I. Introduction
The new mechanism of the Committee on Economic, Social and Cultural Rights, which it utilizes to receive individual complaints and issue views, is a tool with tremendous potential for the promotion of human rights globally. It establishes an individual complaint mechanism for important human rights such as the right to work, adequate standard of living, education and health. In many states, particularly if they are not party to one of the three regional human rights systems, this mechanism is the only way for individuals to hold their state accountable for violation of those rights. However, of the 171 states that are parties to the International Covenant on Economic, Social and Cultural Rights (hereafter, ‘the Covenant’), only 24 have ratified the Optional ProtocolFootnote 1 creating the individual complaint mechanism, and 25 more have signed the protocol but have not ratified it. Under these circumstances, the way the Committee interprets and adjudicates the broad articles of the Committee is crucial – both to increase the effective protection of these human rights and to show states, individuals and civil society actors that this mechanism is normatively desirable. In other words, questions and critiques of the Committee’s interpretation relate not just to its legality, but also to its legitimacy (or lack thereof). So what makes human rights interpretation legitimate, and how should the Committee interpret and adjudicate the broad articles of the Covenant?
The literature on the adjudicatory practice of UN treaty bodies has developed rapidly in recent years – in particular, several edited volumes covering the various reporting processes (General Comments and Views) for each treaty body (see in particular Keller, Grover and Ulfstein Reference Keller, Grover, Ulfstein, Keller, Ulfstein and Grover2012; Moeckli, Keller and Heri Reference Moeckli, Keller, Moeckli and Heri2018). One important issue of human rights interpretation is shared across the general (UN treaty body) and specific (Committee) levels: its oscillation between two conflicting strands of interpretation with which the Committee must come to grips – a ‘moral reading’ of treaty provisions (e.g. through the teleological approach) and a ‘state consent’ approach (e.g. through the textual approach). While the Vienna Convention on the Law of Treaties of 1969 (hereafter, ‘the Vienna Convention’) lists these as acceptable approaches, it does not indicate how to choose between or balance them. To resolve this issue, many contributions in the literature support what we call the ‘Vienna Convention process’ suggesting that interpreters should take into account all interpretive options contained in Articles 31–33 of the Vienna Convention and, more importantly, better explain their interpretive findings in order to persuade their subjects. This would increase legitimacy and particularly compliance. This understanding reduces all questions of legitimacy to sociological legitimacy based on acceptance and compliance. However, from the perspective of normative legitimacy, it is not clear why this approach is most suitable. Empirically speaking, it is not clear that it in fact does increase compliance.
In this article, we argue conversely that the individual communication mechanism should be evaluated in normative terms as legitimate authority. Following Joseph Raz’s (Reference Raz1986, Reference Raz2006) seminal idea, legitimate authority consists of exclusionary and content-independent reasons to comply. On this view, subjects would not evaluate the content of the treaty body’s interpretive findings directly, but rather rely on the legitimate authority of the issuing institution. We argue that relying on the understanding of legitimate authority makes the characteristics of the institution (that claims authority), and particularly its procedures, essential. To establish which procedure is appropriate for human rights interpretation, one has to step back and ask what reasons we have for valuing international human rights law and hence how such treaties should be interpreted to maintain and strengthen their claims for deference.
While often confined to matters of interpretation and the various established approaches to it within positive international law, the tension between morality and the consent-based approach reveals a deep and genuine philosophical problem of reconciling personal and political autonomy. The morality approach privileges the personal autonomy of right-holders, while the consent-based approach places emphasis on the political or collective autonomy of the state. An autonomy-based understanding of legitimate authority takes both personal and (collective) political autonomy to be relevant. We argue that a clear and consistent procedure that addresses this tension is more likely to generate content-independent reasons than applying (and explaining) the various interpretive options of the Vienna Convention.
We suggest that one of the most suitable procedures in the case of human rights is proportionality analysis. This is because the latter provides a systematic procedure to balance between these two core foundations of human rights. The classical formulation of the proportionality test requires that interferences with rights are permissible if and only if (1) the state party pursues a legitimate aim; (2) there is a rational connection between the aim and the means employed; and (3) there is no less invasive alternative. We argue that proportionality hence offers the procedural framework appropriate to the task of attenuating the tension between personal and (collective) political autonomy while accommodating the specificities of economic, social and cultural (ESC) rights. Political autonomy is factored into the Committee’s proportionality analysis, as rights restrictions are possible if they benefit the ‘welfare of society’. This applies, for example, to the question of potentially restricting the right to adequate housing (Article 11 of the Covenant) in the name of austerity measures, as we shall see in the last section of this article. By arguing for an account of legitimate authority, this article aims to propose a coherent way to improve the Committee’s normative legitimacy. While other important procedural aspects are relevant for the effectiveness of the Committee that also affect the latter’s legitimacy, such as improving communication with complainants and addressing the case backlog and delays in processing individual complaints (Çalı and Galand Reference Çalı and Galand2020), this article limits itself to the interpretation of the Committee and the legitimacy issues associated with that.
The argument proceeds in three steps. First, we reconstruct the debate between international legal scholars over how to evaluate the adjudicatory practice of UN treaty bodies with particular emphasis on the Committee. By doing so, we aim to critically analyze how the Vienna Convention+process view is supposed to enhance legitimacy. Second, we show why a fully sociological account of legitimacy is insufficient and explore the normative assumptions of the Vienna Convention+process. On this basis, we propose an autonomy-based normative understanding of legitimacy. Third, in order to confer legitimate authority to the Committee, a clear procedure needs to be followed – one that addresses the tension to which the Vienna Convention itself points but that it does not systematically resolve. We show that proportionality analysis is just such a procedure and that it is able to generate the kind of content-independent reasons that are distinctive of legitimate authority.
II. The Vienna Convention as legitimacy enhancing
In this section, we introduce the evaluative framework that has come to predominate in the literature on the legitimacy of UN treaty bodies with particular application to the Committee. In order to do that clearly, one needs a grasp of the very concept of legitimacy, which is undoubtedly multifaceted. We start with the well-established distinction between sociological and normative legitimacy, then zoom in on the debate that has occupied the research landscape around the legitimacy of UN treaty bodies in general and the Committee in particular.
Going back to Max Weber (Reference Weber1964), sociological legitimacy is generally defined as the belief in the legitimacy of a system (‘Legitimitätsglaube’). However, this sociological aspect of legitimacy can be distinguished from normative legitimacy, which reflects the moral justifiability of political power or authority. Often, normative legitimacy is conceptualized as the right to rule (Raz Reference Raz1986). Questions of legitimacy arise in the context of institutions that claim authority or demand compliance. Therefore, both notions of legitimacy are connected to questions of compliance on the part of those subjected to authority. While sociological legitimacy answers the question ‘Do they believe that they should comply?’, normative legitimacy addresses the question ‘Should they comply?’ In other words, normative legitimacy concerns what the sociological beliefs are about. For example, states may think they are bound by an international treaty because they have signed it, establishing its sociological legitimacy. To establish the normative legitimacy of the treaty, one needs to know whether signing it is sufficient to bind states, morally speaking. A certain level of sociological legitimacy is necessary for institutions to function well, and therefore also for their normative legitimacy (Buchanan Reference Buchanan2002). Yet reducing legitimacy solely to its sociological aspects is problematic. When understood exclusively in sociological terms, legitimacy is seen to be a social fact about held beliefs that are disconnected from the reasons for holding these beliefs in the first place. Such reasons may have to do with the characteristics of the regime or institution, among others. According to David Beetham (Reference Beetham1991: 11), ‘a given power relationship is not legitimate because people believe in its legitimacy, but because it can be justified in terms of their beliefs’. Thomas Franck (Reference Franck1990: 713), while considering legitimacy’s pull toward compliance, is careful to define it as ‘a property of a rule or rule-making institution’. This means that, in order to understand legitimacy fully, we have to understand the normative grounds that an institution provides to comply with its rules – and this equally applies to rules governing the adjudicatory process of the Committee.
