This article critically examines whether, and how, five United Nations human rights treaty monitoring bodies deal with situations in which human rights have been interfered with by non-State actors. The article uses the concepts of ‘direct’ and ‘indirect’ horizontal effect of international human rights law. An in-depth comparative analysis reviews the monitoring bodies’ general comments and views on individual communications from the bodies’ establishment until August 2017. The analysis identifies very limited evidence of ‘direct’, and two main kinds of ‘indirect’ horizontal effect of human rights applied by the bodies. First, it finds that while the bodies do engage with the ways in which non-State actors can interfere with human rights, they predominantly focus on the positive and procedural obligations of States. Second, it finds that where non-State actors are sufficiently ‘State-like’ in their nature and actions, they may be re-categorised as public actors for the purposes of human rights.
According to the prevalent approach in legal science, the responsibility for protecting individuals’ human rights lies in the positive (vertical) obligations of the State. Non-State actors are to be regulated and controlled by the State through its domestic legal system. However, many non-State actors are in a position to greatly affect individuals’ enjoyment of their human rights.1 This is particularly true of non-State entities carrying out public functions, having control over an area of territory or even being in a position to direct States in the adoption and implementation of certain domestic laws and policies. However, single individuals may also be in a position of relative power over (for example) other family members, employees, or members of a different social class, placing them in a position to more easily affect another individual’s rights.
The traditional human rights paradigm in the international legal discourse was founded on a relationship of dependence and trust on behalf of individuals towards States as the primary subjects of international law2; human rights provided them with ‘fundamental guarantees and standards of legal protection’ against potential abuses of State power.3 The focus of international human rights law has therefore been on the actions of States; private actors would normally fall within the remit of domestic criminal laws4 or private laws,5 rendering international governance of their actions unnecessary. Nevertheless, the ‘new-medievalism’ of international relations6 is steadily replacing the Westphalian, one-dimensional State-centric model of the international legal order with a multi-layered system.7 As a by-product of globalisation, this is creating a ‘neo-feudal’ society in which power and influence are distributed amongst various actors.8 But States, originally targeted for obligations because of their socio-economic and legal power monopoly over individuals ‘in the absence of legal restraints’,9 are continuously losing ground to private actors.10 It is increasingly evident that despite trusting States to protect individuals from interference with their rights by third parties, domestic laws are not (always) sufficient or effective in governing the actions of non-State actors vis-à-vis human rights.11
This encroachment of non-State actors is not confined to practical effects; whilst under the current legal framework the State remains the primary actor, some non-State actors have now obtained a certain political or even a ‘law-making’ role.12 States are no longer the only actors adopting instruments aimed at implementing or developing international law (including aspects of international human rights law). Non-State actors are becoming increasingly involved in global governance,13 whether through contributions to international law itself, or through (self-)regulation by private bodies.14 As a result, non-State actors are beginning to entrench themselves in the international human rights regime. Non-State actors are also becoming increasingly entrenched in the regime by other actors, including scholars and civil society. For example, there now exist many studies, projects and initiatives examining the ways in which non-State actors could be held accountable for interfering with the enjoyment of human rights.15 There are also more and more non-legally binding instruments being adopted, including by the un itself, to encourage non-State actors to respect human rights and States to more rigorously regulate non-State actors (a popular example being the un Guiding Principles on Business and Human Rights16). Furthermore, international human rights treaty monitoring bodies have found themselves faced with many cases in which the direct causal responsibility for harm to human rights lies with a non-State actor, rather than a State.17 In these cases, the supervisory bodies have often interpreted and applied human rights standards vis-à-vis non-State actors (see Section 4).
A major question within the field of international human rights law is whether the answer to human rights interference by non-State actors should be addressed through the international legal framework, or whether non-legal approaches would better protect individuals. In other words, do we need to look beyond the current (albeit still developing) legal framework to achieve de facto horizontal effect? In order to answer this question, it is necessary to have a very thorough understanding of how the current international human rights law system deals with non-State actors. This of course requires an examination of international human rights obligations per se, international human rights treaties, and the application of these treaties in practice. This article undertakes a comparative analysis of the latter, focusing on the interpretation and application of human rights vis-à-vis non-State actors by five core United Nations human rights treaties supervisory bodies. The analysis examines and compares the different ways in which the five monitoring bodies deal with situations and cases where the harm to an individual’s human rights was directly caused by a non-State actor. The analysis produces two main results: (1) identification of whether (and if so, which) different kinds of horizontal effect are employed by the monitoring bodies; and (2) identification of whether (and if so, how) the monitoring bodies apply the types of horizontal effect differently to distinct kinds of non-State actors (e.g. individuals, private companies and non-State armed groups).
Before moving on to the methodology of the article, a quick note must be made regarding its place in academic literature. Much research has already been conducted into the ‘horizontal effect’ of international human rights law; indeed, a veritable abundance of literature can already be found on ‘non-State actors and human rights’.18 The present article uses much of this literature in setting out its theoretical framework, particularly that regarding the direct and indirect horizontal effect of human rights and the typology of States’ obligations.
Some scholars, most notably Andrew Clapham, have published analyses of horizontal effect at the international level.19 The present article seeks to both draw and build upon such literature, conducting a more thorough analysis in the practice of international human rights monitoring bodies. Clapham’s seminal book on ‘Human Rights Obligations of Non-State Actors’, published in 2006, includes a chapter on ‘Selected human rights treaties’.20 The chapter takes the treaties themselves as a starting point, looking into several pertinent examples of general comments and individual complaints that deal with the interpretation and application of the six treaties he analyses. However, a more comprehensive and systematic (as well as up-to-date) analysis of the general comments and cases from human rights monitoring bodies has yet to be carried out. The present article seeks to fill this gap.
2 Research Methods and Design
The article examines the ways in which the United Nations human rights treaty monitoring bodies established in connection to five ‘core’ United Nations human rights treaties, have interpreted and applied human rights obligations in relation to non-State actors. Generally speaking, the bodies express their interpretations of various provisions of the respective human rights treaties through ‘general comments’21 and apply the provisions through their respective individual complaints procedures, in the form of ‘views’ (see below). General comments are one of the most well-known outputs of human rights treaty bodies. While the source of Committees’ mandates may differ,22 the adoption of general comments is a common practice of each of the bodies examined.23 General comments and views of the treaty bodies are not legally binding,24 but have been repeatedly found to be of high interpretative value.25 Not every State Party to the treaties takes this approach towards general comments and views.26 This does not negate their importance for the task at hand, however; the outputs of the bodies remain very important for determining how, at the international level the horizontal effect of human rights is (whether implicitly or explicitly) discussed and applied.
Views of the Committees regarding individual complaints are adopted by virtue of the individual complaints/communications procedures for each of the bodies examined. For the most part, the authority to hear individual complaints derives not from the main human rights treaty itself, but from an additional (and optional) protocol.27 There is no system of precedence within the jurisprudence of the treaty bodies, which means that the scope and application of rights may change over time and the impact of the decisions could be affected.28 As with general comments, the Committees’ views are not legally binding, although they have been said to be of ‘great weight’ because the bodies are ‘established specifically to supervise the application’ of the relevant treaties.29
The following analysis looks at general comments from the establishment of the treaty bodies until August 2017, and views on individual complaints from the time of activation of the individual complaints procedures until August 2017.
The monitoring bodies chosen for the analysis are: (1) the Human Rights Committee; (2) the Committee on Economic, Social and Cultural Rights; (3) the Committee on the Elimination of Discrimination against Women; (4) the Committee against Torture; and (5) the Committee on the Elimination of all Forms of Racial Discrimination. Although by now there are nine ‘core’ human rights treaties,30 the five monitoring bodies examined were chosen for analysis due to their substantive contributions in the context of non-State actors and human rights obligations and on the nature of human rights obligations more generally (see, e.g. the Committee on Economic, Social and Cultural Rights31). Further, the Human Rights Committee and the Committee on Economic, Social and Cultural Rights were chosen for the reason that they supervise the implementation of the ‘twin’ human rights Covenants – the International Covenant on Civil and Political Rights and the International Covenant on Economic, Social and Cultural Rights. The Covenants were adopted at the same time, in 1966, and laid down what was then the full range of international human rights. The Committee on the Elimination of Discrimination against Women was chosen as a good example of bodies that monitor human rights treaties drafted and adopted for the protection of a particular vulnerable group – in this case, women.32 Finally, the Committee against Torture and the Committee on the Elimination of all Forms of Racial Discrimination were chosen as examples of bodies that monitor the implementation of subject-specific human rights treaties – the subjects here being torture and racial discrimination. Including this range of bodies in the analysis should allow for a broad understanding of how horizontal effect is applied in practice within the international human rights system. For reasons of space, the output of the monitoring bodies in relation to State reporting procedures (i.e. concluding observations) have been excluded from the analysis in this article. The sheer volume of concluding observations makes it impractical to analyse them within the framework of this article, and the results obtained from the general comments and views of the Committees are considered to be substantive enough without the analysis of additional documents.
The examples analysed in the following sections were found using two different methods. First, a database compiled by scholars, ‘Bayefsky’, was relied upon, which aims to ‘enhanc[e] the implementation of the human rights legal standards of the United Nations’.33 The database includes a large collection of documents containing references to particular subject-matters in the supervisory bodies’ application of human rights treaties. The document relied upon for this article lists (and contains extracts of) which general comments and views of the bodies include reference to ‘Public and Private Actors’.34 The document was used as a starting point to identify which general comments and individual communications would be relevant for the analysis. Each general comment or communication identified in the Bayefsky document was then searched manually, using the terms ‘non-State actor’, ‘private actor’, and ‘positive obligations’. Where this led to information regarding how the relevant monitoring body viewed or applied human rights obligations vis-à-vis non-State actors, the general comment or individual communication was used in the analysis. In some cases, however, the reference was cursory and was not substantive or informative enough to contribute to the analysis, and was therefore excluded from the analysis. Although extensive, the database is not exhaustive as it does not include general comments and views after 2005. For this reason, the individual general comments and jurisprudence of the relevant bodies from 2005–2017 were manually searched using the three terms above, to identify any new references to non-State actors. This search was conducted using the Office of the High Commissioner for Human Rights’ treaty body database.35 Given the nature of this method, it is possible that some views or general comments that have a less explicit (but perhaps still substantive) reference to non-State actors may have been missed. Nevertheless, it is believed that the documents discussed provide enough material to gain an overview of the way in which each of the five bodies deal with the horizontal effect of human rights.
The analysis itself was conducted from the viewpoint of ‘horizontal effect’, and due to the findings, turned out to be exclusively focused on indirect horizontal effect. In particular, the State obligation to protect human rights (and the encompassed duty of due diligence) are frequently referred to in the analysis as a starting point (see Section 3.3). Other than this, the analysis has been made as ‘clean’ as possible, referring as far as possible to the language of the monitoring body itself, reserving any categorisation of the type of approach used until the reflective comments in Section 5.
The article begins with an explanation of the theoretical framework within which the analysis was undertaken (Part 3), explaining in particular the concept of horizontal effect. From there, the analysis proper begins (Part 4), which is structured by monitoring body. First, a Committee’s general comments are examined, before moving on to the Committee’s ‘views’. Brief critical reflections on the practice of each body are made before moving on to the next monitoring body. Part 5 constitutes the article’s comparative analysis, which arranges the findings first according to certain ‘types’ of horizontal effect that can be identified, and second, according to the ways in which the bodies have dealt with different kinds of non-State actor. This Part 5 also includes a brief critical comment on the Committee’s reasoning in the practice analysed. Final conclusions are drawn in Part 6.
3 Theoretical Framework
3.1 Vertical vs. Horizontal Effect
It is common knowledge among international lawyers that the nature of international human rights obligations is vertical – being owed by the State (as obligation-holder) to the individual (as beneficiary).36 The term ‘vertical’ demonstrates that the State, as the primary subject of and regulated by international law, is placed on a ‘higher’ playing field than the individual, who typically operates within a State and is therefore regulated by national law. The vertical effect of human rights law stems predominantly from the fact that, as the original subjects of international law, only States can become party to human rights treaties and therefore be legally bound by their obligations (although Jan Hessbruegge emphasises that this incapacity does not exclude obligational relationships in the non-State sphere37). Furthermore, only States can be the subject of individual complaints or cases before international human rights treaty monitoring bodies and regional human rights courts. Despite the increased power and influence of non-State actors, there have been no real correlative developments in (binding) international human rights law allowing for obligations to be applied between two (or more) non-State actors. The application of international human rights law between non-State actors is known as the ‘horizontal effect’ of human rights and will form the basis of the legal analysis undertaken in this article.
The terms ‘direct horizontal effect’ and ‘indirect horizontal effect’ are very often used in the fields of constitutional law and private law, particularly within the European context.38 Much of the debate surrounding horizontal effect comes from, in particular, Germany and the United Kingdom.39 However, the terms are also used at the international level, with scholars such as John H Knox discussing the ‘horizontality’ of international human rights law, and Thomas Gammeltoft-Hansen considering the ‘direct’ and ‘indirect’ responsibilities and obligations of non-State actors.40 Although rarely, the term ‘direct horizontal effect’ has also been used by one of the un human rights treaty monitoring bodies (see Section 4.1.1).
3.2 Direct Horizontal Effect
The direct horizontal effect of human rights treaties would ‘la[y] duties directly upon a private body to abide by its provisions and mak[e] breach of these duties directly actionable at the instance of an aggrieved party’.41 In other words, it would place non-State actors under direct and explicit obligations to respect, protect and/or fulfil human rights (see Section 3.3.1).
Direct horizontal effect is sometimes discussed from the perspective of a victim of a human rights violation, in which case it is considered to have two components – substantive and procedural. Substantive horizontal effect would enable individuals to claim violations of rights owed to them by non-State actors, whilst procedural horizontal effect would allow an individual to ‘enforce his fundamental rights against another individual’.42 At the international level, this would challenge the existing rule that complaints of human rights violations may only be brought before human rights monitoring bodies (and for the most part, human rights courts) by individuals against States.43 In today’s international human rights framework, this is not possible. As it stands, neither substantive nor procedural direct horizontal effect can be found in international human rights law. This has been reiterated many times, for example by the Human Rights Committee in General Comment 31: ‘obligations are binding on States and do not, as such, have direct horizontal effect as a matter of international law’ (see Section 4.1.1).44
Although this is the current state of play, there have been significant strides towards direct horizontal effect for business enterprises. Most notably, in July 2014 the un Human Rights Council adopted Resolution 26/9, through which it established an open-ended, intergovernmental working group with the mandate ‘to elaborate an internationally legally binding instrument to regulate, in international human rights law, the activities of transnational corporations and other business enterprises’.45 The working group has made considerable progress, with its latest report showing that a detailed framework and overview of the contents of a treaty on business and human rights has been developed.46 However, it is it likely to be quite some time before a final version of the treaty has been adopted and gained enough State ratifications to enter into force.
