Accounts of vulnerability within positive human rights obligations

Abstract

Accounts of human beings as vulnerable have provided powerful reposts to liberal individualism in recent decades. Concurrently, the European Court of Human Rights’ jurisprudence on Convention states’ positive obligations often obliges public authorities to address particular vulnerabilities. These developments reflect elements of different theoretical accounts of vulnerability but lack a coherent approach to the human subject. Exploring the impact of this in the UK Supreme Court’s jurisprudence, we evaluate two case studies in which positive obligations have been imposed on the police; (1) public order in the context of inter-community tensions in Northern Ireland (DB v. Chief Constable of Police Service of Northern Ireland) and (2) police investigations in regard to serial sexual offending (Commissioner of Police of the Metropolis v. DSD). This jurisprudence illustrates how some domestic judges are supplying their decisions with rationalisations which are lacking in the European Court’s case law.

1 Introduction

At the turn of the millennium feminist scholars had good reason to conclude that codified human rights instruments held out false promises for women, and to find little transformative in the enactment of the United Kingdom’s Human Rights Act (McColgan, 2000). The last two decades, however, have seen perceptible shifts in human rights jurisprudence. ¹ The European Court of Human Rights has increased its reliance on the language of vulnerability, a concept deeply embedded in the feminist scholarship of Martha Fineman (2008) and Judith Butler (2004; 2009). This jurisprudence is, moreover, increasingly becoming intertwined with the Court’s emphasis of the state’s positive obligations under Articles 2, 3 and 8 of the European Convention for Human Rights (ECHR). For Sandra Fredman, who also removes her account of human rights from liberal conceptions of individualism and the neutrality of the state (2008, pp. 18–23), these currents within contemporary jurisprudence are combining to challenge the notion that ‘freedom should be seen only as an absence of deliberate State interference’ (2008, p. 11). In this article we explore how human rights approaches which draw upon concepts of human vulnerability are reshaping the obligations upon the police in public order scenarios and in the investigation of crime. Beyond the beginnings of these connections within Strasbourg jurisprudence, more developed invocations of a vulnerable subject are playing an expansive role in the UK Supreme Court’s approach to policing.

We first provide an overview of different accounts of vulnerability and the extent to which aspects of these have been embraced within the European Court of Human Rights’ jurisprudence. This section is not intended to provide a holistic social and intellectual history of the development of the rights under the ECHR, but to provide the context for one particularly significant shift in how those rights are understood. The Court’s engagement with accounts of vulnerability to reconceptualise understandings of human rights is multi-faceted, extending across its operation of the proportionality test and use of the margin of appreciation (Peroni and Timmer, 2013, p. 1063). For the purposes of this article, however, we explore its specific significance for the developing range of positive obligations under the Convention. Having explored the changing nature of Strasbourg’s jurisprudence and attempted to locate it within distinct theoretical accounts of vulnerability, the second part of our article explores the impact of this shift upon the positive obligations which apply to UK policing. As case studies, we examine how the Holy Cross ² and Flags Protest ³ cases reshaped public order policing and how the nature of Northern Ireland’s post-conflict society is informed by general understandings of vulnerability. We thereafter explore how positive obligations shaped the outcome of Commissioner of Police of the Metropolis v. DSD and another, ⁴ and established a new base line for criminal investigations relating to serious harms to individuals.

This case-study approach, exploring these shifts within the context of particular policing practices (Bryman, 2016, p. 48), enables a detailed investigation and analysis of how the courts approach a concept like vulnerability which, ‘can only be studied or understood in context’ (Gillham, 2000, p. 1). These particular cases have been selected because of their transformative potential for shaping police accountability in relation to breaches of human rights, and because they encompass the development of positive obligations across a spectrum of rights, from the strict requirements of Article 3 ECHR to the context of Article 8 ECHR as a qualified right. The case law around positive obligations remains fluid, and therefore any effort to track the influence of a generalised conception of vulnerability within judicial thinking requires a detailed analysis of particular strands of case law. Moreover, as Martha Fineman emphasises, scholarship which engages vulnerability theories requires a focus on societal institutions, oppressive relations and the individual as always being socially situated (2008, p. 13), meaning that the cases we explore can only be understood in context. Holy Cross and Flags Protest must be understood in the context of post-conflict policing after decades of political violence and the inter-community tensions in Northern Ireland. DSD cannot be effectively evaluated without a grounding in sexual violence and its place and role within structurally oppressive relations, alongside the extensive critiques of policing and police responses to sexual violence.

Our analysis is based on the theoretical approach that a case study can generate ‘contextualised explanation’ which does not trade-off internal validity for thick description, as other approaches to case studies can be prone to do, and ‘acknowledg[es] the complexity of the social world, the bounded scope and contingency of causal relationships, and the simultaneous operation of multiple interaction effects’ (Welch et al., 2020, pp. 173, 204). There is, however, the resultant potential for cases studies to be indeterminate if more than one explanation could be offered for a particular point (King, Keohane and Verba, 1996, pp. 208–211). Indeed, the Supreme Court judgments we analyse rely on such sparse reasoning that they occlude the reconceptualisation of human rights at work in these cases and, as such, could arguably be open to alternative interpretations. Nevertheless, as we work through these cases from first instance, ground them in context and in relation to the wider jurisprudence of the European Court and vulnerability scholarship, our approach enables us to demonstrate that the courts are not simply invoking positive obligations but are wrestling with the implications of vulnerability theory for human rights.

The UK courts’ engagement with accounts of vulnerability is arguably an effort to supply a theoretical basis for positive obligations which has been limited within Strasbourg case law, which in itself necessitates further analysis of why Strasbourg continues to use the concept of vulnerability so loosely and whether the UK courts are outrunning the Convention organs which is beyond the scope of this article. We instead seek to understand these developments in light of the pioneering work of Butler, Fineman and Fredman, and determine that these case studies illustrate that the UK courts are moving towards an ontological approach which is most compatible with Fineman’s account of vulnerability. We also indicate the potential for Strasbourg case law to be developed in ways which more effectively embed a vulnerable subject within ECHR rights.

2 Human rights, vulnerability and positive obligations

2.1 Human rights and the liberal subject

At first thought, it might appear axiomatic that the ECHR was concluded, in the aftermath of World War 2, precisely to respond to human vulnerability to harm and the precariousness of life (Grear, 2020, p. 156). In the post-war context, however, the European Convention’s drafters were primarily concerned with protecting the individual from intrusive state interference with their civil and political rights, typically grounded upon narrow conceptions of dignity and autonomy (Mahoney, 1998, p. 4). Such negative accounts of civil and political rights are often intuitively attractive when viewed from the stand point of financially secure, able-bodied white men (Macpherson, 1962). ⁵ When people from this group were overwhelmingly responsible for drafting the ECHR, in the shadow of fascism and communism, the state was unsurprisingly presented as a threat to be constrained by legal bonds (Heri, 2021, p. 24).