Now, we need to have a better grasp of the kind of argument that actually prevails in the scholarly debate on the adjudicatory practice of UN treaty bodies. A review of the literature reveals the number of references to, and reliance on, the Vienna Convention as the authoritative and positive set of secondary rules to interpret human rights covenants and conventions. Moreover, as we shall see, it is also a tool to bolster the bodies’ sociological legitimacy. In their investigation of the General Comments of UN treaty bodies, Keller and Grover (Reference Keller, Grover, Ulfstein, Keller, Ulfstein and Grover2012: 167) put the point quite clearly: ‘the legitimacy of General Comments seems to benefit from reasoned statements that, expressly or implicitly, adhere to secondary rules of interpretation’.
That the Vienna Convention operates as a starting point for interpretation should not be a surprise. Articles 31–33 of the Vienna Convention are understood as the relevant and authorized guidelines for interpreting all international treaty law; the various UN human rights covenants and conventions form part of that body of law. Further, Articles 31–33 are generally understood as reflecting international customary legal rules following the list of authorized sources in the Statute of the International Court of Justice (ICJ), Article 38(1). It is certainly beyond the scope of the present article to address all the relevant Vienna Convention articles and reconstruct how legal scholarship has assessed them in the context of human rights law. Rather, we suggest focusing on Article 31(1) in particular for the sake of the discussion, since this Article defines various approaches to interpretation. To recall, Article 31(1) requires treaties to be interpreted ‘in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose’. Most importantly, this passage delineates three different approaches to interpretation that scholars routinely use across the various sub-fields of international law: the reference to ‘ordinary meaning of the terms’ grounds an approach labeled textual; the reference to context grounds an approach called evolutive; and the reference to ‘object and purpose’ grounds an approach called teleological. There has been a long-standing debate over whether a hierarchy exists between these three approaches in international legal scholarship, what each approach encompasses and how exactly they relate to each other (Çalı Reference Çalı and Hollis2018). Notwithstanding this general debate, there has also been a more localized debate on how these approaches to interpretation should apply to human rights norms. We shall focus on this debate for the rest of the discussion.
Human rights law poses particular problems for the application of these secondary rules. Indeed, while ‘the classical methods of interpretation in the Vienna Convention have not lost any of their relevance in international human rights law’ (Schlütter Reference Schlütter, Keller, Ulfstein and Grover2012: 277), many scholars concurrently observe that the Vienna Convention framework is considerably limited – and, some suggest, deficient – in guiding interpretation. As a result, it has been argued that interpretation should benefit from special approaches that do not squarely fit the Vienna Convention framework. For example, the principle of effectiveness suggests that one should privilege an interpretation that gives meaning and effect to all the terms of the treaty, which implies that these rights should be read as having an effective impact on the life of right-holders. In order to explain this divergence from the standard Vienna Convention framework, most scholars point to the peculiar non-reciprocal structure of human rights obligations: only the state party owes these rights to individuals, but not the other way around – unlike the traditional reciprocity of treaty obligations. Birgit Schlütter (Reference Schlütter, Keller, Ulfstein and Grover2012: 264) puts the point clearly: ‘the relationship of the obligation is vertical. Human rights are also phrased to protect the individual from the state, so that the rights, by their very nature, subvert the concept of state sovereignty.’
It is interesting to reconstruct the range of implications that legal scholars draw from this special structure for interpretation. Overall, it appears that human rights scholars acknowledge that this specialty calls for some special interpretive methodology, but whether that implies departing from the Vienna Convention framework is less clear. That is, one may argue that the Vienna Convention rules themselves need to be applied in a particular fashion in human rights cases:
this development is perhaps a reflection of the particular dynamism inherent to human rights law. It deals directly with the relationship of the individual and the state and is thus one of the areas of law where a change in social realities and conditions can exercise a direct influence on the applicable law. (Schlütter Reference Schlütter, Keller, Ulfstein and Grover2012: 317)
The precise path from norm to method can take different forms. In other words, the vertical structure of human rights law implies privileging one of the Vienna Convention approaches, but the human rights specialty remains within the confines of the Vienna Convention. It is about balancing these authorized methods accordingly, not about departing from them. Indeed, several scholars have argued that in fact these special methods are not that different from the Vienna Convention framework. Schlütter (2016: 317), for instance, holds that ‘the treaty bodies generally follow the rules of the VCLT and supposedly special methods, such as the principle of dynamic treaty interpretation and the principle of effectiveness, also fit well within the concept of the VCLT’.
To say that an interpretive framework such as the Vienna Convention is limited does not necessarily entail questioning the referential role it performs. Among defenders of specializing, none is advocating another realm of interpretation than the one established by the Vienna Convention. This is important for the purpose of categorizing the realm of reasons that prevail in the current discussion. Moreover, some scholars even question the very specialty of human rights norms compared with other international norms in that respect. With reference to the teleological approach, it has been argued that
it is not clear why, simply because human rights treaties pursue an important objective, the teleological element should assume a greater importance than when it comes to, say, investment treaties. The object and purpose of every treaty is, when compared to other treaties, special. (Moeckli Reference Moeckli, Keller, Moeckli and Heri2018: 50)
When it comes to the broader practice of international courts and tribunals, Bjorge (Reference Bjorge2014: 36) argues that ‘the jurisprudence of international courts and tribunals show that the object and purpose is, together with the intentions of the parties, the prevailing elements for interpretation in any type of treaty’.
What does this overview tell us about the predominant notion of legitimacy? For the sake of our argument, it is enough to indicate that the arguments remain confined to the realm of the Vienna Convention. The notion of legitimacy mainly used in the context of adjudicatory practice of UN treaty bodies is both sociological and content-dependent. First, legitimacy in this debate is understood in sociological terms. Sociological legitimacy refers to the acceptance of, or in some cases even simply compliance with, a norm (in our context, Views under the new complaint mechanism or General Comments issued by the Committee) of the addressee (state parties). A wide range of authors suggest that privileging these arguments is likely to improve the sociological legitimacy of treaty bodies. That is, it is assumed that the adjudicatory practice of UN treaty bodies should be confined to the Vienna Convention for that purpose. In other words, if the objective is to increase the acceptability of their adjudicatory practice on the basis of beliefs about them being morally binding, UN treaty bodies should rely (more) on the canons of interpretation that are internal – as the Vienna Convention is the authoritative set of rules – to the discipline of international law. Second, legitimacy is understood as content dependent. This means that it is not the source or procedure of a decision that generates this acceptance and compliance but its content. It is assumed that the outcome of the interpretive process issued by the institution matches the view held by the subject and may therefore motivate the subject to take action (in our context, whether states conform to Views or General Comments in the domestic legal order).Footnote 2 In this sense, the state party qua subject evaluates the reasons given by the treaty body. Importantly, a treaty body may also persuade its subject(s) through its interpretive reasoning. It is argued that by basing their interpreting findings on the Vienna Convention, subjects of treaty bodies would have more reasons, and are more likely to be persuaded, to comply with them. This sociological and content-dependent account of legitimacy is important to keep in mind for the second and third parts of the article, where we defend an alternative, content-independent account of legitimate authority.