Slightly earlier developments towards direct human rights obligations for businesses occurred through the adoption of the un Guiding Principles on Business and Human Rights (ungps).47 The ungps were drafted by John Ruggie in his capacity as Special Representative of the Secretary-General on the issue of human rights and transnational corporations and other business enterprises, and were unanimously endorsed by the un Human Rights Council in 2011.48 Significantly, the ungps contain a direct responsibility (as opposed to an obligation) for businesses to respect human rights, as well as a responsibility to conduct human rights due diligence.49 Although the ungps are not legally binding, they have had a tremendous impact; both States and businesses have taken concrete action towards the implementation of the ungps. This includes, for example, the adoption of ‘National Action Plans’50 and national legislation by States (see below) and measures such as human rights impact assessments, human rights policy statements, reporting and training by businesses.51
The Organisation for Economic Cooperation and Development Guidelines for Multinational Enterprises (oecd Guidelines)52 have also helped to pave the way to direct horizontal effect for businesses. The oecd Guidelines were first adopted in 1976 and have been reviewed several times. The most recent review took place in 2011 and resulted in the addition of a new chapter on human rights which is consistent with the ungps, strengthening the commitment to human rights already included in the oecd Guidelines.53 As with the ungps, the oecd Guidelines are not legally binding. However, they require adhering States to establish a ‘National Contact Point’ (ncp), making it the only instrument on the responsibility of businesses that has a ‘built-in grievance mechanism.’54 ncps are mandated to ‘provide a mediation and conciliation platform for helping to resolve cases’ of non-compliance with the oecd Guidelines.55 Since the 2011 revision, the number of ncp cases dealing with human rights has increased dramatically.56 However, even before this, some ncps referred to international human rights treaties in their ‘Final Statement’ on a case.57 While important, this is not altogether surprising, since the previous version of the Guidelines (adopted in 2000) provided that companies should ‘respect the human rights of those affected by their activities consistent with the host government’s international obligations and commitments’,58 and explicitly referred to the human rights legal framework.
Perhaps the ungps’ and oecd Guidelines’ most significant contribution to (binding) direct horizontal effect has been their influence on legislation adopted at the national and European level. Notable examples can be found in the United Kingdom. For example, both The Companies Act 2006 (Strategic Report and Directors’ Report) Regulations 2013 and the Modern Slavery Act 2015 contain provisions requiring certain businesses to disclose information related to human rights.59 In the Modern Slavery Act in particular, this extends, for businesses over a certain size, to information as to ‘what action they have taken to ensure there is no modern slavery in their business or supply chains’.60 Further examples of legislation influenced by the ungps can be found in France and the us.61 A recent law passed in France requires certain companies to make ‘vigilance plans’ that must, inter alia, include ‘reasonable vigilance measures to identify risks and prevent serious violations of human rights and fundamental freedoms’.62
Further developments towards direct horizontal effect have been made through the jurisprudence of several national legal systems. A full discussion falls outside the scope of this article due to its international focus, but it is interesting to see that some courts have held private owners of publicly accessible spaces to be directly bound by the fundamental rights to freedom of expression and freedom of assembly.63 In effect, the private owners’ enjoyment of their property rights have been limited in order to allow individual/s to exercise their right to freedom of expression/assembly.64 Other courts, though not going this far, have engaged in discussions concerning how much discretion private property owners have for what concerns the enjoyment of freedom of expression/assembly on their property and whether public property owners enjoy the same level of discretion.65 Similar case law can also be found at the regional level, although in these proceedings the State has remained the ultimate obligation-holder rather than the relevant private actor, thereby ruling out direct horizontal effect.66
At the international level, although not within the realm of international human rights law, there have also been cases in which human rights obligations have been upheld against private actors. For example, in a significant case in 2014 the European Court of Justice balanced the rights and interests of an internet search engine operator (Google) against those of a ‘data subject’ (an individual).67 In applying the relevant law (the Data Protection Directive68) it considered the Charter of Fundamental Rights of the European Union.69 It found that unless interference with the data subject’s rights to privacy and data protection could be justified, they would ‘as a rule’ take preference over the rights and interests of the internet search engine operator and the general public.70
Similar cases have also been decided by private dispute arbitrations or example, the World Intellectual Property Organization has upheld the right to freedom of expression against a private corporation and has interpreted the policy to be applied in the case by reference to international human rights law.71 The International Centre for the Settlement of Investment Disputes which hears investment disputes between States (acting as private actors) and private actors has also considered human rights law in some decisions.72 However, this remains an exception rather than the rule, and scholars have pushed for further consideration of human rights in investment arbitration,73 warning against ‘considering international investment law in a vacuum’.74 For example, arguing for greater harmonisation between international human rights and international investment law, Bruno Simma and Theodor Kill suggest using international human rights law as ‘external rules’ when interpreting investment treaties.75
Finally, in recent years there have been several instances of international bodies holding non-State actors (in particular non-State armed groups) to be bound by some jus cogens norms. For example, in 2012 the Commission of Inquiry on the Syrian Arab Republic (the body established by the un Human Rights Council to investigate alleged violations of international human rights law since March 201176) in the country stated that ‘at a minimum, human rights obligations constituting peremptory international law (jus cogens) bind States, individuals and non-State collective entities, including armed groups.’77 This has been reiterated by the un Mission in the Republic of South Sudan, which found the same obligations to bind non-State armed opposition groups as well as States.78 These claims are strengthened by international criminal law, which enables members of non-State armed groups to be held individually responsible at the international level for violations of some jus cogens obligations.79 However, in this context the obligations are not framed as part of human rights, but criminal law, and therefore cannot be considered true examples of direct horizontal effect of international human rights law.
Further developments towards direct horizontal effect are buttressed by scholars via the suggestion of new theories or bases for direct horizontal effect.80 Specifically, some literature suggests focusing on our understanding of human rights themselves, as opposed to human rights law, in order to justify placing direct human rights obligations on non-State actors.81 Ibrahim Kanalan is one proponent of this approach, arguing that many previous theories of direct horizontal effect fall short because of their focus on staying within the confines of the international legal framework.82 He suggests a new concept based on the ‘normative power of human rights and the consideration of the functional differentiation of society’.83 Other scholars have taken a broader perspective, developing theories of shared responsibility between State and non-State actors,84 as well as ‘multi-duty bearer regimes’.85 A discussion of the projects falls outside the scope of this article, but they are extremely interesting and show that scholars are working to fill the ‘accountability gap’ arising from the lack of direct horizontal effect in new ways.86
All of these examples show that at the national, regional and international levels, real developments have been made in various contexts towards the direct horizontal effect of human rights. Nonetheless, most of the concrete, binding developments have taken place outside of international human rights law, the direct application of which between non-State actors remains extremely limited.
3.3 Indirect Horizontal Effect
Due to the lack of direct horizontal effect of international human rights, a concept of ‘indirect horizontal effect’ has emerged. In a situation of indirect horizontal effect, it is the State, not the responsible non-State actor, against whom the victim brings a claim of an interference with their human rights. As such, the State is also the entity that is (if the claim succeeds) legally recognised as being responsible for the harm suffered by the victim. This is regardless of the fact that the act interfering with the human right was done by a non-State actor. Essentially, this results in a diagonal application of human rights, through a State’s direct obligation to protect individuals from the harmful actions of other non-State actors (see below). Under this construct, while the State remains directly responsible, indirect obligations, which derive from international law, are imposed on non-State actors; it may well be that through the State’s fulfilment of its obligation to protect, a non-State actor is under an obligation to adhere to certain human rights standards imposed by national law (thereby receiving direct obligations at the national level as well).87 However, this article is concerned with horizontal effect at the international level. Because of the construction of indirect horizontal effect based on the obligation to protect (see Section 3.3.1), this article’s analysis could be seen as an examination of the obligation to protect, as well as indirect horizontal effect as such (from the perspective of non-State actors).
The analysis also deals to some extent with State responsibility, in particular the International Law Commission’s ‘Draft Articles on Responsibility of States for Internationally Wrongful Acts’ (dasr).88 The dasr, although never becoming a binding treaty, are now widely considered to constitute customary international law and have been widely applied in practice.89 Published together with commentary on each article, the dasr specify under what circumstances a State can be held responsible for a violation of international law (that there is an act or omission that ‘(a) is attributable to the State under international law; and (b) constitutes a breach of an international obligation of the State’).90 Chapter ii dasr specifically details how/when conduct can be attributed to the State. The commentary to Chapter ii dasr explains that although as a general rule, only the actions of State agents can be attributed to the State, it may also be ‘responsible for the effects of the conduct of private parties, if it failed to take necessary measures to prevent those effects.’91 In particular, the following conduct may be attributed to the State: (i) the conduct of ‘persons or entities exercising elements of governmental authority’ (e.g. privatised corporations that retain public or regulatory functions);92 (ii) conduct ‘directed or controlled by a State’;93 (iii) conduct ‘carried out in the absence or default of the official authorities’;94 (iv) conduct of insurrectional or other movements;95 and (v) conduct ‘acknowledged and adopted by a State as its own’.96 While the analysis in Section 4 does not focus on the dasr, it is certainly interesting to bear them in mind; some parallels can be drawn between the dasr and the reasoning of the human rights treaty monitoring bodies, which sometimes refer to the ‘attribution’ of non-State actor conduct to the State in their application of indirect horizontal effect. However, as will be shown below, the bodies rarely mention the dasr explicitly.
Because the focus here is on the international level, the definition of indirect horizontal effect adopted above differs somewhat from popular definitions in the national or European context – Gavin Phillipson, for instance, defines indirect horizontal effect as meaning that ‘whilst the rights cannot be applied directly to the law governing private relations and are not actionable per se in such a context, they may be relied upon indirectly, to influence the interpretation and application of pre-existing law’.97 This definition is more relevant at the domestic level, where it is possible to have cases in which both parties are non-State actors, and will not be applied in this article. Under Phillipson’s understanding, it would fall to the national judiciary to apply international human rights standards when giving judgments, even when they are dealing with a case that only involves non-State actors.98
Despite some academics believing that the distinctions between direct and indirect horizontal effect are simply a matter of semantics,99 there is quite a large practical difference between them,100 particularly for the victim of a violation. There is a crucial ideological distinction between seeing the actual perpetrators of the violations as ‘real’ subjects of international law as opposed to holding only the State responsible. At some point, we have to let go of the habit of seeing exclusively the State as some kind of ‘parental’ figure, wholly and singularly responsible for guaranteeing the welfare of its citizens,101 and allow the non-State actors – often more akin to adolescents (torn somewhere between the dependent individual, and the autonomous, independent nation-State) – to take on responsibility for their own actions. The perpetual obstacle here is determining at which point this line should be drawn, and how. Should there be a method akin to the ‘Gillick competence’ test developed in uk common law to determine this by applying established criteria?102 How can standards be differentiated according to different types of non-State actor? Is it legitimate to establish international norms to force non-State actors to operate in a certain way? A discussion of all of the possibilities falls outside of the scope of this article, but these questions should be borne in mind when reading the analysis.
Indirect horizontal effect has been denounced by some as not being a particularly useful modus operandi. Thomas Bennett has declared it to be ‘an inexact method for the supposed protection of what are meant to be the basic, fundamental rights of individuals’.103 This point deserves due credit. On the one hand, the consequences of successful indirect horizontal effect cases can be to further human rights protection and to require the State to adopt more effective measures to protect individuals from harm by non-State actors, thereby going further than a simple case-by-case impact, resulting in broader improvements for human rights protection. On the other hand, the potential impacts of indirect horizontal effect are constrained by the restrictions of the powers of courts and supervisory bodies to interpret international human rights law in a legitimate way (i.e. according to the confines of the Vienna Convention on the Law of Treaties104). The courts, in abiding by these limits, are often unable to provide victims of human rights interference with effective remedies (or even hear cases concerning non-State actors) which would otherwise be available had the perpetration been done by a different actor.
3.3.1 The State Obligation to Protect Human Rights
The obligation to protect very commonly forms the basis of indirect horizontal effect and was alluded to above. It will not be discussed in detail here, but rather introduced, to provide context for the practice of the monitoring bodies below that relies on States’ positive obligations.
The obligation to protect human rights forms one part of tripartite typology of State obligations. The typology consists of the obligations to respect, protect and fulfil human rights, and was proposed in its current form by Asbjørn Eide.105 According to Eide, the protective function of human rights is the most important of all.106 Indeed, the obligation to protect has been afforded a huge amount of attention by legal scholars and practitioners alike.
In concrete terms, the obligation to protect ideally results in the prevention of interference by non-State actors through the incorporation and/or implementation of human rights treaty standards in a State’s domestic laws and policies. This has been interpreted to require States to take immediate steps to ensure that violations by the State, its agents, and non-State actors are prevented.107 It should also include providing access to impartial legal remedies in the case of any violations, regardless of the identity of the perpetrator.108 As an example, Principle 6 of the Maastricht Principles on Extraterritorial Obligations of States in the Area of Economic, Social and Cultural Rights109 explained the obligation as meaning that the failure of States to ensure that private employers comply with basic labour standards may amount to a violation of the right to work or the right to just and favourable conditions of work (protected under Article 7 of the International Covenant on Economic, Social and Cultural Rights).110 In the context of economic, social and cultural rights, the un Handbook for National Human Rights Institutions states that the obligation to protect requires States to take ‘active measures to protect all persons from racial or other forms of discrimination, harassment and the withdrawal of services’.111
A major aspect of the State obligation to protect human rights is the duty of due diligence, which exists in several areas of international law. Due diligence is an obligation of conduct rather than result, meaning that it is more the tangible effort, and ‘progressive’ steps made by States which fulfil the obligation, rather than the success of these actions per se.112 In the human rights context it ‘requires action reasonably calculated to realize the enjoyment of a particular right’,113 whereas an obligation of result would require ‘States to achieve specific targets to satisfy a detailed substantive standard’.114 At first sight, this may appear to be a less demanding or even effective type of obligation, but in reality it is necessary in situations where the State may not have automatic control over private actors, making the actual realisation of rights unrealistic (perhaps a notion connected to the ‘minimum core obligation’ which recognises the differing capacities of States to fulfil rights immediately, and allowing their progressive realisation).115
Joanna Bourke-Martignoni’s evaluation of due diligence in the context of violence against women appears to suggest that the duty would also extend to addressing the causes of human rights violations by non-State actors,116 as well as introducing domestic laws to ensure the effective investigation and redress when they have occurred. This supports the idea that one of the main tenets of the obligation to protect is the prevention of interferences by non-State actors; as the saying goes, prevention is better than cure. In the context of the prohibition on the use of child soldiers, Radhika Coomaraswamy also formulated, in her previous capacity as Special Rapporteur on Violence against Women, a ‘checklist’ of measures which would fulfil the duty of due diligence, including ‘appropriate measures in the field of education and the media to raise awareness’,117 thus entailing quite extensive steps to be fulfilled.
As a principle in international law more generally, due diligence may depend on the foreseeability of a non-State actor’s conduct. As Robert Barnidge explains, it could involve particular knowledge of the State (i.e. of some intended harm, which they could have prevented but failed to do so).118 The duty must also be exercised in relation to the transboundary effects of acts by private individuals within a State’s jurisdiction, as was held in the Trail Smelter case.119 The practical extent of the duty differs according to prevailing circumstances, and according to the resources available to a State. This is because the duty is context-dependent, meaning that the scope of actions required by the State is dependent on the situation on the ground. For example, in the context of diplomatic law, the usual degree of (host-State) police presence required outside an embassy would be low (if required at all), requiring little State action. However, in the event of a protest or riot aimed at the embassy and taking place in its vicinity, more police action would be required to control the situation and prevent unlawful behaviour.120 Similar standards can be said to have developed in the human rights context, which will be seen through the analysis in Section 4.
