justified, given the (exceptional) circumstances. After all, a finding that an action constitutes a deprivation of liberty is only the first step in finding a violation of Article 5, and such a finding would not mean that lockdowns were unlawful or ought to be struck down; the measures could, in contrast to the situation in Austin, surely be justified at the second stage, as a proportionate measure ‘for the prevention of the spreading of infectious diseases’ under Article 5(1)(e).

Perhaps not. As Greene has argued,Footnote 118 the scant case law on Article 5(1)(e)Footnote 119 suggests that deprivations under this limb are judged according to a fairly exacting standard. For example, detention under this limb must truly be a measure of last resort, ideally following consideration and trialling of less severe options.Footnote 120 Perhaps most importantly, detention may need to be justified according to the specific situation of individuals affected (suggesting that justification must be assessed with regard to an individual's circumstances, and that a blanket measure such as a general lockdown may therefore fail to meet this standard).Footnote 121 Passages in the case law in relation to other parts of Article 5(1)(e), such as the provision allowing detention of those with unsound mind, might even suggest that this provision may only be applicable to those who are actually infected with a disease (or that it would be considerably more difficult to justify the use of Article 5(1)(e) with respect to them compared to those who were positively identified as carriers of the disease).Footnote 122

At best, then, the applicability of Article 5(1)(e) to the kinds of blanket lockdowns implemented during the pandemic is questionable. The global outbreak of COVID-19, and the national lockdowns which followed since 2020 are far removed from the paradigm situations of deprivations of liberty which Article 5 was probably intended to govern. In particular, it is hard to imagine that when the court in Einhorn, attempting to set out general principles relating to Article 5(1)(e), would have had in mind the seemingly novel possibility that this limb would, in future, be relied upon to limit the spread of a deadly disease in the context of a global pandemic. Indeed, that Article 5 seems particularly ill-suited to such novel circumstances may be one reason why courts have been keen to dismiss claims against the lawfulness of lockdowns at the initial interference stage, rather than the more uncertain justification and proportionality stage, much like they did in the Austin cases.

Yet, the fact remains, that in order to do so, they adopted an unnaturally narrow view of what constitutes a deprivation of liberty (going against the expansive definition principle), relying, in practice, on the convincing purpose behind the deprivation (going against the exclusive justification principle and irrelevance of purpose principle). Once again, the interpretation and application of Article 5 is conducted in a necessarily artificial way in order to achieve what is considered a just result.Footnote 123


In each of the above examples, judges were confronted with actions which, on the face of it, would constitute a deprivation of liberty, but could not be easily justified under the Article 5 framework. On the basis that those actions ought not to be considered a breach of the right to liberty (heightened, perhaps, by the particular associations and baggage attached to the charge of not just acting unlawfully, but breaching someone's fundamental human rights), the framework is adapted so as to avoid the need to find a breach on the facts.

This is done in two main ways. The first is by redefining, enlarging or otherwise modifying the closed list of justifications in Article 5(1)(a)–(f) so as to accommodate the justification in question (as was the case with preventive and protective detention described in Sections IV–V and, perhaps most explicitly, incorporating further defences in the context of armed conflicts as described in Section VI). The second, wary that a deprivation could not easily be justified under Article 5(1)(a)–(f), is by artificially narrowing the definition of what constitutes a ‘deprivation of liberty’ in the first place, removing conduct from the ambit of Article 5 altogether (as was the case with police kettling described in Section III and responses to the COVID-19 pandemic described in Section VII).

The end result of both methods is the same—the orthodox approach to Article 5, with the exhaustive justification principle at its centre, is sidelined, watered down or outright undermined: Stark is correct in suggesting that the justifications listed in Article 5(1) are often treated, in practice, as if they permitted an all-purpose, general proportionality mechanism.Footnote 124

It should, of course, be noted that this article is not the first to draw attention to the problems with judicial reasoning in Article 5 cases. The result in Austin has been heavily criticised by academics: Ashworth and Zedner, for example, have described what they consider to be ‘result-pulled reasoning’Footnote 125 and Stone has charged the court with reasoning backwards from a desired result.Footnote 126 The argument here is that Austin provides just one example of this phenomenon. Indeed, the above list of cases is intended to be illustrative rather than exhaustive; there are certainly other accounts of judges departing from the Article 5 framework in order to achieve what they consider a just outcome. For example, Fenwick and Fenwick contend that in cases concerning control orders,Footnote 127 judges would ‘reinterpret’Footnote 128 and ‘recalibrat[e]’Footnote 129 Article 5 so as to ensure compatibility. Similar arguments have been made in relation to the police's stop-and-search powers,Footnote 130 and a further claim could even be made regarding detention in the mental health context.Footnote 131 Examples abound of judicial creativity producing dubious reasoning in the context of Article 5.

This is not a completely clandestine enterprise. Authorities on Article 5 constantly reiterate that it must be interpreted and applied in a ‘practical’ way which does not impose inappropriate burdens on the State.Footnote 132 It seems that this need can sometimes trump the fundamental principles behind the right. It has also been stated that the ‘underlying principle[s]’ behind Article 5 may be more important than complying with each specific limb and justification.Footnote 133 Yet, whatever underlying principles that the courts may have had in mind in the cases described above were, in fact, deployed in a manner which undermined the even more fundamental principles described in Section II. Generic recourse to the promotion of unspecified principles in this way is unsatisfying and self-defeating.

Perhaps the clearest acknowledgement of the reality of the modern approach to Article 5 comes from Lord Neuberger's concurring judgment in Austin. In a refreshing departure from the sometimes delphic reasoning employed in Article 5 cases, he said this:

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 … I consider that the fact that the restraint in the present case could not be justified under any of the exceptions … 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 to be open to the police to act as they did in the instant circumstances, without infringing the article 5 rights of those who were constrained.

In other words: it is precisely because the application of the usual approach to Article 5 would lead to an outcome—here, that the kettling in that case would fall foul of the Convention—which would be unpalatable (‘odd’), that the usual approach should be departed from (in this instance, ruling that the relevant actions ‘did not amount to detention within Article 5 at all’). Lord Neuberger's statement is a rare admission that when it comes to Article 5, at least in hard cases, it is pragmatism which rules.Footnote 134


The above sections have sought to illustrate the disconnect between the endorsement of the exhaustive justification principle in theory and the sidelining of this principle in practice. This section asks what ought to be made of this.