This abstract human rights subject, Anna Grear explains, ‘is relatively invulnerable compared to the quintessentially vulnerable others of human rights, but uniquely vulnerable to the kind of injury most significant to the liberal subject of rights’ (Grear, 2020, p. 158). As such, for women and other oppressed groups this conception of the human, freedom, and corresponding rights does not accommodate the violence and abuses common to their lives (Marshall, 2008a, p. 340). In addition, the ‘quintessentially vulnerable others’ have often been excluded from and marginalised in public life, embedded in what the liberal state deems the ‘private sphere’ and relations of care into which it is loath to intrude. Feminists have thus highlighted the relational aspects of human life (Nedelsky, 2011), and the co-constitutive relationship between the individual and society (Fredman, 2008, ch. 1; Nussbaum, 1999, ch. 2). This picture is a far cry from the atomistic, rational, and disembodied liberal subject. Within this frame it is clear that social and institutional relations and structures affect individuals’ participation in public life, access to public goods, and opportunities. As such, Fredman argues, individual freedom cannot be guaranteed purely by protecting individuals from state interference but rather requires ‘a positive duty on the State to ensure the provision of a range of options, of public goods and the framework within which human relationships can flourish’ (Fredman, 2008, p. 18). In addition, oppressed and marginalised groups are more likely to be subject to harms by non-state actors, for example, men perpetrating sexual and domestic violence against women, which are also a cause and consequence of social and institutional inequalities. This means that to address inequalities and to protect all human beings there must be positive obligations on the state to protect against harms caused by non-state actors (Radačić, 2008, p. 275).

Michael Oakeshott regarded post-war human rights arrangements as diverging from legal protections grounded in valorised accounts of ‘rugged individualism’. Even if this shift was at best nascent in the Universal Declaration of Human Rights and the subsequent ECHR, it nonetheless alarmed him (1947–8, p. 490):

‘[T]hese children of our own flesh have been returning to us, disguised in a foreign dress, the outline blurred by false theory and the detail fixed with an uncharacteristic precision. What went abroad as the concrete rights of an Englishman have returned home as the abstract Rights of Man, and they have returned to confound our politics and corrupt our mind.’

Oakeshott’s concern over the looming displacement of ‘the concrete rights of an Englishman’ with ‘foreign’ human rights protections continues to motivate opposition to the ECHR in the UK. Nor was this a mere rhetorical flourish. Oakeshott actively set the rise of the ‘anti-individual’ he found in human rights instruments against the era in which ‘every subject was secured of the right to pursue his chosen directions of activity as little hindered as might be by his fellows or by the exactions of government itself, and as little distracted by communal pressures’ (Oakeshott, 1961, p. 156). ⁶ And yet, notwithstanding these complaints, the efforts to universalise post-war human rights treaties continued to rely on a disembodied and rationalistic subject. ⁷

Indeed, generalised engagement with more complex conceptions of human relationships and vulnerability have been slow to develop under general human rights instruments (Paglione, 2006, p. 123). For many commentators the ECHR continues to prioritise individual autonomy and the protection of ‘an individual’s freedom of choice as to how to live their lives’ (Douglas, 2018, p. 362; Möller, 2012, pp. 63–65). To the extent that the ECHR imposed positive obligations on state parties, these depended upon obligations to secure Convention rights for all persons within a state’s jurisdiction (Article 1) and to provide effective domestic remedies for breaches of rights (Article 13), ⁸ and the Court was left to grapple with the limits of the duty on states to make rights practical and effective. In doing so Strasbourg has since developed positive obligations in its interpretation and application of ‘virtually every Convention Right’ (Lavrysen, 2016, p. 4; Starmer, 2001, p. 139), but this remains a work in progress. The realisation that, ‘[d]ue to growing standards of human dignity, the Court can no longer apply post-Second World War approaches to current pressing human rights issues’, ⁹ has yet to be fully embodied in its jurisprudence.

2.2 The development of positive obligations in ECHR law

Despite an initial emphasis on securing a negative conception of freedom, the Strasbourg Court has come to recognise that ‘the protection of human rights requires the state to take appropriate measures to safeguard these rights from violation by others’ (van Kempen, 2013, p. 16). To do so, positive duties have been extended, underpinned by shifts in the Court’s conception of the individual and their relationship to the state, with the Court paying close attention to the particular vulnerability, or ‘constructed disadvantage’, of certain rights holders (Peroni and Timmer, 2013, p. 1062). Such interpretations, however, will not necessarily apply evenly across the states party to the ECHR because the interpretation and application of the ECHR varies at a national level under the influence of interpretative tools such as the margin of appreciation (Føllesdal, 2018). In addition, the extent and scope of positive obligations can depend on the context; for instance, law enforcement agencies’ obligations vary depending upon the rights at issue and their interaction with the established purposes of criminal justice (Dickson, 2020, pp. 232–233).

Strasbourg first identified positive obligations upon police activities in relation to the right to life (Article 2) and freedom from torture or inhuman or degrading treatment (Article 3). Although primarily framed as prohibitions upon state conduct which interferes with Articles 2 and 3, the Court recognised that a requirement for effective investigations into potential infringements by state actors is needed to prevent abuse. ¹⁰ The Court subsequently recognised that states were obliged to protect against breaches of Articles 2 and 3 by non-state actors in certain circumstances. ¹¹ Beginning with Osman v. United Kingdom, cases framed this obligation upon the authorities as a requirement to take ‘reasonable steps’ to respond to a known ‘real and immediate risk’ to someone’s life or that they could be subject to inhuman or degrading treatment. This, however, has been far from an easy standard for claimants to meet and the Court ultimately dismissed the claim in Osman, leaving commentators to conjecture about the scope of these duties (Burton, 2009, p. 286). The Court’s jurisprudence was nonetheless emphasising that states must maintain both a legal framework and policing arrangements sufficient to ensure an acceptable level of protection against breaches of these rights.

Challenges to the effectiveness of investigations increasingly drew upon Article 8’s protection of private and family life. ¹² In MC v. Bulgaria the Court accepted that Articles 3 and 8 protected ‘essential aspects of private life’ (para. 150), and that a state’s obligations to secure these rights extended beyond the enactment of laws to criminalise the abuses of these rights inherent in cases of rape to also oblige the state to adequately investigate and prosecute such allegations (paras. 150 and 153). MC’s challenge was, for the Court, a showcase instance of the systemic failings in Bulgaria’s criminal justice system with regard to protecting the rights of women who had been raped (Conaghan, 2005). As fresh challenges drew the Court into considering whether this approach applied to cases of domestic violence, judgments began to emphasise the vulnerability of individuals subject to abuse by their partner as part of the explanation for why Article 3 ECHR imposed positive obligations on the authorities to protect the bodily integrity of the individual in such cases. It was for this reason that, in Rumor v. Italy, for example, the Court identified a positive obligation to provide ‘effective protection of, inter alia, an identified individual or individuals from the criminal acts of a third party, as well as reasonable steps to prevent ill-treatment of which the authorities knew or ought to have known’. ¹³ This application was nonetheless unsuccessful, with the Court emphasising that it was ‘not the Court’s role to replace the national authorities and to choose in their stead from among the wide range of possible measures that could be taken to secure compliance with their positive obligations’ (para. 59; see also McQuigg, 2015, pp. 1013–1014).