Let us now illustrate this understanding of legitimacy with reference to the specific practice of the Committee. In a notable contribution, Daniel Moeckli (Reference Moeckli, Keller, Moeckli and Heri2018) argues that although the Vienna Convention rules establish the relevant framework for interpretation, there is too much disparity and oscillation in the current practice of the Committee – disparity in that the methods used differ too much from one another, and oscillation insofar as there is an excessive amount of swinging between them. This prevents the Committee from generating the added value in terms of legitimacy that it could do (and indeed should do, following the normative implication above). This approach fits the standard of sociological and content-dependent legitimacy based on the Vienna Convention. It argues that one should focus on generating ‘an empirical evaluation [that] must be undertaken to identify the qualities of a decision-making process that lead the relevant actors to regard it as adequate and fair’ (2018: 66). While these qualities may be found in other fields of law and practical reasoning, again they should apply with respect to the Vienna Convention rules.
Moeckli’s (Reference Moeckli, Keller and Heri2018) argument starts with the observation that the practice of the Committee is dominated by two Vienna Convention-compatible approaches: ‘morality’ and ‘state consent’. ‘Morality’ reflects the teleological method comprised in Article 31 Vienna Convention – the Committee has, for example, relied on the principle of dignity found both in the Committee’s Preamble and the Universal Declaration of Human Rights to justify using the teleological approach. Meanwhile, ‘state consent’ echoes the recourse to the preparatory works of the Committee authorized by Article 32 of the Vienna Convention. This illustrates the oscillation between the two approaches in the Committee’s definition of the ‘minimum core obligations’: at times, the Committee takes the expansive view that the obligations ought to be met regardless of the respondent state’s available resources. At other times, though, it acknowledges that resource limitations could justify the failure of meeting these core obligations. In Moeckli’s (Reference Moeckli, Keller and Heri2018: 65) view, this oscillation is problematic for legitimacy, since it fails to meet the basic process values that are central to what he calls the ‘interpretive community’, which refers to a very broad definition of the relevant audience of UN treaty bodies.Footnote 3 This claim, we should note, is not derived from the commitment to remain within the realm of the Vienna Convention.
As a result, in order to improve sociological legitimacy, this predominant approach does not simply call for more of the Vienna Convention; rather, it recommends better organizing the application of these various interpretive elements authorized by the Vienna Convention: ‘there may be perfectly good reasons to give a teleological approach priority over a textual interpretation and, even more so, over an interpretation according to original intent. However, if an interpretation is to be regarded as legitimate, the various interpretive elements should at least be dealt with’ (Moeckli Reference Moeckli, Keller, Moeckli and Heri2018: 66, our emphasis). In other words, the various permitted elements of the Vienna Convention framework should be more systematically examined in the interpretive process. Not only should the interpreter address all the interpretive options, they should also better explain the logical reasoning underlying the findings and means of interpretation in a transparent manner. On this view, the very principle of transparency, for instance, does not have an independent normative value; rather, such a principle is valuable insofar as it increases the community’s perception of these findings as worthy of support: ‘an interpretation will only appear legitimate if it is reached by adhering to the principles agreed upon by the interpretive community for this very purpose. The game must be played by the rules’ (2018: 66). Thus, in a nutshell, on this view the recipe for improving legitimacy is to remain within the Vienna Convention rules but to combine their application with a rigorous application of process values. These process values should further help persuade the relevant community of the legitimacy of the Committee’s interpretive findings. Here we refer to this proposal to combine the Vienna Convention rules with process values as ‘Vienna Convention+process’.
Before moving to a critical analysis of the Vienna Convention process view as a form of sociological legitimacy, it is important to explain why the emphasis on the Vienna Convention as legitimacy enhancing is so strong in the literature. One may concede here that various functions of the Committee should always be placed within the broader institutional context in which the Committee operates and also within the framework of its limited legal authority. That is, the findings of the committee have always depended significantly on the cooperation of state parties to the Committee. Although this applies to the interpretive decisions of all treaty bodies to varying degrees, their decisions are not legally binding as a matter of international law, which may justify such interpretation. Moreover, unlike the Human Rights Committee (HRC), for a long time the Committee had to rely on other interpretive instruments (General Comments, Letters to State Parties, Concluding Observations and so on) to establish its legitimate role before it could deliver Views through an individual complaint mechanism. The literature on the history of the Committee makes clear that its influential contribution was through General Comments and Concluding Observations, which were endorsed by a variety of international organizations (Moeckli, Keller and Heri Reference Moeckli, Keller and Heri2018). This record could strengthen the pull toward using an interpretive framework that does not depart from established standards in international law when the Committee examines individual complaints and renders Views. From the perspective of the Committee’s history and its always-fragile authority, reliance on established and authorized standards of the discipline of international law remains attractive.
Yet, as we will show in the next section, the sociological legitimacy that might be enhanced is not only empirically dubious; it is also conceptually insufficient to exhaust the realm of legitimacy claims – in particular, its normative dimension. This is a crucial point that the literature on the Committee’s legitimacy tends to overlook. Furthermore, the application of Vienna Convention standards to individual circumstances through the complaint mechanism highlights an under-noticed tension between the two Vienna Convention-derived approaches outlined earlier: the ‘moral’ reading and the ‘state consent’ reading. We believe that this tension is serious, and that a failure to address it may also affect the sociological legitimacy of the Committee’s findings in the long run, as we explain below.
III. From sociological to normative legitimacy
The article so far has attempted to reconstruct the concept of legitimacy that underlies the debate on the legitimacy of the Committee’s approach to interpretation. Having highlighted the predominantly content-dependent and sociological approach to legitimacy in this field, the next step is to critically engage with that approach. We begin by scrutinizing the Vienna Convention+process by pointing to some assumptions that, in our view, are not fully addressed. We then further unpack the distinction between the ‘morality’ and ‘state consent’ approaches that drives that argument. We believe that this tension is distinctive of human rights law, as it is far higher here than in other fields of international law. The reason underlying this distinctiveness, we believe, is that the morality and state consent approaches refer to two important moral notions: personal and political autonomy, respectively. The legal literature reviewed above does not sufficiently appreciate these notions precisely because it is still heavily influenced by positivism (which coincides with a sociological account of law on a Hartian view) and of which the reliance on the Vienna Convention is an instance. In contrast, we suggest opening up the theoretical horizon of research to address personal and political autonomy as underpinning interpretation and re-evaluate the practice of the Committee in that light.