Due diligence has been applied as a principle under international human rights law by many monitoring bodies, examples of which will be discussed below. While it has been interpreted and applied slightly differently by the various bodies, a very common expression of the duty is that to ‘prevent, investigate and punish’ human rights violations by private actors.121
It is important to be aware that the changes taking place in relation to globalisation, privatisation and the increased power of non-State actors may affect the way in which due diligence develops. On the one hand, the changing nature of the relationship between certain non-State actors, the State and individuals, and the taking on of more traditionally State functions by non-State actors (i.e. through the provision of state services), may mean that the scope of the obligation to protect and the duty of due diligence will expand accordingly. As a State delegates or outsources more activities to non-State actors, it may be necessary for a wider-ranging duty to emerge. This may already be happening in relation to some actors. The ungps, for example, detail the duties that States should observe in relation to businesses. The Principles provide for concrete measures that States should take to protect individuals from the harmful actions of businesses. For example, they require States to ensure that there are effective mechanisms in place for individuals to gain redress for human rights-related harm they have suffered at the hands of businesses.122 We see similar standards in the oecd Guidelines for Multinational Enterprises,123 through the aforementioned requirement that adhering States establish National Contact Points to deal with situations in which a multinational corporation is believed to have negatively affected an individual’s human right.124 More recently, a general comment on business and human rights was adopted by the un Committee on Economic, Social and Cultural Rights, explaining in quite some detail the due diligence obligations (among others) of States vis-à-vis business enterprises (See Section 4.2.1).
On the other hand, the new levels of influence being reached by some non-State actors could mean that the obligation to protect becomes increasingly difficult to fulfil. For example, again in the context of multinational corporations, it may be difficult to obtain information relating to their daily operations in order to prevent human rights interference. Non-governmental organisations trying to improve the human rights impact of corporations in the coal industry have often found that corporations are unwilling to share information with States, or with the public at large.125 Short of States obliging corporations to provide information, it may be difficult to determine which measures need to be taken by States under the duty of due diligence in order to ensure that the daily operations of such corporations do not infringe the human rights of (for example) their employees. A worrying trend of political influence of corporations may even mean that the State is unable to take the requisite measures to oblige a corporation to publicise information without the threat of losing the investment of the corporation within their territory.126
In addition, the vast range of non-State actors and the circumstances in which they operate may make it impractical to develop concrete standards that States should follow under the duty of due diligence. Indeed, this may not even be desirable given that the contextual nature of the duty suggests a case-by-case analysis of what States must do. The fact that due diligence is a duty of conduct rather than result is also relevant again here. Although reasonable (it has to be acknowledged that States do not have an unlimited capacity to act, particularly due to resource restraints), this underlines the major setback of the obligation to protect in achieving its aim of protecting individuals from the harmful actions of non-State actors. A victim who brings a complaint against a State on the basis of the obligation to protect will not be able to gain any kind of redress for the violation of their rights if the State has taken all feasible measures and still failed to regulate the actions of the non-State actor to the effect that they respect human rights. The lack of direct horizontal effect precludes the victim from getting a remedy vis-à-vis the direct perpetrator of the harmful acts, and the duty of due diligence precludes them from achieving indirect horizontal effect, whereas if the case had been brought against the State which had, itself, committed the same actions as the non-State actor in this scenario, the victim would have been able to gain some kind of redress. Further critique of indirect horizontal effect will be left to the following sections’ analysis of horizontal effect in practice.
4 Horizontal Effect of International Human Rights in International Practice
4.1 Human Rights Committee
The Human Rights Committee (HRCtee) is the designated treaty monitoring body for the International Covenant on Civil and Political Rights (iccpr).127 The next sections will analyse first the Committee’s general comments, and then its views on individual complaints, to see whether and how it has dealt with situations in which non-State actors have interfered with the enjoyment of human rights (i.e. whether and how it engages with the horizontal effect of human rights).
4.1.1 General Comments of the Human Rights Committee
The HRCtee ‘takes its authority … from article 40, paragraph 4, of the Covenant, which provides that it may transmit “such general comments as it may consider appropriate” to all States parties’.128 The general comments of the HRCtee apply to all State Parties to the iccpr. The HRCtee has referred to various types of horizontal effect in its general comments, the most pertinent examples of which will now be discussed. It should be mentioned at the outset that the HRCtee’s interpretations of the iccpr have often been heavily criticised for its forthcoming and at times overreaching nature,129 which ‘preclude[s] any claim that the assertions made […] represent an international consensus of any kind’.130 Although the Committee may at times appear to overstep the mark, its interpretations related to non-State actors and horizontal effect do not go beyond the scope of interpretation allowed by Article 31 of the Vienna Convention on the Law of Treaties.131
The most explicit discussion of horizontal effect in relation to the iccpr by the HRCtee can be found in General Comment No. 31.132 The general comment focuses on the definition and scope of obligations found within the iccpr. In the context of obligations for non-State actors, the Committee acknowledged that ‘as a matter of international law’, the obligations within the iccpr ‘do not, as such, have direct horizontal effect’.133 However, it did go on to elaborate how indirect horizontal effect can be applied to obligations contained in the Covenant. Commenting upon Article 2(1) iccpr (declaring the nature of State obligations), the HRCtee held that States do indeed have positive obligations which require them to protect individuals ‘against acts committed by private persons or entities that would impair the enjoyment of Covenant rights’.134 Specifically, the Committee determined the positive obligations to mean that ‘State Parties’ permitting or failing to take appropriate measures or to exercise due diligence to prevent, punish, investigate or redress the harm caused by such acts by private persons or entities’ could result in a violation of a Covenant right.135 This clearly reflects States’ obligations to protect human rights from interference by third parties, in particular the standard of due diligence introduced above.
However, the statements in General Comment No. 31 were not positively received by some States. The Unites States of America (us), for example, published observations on the comment condemning the Committee’s assertions as ‘sweep[ing] too broadly and categorically’.136 Rejecting the idea put forward by the Committee that all human rights within the Covenant contain positive obligations (drawn from the Committee’s use of Article 2(1) as the basis for positive obligations), the us believed that each right must be considered separately in order to determine whether, and to what extent, the State could be expected to take positive action to ensure its protected enjoyment.137 This more restrictive view of the us does not correspond very well with the views of other human rights bodies as to the positive obligations and the extent of the duty of due diligence under international human rights law (see below). The argument appears to value distinguishing rights from one another on the basis of rights, rather than obligations. This seems to clash with the tripartite typology of human rights (which by now has been widely accepted as applying equally to all human rights).138 The reluctance could stem from a (now outdated) belief that only economic, social and cultural rights require positive action by States, while civil and political rights should be equated with negative obligations only (i.e. an obligation to refrain from interfering with the enjoyment of rights). The us objected, in particular, to the consequence of the general comment that States would have a positive obligation to protect individuals from torture by private actors (this does not preclude, however, that they would be willing to accept a positive obligation to protect individuals from torture by public actors).139 The us found it paradoxical that such an effect could be read into the prohibition of torture provided in Article 7 iccpr, without the Convention against Torture (which was subsequently specifically adopted for combatting torture) explicitly including a positive obligation. The argument here referred to the requirement in Article 1 of the Convention against Torture (cat) that torturous acts be done ‘by, at the instigation of or with the consent or acquiescence of a public official’, with no mention of non-State actors.140 As will be explained in detail below, the Committee against Torture has itself interpreted Article 1 cat in the same way as the HRCtee, allowing for some degree of indirect horizontal effect.
Although at first sight the statement in General Comment No. 31 seems to be quite far-reaching, the HRCtee may have curtailed its effect. The Committee restricted the extent of due diligence by stating that States’ positive obligations only apply to rights ‘so far as they are amenable to application between private persons or entities’.141 The lack of explanation of what is meant by this, however, dims the potential of this caveat to mollify States like the us. Because the HRCtee’s general comments are not legally binding, for the time being States may be able to evade (binding) legal responsibility142 for positive obligations under the iccpr, especially for those rights in relation to which the text of the Covenant does not expressly place such an obligation on States (unless and until such time as this interpretation becomes a matter of customary international law).143
General Comment No. 31 does give an important overview of the HRCtee’s interpretation of human rights obligations within the iccpr as a whole. However, the Committee has also published general comments pertaining to specific rights, which also involve horizontal effect and actually pre-date General Comment No. 31. It may be possible, therefore, for the earlier comments of the Committee to clarify the meaning of its more general observations on Article 2. The following discussion will shed light on which rights the Committee seems to view as being ‘amenable to application between private persons or entities’.
The rights-specific General Comments referring to horizontal effect of iccpr rights include Comments 7, 16, 18, 20, 27, and 28.144 The comments in these documents range from direction as to what standards States are expected to meet in relation to the relevant rights in the Covenant, to how States should fulfil their obligation to protect individuals, or calls for States to improve and provide evidence of the measures they take to protect individuals.
For example, General Comment No. 7 explicitly mentioned the State obligation to protect individuals from torturous acts, stating that ‘the scope of protection required goes far beyond torture as normally understood’, and, significantly, that ‘it is also the duty of public authorities to ensure protection by the law against such treatment even when committed by persons acting outside or without any official authority’.145 This strong proclamation was superseded by more tempered wording by General Comment No. 20 which replaced No. 7. Nevertheless, General Comment No. 20 still requests that in their periodic reports to the Committee, State Parties provide information on the ‘criminal law which penalize[s] torture and cruel, inhuman and degrading treatment or punishment’ by any actor, including by private persons.146 It further requires that complaints of torture ‘must be investigated promptly and impartially by competent authorities’.147 While there is no specific mention of this relating to complaints against private persons, it can be assumed from the previous quotation that this would be expected, since the Committee sees the State as being responsible for protecting individuals from torture by private actors.
Regarding the right to privacy, the HRCtee stated in General Comment No. 16 that the ‘right is required to be guaranteed against all such interferences and attacks whether they emanate from State authorities or from natural or legal persons’148 – an implicit reference to the obligation to protect. The Committee went on to explain what kind of measures this required. In particular, it mentioned that States must regulate by law the ‘gathering and holding of personal information on computers, data banks and other devices’ by public and private actors, ensure that individuals can ‘ascertain which public authorities or private individuals or bodies control or may control their files’ and ‘have the right to request rectification or elimination’ of files collected or processed unlawfully or that contain incorrect personal data.149 Here, the HRCtee has gone quite far, compared to its comments regarding other rights, in detailing how the right to privacy should be safeguarded against interference by private actors.
The HRCtee also mentions specific measures to be taken vis-à-vis non-State actors in General Comment No. 18 on the subject of non-discrimination. The HRCtee required States to provide evidence of their ‘legal provisions and administrative measures directed at diminishing or eliminating’ discrimination carried out by private actors.150 General Comment No. 28 on the equality of rights between men and women requests States to ‘report on any laws and public or private actions that interfere with the equal enjoyment by women of the rights under article 17, and on the measures taken to eliminate such interference and to afford women protection from any such interference’.151 These comments clearly impose the obligation to protect upon States, requiring due diligence to be taken by implementing laws to prevent and punish interference with the enjoyment of (particularly women’s) rights by private actors.
It is very interesting that the majority of the instances in which the Committee has spoken of or implicitly applied indirect horizontal effect of rights within the iccpr have been in relation to groups that are generally considered to be ‘vulnerable’. In particular, the Committee has upheld the obligation to protect with regards to women and individuals deprived of their liberty or freedom of movement. Another example is General Comment No. 27 on freedom of movement, which demonstrates the importance that the HRCtee places on the obligation to protect to women as a vulnerable group. In this comment, the Committee opined that individuals must be ‘protected not only from public but also from private interference’, the ‘obligation to protect [being] particularly pertinent’ in relation to women.152
It could be argued that General Comment No. 31 attempts to take this obligation out of the realm of the enjoyment of specific rights by particular groups by upholding indirect horizontal effect more generally. To identify the Committee’s approach to horizontal effect in a more holistic manner, the jurisprudence resulting from its individual complaints procedure under Article 1 of the Optional Protocol to the iccpr will also be briefly assessed, in so far as it adds value to the findings of the General Comments.
4.1.2 Views of the Human Rights Committee
Pursuant to Article 1 of the Optional Protocol to the iccpr, the HRCtee can hear individual communications ‘from individuals subject to [a State Party’s] jurisdiction who claim to be victims of a violation by that State Party’.153 The Optional Protocol entered into force in 1976, and has resulted in almost 3,000 communications being brought before the Committee.154 Of those cases in which a ‘view’ was adopted, it is possible to see evidence of indirect horizontal effect.
For example, the case of B.d.B. v. The Netherlands concerned non-discrimination under Article 26 iccpr.155 The communication was actually declared inadmissible by the Human Rights Committee, but remains interesting due to the Committee’s comments regarding the author of the alleged human rights violation. The claimants alleged that a non-State actor, the Industrial Insurance Board for Health and for Mental and Social Interests, had discriminated against them in relation to social security contributions.156 The Netherlands wished to rely on the fact that the Board is a non-State actor as one of its arguments against admissibility of the complaint. The main argument was that because the Board is an independent body (established merely to ‘implement social security legislation’),157 the Dutch State could not ‘influence concrete decisions’ of the Board.158 Because of this, the Netherlands argued that it could not be held responsible for discrimination on behalf of the Board. The HRCtee, however, had a different point of view. It observed that the Dutch State was ‘not relieved of its obligations under the Covenant when some of its functions are delegated to other autonomous organs’.159 Through this statement, the Committee seems to imply that the Netherlands remains responsible here because the conduct of the non-State actor could be attributed to the State. Although the State was not directly controlling the Board, it had delegated traditionally ‘public’ functions to it, and thus would have retained responsibility for any discrimination that would have arisen from the Board’s decisions (had the case been assessed on its merits).
The reasoning in B.d.B. v. The Netherlands was later reiterated and to some extent also clarified by the Committee in Cabal and Pasini v. Australia.160 The latter communication was brought in relation to alleged discrimination by a privatised prison. The HRCtee again held that ‘contracting out to the private commercial sector of core State activities which involve the use of force and the detention of persons’ does not diminish the responsibility of the State; it remains an obligation of the State to ensure that the rights of individuals are protected vis-à-vis these private actors, the conduct of whom could be attributed to the State.
The HRCtee has discussed the obligation to protect in other cases that take place in a ‘quasi-public sphere’. In discussing the admissibility of the case of Nahlik v. Austria, for example, the HRCtee explicitly referred to the State’s obligation to protect individuals from discrimination ‘among private parties’.161 This seems to reiterate General Comment No. 31, but its significance for horizontal effect is weakened due to the ‘quasi-public’ context. The case concerned alleged discrimination that had occurred through a private agreement. The Austrian State tried to rely on this fact to have the communication declared inadmissible. However, the HRCtee reaffirmed that States must protect individuals from discrimination ‘within the public sphere or among private parties in the quasi-public sector of, for example, employment’.162 Additionally, because agreement had to be confirmed by a public body before it could enter into force, the State was deemed to be in a position to protect the individual from discrimination through the agreement. Ultimately, the case was declared inadmissible on other grounds. Nevertheless, the statement made about private parties does demonstrate that the HRCtee believes non-discrimination to be one of the rights ‘amenable to application’ between private actors.
