Do Guarantees of Non-Recurrence Actually Help to Prevent Systemic Violations? Reflections on Measures Taken to Prevent Domestic Violence

This article assesses the effectiveness of guarantees of non-recurrence to prevent the recurring, systemic violation of domestic violence. I consider whether there is any evidence to demonstrate that the measures ordered in response to state due diligence failings concerning domestic violence produced a change that that can be understood as effectively contributing to prevention. This is done principally by reviewing reports on the implementation of domestic violence judgments and decisions issued by the Inter-American and European regional human rights systems and by the UN CEDAW Committee. However, as is shown, for the most part these bodies have carried out only superficial scrutiny of the effectiveness in practice of measures undertaken by states to guarantee non-recurrence. Thus, I identify a number of areas which merit further cross-disciplinary study.


Introduction
Guarantees of non-recurrence are steps taken in response to a violation to prevent them from happening again to the same or other victims similar to them. They form part of states' obligation to ensure continuous and effective protection of human rights 1 and have been recognised as one of the forms of reparations ordered by a court, arbitral body or other adjudicative body, 2 in response to a finding of responsibility for a human rights violation when there is a risk of repetition and when reestablishment of the prior situation is not considered sufficient. Under the law of state responsibility, these guarantees can also form part of a state's obligation to address responsibility for an internationally wrongful act. 3 Guarantees of non-recurrence have also been incorporated by governments into transitional justice and other forward looking reform processes. 4 Much has been written about the nature of a state's obligation to afford guarantees of non-recurrence and the content of such guarantees. 5 Elsewhere, I have written on the extent to which guarantees of non-recurrence ordered as part of a judicial remedy can contribute to broader societal objectives related to widespread or systemic violations. 6 It is not obvious for a court judgment to address simultaneously the needs of victims affected by the conduct which forms the basis of the award and the much wider, societal factors that gave rise to that conduct. 7 Part of the challenge stems from the fact that usually, the reparations ordered (even for guarantees of nonrecurrence) derive from recommendations put by individual petitioners whose vantage point is inevitably the specific violations they suffered as opposed to, necessarily, the wider, systemic or structural problems which gave rise to those violations.
In this article, I assess the effectiveness of guarantees of non-recurrence ordered to prevent the recurring, systemic violation of domestic violence. Domestic or intimate partner violence is physical, sexual, emotional, economic or psychological actions or threats of actions that frighten, intimidate, terrorize, manipulate, hurt, humiliate, blame, injure, or wound, and can progress to killings. A 2018 study by the World Health Organization indicates that 26% of women worldwide have been subjected to physical and/or sexual violence from a current or former husband or male intimate partner at least once in their lifetime. 8 Domestic violence is a gendered phenomenon that invariably targets women but can happen to anyone of any race, age, sexual orientation, religion, or gender. I consider whether there is any evidence to demonstrate that the measures ordered in response to state due diligence failings concerning domestic violence produced a change that that can be understood as effectively contributing to prevention. Courts tend to order measures to guarantee non-recurrence with only limited regard to the prospects for them to be effective. The bodies deciding whether those measures have been implemented likewise have had only limited regard to whether the measures adopted have been effective in practice. This is a topic that has not been considered sufficiently in the literature, partly because of the inherent methodological challenges. Understanding whether guarantees of non-recurrence have been effective requires an assessment of the extent to which the measures impacted states' behaviour, and whether the violations stopped or were significantly reduced. This can be hard to assess in real-time. It is difficult to isolate the variables to determine whether a measure was actually successful in its goal of prevention, or whether the absence of recurrence was due to unrelated factors or, conversely, whether a recurrence was due to the ineffectiveness of preventive measures, or unrelated factors. Particularly challenging is the process of establishing a causal link between the measures and their intended short-term and longerterm outcomes.
I begin my analysis by considering notions of effectiveness and thereafter applying these to measures ordered with the aim of guaranteeing non-recurrence for domestic violence. I then assess the methodological challenges, considering whether the bodies mandated to assess the effectiveness of the measures ordered as guarantees of non-recurrence or adopted as part of the enforcement of the judgment have engaged in any evaluation of the outcomes, whether by considering if the ways in which the measures were implemented impacted on their effectiveness and whether there was any assessment of the outcome (e.g., reduction in scale or intensity of violations; greater clarity of obligations; increased compliance with obligations; more systematic sanctions for non-compliance). This is done principally by reviewing reports on the implementation of domestic violence judgments and decisions issued by the Inter-American and European regional human rights systems 9 and by the Committee established to monitor compliance with the Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW Committee).
