Constitutional Court Review Is Devoted to Analysing the REVIEW COURT CONSTITUTIONAL Work of the South African Constitutional Court in the Year Preced- Ing Publication

CCR CCR

VOLUME 1 • 2008

The annual Constitutional Court Review is devoted to analysing the REVIEW COURT CONSTITUTIONAL work of the South African Constitutional Court in the year preceding publication. Every volume contains two long essays, each with CONSTITUTIONAL replies, subject-specific articles, and case comments. This inaugural 2008 volume examines the major themes and currents reflected in COURT the decisions handed down by the Court in 2007. REVIEW VOLUME 1 • 2008 • 1 VOLUME

Lead essays/respones On the uses of interpretive charity: Some notes on application, avoidance, equality and objective unconstitutionality from the 2007 term of the Constitutional Court of South Africa CLoSA Frank I Michelman ...... 1 Constitutional Law of South Africa The Constitutional Court, court watchers and the Commons: A reply to Professor Michelman on constitutional dialogue, ‘interpretive charity’ and the citizenry as sangomas Tshepo Madlingozi ...... 63 Normative pluralism and anarchy: Reflections on the 2007 term AJ van der Walt ...... 77 Konrad Legal subsidiarity and constitutional rights: A reply to AJ van der Walt Adenauer Karl Klare ...... 129 Stiftung Articles ‘Oh, what a tangled web we weave ...’ Hegemony, freedom of contract, good faith and transformation – towards a politics of friendship in the politics of contract Jaco Barnard-Naudé ...... 155 Clearing the intersection? Administrative law and labour law in the Constitutional Court Cora Hoexter ...... 209

South African Institute for Advanced Constitutional, Public, Human Rights and International Law Case comments Sustainable development in practice: Fuel Retailers Association of Southern Africa v Director- General Environmental Management, Department of Agriculture, Conservation and Environment, Mpumalanga Province Loretta Feris ...... 235 Fuel Retailers, sustainable development & integration: A response to Feris Dire Tladi ...... 255 Cultural and religious accommodations to school uniform regulations Patrick Lenta ...... 259 The case for religious inclusivism and the judicial recognition of religious associational rights: A response to Lenta Iain T Benson ...... 295 Strap ad FINAL.indd 1 5/27/09 9:57:58 AM Media freedom and the law of privacy: NM & Others v Smith & Others Pretoria University Law Press (Freedom of Expression Institute as amicus curiae) Glenn Penfold & Dario Milo ...... 311 PULP Wanted: A principled approach to the balancing of policy considerations Steenkamp NO v www.pulp.up.ac.za Provincial Tender Board, Eastern Cape Sanele Sibanda ...... 335 Severing the umbilical cord: A subtle jurisprudential shift regarding children and their primary caregivers ISSN: 2073-6215 PULP Ann Skelton ...... 351 Editors Stu Woolman, Associate Professor, University of Pretoria Theunis Roux, Professor, University of New South Wales Danie Brand, Senior Lecturer, University of Pretoria

Editorial Board Laurie Ackermann, Constitutional Court Justice Emeritus Mary Arden, Lady Justice, Court of Appeal of England & Wales Danwood Chirwa, Associate Professor, University of Cape Town Sujit Choudhry, Professor, University of Toronto Christian Courtis, Human Rights Officer, United Nations Office of the High Commissioner for Human Rights Javier Couso, Professor, Universidad Diego Portales, Chile Charles Fombad, Professor, University of Botswana Nicole Fritz, Director, Southern Africa Litigation Centre Karthy Govender, Professor, University of KwaZulu-Natal Michelo Hansungule, Professor, University of Pretoria Karl Klare, Professor, Northeastern University Heinz Klug, Professor, University of Wisconsin Sandy Liebenberg, Professor, Stellenbosch University Frank Michelman, Professor, Harvard Law School John Mubangizi, Professor, University of KwaZulu-Natal Christina Murray, Professor, University of Cape Town Charles Ngwena, Professor, University of the Free State Enyinna Nwauche, Associate Professor, Rivers State University Cheryl Saunders, Professor, University of Melbourne Dire Tladi, Legal Counsel, South Africa, United Nations AJ van der Walt, Professor, Stellenbosch University CONSTITUTIONAL COURT REVIEW (2008) 1

2009 Constitutional Court Review (2008) 1

Published by: Pretoria University Law Press (PULP) The Pretoria University Law Press (PULP) is a publisher at the Faculty of Law, University of Pretoria, South Africa. PULP endeavours to publish and make available innovative, high-quality scholarly texts on law in Africa that have been peer-reviewed. PULP also publishes a series of collections of legal documents related to public law in Africa, as well as text books from African countries other than South Africa.