Shortcomings of the fully sociological approach of legitimacy
This distinction between sociological and normative legitimacy helps to better identify the shortcomings of the prevailing sociological approach. Three points deserve mention here. First and foremost, legitimacy assessments cannot be confined to sociological considerations. The Committee’s legitimacy is important not only because it seeks to motivate states under conditions that lack legally binding power and coercive authority, but also and simply because claims to authority need to be justified (e.g. Scherz Reference Scherz2019). The Committee’s interpretations only have to be complied with if they are legitimate in the normative sense. Only by understanding the normative reasons that justify claims to authority is it possible to understand how certain institutions create grounds for compliance that go beyond mere self-interest and coercion.
Second, but related, a purely sociological understanding of legitimacy loses sight of these important normative underpinnings – and with it a theory that can explain why states and individuals hold certain beliefs and whether they should do so. On such a view, the principles and methods valued by the interpretive community itself (‘Vienna Convention+process’) are assumed to be legitimacy enhancing. In other words, they are legitimate because they are believed to be so. Yet this does not tell us anything about the reasons for holding these beliefs in the first place. That is precisely where the sociological approach falls short. It assumes that if the treaty bodies were to follow the Vienna Convention+process approach, their subjects could align with the treaty body’s reasoning. A more specific version of this generic argument has to do with the two approaches to Vienna Convention-inspired methods: ‘morality’ and ‘state consent’. In our view, the assumption that the Vienna Convention+process will increase legitimacy comes at the price of underestimating the conceptual tension between the ‘morality’ and ‘state consent’ approaches. In other words, whether state A and state B (both members of the relevant interpretive community) would agree with the choice of one or the other by the treaty body depends on whether the antagonism between these approaches can be reconciled. Moeckli (Reference Moeckli, Keller, Moeckli and Heri2018: 64) acknowledges this point to some extent, noting that
it is important to acknowledge that morality and State consent are not strict opposites: an interpretation that appears to be morally appropriate will often be one that finds the support of States, and vice versa. Nevertheless, sometimes morality and State consent do pull in different directions.
Presumably, if the two approaches pull in different directions, it is likely that the subjects’ perception of these methods will itself pull in different directions too, which then prevents the overall increase of perceived legitimacy on the subjects’ part in the first place.
Third, even if we assume that what should be fostered is sociological legitimacy, we need empirical studies to establish whether the Vienna Convention+process does in fact do this. Yet this view offers no such empirical evidence in this regard. In particular, the three requirements that Moeckli (Reference Moeckli, Keller, Moeckli and Heri2018) suggests in order to improve the interpretive process (coherence, adherence and transparency) are provided without reference to any empirical studies that show how these interpretive characteristics improve the belief in an institution’s legitimacy or states’ compliance with it. This is not surprising as, empirically speaking, the effects of reasoning of opinions on the acceptance and compliance of states is not straightforward and complex to measure (e.g. Bonneau et al. Reference Bonneau, Kelly, Pronin, Redman and Zarit2017; Farganis Reference Farganis2012). There seem to be some effects of opinion and remedy clarity (e.g. Staton and Romero Reference Staton and Romero2019; Stiansen Reference Stiansen2019). Yet these do not necessarily signal that they are more convincing, but rather that opinion clarity in part makes it easier for states to know what they have to do in order to comply and it enables pro-compliance actors to build up public pressure. Overall, the dynamics of compliance and different strategies in interpretation are complex.
Fourth, the Vienna Convention+process view rests on normative premises because it is not clear that this view can stay within the sociological legitimacy framework all the way down. Interestingly, when it comes to explaining the three procedural requirements, Moeckli (Reference Moeckli, Keller, Moeckli and Heri2018) seems to invoke distinctively normative grounds for supporting these particular procedural qualities without making it explicit. He refers to ‘virtuous interpretations’ (2018: 66) and notes that ‘the rules of the game cannot be changed in the middle of the game’ (2018: 67). Yet, it remains unclear from this analysis why it is wrong and how the proposed procedures are relevant to legitimacy. On a fully sociological understanding, changing the rules of the game during the game is wrong because people believe it to be wrong and will not comply. What is meant here is not merely that not changing the rules will create more compliance, but rather that it is normatively wrong to change the rules in the middle of the game, and therefore also generally perceived as such. We believe that by turning to normative legitimacy, it becomes clearer why procedures are particularly important in generating legitimacy.
Normative legitimacy: Justifying authority
Having identified the shortcomings of the content-dependent and sociological account of legitimacy as it applies to the adjudicatory practice of the Committee, we now offer our own account of legitimacy. As outlined above, questions of legitimacy arise in the context of an authority that demands compliance. Raz (Reference Raz1986: 46) notes that claims of authority are absolute claims: do as I say. In this sense, they are exclusionary, as they constitute second-order reasons for the prescribed action that exclude at least certain other first-order reasons against that action (1986: 46). Second, if an institution has legitimate authority, it creates content-independent reasons for compliance (1986: 35). This means that an authority’s decisions or rules are normatively binding even if they are mistaken.Footnote 4 Why is content-independence important? Because it has to do with the function of legitimacy, namely to create a focal point for reliable coordination which requires consistent deference in order to provide stable expectations. If we are interested in the normative legitimacy of an institution, then we are interested in the question of whether someone should comply with its rules or directives, particularly if these rules do not promote one’s immediate self-interest (e.g. Buchanan Reference Buchanan, Sobel, Vallentyne and Wall2018; Franck Reference Franck1990: 21). Therefore, legitimacy cannot rely on the evaluation of the merits of its decision in individual cases. This would undermine the service that authoritative institutions provide – that is, stable coordination beyond self-interest (and coercion). In other words, it cannot rely on content-dependent reasons, since this would give too much weight to self-interest and undermine stability.
The next question is how an institution more precisely creates content-independent reasons for compliance. Raz famously suggests that a person or institution has authority if they fulfil the so-called normal justification thesis (NJT). The NJT holds that the authority helps its subjects better comply with reasons that apply to them by abiding by the authority’s directive rather than by evaluating the directive themselves.Footnote 5 Generally, Raz’s approach, and the NJT in particular, are understood as an instrumental account of legitimacy, because this requires evaluating the overall instrumental benefit that the institution produces. On this reading, it does not matter how the institution reaches decisions that make the individual better comply with reasons; all that matters is that it is better on average. However, newer readings of Raz suggest that procedural elements, such as the fairness of the democratic procedure, can also be seen as reasons in the NJT (e.g. Besson Reference Besson, Føllesdal, Schaffer and Ulfstein2014; Viehoff Reference Viehoff2014).Footnote 6 Since it is specifically the characteristics of institutions that create content-independent reasons to comply, the procedures that institutions apply are an essential source of content-independent reasons. Further, if legitimate authority is central to coordination, it is also decisive that the normative grounds for legitimacy are epistemically accessible and assessable. Footnote 7 In order to define which procedures generate content-independent reasons along those lines, we need a normative account of the grounds of legitimacy to start with. While the literature often distinguishes between instrumental, consent-based and democratic conceptions of legitimacy (e.g. Peter Reference Peter2010), we propose an autonomy-based conception of legitimacy that we take to be fundamental for the others.
