INTRODUCTION
In chapter one, I discussed the three conceptual assumptions underlying the default vertical approach to constitutional rights. However, it is almost trite to observe that contemporary conditions do not resemble those under which the default vertical approach emerged.
The emergence of ‘new power centres and mechanisms of oppression’[341] has necessitated departures from default verticality and triggered the application of constitutional rights to relationships involving non-state parties. In chapters two and three, therefore, I discussed some of the archetypal ways in which courts have accomplished this. I argued that each of the dominant archetypes is unsatisfactory, as it remains bound to one or more of the three underlying assumptions of default vertical- ity: sovereignty, abstract freedom and individual responsibility. In concluding chapter three, I argued that, nonetheless, two important insights from the existing archetypes could form the basis of an alternative approach. The first is that default verticality leaves certain crucial human interests unprotected, in cases where the state is uninvolved. The perspective of the rights bearer, thus, ought to be central to any approach that departs from the default (Jean Thomas calls this the ‘vulnerable valuable interests’ diagnosis[342]). The second, related insight is that the background relationship between the rights bearer and the (putative) rights violator is an important factor in determining whether and, if so, how the rights framework is to be applied to the impugned conduct or practice. In chapter four, I explored two contemporary models - proposed by Johan van der Walt and Jean Thomas - that use these insights (to different degrees), to articulate versions of bounded direct horizontality that are no longer tied to the assumptions of default verticality. I argued that individually, Van der Walt and Thomas’s accounts represent promising approaches to the problem, but also fall short in certain respects. However, Thomas’s focus on the relational structure between the private parties and Van der Walt’s insight about understanding private relationships contextually are both integral elements in a successful account of bounded horizontality. In this chapter, drawing upon those insights, I will develop the institutional model of horizontality.The institutional model of bounded horizontality examines (allegedly rights- violating) acts or practices between two parties in the context of the background institutional relationship between them (ie the parties' relative locations within social, economic or cultural institutions). It argues that where there exists a difference in power between the parties, enabling one to violate certain constitutional rights of the other, and where such difference flows from the parties' relative locations within the institution, then those rights should be horizontally enforced.
Therefore, an account of the institutional approach requires the following: first, a working understanding of what constitutes an ‘institution' for our purposes; and secondly, the link between the institutional differences in power on the one hand and the nature of the rights and remedies that are enforced on the other.
Let me, at the outset, offer a tentative articulation of the concept of an ‘institution', which I will attempt to refine in the course of this chapter. I begin by noting that there is extensive philosophical and sociological literature that discusses the idea of an institution.[343] Drawing from these accounts, I suggest that there are three salient features that constitute an ‘institution' in general. First, there must be convergence of behaviour, by multiple actors, over a sustained period of time (creating what might be understood as ‘norms', ‘patterns of conduct' or ‘conventions'). Secondly, an institution is not simply an aggregate of individual converging practices, but has a ‘coherence' and ‘unity' of its own. And thirdly, an institution assigns roles, functions and powers to individual actors, often in a hierarchical fashion.
Derek Layder's articulation perhaps sums up the concept best: a set of ‘(reproduced) social relations which have an ongoing, organised and relatively enduring quality to them'.[344]A quick caveat: while these three features indicate that an institution is not reducible to individuals and individual acts, it is nonetheless constituted by individuals. Thus, when, in the succeeding pages, I discuss the concept of impersonal institutional power, it is important to keep in mind that this does not entail anthropomorphising or reifying abstract social structures. As succinctly observed by Moishe Postone, institutional power can be best understood as the ‘domination of people by abstract social structures that people themselves constitute';[345] or, to put it even more clearly, ‘power is implicated in the enduring structural relations that characterize a society and is exercised by intentional human agents who participate in these relations'. (emphasis added)[346]
Now, while these three characteristics are sufficient for us to intuitively grasp the meaning of the term ‘institution', there are two additional features that are necessary in considering which institutions are relevant for the purpose of applying a horizontal rights framework. The first is that an individual does not ordinarily have the ability of exiting from an institution. So, for example, a particular school or hospital might be an ‘institution' in the sense that sociological theory understands it. However, it is not an institution for the purposes of the horizontal rights framework. On the other hand, as we shall see below, race, or caste, does constitute an ‘institution' in the relevant sense.
The second, linked feature is that the hierarchies of domination and subordination created by and through institutions are permanent, enduring and - again - features that a particular individual cannot ordinarily overcome. This point is articulated by Derek Layder, who uses the term ‘structural power', but broadly refers to the same phenomenon that this book calls ‘institutional power':
structural power is relatively independent of agency, but that agency, for example in the form of exercises of power, is always subject to the influence of structural power.