In Opuz v. Turkey, ¹⁴ the European Court had previously found breaches of Articles 2, 3, 8 and 14 ECHR where the police had failed to protect the applicant and her mother from domestic violence. Here, Strasbourg emphasised the vulnerability of the applicant – due to past experiences of abuse, her fear of further violence, and her position as a woman in south-east Turkey (para. 160) – as relevant to establishing that the minimum level of severity for Article 3 was met. ¹⁵ These judgments, considered together, indicate that the ECHR requires that states provide people within their jurisdiction with adequate protections for their fundamental personal interests against abuse. The Court has furthermore connected its approach to positive obligations to interpreting the ECHR rights in light of treaties such as CEDAW ¹⁶ and the UNCRC, ¹⁷ which draws these treaties into counterbalancing the established narratives of the ‘rights holder’. For all that the Strasbourg Court has experimented with the language of vulnerability, this jurisprudence does not necessarily align with any of the developed accounts of vulnerability in scholarship. In the next section we therefore explore some of the main theories of vulnerability and its relationship to the human subject, before addressing the extent to which these are manifested in Strasbourg jurisprudence.

2.3 Vulnerability and the conception of the human

Butler conceives of human life as being characterised by ‘a common human vulnerability’ (Butler, 2004, p. 30). However, not everyone is vulnerable in the same way, to the same extent, or experiences it as such because, Butler emphasises, ‘our precarity is to a large extent dependent upon the organization of economic and social relationships, the presence or absence of sustaining infrastructures and social and political institutions’ (2004, p. 20). Attending to vulnerability, as Susan Dodds puts it, ‘changes citizens’ ethical relations from those of independent actors carving out realms of right against each other and the state, to those of mutually-dependent and vulnerably-exposed being whose capacities to develop as subjects are directly and indirectly mediated by the conditions around them’ (2007, p. 501). This reflects a positive conception of freedom which negative state obligations alone cannot secure. Butler, however, is sceptical as to whether the state can provide an effective response to human vulnerability.

Martha Fineman adopts a similar approach to the human subject but a different view of the role of the state. She challenges the independent and autonomous atomistic individual of traditional liberalism and its negative conception of freedom. She also sets out a conception of human vulnerability which is a ‘universal’ and ‘constant’ aspect of the human condition (Fineman, 2008, pp. 1–2). Her particular take on this is to centre the vulnerable human in law and politics to demand a more responsive state, and to work toward a renewed vision of substantive equality (Fineman, 2008, p. 2). Moreover, she argues that vulnerability should not be seen as a pejorative concept, in that individuals’ reliance on, care for, and engagement with others can entail empathy, social connections, pleasure and intimacy (Fineman, 2008, p. 10). For Butler, acceptance of the self as vulnerable is a way to empathise with vulnerable others, providing a basis for a more ethical politics and global justice (2004).

Butler and Fineman are not the only theorists to develop a conception of human vulnerability and consider the implications for politics, law, and social life. This growing field is not without critiques. There are numerous examples of a group or individual being identified as vulnerable to justify paternalism, restricting autonomy rather than ensuring care and support which enhances freedom (Brown, 2011; Dunn, Clare and Holland, 2008; Munro, 2017; Murphy, 2009). Moreover, employing vulnerability can have prescriptive force which could reinforce paternalism and increase the possibilities for abuse (Murphy, 2009, p. 69). Identifying a particular group as vulnerable can also position it against independent and autonomous individuals, thus not challenging traditional liberalism nor recognising vulnerability as an inherent aspect of the human condition (Diduck, 2014, p. 95).

These debates have played out in the context of human rights. Turner has argued that a vulnerability approach could radically transform the foundations of human rights because human rights structures are a way to limit vulnerability, and to hold states accountable for abuses and failures to protect (Turner, 2006). Rather than protecting against vulnerability and instead encountering it and responding to it, Butler says that ‘[t]he recognition of shared precariousness introduces strong normative commitments of equality and invites a more robust universalizing of rights that seeks to address basic human needs for food, shelter, and other conditions for persisting and flourishing’ (Butler, 2009, p. 29). However, for all that we have tracked the Court engaging with the concept of vulnerability in its jurisprudence, it is predominantly employed as a ‘binary characteristic of specific individuals and groups’ rather than as a universal aspect of the human condition (Heri, 2020, p. 96). This group vulnerability, Lourdes Peroni and Alexandra Timmer argue, is neither at odds with the autonomous liberal subject nor with the vulnerable subject (Peroni and Timmer, 2013; Xenos, 2004). If this is the limit of vulnerability in jurisprudence, however, then it is the antithesis of the empowerment sought by Fineman and Butler, leaving Corina Heri sceptical about whether much of the Court’s positive obligations jurisprudence has provided an effective realisation of Fineman’s thought (Heri, 2021, pp. 133–136).

Timmer nonetheless sees the potential for Strasbourg to move beyond a ‘vulnerable group’ approach toward a conception of vulnerability which reflects Fineman’s, on the basis that the Court is increasingly highlighting human dignity as a justification for recognising positive obligations (Timmer, 2013). ¹⁸ The Court’s reliance upon this rationale is notable in cases involving countries which have embedded the principle of human dignity within their constitutional orders. ¹⁹ This shift, for Heri, has transformational potential for the operation of human rights; ‘the interdependence of different human rights means that vulnerability and capabilities theories can transform the application not just of a particular provision, but of the human rights framework as a whole’ (Heri, 2021, p. 187). The development of positive obligations on the state to prevent rights violations furthermore illustrates the Court recognising that people are vulnerable not simply to abuses of state power but also to harms inflicted by non-state actors, the existence and extent of which is shaped by particular social and political conditions (Butler, 2004, pp. 28–29). The extent to which Strasbourg could be described as expanding the scope of positive obligations in a way which encompasses a more vulnerable subject thus remains open to debate, but this debate is not without limit. Even expansive accounts of positive obligations jurisprudence remain hedged with constraints.

2.4 The constraints on positive obligations

The development of positive obligations, informed by the European Court’s concerns that particular individual or group vulnerabilities require protection by states, has been subject to trenchant critiques. For some, positive obligations are an unacceptable example of the Court overstepping the boundaries of the Convention and engaging in the creation of human rights (Ashworth, 2014, p. 198). For others, these developments undermine the concept of human rights; ‘the fact that certain conduct is contrary to the value on which a human right is based does not ipso facto mean that this particular conduct falls within the scope of that human right or its corresponding obligations’ (van Kempen, 2013, p. 18). Further critiques regard the European Court’s invocation of positive obligations as adding an indeterminate element into human rights cases, against a backdrop in which the already ‘often nebulous content of human rights prescriptions’ presents challenges for ‘a positivist ideal of the rule of law’ (Campbell, 1999, p. 9).

As a result, case law asserting positive obligations comes couched in anxious concerns over whether the concept of human rights has ‘lost its meaning’. ²⁰ Even supportive scholarship has acknowledged that ‘the Court has already shown a substantial degree of creativity in producing the concept of positive obligations in the first place’ (McQuigg, 2015, p. 1024). In contemporary human rights debates, any suggestion of ‘human rights inflation’ draws with it accusations that judges are disregarding the boundaries of their role (Dzehtsiarou, 2021, p. 4). Positive obligations are, however, better understood as being unpacked rather than created by the European Court. The positive/negative dichotomy around human rights has long obscured the reality that the recognition of rights brought with it concomitant obligations (Hohfeld, 1913; Shue, 1996). It is unsurprising that the Court took quite some time to recognise this reality given the baggage of liberal orthodoxy on rights. Drawing on Articles 1 and 17, Brice Dickson concludes that the ECHR ‘presupposes that Council of Europe states will actually do things to “secure” Convention rights and to save them from “destruction”’ (Dickson, 2010, p. 204). Far from constituting a threat to the decision-making capacity of elected institutions, Fredman highlights that these developments seek to enhance the legitimacy of such decision making, by ensuring ‘that all have the ability to participate equally in democratic decision making’ (Fredman, 2008, p. 4). Fineman argues that in her vulnerability approach the state has an ‘affirmative obligation’ to either justify different treatment or remedy it, which in this context would mean to ensure that the Convention rights apply meaningfully to all (2008, p. 22).