The use of social science methods to scrutinise the short and long-term effectiveness of measures ordered to guarantee non-recurrence is warranted. 10 The goals of this paper are more modest. I consider the extent to which decisions on whether judgments involving guarantees of non-recurrence have been implemented are sufficiently datadriven. For the most part they are not. Thus, I identify a number of areas which merit further cross-disciplinary scrutiny which may be relevant for the ongoing assessment of the execution of judgments involving guarantees of non-recurrence.

What Makes a Measure 'Effective'?
Effectiveness in human rights parlance is most commonly applied to the procedural aspects of rights. Most human rights treaties and declarative texts recognize that when an obligation is breached, there is a right to an effective remedy. 11 A remedy is understood as effective when a person claiming to be a victim of a human rights violation has access to independent and competent authorities that are capable of determining the truth of what happened and fairly deciding upon a claim of violation of their rights. 12 Remedies must be prompt, accessible and capable of offering a reasonable prospect of success. 13 A remedy must also be capable of adequately addressing the rights violation and able to grant appropriate relief. 14 Moreover, to be effective, the afforded relief cannot be discretionary; the competent authorities must be obliged to enforce and implement reparations when ordered.
Notions of effectiveness have also been applied to the substance of the relief -the reparations measures themselves, though what is understood to be effective invariably has been framed in general terms.  Russia, ECtHR, Appl. No. 42502/06, 11 December 2008, para. 100;The Social and Economic Rights Action Center and the Center for Economic and Social Rights v. Nigeria, Comm. No. 155/96 (ACommHPR, 30th Ord Sess, 13-27 October 2001);Silver v. United Kingdom, Appl. Nos. 5947/72, 6205/73, 7052/75, 7061/75, 7107/75, 7113/75 and 7136/75 (ECtHR, 25 March 1983)  the quantum and quality of inter-State claims is 'full,' as needing to wipe out all the consequences of the illegal act and reestablish the status quo ante. 15 Human rights treaties and declarative texts and the judgments and opinions that interpret these texts tend to use descriptors such as fair, adequate or effective, used either singly or grouped together, 16 appropriate, 17 proportionate to the harm 18 and equitable. 19 These help to clarify what is required, particularly when re-establishing the status quo ante is impossible and it is impractical to precisely quantify the harm. 20 Guarantees of non-recurrence have been incorporated into this analysis. The UN Human Rights Committee in its General Comment 31 explains that '… the purposes of the Covenant would be defeated without an obligation … to take measures to prevent a recurrence of a violation of the Covenant. … Such measures may require changes in the State Party's laws or practices.' 21 As was stated by de Greiff, then Special Rapporteur on the promotion of truth, justice, reparation and guarantees of non-recurrence, '[b]ecause guarantees of non-recurrence are a function that can be satisfied by diverse measures, there is no such thing as a general non-recurrence policy that will be equally effective in all contexts. An effective policy designed to prevent systemic violations will need to adjust form to function and choose the proper measures.  [152]. See also ARS Arts 31, 34 and commentaries thereto; Basic Principles and Guidelines, 18, which describes 'full and effective' reparation. Note that in the LaGrand case, which concerned breaches of Art. 36 or the Vienna Convention on Consular Relations, Germany requested a guarantee that the opposing party would not repeat these violations in the future and would ensure that both its domestic law and practices allowed for an effective exercise of the rights included in the Convention. The ICJ acknowledged the existence of specific guarantees that would trigger amendments in laws or policies, but does not explain what particular steps would need to be taken to comply with an order of guarantees of non-recurrence [LaGrand Case (Germany v Diverse forms of guarantees of non-recurrence have been ordered or have formed part of the consideration of the execution of judgments in domestic violence cases, with variable consideration given as to what would make such measures effective, other than that they should reflect the particular contexts of the violations.

Domestic Violence as a Recurring and Systemic Violation
This section considers the practise of domestic violence as a recurring and systemic violation. Following a brief introduction, the section considers the practise of the Inter-American Commission and Court of Human Rights, the European Court of Human Rights and the CEDAW Committee.
Recurring, systemic violations stem from significant, usually long-term and intersectional problems, which are both complex and structural in character. Often the violations comprise composite acts, 'an accumulation of identical or analogous breaches which are sufficiently numerous and inter-connected to amount not merely to isolated incidents or exceptions but to a pattern or system.' 23 Sometimes, recurring or systemic violations are so engrained that they are not thought of as violations and are therefore not subject to adjudication because the conduct has become normalised. This phenomenon of normalisation is often present in domestic violence cases. 24 Domestic violence is pervasive in all parts of the world and is rooted in sex and genderbased discrimination. 25 The state is responsible for such violence where it has failed to exercise due diligence to prevent and respond adequately to such acts occurring. When its responsibility is engaged, the state is obliged to afford reparations to remedy its failings.