Autonomy-based legitimacy
Normative legitimacy judgements of authoritative institutions assess whether they sufficiently protect the conditions of autonomy of those subject to the rules and those affected by them. The notion of autonomy employed here considers two classical dimensions of personal and political (collective) autonomy (Forst Reference Forst2012: 125–37; Habermas Reference Jürgen1996: Chapter 3). Personal autonomy is understood as the individual’s capacity or freedom of choice, while political autonomy is the capacity to act as an equal co-author in political decisions. Demands for compliance have to be justified because they pose particular threats to autonomy. First, compliance restricts the set of options available – that is, the freedom of choice decisive for personal autonomy. Second, compliance with rules also restricts those subjected as norm-givers – or, in other words, their political autonomy as self-legislative agents. This reliance on individuals’ self-chosen reason is warranted if we think of persons in a Kantian sense as ‘self-originating sources of valid claims’ about what should be done (Rawls Reference Rawls1980: 543; Rawls Reference Rawls1993: 32). It is important to note on this view that the autonomy of individuals is never unrestricted autonomy. Rather, it is always already constrained by the autonomy of everyone else – in this sense, it is relational equal autonomy that we are considering here.Footnote 8 Such rules have to be justified to those subjected to them by ensuring their compatibility with equal personal autonomy, in particular basic rights that robustly ensure freedom of choice, and political autonomy, ideally realized through democratic participation. Both these sides of autonomy are equally important – or as Habermas (Reference Jürgen1996) puts it, they are ‘co-original’ (Schaffer Reference Schaffer2015). In this sense, our account can be understood as what Nicole Roughan (2013) describes as conjunctive, taking both substantive and procedural justifications into account.Footnote 9 Therefore, the legitimacy of international institutions and human rights bodies in particular depends on both how they respect and protect human rights and the decisions of states, representing the political autonomy of their citizens.Footnote 10
This incursion into the notion of normative legitimacy allows us to better understand the extent to which doctrines or methods of interpretation that abound in the literature on interpretation of UN treaty bodies – mostly used at the level of sociological legitimacy – can be viewed through the lens of normative considerations pertaining to autonomy, both personal and collective. Of course, we need to again fully appreciate the paradigmatic shift that occurs here: interpretation reflects a paradigm of legal and positivized authority – and ‘within the bounds of its authority, an institution is not regularly required to justify its actions, but merely to identify the legal source of its authority to act’ (Gardbaum Reference Gardbaum, Huscroft, Miller and Webber2014: 202). As we shall see later, proportionality reflects a culture of justification based on further, properly normative reasoning – ‘an additional or second stage of scrutiny for all government action, in which it must provide substantive justification in terms of public reason for what it has done’ (2014: 202). From the perspective of dual autonomy just outlined, there are important normative reasons for legitimacy that would support an extensive interpretation of human rights norms (personal autonomy), but equally important normative reasons to respect state consent (political autonomy). This helps observe that the two methods of interpretation that Moeckli (Reference Moeckli, Keller, Moeckli and Heri2018) labels ‘morality’ and ‘state consent’ – and the oscillation between these two methods in the practice of the Committee – might be explained by their normative significance, which might in turn explain why they are sociologically prevalent. This would connect the reasons that subjects have (sociological legitimacy) to the reasons they should have (normative legitimacy). By understanding legal and interpretive processes based on normative considerations, we thereby also create reasons for compliance and exert a pull towards compliance. As such, it becomes clear that interpretation requires a procedure that can accommodate – or at least attenuate potential tensions between – both sides of autonomy.
Is the Vienna Convention+process enough for content independence?
The general idea that interpretation needs to be restricted in certain ways – for example, by the Vienna Convention – seems to fit rather well the outlined concept of content-independent legitimacy. Of course, this supposes that we exclusively operate within the realm of normative and not sociological legitimacy. In schematic terms, the question becomes whether the Vienna Convention+process approach, taken normatively, would be likely to generate the kind of content independence required for legitimate authority. We argue, however, that the Vienna Convention does not have in-built procedural elements to address the tension between personal autonomy (the ‘morality’ method) and political autonomy (the ‘state consent’ method) that would be required to generate content-independent reasons. This is a point we mentioned in our critical analysis of the Vienna Convention+process approach: it is not only that there is virtually no firm agreement on when and where to apply one or the other method; in this respect, the quarrels of international lawyers are unlikely to cease. One can also wonder whether the very project of a procedure informed by the Vienna Convention is achievable. In a contribution to a recent volume on the methodology of human rights research, Martin Scheinin (Reference Scheinin, Andreassen, Sano and McInerney-Lankford2017: 21) holds ‘there is considerable flexibility, or even ambiguity, in determining a proper methodology for the interpretation of human rights law’. Nevertheless, he concludes (2017: 26) that ‘different lawyers may propose different solutions in a dispute or to a legal problem, but they should all share a commitment to the same methodology’. This article does not claim that the quest for a Vienna Convention-based methodology should necessarily be abandoned, yet it is equally clear that there is no such methodology available to operate as a procedure – and that procedural quality is required for our approach to an autonomy-based legitimacy. In other words, there is far too much indeterminacy in the current debate around how the Vienna Convention-based methods should apply. As we have argued, this partly has to do with, the inherent tension between personal and political autonomy. Furthermore, we also believe that excluding the use of moral reasoning – or applying it selectively (for instance, only in the teleological method) – may further complicate the project of a methodology based on interpretive methods inspired by the Vienna Convention. As we have seen, the two methods are difficult to conceptually reconcile a priori, and these conceptual constraints do not just apply to moral reasoning – they apply to sociological or positivist reasoning as well. This becomes clearer when one reconstructs the ‘morality’ and ‘state consent’ methods through the lens of autonomy.
The same kind of considerations applies to the procedural requirements of the Vienna Convention+process view (adherence, coherence and transparency). At first glance, these criteria could precisely operate as procedural and content-independent constraints on authority. First, adherence, or ‘playing by the rules’, is not a good normative reason as such. How much legitimacy can be created by rule-following depends first on how normatively desirable the applied rules are. If the playing field is not level, or certain players are allowed to play with their hands as well as their feet while others are not, changing the rules might be called for rather than playing by them. Second, however, we might think that any restriction on interpretive discretion might be restricting the power to arbitrarily decide. In that sense, it limits the undue restriction of the autonomy of those subjected to the ruling. Similarly, coherence should not just be based on the idea that ‘the rules cannot be changed in the middle of the game’, which again relies on the value of the rules in question. Rather, one of the main reasons why procedures are normatively valuable is because they create reliability. If an institution is not coherent and therefore reliable, it has no consistent characteristics by which it can be judged, and as such it cannot create any content-independent reasons necessary for legitimacy. Finally, regarding transparency, it is not clear why the rules have to be applied in a transparent manner. Moeckli (Reference Moeckli, Keller, Moeckli and Heri2018) argues that by better explaining its interpretive processes, the Committee can persuade the interpretive community. However, this is a content-dependent reading of interpretation. Nonetheless, transparency also remains fundamental for generating content-independent reasons. In order to know that there is a process – or, in other words, to assess the characteristics of an institution, this process has to be assessable and transparent. To conclude, we therefore agree with the three procedural values proposed by Moeckli, but for different reasons. Here again, it seems crucial to disentangle sociological and normative legitimacy to explain that these procedural values are in fact relevant at the normative level of argument.