First, why is this so problematic? Perhaps the most obvious issue, at least to lawyers, is the idea of judges advancing beyond their role, answering the question of what ought to constitute a deprivation of liberty according to their own wisdom, rather than according to what Article 5 of the ECHR requires. The adoption of an interpretation of Article 5 which strays significantly from its text may be considered to trespass upon the intentions of its drafters and threaten the democratic legitimacy conferred by its adopters. Thus, judges can be charged with crossing the cardinal line between application of the law and creation of it, a breach of the all-important legislative boundary. Of course, proponents of ‘slot machine jurisprudence’—whereby judges are limited to merely declaring the law—are few and far between.Footnote 135 Most would allow for some kind of pragmatic, evolutive approach to the interpretation of legal sources. However, that does not necessarily bestow legitimacy to the approach adopted in the cases set out above.

The interpretation of Article 5 as described above is also based on achieving outcomes which are unlikely to attract universal moral assent. The use of ‘kettling’ as a form of protest response and crowd control has been strongly criticised.Footnote 136 The use of preventive and protective detention may have merit, but their recognition sits uneasily with the continued expansion of police powers (and, in particular, evidence of discriminatory use of such powers).Footnote 137 There are loud dissenting voices concerned about the effect of lockdowns.Footnote 138 Yet the Article 5 framework was modified to facilitate each of these. Noting these criticisms is not necessarily to endorse them, or to contend that, if accepted, they would outweigh the potential positives stemming from allowing a deprivation of liberty in each case. It is simply to note that even if it is accepted that judges must make and develop law to some extent, the kind of changes made in these cases, grounded in moral considerations, are bound to be questioned.

Another concern applies in particular to the situations outlined in Sections III and VII—those who are cordoned and kettled, those who are unable to leave their homes due to their health concerns and so on. Due to a judicial sleight of hand, such individuals are not receiving reduced protection, but are deemed to fall outside of the scope of Article 5 altogether, receiving no protection. States are not required to justify infringements at all. There is no requirement that the infringement be justified ‘by law’. The additional protections in Article 5(2) to 5(5) do not apply to them. Even if it is considered that kettling and lockdowns should not fall afoul of Article 5(1), it seems unfair that those subject to such measures should not enjoy the additional Convention safeguards which they would otherwise be entitled to. The situation does fare better, in theory, for those affected by rulings in Sections IV–VI, whose deprivations of liberty were considered justified rather than not engaging Article 5 in the first place, but the applicability of these further protections in their case remains uncertain and patchy.

Finally, there is also a concern with the kind of changes being made in the examples above—judges are modifying the law by loosening, and ultimately weakening, the strict provisions of Article 5 (either by artificially narrowing what constitutes a deprivation of liberty or by artificially widening what can justify it). The result is, as commentators have put it, a dilution of protections under that rightFootnote 139 and an ‘enlarged scope for coercive state action against individuals’Footnote 140 going forwards. Towards this, there should, of course, be caution.

In sum, the approach to Article 5 set out above raises legitimate concerns relating to the appropriate role of judges in human rights cases, the relevance of personal morals in the case law, the specific moral choices being infused into law, the lack of safeguards applicable to some, and the implications of these developments for future cases. So, what should be done about all of this? Certainly, it is plausible to argue that some, or indeed all, of the cases above were wrongly decided. Many judges, in dissenting opinions, have said as such.Footnote 141 Over time, cases could be revisited or over-ruled. However, the cases outlined in this article are neither exhaustive examples of Article 5 being interpreted in a manner which undermines its orthodox principles, nor is it very likely that this trend will cease and there will be seen a return to the standard approach. The genie is seemingly out of the bottle. Judges adopted the approach they did in each case willingly; their doing so was no accident or mere legal error spotted with hindsight—the decisions were made intentionally. Judges had concerns, so serious that they compelled at times a fairly radical reinterpretation of Article 5. It is unlikely that these concerns will simply go away.

Perhaps more useful, then, is considering whether some kind of reform could be pursued, so as to accommodate, at least to some extent, some of the concerns leading to the problems identified above.Footnote 142 Reform of the Convention could of course be pursued. Amending Article 5 to include a more general proportionality-based approach would fit easily alongside the approach taken in relation to Articles 8–11 and elsewhere. Indeed, Article 5 was originally going to include a very similar proportionality measure to those rights before a change in drafting at an early stage.Footnote 143 An alternative solution might be to amend or extend the grounds listed in Article 5(1)(a)–(f) so as to accommodate further specific justifications.Footnote 144 Both changes would be likely to result in a greater emphasis being placed on the need to ensure a fair balance between that aim and the measures adopted,Footnote 145 and would discourage judges from finding that measures which so clearly engage Article 5 on the face of it fall outside of its scope: there would be no need to do so if the justification for the measure could be subject to a more holistic assessment.

However attractive this may be from a legal point of view, the likelihood of this kind of reform taking place is virtually zero. It would be nearly impossible to mobilise support for an amendment to the Convention to modify the contours of Article 5.Footnote 146 There is surely very little appetite for drafting and supporting what is essentially a fairly technical amendment, put forward to achieve a result which the court has considered itself capable, however messily, of achieving anyway. As is so often the case, the neatest academic solutions to a problem fall away in the face of political reality.

If Article 5 itself cannot be amended easily, perhaps for answers, elsewhere in the Convention should be consulted. Article 5 of the ECHR is not totally absolute—unlike, say, the right to life under Article 2 of the ECHR, States can derogate from their Article 5 obligations under Article 15 of the ECHR.Footnote 147 When issues arise which, exceptionally, warrant depriving someone of the liberty in circumstances which do not fall under Article 5(1)(a)–(f), it may be fruitful to encourage States to derogate from the Convention, rather than asking courts to engineer an inauthentic interpretation of the Convention to justify such an action. In such a situation, exceptional cases are removed from the scope of the right altogether, mitigating the dilution of Article 5 as described above.Footnote 148

Indeed, derogations from various rights under the Convention, including Article 5, have been historically made in the context of armed conflict;Footnote 149 more recently, a raft of derogations were made by States during their initial response to the COVID crisis.Footnote 150 Both contexts raise issues for the orthodox approach to Article 5, as outlined above.