Owing to the systemic and structural factors that underpin domestic violence as an embodiment of violence against women, there is a need for reparations to address the unequal and discriminatory social structures that contributed to the violence so as to avoid recurrence. 26 Thus, reparations should be 'transformative'; it 'must go above and beyond the immediate reasons and consequences of the crimes and violations; they must address structural inequalities that negatively shape women's and girls' lives'. 27 Nevertheless, the detail of what precise steps states must take to meet this obligation of transformation is lacking. This is a problem that was also noted by Rashida Manjoo, when serving as UN Special Rapporteur on Violence Against Women: Treaties that establish obligations to protect against rights violations often require States to take appropriate measures, without explicitly defining what measures are 23 Ireland v. The United Kingdom, App. No. 5310/71 (ECtHR, 13 December 1977) para. 159. 24 Copelon (1994 appropriate. Similarly, due diligence standards require States to exercise whatever diligence is due; they do not define such diligence. 28 Manjoo indicates that despite the acknowledgment by most Member States that violence against women is the most prevalent human rights violation facing countries, 'this acknowledgement has not led to the adoption of necessary solutions that are coherent and sustainable, and which would lead to elimination of all forms of violence against all women. In fact, the view from civil society is that the prevalence rates are increasing and also manifesting in new forms in many parts of the world. Also, that impunity for both perpetrators and State officials who fail to protect and prevent violence against women continues to be the norm.' 29 In the field of transitional justice, Gilmore, Guillerot and Sandoval argue that 'transitional justice experiences are yet to deliver on guarantees of non-repetition with a gender angle in order to subvert inequality and are often outside [domestic reparation programmes]'. 30 In cases involving spousal or partner violence against women, courts and/or the bodies tasked with assessing the implementation of judgments, have ordered or approved respectively, a variety of measures aimed at guaranteeing non-recurrence.

The Inter-American Commission and Court of Human Rights
Domestic violence has received regular attention within the Inter-American system. The Convention of Belém do Pará was the first regional convention to outlaw specifically violence against women including physical, sexual and psychological violence 'that occurs within the family or domestic unit or within any other interpersonal relationship, whether or not the perpetrator shares or has shared the same residence with the woman, including, among others, rape, battery and sexual abuse'. 31 In 2014, the Committee of Experts of the Follow-up Mechanism to the Belém do Pará Convention (MESECVI) issued a guide on the implementation of the Belém do Pará Convention which provided detailed analysis of the types of measures required to guarantee non-recurrence for structural and systemic violations. 32 It noted that '[w]hat is needed are results-based indicators of how these State policies can become reparations from a gender perspective, in that: (i) they question and are able to modify the status quo that causes and maintains violence against women, through special measures; (ii) they constitute clear progress in overcoming the formal and de facto legal, political, and social inequalities that cause, foster or reproduce gender-based discrimination; and (iii) they 28 Ibid., para. 16. 29 Ibid., para. 43. 30 Gilmore (2020), p. 17. 31 Inter-American Convention on the Prevention, Punishment, and Eradication of Violence against Women (Convention of Belém do Pará) (adopted 9 June 1994, entered into force 5 March 1995), 33 ILM 153433 ILM (1994 Art. 2(a). 32 MESECVI, Guide to the application of the Inter-American Convention on the Prevention, Punishment and Eradication of Violence against Women (Belém do Pará Convention), OEA/Ser.L/II.6.14 (2014). See in particular, sensitize public officials and society on the impact of discrimination against women in the public and private spheres.' 33 The Inter-American Commission on Human Rights has addressed domestic violence in their caselaw, with an important focus on guarantees of non-recurrence to address the structural and systemic aspects of violations.
Maria da Penha Maia Fernandes v. Brazil 34 concerned the abuse and attempted murder of the victim by her husband, which resulted in her becoming a paraplegic. The Inter-American Commission determined that the lengthy delay in the prosecution of Maria's husband violated her right to fair trial and judicial protection. Also, it recognised Brazil's actions as part of a general pattern of tolerance of domestic violence, 35 contrary to Article 7 of the Belém do Pará Convention. In addition to the specific recommendations flowing directly from the violation, the Commission recommended that Brazil adopt measures to demonstrate its condemnation of domestic violence, such as training and awareness raising of the judiciary and specialized police; reducing the timeframe for criminal proceedings; establishing alternatives to judicial mechanisms; and increasing the number of special police stations. 36 Brazil adopted a series of measures, including the promulgation of Law 11,340 of 7 August 2006 (the 'Maria da Penha Law'), which criminalizes domestic and family violence against women, creates special courts on domestic and family violence and offices of ombudspersons for women, among other measures. 37 The Commission classified the implementation of its recommendations as 'partially complied with'. 38 It convened a public hearing in 2011 to discuss the obstacles for the effective implementation of the Maria da Penha Law and in 2019 produced a Technical Opinion to further support implementation. Since then, the Commission has remained engaged in reviewing updates supplied by the State and the petitioners. 39 Jessica Lenahan (Gonzales) v. United States 40 concerned the killings of three girls by their father despite the mother having a restraining order against him in place. She contacted the police repeatedly to report her daughters as missing, that she thought the girls' father took them, and that this was a violation of the restraining order. Many hours later, the father drove to the police station, and there was a shootout between him and the police; the bodies of the three girls were found in his car. The mother filed a suit against the police for failing to enforce the restraining order which was ultimately appealed to the US Supreme Court which held that the police had no specific obligation to enforce a restraining order. 41 The Inter-American Commission determined that.