IV. Proportionality and legitimate authority
The last section aimed to transition from a content-dependent sociological account of legitimacy to a content-independent normative account. The key distinction is that the normative account seeks to justify the claim to authority that an institution makes – in this case, the adjudicatory rulings of the Committee – and justify how its subjects’ autonomy ought to be restricted. We argued that, in order to confer legitimate authority to the Committee, there needs to be a clear procedure to attenuate the tension that the Vienna Convention itself points to but does not systematically address. This is necessary for the authority to offer content-independent reasons to its subjects. This emphasis on procedure is not completely absent from the sociological account; however, its normative dimension is not fully fleshed out.
In this section, our aim is to connect the discussion of autonomy-based legitimate authority to the adjudicatory context of the Committee surveyed in the first part of the article. The core claim of this section is that proportionality testing is better equipped than conventional methods of interpretation to play the role expected of a legitimate authority, namely to provide content-independent reasons to their subjects in the context of human rights law interpretation generally, but even more so in the context of the Committee. First, this argument requires showing that the proportionality test can be described as a procedure and how proportionality can best attenuate the tension between personal and political autonomy and, as such, improve legitimacy in a normative sense. Second, it implies explaining how proportionality operates in the practice of the Committee in order to show, in fine, how proportionality qua procedure could improve the normative legitimacy of that specific practice. This second step is important because in that practice, proportionality does not operate exactly as it does in the case of civil and political rights. As we will explain, economic, social, and cultural rights are positive rights that need to be ‘progressively realized’. However, proportionality is far from absent. As Katharine Young (Reference Young, Jackson and Tushnet2017: 13) puts it, ‘the principle of proportionality, without the structured test, has found a home in economic and social rights adjudication’.
The function of the proportionality test
Let us start with the general function of the proportionality test. In short, the function of proportionality in human rights law is to establish the conditions for interfering with one or several protected rights. This implies that the proportionality test applies when courts or adjudicators have already established an interference with one or several rights based on established interpretive principles, as we explain further below. They consider proportionality by turning to the normative grounds that the respondent state had to interfere with these rights. However, not all rights are subject to the proportionality analysis. Classically, rights in constitutional and human rights interpretation are divided between non-derogable and derogable rights. This distinction is clear from the text of the treaty, convention or covenant, although judges may depart from the strict text to extend proportionality to non-derogable rights. For instance, freedom from torture and inhuman or degrading treatment or punishment, and the right not to be held in slavery or servitude, are famously ranked as non-derogable rights, which implies that the respondent state cannot interfere with them under any circumstances. In contrast, derogable rights are those that may be interfered with under very specific conditions. Indisputably, ESC rights belong to the second category, which makes them particularly interesting for proportionality analysis. The proportionality test provides judges and adjudicators with a specification of these conditions, the cumulative fulfilment of which leads courts to grant leeway to the state in a case-specific, ad hoc fashion.
There are multiple formulations of the proportionality test, depending on the constitutional or human rights system under consideration. It is generally taken to involve four distinct steps. Once an interference has been found prima facie, the court or adjudicator examines whether the policy or law interfering with one or several rights pursues a legitimate aim – which typically refers to a range of collective or public goods, such as national security, public health or morals. It is important to note that this first step is hypothetical: the courts are examining ‘whether there are any interests which are candidates for justifying the interference in the sense that it is not entirely implausible that they will at least be rationally connected to the policy’ (Möller Reference Möller2012: 181). Second, it scrutinizes whether there is a rational connection between the law and the pursuit of the aim – that is, whether applying that law actually brings the realization of the aim closer. Third, the law must be the least invasive for the individual, but at least as effective as the alternatives. Fourth, the interference should respond to a pressing social need and must not impose a disproportionate burden on the right-holder – a step often called proportionality stricto sensu or balancing. At this stage, there is proper weighing between the aim and the interference. Again, if the court or adjudicator establishes that the respondent state has passed each and every step (failing one is sufficient to find a violation), it may find that there was no violation of this or that right. In practice, respondent states usually pass the first step of the legitimate aim easily – if not with one of the listed aims, courts and bodies will accept that the aim was the ‘protection of the rights of others’, which amounts to aggregating the burden inflicted on one or more individuals as a result of the interference. Adjudicators will then weigh that aggregate against the interference with rights and their underlying interest(s) – for instance, whether deporting an illegal migrant has a disproportionate impact on the migrant’s right to family life. The literature has made clear that ‘balancing’ or ‘weighing’ are misnomers: very often, it is not possible to balance or to weigh at all, simply because the impact of one or the other interest at stake is itself not measurable in a way that it could be balanced against the competing interest. Endicott (Reference Endicott, Huscroft, Miller and Webber2014: 316) explains this point by stating that
for any given articulation of the public purpose (but there is none in these cases), there would be no rational basis for deciding just how serious the effect must be on the person’s family life before it would outweigh the public benefit of the deportation.
The indeterminacy of the balancing test is, however, less prevalent in the context of the ESC rights. As we shall see later, the Committee is not required to conduct the balancing stricto sensu and concentrates mostly on the steps of legitimate aim and necessity, which alleviates the indeterminacy that comes with the balancing phase.
Proportionality, reasonableness and justifiability
Proportionality is about the most reasonable option, not the ultimately correct one. That is why we opt for ‘attenuate’ rather than ‘solving’ the conflict crystallized in the proportionality test: did the state interfere while pursuing a rights-protective collective aim? Did the interference actually serve that aim? And was there any less-invasive measure available that would have achieved the same outcome? This helps us to understand that the typical questions of proportionality structure and discipline rather than determine or solve the outcome of a court or adjudicator’s review process. It is content independent in the sense that, irrespective of the actual variables under scrutiny, the court or adjudicator must address these questions. Describing these various steps helps to introduce the claim that proportionality can operate as a procedure in the sense specified in the last section of the article. Earlier, we established that content-independence has to do with the features of the institution, not the content of its directive. Rather, the content-independent function of proportionality is not to solve this or that particular conflict, but instead to attenuate any of these conflicts and tensions by providing a rule-based structure of reasoning. In the words of Kaï Möller (Reference Möller2014: 222), ‘properly understood, proportionality does not provide, nor does it claim to provide, a ‘shortcut to moral truth’ – such shortcuts do not exist. Rather, its value is that it helps judges identify and address all morally relevant considerations when resolving a rights issue’. As we further explain in the next section, this role can apply to the context of ESC rights, even if the nature, content and justification of these rights are not identical to civil and political rights. Once courts and treaty bodies establish general standards of protection, which the Committee has been pioneering in the field of ESC rights through its doctrine of ‘minimum core obligations’, proportionality can serve the same function of structuring the body’s reasoning.