In contrast, the right to participation in the public life of the nation (protected by Article 25 iccpr) does not seem to be considered as amenable in the same way. In the case of Karakurt v. Austria the HRCtee emphasised that Article 25 could ‘not extend to cover private employment matters’.163 Interestingly, in another case involving Article 25 (Arenz v. Germany), the Committee did suggest that the right to freedom of religion was subject to the ‘obligation to ensure’ all rights in the iccpr, including Article 25, and in relation to private associations as well as State actors.164 The complaint centred on the fact that a religiously-founded political party had denied the claimants membership of the party due to their membership of a different ideological organisation. The claimants argued that through upholding this decision, the German national courts had interfered with their right to take part in public affairs. The Committee rejected this argument, however, on the grounds that it was not the function of the Committee (not being an appeals court) to re-evaluate the German national courts’ application of German domestic law, as long as there was no arbitrariness or denial of justice.165 Despite not going to the merits, this case does show, to a limited degree, that a right so inherently linked to the public sphere can also be amenable to relationships between private actors. Because the applicants did not focus on this in their communication, however, the HRCtee restrained itself from ‘address[ing] the broader issue of what legislative and administrative measures’ must be taken to ensure the enjoyment of Article 25.166
4.1.3 Critical Remarks on the Jurisprudence of the Human Rights Committee
It may be concluded from the above examples that the HRCtee does indeed engage with the idea that non-State actors could cause infringements of human rights. So far, the Committee has not been particularly radical in its application of the obligation to protect human rights. Despite the backlash surrounding General Comment No. 31, it appears to have been based on a line of previous general comments within which the Committee elaborated on States’ positive obligations to protect human rights from both public and private actors. It cannot be said, however, that the Committee has been as extensive in its views on individual communications. It is likely that this is because of the restrictions of what the Committee could discuss imposed by the subject matter of the complaint itself (e.g. in Arenz v. Germany) or simply because the cases in which issues of horizontal effect have arisen have tended to be declared inadmissible, thus not warranting a detailed examination or application of the relevant provisions vis-à-vis non-State actors.
What is clear is that when the HRCtee engages with the obligation to protect, it requires certain measures to be taken by States to ensure that human rights protection is engrained on an institutional level within States (i.e. within their legal or administrative frameworks). As with other human rights monitoring bodies, when faced with human rights interference by a privatised actor or with a situation where the State has delegated certain ‘public’ tasks, the HRCtee has not hesitated to attribute the actions to the State. It also has not shied away from applying States’ obligation to protect in relation to rights that are often affected within private relationships, such as privacy and non-discrimination, although the HRCtee seems to tread more carefully with rights that are of a more public nature (e.g. Article 25 iccpr). In terms of concrete indirect obligations for non-State actors at the international level (through direct obligations imposed by domestic law), the hrc has remained quite general. The actual standards that States are expected to hold non-State actors to prevent them from interfering with human rights are very vague. Most aspects of the obligation to protect that are discussed by the Committee go either to procedural obligations (i.e. due diligence), or simply state that, for example, non-State actors should not be permitted to discriminate. The HRCtee therefore predominantly stays within the realm of State’s direct obligations rather than non-State actors’ indirect ones.
Ultimately, whatever approach the HRCtee takes in its interpretation and application of the iccpr, it remains limited by the international human rights framework, its non-binding nature and, regarding the individual communications procedures, the very cases that appear before it.
4.2 Committee on Economic, Social and Cultural Rights
The Committee on Economic, Social and Cultural Rights (CteeESCR) has made quite widespread implicit references to indirect horizontal effect in relation to the International Covenant on Economic, Social and Cultural Rights (icescr).167 As with the HRCtee, the CteeESCR’s general comments will be addressed before moving on to its jurisprudence.168 The CteeESCR is not technically a ‘un treaty body’ in the same sense as the other bodies discussed here, but will be treated as an equivalent body.169
4.2.1 General Comments of the Committee on Economic, Social and Cultural Rights
The CteeESCR has consistently applied the tripartite typology of human rights, and focuses most of its consideration of indirect horizontal effect on the State obligation to protect.170
Even before the CteeESCR engaged with the full typology explicitly, it was applying some degree of indirect horizontal effect by applying States’ protective obligations regarding human rights. For example, in General Comment No. 5 on persons with disabilities, the CteeESCR emphasised that the private sphere must be appropriately regulated by the State to ensure the protection of people with disabilities from inequitable treatment.171 This could (and indeed has by some scholars) be seen as an ‘obligation to regulate’ private actors.172 Going into more detail, CteeESCR explicitly mentioned that ‘it is essential that private employers, private suppliers of goods and services, and other non-public entities be subject to both non-discrimination and equality norms in relation to persons with disabilities’ – an idea that was reaffirmed in General Comment No. 14 on the right to health.173 The statement in General Comment No. 5 cannot be read as implying that the CteeESCR is in favour of subjecting private actors directly to international human rights law. Rather, it requires States to adopt national norms concerning economic, social and cultural rights which must be applied within the private sphere (i.e. in private relationships). Indeed, the CteeESCR emphasised that the ‘ultimate responsibility’ for ensuring that people with disabilities’ Covenant rights are complied with lies with the State.174 The CteeESCR appeared to be particularly concerned that this was the case in relation to privatised public services – a concern consistent with that expressed by the HRCtee above. Indeed, very similar language was used by the CteeESCR here as by the HRCtee in B.d.B. v. The Netherlands; the CteeESCR pointed out that delegation of activities by the State to private actors does not ‘absolve’ the State of its international obligations, whereas the HRCtee explained that the same did not ‘relieve’ a State of its obligations.175
In the context of the right to food, the CteeESCR made similar claims regarding States’ obligation to protect and the fact that this necessitates State regulation of the private sphere, in General Comment No. 12.176 Here, the CteeESCR reiterated that ‘[v]iolations of the right to food can occur through the direct action of States or other entities insufficiently regulated by States’, confirming that violations can occur when the harm is directly caused by non-State actors.177 A subsequent general comment on the right to water extended the requirement of a regulatory regime of private providers of public services to include a punitive aspect in case of non-compliance by the private actor.178 Failing to adopt an effective regulatory system would, according to the Committee, result in a violation of the obligation to protect, and therefore of Covenant rights.179 A similar statement can be found again in General Comment No. 23 on the right to just and favourable conditions of work. In the context of States’ obligation to protect human rights, the CteeESCR emphasised that States must ‘tak[e] steps to prevent, investigate, punish and redress abuse through effective laws and policies and adjudication.’180 Taken together, these general comments not only create a framework for the obligation to regulate private actors, but also expound the Committee’s view of what the duty of due diligence requires of States. Curiously, the CteeESCR did not mention an explicit duty of ‘due diligence’ until General Comment No. 16, and since then, it has not been consistent in using the term. This is not necessarily significant, as it can be argued that by providing States with guidance as to how they can fulfil the obligation to protect individuals, and what measures they should take for the prevention and punishment of non-State actors for certain conduct (such as the inclusion of the punitive aspect of regulatory regimes), the Committee is implicitly referring to due diligence, or at least leads to comparable obligations. This is particularly persuasive when the explicit language in different general comments is compared. Explicit reference was made in General Comment No. 16, in which the CteeESCR calls upon States to ‘act with due diligence to prevent, investigate, mediate, punish and redress acts of violence against them by private actors’.181 This is almost identical to the language of General Comment No. 23, quoted above. The reason for the inconsistency is unclear, although an explicit use of ‘due diligence’ returns repeatedly in the Committee’s most recent general comment on business and human rights (see below). As with the HRCtee, the CteeESCR’s treatment of non-State actors so far does not contain concrete or detailed standards against which they should be held, even at the national level. Rather, because of the focus on due diligence and regulation, there is again an emphasis on States’ procedural obligations. It could be assumed that the substantive standards to which non-State actors should be held therefore match those of the State (at least in terms of defining particular terms, such as ‘access’ to certain services, or in determining whether, for example, discrimination has occurred). On the other hand, it could also be assumed that the standards fall within States’ margin of appreciation.
The CteeESCR has also discussed potential responsibilities of non-State actors themselves, for example in General Comment No. 12 on the right to adequate food. The CteeESCR emphasised (again) the fact that only States retain ultimate accountability for ensuring the right to food.182 Significantly, though, it also highlighted that all members of society have a role to play in the realisation of this right.183 This would seem to go further than the (generally accepted) assertion that non-State actors may have an international responsibility (although not a legal obligation) to respect human rights. The statement suggests that private actors may have a more ‘positive’ role vis-à-vis the right to food. While this is suggested by the wording of the comment, the CteeESCR has not made it explicit. In fact, in General Comment No. 12 the CteeESCR seemed to be more concerned with the actions of private businesses interfering with human rights and the State’s obligation to prevent this, rather than any positive action that private businesses themselves could or should be making to contribute to the realisation of the right to food. Although not going into the positive actions of business vis-à-vis human rights realisation, the CteeESCR has gone so far as to explain how businesses should act so as to avoid interfering with the right to food (i.e. by operating ‘within the framework of a code of conduct conducive to respect of the right to adequate food, agreed upon jointly with the Government and civil society’).184 This goes further than the comments discussed until this point, as it makes explicit reference to the possibility that non-State actors could have some responsibilities to take action, at least in terms of a responsibility to respect the right. The CteeESCR repeated the relevant passage almost verbatim in its General Comment No. 14 on the right to health.185 The mention of a collaboration between a business and the State, and of the ultimate responsibility of the State, does temper the horizontal effect within the general comments to some extent though, and (as with General Comment No. 5) cannot be read as imposing direct (and in any case not binding) international obligations on non-State actors. Thus, this Comment remains an example of indirect horizontal effect.
Although the comments thus far have actually been relatively extensive in dealing with non-State actors, a noticeable shift occurred in General Comment Nos. 14 and 15, which include a section specifically entitled ‘Obligations of actors other than States’. This suggests that the CteeESCR does in fact envisage more direct, and possibly even binding, human rights obligations for non-State actors. In the two comments the Committee mentions, for example, obligations of international organisations that have ties with the right to health and water, respectively (naming, for example, the World Health Organization, the Food and Agricultural Organization and unicef).186 The obligations included mostly involve giving support and aid to States in the implementation of their Covenant rights, although international organisations (including international financial institutions) should also incorporate ‘human rights law and principles in the[ir] programmes and policies’.187 Although the Committee does actually say that the obligations of the international organisations fall under the scope of human rights obligations, it is significant that in General Comment No. 15 the CteeESCR notes that the cooperation and support of the international organisations will be considered when assessing how capable States are of realising rights within their jurisdiction,188 thereby implying that the obligations of the international organisations are related to the realisation of human rights, and could have a bearing on the standards expected of states in terms of their obligation to fulfil human rights. The fact that these obligations are only mentioned as existing for international organisations could be telling here, especially as other non-State actors (e.g. privatised companies) often contribute to the obligation to fulfil human rights by providing certain services without having an obligation to do so.189 We normally assume that non-State actors are dealt with within national legal frameworks. However, international organisations, by their nature, operate in the international sphere, and are directly subject to some international laws (although not to date those found in human rights treaties). This makes it less of a stretch to imagine that the CteeESCR was directly involving international organisations in human rights realisation, particularly when the operations of many international organisations are connected to human rights in some way (e.g. the World Bank, or even the un itself).
Significantly, though, in more recent general comments, the CteeESCR has even taken one step further to say that private actors do actually have an ‘obligation’ (rather than a responsibility) to respect certain human rights. In General Comment No. 22 on the right to sexual and reproductive health, the CteeESCR requires State Parties to avoid taking retrogressive measures that could, inter alia, ‘reduce oversight by States of the obligation of private actors to respect the right of individuals to access sexual and reproductive health services’.190 The language used here is significant, although admittedly it is not clear whether the ‘obligations’ envisaged for private actors exist at the national or international level (the assertion having been made in the context of the State’s obligation to protect human rights). Although the statement is quite narrow in terms of substance, the broad term ‘private actors’ suggests that all private actors have an obligation not to interfere with access to sexual and reproductive health services. Whether or not the statements can really be read as an understanding that non-State actors do have human rights obligations, they still cannot result in legally binding obligations at the international level. As explained, general comments do not have a legally binding nature and cannot create obligations themselves, but rather constitute the (arguably authoritative) opinion of the CteeESCR as to what standards the Covenant rights demand.
The CteeESCR has very recently adopted a general comment on business and human rights (General Comment No. 24),191 in which it explains in more detail what obligations States have regarding human rights interference by businesses. Interestingly, the comment includes measures under the obligations to respect, protect and fulfil human rights, rather than just the obligation to protect. Although the document is directed predominantly at States, its relevance for ‘the corporate sector’ is also emphasised, as it seeks to help them ‘in discharging their human rights obligations and assuming their responsibilities’.192 Although the language of ‘obligations’ as well as responsibilities here is striking, it is unclear whether the CteeESCR is referring to obligations of businesses at the national or international level, and the CteeESCR later clarifies that it ‘only deals with the conduct of private actors … indirectly.’193 Of course, the comment is also of interest to individuals seeking redress against States for human rights interference by businesses.
This is the first time that the Committee has explicitly referred to the dasr. It seems that for the most part, the monitoring body seems to mention attribution in connection with the obligation to protect human rights, but without explaining how this fits within the framework of the dasr (or indeed whether it actually needs to). The fact that they are quoted in the general comment is therefore significant, and it will be interesting to see whether reference is also made in future general comments.
(a) if the entity concerned is in fact acting on that State party’s instructions or is under its control or direction in carrying out the particular conduct at issue, as may be the case in the context of public contracts; (b) when a business entity is empowered under the State party’s legislation to exercise elements of governmental authority or if the circumstances call for such exercise of governmental functions in the absence or default of the official authorities; or (c) if and to the extent that the State party acknowledges and adopts the conduct as its own.195
Following the tripartite typology of human rights, the CteeESCR details what States should do under each category of obligation. Under the obligation to protect human rights, this includes:
- –‘prevent effectively infringements of economic, social and cultural rights in the context of business activities’196;
- –‘adopt legislative, administrative, educational, as well as other appropriate measures to ensure effective protection against Covenant rights violations linked to business activities’197 (e.g. the adoption of criminal sanctions, imposing due diligence standards on businesses and conducting ‘direct regulation and intervention’ where necessary)198; and
- –to put in place ‘effective monitoring, investigation and accountability mechanisms’ to enable ‘those whose Covenant rights have been violated in the context of business activities’ access to effective remedy.199
The CteeESCR is explicit in laying down its general understanding of the obligation to protect, which it understands as being violated by ‘a failure by the State to take reasonable measures that could have prevented the occurrence’ of a violation caused by a private actor, even when the private actor was not the sole cause of the violation and even (as long as it was reasonably foreseeable) when the State did not foresee the violation.200 The comment also provides considerable detail regarding precise legislative, administrative and other measures that could or should be adopted (and enforced) by States to regulate and monitor businesses’ effects on economic, social and cultural rights in certain contexts (e.g. in relation to the tobacco industry, the housing market, the education and employment sectors, among others).201 In this regard the CteeESCR also reiterates the obligation to regulate non-State actors, especially in the context of privatisation and the private provision of public services.202 As well as explaining the obligations of States the CteeESCR elaborates on the due diligence standards that States should hold businesses to, including to ‘act with due diligence to identify, prevent and address abuses to covenant rights’ by their subsidiaries and business partners.203 What can be seen from the general comment is that the CteeESCR applies indirect horizontal effect predominantly through States’ obligation to protect human rights (including due diligence), but also through attributing non-State conduct to States in certain situations. Interestingly, the other monitoring bodies also tend to mention (without explaining in legal terms) attribution in the context of the obligation to protect.