[t]he state apparatus was not duly organized, coordinated, and ready to protect these victims from domestic violence by adequately and effectively implementing the restraining order at issue; failures to protect which constituted a form of discrimination in violation of Article II of the American Declaration. These systemic failures are particularly serious since they took place in a context where there has been a historical problem with the enforcement of protection orders; a problem that has disproportionately affected women-especially those pertaining to ethnic and racial minorities and to low-income groupssince they constitute the majority of the restraining order holders. Within this context, there is also a high correlation between the problem of wife battering and child abuse, exacerbated when the parties in a marriage separate. Even though the Commission recognizes the legislation and programmatic efforts of the United States to address the problem of domestic violence, these measures had not been sufficiently put into practice in the present case. 42 In addition to the specific recommendations pertaining to the case, the Commission recommended US authorities inter alia to investigate the systemic failures that took place related to the enforcement of the protection order as a guarantee of their nonrecurrence; determine the responsibilities of public officials for violating state and/ or federal laws, and holding those responsible accountable; and to make mandatory the enforcement of protection orders and other precautionary measures to protect women from imminent acts of violence. It also called on US authorities to continue adopting public policies and institutional programs aimed at restructuring the stereotypes of domestic violence victims, and to promote the eradication of discriminatory socio-cultural patterns that impede women and children's full protection from domestic violence, including training and awareness raising. 43 While a range of practical, institution-strengthening, guidance and training measures and legislative reforms were undertaken, the Commission classified the implementation of its recommendations as 'partially complied with' and continues to monitor compliance. 44 It convened a public hearing on implementation on 27 October 2014, with the participation of the parties and the then UN Special Rapporteur on Violence against Women. 45 In both the Maria da Penha Maia Fernandes and Jessica Lenahan cases, the Inter-American Commission has played an important role in identifying relevant measures to guarantee non-recurrence and serving as a catalyst for wider advocacy on the implementation of such measures in both states. In both cases, the Commission recommendations remain only partially complied with. In Brazil, this appears to be a general challenge related to the overall implementation of the 'Maria da Penha Law' whereas in the US, the problem stems more from the apparent limitations of the federal system of government to engage in all the recommended policy and legislative reforms. What is striking however, is the dearth of cases pertaining to domestic violence to reach the Inter-American Court. In the Jessica Lenahan case, as the USA has not ratified the American Convention on Human Rights giving the Inter-American Court competence over it, proceedings were destined to end at the Commission where findings do not have binding effect in the sense of a court judgment. Nevertheless, as shown, the Inter-American Commission has been active in engaging both the USA and Brazil on compliance with its recommendations; this, accompanied by significant civil society advocacy, has helped to foster a continued dialogue on the need to adopt measures to guarantee non-recurrence.

The European Court of Human Rights
The European Court of Human Rights (ECtHR) has had frequent occasion to address domestic violence, finding violations of the right to life, freedom from torture and other prohibited ill-treatment, the right to respect for private and family life and freedom from discrimination in numerous judgments. 46 Following the coming into force of the Istanbul Convention, 47 the ECtHR has taken the stipulations of that Convention into account when assessing positive obligations and the adequacy of preventive measures taken by states parties to that Convention, as well as the recommendations and associated analytical work of the Group of Experts on Action against Violence against Women and Domestic Violence (GREVIO), the independent body responsible for the monitoring of the Istanbul Convention. 48 Nevertheless, and despite recognising the systematic and structural nature of domestic violence in a number of cases, the European Court has refrained from indicating the type of general measures that might be taken in order to put an end to the situation it has found to exist. 49 Thus, discussion on general measures to guarantee non-recurrence has proceeded solely at the Committee of Ministers (CoM), the Council of Europe political organ tasked with supervising the execution of judgments of the European Court of Human Rights.