It is important to explain that defending the benefit of proportionality against interpretation along the lines just outlined finds some support among proportionality theorists in constitutional law scholarship. Surprisingly, the discussion among international lawyers rarely uses this resource. Generally, there is wide consensus that the rise of proportionality in constitutional interpretation reflects a structural evolution of the very purpose of interpretation from a ‘culture of authority’ toward a ‘culture of justification’ (Cohen-Eliya and Porat Reference Cohen-Eliya and Porat2013) that requires state actions not only to be grounded in distinctively legal and positive authority (typically, through canons of interpretation), but also shown to be justifiable from a distinctively normative standpoint. The proportionality test introduces the analytical framework (or the procedure) to fulfil that properly justificatory function. A variety of comprehensive normative approaches inform the conduct of the test in that literature. Mattias Kumm (Reference Kumm2010: 123), for instance, uses justifiability as the overarching criterion:
proportionality based judicial review institutionalizes a right to justification that is connected to a particular conception of legitimate legal authority: That law’s claim to legitimate authority is plausible only if the law is demonstrably justifiable to those burdened by it in terms that free and equals can accept.
By justifiability, Kumm essentially means non-perfectionist and non-consequentialist reasons; this crucially reflects autonomy concerns. Möller (Reference Möller2014: 182) also understands proportionality as ultimately addressing whether a conflict between issues of autonomy has been resolved in a reasonable way: ‘the proportionality test is the doctrinal tool which enables them to carry out this task’. One may object that autonomy is a primarily individualistic notion, which raises concerns about it being parochially and predominantly Western. However, political autonomy also attributes value to the collective decision of how to realize the public good in a particular society. In addition, the context of ESC rights interestingly adds nuance to the matter by giving more deference to the state in determining the measures necessary to (progressively) realize the right in question.
Proportionality in the field of ESC rights
The relatively uniform use of the proportionality test in the field of civil and political rights should not obscure the specificities of proportionality in the field of ESC rights in general and in the practice of the Committee in particular. Not only are the test’s wording and structure less systematized than in the former field, but the very content of rights is significantly more malleable. The lack of a clear and structured test is reflected in the fact that the Committee has a slightly different task in terms of reviewing potential violations as opposed to, say, the Human Rights Committee reviewing the ICCPR or the European Court of Human Rights (ECtHR) reviewing the European Convention on Human Rights (ECHR). That is because, under the Committee, the Committee has to review two types of obligations: whether states have taken the necessary actions that the Covenant requires (what is commonly called ‘minimum core obligations’) on the one hand; and whether states have progressively realized these rights on the other:
Each State Party to the present Covenant undertakes to take steps, individually and through international assistance and co-operation, especially economic and technical, to the maximum of its available resources, with a view to achieving progressively the full realization of the rights recognized in the present Covenant by all appropriate means, including particularly the adoption of legislative measures. (Article 2(1) ICESCR)
There is a clear ambiguity in this proposition: states have to take reasonable steps to the maximum of their resources. This is explained by the fact that the content and scope of ESC rights are not defined prior to proportionality review and/or based on established interpretive principles. Drawing on the South African context, Young (Reference Young, Jackson and Tushnet2017: 21) explains that ‘by integrating the analysis of a right’s progressive realization, within the state’s available resources, in the same step as defining the right, there is no standalone content, inflated or otherwise’. The Committee explicitly acknowledged that ‘ESC rights will generally not be able to be fully achieved in a short period of time’ (General Comment No 3, supra 28, para 9). It should also be noted that taking steps to the maximum of resources can take the form of both process (budgetary measures, legislative measures) and outcome measures.
Correlatively, there should not be retrogressive measures, as this would similarly impede progressive realization (General Comment No 14, supra 28 at para 32). Further, and reflecting a general and structural pattern across universal and regional human rights systems, states are free to design and implement the measures appropriate to meet this threshold (subsidiarity). This is further reinforced in the context of ESC rights, given the context-dependent character of the measures for (progressively) realizing these rights. The Committee, however, reviews the effectiveness of these measures: ‘the ultimate determination as to whether all appropriate measures have been taken remains one for the Committee to make’ (General Comment No 9, supra 48 at para 4).
From this brief incursion, one can observe that while proportionality review is weak in the ESC context – which is why the Committee itself calls it a ‘reasonableness’ test – its basic function can remain the same for civil and political rights, namely to structure and discipline the review process. The reference to ‘reasonableness’ is found in Article 8 of the Optional Protocol to the Covenant, which precisely implements the individual communication mechanism. The relatively scarce literature on proportionality and social rights also attests that proportionality is ‘a method that can protect the substance of rights not by excluding them from conflict, building a wall around them, but by ensuring that this conflict is conducted in accordance with the constitution’ (Contiades Reference Contiades2012: 212). This is particularly important in the context of ESC rights and the core minimum obligations. It is in this context that proportionality – and its legitimacy-enhancing role – will best operate, in particular with respect to ‘non-retrogressive’ cases where core minimum obligations are at stake. These cases depict a situation where there is a prima facie violation of the provision – very much like in the context of civil and political rights. However, the ‘progressive realization’ cases, which concern the moving up of these standards, often targeting the worst off in society, are not necessarily excluded from proportionality analysis as we understand it. This can also be apprehended through our distinction between individual and collective autonomy. That is, these cases highlight a greater emphasis on collective autonomy where deference is conferred to the state so as to determine which measures are best suited to realize the core content of the right progressively. Still, proportionality considerations matter: the Committee still examines whether there is a rational connection between the right at stake and the measures taken by the state, and whether these measures are effective – hence reasonable.
One may therefore remain agnostic as to the ultimately different nature, content and scope of (justiciable) ESC rights compared to civil and political rights, while arguing that the role of proportionality across the landscape of rights remains the same. In order to see this better, it is helpful to look at the express limitation clauses in Article 4 and 5 of the Covenant and how the Committee has to apply them. Most importantly, these articles establish that only limitations pursuing the aim of promoting the welfare of society are acceptable – in other words, only measures that are ESC rights-enhancing in an aggregated sense can be accepted. Article 4 requires that any restrictions on rights must be ‘determined by law’ as well as ‘compatible with the nature of these rights and solely for the purpose of promoting the general welfare in a democratic society’. As we shall see below, requiring that restrictions to rights pursue the general welfare (and serve the ESC rights of others as an aggregate) conceptually corresponds to the legitimate aim of the ‘rights and freedoms of others’ found in adjudicating civil and political rights. This also means that there is no distinct step of balancing that presupposes having a clearly established right. As we shall we see shortly, the test focuses on the following steps: legitimate aim, rational connection and necessity. And that is precisely why the test is labeled ‘reasonableness’. We draw out the implications of this more limited test in the concluding part of this section.
To provide just one example of how ‘reasonableness’ operates in the practice of the Committee, let us survey the 2013 case of Mohamed Ben Djazia and Naouel Bellili v Spain. In this case, the Spanish government evicted Mohamed Ben Djazia, Naouel Bellili and their two minor children from the home they had rented in Madrid after the contract expired. The government also failed to provide a plan for emergency housing while Mr Ben Djazia had repeatedly applied for social housing for more than a decade. The applicants claimed that their right to adequate housing (Article 11 of the Covenant) was violated. As explained above, proportionality considerations start where the state needs to justify the measure of eviction. On this point, the Committee found that the state did not evidence that Mr Ben Djazia failed to meet the criteria for social housing, but that, ‘rather than actively searching for housing, simply expected that it would be provided by Social Services, even when the eviction was imminent’ (para 4.5). In the Committee’s view, however, the question immediately becomes whether there was an ESC right-enhancing aim underlying the interference:
The Committee considers the State party’s arguments as insufficient to demonstrate that it has made all possible efforts, using all available resources, to realize, as a matter of urgency, the right to housing of persons who, like the authors, are in a situation of dire need. For example, the State party did not explain that denying the authors social housing was necessary because it was putting its resources towards a general policy or an emergency plan to be implemented by the authorities with a view to progressively realizing the right to housing, especially for persons in a particularly vulnerable situation (para 17.5).