4.2.2 Views of the Committee on Economic, Social and Cultural Rights
The CteeESCR has the competence to hear individual complaints pursuant to the Optional Protocol to the International Covenant on Economic, Social and Cultural Rights.204 However, as the Optional Protocol only entered into force in 2013, there is relatively little jurisprudence from the body as compared to the other un human rights monitoring bodies.205 The communications that have been considered by the Committee so far have not dealt with the horizontal effect of human rights, and will therefore not be discussed here.
4.2.3 Critical Remarks on the Jurisprudence of the Committee on Economic, Social and Cultural Rights
Overall, the CteeESCR’s general comments are quite telling of its attitude towards indirect horizontal effect. It has repeatedly referred to the State obligation to protect human rights, requiring specific action to be taken to regulate and control the actions of non-State actors, particularly those with a connection to the State (i.e. privatised companies) and actors that have been delegated certain public tasks. The fact that the Committee has adopted comment specifically on business and human rights is significant. Regarding the CteeESCR’s interpretation and application of the obligation to protect, the same can be said as for the HRCtee – although there are now detailed recommendations for States to take certain action to protect human rights, it is not clear to which standards exactly non-State actors must be upheld at the national level. What is clear from its general comments is that the CteeESCR has a strong belief that non-State actors do have at least a role, and perhaps even obligations, within the international human rights framework. Thus, while the Committee has mostly confined itself to interpreting icescr in a way that results in indirect horizontal effect, it has gone further than the HRCtee in alluding to direct horizontal effect.
Significantly, the Committee has even suggested that some non-State actors have a more active or positive role to play in the realisation of rights, with an obligation to help States with their own implementation. This already pushes the boundaries of the international human rights framework, which does not envisage human rights obligations for non-State actors. Getting to the point where the Committee could legitimately (i.e. within its mandate) and explicitly detail human rights obligations for non-State actors (beyond those to be imposed within domestic law) would require a change of the international legal framework. For such statements to have a practical effect, State Parties to icescr would then need to act upon them. Of course, in theory, the Committee could choose to go down this road without a change in the framework or its mandate, but this is likely to be met with much resistance from disgruntled States that could choose to simply ignore the CteeESCR. Until the international human rights framework is expanded from a State-centric system, or alternatives are found outside of the legal framework, it is unlikely that more horizontal effect will be achieved in relation to the Covenant on Economic, Social and Cultural Rights.
4.3 Committee on the Elimination of Discrimination against Women
The Committee on the Elimination of all Forms of Discrimination Against Women (CteeEDAW) has also faced situations necessitating an application of horizontal effect of human rights. The following sections will assess the extent to which the CteeEDAW has applied this obligation in practice, looking first at its general recommendations before discussing its relevant jurisprudence.
4.3.1 General Recommendations of the Committee on the Elimination of Discrimination against Women
Significantly, non-State actors are mentioned explicitly in the Convention on the Elimination of All Forms of Discrimination Against Women (cedaw) itself. Article 2(e) of the Convention imposes a general obligation on State Parties to ‘take all appropriate measures to eliminate discrimination against women by any person, organization or enterprise’.206 General Recommendation No. 28 on the core obligations of State parties under Article 2 affirmed that the provision embodies a positive obligation for States to ensure that women are not subject to discrimination by non-State actors, including by ‘national corporations operating extraterritorially’ (i.e. similarly to the standards upheld in the Trail Smelter case regarding transboundary harm).207 In this Recommendation, the CteeEDAW also mentioned the attribution of non-State actors’ conduct to the State in some situations, although it did not explain when this would be the case or refer to the dasr. Rather, it simply noted that States are ‘thus’ obliged to take ‘appropriate measures’, including ‘the regulation of the activities of private actors with regard to education, employment and health policies and practices, working conditions and work standards, and other areas in which private actors provide services or facilities, such as banking and housing’.208 Again, this seems to either conflate ‘attribution’ as a concept within public international law with the obligation to protect, or to use it as a separate term. The approach of the CteeEDAW in this respect is unclear. As with the CteeESCR, the obligation to regulate is clear here, and seems to apply to all private actors that are providing services of a public nature.
Similar obligations were mentioned in CteeEDAW’s General Recommendation No. 19 on violence against women. In this recommendation the CteeEDAW emphasised the importance of States ‘tak[ing] appropriate and effective measures to overcome all forms of gender-based violence, whether by public or private acts’.209 It had previously emphasised that discrimination against women ‘is not restricted to action’ by State actors, but that ‘States may also be responsible for private acts’ if they do not act with due diligence.210 Indeed, the general recommendation lists instances in which violence against women can ‘result from the acts or omissions of State or non-State actors’, including, inter alia, where a State has delegated public tasks, e.g. security within places of detention, to private actors.211 In specific recommendations, the Committee suggested action that States should take relating to particular non-State actors, such as the media, people in the workplace and family members. The measures to be taken included the adoption and implementation of legislation, provision of services, and other ‘preventive, punitive and remedial measures’ which States should report to the Committee about.212 The wording here suggests that taken together, the measures in the recommendations would constitute the precise standards of due diligence expected by the CteeEDAW. The Committee also mentioned due diligence in Recommendation No. 19 in its explanation of how the State could be held responsible for the conduct of non-State actors (i.e. when they fail to ‘act with due diligence to prevent violations of rights or to investigate and punish acts of violence, and for providing compensation’213). Interestingly, in this instance, the CteeEDAW explained such responsibility without mentioning attribution.
In General Recommendation No. 24, the CteeEDAW (like the HRCtee and the CteeESCR), expressed concern about States trying to pass their human rights obligations to private actors when States delegate what are traditionally public functions to these actors.214 The CteeEDAW reiterated the opinion of the HRCtee and the CteeESCR that States ‘cannot absolve themselves of responsibilities’ by delegating public tasks.215 This has the effect of ruling out any direct horizontal effect of human rights contained within cedaw that would occur if responsibility could be passed onto the non-State actors. However, the widespread concern of harmful actions of private actors operating in the ‘public’ sphere also serves to emphasise and recognise the impact that such non-State actors can have on human rights. Still, though, there is no explanation of what substantive obligations States should impose on non-State actors in order to ensure that they do not cause harm to individuals’ human rights. The CteeEDAW’s unease also reflects qualms regarding the attitude of States (although no examples were provided by cedaw), that delegating functions to private actors would mean that the State no longer has to deal with the human rights aspects related to that function. What is not addressed by the CteeEDAW, or indeed the other monitoring bodies addressed so far, is the possibility of concurrent or shared responsibility for non-State actors that are delegated State functions – it could be possible for both the State to retain responsibility as well as confer a degree of responsibility on the non-State actor.216 So far, as we have seen with the two previous treaty bodies discussed, this has only been deemed to be appropriate at the national, and not the international, level. Indeed, at the international level it is not considered possible for States to ‘replac[e], legally delegat[e] or chang[e]’ their de jure responsibility when they outsource certain tasks to non-State actors.217
Further explanation of the State obligation to protect women was provided in General Recommendation No. 34.218 The CteeEDAW adopted the same approach as the CteeESCR by imposing an obligation to ‘regulate the activities of domestic non-State actors within their jurisdiction, including when they operate extraterritorially.’219 Specifically, reiterating its General Recommendation No. 28, the CteeEDAW requests States to ‘prevent any actor under their jurisdiction, including private individuals, companies and public entities, from infringing or abusing the rights of rural women outside their territory’.220 As well as the obligation to regulate, General Recommendation No. 34 includes implicit reference to due diligence vis-à-vis non-State actors, requiring States to ‘[t]ake effective measures aimed at preventing, investigating, prosecuting and punishing acts of violence against rural women and girls, including migrant rural women and girls, whether perpetrated by the State, non-State actors or private persons’.221 Similar to its approach in Recommendation No. 19, the CteeEDAW goes on to elaborate more precise recommendations to protect human rights and uphold due diligence (this time, in relation to different subject areas, e.g. ‘adequate living conditions’).222
While the obligation to protect is paramount in situations where private actors carry out State functions (and indeed more generally in the purely ‘private’ sphere as well), the indirect horizontal effect being applied unfortunately does not go very far in preventing human rights interference by actors outside of the control of the State, in situations where domestic law and policies are not effective in manipulating the behaviour of private actors (as explained in Section 3.3.1, due diligence is an obligation of conduct, not result). Interestingly though, in General Recommendation No. 25, the CteeEDAW noted that merely averring ‘powerlessness’ or succumbing to ‘predominant market or political forces’ is not enough for States to justify a failure to fulfil their obligation to protect women’s rights from the actions of private actors.223 In doing so, the Committee reaffirms that cedaw holds States accountable for the actions of private actors such as businesses, which are in a position to assert influence over the State.224
Further reiteration of this, provided in some detail, can be found in the CteeEDAW’s General Recommendation No. 30.225 This Recommendation deals with the treatment of women during and after armed conflicts, and goes quite far in detailing the obligations of States to act with due diligence as well as (crucially) the behaviour expected of non-State actors towards women during these times.226 Such recommended action includes refusing to reduce the protection afforded to women in order to mollify non-State actors,227 engaging with non-State actors,228 and helping non-State actors to act in a manner consistent with the Convention, in particular to ‘address and assess’ risks of human rights violations.229 This standard could also be included, for example, in the codes of conduct of businesses alluded to by the CteeESCR above. The context of armed conflict here arguably allows the CteeEDAW to go further in describing the obligations of non-State actors, because under international humanitarian law non-State actors are subject to direct obligations.230 Breaches of some of these obligations, as well as ‘gross violations of human rights’ by non-State actors (an anomaly in terms of language), can result in these actors being held (individually) criminally responsible at the international level.231 By mentioning these obligations and the connections that can be made between international humanitarian and human rights law during armed conflicts, the CteeEDAW allowed itself space to mention the human rights obligations of non-State actors during (or after) armed conflicts in their own right. The actors specifically targeted by the Recommendation were non-State armed groups,232 which the CteeEDAW explicitly noted as having an obligation to respect human rights (particularly when they have an ‘identifiable political structure’ and ‘significant control over territory and population’).233 Notably, this assertion was made after the CteeEDAW acknowledged the fact that non-State actors cannot become party to international human rights treaties.234 Unfortunately though, the strength of the comments is diminished by the lack of legal basis provided by the CteeEDAW to justify how non-State actors could be said to hold these obligations. Nevertheless, the Committee continued not only to make recommendations to States in the general recommendation, but also directly to non-State actors. For example, besides calling on non-State armed groups to respect women’s rights,235 the CteeEDAW (like the CteeESCR did in relation to private providers of public services),236 suggested that the groups ‘commit themselves to abide by codes of conduct on human rights’.237 Again, the common downfall of treaty bodies’ documents comes into play here, and the non-binding nature of General Recommendation No. 30 significantly reduces the impact that the recommendations may have in practice. Yet, the statements clearly show that the CteeEDAW is willing to entertain the notion of at least human rights responsibilities for certain non-State actors.
4.3.2 Views of the Committee on the Elimination of Discrimination against Women decisions
Many of the complaints filed with the CteeEDAW involve action taken by non-State actors as well as (inaction) by State actors. For reasons of space, some examples will be chosen which show clearly the ways in which the CteeEDAW engages with the private sphere and interference with human rights by non-State actors. The cases used rely to some extent on the provision in Article 2(e) cedaw, which requires States to eliminate discrimination against women by private as well as public actors.
First, the application of due diligence by the CteeEDAW can be clearly seen in its jurisprudence relating to domestic violence suffered by women at the hands of private actors. The vast majority of the cases before the CteeEDAW involving horizontal effect concern instances of domestic violence, and failures of the State Party to provide effective protection for women from (for example) their partners. An oft-quoted example of this is the case of A.T. v. Hungary.238 In this case the Committee explicitly adopted the tripartite typology of human rights and the duty of due diligence.239 The author of the complaint had repeatedly suffered domestic abuse and threats had been made against her children by her former partner. Despite the author bringing several civil and criminal proceedings against the husband, the Hungarian State had failed to protect both her and her children.240 Quoting its comments in General Recommendation No. 19 on due diligence, the Committee found Hungary to have violated the woman’s rights due to its failure to effectively protect her from her former common law husband.241
A similar conclusion was reached by the CteeEDAW in the recent decision of Angela González Carreño v. Spain,242 in which the author had also suffered repeated abuse at the hands of her partner, culminating in the murder of her daughter by him.243 The Committee’s focus in this case was also on due diligence, looking specifically at Spain’s failure to conduct an investigation or inquiry into the situation being suffered by the author and her daughter and failing to provide adequate supervision, despite over thirty complaints and requests made by the author to the State asking for protection.244 Rather than fulfilling its duty of due diligence, the CteeEDAW found that Spain had made assumptions without analysing the specific situation of the author. In the context of the complaints made by the author in this case, the Committee’s approach could be compared to the aspect of ‘foreseeability’ of harm by a non-State actor, as explained in Section 3.3.1. For example, the murder of the child occurred during an unsupervised visit of the abusive partner, which the domestic court had granted following its assumption that it was always better for children to have contact with their father, ignoring (or deeming irrelevant) the abusive history of this particular father. In this case, as well as previous cases, the CteeEDAW went on to provide specific action that should be taken by the respondent State in order to fulfil its positive obligations under the Convention in the future (as well as measures to provide the author with redress). For example, the Committee requested that Spain ‘[s]trengthen application of the legal framework to ensure that the competent authorities exercise due diligence’ and ‘[c]onduct an exhaustive and impartial investigation’. This emphasis on investigations was also apparent in A.T. v. Hungary, with CteeEDAW requesting Hungary to ‘[i]nvestigate promptly, thoroughly, impartially and seriously all allegations of domestic violence and bring the offenders to justice’.245 As the HRCtee and CteeESCR, the CteeEDAW appears to focus mostly on the State’s procedural obligations under the obligation to protect, again limiting the extent to which its approach can really be considered to be one of ‘indirect horizontal effect’.
Outside of the context of domestic violence, the CteeEDAW has also applied a duty of due diligence. As the HRCtee and the CteeESCR, the CteeEDAW has used the duty of due diligence to attribute the actions of private providers of public services to the State and find a violation of rights within cedaw. In Alyne da Silva Pimentel Teixeira (deceased) v. Brazil, for example, a complaint was brought on behalf of a woman who had died in a private health institution during her pregnancy. The Brazilian State tried to argue that because the health institution was a private actor, its
professional negligence, inadequate infrastructure and lack of professional preparedness’ could not be attributed to the State.246 However, Brazil did acknowledge ‘shortcomings in the system used to contract private health services and, by extension, the inspection and control thereof’. The Committee used this to find that Brazil had failed to fulfil its duty of due diligence in line with Article 2 cedaw.247 The CteeEDAW also reiterated its position taken in previous decisions and general recommendations that ‘the State is directly responsible for the action of private institutions when it outsources its medical services’ and ‘always maintains the duty to regulate and monitor private health-care institutions’.248 In other words, Brazil could not delegate its legal responsibility for the protection of human rights through the act of delegating certain public services (i.e. healthcare) to private institutions. Again, we see here reference to an obligation to regulate privatised services.