In Kurt v. Austria, 50 a violent husband against whom a barring and protection order had been issued, killed his eight year old son as the ultimate form of punishment against his spouse. The majority of the Grand Chamber (10 -7) rejected the applicant's claim of a violation of the right to life on account of the authorities' failure to ensure the protection of her son's life (concluding that the authorities could not have known that the applicant's husband would kill him). While the outcome has been criticized, 51 the case is important, as it afforded the Grand Chamber the opportunity to consider the measures a state must put in place to protect against domestic violence. The Grand Chamber noted in particular the importance of risk assessments, including an assessment of the reality and immediacy of any risk to life in the particular context of domestic violence, noting that imminence in the context of domestic violence should reflect 'the common trajectory of escalation in domestic violence cases'. 52 The risk of lethality must be autonomous (not solely based on the victim's perception of the risk), proactive and comprehensive, 53 and the operational measures taken by the state must be adequate and proportionate to the level of the risk assessed. 54 As Kurt v. Austria did not result in a violation, there was no process before the CoM. However, the CoM has been active in assessing the adequacy of measures taken by states to guarantee non-recurrence in a number of other domestic violence cases. These assessments present a mixed picture. At the time of writing, a number of cases, such as Talpis v. Italy, 55 Volodina v. Russia 56 (three grouped cases) and Opuz v. Turkey 57 (four grouped cases) remained under scrutiny subject to the CoM's enhanced procedure. Opuz and Talpis are at an advanced stage of scrutiny. Talpis involved several episodes of violence against the applicant and her children by her husband, ultimately resulting in the husband murdering the son and attempting to kill and injuring the applicant. The CoM has been generally positive about the steps Italy has taken in response to the judgment but has insisted that Italy provide comprehensive statistical data in particular to demonstrate that its response to domestic violence allegations is adequate, effective and timely, given that timeliness -police questioned her seven months after she filed the complaint -was a key factor in the judgment. 58 Further data has also been sought by the CoM on protection orders, criminal proceedings on domestic violence and harassment, risk assessment and management, capacity-building for law enforcement agents and the judiciary, and anti-violence centres and women's shelters. 59 Opuz v. Turkey involved repeated attacks perpetrated by the applicant's husband (the applicant's mother was 51

ultimately killed by him). As part of the general measures taken to implement the judgment, Turkey introduced new legislation in 2012, Law No. 6284 on the Protection of Family and Prevention of Violence against Women.
The law provides for a range of protective and preventive measures including shelters for victims, financial aid, psychological, professional, legal and social guidance and counselling services as needed, temporary protection measures in life threatening circumstances and day care provision to aid with integration. 60 Turkey issued a directive on the implementation of Law No. 6284 in 2019, to aid with coordination between the competent government agencies. It also established a special bureau of domestic violence and violence against women within the Public Prosecution Offices. 61 Turkey also signalled an array of capacity building measures and training for public officials that had been taken, as well as measures to aid with the sharing of data between government departments. The CoM welcomed the adoption of legislative measures and high-level public statements issued by Turkey highlighting its commitment to eradicate domestic violence, however expressed concern about the persistently high number of domestic violence and femicide victims in Turkey and invited the Government to provide further information on a range of measures aimed at preventing domestic violence including the alternative solutions to shelters for women who were unable to stay in them, practical delays to issuing and serving injunctions, the criteria to determine the length of preventive imprisonment and the types of sentences imposed for domestic violence. It also invited the authorities to work with local NGOs to collect data on femicide victims in order to improve understanding of the factors contributing to such cases. It also stressed the importance of the Istanbul Convention 'as a systemic and structural instrument for resolving issues related to domestic violence and violence against women' (which Turkey has subsequently decided to end its ratification). 62 In contrast, there are many domestic violence cases which are no longer subject to CoM scrutiny (because the CoM has deemed the measures taken by the respective states to be satisfactory). For example, in Bevacqua and S v. Bulgaria,63 the applicant, who claimed she was regularly battered by her husband, left him and filed for divorce. Her requests for a criminal prosecution were rejected on the ground that it was a 'private matter' requiring a private prosecution. The Court determined that the authorities' view that the dispute was a 'private matter' was incompatible with their positive obligations to secure private and family life. Before the CoM, Bulgaria explained that it had introduced several new laws since the events to address the problems identified, in particular, amending the code of civil procedure to speed up the examination of interim custody measures, introducing a new Protection Against Domestic Violence Act which gives new powers to issue injunctions and arrest and prosecute and to rehabilitate victims, and strengthens co-operation between 60 Action Plan: Communication from Turkey concerning the group of cases of OPUZ GROUP v. Turkey (Committee of Ministers, 1390th meeting, December 2020) paras. [17][18][19][20][21][22][23][24][25][26][27][28]paras. 46,47. 62 390th meeting (1-3 December 2020) (DH) -H46-24 Opuz group v. Turkey. 63 Bevacqua and S v. Bulgaria, Appl. No. 71127/01 (ECtHR, 12 June 2008). authorities and non-governmental organisations. 64 As a result, without scrutinising the manner in which the laws were being implemented, and seemingly, without having considered the UN Human Rights Committee's 2011 concluding observations on Bulgaria's state party report (in which the Committee expresses regret at 'the low number of cases of domestic violence, in particular against women, that are actually brought to justice and sanctioned'), 65 or the CEDAW Committee's July 2012 concluding observations (expressing 'serious concern about the high prevalence of domestic violence, the persistence of sociocultural attitudes condoning such violence and its underreporting', its particular concern 'about the absence of specific provisions criminalizing domestic violence and marital rape, the lack of criminal prosecution of violence within the family and the failure by the judiciary to follow the practice of shifting the burden of proof to favour victims…' and further concern 'about the scarcity of and insufficient funding for shelters for women victims of domestic violence'), 66 in December 2012, the CoM deemed the execution of the judgment complete. 