Here one should emphasize the scope of the Committee’s review. The Committee did not scrutinize the (democratically valid) budgetary decisions that determined the level of resources available at the moment of the interference, which reflects the Committee’s responsiveness to political autonomy. Rather, it argued that the state failed to explain how the ground for interference served the general welfare of society, which clearly echoes the legitimate aim step which, as Thorburn explains, determines at least that we are dealing with ‘an entity that has the proper standing – as a public authority pursuing a legitimate public purpose – to make a claim of justification’ (Thorburn Reference Thorburn, Dyzenhaus and Thorburn2016: 308). This deficit grows with the fact that the case included children, who are standardly taken to be vulnerable persons under human rights law – ‘a baseline assessment of the gravity of certain laws and policies on the most vulnerable, including the most economically vulnerable’ (Young Reference Young, Jackson and Tushnet2017: 19). Also, and relatedly, the Committee established that the state did not provide the applicants with an alternative housing option since this would have resulted in splitting up the family by placing members in different shelters:
if the authors had accepted this offer, the family would have been split up, in violation of the State’s duty to grant the greatest and widest possible protection to the family, as the foundation of society, in keeping with article 10(1) of the Covenant. In this regard, the State party has not explained to the Committee why no other options were available to the authors. (para 17.8)
This part of the decision echoes the step of necessity that forms part of the classical test: the state could have opted for another, least invasive option when interfering with the right.
Having explained how proportionality review can operate as a procedure, we are now able to specify the last prong of our core argument. Proportionality can operate as a content-independent but rule-based procedure that can best attenuate the tension between personal and political autonomy. In that respect, we are not deviating from the core notions that permeate the interpretive process as a whole: whether it is interpretation stricto sensu or proportionality, we have shown that the same normative considerations, grounded in autonomy, underpin the various parts of the treaty body’s review process. The argument is not to abandon one for the other, or to replace one with the other. Rather, the argument is that, considering its procedural pedigree, proportionality is better suited than interpretation to offer the kind of reasons expected from a legitimate authority. Drawing from the case just reviewed, the tension between the right to housing (personal autonomy) and the overall welfare policy of the society (political autonomy) could very well be built into an exercise of interpretation, which itself may require conducting proportionality analysis. The proportionality test – usually located at the end of the review process – merely systematizes and attenuates the tension between personal and political autonomy. The prevalence of proportionality indicates that personal and political autonomy are the ultimate and normative building blocks of human rights interpretation. Proportionality requires that, in the event that limiting personal autonomy is necessary, the state has ensured that such limitation is best grounded in political autonomy and minimally restricts personal autonomy. Again, that procedural operation is content independent and can apply in each and every case reviewed by the Committee.
V. Conclusion
This article has sought to bring two seemingly distinct disciplinary approaches to the interpretation of the Committee into closer dialogue through the shared notion of legitimacy. Legitimacy is widely used across disciplines, and therefore usually provides a useful point of departure to critically reflect on authoritative institutions. One disciplinary approach is the conventional toolbox that the Vienna Convention has usually offered – and this is explained by the fact that the Committee is a creation of international law. The article began by reconstructing how international lawyers have envisioned applying this toolbox to human rights law in general and to the practice of the Committee in particular. This reconstruction led to the view that applying the Vienna Convention’s canons can best enhance the perception on the part of the relevant interpretive community that the interpretive outcome of a treaty body’s review is worthy of support. On this view, legitimacy is sociological and content-dependent. Of course, it is not surprising to find that the Vienna Convention almost always constitutes the starting point of a principled reflection on methodology in human rights interpretation, as it is the positivized starting point of international interpretation tout court.
Yet, as Scheinin (Reference Scheinin, Andreassen, Sano and McInerney-Lankford2017: 21) puts it in the context of Articles 31–32 of the Vienna Convention, ‘these two obvious steps represent just the beginning of the journey’. We have suggested that there might be a destination envisioned, but no roadmap showing how to get there. The second step of the article critically engaged with the predominant disciplinary approach to interpretation. We followed two lines of thought. The first was that sociological legitimacy is not all there is to legitimacy, since the sociological account implies disregarding the reasons subjects have or should have to perceive an authority as legitimate. Normative legitimacy fills this gap by asking what reasons subjects should have in the first place. The second line of thought is that the prevailing Vienna Convention-informed methods that oscillate in the Committee’s practice in fact crystallize the prevalence of two moral notions: personal and political autonomy. Such notions cut across human rights systems – for example, political autonomy through the margin of appreciation in the practice of the ECtHR. As these notions are conceptually distinct and may clash, it is difficult to establish a clear procedure for human rights interpretation based solely on them. Indeed, what further criteria help to determine when one type of autonomy should prevail over the other? One could even think that the very fact that these notions are morally significant explains their prevalence at the legal and adjudicatory levels. From the standpoint of legitimate authority, one should strive to find a clear procedural path that could generate content-independent reasons.
The third and final step was to develop another principled approach to interpretation that not only builds upon the findings of the first step, but also offers amelioration in procedural terms. We hope to have shown that legal interpretation and political theory have far more in common than is usually thought. Defending this alternative approach required first adjusting the level of argument to normative considerations, which necessitated an explanation of how to address the indubitably difficult and recurrent tension between personal and political autonomy in principled and procedural terms. We argued that the proportionality test is better suited to address the very problem that the Vienna Convention-based approaches pose, namely to attenuate – but not solve – the tension between personal and political autonomy. Proportionality requires that, in the event that limiting personal autonomy is necessary, the state has ensured that such limitations are grounded in political autonomy and restrict personal autonomy as minimally as possible. Again, this procedural operation is content independent – it is not necessarily connected to the actual article of the Covenant under scrutiny. This article therefore paves the way for an approach that maximizes the prospect for generating the kind of reasons needed to confer legitimate authority to UN treaty bodies, in particular the Committee.
Acknowledgements
We would like to thank participants in the workshop ‘Within the realm of the possible: Reforming the UN treaty bodies’ individual communication mechanisms’ at the Hertie School of Governance, Berlin, in November 2019, where a draft of this article was presented. We are particularly grateful to Andreas Zimmermann, Basak Cali, Andreas Von Staden for very helpful remarks. Special thanks to Andreas Follesdal for written comments on the draft. Further, we would like to thank colleagues at the University of Glasgow School of Law’s Staff Seminar where the paper was also presented in January 2020. Many thanks to George Pavlakos, Lea Raible, Tria Gkouvas, Toni Marzal and Jill Robbie. Finally, we are very grateful to two anonymous reviewers of the journal for extremely helpful and constructive comments.
Funding Statement
This article was partly supported by the Research Council of Norway through its Centres of Excellence Funding Scheme, project number 223274 – PluriCourts, The Legitimacy of the International Judiciary.
