Among the numerous national constitutions adopted throughout the twentieth century, that of South Africa stands out for the transformation to which it aspires. Frank Michelman goes so far as to describe it as “postliberal.”Footnote ¹ In the early 1990s, the long-dominant National Party, which had perpetrated Apartheid, and the African National Congress (ANC), founded in response to the racist system, approached the negotiating table to reconstitute the country. Horizontal application fit naturally in this transformative project. As Apartheid had permeated all spheres of life, so too would such values as dignity, equality, and freedomFootnote ² need to suffuse even private life. In this understanding, the constitution could not but call for parity between public and private spaces and hold at least some private actors responsible for these new commitments. A certain egalitarian bent consequently characterized much of the discourse, as distinct from more traditional discourses that emphasize classical political rights over socioeconomic rights, and vertical relationships over horizontal.

Whereas Germany’s adoption of horizontal application came through the Federal Constitutional Court’s decision in Lüth, much of the groundwork for horizontal application in South Africa was laid in the deliberations preceding the adoption of the Constitution. Ultimately, the text of the Final Constitution itself provides for broad application of horizontality. Indeed, Nick Friedman describes horizontality as practically necessary given the nature of the constitutional project. He states:

Republican principles can easily be read into this account of horizontality and the South African constitutional project – the recognition of duties vis-à-vis community, moral duties of one to the other, a shared future. Also striking is the way Friedman attributes dignity (a concept with a long republican pedigree) to duty-bearers and rights-bearers alike. There is a sense here in which the maintenance and execution of constitutional duties presupposes membership in a common project. In this light, the South African Constitution and the large-scale malaise it aims to ameliorate make clear space for horizontality and a republican interpretation of horizontality at that.

What specific role, then, does this chapter play in the book’s project to interpret the practice of horizontal application through the lens of republican political theory? First, that the South African project aimed at thoroughgoing transformation in the way and to the degree it did sets it apart from other constitutional contexts this book considers. Of course, Germany’s Basic Law can and has been described as transformative.Footnote ⁴ The German Federal Constitutional Court accordingly adopted horizontal application on the basis that constitutional principles create an “order of objective moral and legal principles” that “radiates” to both public and private spheres of life, as described in Chapter 5.Footnote ⁵ However, the Constitutional Court set limits on horizontality in the realm of equality rights, limits which only began to be broached some fifty or sixty years after Lüth was decided. In contrast, issues directly connected to equality, such as housing and education, have gained increasing salience in South African constitutional politics. While both the German and South African Constitutions have been dubbed transformative, the South African Constitution’s explicit and broad provision for horizontal application arguably points toward a different conception of transformation and thus provides different ground on which to read horizontality as republican. From early initiatives of the ANC to later decisions of the Constitutional Court, South African constitutionalism proves quite receptive to republican interpretations and often echoes republican values.

However, some scholars and political observers take issue with this characterization of the South African Constitution as transformative. Emile Zitzke argues that what are often seen as transformative aspirations in South African private law and human rights do not ultimately address the problems arising from the country’s colonial past. Specifically, the new order’s Eurocentric roots inhibit full decolonization, a level of transformation that might be possible if African concepts were employed as meaningful reference points instead.Footnote ⁶ Perhaps more critically, Joel Modiri emphasizes how, as a product of negotiation and compromise, the new constitution actually preserves the interests and powers initially secured through racial oppression.Footnote ⁷ The space of this book does not allow for the full discussion these critiques merit. At the same time, the fact that the book’s project is primarily analytical and theoretic, rather than normative, allows some liberty to take the South African Constitution at face value – to consider the transformation it does imply, though a different and more transformative project may be conceivable. As this book maintains, horizontality exists within the liberal constitutional milieu even as it marks a kind of digression from certain renderings of that milieu. At the same time, it seems safe to say that, as a descriptive matter, the South African Constitution does mark some real change from the prior order, though that change falls short of certain normative assessments.Footnote ⁸

Even within the scope of this project, however, fears that the new order is not sufficiently transformative may be borne out in the way constitutional actors reconcile horizontal application with continuities from the prior order. Although key framers were clear, and the constitutional text explicit, in providing for horizontality, certain factors, both persisting from the prior system and emerging in the new, seem to countervail against this practice in particular instances. Indeed, differently situated constitutional actors have related differently to horizontality. As institutions and actors bring their various interests to the constitutional political table, it is to be expected, even in what is arguably an optimal context, that different actors may be more or less disposed to develop horizontal application. This chapter thus follows the same dynamic as previous chapters in finding more compelling republican interpretations in certain moments and spaces than others. While we ascertain their full weight in the light of South African history, law, and politics, these considerations offer important insights into horizontality more generally as well.

This chapter proceeds by uncovering republican interpretations of horizontality in various South African constitutional debates, paying particular attention to founding debates and the early landmark case, Du Plessis v. De Klerk, before turning to more recent developments. This chapter also addresses those factors that, in South Africa, have seemed to countervail against horizontal application. They include persistent priorities of the formerly dominant National Party, oppositional institutional interests not unlike the early debates between private and public lawyers in Germany, and a highly formalist legal culture. As in foregoing chapters, attention to these factors is not intended to set up a developmental argument, but to set the full contextual scene on which to analyze constitutional discourses surrounding horizontality and bring the republican lens to bear. Insofar as such factors are inseparable from, and even constitutive of, constitutional politics, so too will they bear on a republican interpretation of this constitutional practice.

Foundations for a Republican Interpretation: Ubuntu and Democracy

While those involved in the constitutional deliberations would not necessarily have described the constitutional project in terms of civic republicanism, communalistic elements have long figured into South African political culture. Specifically, the concept of ubuntu dates back to precolonial times, encompassing a rootedness in community and inescapable connectedness to neighbor. Praeg describes it as “a political economy of obligation”Footnote ⁹ and a “logic of interdependence.”Footnote ¹⁰ Markedly different from political narratives focusing on the individual, ubuntu speaks to an interconnectedness and capacity for friendship among people as human beings. The Constitutional Court has made some use of the concept in its judgments, with Justice Mokgoro describing it as “an idea based on deep respect for the humanity of another”Footnote ¹¹ and “part of our rainbow heritage.”Footnote ¹²

Some object that ubuntu is hollowed out by attempts to apply it to the liberal structures of South African constitutionalism.Footnote ¹³ The Constitution simply does not admit space for ubuntu’s full-blown communalism, the argument goes, nor can ubuntu be employed to support such typically liberal structures as rights, courts, and the like. The very fact that courts are entrusted with enforcing the Constitution poses an obstacle in that they tend to understand conflicts as clashes of rights in contrast with the community-mindedness that ubuntu espouses.Footnote ¹⁴ Others have been more optimistic with respect to ubuntu’s contributions to South African constitutionalism.Footnote ¹⁵ We see such optimism in Justice Sachs’s judgment in Port Elizabeth Municipality v. Various Occupiers,Footnote ¹⁶ a case in which a white landowner attempted to evict extremely poor black families who had lived on the land for years. Drawing on ubuntu, Justice Sachs required the parties to “engage with each other in a proactive and honest endeavor to find mutually acceptable solutions” before resorting to such adversarial measures as eviction. He explained:

The Court would rely on similar reasoning in such later cases as President of South Africa v. Modderklip,Footnote ¹⁸ discussed later, actually shying away from the language of rights in favor of a call for neighborliness.

For some, the concept of ubuntu has thus served as a counternarrative or supplement, aligning South African constitutionalism with the maxim that “[a] person is a person through other persons.”Footnote ¹⁹ Put differently, ubuntu potentially brings something new to the constitutional table in the very act of bringing something quite old, not unlike this book’s republican interpretation of horizontality. What Chapter 2 argues to be republican features of horizontal application bring to constitutionalism an understanding of the polity conceived as a whole, as well as an understanding that citizens have duties to one another. In a similar way, ubuntu conjures a conception of constitutionalism that transcends the individual, evinced in Justice Sachs’s judgment in Port Elizabeth Municipality. He explains that “those seeking eviction should not rely on “concepts of faceless and anonymous squatters automatically to be expelled as obnoxious social nuisances,”Footnote ²⁰ and that “those who find themselves compelled by poverty and homelessness to live in shacks on the land of others, should be discouraged from regarding themselves as helpless victims, lacking the possibilities of personal moral agency.”Footnote ²¹ Thus, Justice Sachs concludes, “Wherever possible, respectful face-to-face engagement or mediation through a third party should replace arms-length combat by intransigent opponents.”Footnote ²²

Ubuntu arguably offers a philosophic tradition, already existing in South Africa, to buttress a republican interpretation of horizontal application. While not self-consciously or explicitly connected with republicanism, ubuntu tracks certain republican ideals in its calls for mutual respect among citizens.Footnote ²³ And indeed, the Court invokes ubuntu in crucial horizontality cases, including City of Johannesburg v. Blue Moonlight PropertiesFootnote ²⁴ and AB and Another v. Pridwin Preparatory School,Footnote ²⁵ discussed later. The point of comparing ubuntu and republican ideas is not to create a false equivalence between them, but rather to sketch the full contextual space that could admit a republican reading of horizontal application. Indeed, ubuntu is a primary reason there is more such space for a republican reading of horizontality in South Africa than in nearly any other context in the world. As Chapter 2 explains, horizontal application seems to call for a new theoretical reference point that this book sums up in the concept of the common good – the sheer idea that there is a good of the community of which to speak, one that extends beyond individual rights to encompass some consideration of shared ends. However, what ultimately is contained in, or what substance is meant by, the common good is at least partly contingent on the politics of the place.Footnote ²⁶ Likewise, ubuntu emphatically is not about individual rights but rests on the essential nature, even the priority, of community. Therefore, in addition to tracking a similar orientation toward the communal, the concept of ubuntu is noteworthy as it potentially offers some substantive content, even if at a high level of abstraction, to understand what exactly the common good might involve in this context.Footnote ²⁷

In addition to this age-old principle of ubuntu, early proposals of the ANC also point toward something of a communal approach to treating South Africa’s pathologies. Democracy would become the ANC’s constant refrain, as the party put its faith in the change it hoped would follow from enfranchising the population’s actual majority. Moreover, the cooperation even of private actors within the public project was key. In a certain sense, bringing public and private standards under a more common regulative umbrella could serve to democratize law and politics, rather than exempting one corner of law to follow a different, perhaps less demanding standard.Footnote ²⁸ In this democratized understanding, no space of society exists beyond the fundamental law to which “the people” have ascribed their consent, provided that the people are in fact consenting to and creating that law.