Although the developing approach to positive obligations is grounded in what Andrew Ashworth describes as ‘a fair view of the implications of Article 1 of the Convention’ (Ashworth, 2014, p. 209), it is nonetheless an approach which is still being extrapolated by the Court, and which therefore gives rise to further concerns regarding the reach of positive obligations. For some, they illustrate how human rights ‘no longer serve to control and restrain the power of the state, but … to an important degree they also legitimize and even require the use of that power’ (van Kempen, 2013, p. 19). Going further, Butler is a ‘staunch critic’ of the state which she sees as ‘regulatory institution intent on consolidating its own power’ (Lloyd, 2007, p. 147; with reference to Butler, 2000). On one level, human rights have always been a legitimating tool for states, indeed they have been said to be ‘becoming a universal vocabulary of political legitimacy’ (Kennedy, 2005, p. 328), so it would be surprising if states had made no effort to harness positive obligations in this way. It is nonetheless a mischaracterisation to present the positive obligations which have developed under the ECHR as imposing ‘coercive duties on the state to criminalise, prevent, police, and prosecute harmful acts’, thereby permitting or even obliging states to slip the bounds of other human rights obligations (Lazarus, 2012, p. 136). The European Court has never formulated a positive obligation in such a way as to create an opportunity for public bodies to use powers to perpetrate unjustified infringements of rights. ²¹

The final element of this trifecta of concerns is the supposed indeterminacy of positive obligations. Indeterminacy is always a facet of developing jurisprudence, and so these concerns were inevitably going to be found alongside accusations of judicial creativity and overreach. Even commentators who are by no means hostile to these developments note the European Court’s ‘vacillation’ over certain elements of positive obligations (Ashworth, 2014, p. 206) and the incomplete indicators as to when it will draw upon group/individual vulnerability (Peroni and Timmer, 2013, p. 1064). This indeterminacy, however, does not occlude the relevant principles. Fredman, indeed, acknowledges the lynchpin case of Osman as doing an effective job at identifying the principles at play with regard to positive obligations to protect the right to life, with subsequent jurisprudence having brought increasing determinacy to the interaction between these principles, ‘each of which requires optimization in the light of legal and factual possibilities’ (Fredman, 2008, p. 76). This mapping of the scope of positive obligations has, moreover, taken place against a backdrop of the European Court’s acknowledgement of decision maker discretion and the limits of international-level oversight. The Court has limited means at its disposal to enforce positive obligations (Chevalier-Watts, 2010, p. 721), but it is developing a space in which domestic courts can operate free from the constraints of the margin of appreciation. The following section assesses the responses of the domestic courts in the UK to this shift. We examine the application of positive obligations to the policing of public order and criminal investigations, focusing on these examples because of their transformative potential for police accountability in relation to breaches of human rights and because they illustrate how judges are drawing accounts of vulnerability into the application of positive obligations.

3 Policing and positive obligations

Shifts in understandings of the authorities’ responsibilities under the ECHR took a considerable time to have any appreciable impact on policing in the UK. Some of the forms of legal accountability, such as the offence of misconduct in public office, could be applied to failures of the police to perform their general duty to society. But the strictures upon the offence, requiring police neglect to be wilful and the broad discretion available to the police in the exercise of their role, meant that it was only applicable to the most egregious police misconduct in the prevention of crimes. ²² With regard to negligence liability, Claire McIvor noted the impact of the Courts’ ‘rather rose-tinted view’ of policing in the UK (2010, p. 133), which prevailed into the last years of the twentieth century. ²³ This reflexive acceptance that the Peelite bromides of impartial community policing were invariably embodied in policing practice stymied the development of meaningful duties towards the victims of crime. ²⁴ As for the impact of positive obligations to protect the human rights of victims of crime, the courts recognised Strasbourg’s test in Osman as being clear and objective, ²⁵ but applied it in a restrictive manner. For Mandy Burton, ‘the duty to protect is emptied of much of its content if it refers only to the information the police had and not information which they ought, with due diligence, to have acquired’ (Burton, 2009, p. 287).

The Courts have since, however, had regular opportunities to rethink their approach to police obligations. In cases such as Michael v. Chief Constable of South Wales, ²⁶ the majority of judges on the UK Supreme Court have continued to restrict the operation of tort in this sphere (Godden-Rasul and Murray, 2024), placing considerable weight on the concern that the public interest will not be served by police practice being shaped by the prospect of civil claims (para. 121, Lord Toulson), despite a dearth of evidence to support this point (para. 186, Lord Kerr (dissenting)). The Supreme Court instead permitted a Human Rights Act claim to proceed in Michael on the basis of the police’s positive obligations. This might appear to be difficult to square with the majority’s reasoning with regard to tort, but there appears to have been an expectation that the constraints upon positive obligations and the limitations upon damages in Human Rights Act claims would lessen the impact of this approach upon police practice (Hyde, 2016). If this was the Supreme Court’s position, it was to shift in the following cases.

4 Public order policing in Northern Ireland

The Northern Ireland conflict left behind the legacies of both widespread use of political violence and human rights abuses. By 1998, large sections of society in Northern Ireland lost confidence in the authorities’ commitment to administer justice in an impartial manner, a distrust borne out of miscarriages of justice, extra-judicial killings, and collusion between parts of the security apparatus and paramilitary groups (Campbell and Turner, 2008, pp. 376–382). The decades of political violence had also profoundly affected policing in Northern Ireland. The legislation implementing the Patten Commission proposals was intended to provide for a reset for post-conflict policing in Northern Ireland, with the Royal Ulster Constabulary (RUC) being reformed as the Police Service of Northern Ireland (PSNI) (Patten, 1999) and the development of explicit statutory policing duties (Police (Northern Ireland) Act 2000, section 32(1)). New human rights oversight arrangements for Northern Ireland were also put in place through the introduction of the Northern Ireland Human Rights Commission (Northern Ireland Act 1998, section 69).

The Patten policing reforms took effect against a backdrop of heightened inter-community tensions in parts of Northern Ireland through the summer and autumn of 2001 (Jarman, 2002, para. 70). The attacks on pupils and family members making their way to the Holy Cross Primary School in the Ardoyne area of North Belfast during this period was an event which captured world attention, emphasising the precarious state of Northern Ireland’s peace process. Police mounted an operation to ensure children could reach their school by the main entrance using Perspex barriers lined by police and soldiers. The police did not take immediate steps to arrest the groups throwing stones, urine-filled balloons and even explosive devices at the children and their protective screen. This was a deliberate operational choice, grounded in concerns that such an approach would spark even more widespread disorder. The disorder continued for months.