Power structures are relatively independent of agency in the sense that although they are social products, and thus the outcome of human action at specific historical junctures (in the form of individual or collective struggles, conflict, diplomacy, reform etc.), once established, such structures develop over time to attain an existence which sediments them as constraints on future agency. (emphasis added)[347]Thus, the primary concern of the institutional approach revolves around differences in power that exist within and because of institutions, how these differences in power enable some private parties to violate the rights of others and how that can be mitigated or addressed through a horizontal rights framework. However, it is not any and every difference in power that attracts the horizontal rights framework. The individual's inability to exit from, or overcome, these power differences because of the manner in which they are embedded in certain kinds of institutions is also a key determining factor in the application of horizontal rights.
The concept of the institution, therefore, serves as a boundary condition for triggering the application of horizontality. At this stage, let me flag two possible preliminary objections: vagueness and width. I shall address both objections briefly here, with the caveat that a more complete response will take the form of demonstrating how the institutional approach will work in practice, in this chapter as well as in part II of this book.
The first objection might be that the concept of the institution is too vague for practical application (and, relatedly, creates challenges for the rule of law and legal certainty if courts are called upon to determine such definitional questions on a case-to-case basis). To this, a brief answer may be made: I believe that the elements of an institution, as outlined above, are no more vague than concepts that courts do apply on a regular basis - such as ‘personal liberty’ and ‘equality’ - and whose content they flesh out through precedent, incrementally and over time.
Indeed, Langille and Davidov make a very similar point about the labour law category of the ‘employment contract’ (which I shall discuss at some length in the next chapter). Responding to criticism that courts have failed to adequately define what constitutes an employment contract, Langille and Davidov rely on Wittgenstein’s concept of family resemblances to argue that ‘not all definitions have to be of the analytical variety’.[348] Indeed, ‘institutions’, as set out above, do fall within the concept of ‘family resemblances’, which - Wittgenstein notes, in the context of the concept of ‘games’ - are best understood as ‘network(s) of similarities overlapping and criss-crossing’.[349] Langille and Davidov go on to point out that in legal reasoning, context gives colour to the meaning of terms.[350] The task of this chapter, and the succeeding ones, will be to provide that context while demonstrating how the content of the concept of an institution can be moulded through adjudication.With respect to concerns around the rule of law and legal stability, as we shall see in this chapter as well as part II, the institutional approach does not seek to supplant private law or legislation, or to turn all legal disputes into ‘higher constitutional law review’. The institutional approach acknowledges that there can be reasonable differences when it comes to the specific contours of a right, which are to be resolved, in the first instance, by legislation. The institutional approach steps in when there are gaps, or when legislation is evidently insufficient to protect rights. One of its important roles is to inform judicial approaches towards interpreting existing legal frameworks. We shall see this in practice in part II, when considering the application of the institutional approach to specific legal problems.
Secondly, in addition to the problem of vagueness, it may be objected that, notwithstanding the definitional stipulations set out above, the concept of the institution is a wide one, and takes in a broad swathe of human relationships.
This is true. The initial threshold for attracting the institutional approach - ie the existence of an institution - is a relatively low one. However, a few things ought to be noted about the width of the concept. The first is that while the existence of an institution is a necessary condition for horizontal rights application, it is not a sufficient one. It still needs to be shown that the (private) duty bearer exercises power over the rights holder, that the difference in power flows from their relative institutional positions, and that it enables the former to violate the rights of the latter (indeed, as we shall see, the two enquiries - the existence of the institution and the existence of structural power hierarchies - are often bound up with each other, making the threshold not quite so low as it may initially seem).For example, in the next chapter, it shall be argued that the labour market is an institution for our purposes. It does not follow, however, that every constitutional right is ipso facto applicable between the worker and the employer. Rather, the institutional approach limits itself to a set of discrete rights and remedies, flowing from the power differential between the parties, and the horizontal application of each of these rights needs to be separately justified.
Indeed, such an approach, where the necessary-but-not-sufficient threshold for triggering the application of rights is a low one, is familiar to constitutional adjudication.[351] It is important to reiterate that the satisfaction of the threshold condition, ie the existence of an institution, is only a first step. The identification of power differences, the rights at issue and, of course, the question of whether remedies are justified in a specific case (taking into account the rights of the other party as well) all remain to be answered.