Giving voice to these commitments, among others, the ANC Constitutional Committee drafted its 1991 Bill of Rights for a New South Africa even prior to the negotiations and convening of the Constitutional Assembly.Footnote ²⁹ While this document would not be adopted wholesale into the Final Constitution, it offers a clear articulation of the solutions key actors envisioned would solve South Africa’s problems. Relevant to the present project, Article 14(1) states, “In its activities and functioning, the State shall observe the principles of non-racialism and non-sexism, and encourage the same in all public and private bodies.” In a similar vein, Article 14(3) continues, “The State and all public and private bodies shall be under a duty to prevent any form of incitement to racial, religious or linguistic hostility and to dismantle all structures and do away with all practices that compulsorily divide the population on grounds of race, color, language, or creed.” And finally, Article 16(3) on Enforcement states, “The terms of the Bill of Rights shall be binding upon the State and organs of government at all levels, and where appropriate, on all social institutions and persons.” Such early proposals clearly imply a republicanesque diagnosis of the situation and republican prescription for what a new, democratic South Africa would require. References to how the state should direct both public and private action, as well as to the duties of private bodies to dismantle divisive social structures, contemplate something like horizontal application. Moreover, such formulations seem to conceive of these duties not as separate from or supplementary to the public project but as constitutive of a vision or good that could only be achieved in common.

The proposed Bill of Rights continues with Article 15 establishing the possibility that legislation might limit the exercise of rights, and beginning to outline the conditions under which such limitations could occur for the sake of maintaining “an open and democratic society.”Footnote ³⁰ As Chapter 2 suggested, limitation clauses are arguably a precondition for horizontal application, insofar as they acknowledge that individual rights might give way to broader ends, whether to the rights of others or to certain communal goods. Such limitations may not amount to full-blown duties, necessarily, but they do entail some concession of limits to exercising one’s own rights. Following the Canadian Charter as an example, the South African drafters, therefore, took the crucial step to preserve limitations analysis in both the 1993 Interim Constitution (Chapter 3(33)) and the Final Constitution (Chapter 2(36)).Footnote ³¹ By the Interim Constitution’s adoption, the early articulations of horizontality in the ANC’s 1991 Bill of Rights, mentioned above, had been dropped, with no mention of private bodies in Chapter 3(7) on the “Application” of rights. While the ANC’s vision offers robust ground for a republican interpretation in early proposals for horizontal application, the omission of horizontal application from the 1993 text likewise reveals the presence of other forces in the constitutional story running up against this republican interpretation.

The Interim Constitution: A First Step “to Make the Best Constitution in the World”

The process by which South Africa adopted its new constitution is well known.Footnote ³² The National Party (what historically had been the white Afrikaner ethnic nationalist party) and the ANC (founded for the purpose of ending Apartheid) agreed to pursue a two-stage constitution-making process, beginning with an Interim Constitution negotiated between the parties that would set the terms for a Final Constitution. Ultimately the Constitutional Court would prescribe nine changes to the draft text before certifying the country’s new vision in the Final Constitution. The choice to conduct a multistage process was a conscious attempt to navigate an issue ubiquitous in constitution-making but especially salient in the South African context, namely, the challenge of balancing majoritarian rule with the protection of minority rights.Footnote ³³ In South Africa, this meant facilitating the transition to democracy by abolishing racial elections and granting universal suffrage, a move likely to cast white populations in the role of political minority in the new government. The Interim Constitution was thus an effort to begin the constitution-making process by setting initial terms on which all the major parties could agree.Footnote ³⁴

As the ANC came to the negotiating table with such aspirations as they had articulated in their 1991 Bill of Rights, so too did the National Party come with goals of its own, namely, to secure a space for itself in public and private life as it assumed minority status.Footnote ³⁵ On this basis, National Party representatives initially advocated for group rights in the new constitution, specifically for the recognition of the Afrikaner population as a group with such rights as to education in its own language.Footnote ³⁶ The concept of group rights was met with great suspicion, however, from both domestic and international voices. With the possibility of group rights roundly rejected, the National Party ultimately changed tack to pursue individual rights instead, especially strong property rights,Footnote ³⁷ as well as a high degree of autonomy in private spaces and at local levels. Indeed, local governments were not even discussed at this stage of the deliberations and would remain racialized through the 1990s.Footnote ³⁸

This early stage of constitution-making was an exercise in pragmatism and compromise, arguably to a degree exceeding many other nations’ processes. The goal at this interim stage was simply to commence a conversation the parties would be willing to continue into subsequent years. As a practical document, the Interim Constitution rolled back many facets of the ANC’s early proposals. The proposed practice of horizontal application, for example, was dropped. Relatedly, the text shows some shift in emphasis toward such rights as to property, departing somewhat from the vision of the 1991 Bill of Rights. In the terms of this book, that which might have supported republican understandings – say, a vision of ends related to elevating the socioeconomic status of black populations – was effectively dropped at this stage to accommodate other rights and interests.Footnote ³⁹ Recounting this dynamic in his memoir, Justice Sachs explains how “[t]he dialectic of legal development was such … that concepts intended to assuage the anxieties of the whites inevitably aroused the concerns of the blacks.”Footnote ⁴⁰

Although direct horizontal application was omitted from the text of the Interim Constitution, the subject did arise in the deliberations. Mindful of Apartheid’s deep-seated legacy in the social fabric, this was a natural question for ANC representatives to consider as they aimed beyond formal rights protections at thoroughgoing transformation.Footnote ⁴¹ Insofar as private actors would maintain great power, framing rights as only the state’s responsibility, rather than shared by citizens or other private actors, might have impractically constrained the larger project.Footnote ⁴² Interestingly, however, the prospect of applying rights horizontally was debated both among and within parties involved in the deliberations of the Interim Constitution.Footnote ⁴³ Despite some consensus that the new constitution was bound to bring sweeping change, less clear was the role of horizontal application in this project. Members across the parties feared the real possibility that a clumsy formulation of horizontal application would have far-reaching and unintended consequences.

Noting the global trend toward horizontality, the Technical Committee on Fundamental Rights included horizontal application in the original proposal for the Interim Constitution.Footnote ⁴⁴ Unsurprisingly, representatives of the National Party and the libertarian Inkatha Freedom Party were wary of this provision, pushing instead for a vertical model to obligate only state actors and not change so drastically the status quo that privileged whites.Footnote ⁴⁵ Moreover, the Communist Party also expressed hesitation about horizontal application, but their criticism stemmed from fear of its bidirectionality, or the possibility that corporations could claim rights against private individuals, too.Footnote ⁴⁶ In such an understanding, horizontal application could be abused by “economically powerful institutions in areas which do not properly fall within the ambit of an instrument for human rights protection.”Footnote ⁴⁷

Despite early articulations of horizontal application in the ANC’s 1991 Bill of Rights and the Technical Committee’s proposal, a gap emerged even among those who favored significant change under the new constitutional dispensation. For example, in Spitz’s telling, Halton Cheadle and Albie Sachs disagreed on this question, Sachs counting himself among those who thought horizontal application necessary to prevent the privatization of Apartheid.Footnote ⁴⁸ Spitz also attributes this divide to various party members’ consultation with such law professors as Laurence Tribe. Apparently, Tribe cautioned the committee on rights against horizontal application, for the common reason that legislatures, rather than courts, are better suited to regulating private relations.Footnote ⁴⁹ Likewise, Chief Justice Michael Corbett commented on a draft expressing the judiciary’s opposition for similar reasons, asserting that horizontal application would bring judges into policy debates that ought to happen in politically accountable branches of government. On the other hand, others’ worries cut in the other direction, that cementing horizontal application of particular rights in the Interim Constitution would prevent courts from employing broader interpretive powers to apply additional rights horizontally in the future.Footnote ⁵⁰

Key drafter Halton Cheadle initially worried that horizontal application would empower “judges, particularly those appointed by the Nationalist régime to intervene in areas in which a democratically elected Parliament should be legislating.”Footnote ⁵¹ The fact that the ANC presumably would become the majority party in the new parliament sheds further light on the hesitance of party affiliates to hand over additional power to the courts. As a practitioner of labor law, Cheadle also worried that horizontal application would bring too much within the purview of the Constitutional Court to the detriment of those issues and institutions dedicated more specifically to fair labor practices. Indeed, his concern was not unlike that of some German practitioners of labor law in the years leading up to Lüth. Spitz suggests that such worries over horizontal application were symptomatic of the presence and different preoccupations of academics and practitioners in the Assembly.Footnote ⁵² The academics were generally convinced that some provision of horizontal application was essential to realizing constitutional values writ large. In contrast, the practitioners initially were more inclined to preserve traditional distinctions in law, often for the simple reason that they were not clear on how this novel constitutional mechanism would operate in practice, but could see the potential repercussions of a careless execution.

Some of these apprehensions of what horizontal application might amount to seem to conjure up possible interpretations of horizontal application, beyond a republican one. The concerns of Professor Tribe, Chief Justice Corbett, and Halton Cheadle, for example, give voice in different ways to the objection, taken up in Chapter 2, that horizontal application’s reliance on courts may be antirepublican. Such institutional questions may be tied to substantive matters, moreover – for example, whether the courts could be trusted to place due weight on such issues as fair labor, or instead would give priority to more traditional rights. A picture begins to emerge of how horizontal application could fit interpretations that are not clearly republican by keeping rights and the individual at the center of the inquiry, rather than initiating a kind of reordering to prioritize certain ends and community.Footnote ⁵³ Put differently, the fact that private actors may come to have duties to one another may not necessarily entail a common good that has assumed priority, but may in fact be designed to preserve the individual and individual rights above all.

In this way, these debates over the Interim Constitution reveal the question of whether horizontal application would ultimately be employed to further the ANC’s early articulations of a new constitutional vision, or simply serve as another means by which private interests were furthered – whether the old interests connected with the Apartheid regime or a new set of private interests altogether. Indeed, this is not unlike the criticisms discussed earlier about the way ubuntu has come to be coopted by a rights-centric narrative. It is worth noting, however, that in these debates, neither the National Party nor corporate interests ever pushed for horizontal application. And indeed, those corners of the ANC and its allies who voiced concerns were presumably thinking through more extreme and marginal scenarios of how this practice could be employed. Ultimately, the ANC and its allies came to support horizontal application, judging that the capacity of this practice to achieve the ends of the ANC’s constitutional vision outweighed any potential threats.

Despite the balance ultimately weighing in favor of horizontality, the rights committee led by CheadleFootnote ⁵⁴ would adopt a vertical model for the Interim Constitution, omitting reference to direct horizontal application while still admitting something like indirect horizontality.Footnote ⁵⁵ Section 7(1) in the chapter on rights, for example, does not list the judiciary among the state organs bound by constitutional rights. Du Plessis and Corder suggest that this provision should, nevertheless, be read as including the judiciary if it is to cohere with other parts of the Constitution.Footnote ⁵⁶ Specifically, they mention Section 4(2)’s statement of the Constitution’s supremacy, binding “all organs of state at all levels of government.”Footnote ⁵⁷ The Interim Constitution did not embrace a robust practice of horizontal application, but also did not preclude it altogether. Presumably, as was its stated goal, it aimed at a kind of middle ground in light of the need for consensus both within and between parties. The Interim Constitution remained uncommitted to direct horizontal application, but also took steps to avoid a strict state action doctrine of the American stripe.