Whilst the necessity of derogation in different circumstances is of course worth considering,Footnote 151 the idea that derogations could be effectively deployed in such a way to avoid the problems outlined above is impractical and unrealistic. Derogations are rarely made in practice.Footnote 152 This is by design—there are fairly strict requirements which must be satisfied for derogations to be valid and effective: there must be a ‘public emergency threatening the life of the nation’, and any measures adopted must be ‘strictly required by the exigencies of the situation’;Footnote 153 this exacting threshold will not always be met, even in paradigm cases such as wartime and during a pandemic.Footnote 154 Certainly, as things stand, derogation for the purposes of the exercise of ordinary police powers, or for routine crowd control, would be very difficult, if not impossible, to achieve.Footnote 155 Expecting States to derogate in such a manner has rightly been described as ‘unreal’.Footnote 156

There are also downsides to an increased recourse to derogation: as Lord Wilson pointed out, doing so in the context of armed conflict ‘would leave the human rights of those caught up in the conflict far less protected’.Footnote 157 The same goes for each of the categories of people affected by the measures set out above. It may be considered better to retain the general protections of Article 5 and ‘tweak’ its application at the edges, in the cases described above, rather than to remove those protections altogether.Footnote 158

As such, it seems there remains a problem without a neat solution: the current approach, under which a very small number of justifications are permitted for interfering with Article 5, is being jettisoned in practice in order to achieve a desired result, raising clear issues for judicial legitimacy, rights protection standards and the fairness of the approach to potential detainees in concrete cases. However, when it comes to finding a better route, the cleanest option—amending Article 5 in some way—is all but impossible. Further, derogation is impractical, presents problems, and would in any case only ever apply some of the time (there is no possibility of derogation when it comes to the ordinary application of police powers, for example).

A. A Reconsideration: Diplomacy, Pragmatism, Morality

Perhaps, then, with no real alternative, the reality may have to be that judges will not always apply Article 5 in an orthodox manner; they will, consciously or otherwise, apply it in such a way so as to achieve certain outcomes, even if this involves some tinkering of that right around the edges. This may be regrettable. However, this is not the only perspective that ought to be considered: there may be, after all, some good reasons to employ bad legal reasoning.

For example, a pragmatic approach may contribute, at least in part, to a strategy of avoiding backlash, particularly from politicians, and especially from those national politicians primarily responsible for the maintenance and enforcement of human rights laws in practice.Footnote 159 Many State actors, including those in the UK, have been openly hostile to the ECHR and human rights in general, complaining about, amongst other things, the perceived expansion and ‘over-reach’ of rights concepts; the constraining effect on governance; and what is considered to be an unjustified privilege afforded to categories of people such as prisoners and immigrants.Footnote 160 Whether or not these claims are credible, they are often taken seriously. If a State considers that the ECHR prevents police forces from being able to protect the public effectively, or that it places some bureaucratic stranglehold on essential wartime operations, or that it hampers efforts to prevent, and mitigate the effects of, an unprecedented global pandemic, there would probably follow some severe political backlash, especially if that State or its relevant actors already hold significant concerns about the Convention and its application.Footnote 161 Political backlash can take many forms, from a fairly mild censuring by politicians to more vindictive efforts to reduce or remove the offending altogether.Footnote 162 A judgment which carefully avoids outcomes unpalatable to those hostile to the Convention, even if dubious legal reasoning must be employed in order to get there, may reduce the potential for some of the more extreme backlash and, in effect, contribute to the continuance of an effective human rights regime at a domestic level.Footnote 163 In this way, judges who favour the current framework may consider a small—relatively speaking—concession (Article 5 protections being denied to, eg targets of kettling) justifiable in order to ensure a greater level of protection overall (an absence of backlash resulting in less danger of Article 5 protections being reduced or removed across the board).Footnote 164

Politicians and legislatures are not the only potentially antagonistic actor when it comes to human rights judgments: judges of the European Court of Human Rights must also consider the position of national courts. Despite having a generally amiable relationship, clashes between the Strasbourg Court and domestic courts when it comes to the interpretation and application of Convention rights have occasionally arisen. The UK and Strasbourg judiciaries have sometimes clashed in this respect.Footnote 165 There are potentially good reasons for the Strasbourg Court to try and ensure as much co-operation between itself and the judiciaries of member States as possible, not least the pursuit of a uniform conception of rights across Europe.

When faced with a national judiciary which appears to be bending the rules of Article 5, a judge at the Strasbourg Court may ask themselves whether telling the domestic court that it is wrong—with all the legal and political consequences this entails—is really worth it. Indeed, some of the case law examples set out above evidence some kind of conflict avoidance. The decision in Austin v UK, removing kettling from the sphere of Article 5, was preceded by a robust judgment holding similarly from the House of Lords.Footnote 166 Various passages from the domestic case were replicated and endorsed in the Strasbourg decision. Similarly, the Strasbourg Court's most definitive ruling on protective detention drew very heavily from a decision of the UK Supreme Court which was very critical of the orthodox position.Footnote 167 In both cases, the traditional approach to Article 5 was considered, and rejected, in favour of the contrary position adopted by UK courts. Rightly or wrongly, avoiding a drawn-out conflict between a national court—especially one which exerts considerable influence and has a history of diverging from Strasbourg under certain circumstancesFootnote 168—would understandably seem attractive to the judges of the Strasbourg Court.Footnote 169 Strasbourg may want to pick its battles to some extent. The approach to the exceptions to Article 5 seems to be one battle it does not consider necessary to wage at this time.

So much for diplomacy and pragmatism. There is one remaining justification for the approach worth returning to: a simple moral one. Justifying deprivations of liberty in circumstances outside of the closed list in Article 5 may be pragmatic; it may be politically savvy; but it also might be a substantively good thing to do.

It may be thought that preventative detention may sometimes be necessary. Indeed, it may also help to ensure that other human rights obligations are secured (for example, States are under a positive obligation to prevent domestic violence from occurring in certain circumstances).Footnote 170 It may be believed, as many of course do, that lockdowns are both necessary and effective in combating the spread of COVID-19. It may therefore be agreed with the judges in some of the cases above that Article 5 ought to allow detention in certain circumstances outside of those proscribed in Article 5(1)(a)–(f). However, even if this is not agreed to, is there not some conceivable circumstance under which a deprivation of liberty may be considered justifiable for reasons outside of those specified? In Austin, Lord Neuberger put forward the hypothetical example of a public authority becoming aware of ‘a deranged or drunk person … on the loose with a gun in a building’;Footnote 171 under such circumstances, if the authority ordered the general public to hide in a confined, tight space, until the incident subsided, surely this is justifiable, even if it cannot be justified under the proscribed grounds of Article 5?

This hypothetical example may be too remote to be of great assistance, and it may be concluded that the core Article 5 protections should not be adjusted for cases at the very margins. After all, hard cases make bad law. But the point remains: if even some hypothetical example can be considered whereby a deprivation of liberty would be warranted in circumstances falling outside of the grounds in Article 5(1)(a)–(f), is there already therefore a moral concession to the judges deciding the cases above?

Are any of these considerations—diplomatic, pragmatic, moral—enough to make a convincing argument that the judges were right to depart from the orthodox approach, in any, or all, of the cases above? Or should judges close their eyes to these concerns and simply apply the law in its most straightforward iteration? In this article attention has been drawn to the importance of these questions and presented some competing considerations. The reader is now left to decide the answers for themselves.