For more information on PULP, see www.pulp.up.ac.za

Printed and bound by: ABC Press Cape Town

To order, contact: PULP Faculty of Law University of Pretoria South Africa 0002 Tel: +27 12 420 4948 Fax: +27 12 362 5125 [email protected] www.pulp.up.ac.za

Cover: Yolanda Booyzen, Centre for Human Rights

ISSN: 2073-6215

The Constitutional Court Review forms part of the Rule of Law in Africa Project funded by the World Bank. CONSTITUTIONAL COURT REVIEW

The Constitutional Court Review is a once-a-year journal dedicated to the analysis of the Constitutional Court’s decisions of the previous year. Its purpose is to provide a platform for high-level academic engagement with the jurisprudence of the South African Constitutional Court.

To this end, each issue of the Review contains two lead essays exploring broad themes arising from a given year’s jurisprudence (each ± 20 000 words), each with its own response (± 5 000 words); a number of shorter subject-specific articles (each ± 10 000 words); and several case comments that engage more narrowly with a given decision of the Constitutional Court (each ± 5 000 words).

Lead essays are solicited by the editors, as are some of the subject- specific articles and case notes, but for the remainder unsolicited contributions are invited. Such contributions must be sent to the editors at [email protected] in MS Word format on or before 31 May of the year following that on which a contribution focuses. Contributions will only be considered if they follow the house style, available at www.pulp.up.ac.za.

iii SUBSCRIPTION

Subscriptions through agents or bookshops: R380-00 (including VAT but excluding postage and packaging)

Subscriptions directly through PULP: R292-00 (including VAT but excluding postage and packaging)

Subscriptions by registered students directly through PULP: R220-00 (including VAT but excluding postage and packaging)

Postage and packaging charges are: R45-00 per subscription

Subscriptions should be directed to the publisher: PULP - Pretoria University Law Press Faculty of Law University of Pretoria Pretoria 0002 E-mail: [email protected] Website: http://www.pulp.up.ac.za

iv TABLE OF CONTENTS

Lead essays/responses

On the uses of interpretive charity: Some notes on application, 1 avoidance, equality and objective unconstitutionality from the 2007 term of the Constitutional Court of South Africa Frank I Michelman

The Constitutional Court, court watchers and the Commons: 63 A reply to Professor Michelman on constitutional dialogue, ‘interpretive charity’ and the citizenry as sangomas Ts h e p o M a d l i n g o z i

Normative pluralism and anarchy: Reflections on the 2007 term 77 AJ van der Walt

Legal subsidiarity and constitutional rights: A reply to 129 AJ van der Walt Karl Klare

Articles

‘Oh, what a tangled web we weave ...’ Hegemony, freedom of 155 contract, good faith and transformation — towards a politics of friendship in the politics of contract Jaco Barnard-Naudé

Clearing the intersection? Administrative law and labour law 209 in the Constitutional Court Cora Hoexter

Case comments

Sustainable development in practice: Fuel Retailers Association 235 of Southern Africa v Director-General Environmental Management, Department of Agriculture, Conservation and Environment, Mpumalanga Province Loretta Feris

Fuel Retailers, sustainable development & integration: 255 A response to Feris Dire Tladi

v Cultural and religious accommodations to school uniform 259 regulations Patrick Lenta

The case for religious inclusivism and the judicial 295 recognition of religious associational rights: A response to Lenta Iain T Benson

Media freedom and the law of privacy: NM & Others v Smith & 311 Others (Freedom of Expression Institute as amicus curiae) Glenn Penfold & Dario Milo

Wanted: A principled approach to the balancing of policy 335 considerations. Steenkamp NO v Provincial Tender Board, Eastern Cape Sanele Sibanda