Secondly, concerns around the width of the concept are mitigated when we consider the role that the institutional approach will play in adjudication. As pointed out above, the institutional approach exists to complement existing law, not supplant it. The width of the concept, therefore, does not necessarily translate into width of adjudication, where a vast new area of human relations is brought within the scope of regulation for the first time. This will be demonstrated in the final section of this chapter and in part II.
Finally, it ought to be reiterated that the width of the concept does not mean that it lacks analytical purchase, or that it cannot serve a useful role in adjudication. Once again, this chapter and part II will be dedicated to showing how the institutional approach - with the concept of the institution as an integral element - is both theoretically sound and practically useful.
To sum up my overview of the institutional approach: an ‘institution’ for the purposes of the horizontal rights framework has a conglomeration of characteristics. Consequently, it might not always be possible to define and identify institutions in advance. Rather, the presence - and combination - of the characteristics discussed above will often serve as evidence that practices that initially appear to be individualised rights violations are actually embedded in an institutional relationship, and require the application of horizontal rights. Indeed, as I shall argue in this chapter, the concrete resolution of both issues - the articulation of an ‘institution' and the question of rights and remedies - must be context- and jurisdiction-specific. My task here is to develop the contours of the approach and illustrate its working in one specific jurisdiction.
I attempt to accomplish this in the following manner. I begin in section II with some context setting, identifying a broader legal and philosophical tradition that explores the nature of institutional power, and the relationship between questions of justice on the one hand and the role of institutional power on the other. The institutional approach, I shall show, takes this enterprise forward by drawing a conceptual connection between institutional argument and direct horizontality (section II).
I then move on to doctrine. Section III of this chapter demonstrates how, in some of the jurisdictions that we have considered in detail in chapters two and three - the USA, South Africa and Germany - judges have, at various times, articulated the beginnings, or proto-versions, of the institutional approach. The purpose of sections II and III, therefore, is to show that the ideas underlying the institutional approach have not come out of the blue (either in legal theory or in legal doctrine), but have been part of an undercurrent in judicial discourse across jurisdictions (section III).
In the heart of this chapter, I shall then consider the judicial development of the institutional approach under the Indian Constitution in some detail. The Indian Constitution has three relevant horizontal rights provisions - Article 15(2) (non-discriminatory access to shops), Article 17 (prohibition of ‘untouchability') and Article 23 (prohibition of ‘forced labour'). I shall look at three court cases where an allegedly rights-violating practice was at issue, but which did not appear to fall within a strict understanding of these provisions (discrimination in schools, gender-based exclusion of entry to temples and non-payment of minimum wages, respectively). However, by understanding the practices in question as both constitutive and reflective of institutionally mediated differences in power (the institutions in question being caste, patriarchy and the labour market), and by understanding the constitutional provisions as targeting precisely these institutions, the Indian Supreme Court brought them within the ambit of the horizontal rights framework. The Court's interpretation, I shall argue, is justified in the context of Indian constitutional history, where the power of economic, social and cultural institutions has long been viewed as a threat to rights, and resistance to which has been framed in the language of rights. I will end this part of the chapter by reflecting on how these decisions help us to understand the application of the institutional approach in practice (section IV).
Having grounded the institutional approach in both judicial doctrine and legal philosophy, I then move on to consider two issues: first, the ways in which the institutional approach succeeds where the alternatives discussed in chapters two and three failed, by successfully jettisoning the underlying assumptions of default verticality. The institutional approach does not depend upon a unitary idea of sovereignty. Rather, it relies upon an embedded concept of freedom, which takes into account the background socioeconomic situation of the parties. It also understands that in situations where there exist abiding and pervasive differences of institutional power between parties, they should not be treated as existing on the same ‘horizontal’ plane, equally responsible for shaping the contours of their relationship (section V).
Secondly, I shall address the relationship between the institutional approach and existing regimes of private law, by examining three possible interfaces between the two. The first, and most common, situation is where there exists a constitutional right that is to be applied horizontally, as well as an existing legislative or common law regime. The constitutional right, in such a situation, complements and informs the interpretation of the legislative regime, and in certain circumstances acts as a touchstone upon which the constitutionality of private law may be tested (if required). The second situation is where a private law regime that imposes obligations upon private parties is challenged by those very parties, on the basis that it curtails their (primarily autonomy-based) rights and freedoms. The institutional approach, through its horizontal rights provisions, will play a role in determining whether the legislature has successfully ‘balanced’ the competing rights at play in regulating a set of private relationships and practices. The final situation is where there is a legislative vacuum, or where private law simply does not exist. I shall show that in such cases, where possible, constitutional right can be enforced directly (section VI).
II.