In a speech in 2014 given to an audience in the USA, Lord Wilson opined:

Unwisely the drafters of the European Convention had limited the power to deprive someone of liberty to six specific situations … [t]he drafters would have done better to cast the power within more general limits, like the reference to ‘due process of law’ in your 5th and 14th Amendments.Footnote 172

This article has drawn attention to cases which may well evidence Lord Wilson's claim that the drafting of Article 5 was indeed ‘unwise’, evidenced by the fact that the fundamental principles which follow from this drafting, namely the expansive definition principle, exhaustive justification principle and irrelevance of purpose principle, if followed and applied consistently, would have led to results that judges simply could not countenance. Rather, elastic judicial reasoning has allowed them to sanction certain actions which would otherwise have fallen afoul of the Convention.

Perhaps Lord Wilson is right, and a more general limitation clause would have allowed for these concerns to be addressed in a different, more straightforward, manner. After all, exactly this kind of general limitation clause qualifies restrictions on freedom of movement under Article 2 of Protocol 4.Footnote 173 At this stage, however, reform seems a remote prospect; the ‘unusually inflexible’Footnote 174 framework of Article 5 is very likely to be here to stay.

Whilst the actions set out in this article—police kettling, preventive detention, protective detention, the taking of prisoners of war in certain types of armed conflict, and the myriad of lockdown measures imposed in response to COVID-19—arguably breach Article 5 as it has traditionally been understood, courts have, one way or another, held the opposite. There are some reasons to be concerned about this: such an approach risks undermining the values of certainty and predictability; the newly identified judicial defences lack democratic legitimacy and are unlikely to attract universal support on moral terms; and there is a real prospect of the door being opened to ‘rights dilution’, now or in the future. At the same time, a more flexible approach, even if it is somewhat artificial, may be considered necessary to avoid backlash and pushback from political bodies; may help foster positive relationships with other courts and judicial bodies; and may deliver a substantive outcome which falls in line with practical reality and good moral sense.

Judicial creativity is nothing new; neither is the charge of conducting results-based reasoning. Whether it is liked or not, there is a strong argument that judges will always engage in results-oriented reasoning to some extent. Within Article 5, as with all articles in the Convention, and indeed as with all laws requiring judicial interpretation, there is space—perhaps a limited space, but space nonetheless—for judges to manoeuvre their way towards preferred outcomes, especially in hard cases like those described above. Whether it is justified in all of the circumstances described in this article remains highly doubtful, even considering the potential diplomatic benefits. However, the personal element of judicial decision-making is difficult to counteract; when it comes to Article 5, at least in the most difficult cases, pragmatism wins the day. The nature of judicial reasoning, so long as it is conducted by human beings, may render this practically inevitable. Indeed, as has been shown, even the very precise, closed list of exceptions in Article 5 has been, in the end, no match for judicial ingenuity.


The author wishes to thank Shona Wilson Stark for her helpful comments on an earlier draft.


1 Medvedyev v France (2010) 51 EHRR 39, paras 76, 117.

2 Denis v Belgium (2022) 74 EHRR 8, para 123.

3 See eg Medvedyev (n 1) para 78.

4 Stafford v United Kingdom (2002) 35 EHRR 32, para 68.

5 Airey v Ireland (1979–80) 2 EHRR 305, para 24. For application in the Article 5 context, see Magee v United Kingdom (2001) 31 EHRR 35, para 90.

6 Manole and Postica v Moldova App No 4711/07 (ECtHR, 29 June 2021) paras 68, 70–72.

7 Mooren v Germany (2010) 50 EHRR 23, para 72; JN v United Kingdom App No 37289/12 (ECtHR, 19 May 2016) paras 76–80; Nasirov v Azerbaijan App No 58717/10 (ECtHR, 20 February 2020) para 47.

8 Unlike, for example, Articles 8–11, proportionality is not included in the text of Article 5. However, courts have infused a proportionality-style assessment into its case law: see, eg Beghal v DPP [2015] UKSC 49, para 119.

9 Medvedyev (n 1) para 78; Denis (n 2) para 124; Archer v Commissioner of Police of the Metropolis [2022] EWCA Civ 1662, para 71.

10 Merabishvili v Georgia (2017) 45 BHRC 1, para 298.

11 A v United Kingdom (2009) 49 EHRR 29, para 171.

12 Al Husin v Bosnia App No 3727/08 (ECtHR, 7 February 2012) para 64.

13 Baisuev and Anzorov v Georgia App No 39804/04 (ECtHR, 18 December 2012) para 60.

14 Khlaifia v Italy App No 16483/12 (ECtHR, Grand Chamber, 15 December 2016) para 71.

15 Jendrowiak v Germany (2015) 61 EHRR 31, para 38.

16 Medvedyev (n 1) para 81; Baisuev and Anzorov (n 13) paras 60–61.

17 P v Cheshire West and Chester Council [2014] UKSC 19, paras 28, 35, 82; Khlaifia (n 14) paras 58–59, 71.

18 Guzzardi v Italy (1981) 3 EHRR 333, para 92; Medvedyev (n 1) para 73. Applied in UK courts in Secretary of State for the Home Department v JJ [2007] UKHL 45, paras 16, 58; Secretary of State for the Home Department v AP [2010] UKSC 24, paras 1, 25; and see generally, Stark, S, ‘Deprivations of Liberty: Beyond the Paradigm’ (2019) PL 380, 396–8Google Scholar.

19 Austin v United Kingdom (2012) 55 EHRR 14, Dissenting Opinion of Judges Tulkens, Spielmann and Garlicki, para 4. The approach of the majority in Austin with regards to relevance of purpose is detailed below in ‘Kettling and Crowd Control’ (Section III).

20 See eg Rozhkov v Russia (No 2) App No 38898/04 (ECtHR, 31 January 2017) para 74.

21 P v Cheshire West (n 17) para 84.

22 A similar principle guides the interpretation and application of Article 3 of the ECHR, and in the context of that right judges have similarly repeated the mantra that consideration of the purpose behind a given act ought to be irrelevant to the question of whether that act breached Article 3. Yet, much like with Article 5, it is arguable that the purpose for which a measure is adopted does, in fact, play a role in determining whether Article 3 may be breached in practice: see Mavronicola, N, ‘Crime, Punishment and Article 3 ECHR: Puzzles and Prospects of Applying an Absolute Right in a Penal Context’ (2015) 15(4) HRLR 721Google Scholar and Graham, L, ‘Jeanty v Belgium: Saving Lives Provides (another) Exception to Article 3 ECHR’ (2021) 21(1) HRLR 221Google Scholar.