67 This superficial scrutiny of the general measures adopted by the state is simply lip-service to the obligation to put in place effective measures to address the structural and systemic failings which led to the violation and to guarantee non-recurrence. Similarly, in Kontrová v. Slovakia, 68 in which the applicant's husband killed their two children and himself and which resulted in a finding of a violation of the right to life among other provisions, Slovakia informed the CoM that it viewed the matter to be an isolated incident. It also informed the CoM of its publication and circulation of the judgment and of existing procedures for victims to claim compensation for personal integrity. Accordingly, without further scrutiny, the CoM deemed the execution of the judgment complete. 69 The CoM's decision to close its examination does not appear to take account of the CEDAW Committee's 2008 concluding observations expressing concern 'at the high rate of violence against women and girls, including homicides resulting from domestic violence', the lack of information on 'support to women victims of violence, and the allocation of sufficient financial resources to programmes aiming at combating violence against women', among other concerns. 70 Similar to the Bevacqua and S and Kontrová cases, in Valiulienė v. Lithuania, 71 the CoM only provides cursory scrutiny of the measures taken by Lithuania to address the structural and systemic factors underpinning the violation. 72 In Valiulienė, the applicant applied to a district court to bring a private prosecution after being beaten by her partner on five separate occasions. The investigation was discontinued because new legislation meant that prosecutions in respect of minor bodily harm had to be brought by the victim privately unless specific exceptions applied. When the applicant tried thereafter to bring a private prosecution, this was refused as in the meantime the action had become time-barred. The Court determined that the applicant was denied adequate protection, though it refrained from commenting on whether it was appropriate for domestic violence cases to have to be pursued by way of private prosecution. 73 Before the CoM, Lithuania's focus was on the technical breach in the case which resulted in the applicant's private prosecution being thrown out. It explained that the Constitutional Court had ruled as unconstitutional the dismissal of criminal proceedings after the expiry of a statutory limitation period for criminal liability. Further, the Prosecutor General's office and the police instituted a series of trainings on domestic violence and several recommendations designed to address the weaknesses in the system of pre-trial investigations of domestic violence were approved or amended, such as appointing specialist prosecutors for domestic violence cases, expediting the timeframe within which a prosecutor must lodge an application concerning domestic violence and for the application of coercive measures (such as restraining orders) and improving the procedure for assessing victims' specific protection needs. Various guidelines were also issued for police on the enforcement of court orders pertaining to domestic violence and related matters relating to the role of police officers in domestic violence cases. Also, it explained that data had been and would continue to be collected on the issuance of protective measures in pre-trial investigations concerning domestic violence. 74 On the basis of Lithuania's report, the CoM declared the examination closed. 75 The robustness of the CoM's scrutiny of measures taken by states to guarantee non-recurrence has depended in large part on the independent information it had at its disposal to assess the veracity of the measures taken by the respective states. 73 Valiulienė v. Lithuania, para. 85. This is addressed in the concurring opinion of Judge Pinto de Albuquerque: 'in most cases, to place the victim of domestic violence in the unbearable quandary of having to decide for herself whether she wants to harm the family/intimate relationship through private prosecution is to perpetuate the subordinate position of the victim, and therefore, the violence itself, because she is evidently not in a position of freedom to make that choice due to her state of dependency on the offender. In other words, 'the requirement of a victim to act as a private prosecutor, which reflects the misconception of violence between members of a family/intimate relationship as "private business", is not compatible with the above-mentioned international obligation to protect' (emphasis in original). 74 Communication from Lithuania concerning the case of Valiulienė v. Lithuania (Appl. No In both Opuz and Talpiz, scrutiny included rule 9(2) submissions 76 by non-governmental organisations, review of recommendations by GREVIO, concluding observations on state party reports by the Committee on the Elimination of Discrimination Against Women (CEDAW Committee). In Opuz, the CoM also had at its disposal the report of the Turkish Parliamentary Commission of Investigation, 77 which had been established to investigate the causes of violence against women in Turkey and measures to address it. Also, marked improvement can be observed in the scope and robustness of the CoM's scrutiny following the coming into force of the Istanbul Convention; though this Convention has also influenced the degree of specificity with which ECtHR judges has referred to states' positive obligations in the course of their rulings in domestic violence cases. While to date the ECtHR has refrained from issuing Article 46 rulings (where the Court indicates specific remedial measures, often as regards systemic problems, without invoking the pilot judgment procedure) in domestic violence cases, the fact that states' positive obligations are typically at the forefront of the analysis of violations in these cases has provided a natural blueprint for the CoM as to what states should undertake by way of general measures. This deepening understanding of positive obligations -not only the positive obligation to enact legislation, but also to put in place and ensure implementation of a variety of multidisciplinary measures necessary to tackle the structural antecedents of domestic violence -mirrors the CoM's intensified engagement to gauge the extent to which general measures implemented by states are fit for purpose.