Amid the general caution surrounding horizontal application at this early stage, the committee also agreed to what became known as “seepage provisions” in Section 35(3). This section states: “In the interpretation of any law and the application and development of the common law and customary law, a court shall have regard to the spirit, purport and objects of this Chapter.”Footnote ⁵⁸ This formulation would not be unusual as justification for the horizontal application of rights in other contexts. Indeed, it resembles the reasoning of the German Federal Constitutional Court in Lüth, that the Basic Law’s objective order of values necessarily informed the interpretation of all other areas of law. To this extent, the Interim Constitution left an opening for some form of horizontal application and even laid groundwork for constitutional rights to trump customary law.Footnote ⁵⁹ But while it required judges to take constitutional principles into consideration, it did not provide any mechanism that would require judges to take the affirmative step of reassessing private law.Footnote ⁶⁰

More than a mere placeholder, the Interim Constitution came with the promise that the National Party’s voice would maintain a significant level of volume even as the ANC grew louder. Following the adoption of the Interim Constitution in November 1993 was the first nonracial election in April 1994, allowing the ANC to assume its status as, in fact, representing the majority of South Africans after some 350 years of disenfranchisement. Even in later stages of the process, however, the Constitutional Assembly offered protection to the National Party, such as through high thresholds for adopting certain provisions into the Final Constitution, including the Bill of Rights, which required a noteworthy 75 percent affirmative vote.Footnote ⁶¹

The Court at a Crossroads: Du Plessis v. De Klerk

In the years the Interim Constitution was effective, from 1993 to 1997, the question of horizontal application remained a live one.Footnote ⁶² Those working on Theme Committee Four on Fundamental Rights in the Constitutional Assembly and others in the ANC parliament began to coalesce around the earlier consensus reflected in the 1991 Bill of Rights regarding the need for something like horizontal application. Indeed, there was a sense of urgency that they had to get this issue right in the Final Constitution. The drafters also viewed the horizontal application of rights against private actors as a way to signal to the larger populace that the Constitution would meet their material needs.Footnote ⁶³ National Party members still preferred a vertical model, but chose instead to direct their energies toward protecting property rights, not least because their status had changed from the dominant to opposition party by this later stage.Footnote ⁶⁴ The constitutional drafters thus seemed to reach some settlement on the question of horizontality, reintroducing language that approximated provisions from the 1991 Bill of Rights. The questions that remained about horizontal application were more practical in nature, about specific remedies and finding a formulation sufficiently precise so as not to give courts a carte blanche.

In the meantime, the newly established Constitutional Court contended with the ambiguous language of the Interim Constitution, when it decided Du Plessis v. De Klerk. This judgment has received much attention since it was decided in 1996, largely because the majority assumed something like the conventional understandings of public and private spheres, and public and private law, in its interpretation of the Interim Constitution. In short, the Court opted for a more moderated indirect horizontal effect of the German variety rather than the direct horizontal effect for which many had been arguing. Indeed, this case and its accompanying judicial rationale made ANC drafters uneasy, contributing to their urgency to follow through with stronger, more explicit provision for horizontal application in the Final Constitution.Footnote ⁶⁵

In Du Plessis, a newspaper reported that South African citizens had been transporting weapons to rebel forces in Angola via covert flights. The newspaper suggested that such private air operators as Gert De Klerk were intentionally fueling the Angolan civil war in order to make a profit. De Klerk sued the newspaper for defamation, arguing that these articles had damaged both his reputation and his business.Footnote ⁶⁶ The decision itself hinged on whether the Interim Constitution could be applied retrospectively insofar as the newspaper (the plaintiffs) sought to rely on Section 15, protecting the freedom of speech and expression, including “freedom of the press and other media.” The fact that the Interim Constitution was not adopted until after the articles were published and damages were incurred posed some difficulty to their argument. Implicit in this question of retrospectivity, moreover, was the further question of whether an article of the Constitution could even be brought to bear on a private relationship in the way the newspaper argued. The case made its way to the Supreme Court of Appeal, which decided against the newspaper. Specifically, the Court decided that the Interim Constitution could not apply retrospectively and that the Bill of Rights of the Interim Constitution only had vertical effect.Footnote ⁶⁷ In plain terms, only the state could be said to violate rights such as to freedom of expression; private actors and spaces existed beyond this standard.

The Supreme Court of Appeal was considered to be the final court of appeal in all areas of ordinary law. The question of both retrospectivity and horizontality in the case at hand seemed to implicate constitutional questions, however. Crucial to the outcome of this case was the question of how the common law related to the Constitution: Were such aspects of the common law, such as the rule governing defamation, ultimately subject to the Article 15 commitment to free expression? With such questions straddling the divide between ordinary and constitutional law, the case went to the Constitutional Court. The Constitutional Court, in turn, agreed with the Supreme Court of Appeal that the Interim Constitution could not apply retrospectively. While this judgment alone would have been sufficient to determine the case’s outcome, the Court took the additional step of deciding the question of horizontal application as well.Footnote ⁶⁸ Justice Kentridge cited the Interim Constitution’s aforementioned Section 7 to argue that the Constitution included no basis for direct horizontal effect; indeed, Section 7 only obligated the executive and legislature to uphold the Bill of Rights. The Court thus reasoned that the text of the Interim Constitution provided a foundation for a kind of indirect horizontal effect that closely resembled the German practice of Drittwirkung. Indeed, Kentridge suggested that horizontal effect in the Interim Constitution resembled the German model more than any other approach to horizontal effect in a national constitution. In this understanding, the Bill of Rights of the Interim Constitution obligated private relationships only insofar as statutes attempted to regulate them.

South Africa was different from Germany, however, in that it maintained a system of common law. This common law, the Court concluded, was not subject to the Constitution in the same way, largely tracking the Canadian take on this question.Footnote ⁶⁹ Of course, Section 35(3) of the Interim Constitution required that a court should have regard for “the spirit, purport and objects” of the Bill of Rights in interpreting any law, and the provision explicitly included the common law in this.Footnote ⁷⁰ Nevertheless, Justice Kentridge suggested that the Constitutional Court was not up to this task, explaining that it simply did not have the capacity to balance matters of common law in the way that would be required in applying horizontal effect. In particular, he worried that in deeming some rule of common law incompatible with the Constitution, the Constitutional Court would have to make law to fill the resulting gap.Footnote ⁷¹ In addition to actions of parliament in the realm of statutory law, the ordinary courts and particularly the Supreme Court of Appeal were better equipped to apply horizontal application in their “routine common law interpretive work.”Footnote ⁷² Kentridge’s argument is premised on a specific understanding of law and the work common law does to balance private relationships. In Kentridge’s telling, these new constitutional rights were likely to resist balancing in the particular ways that common law decision-making prescribed.Footnote ⁷³ In this way, the Constitutional Court’s initial formulation of horizontal application was one that preserved traditional boundaries of law, largely intended to isolate private relationships from constitutional obligations directed to the state.Footnote ⁷⁴

Early critics of the Du Plessis decision, Stuart Woolman and Dennis Davis, suggest that the Interim Constitution pointed just as easily, if not more so, to the opposite conclusion. They argue that the Interim Constitution supplies good evidence that the drafters did mean to govern all aspects of law and life, including the common law, as evinced by Section 35(3).Footnote ⁷⁵ The very fact of Justice Albie Sachs’s concurring opinion and, especially, Justice Kriegler’s dissenting opinion in Du Plessis illustrate the plausibility of such alternative reads of the Interim Constitution. We see this tension in Justice Sachs’s opinion, in particular. On the one hand, as a core member of the ANC in the decades prior, he recognizes the capaciousness of the South African constitutional project. He states:

In this way, Justice Sachs expressed his understanding that the purpose of the South African Constitution differed from the conventional model that prioritized negative rights and sought only to protect against government interference. Indeed, he cites “the circumstances in which [the] Constitution came into being,” rooting its normative commitments in a larger understanding of the history of the South African polity. But this understanding was not enough ultimately for Justice Sachs to depart from the majority, as he saw this consideration of constitutional purpose as “not the issue” of the case at hand.Footnote ⁷⁷ He explains his more immediate concern that the enforcement of constitutional commitments appropriately depends on the actions of parliament and the Supreme Court of Appeal.Footnote ⁷⁸ Moreover, following Kentridge, he questions whether the Constitutional Court is even equipped to undertake the sort of “social, political, and economic questions” that accompany horizontal application when individuals’ rights and duties are weighed against each other.Footnote ⁷⁹ In this way, Sachs straddles the line between the concerns to which Kentridge gives voice in the majority opinion and acknowledgment of certain normative commitments that would seem to accommodate a larger role for the Constitutional Court.

That “the most radical member of the court”Footnote ⁸⁰ could not bring himself to endorse horizontal effect in this instance is revelatory of the sort of crossroad the Court faced in Du Plessis v. De Klerk, still operating under the Interim Constitution and well aware of the disconnect between aspiration and the ability to implement those aspirations in local settings. The tension with which Sachs wrestled – of a constitutional vision that clearly implicated the polity as a whole but did not decidedly abandon certain forms of liberal thinkingFootnote ⁸¹ – was not as salient for dissenting Justice Kriegler. After “castigating the majority,”Footnote ⁸² Kriegler states:

Kriegler had little patience for the sorts of distinctions that Kentridge and Sachs maintained in their own opinions. Indeed, Section 35(3) and others that Kriegler cites suggest that the “spirit, purport, and objects” of the Constitution govern all law and life in the polity, regardless of the particular institution or court that happens to be acting. Even beyond such specific provisions, however, Kriegler insists on the need to consider the South African constitutional vision as a whole. Due appreciation for the power relationships embedded in South African society and the Constitution’s frontal assault on these public and private power structures do not permit the queasiness that Kentridge and others display with respect to the intervention horizontal effect would entail. Kriegler brings his rejection of liberal presuppositions into sharp relief when he accuses the majority of “preying on the fears of privileged whites, cosseted in the past by laissez faire capitalism thriving in an environment where the black underclass had limited opportunity to share in the bounty.”Footnote ⁸⁴ Laissez-faire priors may work for other polities, but it is not the basis of the new South African polity, Kriegler argues, and in fact represents much of what the Constitution is combating.