23 In particular, much like with the rights under Articles 8–11 of ECHR, interferences with freedom of movement can be justified, where proportionate, by one or more legitimate aims listed in the relevant Article. It is therefore much easier for a State to justify an interference with a right under Article 2 of Protocol 4 compared to Article 5.

24 A further reason for bearing in mind the distinction between the two rights is that a number of States, including the UK, have not ratified Article 2 of Protocol 4, whereas all States are required to respect Article 5.

25 Stark (n 18) 383–4.

26 De Tommaso v Italy (2017) 65 EHRR 19, para 80.

27 In P v Cheshire West (n 17) para 37, the UK Supreme Court (UKSC) suggested the following ‘acid test’ with regards to determining whether a deprivation of liberty has arisen: ‘the person is subject to continuous supervision and control and the person is not free to leave’. See also Lancashire County Council v G [2020] EWHC 2828 (Fam) para 41 and Barking and Dagenham LBC v A [2019] EWHC 2017 (Fam) para 36. The definition is not intended to be exhaustive and is most apt for cases concerning liberty in the hospital detention context.

28 See D Bonner, ‘Checking the Executive? Detention Without Trial, Control Orders, Due Process and Human Rights’ (2006) 12(1) EPL 45, 62–70. For examples, see Secretary of State for the Home Department v JJ (n 18) and Secretary of State for the Home Department v AP (n 18).

29 Manole and Postica (n 6) para 67. See also MA v Cyprus App No 41872/10 (ECtHR, 23 July 2013) para 190; Zelčs v Latvia App No 65367/16 (ECtHR, 20 February 2020) paras 36, 40; Foka v Turkey App No 28940/95 (ECtHR, 24 June 2008) para 74.

30 Kasparov v Russia (2018) 66 EHRR 21.

31 Gillan v United Kingdom (2010) 50 EHRR 45, para 57. Particularly if there is a significant degree of coercion involved: see Foka (n 29) paras 77–79.

32 Engel v the Netherlands (1979–80) 1 EHRR 647, para 57; Al-Jedda v United Kingdom (2011) 53 EHRR 23, para 99; Medvedyev (n 1) para 78; Secretary of State for the Home Department v JJ (n 18) para 5.

33 Whilst there are also a number of corollary rights in the text of Article 5 (such as the requirement that anyone arrested should be made aware of the reasons for this, that anyone arrested should be brought ‘promptly’ before a judge and that any deprivation should be open to challenge: see Article 5(2), 5(3) and 5(4) of the ECHR respectively), each of these rights are dependent on a deprivation of liberty being made out in the first place.

34 Austin v Commissioner of Police of the Metropolis [2009] UKHL 5, para 8.

35 ibid, para 51.

36 ibid and see also the decision of the Court of Appeal: [2007] EWCA Civ 989 EWCA. nb. at first instance, Tugendhat J did recognise that there had been a deprivation of liberty on the facts, albeit one justified under Article 5(1)(c): [2005] EWHC 480 (QB).

37 Austin (n 19).

38 Austin (n 34) para 34 (Lord Hope) cf Lord Walker's concerns, paras 43–44.

39 See, in particular, Austin (n 34) para 63 (Lord Neuberger), comparing cordons used for the prevention of crime with police tactics used for malicious or punitive purposes.

40 Austin (n 19) paras 58–59.

41 Stone, R, ‘Deprivation of Liberty: The Scope of Article 5 of the European Convention on Human Rights’ (2012) 1 EHRLR 46Google Scholar, 56.

42 Ashworth, A and Zedner, L, Preventive Justice (OUP 2014) 61CrossRefGoogle Scholar.

43 Commissioner of Police of the Metropolis v ZH [2013] EWCA Civ 69. See references to ‘good faith’ and ‘justification’ in cases such as Castle v Commissioner of Police of the Metropolis [2011] EWHC 2317 (Admin) para 69 and R (Moos) v Commissioner of Police of the Metropolis [2012] EWCA Civ 12, paras 57, 59, 60.

44 Commentators have noted that the approach falls out of line with cases handed down both before it (see D Mead, ‘Of Kettles, Cordons and Crowd Control – Austin v Commissioner of Police for the Metropolis and the Meaning of Deprivation of Liberty’ (2009) 3 EHRLR 376, 388–98) and after it (see Stark (n 18); see P v Cheshire West (n 17) and Welsh Ministers v PJ [2018] UKSC 66, para 22).

45 This seems to align with the approach of the European Court of Human Rights: see Austin (n 19) paras 59, 65–67. As Mead has noted, importing a consideration of purpose into the initial engagement question arguably makes the exhaustive list of proscribed justifications redundant, as justifications beyond those listed in Article 5(1)(a)–(f) can, in theory, be used to justify an interference in practice: Mead, ibid 393.

46 Re D [2019] UKSC 42, para 119 (Lady Arden).

47 This seems to align with the approach of the House of Lords: see Austin (n 34) para 27 (Lord Hope).

48 Stark (n 18) 382.

49 Guzzardi (n 18) para 102.

50 Al-Jedda (n 32) para 110; see, amongst many others, Al Husin (n 12) para 65; Ječius v Lithuania App No 34578/97 (ECtHR, 31 July 2000); Ciulla v Italy App No 11152/84 (ECtHR, 22 February 1989).

51 A v UK (n 11).

52 Ostendorf v Germany (2013) 34 BHRC 738.

53 ibid, para 82. A majority of the court found that Article 5(1)(b) could be used to justify the detention, in light of an imminent breach of the peace on the facts.

54 ibid, Dissenting Opinion of Judges Lemmens and Jäderblom.

55 R (Hicks) v Commissioner of Police of the Metropolis [2017] UKSC 9.

56 The UKSC suggested that the majority judgment in Ostendorf did not form part of a clear and constant line of case law: ibid, para 32. But they also thought that the reasoning behind the orthodox view was weak and did not lead to sensible conclusions: ibid, paras 34–35, 38–40. See Graham, L, ‘The Modern Mirror Principle’ (2021) PL 523, 529 and 531Google Scholar.

57 S v Denmark (2019) 68 EHRR 17.

58 ibid, para 118.

59 Al Husin (n 12) para 65; Jendrowiak (n 15) paras 35, 38–39.

60 Although note, as the Court has pointed out, this change ‘does not permit a policy of general prevention directed against individuals who are perceived by the authorities as being dangerous or having the propensity to commit unlawful acts’, chiefly because such a policy would fall foul of other requirements in the Court's case law which must be shown in order to justify arrest, for example that the commission of an offence must be imminent: Kurt v Austria (2022) 74 EHRR 6, para 186.

61 S v Denmark (n 57) para 137.

62 ibid, para 137.