The policing operation at Holy Cross became the subject of prolonged litigation in which one of the families subject to the attacks challenged the adequacy of police efforts to secure their rights to life and not to be subject to inhuman and degrading treatment. The Northern Ireland Human Rights Commission (NIHRC) supported the litigation. Whereas the Patten reforms of themselves would struggle to break down deep-seated mistrust of the police, many thought at the time that the recognition and active assertion of the authorities’ positive obligations might provide a means of building confidence in the new policing arrangements across Northern Ireland’s communities (Bell, 2017, pp. 303–305). In North Belfast in particular, official investigations suggested a widespread desire for ‘prompt and decisive action to prevent potential riot situations escalating’ (Jarman, 2002, para. 75).

The Holy Cross claims were, however, rejected at every stage of the litigation before the domestic courts. In the Northern Ireland High Court, Kerr LCJ was quick to foreground the limits to positive obligations arising from the ECHR and to highlight the broad discretion the police enjoyed in public order situations:

‘Those who had to decide how to deal with this protest were obliged to have regard to the effect that their decisions might have in the wider community. It is not difficult to understand that an aggressive, uncompromising approach to the protest might have been the catalyst for widespread unrest elsewhere. It is precisely because the Police Service is better equipped to appreciate and evaluate the dangers of such secondary protests and disturbances that an area of discretionary judgment must be allowed them, particularly in the realm of operational decisions.’ ²⁷

The House of Lords ultimately endorsed this analysis. ²⁸ The claimants’ vulnerability in the context of the disturbances was mentioned in passing in a number of judgments, with Lord Carswell emphasising that police officers ‘placed themselves as a shield between a hostile and dangerous crowd and a small group of vulnerable people’. ²⁹ But it did not play a significant part in judicial analyses of the claim, other than that of Baroness Hale. ³⁰ For Gordon Anthony, whereas Osman v. United Kingdom had ‘raised many searching questions about state responsibility in the era of the HRA’ the cases which culminated in the House of Lords’ Holy Cross decision, ‘have resolved some of those in the context of policing’ (2009, p. 551). The resolution was, expressly, to dampen the impact of positive obligations on policing of public order situations. For the NIHRC this approach to the policing of public order situations amounted to simply taking the police at their word, ³¹ in marked contrast to other circumstances in which the courts had undertaken active assessment of whether decision makers had ‘walked the rights walk’ and not simply ‘talked the rights talk’ (Knight, 2007, p. 226). Perhaps in part for daring to point this out, the Commission’s intervention was subject to scathing criticism by Lord Hoffmann. ³²

This outcome would be the subject of a subsequent challenge before Strasbourg. ³³ The European Court accepted that the demonstrators’ actions had interfered with the applicants’ Article 3 rights and that the police had been forewarned of this situation. ³⁴ This finding required the Court to address whether the actions of the police fulfilled their duty to take all reasonable steps to uphold these rights. Putting considerable weight on the fact that forty-one officers sustained injuries in their efforts to protect the children, the Court found that the PSNI had fulfilled the positive obligations upon it:

‘[T]he police followed a course of action which they reasonably believed would end the protest with minimal risk to the children, their parents and the community at large. The risks which concerned the police were not, in fact, purely speculative. Violence had been erupting throughout the city over the summer, often at great speed and with little prior warning. Moreover, the police had intelligence which suggested that a more direct approach could increase the risk to the parents and children walking to the Holy Cross School, lead to further attacks on Catholic schools and also result in increased violence in north Belfast’ (PF and EF v. United Kingdom, para. 43).

This decision illustrates that, when it comes to the positive obligation to protect individuals under Articles 2 and 3, the Court will often seek evidence of a considered and meaningful response by authorities, but suggests that once this has been demonstrated the Court will not be drawn into the operational choices confronting police.

As in the domestic courts, the relegation of the claimants’ vulnerability to the margins of this decision underlines that Strasbourg was paying attention to another factor which has shaped parts of its case law. The Court framed the facts of Holy Cross as a ‘highly charged community dispute’ (para. 45). ³⁵ James Sweeney identified a transitional element in the application of the margin of appreciation to post-Communist states following their accession to the Council of Europe; ‘the discretion devolved to the respondent states in each was checked in order to guarantee loyalty to the concept of human rights thinly constituted’ (Sweeney, 2005, p. 474). In much the same way, the Holy Cross case granted a degree of leeway to the authorities in light of the state of Northern Ireland’s peace process. This is perhaps best summarised by Baroness Hale’s conclusion in the House of Lords, subsequently quoted by the European Court, that it is ‘all too easy to find fault with what the authorities have done, when the real responsibility lies elsewhere’. ³⁶ Although this might superficially appear to be an understandable sentiment, it nonetheless undermines the very concept of the police being subject to positive obligations to intervene against human rights abuses.

Notwithstanding this outcome, Holy Cross marked a turning point in the human rights context of the policing of public order situations in Northern Ireland. Prior to this decision, the interests of individuals affected by public disorder were subsumed into the general police duty to uphold law and order. After Holy Cross, despite the leeway which the courts afforded to the police in the exercise of their duty in the specific transitional context of Northern Ireland in 2001, the legal rights of those affected by disorder became an increasingly important element of police decision making, with consideration of specific vulnerability being a required part of this analysis. ³⁷ Beyond this particular case, the courts increasingly emphasised that in public order cases ‘operational discretion is not sacrosanct … [and] cannot be invoked by the police in order to give them immunity from liability for everything that they do’. ³⁸

In 2012 Belfast City Council took the decision to limit the number of days on which the Union flag would be flown over Belfast City Hall each year. Demonstrations over this decision in late 2012 and early 2013 spilled over into violent clashes in the Short Strand area of Belfast. A judicial review was launched by a Short Strand resident which challenged the refusal of the police to prevent the demonstration marches, in line with the Public Processions (Northern Ireland) Act 1998’s criminalisation of organising or participating in unnotified processions (Public Processions (Northern Ireland) Act 1998, section 6(7)). At issue was whether the approach adopted by the PSNI breached the positive obligations they owed under Article 8 ECHR to protect the claimants’ private and family life. Leaning upon Holy Cross, the PSNI intoned that they ‘have considered the impact of the protests upon the general and Short Strand community as well and the rights of the protestors, and sought to strike a balance based upon their analysis of how best to police the situation’. ³⁹ This gave all the appearance of a boilerplate assertion and little effort was made to substantiate the claim before the High Court. Treacy J noted that ‘there is a complete lack of information as to why the police repeatedly permitted violent loyalist “protestors” to participate in illegal marches both to and from Belfast City Centre on every Saturday between 8 December and 14 February and why they permitted those marches to pass an interface’. ⁴⁰ Considering the importance of the strictures imposed on processions in Northern Ireland, he concluded that the police operation was characterised by ‘unjustified enforcement inertia’. ⁴¹

In reversing this decision, the Northern Ireland Court of Appeal reaffirmed the importance of ‘the enormous difficulties for those policing modern societies in circumstances of community conflict and heightened tension’. ⁴² When the case subsequently reached the Supreme Court, the appellants must have thought the worst when Lord Kerr, the first judge to hear the Holy Cross claims, opened his judgment of the Court with the acknowledgement that ‘[t]he series of demonstrations and marches that the protest involved presented the Police Service with enormous, almost impossible, difficulties’. ⁴³ Notwithstanding those difficulties, Lord Kerr found that the ‘absence of a more proactive approach was due to a concatenation of unfortunate circumstances’. ⁴⁴ High on the list of those circumstances was that the PSNI had placed ‘too great an emphasis on the possible article 11 rights of protesters’ relative to their obligations to the Short Strand residents. ⁴⁵ There is much that is left in the white space of this judgment and there is no explicit reference to vulnerability within it. Instead, Lord Kerr concluded that given those established facts from the hearing the Court of Appeal should have been more reticent about reversing Treacy J’s decision. ⁴⁶ The Supreme Court’s endorsement of the first instance decision nonetheless affirms the need for the police to do more than acknowledge the positive obligations they owe to those affected by disturbances, especially in the context of offences relating to unnotified processions which were enacted in recognition of acute vulnerabilities in the context of society in Northern Ireland.