Severing the umbilical cord: A subtle jurisprudential shift 351 regarding children and their primary caregivers Ann Skelton

vi ON THE USES OF INTERPRETIVE ‘CHARITY’: SOME NOTES ON APPLICATION, AVOIDANCE, EQUALITY AND OBJECTIVE UNCONSTITIONALITY FROM THE 2007 TERM OF THE CONSTITUTIONAL COURT OF SOUTH AFRICA*

Frank I Michelman*

Charity is forced on us; whether we like it or not, if we want to understand others, we must count them right in most matters.1

1Introduction

To hear Stu Woolman tell it, disturbing lapses and weaknesses — an apparent ‘lack of analytical rigour’ suggesting what could be a ‘penchant for outcome-based decision-making’ — have been showing up recently in the work of a Constitutional Court whose prior record of performance has deservedly garnered widespread applause.2 Woolman cites as evidence three decisions from the Court’s work in the year 2007: Barkhuizen,3 Masiya,4 and NM,5 and suggests that his reactions to these decisions are widely shared among South Africa’s well-informed Court-followers.6 He makes a worthy, illuminating, formidable case, one that the Court would do well to consult and ponder.

* I am indebted to Stu Woolman, and to two anonymous referees, for exceptionally helpful, trenchant comments. I hope that what follows shows the benefit of their efforts, even if I remain incorrigible on some points of difference. ** Robert Walmsley University Professor, Harvard University. 1 D Davidson Inquiries into truth and interpretation (1984) 197 (Davidson Truth). 2 S Woolman ‘The amazing, vanishing Bill of Rights’ (2007) 123 South African Law Journal 762 (Woolman ‘Amazing’). 3 Barkhuizen v Napier 2007 7 BCLR 691 (CC). 4 Masiya v Director of Public Prosecutions 2007 5 SA 30 (CC); 2007 8 BCLR 927 (CC) (‘Masiya’). 5 NM v Smith 2007 5 SA 250 (CC); 2007 7 BCLR 751 (CC) (‘NM’). 6 Woolman ‘Amazing’ (n 2 above) 762 (‘chattering classes’).

1 2 Uses of interpretive ‘charity’

Formidable is not, however, conclusive. For reasons I shall come to shortly, I have chosen to devote this space to seeing what might be said on the other side, specifically with regard to NM and Masiya, on behalf of a Court that, I quite agree with Woolman, has left itself with a lot of explaining to do. The controlling opinions in these cases are indeed, as Woolman says, ‘thinly reasoned’,7 if by that we mean they are in some respects insufficiently explained. It is, however, another question whether these cases have been wrongly or irresponsibly managed, as measured by reasonably discoverable, valid considerations of law and legal administration.

In particular, I shall be questioning Woolman’s diagnosis from these cases of ‘a court uncomfortable with the direct application of the specific substantive provisions of the Bill of Rights’ and ‘in full flight from any meaningful engagement with Chapter 2 of the Constitution.’8 Whether a wider survey of the jurisprudence would warrant an over-all diagnosis of an excessive flight from substance is a question on which I hazard no judgment here. All I say here is that NM and Masiya do not, to my eye, support the diagnosis, nor is that, in my view, the best way for us to regard these cases. One feature common to both is the Constitutional Court’s seeming gravitation to its inherent power to develop the common law in terms of Constitution sections 173 and 39(2) — as opposed to its judicial review power in terms of sections 8 and 172(1) — when undertaking modification of common law rules under pressure from the Bill of Rights. Woolman believes the Court moves too freely to the inherent power. He associates that tendency, as symptom or cause (or both), with excessive flight from substance. I aim to raise a doubt about any such connection.9

The main controversy over Masiya appears to me to turn, at bottom, more on a point of substantive disagreement between the Court and Woolman than on any notable disregard for Bill-of-Rights substance on the Court’s part. The controversy over NM is more complicated, and more centrally my concern in these pages. It certainly is true that Woolman and the Court divide over when, if ever, the Constitutional Court ought to resort to an ‘indirect’ instead of a ‘direct’ application of the Bill of Rights to a common law doctrine or rule — the Court, in Woolman’s view, making far too much use of