63 See Eiseman-Renyard v United Kingdom (2019) 68 EHRR SE12, paras 36–38 and Archer (n 9) para 91(ii).

64 See Macken, C, ‘Preventive Detention and the Right to Personal Liberty and Security Under Article 5 ECHR’ (2006) 3 IJHR 195Google Scholar.

65 It is true that, as was pointed out by the UKSC in Hicks, there are some cases in the early jurisprudence of the European Court of Human Rights which might be said to reflect, to some extent, the approach set out in S v Denmark, such as Lawless v Ireland (No 3) (1961) 1 EHRR 15 and Brogan v United Kingdom (1988) 11 EHRR 117. However, in the view of the author, even if these early cases could be said to constitute an authoritative line of case law, it is nonetheless one which has clearly been overtaken by the Court in more recent jurisprudence. Whether the Court's approach in S v Denmark is framed as an entirely new development or something like course correction, it still represents a clear change in the law.

66 Khlaifia (n 14) para 71.

67 Archer (n 9).

68 ibid, para 41.

69 ibid, paras 112, 120.

70 ibid, para 115.

71 The Court of Appeal at one point insisted that the abstract compatibility of the provision in the Police and Criminal Evidence Act was ‘not the central issue’ between the parties (ibid, para 115) as the most important issue was whether the detention could be justified on the facts. This is unconvincing—from the very outset the Court accepts that the case involves a challenge to the legislation itself, rather than just its application in specific circumstances: ibid, paras 1, 3, 111.

72 ibid, paras 123, 126.

73 ibid, para 123. A thorny question arises here as to how judges should approach questions of incompatibility under the Human Rights Act when the provision may operate compatibly in some cases but not others: see S Wilson Stark, ‘Section 4 of the Human Rights Act 1998: Still Standing or Standing Still?’ (UK Constitutional Law Association Blog, 16 November 2022) <>.

74 At one point, the Court in Archer warned that ‘“own protection” cannot justify the detention of an individual in circumstances where none of the limbs of Article 5(1)(c) are made out’ (n 9, para 115), before going on to uphold a provision which does exactly that.

75 A Sanger, ‘Review of Executive Action Abroad: the UK Supreme Court in the International Legal Order’ (2019) 68(1) ICLQ 35, 45–7; S Wallace, The Application of the European Convention on Human Rights to Military Operations (CUP 2019) 157–65; KL Yip, ‘The Weakest Link: From Non-Derogation to Non-Existence of Human Rights’ (2017) 17(4) HRLR 770; C Landais and L Bass, ‘Reconciling the Rules of International Humanitarian Law with the Rules of European Human Rights Law’ (2015) 97 IRRC 1295; E Stubbins Bates, ‘Hassan v. The United Kingdom (Eur. Ct. H.R.)’ (2015) 54(1) ILM 83.

76 Al-Skeini v United Kingdom (2011) 53 EHRR 18, paras 133–137.

77 ibid, paras 138–140; Chiragov v Armenia (2016) 63 EHRR 9, para 168.

78 Al-Skeini, ibid, para 132.

79 See eg ibid, paras 143–150; Al-Jedda (n 32) paras 74–86.

80 Hassan v United Kingdom (2014) 38 BHRC 358, para 97.

81 Al-Saadoon v United Kingdom (2010) 51 EHRR 9, para 128; Al-Jedda (n 32) para 108. Otherwise, on occasion courts have rejected the argument that there existed any clash between the powers and obligations under the ECHR and other international law documents: see eg Al-Jedda, para 109 and Iraqi Civilians v Ministry of Defence [2014] EWHC 3686 (QB) paras 23–28.

82 Hassan (n 80).

83 A useful summary of the reasoning and findings in Hassan can be found in Abd Ali Hameed Al-Waheed v Ministry of Defence [2017] UKSC 2, paras 51–58 (Lord Sumption).

84 Hassan (n 80) para 103.

85 ibid, paras 104, 108–111.

86 R Ekins and G Verdirame, ‘Judicial Power and Military Action’ (2016) 132 LQR 206.

87 cf Al-Waheed (n 83) para 68 (Lord Sumption).

88 Al-Saadoon v Secretary of State for Defence [2016] EWCA Civ 811, para 184 (Lloyd-Jones LJ).

89 Hassan (n 80) para 104.

90 Al-Waheed (n 83).

91 See Alseran v Ministry of Defence [2017] EWHC 3289 (QB) para 89.

92 Al-Waheed (n 83) para 132.

93 ibid, paras 111 (Lord Sumption) and 140 (Lord Wilson). For criticism, see A Habteslasie, ‘Detention in Times of War: Article 5 of the ECHR, UN Security Council Resolutions and the Supreme Court Decision in Serdar Mohammed v Ministry of Defence’ (2017) 2 EHRLR 180, 183–8.

94 Al-Waheed, ibid, para 68.

95 ibid, paras 68, 84.

96 ibid, paras 305–316.

97 ibid, para 84 (Lord Sumption).

98 Especially at the early stages of the pandemic, see eg M Smith, ‘Public Overwhelmingly Backs the Government's New Measures to Tackle Coronavirus’ (YouGov, 24 March 2020) <>.

99 This article does not consider the challenges lodged to other, more specific measures adopted in response to lockdowns, some of which may invoke Article 5 alongside other rights, eg inter-State travel restrictions (R (Khalid) v Secretary of State for the Home Department [2021] EWHC 2156 (Admin); R (Hotta) v Secretary of State for Health and Social Care [2021] EWHC 3359 (Admin)), limits on the way in which legal proceedings are conducted (Bah v the Netherlands App No 35751/20 (ECtHR, 22 June 2021)), closure of places of worship (R (Hussain) v Secretary of State for Health and Social Care [2022] EWHC 82 (Admin)) and the implementation of so-called ‘vaccine passports’ (Re O'Murchu [2022] NIQB 13).

100 See, eg D Feldman, ‘Counter-Infection Methods and ECHR Article 5’ (2020) 25(2) JR 80, 91: ‘Does ordering people to remain at home (with exceptions) under threat of coercion and penalties deprive them of their liberty? On the face of it, yes.’; A Greene, ‘Derogating from the European Convention on Human Rights in Response to the Coronavirus Pandemic: If Not Now, When?’ (2020) 3 EHRLR 262, 267–9; T Hickman, ‘The Coronavirus Pandemic and Derogation from the European Convention on Human Rights’ (2020) 6 EHRLR 593, 602.

101 De Tommaso (n 26) para 85.

102 ibid, paras 49, 85, 88; Secretary of State for the Home Department v JJ (n 18) para 24; Secretary of State for the Home Department v GG [2016] EWHC 1193 (Admin) para 36.