It is therefore difficult to make sense of this outcome without concluding that the judges involved had moved beyond orthodox conceptions of the active rights holder being protected against malign state intervention. More than that, the circumstances of this case suggest that a rich account of vulnerability was at work, one that recognised that the claimants might not be inherently vulnerable because of particular characteristics, but that their rights nonetheless needed the protection of the state in this instance. This indicates that vulnerability was being conceived, as Fineman advocates, as manifesting along a spectrum, and not solely being the product of the sorts of special characteristics identified in much of Strasbourg’s vulnerability jurisprudence and at work in Holy Cross (Fineman, 2008). It was, significantly, reached with little groundwork for the Supreme Court to draw upon in Strasbourg jurisprudence. But as he made clear in other judgments, notably his dissent in Ambrose, Lord Kerr saw little impediment in the limited account of vulnerability in Strasbourg jurisprudence; ‘[i]f the much vaunted dialogue between national courts and Strasbourg is to mean anything, we should surely not feel inhibited from saying what we believe Strasbourg ought to find in relation to those arguments’. ⁴⁷

Although these theoretical underpinnings are not explicit, they are inherent in Lord Kerr’s judgment. Indeed, after the stark divide in the Court in Michael, he likely thought the better of articulating these underpinnings of his position. Had he fully articulated his theoretical approach to positive obligations it is difficult to see how, given Lord Hughes’ subsequent position in DSD, which we explore below, he could have carried the Court with him in a unanimous decision. As with the relationship between proportionality and irrationality, ⁴⁸ the wait for a new dispensation to be agreed by the Supreme Court might be long, and it is tempting to think that Lord Kerr did not think this worth the wait. Under his direction the Court, for all that it recognised the complexity of public order policing in Northern Ireland, was prepared to assert positive obligations in a way that made the police’s duties meaningful. That this outcome was reached in the context of an Article 8 ECHR claim, in which the qualified nature of the right makes the realisation of positive obligations more difficult than under Articles 2 and 3 ECHR, makes this outcome even more significant. While this decision does not reflect a full ontological shift to a jurisprudence based upon a vulnerable subject, DB nonetheless saw the Court move beyond an account grounded in vulnerable groups and begin to rethink the boundaries of the liberal subject.

5 Police investigation obligations

Policing investigations into reports of sexual violence have a longstanding track record of failing to meet standards set and of treating complainants poorly and causing them further harm. It is part of the reason why many sexual violence survivors – potentially as high as 95 percent – do not make a police report (Hohl and Stanko, 2015; Ministry of Justice, Home office and Office for National Statistics, 2013; Stern, 2010). For those who do make a complaint to the police, it is likely their case will not go beyond the investigation stage (Kelly, 2002; Kelly, Lovett and Reagan, 2005; Ministry of Justice, Home Office, and Office for National Statistics, 2013, para. 324). Although there have been significant improvements in policies and practices over the last few decades, disrespect, disbelief, privacy invasions and harmful policing practices continue to plague sexual violence cases in the criminal justice system (McGlynn, 2010). As such, victims’ families and survivors have turned to the law to hold the state accountable. However, the UK courts have protected the police from negligence claims on the basis that it is the police who are in the best position to make decisions about resource allocation, priorities, and investigation practices. ⁴⁹ While human rights have offered some possibilities for challenging policing practices, the high thresholds have typically prevented cases succeeding. ⁵⁰ The case of Commissioner of Police of the Metropolis v. DSD and another, however, may have paved the way for more successful challenges against the police and to hold the state to account. The UK courts have achieved this by beginning to reorientate their focus to a more vulnerable subject.

In DSD the key issues were whether the state has a duty, beyond providing adequate legal structures, to investigate Article 3-level harm inflicted by non-state actors and, if so, what the scope of the duty is – specifically, whether operational failures alone suffice to breach Article 3. The majority of the Supreme Court held that there is such a duty and that a breach could be established by egregious operational failures alone (with Lord Hughes dissenting on this point). The facts of DSD and the Metropolitan Police Service’s catalogue of failings are detailed in a first instance judgment by Green J which powerfully illustrates the consequences for sexual violence survivors of police investigative ineptitude and systemic faults. ⁵¹ The applicants, DSD and NBV, were victims of the ‘black cab rapist’, John Worboys, whose modus operandi was to drug women passengers in late night taxi journeys. He would then sexually assault them and leave them intoxicated, disorientated and confused, sometimes with little or no memory of what had happened. They were not, sadly, his only victims. By 2008 when Worboys was apprehended there was evidence that he had carried out over 105 sexual assaults from 2002. However, his arrest was only a result of a routine search in which the Metropolitan Police Service identified connections between a series of sexual assault allegations. The connections were not made earlier, and assaults not prevented, due to extensive systemic and operational shortcomings. Police officers regarded DSD as a drunk or an addict when she first met with them shortly after the assault, against specific policies on drug-facilitated sexual assaults. The police did not record complaints as serious sexual assaults, did not conduct appropriate searches, and did not follow up on evidence. There was improper training, improper supervision and management, and resource misallocation (paras. 247–284 (systemic failures) and 289–311 (operational failures)). That there had been serious systemic and operational issues was never in doubt.

The judgments in DSD are predicated upon identifying the bases for applying Article 3 investigative duties. On one level, the Strasbourg case law is relatively strong. Green J at first instance, Laws LJ in the Court of Appeal and Lords Kerr and Hughes in the Supreme Court all undertake extensive analysis of this case law. Drawing upon MC v. Bulgaria, the Supreme Court majority accepted that ‘states have a positive obligation inherent in Articles 3 and 8 ECHR to enact criminal-law provisions, effectively punishing rape and to apply them in practice through effective investigation and prosecution’. ⁵² The Supreme Court majority affirmed Laws LJ’s conclusion that there is a ‘clear and consistent line of authority’. ⁵³ On a deeper level, however, the underlying justifications are somewhat murkier. In Lord Hughes’ view the ‘European Court of Human Rights starts from a solidly rationalised principle, but then extends it to situations to which the rationale does not apply, without overt recognition of the extension, without formulating any fresh rationale and relying on supposed authority which does not actually support the extension’. ⁵⁴ As such, anything beyond finding the state breached Article 3 where torture or inhumane or degrading treatment is inflicted by a state agent against a citizen is ‘a judicial gloss on the convention’. ⁵⁵ Similarly, Lord Mance concludes that the ‘foundations and rationale’ underlying this case law are ‘shaky’. ⁵⁶ In the face of such discontent, it must be remembered that the rationale for extending the positive obligation is to ensure rights are meaningful. ⁵⁷ Indeed, in a subsequent case involving a police investigation of a rape complaint in Northern Ireland, McAlinden J reaffirmed that a sufficient investigation is necessary to ensure that the right under Article 3 is ‘practical and effective’. ⁵⁸