It is not difficult to see how the questions of Du Plessis could yield such an array of answers. For example, while the Interim Constitution does not explicitly provide for horizontality, certain provisions like the “seepage provisions” in Section 35(3) could be interpreted as a basis for indirect horizontal application. Kentridge found some grounding for traditional liberal commitments in the Interim Constitution, while Kriegler’s more purposive take brought him to a very different conclusion. And indeed, these disparate readings pointed toward a real choice – Robertson describes this as a choice between “thin” and “thick” constitutionalism,Footnote ⁸⁵ while Woolman and Davis understand it as the difference between classical and “creole” liberalism.Footnote ⁸⁶

An additional, and important, reason for the impasse in Du Plessis was the fact that the Interim Constitution maintained separate jurisdictions of the Constitutional Court and the Supreme Court of Appeal. Insofar as the Supreme Court had been the highest court until this time of transition, there was good reason to keep it as supreme in its jurisdiction even after the Constitutional Court was established. Indeed, the continued separation at this stage was indicative of a general distrust of the extent of the old Supreme Court’s loyalty to the new constitution, as well as an initial concession to appease those who worried that the Supreme Court no longer would enjoy primacy in its jurisdiction.Footnote ⁸⁷ This separation of courts preserved in the Interim Constitution provoked the questions we find in Du Plessis concerning the relationship between ordinary law, common law, and constitutional law.

This separation of courts is not all that different from that which informed the German Federal Constitutional Court’s decision in Lüth. Underlying Lüth was the similar question of how the Federal Labor Court, as well as other private law courts, related to the Constitutional Court. In Lüth, the Constitutional Court asserted itself as the primary and final interpreter of the Constitution; however, it did not prevent other institutions from also engaging in constitutional interpretation. Neither did the Constitutional Court presume to have final interpretive authority with respect to the Civil Code. Indeed, its decision to focus primarily on the general clauses, as explained earlier, shows a kind of self-imposed limit and even deference to other apex courts. Michelman sees all of this as evidence that Germany preserved separate systems of law and courts, even as Lüth required that the principles of the Basic Law influence the private law. In the South African context, on the other hand, the separation between the jurisdiction of the Supreme Court of Appeal and that of the Constitutional Court, Michelman thinks, “was always headed for instability.”Footnote ⁸⁸ He goes so far as to describe such a system as a “design error” of the Interim Constitution in need of correction,Footnote ⁸⁹ given the vast transformation the ANC and other constitutional actors sought. And indeed, the Final Constitution of 1996 brought significant change, demonstrating that a choice had eventually been made. The Constitutional Court was confirmed as the final arbiter in all areas of law, including matters of common law, and direct horizontal effect of the Bill of Rights became a constitutional requirement. “Under pressure from the idea of a socially transformative, constitutional bill of rights,” the Final Constitution united the South African systems of law and courts under a common constitutional standard.Footnote ⁹⁰ With the adoption of the Final Constitution, the institutional structures were more clearly harmonized with the political needs entailed by these new substantive commitments.

South Africa as Heir and Foil to German Constitutionalism

Before discussing the specific changes of the Final Constitution and subsequent caselaw on horizontal effect, it is worth fleshing out comparisons between South Africa and Germany, especially as some justices writing in Du Plessis explicitly mention the German case.Footnote ⁹¹ In particular, Justice Laurie Ackermann’s opinion in Du Plessis compares the South African situation with postwar Germany, finding in these apparent similarities justification to develop an understanding of horizontal application that was also similar. Ackermann maintained that the German Basic Law “was no less powerful a response to totalitarianism, the degradation of human dignity and the denial of freedom and equality than our Constitution.”Footnote ⁹² Indeed, he elaborates that a similar stage for the development of horizontal application was set by each country’s troubled history. If indirect (in contrast with direct) horizontal effect was good enough for Germany, Ackermann suggests it ought to be good enough for South Africa, too.Footnote ⁹³

Woolman and Davis, however, insist on important differences between Germany and South Africa that, in their view, Ackermann and others paper over. While the German and South African Constitutions were both responses to serious rights abuses and totalitarian regimes, the states of affairs in each postconflict situation were vastly different. After the Second World War, Germany was a “modern, industrialized and relatively egalitarian society. It was into these less than dire circumstances that the [Basic Law] was born.”Footnote ⁹⁴ On the other hand, Woolman and Davis wrote:

This explanation gives some historical context for Germany’s ability to cling continually to old legal structures, even importing directly many Weimar legal structures. Whereas Du Plessis v. De Klerk initially seemed to put South Africa on that same track of preserving some insulation of private relations from constitutional standards, such expressions as Kriegler’s condemnation of laissez-faire structures pick up on a progressive shift. And indeed, taking a cue from the ambivalence in Du Plessis, the drafters of the Final Constitution made explicit the sheer scope of transformation intended, expanding the Constitutional Court’s jurisdiction and clarifying a doctrine of direct horizontal effect. This way, the drafters enabled future judgments to pursue the transformative constitutional project without the need for any “jurisprudential gymnastics.” The text and institutions of the Final Constitution were intentionally aligned with transformation and, by extension, with a stronger assertion of direct horizontal application.

The decision of whether and how to apply horizontal application came through different fora in Germany and South Africa – through the Constitutional Court and the Constitutional Assembly, respectively. However, in both cases, the express aim was to respond to histories of violence and to influence the broader social order according to constitutional values. To this extent, Ackermann’s comparison in Du Plessis is accurate. The parity with which constitutional values applied across public and private spaces manifested differently in each context, however. As explained with respect to Germany, its long tradition of civil law and system of specialized courts, as well as the persistence of the principle of Privatautonomie, tracks the way in which horizontal application developed after Lüth. Indeed, most cases developed in the context of such classical negative rights as freedom of speech and assembly, with fewer occurring in the realm of equality and antidiscrimination until much later.Footnote ⁹⁶ And when courts and even legislatures did move to bring certain equality rights into private spaces, private interests often resisted. The more traditional rights that make up the content of the German Basic Law as well as certain structural and doctrinal features, therefore, map onto a practice of horizontal application that is in some ways more limited. Likewise, while a republican interpretation still emerges in the German turn toward community, this too appears bounded in certain ways, as discussed in Chapter 5.

Comparing German Drittwirkung with horizontal application in South Africa bears out Woolman and Davis in their desire to lend nuance to Ackermann’s account in Du Plessis. In particular, the advance of horizontal application from the Interim Constitution to the settlement of the Final Constitution, to say nothing of the tension internal to Du Plessis, points to a real disconnect between the priors of some framers and judges and the ANC’s vision articulated in such early documents as the 1991 Bill of Rights and ultimately cemented in the Final Constitution. This was the quandary Du Plessis hit upon and which a comparison of Germany and South Africa brings into sharper relief. The formulation of horizontal effect in the Interim Constitution was comparable to that of the German Federal Constitutional Court in Lüth, despite the fact that the countries actually faced very different problems.

To be sure, classical liberal rights and freedoms are represented in South African constitutionalism. Indeed, the Constitution founds the polity on “[h]uman dignity, the achievement of equality and the advancement of human rights and freedoms.”Footnote ⁹⁷ But these commitments carry unique status and meaning in South Africa. For example, both freedom and equality encompass more than their formal meanings, in that they involve certain material prerequisites as well.Footnote ⁹⁸ Moreover, rather than shy away from certain instances of equality in the context of horizontal effect as in the German case, the South African Constitution actually singles out equality by name as requiring horizontal effect. Chapter 2, Section 9(4) guaranteeing the right to equality, states:

Of course, merely providing that the right to equality or right against discrimination has horizontal effect does not entail that a person’s right to equality will always prevail against the countervailing right in question. (Section 9(5), for example, suggests that there are instances in which discrimination may be fair.) Nevertheless, the very fact that the South African Constitution so prioritizes equality to specify its provision for horizontal application diverges from the German treatment of the same question.

New Clarity in the Final Constitution

The 1996 Constitution established the Constitutional Court as “the highest court in the republic,” with authority to decide constitutional matters as well as any other matter it decides is in its jurisdiction.Footnote ¹⁰⁰ It was within this new institutional context that the framers entrenched direct horizontal effect as a legal-constitutional practice.Footnote ¹⁰¹ In contrast with the ambiguity of the Interim Constitution, the constitutional drafters established horizontal application with uncommon clarity, taking pains to avoid the sort of confusion evident in Du Plessis v. De Klerk.

Several provisions across different sections of the final text collectively establish horizontal effect. In contrast to the Interim Constitution that obligated only the actions of the legislature and the executive to the Bill of Rights, Chapter 2, Section 8(1) of the 1996 Constitution provides that the Bill of Rights binds the judiciary as well. In addition, several entirely new provisions concerning horizontal effect were added. Among them were Section 9(4) on the right to equality, described earlier, as well as Section 8(2):

In stating that the Bill of Rights binds “a natural or a juristic person,” Section 8(2) establishes that rights apply horizontally to obligate both private individuals (natural persons) and other private entities such as firms and corporations (juristic persons). When a panel of experts at the Constitutional Assembly considered this terminology, they found strong reason to bind both natural and juristic persons in view of the goal “not just to root out discrimination by the state but to attack it in its most pervasive form – discrimination as between citizens.”Footnote ¹⁰² This section goes on to state that the nature of the right and the nature of the duty imposed by the right may influence the outcomes of particular cases when balancing one right against another.Footnote ¹⁰³ Therefore, although horizontal effect establishes parity in applying South African constitutional values across spheres, Section 8(2) creates some space for variation in the way rights obligations apply to private actors as opposed to state actors.Footnote ¹⁰⁴

Also in Section 8, the Constitution explains specifically how a court will apply a right horizontally, stating that a court must apply or develop the common law in applying horizontal effect. The provision reads:

This section seems to take Du Plessis v. De Klerk head on, giving all courts, including the Constitutional Court, authority to develop the common law in order to hold private actors accountable for the Bill of Rights. Moreover, as was the case in Section 8 (1), the Constitution acknowledges that some rights will necessarily be limited in the process of balancing. Taken together, these several additions to the Final Constitution mark a clear departure from foregoing renderings of horizontal effect. Indeed, Halton Cheadle had long moved on from any initial concerns to argue the necessity of horizontal effect, drafting Section 8 to provide just the kind of constitutional cause of action the Interim Constitution lacked. How consistently jurists would treat this provision as a break from the Interim Constitution was another matter.

In terms of institutional structures, doctrinal specifics, and constitutional content, the South African Final Constitution cemented a clear shift away from the German model. Insofar as both constitutions implemented some version of horizontal application, constitutional standards came to govern public and private spaces alike. In the terms of this book, the turn toward horizontal application can be interpreted as a turn toward certain communal values over sheer individual rights in both places. Nevertheless, from the outset, the particularities of the South African case seemed to invite constitutional actors to extend horizontal application further into private spaces than in the case of Germany. Moreover, subsequent jurisprudence and legislation contributed to a different scope for the parity and duties resulting from the initial communal turn, as constitutional actors in each place incrementally carved up the private sphere in different ways. Thus, while a republican interpretation applies in both experiences, constitutional actors in South Africa now operate in a constitutional context where that interpretation may ultimately ring with more truth at more levels of society.