7 Woolman ‘Amazing’ (n 2 above) 762, 790 n 51. 8 Woolman ‘Amazing’ (n 2 above) 783. 9 I have nothing to say about another symptom or cause of flight that some might mention, to wit, the Court’s readiness on some occasions to move to the ‘justification’ phase of a Bill of Rights case either without having decided the question of a substantive infringement, see, eg, Christian Education South Africa v Minister of Education 2000 4 SA 757 (CC) para 27, or, perhaps, having found an infringement on the basis of a merely ‘notional’ reading of the right in question. See S Woolman & H Botha ‘Limitations’ in S Woolman et al (eds) Constitutional law of South Africa (2nd Edition, OS, 2005) ch 34 16-18. (2008) 1 Constitutional Court Review 3

‘indirect’. Such a division need not, however — or so I shall contend — reflect any reduced or absent sense on the Court’s part of responsibility to engage with the substance of Chapter Two and its several, rights-naming clauses. It might rather come down to a question of doctrinal good-housekeeping on which nothing of substance depends. The Court and Woolman are differing, I shall suggest, over how best to understand and sort out the respective offices of the Constitution’s two paths to judicial revision of the common law under constitutional pressure: revision as a remedy for constitutional violation pursuant to sections 172(1)10 and 8,11 and revision in the exercise of judicial powers to develop the common law, with a view to promoting the spirit, purport, and objects of the Bill of Rights in terms of section 39(2). That filing-system question, I shall maintain, is entirely distinct from the one about when and how regularly the Court regards itself as on or off the hook for an elucidation of one or another of the specific clauses in the Bill of Rights.

In developing these claims, I shall be quite openly engaged in filling in passages of exposition and explanation that are missing from the Court’s opinions in NM and Masiya, to a degree that may sometimes strike readers as excessively indulgent of the Court, if not as entirely fanciful. In construing and re-presenting the work of the Constitutional Court in these cases, I take myself to be following something akin to what linguists and language-philosophers have called a ‘principle of charity’. ‘Something akin,’ not the genuine article, for this is not a work of philosophy, but rather an intended contribution to a lawyers’ kibitz on the work of the Constitutional Court.

The ‘principle of charity’, Wikipedia tells us,

10 172.(1) When deciding a constitutional matter within its power, a court — (a) must declare that any law or conduct that is inconsistent with the Constitution is invalid to the extent of its inconsistency; and (b) may make any order that is just and equitable ... 11 8.(1) The Bill of Rights applies to all law, and binds the legislature, the executive, the judiciary and all organs of state. (2) A provision of the Bill of Rights binds a natural or a juristic person if, and to the extent that, it is applicable, taking into account the nature of the right and the nature of any duty imposed by the right. (3) When applying a provision of the Bill of Rights to a natural or juristic person in terms of subsection (2), a court — (a) in order to give effect to a right in the Bill, must apply, or if necessary develop, the common law to the extent that legislation does not give effect to that right; and (b) may develop rules of the common law to limit the right, provided that the limitation is in accordance with section 36(1) ... 4 Uses of interpretive ‘charity’

is an approach to understanding a speaker’s statements by interpreting the ... statements to be rational and, in the case of any argument, rendering the best, strongest possible interpretation of an argument.12

Donald Davidson, surely one of the principle’s chief philosophical architects and expositors, also calls it ‘the principle of rational accommodation,’ and summarises as follows: ‘We make maximum sense of the words and thoughts of others when we interpret in a way that optimises agreement.’13 Davidson meant ‘optimise’ as between thinking that the other must be holding to beliefs (and, relatedly, aims) that differ drastically from our own (else he couldn’t have said what he did), and thinking that we must not have heard him right the first time.14

The philosophical underpinnings of the charity principle are far afield from our concerns here and need not delay us. All that matters from that department is the motivation for adherence to the principle, which the moniker ‘charity’ does not very well convey. The aim of interpretive charity is not generosity toward others, or anything like that. It is not to pay homage, deference, or respect to our interlocutors, or to avoid giving offense. It is not to demonstrate our own good manners, or to toe some Goody Two-Shoes line against critiques that are not ‘constructive’. (I hold Stu Woolman’s pull-no- punches style of court-watching to be entirely constructive and admirable.) No, the aim of ‘charitable’ interpretation is not any of those. The aim is to learn. It is aggressively to learn what there is to be learnt from puzzles the interlocutors pose to us, by assuming there is method in their madness and doing our best to ferret that out, using everything else we know or can guess (in part from their likeness and kinship to us) about where they are coming from. ‘To see too much unreason on the part of others’, Davidson says, is ‘to undermine our ability to understand what it is they are so unreasonable about.’15 It is to risk missing issues that might merit our consideration.