The CEDAW Committee
The Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW) does not address specifically domestic violence or violence against women though the CEDAW Committee has interpreted the focus of CEDAW on condemning discrimination against women in all its forms as applying to all forms of violence against women as the most extreme manifestation on a continuum of discrimination. 78 CEDAW Committee's General Recommendation 19 makes specific reference to domestic violence (which it refers to as family violence), terming it 'one of the most insidious forms of violence against women.' 79 This overriding lens of discrimination has impacted the way in which the Committee has interpreted due diligence obligations, often focusing on the prevalence of gender stereotypes which subordinate women to men and assign them a fixed gender role in subverting the 76 Under Rule 9(2) of the Committee of Ministers Rules, national human rights institutions and nongovernmental organizations can be involved in the supervision process by submitting reports to the Committee of Ministers in which States' performance with regard to the execution of judgments is reviewed and assessed and recommendations can be made on how to proceed with the execution process. In this respect, the submissions 'act as a kind of amicus to the Execution Department and the Committee of Ministers.' [Erken (2021). 77 Parliamentary Inquiry Commission's report (May 2015). 78 CEDAW Committee, General Recommendation No. 19: Violence against women (1992), UN Doc. A /47/38 (1993). 79 Ibid., para. 23. exercise of due diligence. 80 Recommendations on non-recurrence have thus incorporated measures to eliminate negative gender stereotypes. 81 The CEDAW Committee's complaints mechanism is established by its Optional Protocol 82 and it has issued several key decisions concerning domestic violence in which it addresses the measures states should take to guarantee non-recurrence. While the CEDAW Committee's findings and recommendations are not strictly binding in the sense of a court judgment, they provide a clear indication as to what is required to comply with the binding provisions of the Convention and how states should rectify violations of the Convention, and in this sense have important weight. According to the Committee's procedures, states parties must submit a written response within six months of receiving the Committee's decision and recommendations, detailing any action taken. 83 The Committee may subsequently invite the state party to submit further information. This may take the form of an update in the State party's subsequent periodic report to the Committee. 84 A. T. v. Hungary,85 the CEDAW Committee's first case on domestic violence, concerned abuse by the complainant's husband over a four year period. The husband eventually moved out of their apartment but returned periodically to abuse the complainant and threatened to kill her and harm their children. She filed civil proceedings to prevent him from returning to the family home, but this was denied on the basis that the evidence to support her assault claims was unsubstantiated (despite medical evidence) and his right to property could not be restricted. No criminal proceedings were entertained nor any restraining orders, and the complainant was unable to avail herself of the protection of a shelter because there were none equipped to meet the needs of her disabled child. The CEDAW Committee determined that Hungary had breached its due diligence obligations to protect the complainant against domestic violence. It also condemned the low priority afforded by national courts to domestic violence matters and Hungary's failure to eliminate the causes of widespread violence against women in the country. Whilst acknowledging the rights of the perpetrators, those rights cannot be allowed to supersede women's rights to physical and mental integrity. In addition to the specific measures stemming directly from the violation, the Committee recommended an array of measures aimed at guaranteeing non-recurrence such as enacting a law prohibiting domestic violence and introducing measures of protection, exclusion orders and support services, protecting effectively persons at risk and investigating promptly all allegations of domestic violence and bringing the offenders to justice and training judges, lawyers and law enforcement officials. 86 In Fatma Yildirim (deceased)  to death. The CEDAW Committee determined that Austria violated the Convention by failing to exercise due diligence to protect her, including by failing to detain her husband in spite of its knowledge of the extremely serious threat he posed to her. It recommended Austria to strengthen implementation and monitoring of its law on violence within the family and to vigilantly and speedily prosecute perpetrators of domestic violence as well as ensure that criminal and civil remedies are utilized in cases where the perpetrator poses a dangerous threat to the victim. In all action taken to protect women from violence, due consideration must be given to women's safety; perpetrators' rights cannot supersede women's human rights to life and to physical and mental integrity. It also recommended that Austria enhance coordination between law enforcement and judicial officers and also to ensure that all levels of the criminal justice system routinely cooperate with non-governmental organizations that work to protect and support women victims of gender-based violence and recommended that training programmes on domestic violence be strengthened. 