What remains of this chapter illustrates a republican interpretation of subsequent arguments surrounding horizontal application in South African constitutionalism. Emboldened all the more by the Court’s ambivalence in Du Plessis, the ANC drafters and those allying with the ANC followed through in their intent to transcend the Interim Constitution by providing an explicit basis for horizontality in the Final Constitution. Nevertheless, much about how horizontal application would develop (or not develop) was left to the discretion of judges and other constitutional actors. As explained earlier, South Africa’s experience displays a potential scope for horizontal application, and thus a depth to the republican interpretation, that exceeds that of the German context. Simply put, the commitments of the political community can and have been understood as encompassing more in South Africa and, on this basis, have allowed constitutional actors to make greater inroads into private spaces – including in such areas as housing, education, and even testamentary decisions. At the same time, constitutional actors in South Africa have not always employed this practice in a maximal way. Judges anywhere will always have an interest in drawing lines and developing limiting principles. This seems all the more likely in the case of horizontal application, which marks a departure from conventional understandings and implicates private actors in new ways, intensifying such well-trodden apprehensions in judicial politics as enforcement and funding of courts’ directives.Footnote ¹⁰⁵ Other countervailing factors, such as a pervading formalism in the legal culture, also figure into the practice of horizontal application. Thus, while the constitutional text seems to allow for a robust, republican understanding of the communal, this same text has been employed in more or less expansive ways in the hands of different constitutional actors.

Early Development and Reversion

Albie Sachs and several other justices who joined Kentridge’s Du Plessis opinion quickly acknowledged the different logics the Final Constitution permitted in Carmichele v. Minister of Safety and Security.Footnote ¹⁰⁶ Carmichele established that the courts had an obligation to develop the common law in the light of the Constitution. The case concerned a man, charged with and jailed for assault, who committed another assault after law enforcement had released him on bail. The victim of the subsequent attack argued that police and public prosecutors had “negligently failed to comply with a legal duty” to protect her from a known aggressor.Footnote ¹⁰⁷ The High Court and Supreme Court of Appeal decided there was no evidence that law enforcement had “acted wrongfully.” And so, the applicant appealed to the Constitutional Court. The Constitutional Court, in contrast, rejected tendencies to distinguish between action and inaction on which, for example, the US Supreme Court had relied in DeShaney v. Winnebago County Department of Social Services.Footnote ¹⁰⁸ Rather, a provision similar (though, importantly, not identical) to Section 35(3) of the Interim Constitution provided the basis for the Constitutional Court’s decision. Section 39(2) of the Final Constitution states, “When interpreting any legislation, and when developing the common law or customary law, every court, tribunal or forum must promote the spirit, purport and objects of the Bill of Rights.”Footnote ¹⁰⁹ In contrast with the counterpart provision in the Interim Constitution, Section 39(2) explicitly obligates every court to attend to constitutional commitments. And so, in a unanimous opinion, the Constitutional Court recognized that it had an obligation to ensure that the common law developed according to such constitutional standards that inhered in the rights as to life, human dignity, freedom, and security.Footnote ¹¹⁰ Writing for the Court, Justice Ackermann cites the German Basic Law again, arguing that the South African Constitution also encompassed an “objective, normative value system,” and that it was within this value system that the common law needed be developed. Ultimately, the Constitutional Court referred the case back to the initial High Court on the basis that the common law admitted of different possible modifications to accord with constitutional values.Footnote ¹¹¹

Some question just how radical a break Carmichele was from Du Plessis. Michelman, for example, argues that the case reveals that the Constitutional Court “internalized some separation,” as between systems of law and courts, according to the same traditional paradigm on display in Du Plessis. Though all South African law had to conform to the Constitution, he explains, the common law was still developed within its own framework, evinced by the fact that the Constitutional Court sent the case back to the High Court. This, Michelman argues, is not all that different from the system of separate courts and, by extension, the indirect horizontal effect Germany maintains.Footnote ¹¹² Indeed, Justice Ackermann’s recurring invocation of Germany suggests its continual influence on at least some corners of the Court. Chirwa, on the other hand, points out that the duty to protect that emerges from Carmichele is a step removed from a typical liberal framework and particularly from the requirement that state action be present in order to enforce constitutional rights. He highlights how in the South African context the state is “liable for an infringement of a constitutional right by a non-state actor if it fails to take ‘reasonable and appropriate measures’ to prevent it.”Footnote ¹¹³

Considering the Constitution’s full treatment of rights in private spaces, and particularly Section 8, the South African model clearly goes beyond German Drittwirkung as private actors can be charged with rights violations directly, and not simply through the distillation of private law.Footnote ¹¹⁴ Though such doctrinal differences are important, equally if not more important are how the terms of debate shift and how such differences are, or are not, reflected in understandings of the Constitution’s role in the larger society. Does it provide a standard for the larger polity, including duties of private actors, and how does this society ultimately look, given the particular commitments of the Constitution? While a preliminary and cautious step, Carmichele does seem to set the doctrinal stage for future expansion of horizontal application. Robertson explains how “a stream of cases” followed Carmichele in 2001, all taking as granted that “nothing should stand in the way of the instantiation of constitutional values in the working of the law.”Footnote ¹¹⁵ Whether this drive for executing the constitutional project would steadily propel the development of horizontal application is a different question.

The year after Carmichele, the Court decided Khumalo v. Holomisa,Footnote ¹¹⁶ representing a concrete step to develop the Constitution’s provision for horizontal application in Section 8. Like Du Plessis and so many other horizontality cases, Khumalo concerned a defamation action. A South African newspaper accused prominent politician Bantu Holomisa of involvement in criminal activities, and he, in turn, sued for damages of defamation. The case came down to the newspaper’s right to freedom of expression (Section 16) against Holomisa’s right to dignity (Section 10). The newspaper (the applicant) argued that the common law rule of defamation needed to be developed further in order to comply with the constitutional right to freedom of expression. In particular, the newspaper argued that, under a proper understanding of the freedom of expression, plaintiffs ought never to succeed in defamation cases “unless they can establish that a defamatory statement was false.”Footnote ¹¹⁷

Writing for the Constitutional Court, Justice O’Regan acknowledged that the freedom of expression had horizontal effect, “given the intensity of the constitutional right in question, coupled with the potential invasion of that right which could be occasioned by persons other than the State or organs of State.”Footnote ¹¹⁸ Nevertheless, this right had to be balanced against the right to dignity, as it also had horizontal effect. Weighing these two against each other, O’Regan concluded that the common law rules of defamation, in their current state, struck a balance that was, in fact, compliant with both of these constitutional rights commitments. She explained that newspapers and other media would only be charged for defamation when they could not establish that “the statement was true and its publication in the public interest, nor that the publication was reasonable in all the circumstances.”Footnote ¹¹⁹ However, she also drew attention to the great limitation that shifting the burden of proof to the plaintiff would entail for the right to dignity since it was sometimes impossible to demonstrate the falsity of a claim. Since the newspaper could establish neither the truth nor the reasonableness of the accusations, the Court decided that the common law rules were sound and favored the right to dignity on balance in this particular case.

In Robertson’s telling, the earlier Carmichele decision had been necessary “to bring the jurisprudence on the development of the common law into line with the much firmer stand taken by the Final Constitution after the weakness of the court’s decision in Du Plessis.”Footnote ¹²⁰ Khumalo thus fortified and built on this earlier decision by extending its logic also to apply to relations between private actors. Indeed, with Khumalo, constitutional commitments came to comprise the very content of those common law rules governing private relations. Moreover, and perhaps more importantly, under the precedent set by Khumalo, the concepts of constitutional rights and duties are not excised as they enter private spaces. Rather, individuals face the prospect that they, too, are accountable for such constitutional rights commitments as freedom of expression or dignity. Justice O’Regan does stipulate the need to consider the intensity and nature of the right before applying it horizontally,Footnote ¹²¹ and other judges have been at pains to emphasize that the processes of determining the constitutional duties of state and nonstate actors are not equivalent.Footnote ¹²² Even with these caveats, however, the process of balancing and ultimate judgment in Khumalo still rest on the recognition that the Constitution is a source to both rights and duties of private actors.

A degree of caution in a court as new as South Africa’s Constitutional Court does not come as a surprise,Footnote ¹²³ even in spite of (or perhaps in light of) the promise horizontal application holds for asserting institutional power and jurisdiction.Footnote ¹²⁴ Theunis Roux describes in the early Court a preference for “context-sensitive balancing,” or a tendency only to decide the case at hand rather than create broadly applicable rules to apply to future decisions.Footnote ¹²⁵ Stuart Woolman notices and laments a similar tendency a decade after the Final Constitution’s adoption, explaining how this practice may cut in either a progressive or regressive direction.Footnote ¹²⁶ To illustrate his concern that the Constitutional Court risked regressing, he cites three cases over a single year in which the Court backpedaled from Khumalo’s initial steps to develop horizontality.Footnote ¹²⁷ Such backpedaling is on vivid display in Masiya v. Director of Public Prosecutions when the Court omitted any reference to Khumalo, instead reverting to Du Plessis v. De Klerk as precedent. Woolman describes a hesitance to engage the substance of the Bill of Rights when formulating duties against private actors. In contrast with the documented rationale for which the framers included Section 8 in the Final Constitution, namely, to hold private actors accountable for the specific rights, the Court relied on the vague bases of dignity, equality, and freedomFootnote ¹²⁸ and on Section 39(2) concerning constitutional interpretation rather than direct horizontal effect. Of course, such provisions are important, even essential, to South Africa’s constitutional project, but they do not perform the function of generating specific duties corresponding to enumerated rights as the Section 8 drafters and ANC representatives had earlier envisioned.Footnote ¹²⁹

The historical record is clear that the ANC intended to effect thoroughgoing transformation in South African law and life, and to do so in part through the mechanism of horizontality. Explained in terms of this book’s republican interpretation, the practice of horizontal application has the ability to recast constitutional rights as larger values and thus becomes a means by which to orient more corners of the polity toward the commitments of the new constitution. Why, then, has the Court sometimes avoided applying the Bill of Rights directly to private actors in the way Woolman recounts? Naturally, a confluence of factors will shape judicial decision-making on any issue. One such factor, although by no means the sole or even most crucial one, is how constitutional actors downstream understand the constitutional project. The discourses emerging in and from different cases offer some insight into the range of possible understandings and how those different understandings map onto the more specific question of horizontal application. That some jurists, such as Justice Ackerman, frequently invoked Germany as a point of reference to understand the South African Constitution is telling. Others, such as Justices Sachs, Madlanga, and Moseneke, have been more inclined to define the project on its own terms.Footnote ¹³⁰ Moreover, important debates emerge around the role judges claim for the courts in light of a transformative Final Constitution. Ought judges to employ a jurisprudential minimalism or formalism so that the legislature may spearhead the constitutional project? Or should the Court itself take charge, assuming as much of the project as its institutional capacities permit to ensure the country makes good on its commitments? Considering South Africa’s practice of horizontality alongside that of other countries, the Constitutional Court maintains a role in the country’s transformative project. Even in its relatively short history, for example, the Court has shown itself increasingly willing to apply socioeconomic rights horizontally,Footnote ¹³¹ not a trivial thing when contrasted with other countries’ more bounded practices of horizontality.