4.3.3 Critical Remarks on the Jurisprudence of the Committee on the Elimination of Discrimination against Women
We can see from the above examples that the approach of the CteeEDAW in its views on individual complaints mirrors its approach taken in General Recommendations. Perhaps because of the wording of Article 2 of the Convention, the CteeEDAW has not wavered in States’ positive obligations in order to lend more protection to women, particularly from situations of domestic violence. Further, it has repeatedly stressed the appropriate measures that State Parties must take to observe its duty of due diligence for what concerns ‘purely’ private relationships (e.g. between two or more individuals) and ‘quasi-public’ relationships (e.g. between an individual and a private public service provider). As did the CteeESCR, the CteeEDAW also examined, to a limited extent, more direct obligations of non-State actors. Its comments on non-State armed groups in General Recommendation No. 30 show that in some situations (primarily, when a group has effective control over some territory), non-State armed groups may have direct responsibilities to at least respect human rights both during and after armed conflicts.
4.4 Committee against Torture
The fact that the un Committee against Torture (CteeAT) has given any attention to the application of the un Convention against Torture (cat) between non-State actors is extremely interesting in light of the definition of torture provided in the cat.249 Article 1 cat appears to explicitly rule out the possibility that torture could be committed by a private person without any involvement of the State. The definition includes as a criterion that an act be committed ‘at the instigation of or with the consent or acquiescence of a public official or other person acting in an official capacity’ for it to be torture. The same criterion applies in relation to acts of cruel, inhuman or degrading treatment, which is prohibited by Article 16 cat.250
From this wording, we could assume that the direct horizontal effect of torture can be prima facie ruled out (as suggested by Alice Edwards).251 However, it does not seem to preclude the possibility of some forms of indirect horizontal effect, relying on State’s positive obligations and attributing the actions of a non-State actor to the State. In the context of Articles 1 and 16 cat this would most obviously be through a private actor acting under the instruction of, with the consent of, or at the acquiescence of a public actor. As Article 1 is prohibitive and therefore inherently negative in nature, it may be argued that Article 1 cat also does not, prima facie, give rise to a positive State obligation to protect which would allow the CteeAT to invoke indirect horizontal effects in its views on individual communications.252 Additionally, as a civil and political, rather than an economic, social or cultural right, some may try to argue that the tripartite typology of human rights does not apply to the prohibition of torture, as it only concerned negative obligations. A related argument was made above by the us regarding positive obligations under the prohibition of torture in the iccpr (Section 4.1.1). The likely success of such an argument is now very low. As Magdalena Sepùlveda notes, ‘[t]oday it is beyond doubt that civil and political rights instruments […] also impose “positive obligations” on the part of States Parties which are often not explicitly contained in the text’253 – ‘attempting to classify every right as either flatly negative or positive, is an “artificial, simplistic and arid exercise”’.254
Indeed, the Committee has been willing to openly apply Article 1 in such a way as to allow indirect horizontal effect of the prohibition using the State’s responsibility to protect individuals from harmful actions by non-state actors. The Committee has also allowed for indirect horizontal effect by treating some non-State actors as State actors by virtue of them carrying out particular ‘governmental’ functions. As was the case with the treaty monitoring bodies examined above, the following examination will look firstly at the Committee’s general comments and then the jurisprudence arising from its views on individual complaints, to determine how the CteeAT understands the obligations in the cat vis-à-vis non-State actors.
4.4.1 General Comments of the Committee against Torture
The attitude of the Committee against Torture (CteeAT) towards horizontal effect initially appeared quite limited. In its first general comment adopted in 1998, the CteeAT emphasised the importance of the public official criterion, and did not mention private actors.255 In its second comment, however, the Committee paid more attention to the acts of non-State actors, in particular the due diligence obligations of States. The CteeAT held the due diligence obligation to apply in instances where States ‘know or have reasonable grounds for believing’ that non-State actors are committing torture or acts of ill-treatment.256 The rationale of the Committee here is that because of the wording of Article 1, States must be under a positive obligation to protect individuals from torturous acts by non-State actors. The ‘consent or acquiescence’ of the State essentially means that if a State fails to take positive measures to protect individuals, it is implicitly permitting, or acquiescing to the harm by the non-State actor.257 This interpretation has vastly changed the landscape of the prohibition of torture within international law, and is also applied by other human rights monitoring bodies (in particular regional human rights courts).258
4.4.2 Views of the Committee against Torture
The interpretation and its application by the CteeAT in its own views on individual complaints undoubtedly affords broader protection to individuals. However, the CteeAT has not actually applied horizontal effect to the full potential allowed by Article 1. Certainly, it is positive that the Committee has addressed the duty of due diligence explicitly, but its interpretation of State ‘consent or acquiescence’ is actually relatively limited.259 For example, in the case of S.V. et al. v. Canada,260 the Committee dealt with complaints relating to torture by the Liberation Tigers of Tamil Elam (ltte) in Sri Lanka. The author feared that if he were to return to Sri Lanka, he would be subjected to torture either by the ltte, or by the State (whom he argued had caused him brain damage on an earlier occasion following allegations by the State that he was actually a member of ltte). The Committee clearly held any consideration of torture by the ltte with no consent or acquiescence of the State to fall outside of the scope of the cat.261 However, it did not provide indications as to what actions they believed could constitute consent or acquiescence. Nevertheless, Edwards has deducted from this case, along with that of G.R.B v. Sweden (below, also dealing with torture by non-State actors) that the CteeAT sees ‘consent and acquiescence’ as amounting to either some knowledge on behalf of the State of the actions by the non-State actor, the State being in ‘general agreement’ with them, or purposefully refusing to act against them.262 This seems to be consistent with the application of due diligence by the previous bodies discussed, although narrower than the obligation to protect more generally. It is unclear whether the knowledge on behalf of the State would also require State investigations into the actions of non-State actors (as is the case with due diligence under other bodies), or refers solely to information provided to the State by the non-State actors themselves.
One major issue that surfaces repeatedly in individual complaints relating to torture is that of women being raped by private individuals. Interestingly, the cases in relation to which these comments have been made do not usually include any involvement of public officials, making them noteworthy studies of whether and how horizontal effect has been applied in practice. Furthermore, these cases demonstrate the limitations in the way that the CteeAT looks upon due diligence.
Since 1986, un special rapporteurs have defined rape as constituting torture.263 The un has also noted that in the last few decades, ‘significant efforts’ have been made to ‘redefine the meaning of human rights’264 in order to answer the feminist critique of those such as Hilary Charlesworth, Christine Chinkin and Shelley Wright that the prohibition of torture is inherently biased against women.265 This is largely due to the fact that the majority of rapes of women happen within private relationships (being, for example, aspects of situations of domestic violence).266 In consequence, several bodies have now held that rape may constitute torture ‘per se’,267 and will always meet the ‘minimum threshold’ required to engage provisions prohibiting torture.268
In the case of G.R.B v. Sweden,269 the author of the complaint feared being subjected to torture by both State and non-State officials upon her return to Peru after having three applications for asylum in Sweden rejected. The fear stemmed from previous instances in which the author was kidnapped and raped by a terrorist organisation in Peru, and the fact that she and her family had been politically active against the Peruvian State in the past, with her father already having been tortured by State authorities. The CteeAT held that the fact that an individual ‘might risk pain or suffering inflicted by a non-governmental entity, without the consent or acquiescence of the Government, falls outside the scope of […] the Convention’.270 The risk of torture by the Peruvian State itself was ruled out by the Committee, mainly because the author had not been politically active for 13 years, and that she had never been subjected to torture by the Peruvian authorities in the past.271 The author’s complaint was therefore unsuccessful. Unfortunately, the CteeAT did not assess Peru’s investigations into the occurrence of rape that the victim suffered at the hands of the non-State actor, having failed to detail what standards the Peruvian State would have to fulfil to be acting with due diligence. This could imply that the CteeAT does not understand the cat as meaning that States Parties are obligated to perform investigations into harm caused by non-State actors.272 Indeed, on a literal reading of the Convention and combined with the State actor requirement in Articles 1 and 16, this is understandable, particularly in light of Article 12 cat. Article 12 places a duty on States to ensure ‘prompt and impartial’ investigations by ‘competent authorities’ when an act or torture can reasonably be believed to have been committed within their jurisdiction. Taking this on face value, it does not appear to place any obligations on States to investigate acts conducted by private individuals. Reading Article 12 consistently with CteeAT’s interpretation of Article 1, however, the obligation to investigate would equally apply where the act of torture reasonably believed to have been committed would be by a private actor, as well as a public official. Indeed, as we have already seen, other human rights bodies have used the duty of due diligence to impose investigative obligations on States without explicit mention of private actors in the relevant treaty provisions. This is also true specifically in relation to the prohibition of torture, which the European Court of Human Rights understands as requiring States to investigate private acts of torture.273
One case in which the relevant State Party (Serbia and Montenegro) was found responsible for not protecting an individual from inhuman or degrading treatment or punishment by private actors was Hajrizi Dzemajl et al. v. Serbia and Montenegro.274 The complaint concerned a group of Romani individuals who, following an incident in which a non-Romani girl had been raped by other Romani individuals, were subject to severe violence by a large group of non-Romani citizens. The group of individuals made threats to the Romani, who were advised by police to leave their homes. After the group began setting the Romani individuals’ homes on fire and hazing them to the ground, the complainants managed to escape, although some of them had been hiding in the basement of their houses when the violence began. The State authorities had been informed of the action being taken, but ‘did not take any appropriate steps in order to protect the complainants, thus implying “acquiescence”’ of the State.275 In its finding of a violation of Article 16 cat, the CteeAT explicitly stated that ‘[a]lthough the acts referred to by the complainants were not committed by public officials themselves, the Committee considers that they were committed with their acquiescence’.276 The CteeAT did not go so far as to explain in general what suffices for actions of non-State actors to be attributed to the State due to the latter’s ‘consent or acquiescence’ – does it require that the State actually knew about the risk to the complainants, as in this case, or is it sufficient that the State should have known, as is the standard followed by the European Court of Human rights and the Inter-American Court of Human Rights?277
A second approach towards horizontal effect can be found in the CteeAT’s jurisprudence, this time treating some non-State actors as public actors according to the functions that they are fulfilling within a State. This approach was famously taken in the case of Sadiq Shek Elmi v. Australia.278 The claimant in this case was a Somali national at risk of being transferred to Somalia by the Australian Government. He argued that by transferring him to Somalia, Australia would be breaching the rule of non-refoulement enshrined in Article 3 cat.279 This rule stipulates that individuals may not be extradited or transferred to a country where they would be at ‘real risk’ of being subjected to torture. A pivotal factor in this case was that the actor at whose hands the claimant feared being subjected to torture upon his return to Somalia was the Somali Hawiye clan, a non-State armed group.280 The nature of cases of non-refoulement is rather special, in that to find a violation of torture by the responding State does not require this State to actually commit acts of torture. Simply putting the claimant in a situation where he is at risk of being subject to torture by transferring him to a second State is enough for the respondent State to violate the prohibition. If torture were to be committed within the destination State, that State could then also be held responsible (assuming that State is also party to the cat). The consequence for horizontal effect in cases like Elmi is not a finding that a non-State actor did/not violate the prohibition of torture itself, but can result in a finding that the acts of a non-State actor could amount to torture, were they to be committed against the claimant after their arrival in the destination State.
The Elmi case is very important here because, despite not involving a claim of torture directly against a non-State actor, the CteeAT acknowledged that actions by non-State armed groups could amount to torture, fulfilling the requirement of State involvement. This was because the non-State armed group in question was ‘exercising certain prerogatives that [were] comparable to those normally exercised by legitimate governments’.281 This seems quite a progressive move by the Committee. Indeed, treating a non-State armed group as a State actor goes further than any of the approaches seen thus far, but the special circumstances of the case demonstrate that in order to protect the applicant, the CteeAT did not have much choice in its approach. At the time, there were several clans in Somalia vying for State control, each having ‘prescribed its own laws and law enforcement mechanisms and [having] provided their own education, health and taxation system’.282 The ultimate lack of control by the State itself over these established groups seems to have been instrumental in the CteeAT’s decision. In contrast, even the abovementioned cases that considered the public function of a non-State actor to be decisive in attributing their actions to the State relied (at least partially) on the fact that the functions had been delegated to the non-State actor in order to find a human rights violation. However, in the Elmi case, it was not a public function that the non-State actor was fulfilling on behalf of the State (as in the cases concerning privatised public services), but a wide range of public activities that the group had taken upon itself in the context of a failed State. The lack of a stable State authority meant that indirect horizontal effect through Somalia’s obligation to protect human rights was not possible; the only avenue open to the CteeAT to prevent the transferral was to find a way to categorise the non-State actor as a State actor for the purposes of the case. Had the Committee disregarded the case based on the non-State identity of the potential torturing entity, it would have resulted in a gap in human rights protection solely because the actors perpetrating the violations could not be held directly to the norms under international human rights treaties.
It is certainly an interesting outcome when compared to other cases heard by the Committee dealing with potential torturous acts being committed by other non-State armed groups, such as S.V. et al. v. Canada, in which the Committee took the approach of upholding the State’s positive obligation to protect human rights. In that case, however, the relevant non-State armed group (the ltte) did not have effective control over an area of territory within Sri Lanka, which had not itself failed as a State. The extreme rarity of the circumstances of a failed State in Elmi has indeed prevented the CteeAT from applying this kind of indirect horizontal effect again. Even in a subsequent case regarding extradition to Somalia of an individual who believed himself to be at a real risk of torture by a non-State armed group, the CteeAT distinguished the case on the grounds that Somalia had by then regained a State authority acting as central Government in the international community.283 The fact that the group in question still controlled a portion of Somalian territory was not considered enough to treat it as a State actor. Curiously, though, in a latter case concerning the ltte, the CteeAT did consider that in cases where ‘the non-governmental entity occupies and exercises governmental authority over the territory to which the complainant would be returned’, it was not necessary to ascertain State consent or acquiescence in future acts of torture by that actor.284 The distinguishing factor for the Committee here seems to be whether the individual will be returned to the territory controlled by the ltte. Further evidence that the CteeAT is not willing to broaden the application of Elmi to non-State actors more generally can be found in the case of V.X.N. and H.N. v. Sweden. The Committee explicitly stated that whether non-refoulement extends to a ‘risk [of] pain or suffering inflicted by a private person, without the consent or acquiescence of the State, falls outside the scope of article 3’.285
It is clear, then, that while the CteeAT will take a case-by-case approach, in general it is not actually willing to apply indirect horizontal effect beyond invoking States’ consent or acquiescence.
4.4.3 Critical Remarks on the Jurisprudence of the Committee against Torture
The CteeAT has engaged to a considerable degree with the horizontal effect of human rights within the cat. On the one hand, the definition of torture it has adopted through general comments and views on individual communications is broad, in that the CteeAT has understood it to include a positive obligation on States to protect individuals from torture by non-State actors. On the other hand, the Committee’s application of the requirement that torture or inhuman or degrading treatment or punishment be done at least with the ‘consent and acquiescence’ of a State actor has so far been restricted to cases in which the State had concrete knowledge of a risk of an individual being subject to this treatment. While this appears stricter than the ‘foreseeability’ standard applied by the CteeEDAW, it could simply be due to the facts of the cases brought before the Committee.
In the case of Sadiq Shek Elmi v. Australia, the CteeAT has certainly shown that in extreme circumstances it is willing to go beyond an application of indirect horizontal effect through States’ obligation to protect, to treat certain non-State actors as State actors to prevent torture. By no means, though, can this be considered a general rule for the Committee. The special circumstances of the case must be borne in mind. Specifically, the context of the failed State, the governmental authority being exercised by the non-State actor and the fact that it was not actually the non-State actor, but rather the State wishing to extradite the applicant, who was being held responsible for a potential violation of human rights. These circumstances considerably temper the possible significance of the decision.