103 Nikolova v Bulgaria (No 2) App No 40896/98 (ECtHR, 30 December 2004) para 53.

104 Secretary of State for the Home Department v AP (n 18) para 15.

105 Buzadji v Moldova (2016) 42 BHRC 398, para 43.

106 Re AF [2018] EWHC 138 (Fam) para 14; Re D [2017] EWCA Civ 1695, para 22.

107 See Gillan (n 31).

108 This is so with regards to general lockdowns; other aspects of the UK's response to COVID, such as quarantine requirements for potentially infections individuals and their households, may engage Article 5 more easily: see Feldman (n 100) 86.

109 Terheş v Romania App No 49933/20 (ECtHR, 13 April 2021). See also a case comment by J Dute and T Goffin, ‘European Court of Human Rights’ (2021) 28(4) EJHL 404, 407.

110 On which, see n 118–n 122 below.

111 Terheş (n 109) para 38.

112 ibid, para 39.

113 ibid, para 40.

114 R (Dolan) v Secretary of State for Health and Social Care [2020] EWCA Civ 1605, para 93.

115 ibid, para 94. See also Suraj v Attorney General of Trinidad and Tobago [2022] UKPC 26, paras 98–100.

116 See Scientific Advisory Group for Emergencies, ‘Scientific Evidence Supporting the Government Response to Coronavirus (COVID-19)’ (2020) <>.

117 E Bendavid et al, ‘Assessing Mandatory Stay-at-Home and Business Closure Effects on the Spread of COVID-19’ (2021) 54(1) EurJClinInvest e13484.

118 Greene (n 100) 268–9.

119 The sole authority of significance is Enhorn v Sweden (2005) 41 EHRR 30.

120 ibid, para 44; Greene (n 100) 268.

121 Greene, ibid 268–9 and A Greene, ‘Derogations, Deprivation of Liberty and the Containment Stage of Pandemic Responses’ (2021) 4 EHRLR 389, 390–1. cf Hotta (n 99) para 18. That case concerned the legality of quarantine hotel arrangements, where Chamberlain J assumed, without deciding, that the measures in question constituted a deprivation of liberty.

122 Greene (n 100) 268, citing Winterwerp v the Netherlands (1979–80) 2 EHRR 387.

123 Strasbourg may revisit this approach in the future; a number of cases challenging various aspects of national lockdowns, or measures akin to lockdowns, have been lodged before the Court: see eg Spînu v Romania App No 29443/20 (communicated 1 October 2020). For a table of cases raising similar issues, see L Graham, ‘Challenging State Responses to the Covid-19 Pandemic Before the ECtHR’ (Strasbourg Observers, 18 October 2022) <>. The European Court is also due to hand down its decision in Pagerie v France App No 24203/16, concerning residence orders put in place in France under the state of emergency declared following the terrorist attacks on 13 November 2015.

124 Stark (n 18) 382.

125 Ashworth and Zedner (n 42) 64.

126 Stone (n 41) 56.

127 For an explanation of control orders, see D Anderson QC, ‘Control Orders in 2011: Final Report of the Independent Reviewer on the Prevention of Terrorism Act 2005’ (2012) <>.

128 Fenwick, H and Fenwick, D, ‘The Case for More Ready Resort to Derogations from the ECHR in the Current War on Terror’ (2018) 4 EHRLR 303, 304Google Scholar.

129 ibid 305.

130 R Edwards, ‘Police Powers and Article 5 ECHR: Time for a New Approach to the Interpretation of the Right to Liberty’ (2020) 41 LiverpLRev 331, 334–5; Stone (n 41) 52–4; in at least two high-profile cases, the European Court of Human Rights found problems with UK anti-terrorism search powers, but in each case, dealt with the issues under Article 8, leaving undecided the question of whether the powers in each case engaged Article 5: see Gillan (n 31) paras 56–57 (cf the conclusions of domestic courts in that applicant's case: [2006] UKHL 12, para 25 and [2004] EWCA Civ 1067, para 45) and Beghal v United Kingdom (2019) 69 EHRR 28, paras 111–113 (cf the conclusion of domestic courts in that applicant's case: Beghal (n 8) para 56).

131 Compare the reasoning of the House of Lords in Re L [1998] UKHL 24 with that of the European Court of Human Rights in HL v UK (2005) 40 EHRR 32.

132 S v Denmark (n 57) para 116; Hicks (n 55) para 29.

133 Austin (n 19) para 56.

134 See generally, Stark (n 18).

135 The language of ‘slot machine jurisprudence’ is taken from Robertson, D, Judicial Discretion in the House of Lords (OUP 1998) 19Google Scholar. cf A Beever, ‘The Declaratory Theory of Law’ (2013) 33(3) OJLS 421.

136 M Mansfield, ‘Our Right to Protest is Under Attack’ (The Guardian, 1 May 2013) <>; L Christian, ‘This Judgment in Favour of Kettling is a Missed Opportunity’ (The Guardian, 15 March 2012) <>.

137 D Lammy, ‘An Independent Review into the Treatment of, and Outcomes for, Black, Asian and Minority Ethnic Individuals in the Criminal Justice System’ (2017) <>; HM Inspectorate of Constabulary and Fire & Rescue Services, ‘Disproportionate Use of Police Powers: A Spotlight on Stop and Search and the Use of Force’ (2021) <>; Yesufu, S, ‘Discriminatory Use of Police Stop-and-Search Powers in London, UK’ (2013) 15(4) IJPSM 281Google Scholar.

138 Including Lord Sumption, former Justice of the UK Supreme Court: see eg J Sumption, ‘You Cannot Imprison an Entire Population’ (The Spectator, 17 May 2020) < />.

139 Feldman, D, ‘Containment, Deprivation of Liberty and Breach of the Peace’ (2009) 68 CLJ 243, 245CrossRefGoogle Scholar.

140 Oreb, N, ‘The Legality of “Kettling” after Austin’ (2013) 76(4) MLR 735Google Scholar, 741. See also Austin (n 19) Dissenting Opinion of Judges Tulkens, Spielmann and Garlicki, para 5.

141 See eg Austin, ibid (Dissenting Opinion of Judges Tulkens, Spielmann and Garlicki); S v Denmark (n 57) (Partly Dissenting Opinion of Judges De Gaetano and Wojtyczek); Hassan (n 80) (Partly Dissenting Opinion of Judge Spano).

142 Some have suggested amending or adapting the definition of coercion applicable to Article 5, eg Edwards (n 130). However, this does not solve the problem identified in this article, which relates to justification.

143 Edwards, (n 130) 352.

144 Indeed, a minority of judges in S v Denmark, unable to sign up to the majority's creative interpretation of Article 5, suggested this route, arguing that it would be more democratically legitimate, and less an abdication of the judicial role, than bending the orthodox approach to achieve a desired result: S v Denmark (n 57), Partly Dissenting Opinion of Judges De Gaetano and Wojtyczek, para 10.