On what basis, then, do people like Woolman criticize the Court as unduly minimalist, even regressive? Such critiques typically assess the Court’s practice less from a comparative vantage point than from a view internal to South Africa. That is, they weigh the Court’s actions against what the Constitution itself seemingly permits, and thus find its development of horizontal application as unnecessarily halting – from the tendency to rely on Section 39(2) rather than Section 8(2), to the invocation of the vaguer constitutional provisions. One might say that the Constitution permits a practice of horizontal application that is all but unfettered, for the very reason that the republican potential of horizontality keeps pace with the reach of the ANC’s constitutional vision. But Woolman’s account illustrates how this same republican potential may be cause for wariness, as well. Many cases, including the crucial Grootboom case discussed later, suggest that the Court was ready to cast the legislature as the primary executor of the ambitious Constitution rather than take on the more positive aspects of the project for itself.Footnote ¹³²

This dynamic confronts the question described at the end of Chapter 2, and elsewhere in this book, regarding courts’ status as sufficiently republican bodies to determine whether and when rights apply to private actors.Footnote ¹³³ The German Constitutional Court may not confront this tension in the same way as the German Basic Law undertakes a project that is somewhat more modest, at least in certain ways and comparatively speaking.Footnote ¹³⁴ As Mathews explains in his book, the German Constitutional Court was actually eager to assert itself against the other branches and other courts. Michaela Hailbronner notably classifies the Basic Law as transformative;Footnote ¹³⁵ however, the German Constitution is more modest in what it asks, both of state and citizen, compared to the South African Constitution. While Germany’s case history suggests a court comfortable with the doctrine (and doctrinal scope) it had a definitive hand in formulating, out of South Africa’s case history emerge both discourses that embrace horizontal application and even illustrate its full republican potential, and discourses that attempt to demarcate boundaries to this practice. Early decisions thus show some unevenness in both judicial discourses and ultimate decisions. Later decisions, particularly pertaining to such areas as housing and education that are integral to South Africa’s transformation, are a bit more consistent in their embrace of horizontal application and in their display of republican concepts.

The Negative–Positive Divide in Housing and Education

Questions of dignity, free expression, and defamation are by no means novel in horizontality jurisprudence. Nevertheless, in securing the Final Constitution’s provision for horizontal application of these rights, Khumalo paved the way for other, more distinctive rights of the South African constitutional order also to obligate private actors. These other rights arguably reached deeper into private spaces and, to this extent, could potentially effect transformation all the more. Decisions on housing, education, and even wills and testaments thus emerged, sectors that by themselves signal horizontality extending further into private spaces than in many other contexts. Moreover, the language and arguments in these later decisions reveal the power of a republican interpretation in this context. Language of “neighborliness” and “shared concern” often seems to replace mention of rights altogether, cemented only more in certain invocations of ubuntu. Thus, amid criticisms of some judges’ inhibitive formalism and conventional understandings, much in these later decisions may be read through the lens of constitutional ends, rather than mere rights, pertaining to a larger community of private actors maintaining duties vis-à-vis those ends.

While finding clear space in more aspirational understandings of South African constitutionalism, many of these new rights commitments threw the courts into somewhat unfamiliar territory. In particular, such socioeconomic rights as health and housing raised new questions about the separation of powers and scarcity of resources, both public and private. In recounting his experience in deciding the important Grootboom case, discussed later,Footnote ¹³⁶ Justice Albie Sachs acknowledged the unique difficulties that come with enforcing something such as a right of access to adequate housing, when so much depends on the actions of and resources of legislatures and even private proprietors.Footnote ¹³⁷ Such difficulties first came before the Court in Soobramoney v. Minister of Health, KwaZulu-Natal (1997),Footnote ¹³⁸ when a terminally ill man by the name of Soobramoney invoked his right not to be “refused emergency medical treatment” under Section 27(3) and his right to life under Section 11 of the Constitution in order to receive renal dialysis from a state-funded hospital. As Soobramoney was of limited means, seeking care from a public hospital was his only option to sustain his life even a little longer. And yet he remained at the back of the line for the necessary care.

While conceding Soobramoney’s right to receive and the state’s duty to provide care, the Constitutional Court argued that it was not in a position to compel the state or the hospital to provide care, potentially at the cost of others requiring medical attention. Given that his was an ongoing issue and not a medical emergency, his right was contingent on what resources might be available at any given time, a calculus the Court simply could not make. Soobramoney died shortly after, and criticisms arose about the apparently contingent nature of socioeconomic rights. That medical care was contingent on the availability of resources and even the patient’s particular financial means struck many as incongruous with what was supposed to be a transformative constitution, and cast initial doubt on the whole enterprise of socioeconomic rights. In some ways, the Court responded in the Grootboom case on housing a few years later, acknowledging its own institutional role in supporting the most vulnerable and even its ability to require other institutions to make “reasonable” efforts toward securing such rights.Footnote ¹³⁹

Housing is a salient and complex issue in South African history. From the outset of European colonialism in the mid seventeenth century through the Apartheid regime of the twentieth century, racially based laws and evictions displaced indigenous peoples leading to widespread impoverishment of black populations.Footnote ¹⁴⁰ Thus, the South African Constitution provided, “Everyone has the right to have access to adequate housing,”Footnote ¹⁴¹ and legislation established various parallel safeguards, from protections for occupiers facing eviction to specialized courts that would see through land reform.Footnote ¹⁴² The Grootboom case concerned a particular settlement, Wallacedene, where only 5 percent of the dwellings had electricity, 25 percent of dwellings had no income at all, and 50 percent of residents were children.Footnote ¹⁴³ Mrs. Grootboom was one among many residing there who had applied for low-cost, state-funded housing, but remained in this increasingly uninhabitable settlement. In the end, the Court decided that Mrs. Grootboom was not entitled to emergency housing, but acknowledged that parliament was charged with taking reasonable action on the issue of housing in general. In the words of Justice Yakoob, “The case brings home the harsh reality that the Constitution’s promise of dignity and equality for all remains for many a distant dream.”Footnote ¹⁴⁴ Housing rights are, thus, aspirational in the fullest sense. Although rooted in the core pathologies of South African history, housing is among the socioeconomic rights that involve such balancing of interests and capacities as to avert full and immediate realization.

While the Constitution specifies that the duty to provide adequate housing falls on the state, the very nature of housing is such that cases frequently involve private relationships. Even in Grootboom the Court recognized that a “right of access to adequate housing also suggests that it is not only the State who is responsible for the provision of houses, but that other agents within our society, including individuals themselves, must be enabled by legislative and other measures to provide housing.”Footnote ¹⁴⁵ Those matters considered by the Land Claims Court, such as land restitution and eviction, raise questions of horizontal effect basically by definition.Footnote ¹⁴⁶ Moreover, several judgments of the Constitutional Court have concerned housing and explicitly address the question of horizontal effect.

In 2005, the Constitutional Court handed down the Modderklip decision.Footnote ¹⁴⁷ The owners of the Modderklip farm permitted some people to settle on the land. However, when thousands occupied the farm, the owners became overwhelmed and offered to sell the land to local authorities to accommodate the new occupants. The authorities refused to purchase and the local police demanded a high deposit to enforce the eviction order Modderklip obtained from a court.Footnote ¹⁴⁸ In the Supreme Court, Justice Harms acknowledged that housing rights could be enforced horizontally in theory, but decided that it could not in the present case.Footnote ¹⁴⁹ On appeal, the Constitutional Court did not technically rely on horizontal application, but instead ruled that the state had failed both in protecting the Modderklip company’s property rights and in securing the occupants’ right to have access to adequate housing.Footnote ¹⁵⁰ Although the Court chose not to apply horizontal effect explicitly, the Court’s remedy still involved balancing rights against each other and, to this extent, necessarily involved cooperation of the parties in securing the rights in question. In particular, the Court argued that eviction was at that point impossible given that, over five years, 40,000 people had come to settle on the farm and formed their own community.Footnote ¹⁵¹ As a remedy, therefore, the Court ordered that the state compensate Modderklip for the use of the land.

Although Modderklip was not responsible for the occupants’ right to have access to housing per se, the owners were, ultimately, still responsible for their housing in a practical sense. Moreover, although the occupants were not ultimately understood as trespassing, Modderklip was still entitled to compensation. Of course, the sheer logistics and financial cost of relocating 40,000 people figured into the solution the Court reached. At the same time, the Court’s judgment is punctuated with republicanesque ideas. As a discursive matter, the heart of this judgment is nothing less than a call for the private actors involved to participate in larger constitutional ends, to cooperate and even fulfill duties vis-à-vis one another. Quoting Albie Sachs in Elizabeth Port Municipality, discussed earlier, Justice Langa recounts the need “to balance competing interests in a principled way and promote the constitutional vision of a caring society based on good neighbourliness and shared concern.”Footnote ¹⁵² In the context of the Modderklip case, the difficulty of the immediate circumstances, to say nothing of the longer history that occasioned these circumstances, asks of the parties these same characteristics of neighborliness and shared concern. Put differently, the Court seems to presuppose that the actors involved identify with and share in the values of the constitutional project. Indeed, the very fact that the Court avoided the language of rights in discussing the relationship between the owners and the occupants brings this point into sharper relief. With this choice, “[t]he Court looked to social and economic norms as reflecting a constitutional vision of solidarity that altered the relation of the property owner to the settlers.”Footnote ¹⁵³ By avoiding the language of rights in favor of expressions of the collective pursuit of constitutional ends, the case may be read as taking on a more republican cadence.