4.5 Committee on the Elimination of Racial Discrimination
The Convention on the Elimination of Racial Discrimination (cerd) includes a potential exception to the vertical nature of the human rights obligations contained within it.286 This can be found in Article 5(e) of the Convention, which has been the subject of a brief General Recommendation (No. 20) by the Committee on the Elimination of Racial Discrimination (CteeERD). It is also interesting that reference to the positive obligations of State Parties to the Convention can be found in Article 4. This provision requires States to take ‘immediate and positive measures designed to eradicate all incitement to, or acts of’ racial discrimination. Significantly, reference is also made here to the possible exception of vertical obligations found in the Universal Declaration of Human Rights (udhr),287 the principles embodied within this instrument having to be taken into account by States in the taking of the aforementioned measures.288 Although no mention of a specific provision from the udhr is given in Article 4, it can be inferred that the principles mentioned are those found in Article 29(2)289 – an inference made explicit by the CteeERD in its General Recommendation No. 15 on Article 4.290 The CteeERD’s recommendations and jurisprudence dealing with potential horizontal effect of the Convention will now be assessed.
4.5.1 General Recommendations of the Committee on the Elimination of Racial Discrimination
As stated, Article 5 cerd was briefly discussed by the CteeERD in General Recommendation No. 20, with explicit reference to non-State actors. The recommendation seems to suggest that private actors themselves could be (partially) responsible for protecting the rights contained in Article 5 ‘and any similar rights’. The CteeERD states that ‘protection may be achieved in different ways, be it by the use of public institutions or through the activities of private institutions’.291 While this may emphasise the private actor’s potential role in protecting human rights, the Committee does not go so far as to say that private institutions should actually be responsible for ensuring protection. Indeed, the CteeERD goes on to identify the obligation to ensure human rights protection as a State obligation, which covers the actions of private actors: where ‘private institutions influence the exercise of rights … the State party must ensure that the result has neither the purpose nor the effect of creating or perpetuating racial discrimination.’292 In another recommendation, the CteeERD has reiterated the fact that States’ obligation to protect human rights refers to the actions of non-State actors as well as public actors. In its General Recommendation No. 30 on the rights of non-citizens, the CteeERD identified such actors as including ‘politicians, officials, educators and the media, on the Internet and other electronic communications networks and in society at large’,293 in relation to whom States must (again under their obligation to protect) take ‘resolute action’ to protect individuals from racial discrimination.294
Further inference to States’ positive obligation to protect human rights from non-State actors is found in General Recommendation No. 15, dedicated to Article 4 of the Convention.295 The recommendation does not elaborate much on the meaning of the ‘positive measures’ mentioned in the article. Interestingly, General Recommendation No. 32 did elaborate on what kinds of measures States are obliged to take in this context, including, inter alia, the adoption of ‘legislative, executive, administrative, budgetary and regulatory instruments’.296 The underlying goal of Article 4 must be to protect individuals from propaganda involving racial discrimination by all actors (reflected in the requirement that States ‘declare an offence punishable by law all dissemination of ideas based on racial superiority’).297 Nevertheless, there was no reference either in the Convention itself, or in the General Recommendation, to the obligation to protect, or to the due diligence duty of States. Indeed, the Recommendations only mention an investigative requirement of States in relation to the ‘national law and its implementation’298 – inherently public matters. This seemingly rendered investigation into private acts outside of the scope of the positive measures to be taken unless this can be read into the implementation aspect. However, in a later General Recommendation also dealing with Article 4 (No. 35), the CteeERD suggested that obliging States to conduct such investigations could be read into the obligation that States make certain conduct ‘punishable by law’.299 This is because simply making certain conduct constitute a criminal offence is not sufficient without its effective implementation.300 This implementation, the CteeERD states, requires investigations of potential offences to be carried out, leading to prosecution when appropriate.301 Taken together with the wording of Article 4, the provision could be seen as reflecting (or being part of) a duty of due diligence to be fulfilled by State Parties.
This assertion is supported by a comparable obligation to that in Article 4, found in Article 3 cerd. This provision requires State Parties to ‘prevent, prohibit and eradicate all practices of racial segregation’. In its General Recommendation No. 19, the CteeERD has read this as placing positive obligations on States, emphasising the possibility that partial segregation arise as an ‘unintended by-product of private persons’.302 The consequence of this – that racial segregation can occur ‘without any initiative or direct involvement by the public authorities’303 – prompted the CteeERD to recommend that States Parties monitor the kinds of situations that could lead to racial segregation.304 Again, despite no allusion to due diligence by the Convention or the CteeERD, this reflects very similar principles to those referred to by the other treaty bodies in their discussions of due diligence. It is curious that the CteeERD has chosen not to follow suit in using the same terminology as other treaty bodies, but this also shows that assumptions cannot necessarily be made as to the effect of bodies’ decisions from the fact that they do not refer directly (or even indirectly) to the horizontal effect of the rights within their jurisdiction. To avoid speculating, it suffices to say that the CteeERD has in effect applied indirect horizontal effect by fleshing out States’ positive obligations and making concrete recommendations as to how they could be fulfilled. These recommendations have, as they did with the other bodies discussed above, been restricted mostly to procedural obligations of States rather than laying down concrete standards against which to hold non-State actors (other than the general terms used). The question therefore remains whether the State should impose obligations or responsibilities at the national level holding non-State actors to the same standards as the State in terms of the substance/content of human rights.
4.5.2 Views of the Committee on the Elimination of Racial Discrimination
Pursuant to Article 14 of the Convention on the Elimination of All Forms of Racial Discrimination the CteeERD can hear individual complaints from individuals regarding alleged violations of the Convention by States that have adopted a declaration accepting its competence to do so. The CteeERD has dealt with horizontal effect on quite a few occasions. This is to be expected, since racial discrimination is something that often happens in private relationships rather than by the State itself. Several examples will be focused on in the following discussion.
The CteeERD made several references to States’ obligation to protect vis-à-vis non-State actors in the case of L.K. v. The Netherlands.305 The applicant, a Moroccan citizen residing in the Netherlands was subject to verbal abuse and threats from a group of Dutch citizens who refused to accept him as a resident on their street, going so far as to sign a petition refusing his acceptance. The applicant complained to the police, who he alleged failed to conduct a thorough or complete investigation into the events. In its opinion, the CteeERD agreed with the applicant that the words and actions of the Dutch citizens constituted ‘incitement to racial discrimination and to acts of violence against persons of another colour or ethnic origin’ under Article 4 cerd, which the State had violated through its lack of an adequate investigation.306 In particular, the CteeERD stated that ‘[w]hen threats of racial violence are made, especially when they are made in public and by a group, it is incumbent upon the State to investigate with due diligence and expedition.’307 The explicit references to non-State actors and due diligence in the context of a State’s obligation to protect human rights suggest an application of indirect horizontal effect by the CteeERD based on procedural obligations of the Dutch State (i.e. to investigate and punish non-State actors accused of racial discrimination). The reasoning of the CteeERD in this case mirrors that in its General Recommendations on Article 4, discussed above.
The obligation of States to conduct a thorough investigation into instances of racial discrimination by non-State actors was also discussed in the case of Habassi v. Denmark.308 The author of the complaint, a Tunisian man with permanent residence in Denmark, had applied for a loan from a private bank, which required him (‘motivated by the need to ensure that the loan was repaid’309) to provide evidence of his Danish nationality before approving the loan (after previously stating that they would accept loan requests from individuals with permanent residence). After his loan application was rejected, with the help of an antidiscrimination organisation the applicant complained to the Danish police, who did not pursue an investigation. The State’s argument was that there was insufficient evidence that an unlawful act had taken place. The CteeERD, however, agreed with the applicant that a human rights violation had occurred. It based its decision on the need for States to investigate whether or not racial discrimination had occurred (whether directly or indirectly), which requires a ‘proper investigation’ into the facts.310 Here, the basis upon which the bank required an individual to prove Danish nationality before granting a loan was in question – although nationality is not an issue of racial discrimination as such, it may have an unintended, indirect effect of racial discrimination.311 Since the police had accepted the bank’s reasons on face value and had not conducted such an investigation, it was found to be in violation of Article 6 cerd (the right to an effective remedy) in connection with Article 2(d) (the definition of racial discrimination). The case therefore goes quite far in requiring States to protect individuals from non-State actors (albeit focusing on procedural requirements). A very similar reasoning and findings were applied by the CteeERD in its opinion on Kashif Ahmad v. Denmark.312 In this case, the applicant alleged that he had been racially discriminated against by a teacher and a headmaster of a school, on school property. He complained that the Danish authorities refused to prosecute the individuals or conduct a thorough investigation into what happened. After a cursory investigation into the facts, the authorities stated that the offensive statements made to the applicant fell outside of the scope of Danish criminal law, and that the applicant could only seek redress through civil law – a finding that the applicant was unable to appeal.313 The CteeERD opined that had a thorough investigation been undertaken, the authorities would have been able to establish whether the applicant was subject to racial discrimination within the meaning of Article 2(1)(d) cerd.314 The CteeERD ultimately found that because of the lack of investigation and ability of the applicant to find out whether his rights had been violated at the national level, the State had failed to effectively protect him from racial discrimination.315
As well as an obligation to investigate instances of racial discrimination by private actors (in a timely manner), the CteeERD has made recommendations that States ‘complete’ their legislation protecting individuals from racial discrimination by private actors. In the case of Lacko v. Slovakia, for example, the Committee recommended such legislative action to ‘guarantee the right of access to public places in conformity with article 5 (f) of the Convention and to sanction the refusal of access to such places for reason of racial discrimination.’316 The case had involved a Romani individual being refused service in a restaurant at a railway station on the basis that he was Romani. The State investigated the situation and prosecuted the culprit, which led the CteeERD to find no violation of the Convention. However, it did not refrain from giving this extra counsel to Slovakia to strengthen its performance of the duty of due diligence.317
4.5.3 Critical Remarks on the Jurisprudence of the Committee on the Elimination of Racial Discrimination
In light of the specific references made within the cerd to various non-State actors, the CteeERD has not actually gone very far in applying the Convention horizontally. This is odd in light of the mention of private actors in the Convention itself – the clear obligation to protect individuals from racial discrimination in the private, as well as in the public sphere could seem to render further consideration of private actors even more necessary. However, it could also be said that there are not as many substantive human rights contained in the cerd that could be upheld vis-à-vis non-State actors, other than those in Articles 4 and 5. Most of the cases that involved a non-State actor found violations of the State of these two provisions, as well as Article 6 which provides the right to effective remedy. Taken together, the clear obligation to conduct thorough investigations into alleged incidents of racial discrimination by private parties, and the obligation to have a legal framework in place to punish such parties when found responsible, strongly reflect at least two elements generally considered to constitute a duty of due diligence (together with the duty to prevent interference by non-State actors). Indeed, the CteeERD has made explicit reference to this duty, although not on a regular basis.
5 Reflections on the Horizontal Effect of International Human Rights in International Jurisprudence
The above discussions show widespread and varied acknowledgements of the considerable role that non-State actors have to play in the enjoyment of human rights. While the international legal framework does not allow for non-State actors to be directly burdened with international legal obligations, the un human rights treaty monitoring bodies have on many occasions upheld the standards within international human rights law against the actions of non-State actors. This has been achieved by the Committees mainly in two identifiable ways, which will be discussed below. The first, and most common approach, is through the invocation of States’ positive obligation to protect human rights from third parties. As repeatedly observed in the analysis, the vast majority of recommendations made by the monitoring bodies concern procedural obligations and explanations of the kind of action that States must take (legal, regulatory, administrative, judicial, etc.) rather than mentioning whether the standards applicable to non-State actors should be assumed to be the same as those applied to States (which are themselves further developed through other general comments and jurisprudence of the bodies). In this sense, it could be said that the analysis is more useful in terms of laying down the content of States’ obligation to protect human rights vis-à-vis different actors and in different situational contexts. The focus could show either a reluctance on behalf of the bodies to address standards (albeit national ones) for non-State actors more directly, perhaps an opinion that a distinction between State and non-State standards is irrelevant, or that it should simply be assumed that the concrete standards for non-State actors at the national level are the same substantive standards for States at the international level.
5.1 The Types of Indirect Horizontal Effect Identified
The requirements of the obligation to protect seem to differ somewhat depending on what kind of non-State actor has interfered with the enjoyment of human rights; the nature of the actor and the relationship they have with the victim of human rights violations appears to be a factor in the way in which monitoring bodies apply indirect horizontal effect. For the remainder of this article, this first approach will be named ‘pure indirect horizontal effect’, to denote what is quite a clear-cut method of attributing the harmful effects of non-State actors to States. The second approach, which is very tentatively taken by the monitoring bodies, also appears to depend on the kind of actor (and the kind of actions being taken by that actor). For the purposes of this article, the second approach will be labelled ‘categorical indirect horizontal effect’. Essentially, it refers to cases in which human rights laws are being applied against what is technically a private actor who is being treated, for the particular instance at hand, as a special kind of public actor’, or a ‘quasi-public actor’. In other words, the private actor is being re-categorised for the purposes of ensuring broader protection of international human rights. The following sections critically discuss the two approaches and the instances in/extent to which they are applied.
5.1.1 ‘Pure’ Indirect Horizontal Effect
As explained, pure indirect horizontal effect holds States responsible for the actions of non-State actors by invoking the State’s obligation to protect human rights. There is a limited degree to which this approach can actually be said to be a form of horizontal effect, given its focus on State obligations. However, when States follow the advice of the monitoring bodies under this approach, it creates an indirect obligation for non-State actors (or imposes a certain standard of behaviour) at the international level, having first resulted in a direct standard of behaviour being imposed on the non-State actor at the national level (whether through legal, administrative or other means).
The analysis above shows that ‘pure’ indirect horizontal effect is consistently applied by the different monitoring bodies, but is sometimes expressed in different manners. One of the main tenets of the obligation, applied by all of the bodies, is an obligation to take legal, administrative or other appropriate measures to protect individuals from non-State actors – to adopt an appropriate framework within the State capable of preventing, investigating and punishing instances of non-State actor interference with human rights. In other words, the monitoring bodies uphold the standards of due diligence. This has not been made explicit by each of the bodies every time that they appear to apply the obligation (especially for the CteeESCR) although the actual standards upheld do appear to be consistent.
The obligation of due diligence has often been applied when the non-State actor interfering with human rights operates in the purely private sphere. This is particularly true of those monitoring bodies dealing with issues of discrimination (i.e. the CteeEDAW and CteeERD), which very commonly occurs in purely private relationships between two individuals. As explained in Section 3.3.1, the duty is one of conduct, rather than result. This has led the monitoring bodies to take an approach of listing possible action to be taken in their general comments, although it is not possible to create an exhaustive checklist of protective measures to be taken to combat every circumstance of human rights interference. The general comments have by now provided quite an extensive array of measures that States should be taking in different subject-areas (e.g. regarding employment, or the provision of services) or in relation to specific rights within their respective conventions.
As well as the duty of due diligence (or perhaps even part of it, as the bodies’ practice is not clear on this point), the obligation to protect includes an obligation to regulate private actors. This has been upheld by each of the monitoring bodies except for the CteeAT, and is particularly elucidated in the context of privatisation or the delegation of ‘public’ tasks to non-State actors, as well as in the ‘quasi-public sphere’ such as employment.