145 Feldman, D, ‘Deprivation of Liberty in Anti-Terrorism Law’ (2008) 67 CLJ 4, 7CrossRefGoogle Scholar: it ‘should be possible to assess … the most appropriate Convention … qualifications … instead of cramming them into (unqualified) Article 5’.

146 An amendment to the Convention must be drafted, then approved by the Council of Europe's Parliamentary Assembly and the Committee of Ministers. So far, no amendment to the content of the original rights in the Convention has been seriously entertained by those bodies. States can also choose not to accede to any such amending protocols (how this would work with an amendment which amends an existing right, rather than adding a new one, is unclear).

147 Alseran (n 91) para 83 (Leggatt J); Hassan (n 80), Partly Dissenting Opinion of Judge Spano, para 9.

148 Fenwick and Fenwick (n 128) 304–9, esp 309: ‘resort to a derogation may have advantages over more stealthy departures from rights standards’; Greene (n 100) 273: ‘Far from protecting human rights, arguing against the necessity for derogations to ensure lockdowns are compatible with the ECHR recalibrates the protection of rights downwards in order to accommodate lockdown measures under the ostensible banner of normalcy.’

149 A helpful list can be found in Appendix 1 to Wallace, S, ‘Derogations from the European Convention on Human Rights: The Case for Reform’ (2020) 20 HRLR 796, 793–6Google Scholar.

150 A list of derogation notices in relation to the pandemic can be found at Council of Europe, ‘Derogations Covid-19’ <>.

151 See Greene (n 100) 269–70 and Greene (n 121).

152 Hassan (n 80) para 101.

153 See Article 15(1) of the ECHR.

154 Particularly with regards to the second requirement: see A v UK (n 11) paras 182–190; Dareskizb Ltd v Armenia App No 61737/08 (ECtHR, 21 September 2021) paras 55–63.

155 Also, the derogation mechanism may not apply straightforwardly in cases concerning the extraterritorial application of the Convention: Al-Waheed (n 83) paras 45 (Lord Sumption) and 163 (Lord Mance) cf Rooney, J, ‘Extraterritorial Derogation from the European Convention on Human Rights in the United Kingdom’ (2016) 6 EHRLR 656, 659–62Google Scholar.

156 Al-Waheed (n 83) para 163.

157 ibid, para 142 (Lord Wilson); see also Hickman (n 100) and Dzehtsiarou, K, ‘Article 15 Derogations: Are They Really Necessary During the COVID-19 Pandemic?’ (2020) 4 EHRLR 359Google Scholar.

158 Sanger (n 75) 43.

159 The importance of a favourable political environment as a precondition for functioning constitutional courts has been noted in the literature: Gardbaum, S, ‘What Makes for More or Less Powerful Constitutional Courts’ (2018) 29 DukeJComp&IntlL 1, 4Google Scholar.

160 Bates, E, ‘The UK and Strasbourg: A Strained Relationship – The Long View’ in Ziegler, K, Wicks, E and Hodson, L (eds), The UK and European Human Rights: A Strained Relationship? (Hart 2015)Google Scholar.

161 See, in different contexts: Madsen, MR, ‘Rebalancing European Human Rights: Has the Brighton Declaration Engendered a New Deal on Human Rights in Europe?’ (2018) 9(2) JIDS 199Google Scholar; Alter, K, Gathii, J and Helfer, L, ‘Backlash Against International Courts in West, East and Southern Africa: Causes and Consequences’ (2016) 27(2) EJIL 293CrossRefGoogle Scholar; Gardbaum, S, ‘Are Strong Constitutional Courts Always a Good Thing for New Democracies?’ (2015) 53 ColumJTransnat'lL 285Google Scholar.

162 Madsen, Cebulak and Wiebusch have usefully distinguished, in this context, ‘ordinary resistance’ or ‘pushback’ (eg political statements disagreeing with the outcome of a particular judgment, or a lack of enthusiasm when it comes to adopting and following it) with ‘extraordinary resistance’ or ‘backlash’ (eg responses which go beyond criticism and seek to undermine the judicial bodies rendering judgment); judges ought reasonably expect to encounter the former but can legitimately resist the latter: Madsen, MR, Cebulak, P and Wiebusch, M, ‘Backlash Against International Courts: Explaining the Forms and Patterns of Resistance to International Courts’ (2018) 14(2) IntJLC 197Google Scholar.

163 A related phenomenon is ‘avoidance’, where judges engineer a solution to a case which sidesteps a key controversy, or otherwise remove the case from the judicial docket altogether, in order to avoid engaging with certain sensitive issues: see Odermatt, J, ‘Patterns of Avoidance: Political Questions Before International Courts’ (2018) 14(2) IntJLC 221Google Scholar; Delaney, EF, ‘Analyzing Avoidance: Judicial Strategy in Comparative Perspective’ (2016) 66 DukeLJ 1Google Scholar. For application to the European Court of Human Rights, see Jackson, M, ‘Judicial Avoidance at the European Court of Human Rights: Institutional Authority, the Procedural Turn and Docket Control’ (2022) 19 ICON 1Google Scholar and Graham, L, ‘Strategic Admissibility Decisions in the European Court of Human Rights’ (2020) 69(1) ICLQ 79CrossRefGoogle Scholar.

164 Commentators have suggested that the approach adopted in some of the cases described in this article was driven, in part, by a desire to avoid backlash: see, eg in relation to Austin: D Mead, ‘The Right to Protest Contained by Strasbourg: An Analysis of Austin v UK & the Constitutional Pluralist Issues it Throws Up’ (UK Constitutional Law Association Blog, 16 March 2012) <> and in relation to the extraterritorial application of the ECHR: Mallory, Conall, Human Rights Imperialists: the Extraterritorial Application of the European Convention on Human Rights (Hart 2019) 61–87CrossRefGoogle Scholar.

165 Bates, E, ‘Principled Criticism and a Warning from the ‘UK’ to the ECtHR?’ in Breuer, Marten (ed), Principled Resistance Against ECtHR Judgments – A New Paradigm? (Springer 2019)Google Scholar.

166 Austin (n 34).

167 Hicks (n 55).

168 See Graham (n 56).

169 See Jackson (n 163).

170 See Kurt (n 60) paras 183–189.

171 Austin (n 34) para 58.

172 Lord Wilson, ‘Our Human Rights: A Joint Effort?’ (Northwestern University, Chicago, 25 September 2018) 8 <>.

173 It bears repeating here that the United Kingdom has not ratified this provision.

174 Al-Waheed (n 83) paras 42 (Lord Sumption) and 142–143 (Lord Wilson).