From an analytical perspective, moreover, attention to the larger constitutional vision articulated in Modderklip, rather than the fact that the Court does not understand itself as applying horizontal effect per se, allows the observer to appreciate the extent to which the actors do retain responsibility here and participate in the constitutional project in the ultimate resolution. Indeed, the articulation of the Constitution’s commitments in terms of rights is only one possible expression of these commitments, seeing as the South African Constitution itself understands the Republic as founded on the more general values of human dignity, equality, freedom, nonracialism, nonsexism, and so on.Footnote ¹⁵⁴ Helen Hershkoff finds this approach of avoiding rights language in other contexts, too.Footnote ¹⁵⁵ This is worth acknowledging for the aforementioned analytical reasons as well as for more substantive reasons to which Hershkoff draws attention. Speaking specifically of Modderklip, she explains:

While Modderklip does not technically employ the language of rights and duties between the private actors, the Court does “reshape a power relationship,” as Hershkoff puts it, to balance their conflicting interests against each other. Whereas, in most countries, Modderklip’s right to property might have controlled the outcome of the case, here the owners’ rights were subject to the broader framework of social welfare norms or, in republican terms, to a particular conception of the common good. This broader normative context does not negate Modderklip’s property rights entirely; nevertheless, it does require a general compliance with the constitutional vision. As these norms thus apply across spheres, private actors such as Modderklip are brought into the fold of the larger constitutional project.Footnote ¹⁵⁷

In the years following Modderklip, the Court still contended with the clash between the rights to private property and housing, often invoking the constitutional ethic of neighborliness and ubuntu of Port Elizabeth Municipality, discussed earlier.Footnote ¹⁵⁸ One such case was Blue Moonlight Properties,Footnote ¹⁵⁹ in which a development company sought to evict eighty extremely poor people living on the property the company had recently purchased. The company was aware of the occupiers’ presence prior to purchasing, and eviction would all but certainly lead to their homelessness. The City of Johannesburg claimed it had resources to provide emergency housing only for those evicted from publicly owned land. Because these people occupied private property, however, the city could not offer any accommodation.

In Blue Moonlight Properties the Court claimed for itself more positive aspects of the constitutional project than it did in Grootboom, when it went to great lengths to accommodate the other corners of government. Ultimately, the Court decided that the constitutional guarantees to equality and housing, in fact, required the city to find accommodation for these occupiers although they were on private property. In the meantime, and more relevant to this book, the property owners themselves had a positive obligationFootnote ¹⁶⁰ pursuant to these same guarantees to equality and housing. Calling for patience, the Court explained that the property owners would have to continue housing these people until city officials could find alternative accommodation.Footnote ¹⁶¹ The Court is clear that the company would not have to provide free housing indefinitely, but as long as was required to ensure these people would not be homeless.

Another salient issue for South Africa’s constitutional project is education, as significant discrepancies in education were part and parcel of the inequality of Apartheid. And indeed, an important horizontality case decided in the same year as Blue Moonlight Properties concerned the right to education. In Juma Musjid (2011), the Court considered explicitly whether a socioeconomic right imposed an obligation on a private actorFootnote ¹⁶² when a private trust took steps to evict a public school that convened on its property. The Constitutional Court decided that the trust had “no primary positive obligation”Footnote ¹⁶³ to provide an education for the students, nor an obligation to make available its property for public use as a school. Nevertheless, the Court found that the trust did have “a negative constitutional obligation not to impair the learners’ right to a basic education,”Footnote ¹⁶⁴ pursuant to Section 29 of the Constitution. The Court concluded that the trust had every right to seek an eviction order, but that the courts were not obliged to grant one. Indeed, in evaluating the eviction request, the lower courts were required to consider “the best interest of the learners”Footnote ¹⁶⁵ and their right to a basic education. Thus, the particular question at issue in Juma Musjid was how to balance the right to a basic education and the right to property.Footnote ¹⁶⁶

Following the reasoning in Khumalo, the Court explained that the horizontal application of rights, including such socioeconomic rights as to education, depended in part on “the intensity of the constitutional right in question.”Footnote ¹⁶⁷ In the context of South African history, the basic right to education did rise to such a level of intensity as to call for a more congruent application to public and private actors alike. The Court explains the particular significance of the right to education in light of the history:

Thus, much like cases concerning housing, the decision to apply this basic right to education horizontally was grounded in the very purpose of the South African Constitution. The crucial nature of the issue of education, the Court held, warranted summoning private actors also to participate in its remediation or at least to cooperate until an alternative venue for the school could be secured. Both the nature of past abuses and the unique position of many private actors to exercise influence in education were cause to involve such institutions as the Juma Musjid Trust in the constitutional project. As property rights had been formulated to disadvantage the black population during Apartheid,Footnote ¹⁶⁹ putting other rights, such as to education, on equal footing ensured that claims to private property did not have undue weight and were calibrated to constitutional ends.Footnote ¹⁷⁰ Moreover, as this decision of the Court to balance these as competing rights departed from the preconstitutional status quo, so too did the decision to cast the issue in terms of the duties of private actors. Aoife Nolan observes how, in Juma Musjid, the Court had the option to employ strategies similar to those in Carmichele and Modderklip that did not invoke constitutional duties of private actors. Instead, however, the Court did speak of duties and found the source of these private obligations in the Constitution itself.Footnote ¹⁷¹

Some fourteen years after the Court first confronted the issue of socioeconomic rights in the Soobramoney case, amici in Juma Musjid expressed concern that this later judgment might mark a kind of apogee for the Court’s jurisprudence in socioeconomic rights. Specifically, they worried the Court would hesitate ever to apply positive rights horizontally since it only applied socioeconomic rights horizontally as a negative matter in Juma Musjid.Footnote ¹⁷² Of course, the Constitution makes no reference to a distinction between positive and negative obligations as a limit on horizontal effect, stating instead that the Court ought to consider “the nature of the right and the nature of the duty imposed by the right.”Footnote ¹⁷³ Nevertheless, distinguishing between positive and negative obligations might have offered a more robust limiting principle to those looking for one – whether to shield any vestiges of those laissez-faire priors Kriegler criticized in Du Plessis, or to offer cover for a Court that, historically, showed reticence to embrace horizontality’s republican logic to extend as far into private spaces as the early ANC vision had aspired.Footnote ¹⁷⁴ For all intents and purposes, Blue Moonlight Properties, decided only a few months after Juma Musjid, imposed positive obligations on a private actor, namely, to accommodate people occupying their property while the city arranged for alternative housing. Nevertheless, it did so without using the words “positive obligation,” ensuring that this remained a live question in South African constitutional politics until at least the 2017 case Daniels v. Scribante and likely beyond as even later cases have shown reticence to speak of positive obligations.Footnote ¹⁷⁵

As explained in Chapter 2, the Daniels judgment decided that a landlord had a (potentially positive) constitutional obligation to ensure that his tenant lived in conditions consonant with human dignity. Ms. Daniels, the tenant, proposed to improve the property in several basic ways at her own expense: leveling the floor, installing a water supply, adding a ceiling, and so on. Despite the fact that Ms. Daniels’s dwelling was basically uninhabitable, the property owner objected on the basis that he might be held responsible for compensating Ms. Daniels for these improvements upon her moving out.Footnote ¹⁷⁶ If he allowed Ms. Daniels to make these improvements, the owner argued, he assumed a duty that could only be described as positive and, thus, beyond the parameters of horizontal application. This was not simply a negative duty as in Juma Musjid, but a positive duty requiring concrete steps on the part of a landlord. The fact that this required more from a private actor than had many past cases was of no moment, however, on the Court’s reading of the Constitution – nothing in Section 8(2) foreclosed the possibility that a socioeconomic right could apply horizontally. Thus, the case hinged on balancing the landlord’s property rights and the tenant’s right to live in dignified conditions, with the Court opting for the latter given Ms. Daniels’s dire circumstances.Footnote ¹⁷⁷ The year after writing the opinion in Daniels, Justice Madlanga reflected in a lecture on the decision’s significance in unreservedly embracing the Constitution’s provision for horizontal application and his hope that the bogeyman of direct horizontal effect as portrayed in Du Plessis had “been slain for good.”Footnote ¹⁷⁸

Describing private obligations in negative or positive terms, as did Juma Musjid and Daniels, respectively, may be a semantic question in certain instances. Perhaps litigants can basically achieve their desired result regardless of how they frame the issue at hand.Footnote ¹⁷⁹ As a theoretical matter, however, this constitutes more than word choice. Indeed, whether obligations of private actors stop where the language of positive rights begins has implications for the reach of a republican interpretation of horizontal application. On the one hand, extending even negative obligations to private actors encompasses them in the communal project. Indeed, their own rights are limited as they make concessions to (negative) constitutional ends that happen to compete and carry out constitutional duties with respect to their fellow citizens. At the same time, to limit the scope of obligations to the negative reflects some effort not to interfere too drastically in private actors’ freedoms – to put limits on the limits, if you will. As explained earlier, limits are entirely predictable as judges stipulate ranges of application for jurisprudential practices. Nevertheless, this particular distinction between negative and positive duties, this potential limit on the content of rights applied horizontality, points to deeper questions (and implicit answers) of how to understand freedom, even if such issues are not actually contemplated by the judges themselves.Footnote ¹⁸⁰

As Chapter 2 explained, much scholarship conceptualizes the crucial difference between liberal and republican notions of liberty as consisting in freedom as noninterference and freedom as nondomination, respectively. From a theoretic perspective, the choice not to apply positive obligations horizontally may be read as an understanding of freedom as noninterference and reluctance to restrict individual autonomy in more intrusive ways. By extension, this position tends to deemphasize the ways private actors use autonomy to restrict other rights (such as socioeconomic rightsFootnote ¹⁸¹) that could actually be constitutive of alternative understandings of freedom, such as nondomination. As judges thus favor a practice of horizontality limited to negative duties on the one hand, or extending it to positive duties on the other, they effectively chart bounds for how far a republican interpretation might reach in a given case or context. On either understanding, we see a bringing of private actors into the communal space. The question becomes in what ways and how much of the private sphere will be encompassed into the common ends of the constitutional project. Scholars such as Sandra Liebenberg argue for the need to take seriously the relationship between socioeconomic rights and equality in the Constitution and thus to resist more traditional understandings that prioritize the negative duties that typically have pride of place in common law.Footnote ¹⁸² And yet, the foregoing cases show the extent to which the Court’s arguments still find different theoretical homes. Later developments show all the more how these theoretic questions are engendered in decisions of the Constitutional Court, as well as the implications of these for a republican interpretation.

A Return to Common Law? Attempting Another Limiting Principle

Just a few months after the Court handed down its judgment in Daniels v. Scribante, it decided Baron v. Claytile Ltd,Footnote ¹⁸³ which again raised questions of horizontal application in the context of housing and eviction. Baron followed Modderklip in the sense that the Court was hesitant to assign a private actor duties to guarantee housing rights. In particular, the Court decided that an employer did not have an obligation to continue providing former employees with housing, insofar as the Constitution designated this duty as belonging specifically to the state. To be sure, realizing a right to have access to adequate housing as a matter of fact is a major endeavor. While the primary obligation rested with the state, however, even after Baron it is conceivable that private actors be asked to cooperate in much the same way as in Modderklip and Blue Moonlight Properties. Indeed, courts still retained the ability to decide whether a particular eviction was “just and equitable.” Therefore, while the duty to provide adequate housing was technically the state’s, a private actor could functionally assume responsibility even for such positive rights as to housing.