What is less clear about the interpretation and application of human rights by the monitoring bodies is to what extent the bodies actually engage with the concept of attribution, at least from the perspective of State responsibility within international law. Although some of the bodies have mentioned that non-State actors’ conduct can be attributed to the State, with the exception of the CteeESCR in its general comment on business and human rights, none of the bodies have explicitly engaged with the International Law Commission’s dasr.318 This begs the question whether the monitoring bodies view ‘attribution’ as simply another word for describing the way in which the obligation to protect works in practice, thus enabling the ‘attribution’ of non-State conduct to States in the situations where the obligation to protect has not been fulfilled. If so, the language used could create confusion amongst international law scholars, aggravated by the explicit reference to dasr in the CteeESCR’s General Comment No. 24.
5.1.2 Categorical Indirect Horizontal Effect
Categorical indirect horizontal effect is the second type that can be identified. This refers to cases in which human rights laws are being applied against what is technically a private actor who is being treated, for the particular instance at hand, as a special kind of public actor’, or a ‘quasi-public actor’. At first sight, this may appear similar to direct horizontal effect, in that it is a private actor that would be ‘violating’ a human right. However, the fact that the responsible non-State actor is deemed to be a public actor in the particular circumstances of the case would allow a case to be brought against the (re-categorised) non-State actor. For the international level, this approach does remain more theoretical than at the national level (e.g. in the United Kingdom),319 as it has still not allowed a full case to be brought against a non-State actor, given the constraints of the human rights treaty framework. Nonetheless, it has been dealt with and taken to a limited extent by the CteeAT.
The CteeAT applied ‘categorical’ indirect horizontal effect in relation to non-State armed groups. Given the nature of the groups and the fact that they are usually engaged in an armed conflict with the State in a particular territory, the public functions that they carry out are not delegated from a ‘pure’ public authority, unlike in some of the other cases discussed above. Instead, the groups take up the responsibility for carrying out the public functions of their own accord. Crucially, in the case of Sadiq Shek Elmi v. Australia where the CteeAT seems to apply this approach, the State was no longer a relevant player – Somalia had failed as a State, which had led to quasi-public authorities taking over traditionally State functions and exercising effective control over ‘State’ territories. The case is admittedly of limited value, because of the particular situation of non-refoulement and the case’s weak precedential value, as explained above. Nonetheless, it opens the door for future general comments and case law to apply a similar approach. This was arguably done by the CteeEDAW in its General Recommendation No. 30.
So far, the limited application of this approach (in terms of number of cases and in relation to which actors it has been applied) makes it difficult to see where the threshold lies for re-categorising a non-State actor as a ‘public’ actor and holding them to international standards (at the international level).
5.2 The Treatment of Different Actors
First, the bodies, both in their general comments and views on individual communications, have upheld States’ obligation to protect individuals from harmful acts by other individuals. This would include, for example, other family members (e.g. CteeEDAW General Recommendation No. 19; A.T. v. Hungary), neighbours (e.g. L.K. v. The Netherlands), and employers (e.g. CteeESCR General Comment Nos. 5 and 14). In cases where such actors were involved, the focus of the treaty bodies was mostly on due diligence obligations of the State. An emphasis has certainly been placed on the obligations to investigate interference with human rights by non-State actors, as well as to prevent and punish the actions, especially where the State is aware of a risk to the individual. In one sense, the cases show a stronger application of horizontal effect, as the relationships between the individuals is often exclusively in the private sphere (e.g. between spouses). However, the actions of the non-State actors themselves are not the subject of much discussion by the Committees, except to the extent that they show that a particular right is engaged. Rather, the focus is (as it should be, according to the international human rights law framework) on the action or inaction of States either to prevent the harm occurring, or in reaction to the harm that occurred, which allows the private actor’s conduct to be attributed to the State. Interestingly, because of the wording of Article 1 cat requiring at least the ‘consent or acquiescence’ of the State for an act to fall within the scope of the Convention against Torture, the CteeAT has seemed to take a strict stance as to when acts of torture by private actors can lead to a violation of the Convention. It seems to be that only when the State knew of a risk of torture would the actions be imputable to the State. This contrasts with the approach suggested by CteeEDAW in Angela González Carreño v. Spain and the approach of other bodies with jurisdiction over the prohibition of torture, which are comparatively broader.
Second, the treaty bodies have applied States’ positive obligation to protect human rights from private businesses (CteeESCR General Comment Nos. 5 and 12) and entities such as banks (e.g. Habassi v. Denmark) and insurance boards (e.g. B.d.B. v. The Netherlands). The horizontal effect here is very similar to that found in the jurisprudence concerning relationships between individuals.
Third, the treaty bodies have applied indirect horizontal effect in relation to private companies or institutions that are carrying out public functions that have been delegated to them by the State. This category of actor has included, for example, privately run prisons (e.g. Cabal and Pasini v. Australia) and private healthcare institutions (e.g. Alyne da Silva Pimentel Teixeira (deceased) v. Brazil). In their discussions of these actors, the treaty bodies have focused on attributing the actions of the non-State actors to the State because of the public nature of the functions that the actors are carrying out and because they were delegated these activities by the State. Here, the main argument appears to be that States cannot give up their own international responsibility by delegating certain functions to non-State actors. The treaty bodies have upheld (within the obligation to protect) an obligation to regulate and supervise the privatised companies, using a failure to do so as the basis for a State violation of a particular right. As well as the obligation to regulate, the treaty bodies have also applied the duty of due diligence to these actors (indeed, it appears to be applied in some form in relation to every kind of actor).
Fourthly, cases of indirect horizontal effect have also occurred where the actor interfering with human rights is a non-State armed group. Here, the Committee against Torture has taken different approaches to horizontal effect. In the case of S.V. et al. v. Canada, the State’s obligation to protect was upheld. In Sadiq Shek Elmi v. Australia, though, the approach taken was to treat the non-State actor as a State actor. This seems only to apply when the group as effective control over the area of land to which an individual is going to be extradited, and/or when there is no effective central State authority within the receiving State. In these cases, it does not appear that there actually has to be any attribution to the State. The CteeEDAW has also dealt with these actors in General Recommendation No. 30, stating non-State armed groups’ obligation to respect human rights during armed conflict (also placing emphasis on those groups with effective control over an area of territory).
Fifthly, suggestions have been made, at least by the CteeESCR (General Comment Nos. 14 and 15), that international organisations may have human rights obligations. This would in theory be an example of direct horizontal effect, but there was no legal basis provided by the CteeESCR and there are no examples of treaty bodies applying human rights treaties to international organisations in individual communications. The obligations could therefore be read as being moral, but not yet legal, in nature.
It is possible to conclude that the vast majority of the jurisprudence of the un human rights treaty bodies involves a connection being made between the State and the private actor concerned. The basis on which to make this connection and attribute the private acts to the State differs slightly between different bodies and depending on the non-State actor involved. The bases remain significantly limited by the current international legal framework. So far, with the exception of the unique case of Sadiq Shek Elmi v. Australia (which, it must be remembered, did not actually apply the Convention obligations to a non-State actor), treaty bodies seem reluctant to push the boundaries too far, at least in their views on individual communications, in which they are careful to invoke a legitimate legal basis. In order to fill gaps in human rights protection arising from situations falling outside of the State’s obligation to protect human rights, either the international human rights framework will have to evolve to cover certain non-State actors, or ways of protecting individuals outside of the confines of the legal framework will have to be strengthened.
Overall, in relation to the kinds of actors that are treated differently, perhaps what is not present in the documents analysed is more telling than what is – the application of the law seems to allow protection from individuals, and from private companies, especially privatised ones, but there are some non-State actors that fall (sometimes completely) out of the mix. The more ‘public’ non-State armed groups and international organisations have been dealt with to a limited degree, which has been particularly significant for the potential direct horizontal effect of human rights. Their treatment has been more direct within general comments, whilst in views on individual communications it has at most been said that these actors could be capable of violating human rights, or have responsibilities to help States in the implementation of their rights.
5.3 Critical Reflections on the Treaty Bodies’ Reasoning
Before drawing conclusions on the application of horizontal effect by the treaty bodies examined, a few comments on the bodies’ reasoning will be made. Although it did not form part of the research of this article per se and will thus not be discussed in detail, it was very interesting to see how the Committees made their decisions in relation to individual complaints, and how they framed their comments in general recommendations and comments. The following comments are made only in relation to the practice included in the analysis, and should not be read as applying to the practice of the treaty bodies generally.320
Each of the five bodies regularly relied upon their own previous practice, both in general comments and views on individual communications. They did refer to each other’s practice as well, although not particularly often.321 The analysis nonetheless showed that at least in relation to those aspects of the bodies’ practice relating to horizontal effect, the interpretations of the bodies have converged to a large extent (notwithstanding the inconsistencies in the use of language).
They also regularly mentioned other international treaties in general comments (particularly the other ‘core’ human rights treaties) as well as, to a lesser degree, the output of international organisations, particularly un agencies and subsidiary organs. However, it was striking to see how rarely most of the bodies relied on sources of international law other than international treaties when interpreting the human rights treaties.322 Indeed, even in instances where one of the parties to an individual communication relied upon the jurisprudence of judicial bodies (e.g. the European Court of Human Rights) or customary international law, the treaty bodies most often declined to mention the sources in their own reasoning.323 It was very rare indeed that a treaty body referred to customary international law. Although several references were made,324 on most occasions the reference was simply to the fact that as well as being bound by the relevant treaty, States were also bound by customary international law that covered the same material. An exception to this was the CteeESCR’s General Comment No. 24, which referred more substantively to customary international law,325 as well as to the dasr,326 which have customary status. In fact, in both its general comments and views on individual communications, the CteeESCR consistently took into account a broader range of sources than the other treaty bodies.327 Perhaps General Comment No. 24 will prove to be part of a broader trend to take into account other sources of law, although it does not seem as though the other bodies have considerably altered their approach to this over the years.
Overall, as also suggested by the previous comments in relation to attribution and the dasr, the reasoning of the human rights treaty monitoring bodies in the practice analysed was sometimes lacking in terms of grounding outcomes in a legal basis. While the treaties being applied were always discussed, the legal reasoning which led to a particular interpretation of the provisions was sometimes extremely minimal. For example, it is unclear whether the treaty bodies follow the Vienna Convention on the Law of Treaties’ rules on interpretation – something that has been discussed more generally.328 This may cause problems concerning the legitimacy of the bodies’ jurisprudence and the willingness of States to implement changes pursuant to general comments and/or views, both of which have already been called into question.329
In turn, such problems may affect the role and impact of the treaty bodies within international law more generally, as well as their place within the community of international courts, tribunals and other adjudicatory bodies. These aspects should also be considered when looking at the significance of the treaty bodies’ practice for what concerns the horizontal effect of human rights. It is important to remember, for example, that unlike other human rights adjudicatory bodies such as the European Court of Human Rights, the treaty bodies do not have binding authority.330 Nonetheless, their practice has been relied upon by binding regional and international adjudicatory bodies (e.g. the European Court of Human Rights, the Inter-American Court of Human Rights and the International Criminal Tribunals for the Former Yugoslavia and Rwanda, respectively),331 although to a lesser degree than by other non-binding bodies (the Inter-American Commission on human Rights, for Example, has referred to the practice on many occasions). Together with the prevalent reference to the practice of treaty bodies in the national context (both in jurisprudence and in the development of new legislation), this gives the findings of the bodies more significance than may be initially expected; the practice may not constitute formal (international) law, but certainly constitutes important guidance for those bodies (whether national, regional or international) that are able to make binding decisions on the same or related matters.
Bearing this in mind, it is a shame that the treaty bodies are not more candid in their reasoning, which currently makes it harder to evaluate whether or not they do justice to the topic of non-State actors and human rights. In particular, a lack of precise explanation of the legal bases for conclusions within general comments (although this may be changing) sometimes obscures the Committees’ views on horizontal effect. This could be clarified through more reference to and engagement with rules and sources of international law and theories of horizontal effect in those instances where the Committees do indeed apply it, as well as more consistent use of terminology. As it stands, the Committees very rarely actually consider ‘horizontal effect’ as a concept, preferring to move directly to the practical measures that should be taken by (predominantly) States to protect human rights. That being said, the bodies have applied horizontal effect to the extent that their mandates and the international legal framework allow. In this respect, the findings of the analysis fit within the prevalent approach in legal science towards horizontal effect – that within the international legal framework as it is, there is no possibility of direct horizontal effect. The findings do not, however, seem to fit within new theories of horizontal effect being developed in literature (e.g. by Kanalan, discussed briefly in Section 3.2). However, the adoption and content of General Comment No. 24 regarding business activities suggests that at least the CteeESCR is willing to keep pace with the international community as it moves towards the elaboration of duties and obligations for business enterprises.
This article has conducted an in-depth analysis into whether and how five un human rights treaty monitoring bodies deal with situations in which the direct cause of a human rights violation is a non-State actor. The analysis was conducted according to the framework of direct and indirect horizontal effect of human rights and in consideration of the confines of the current framework of international human rights law.
It was found that the bodies are still looking mostly at various aspects of States’ positive obligation to protect human rights, which sometimes enables the bodies to attribute the conduct of non-State actors to the State. The current international human rights law framework prevents non-State actors from being dealt with directly at the international level, allowing only States to have direct international obligations. Although with one exception the monitoring bodies themselves do not actually refer to the ‘horizontal effect’ of human rights, their interpretations and applications of treaties clearly acknowledge, on many occasions, the fact that situations of human rights violations can be caused by actions of private actors as well as States. On some (much more limited) occasions the monitoring bodies have explicitly acknowledged a more significant and perhaps even positive role for non-State actors within international human rights law.
The first type of indirect horizontal effect identified in the general comments and the jurisprudence of the monitoring bodies is ‘pure’ indirect horizontal effect – the application of States’ positive obligation to protect human rights from the harmful actions of non-State actors, according to the tripartite typology of human rights. As noted above, this approach has been taken by each of the bodies in relation to a variety of non-State actors. It can generally be considered to consist of an obligation of due diligence, including a specific obligation to regulate, vis-à-vis the conduct of non-State actors. This is an obligation of conduct, meaning that the focus is on the types of action that States take at the national level to prevent, investigate, punish and provide redress for human rights violations caused by non-State actors. For this reason, the requirements and recommendations laid out by the Committees are mostly procedural, rather than explaining the concrete human rights standards which apply to non-State actors at the national level.
The second approach identified was ‘categorical’ indirect horizontal effect, under which non-State actors carrying out particular public functions normally undertaken by the State and having effective control over a certain area of territory are ‘re-categorised’ or ‘re-classified’ as public actors for the purposes of a specific situation. This provides extra protection for individuals in situations where the State may fulfil its obligation of conduct under the obligation to protect, but remains in a position where it is unable to provide de facto protection. So far, this has only been applied by the CteeAT (and arguably also CteeEDAW) and in relation to non-State armed groups.
In light of the non-binding nature of the documents analysed and the current international legal framework, we still cannot actually say that non-State actors currently have any binding obligations concerning human rights at the international level. However, the analysis shows that monitoring bodies are certainly engaging with the impact and effect that non-State actors have on human rights. To a more limited degree, they are also engaging with the prospect of direct responsibilities of non-State actors.
Lottie Lane, llm is a Ph.D. Candidate at the Endowed Chair, Groningen Centre for Law and Governance. Her Ph.D. research focuses on the horizontal effect of international human rights law from a law and governance perspective. The author would like to thank Marlies Hesselman and the two anonymous reviewers for their extremely helpful comments and suggestions.