Nevertheless, Lords Hughes and Mance do have a point – at least when the issue is looked at from a traditional-liberal-negative-liberty-rights point of view. As Lord Hughes says, ‘It is perfectly clear that the primary case of behaviour that breaches Article 3 is where torture or inhumane or degrading treatment is meted out by the state against a person within its jurisdiction.’ ⁵⁹ Read on a traditional understanding of human rights in which citizens need protection from an encroaching state, extending human rights obligations to investigate Article 3-level harm perpetrated by a private individual where the state is not directly involved is arguably a stretch, and especially in relation to an adequate operational response, not just adequate legal structures. Indeed, it has been argued it is possible to derive a positive obligation from the specific wording of Article 2 but not Article 3. ⁶⁰ The Strasbourg Court has nonetheless maintained that the requirements for an effective investigation into Article 3 level harms stem from the case law on effective investigations for breaches of Article 2, ⁶¹ even if it has recognised the different content and underlying rationales of Articles 2 and 3 (Y v. Bulgaria). To be clear, this is not to say that there is no rationale or justification for the development of the Article 3 positive obligation case law and the decision in DSD. But it is to say that this body of jurisprudence is pushing at the boundaries of the Convention rights guaranteed to the liberal subject. While the conception of the human at the heart of this case law is not a fully formed vulnerable subject, the approach nevertheless leads away from the liberal tradition and towards Fineman’s more responsive state.

Before the Court of Appeal in DSD, Laws LJ provides a judgment which emphasises the ‘large canvas’ of values and the ‘fundamentals of a civilised constitution’ embraced by the ECHR (DSD v. Commissioner of Police of the Metropolis, para. 11). Broadly speaking it is about the ‘the rule of law, and the security and protection of people’ (DSD v. Commissioner of Police of the Metropolis, para. 11). Moreover, ‘because the focus of a human rights claim is not on loss to the individual but on the maintenance of a proper standard of protection, the Court is in principle concerned with the State’s overall approach to the relevant ECHR obligation’ (DSD v. Commissioner of Police of the Metropolis, para. 67). As summarised by Joanne Conaghan, Laws LJ is expressing an ideal that ‘normative fluidity’ is part of the nature of human rights laws (Conaghan, 2017, p. 69). This decision is an example of what Moya Lloyd argues is the receptivity of human rights law and its concept of the human to contest, thereby rearticulating ‘those normative conceptions of the human that produce, through an exclusionary process, a host of ‘‘unlivable lives’’ whose legal and political status is suspended’ (2007, p. 100; quoting Butler, 2004, p. xv). Perhaps then, pushing at the boundaries of positive obligations is uncovering a juridical space which allows for alternative conceptions of the human and, in particular, a vulnerable subject.

From first instance in DSD there are references to the notion of vulnerability. Following a detailed overview of the case law, Green J said that whether the applicant was in a ‘vulnerable category’ is a ‘contextual factor’ which may play a role, for example, in determining whether the level of harm has met the Article 3 threshold or whether or not there was an adequate legal system in place. ⁶² Green J does identify the claimants as vulnerable and notes the claimant argued sexual assault victims (mostly women) are a vulnerable group, and that this adds weight to his argument that the state had breached the applicant’s Article 3 rights. ⁶³ If this is where he left this point, it could be chalked up to an endorsement of the European Court’s group-based vulnerability approach and all of its shortcomings. However, Green J continued that it was not necessary to base his findings on the claimants being in a vulnerable group. More significantly, he maintains that ‘almost by definition’ victims of Article 3-level inflicted harm will be vulnerable and, for this reason, ‘[v]ulnerability may be therefore seen as part of the rationale for the existence of the duty in Article 3’. ⁶⁴ Note that Green J does not link this to violence inflicted by a state agent. This approach aligns with Butler when she says:

‘Violence is … a way a primary human vulnerability to other humans is exposed in its most terrifying way, a way in which we are given over, without control, to the will of another, a way in which life itself can be expunged by the wilful action of another … This vulnerability, however, becomes highly exacerbated under certain social and political conditions …’ (2004, p. 13).

This could be read in accordance with the idea that vulnerability is part of the human condition. Any person could be subject to Article 3-level harm by another throughout the embodied life of humans, but the state is implicated in ameliorating and preventing or aggravating and perpetuating these kinds of harms. Put simply, the state is accountable for the unequal distribution of violence. State action or inaction can make some more vulnerable to violence than others. Following on from this, from Fineman’s perspective, a positive duty on the state to adequately investigate crimes inflicting Article 3-level harm makes perfect sense (Fineman, 2008, p. 13).

In the Court of Appeal and the Supreme Court, however, the words ‘vulnerability’ or ‘vulnerable’ are not used. ⁶⁵ There are detailed examinations of the Article 3 case law with the UKSC majority concluding that there is sufficient authority for serious and egregious operational investigation failures alone to breach Article 3 (Lord Hughes dissenting). However, all Supreme Court Justices agreed that the facts very clearly include both systemic and operational failures so there was no need to interrogate in depth what kinds of failures they were, if they were related, and how serious the operational failures were. ⁶⁶ Instead, any details, descriptions and discussions in this regard are mostly quoted from Green J’s judgment. As Lord Kerr explains, ‘It is unnecessary to list all the operational failings. These are set out in admirable and clear detail by the trial judge in his judgment. It is sufficient to refer to a sample of these to explain why I do not accept that these were largely attributable to a flawed structural approach.’ ⁶⁷ Similarly, in the Court of Appeal Law LJ quotes extensively from Green J when setting out the systemic and operational failures of the appellant. ⁶⁸

This is to be expected. The first instance judge does have a different role to the appellate courts and hears the evidence in full. It could be for this reason that Green J’s judgment elucidates the role of vulnerability more clearly. A vulnerability approach necessitates a careful consideration of context to account for the underlying structural oppressive relations which shape the problems in the structures and practices of a legal system. Indeed, Fineman highlights the need for a focus on societal institutions and the individual in their social context when taking a vulnerability approach (Fineman, 2008, p. 13). This can be seen in Green J’s judgment in the way he explains the structural and operational issues. For example, in relation to the structural issues, he highlights the lack of training despite the fact that there were ‘sophisticated guides’ with ‘mature thinking’ about investigating sexual assault, ⁶⁹ which is needed in cases of sexual violence which are underpinned by gender stereotypes and rape myths. ⁷⁰ In terms of the operational failures, for example, he says there was ‘misidentification and misplaced scepticism’ and an automatic assumption that the cab driver was highly reliable. ⁷¹ There was evidence that the claimants did not feel supported or believed and there was an inappropriate focus on the position of the victim rather than the behaviour of which was complained. ⁷² These are just a few of the many problems which have been extensively documented in feminist scholarship and official reports, linked to the underlying gender and other power relations upon which sexual violence is predicated (Brown et al., 2010; Kelly, 2002). It is a reason why negative human rights obligations are insufficient to address the state involvement in maintaining the extent and distribution of sexual violence (Marshall, 2008b, p. 149). Due to their specific functions, the appellate courts eschewed this level of contextual analysis which may in part explain the lack of clarity on whether a vulnerable subject is mobilising the judgments of the majority. As with Green J’s judgment, considerable attention to context and a foregrounding of the experiences of the subject is necessary to expose the vulnerability approach; a doctrinal analysis alone is insufficient.