The adjacent decisions of Daniels and Baron in the 2017 term of the Constitutional Court demonstrate how horizontal application is as much a live issue as it is a complicated one. Baron in particular shows that the Court maintains some limits on the extent to which private actors have constitutional duties. Nevertheless, even in Baron, the Court continued to operate on the basic presumption that private actors could have rights obligations. The later Pridwin Preparatory School caseFootnote ¹⁸⁴ in some ways illustrates a continued trajectory when the Court took yet a further step in applying rights horizontally to conclude that private schools had a constitutional duty to provide a basic education – specifically that administrators could not dismiss students from the school without giving due weight to the children’s right to an education, as through a hearing. The Court argued that this duty existed regardless of any specific terms to which the school and parents contracted. While the Court as a whole seems willing to apply direct horizontality more readily in Pridwin, the judgment still leaves many questions related to horizontal application unanswered and perhaps renders them more ambiguous than before.Footnote ¹⁸⁵ Much of this ambiguity may be traced back to the 2007 case Barkhuizen v. Napier and how the judgments in Pridwin employ this precedent.Footnote ¹⁸⁶

In Barkhuizen, the Court declined to apply constitutional principles directly to challenge the terms of a contract, insofar it was a private contract and not a law of general application.Footnote ¹⁸⁷ Instead, the Court considered whether the terms were reasonable and fair, ultimately giving priority to the principle of “pacta sunt servanda” (that “agreements must be kept”) rather than the competing constitutional rights that might have led to some adjustment in the common law. In his majority judgment, Justice Ngcobo explained that this principle of “pacta sunt servanda” itself “gives effect to the central constitutional values of freedom and dignity. Self-autonomy, or the ability to regulate one’s own affairs, even to one’s own detriment, is the very essence of freedom and a vital part of dignity.”Footnote ¹⁸⁸ These rights, however, are enforced simply through fulfilling the terms of the private contract and preserving existing common law understandings, as opposed to subjecting the contract to other principles arising specifically out of the 1996 Constitution.

In some ways, Pridwin may be read as assuming this position that contracts are insulated from direct constitutional scrutiny. Indeed, some critics, such as Meghan Finn, find that it presumes an even greater degree of insulation than Barkhuizen required. While the prior Barkhuizen case acknowledged the possibility that constitutional principles could apply indirectly to contracts by reshaping common law understandings, Finn explains, the Court’s judgment in Pridwin “perpetuates the sense that there are parallel systems of law: one that is shaped by and tested against the Constitution, and another that is autonomous and can be ignored, rather than constitutionally infused.”Footnote ¹⁸⁹ In other words, although the judgment in Pridwin ultimately did employ horizontal application, it did so solely on the basis that a constitutional right was invoked and, thus, explicitly in play rather than recognizing that constitutional standards might also bear in some way on the terms of a private contract. In Finn’s telling, such a binary framework insulates private contracts to an extent the prior Barkhuizen case did not require, undercutting the full transformative potential of a practice of horizontal application.Footnote ¹⁹⁰

Such doctrinal questions as these cases raise risk getting technical very quickly. To what, then, does all this amount for the purposes of this book? In many ways, this question of the status of contract law vis-à-vis constitutional values echoes the same concerns of the National Party in the negotiations, of the Supreme Court of Appeal after the establishment of the Constitutional Court, of the majority in Du Plessis v. DeKlerk. In later jurisprudential developments such as Pridwin, one can likewise read theoretical concerns about what areas of law and life are (or are not) encompassed in the public project. Even under such a transformative constitution as South Africa’s, private contracts seem to maintain a specially private status in many judgements of the Court, such that they are not so readily subsumed into the Constitution’s transformative project as other areas of law. How horizontality applies to contracts has been contested much more and much longer than the horizontal application of housing rights, for example. In the terms of this book’s republican interpretation, one might read in this history and these judgments a reticence to bring the same parity to encompass certain areas of private life into the constitutional domain. Thus, in a similar spirit to the negative–positive duties distinction discussed in the previous section, the arguments underlying these technical doctrinal developments ultimately implicate the discourses to which a republican interpretation may extend.

That there are lingering questions becomes only more apparent when reading Pridwin’s judgment alongside Beadica v. Trustees,Footnote ¹⁹¹ handed down on the very same day and authored by the same justice, Justice Theron, but undoubtedly different. Invoking the principle found in Barkhuizen of “pacta sunt servanda” (“agreements must be kept”), in Beadica, Justice Theron argues that, in fact, the contract in question ought not to be overturned on the grounds of public policy as the encroachment was not sufficiently severe. Indeed, both the functioning of society and constitutional morality depended on the ability to have one’s contracts enforced. So, the question becomes how to square these two cases, Pridwin and Beadica, seemingly similar in the issues they confront and yet engendering very different discourses. Pridwin involved the right of children to an education, a potentially crucial point given the pattern of horizonal application, and the fit of a republican reading of horizontal application, in the context of housing and education.Footnote ¹⁹² At the same time, Finn and other scholars worry about the way Pridwin confines horizontality to just these kinds of clearcut rights issues. While Beadica ultimately does not employ direct horizontal application, it also does not employ the same “parallel systems of law” that Finn fears, but instead seems to signal a return to Barkhuizen and the possibility of indirect horizontality in contracts.Footnote ¹⁹³

If contracts are a useful area to observe different accounts of the reach of South Africa’s transformative constitutionalism, then testamentary decisions may offer even more insights. To the extent that wills are among the more personal (or private) of documents one could draw up, testamentary decisions constitute important cases within which to find (or not) the kind of republican discourse with which this book is concerned. The way in which constitutional actors discuss wills and testaments relative to the constitutional project may thus reveal different conceptions of just how far into the private sphere certain actors mean to reach in applying horizontal application. And indeed, while varied in their reasoning, the judgments in the 2021 case King v. De Jager prove uncommonly receptive to a republican interpretation.

In King v. De Jager, the Court considered a 1902 will stipulating that the testator’s farms would be passed down only to male descendants for the subsequent three generations. One of the later beneficiaries bequeathed a portion of the original land to his daughters, however, leading some male descendants to file suit. The Court thus confronted the question of whether the limitation that only male descendants could inherit could be enforced in view of the Constitution’s prohibition of discrimination on the basis of gender. In three different judgments, all contrary to the prior decision of the High Court, the Constitutional Court decided that this provision could not be enforced. The first judgment, by Justice Mhlantla, relies on Barkhuizen, arguing for the development of common law and the idea that private testamentary provisions could only be enforced as they complied with public policy, including constitutional values.Footnote ¹⁹⁴ Ultimately, this amounts to a version of indirect horizontality. The majority opinion, penned by Justice Jafta, realized the same outcome that the will ought not to be enforced, but instead reached this result through consideration of Section 9(3) of the Constitution, against unfair discrimination, as well as the Equality Act of 2000, passed with the express purpose of giving greater effect to such constitutional values. Section 8 of the Equality Act understands unfair discrimination specifically as including “the system of preventing women from inheriting property” and “any practice … which impairs the dignity of women and undermines equality between women and men.”Footnote ¹⁹⁵ The third judgment by Justice Victor largely agreed with the majority, writing to emphasize the efficacy of the Equality Act as well as the role of direct horizontality in ensuring that the existing legislation does indeed protect the right in question.Footnote ¹⁹⁶

While largely tracking the reasoning of the majority judgment, Justice Victor’s concurrence offers a great deal from which to draw a republican interpretation.Footnote ¹⁹⁷ She speaks of the need to realize “substantive equality through the lens of transformative constitutionalism.”Footnote ¹⁹⁸ Likewise, she finds that this case is controlled by a capacious interpretation of equality, the Constitution’s very raison d’être,Footnote ¹⁹⁹ in contrast with the impulse to insulate the content of private wills and risk perpetuating Apartheid in private spaces.Footnote ²⁰⁰ More than a mere call for balancing competing rights, Victor’s concurrence thus articulates the republican conviction that even this exceptionally private sphere of action ought to share in the public project and not detract from constitutional ends. She tellingly concludes by connecting this vision with the principle of ubuntu – in her words, “the adage that none of us are free until all of us are free.”Footnote ²⁰¹ Victor goes on to quote Justice Mokgoro’s explanation of the Bill of Rights from a prior case as “an all-inclusive value system, or common values in South Africa” that encompasses, among other values, “group solidarity” and “conformity to basic norms and collective unity.”Footnote ²⁰² While horizontal application inevitably concerns rights, such accounts as these reveal a different take on the phenomenon – specifically that it may also give rise to certain communal ends and duties to contribute to those ends.

Conclusion

Both the German Constitutional Court and the South African constitutional framers introduced practices of horizontal application with the express intent of breaking from their respective pasts and setting course for a future governed by new constitutional commitments. In their jurisprudential histories, this move to horizontality aspires to a certain parity of governing principles across spheres. Nevertheless, the different content of these respective constitutional projects makes for different articulations and applications of horizontality. The German practice of horizontality finds expression through more classical rights and its civil law tradition. While German Drittwirkung still effected parity, initially this parity only extended so far into private spaces before private actors began to resist, as Chapter 5 explained.

On the other hand, the ANC’s constitutional vision for South Africa explicitly aimed to upset background assumptions of more traditional accounts of constitutionalism, and upset these assumptions for the purpose of effecting broad change across spheres. Put differently, the broad practice of horizontal application, even in what are traditionally considered more private spaces, aligns with the ANC’s constitutional project to transform all sectors of society. A republican logic that public values apply to private spaces exists in other areas of constitutional politics beyond horizontality as well, including statutory developments such as ESTA (the Extension of Security of Tenure Act of 1997) and the complex area of customary law. As to the latter, customary law has been elevated as a recognized source of law in South Africa. Nevertheless, the Constitution still requires parity of tribal law with constitutional principles.Footnote ²⁰³ A republican logic thus comes into sharp relief with the reformulation of rules on inheritanceFootnote ²⁰⁴ and female chiefsFootnote ²⁰⁵ according to the Constitution.

From a bird’s eye view, horizontal application seems to be expansive in South Africa relative to other countries. Of course, recent cases reveal continued complexity arising from traditional and formalist elements that persist in South African legal culture. However, when the Constitutional Court has chosen to apply horizontality, its language exemplifies a republican conviction that private entities have a role in the larger constitutional project, in contrast with traditional distinctions between public and private and, arguably, in line with the ANC’s early vision. In this new constitutional terrain, rights assume more ambitious meanings and continually alter the terms in which relationships among citizens are understood.