87 Şahide Goekce (deceased) v. Austria, 88 presented similar facts and resulted in identical recommendations. The deceased's husband threatened to kill her on numerous occasions; requests to detain him were denied, and instead he was given a series of short-term restraining orders. Ultimately, he shot to death his wife in front of their two children. Austria informed the Committee of the steps it took to implement the decisions, which included setting up (state-funded) intervention centres to support domestic violence victims, amending the Code of Criminal Procedure and further amendments planned for the Law on Protection against Violence within the Family, using specifically trained public prosecutors to process domestic violence cases. 89 In another case of long-term domestic violence, V.K. v. Bulgaria, the CEDAW Committee found Bulgaria responsible for the refusal of Bulgarian courts to issue a permanent protection order and recommended Bulgaria to amend its Law on Protection against Domestic Violence so as to remove the one-month time limit and to ensure that protection orders are available without placing undue administrative and legal burdens on applicants and to ease the burden of proof in favour of the victim. It also recommended that Bulgaria ensure the availability of a sufficient number of State-funded shelters and provide support to non-governmental organizations offering shelter and other forms of support to victims of domestic violence, and provide mandatory training for judges, lawyers and law enforcement personnel on the application of the Law on Protection against Domestic Violence. 90 The CEDAW Committee decided to close this and two other cases concerning Bulgaria following payment of compensation to the victims, without having specifically addressed the extent to which Bulgaria implemented the range of preventive measures that it had recommended, 91 though some of the recommendation are taken up by the Committee in its concluding observations on Bulgaria's 4-7th 92 and 8th 93 periodic reports.
In Angela González Carreño v. Spain, which involved extended domestic violence and protracting custody battles and ultimately resulted in the husband killing his daughter and then himself, the CEDAW Committee found Spain to have breached its due diligence obligation to protect the wife and her daughter in light of the known risks posed by the husband. In addition to recommendations specific to the case, it called on Spain to take appropriate and effective measures to ensure that prior acts of domestic violence are taken into consideration when determining custody and visitation rights, strengthen the application of the legal framework to ensure that the competent authorities exercise due diligence to respond appropriately to situations of domestic violence, and provide mandatory training for judges and administrative personnel on the application of the legal framework with regard to combating domestic violence. 94 In August 2018, the Supreme Court of Spain, after recognizing the binding nature of the Committee's views, ordered the Government of Spain to pay compensation to the author in the amount of €600,000. 95 It is unclear which if any measures to guarantee non-recurrence were complied with.
The CEDAW committee's jurisprudence on the fundamental linkages between discrimination against women and domestic violence have played an important role in articulating states' due diligence obligations and identifying discriminatory practices which contribute to the perpetuation of violence. Less advanced is the Committee's work on guarantees of non-recurrence, with many of the recommendations remaining overly simplified and generalized and not subject to rigorous scrutiny at the enforcement stage. Certainly, there are limited means at the disposal of the CEDAW committee to follow up with states on the implementation of recommendations; nevertheless, the failure to make adequate inquiries in this area and look behind states' superficial statements about compliance can undermine the weight of the recommendations.

Conclusions
This article has provided an overview of the practice of key regional and international human rights bodies in recommending measures to guarantee non-recurrence in relation to the systemic violation of domestic violence. Given the synergistic relationship between positive obligations and guarantees of non-recurrence, the measures recommended stem directly from the findings of due diligence failings by the states concerned. This ensures to an extent the appropriateness of the recommended measures to address the actual failings identified in the cases. However, less attention has been placed, aside from the most recent proceedings before the CoM and to an extent the Inter-American Commission, in assessing the practical import of the measures in actually helping to prevent new violations.
For the most part, these bodies have assessed the extent of implementation with only limited (and in some instance no) recourse to data to assess the extent to which measures impacted states' behaviour, and whether the violations stopped or were significantly reduced. Given the extensive caseloads of the bodies tasked with assessing implementation, there is pressure to make rapid assessments of state compliance. This often translates to the evaluation of whether recommendations were actioned, as opposed to whether recommendations produced the desired structural changes in attitudes and behaviours.
More attention by states, by civil society and social scientists conducting independent research, on the collection of statistical data would complement, and indeed enhance the engagement with human rights mechanisms on the enforcement of decisions.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.