Neither the judges of the Court of Appeal nor the Supreme Court judges may have wanted to explicitly articulate that human rights principles are shifting. Indeed, acceptance of the positive obligations on states remains subject to intense debate. So, going further by recognising a human subject who has a more complex and dependent relationship with the state and social relations than has historically been judicially recognised would be bold. Such judicial development could provoke a backlash and a rolling back of current human rights protection. Lord Kerr was therefore coy about bringing more radical underpinning justifications for positive obligations to the fore. If he had done so, perhaps Lord Hughes would not have been in the minority. Regardless of what is on DSD’s surface, the undercurrent is one which pulls in the direction of a vulnerable human rights subject.

Of course, DSD raises questions about the scope and limitations of the operational duty. The majority make clear that isolated operational failures will not constitute a breach of Article 3; rather, the failures must be serious and egregious. ⁷³ In Lord Kerr’s view this is unlikely to ‘herald an avalanche of claims for every retrospectively detected error in police investigations of minor crime’. ⁷⁴ In addition, Lord Kerr highlights that Strasbourg jurisprudence emphasises that the ‘duty is one of means, not one of result’, ⁷⁵ so that, Lord Neuberger explains, judges are not imposing on legitimate areas of police discretion and political resource allocation. ⁷⁶ Generally speaking, there is less detailed analysis of the implications of extending positive duties on the police in DSD – and in Strasbourg jurisprudence more widely – than in the extensive body of tort law cases on the topic. ⁷⁷ In terms of tort cases, judicial authority currently holds that a general investigative duty to the public would, as Lord Hughes summarises, ‘inhibit the robust operation of police work, and divert resources from current inquiries; it would be detrimental, not a spur, to law enforcement’. ⁷⁸ However, because of the private nature of tort law and the public function of human rights law Lord Neuberger and Lord Kerr say different approaches to policing can be justified. ⁷⁹ Put another way, in tort cases judges have tended to defer to police expertise, whereas it is a role of human rights liability to set policing standards. ⁸⁰ On the facts, Green J found that if the operational failings had not occurred the police officers involved in the investigation ‘would have taken steps which would have been capable of identifying and arresting Worboys’. ⁸¹

The decision in DSD, we argue, cannot be squared with the traditional liberal subject of human rights. It is for this reason that Lord Hughes provides a sharp (partial) dissent and Lord Mance has concerns about the foundations of the extension of Article 3 obligations. As in Flags Protest, Lord Kerr did not articulate a theoretical basis for the positive obligations under Article 3 in DSD. Nonetheless, a conceptualisation of the human as inherently vulnerable must be at work within the decision, building upon the first instance judgment of Green J. This approach to the ‘large canvas’ of ECHR values – as identified by Laws LJ in the Court of Appeal – pays attention to the complex relationship between the state, individuals, and social and structural conditions for violence.

6 Conclusion

The position of the UK’s most senior judges evidently shifted from Holy Cross, in which they were averse to interrogating police operational decisions and practices in the context of Article 3 claims, to DSD, which established that this right can be breached by serious and egregious operational failures. Nevertheless, Holy Cross marked the tentative beginnings of more substantive human rights oversight of policing and required these judges to consider vulnerability in new contexts. Indeed, this shift is reflected in Re DB’s Judicial Review in which the Supreme Court’s decision requires more from the police than a nod in acknowledgement of their positive obligations. This culminated in DSD, in which the Supreme Court recognises that to protect individuals from breaches of Article 3 there must be a positive duty to adequately investigate crimes. The context of our case studies matters because public consciousness and legal norms have long diverged as to the extent to which the police are subject to enforceable duties, and the concept of vulnerability is playing a significant role in reshaping the extent of these duties. The cases we have focused upon, moreover, matter because they extend the jurisprudence on vulnerability beyond narrow contexts in which certain individuals or specific groups have been identified by the courts as inherently vulnerable (as seen in rights jurisprudence related to domestic violence). The approach emerging within these cases recognises that anyone is vulnerable to such harms, accompanied by an acknowledgement that some are more vulnerable than others, it is implied, because of social, institutional, and political structures.

The Re DB’s Judicial Review and DSD outcomes are, in short, not compatible with traditional understandings of the liberal ‘rights bearing’ subject being protected from abuses of power by state actors. Instead, these decisions are best understood as giving shape to a vulnerable subject – even if not fully formed – who requires positive action on the part of the state to address the incidence and unequal distribution of violence. The Inspectorate of Constabulary report into the policing of the Sarah Everard vigil illustrates the extent to which public order policing is now conceived in terms of a combination of positive and negative rights protections, even if not always convincingly. The report notes that the police misinterpreted protections for protest rights against the backdrop of the Coronavirus pandemic, but puts this alongside an extended discourse on the Peelite duties of impartial policing and the responsibility of the police to protect the public as a whole (HM Inspectorate of Constabulary and Fire and Rescue Services, 2021, p. 10–11). The public response to the murder of Sarah Everard at the hands of a police officer highlights the impact on women of failings in policing in relation to violence against women. This begs the question which members of the ‘public as a whole’ the police are protecting. For all that the language of police public duties have long been used as a shield for police conduct, the DSD approach to positive duties has found a way past that shield. There is now an opportunity to interrogate and challenge the standards of a police investigation on rights grounds. Positive obligations, expressly limited in terms of the rights of others, provide a vehicle for doing which misty-eyed recitations of Peel’s supposed duties have too long denied.

These outcomes can be explained as partly coinciding with Fineman’s thesis and come amid scholarship encouraging the Strasbourg Court to recognise her account as more compelling than its piecemeal jurisprudence on vulnerable groups (Heri, 2021, p. 28). These approaches, however, have prompted accusations that the European Court has overstepped its bounds in developing positive obligations, complaints which showcase the ongoing push-back against these elements of human rights jurisprudence; ‘it isn’t interpretation to play Haydn when the concert programme and the score call for Mozart’ (Larkin, 2020, p. 15). For all that we identify a different conception of human relationships as inherent in this jurisprudence, its impact is by no means as transformative as such critiques suggest. Our exploration of these cases demonstrates that an approach to positive obligations grounded in vulnerability does not threaten the protections of others’ human rights, but rather ensures accountability within a criminal justice system in which rights are under threat. The focus of existing Strasbourg jurisprudence on criminal justice responses is potentially limiting (in terms of how it conditions carceral state responses), ⁸² but could simply be indicative of a jurisprudence which is still developing. This approach, whatever its shortcomings, does not mandate the form that police action must take or obligate criminal convictions as a mode of response to particular circumstances. Positive obligations instead provide limited requirements which have their most substantive effect where human vulnerability – and its social and political dimensions – provides the basis for breaches of human rights.