There are hazards — normative pitfalls — in the way of such an approach to the construction of judicial opinions. For one: these writings enjoy among us the status of utterances of law by a socially recognised organ of authority to say what the law is. And maybe, therefore, it really does not do for us to go about pretending the writers have said things they did not mean to say (perhaps meant not to say), in our effort to make their writings accord with what we take

12 http://en.wikipedia.org/wiki/Principle_of_charity (accessed 21 June 2008). 13 As above. 14 ‘[T]here is no deciding ... between the view that the Other has used words as we do but has more or less weird beliefs, and the view that we have translated him wrong.’ Davidson Truth (n 1 above) 101. 15 Davidson Truth (n 1 above) 153. (2008) 1 Constitutional Court Review 5 to be legal reason — it being they, after all, and not we, who are the authorised law-sayers.

Lying just ahead are deep jurisprudential waters into which I do not choose to wade on this occasion. I mean only to grant up front the possible dangers from a ‘charitable’ approach, while pointing also to some possible gains. A close examination of NM and Masiya, from a methodological stance of all-but-last-ditch resistance against a conclusion of blatant error, failure of nerve, or flight from responsibility on the Court’s part, can demonstrate, I believe, some benefits of interpretive charity as a working disposition on our part — not necessarily suited, of course, to every moment in history or every judicial performance we may encounter. Such an examination can show how that posture may sometimes aid us in performing our task of getting at the truth of law or at any rate — more modestly — of developing culturally credible resources for the possible guidance of judicial performance in the future, with no less a view (here following Theunis Roux) to the judiciary’s ‘legal’ than to its ‘sociological’ and ‘institutional’ legitimacy.16

2 ‘Direct’ and ‘indirect’ application

Much of what is to follow turns on debates over when, if ever, the Constitutional Court ought to resort to an ‘indirect’ instead of a ‘direct’ application of the Bill of Rights to a common law doctrine or rule. Such debates sometimes risk falling into confusion and misunderstanding, because not everyone conceives or defines the distinction in exactly the same way, and perhaps some of us do not conceive or define it uniformly at every turn or in every context. It will, therefore, behoove us to survey briefly some of the variant senses of this distinction, and, in particular, to specify what Woolman — consistently — means by it in the work of his to which what follows is something of a rejoinder.

So far as I am aware, the distinction in South African legal discourse between direct and indirect application of the Bill of Rights

16 T Roux ‘Principle and pragmatism on the Constitutional Court of South Africa’ (2008) 7 I-CON 106 (Roux ‘Pragmatism’). In Roux’s vocabulary, the ‘legal legitimacy’ of judicial review rises and falls with perceptions of a court’s capacity to decide cases according to forms of reasoning acceptable to the legal community of which it is a part. Legal legitimacy refers to ‘the plausibility (rather than correctness) of a judicial decision or body of decisions according to applicable standards of legal reasoning.’ Thus defined, Roux explains, legal legitimacy is to be distinguished from two other sociological and institutional dimensions of the judiciary’s public standing, which Roux names ‘public support’ and ‘institutional security’ meaning the court’s capacity to resist political attacks on its independence. As above 109. Roux posits that a court presiding over a system of judicial constitutional review does well to act with an eye to all three dimensions of legitimacy. See as above 110-11. 6 Uses of interpretive ‘charity’ first arose in regard to the question of so-called horizontal application of the Bill of Rights under the Interim Constitution (although, as we shall see, it has outgrown its context of origin). In terms of that debate, to apply the Bill of Rights ‘horizontally’ was to bring it to bear decisively, in a case in which (a) there was no state party and (b) it appeared that the pivotal legal doctrine or rule was (or might be) a rule or doctrine of the common law, as opposed to the terms of a statute.17 In such a case, either or both of the following would have to be true: A plaintiff bases a claim for relief on a rule or doctrine of the common law, and/or a defendant bases a defense (including the defense of ‘absolution’ or ‘no cause of action’) on a rule or doctrine of the common law. One or the other (or both) of the parties might question the compatibility with the Bill of Rights of the common law rule or doctrine on which the other relies. If that occurs, and the court rules in that party’s favor on that basis, a horizontal application of the Bill of Rights will have taken place, under the definition we are now considering.

Consider, now, a case such as NM, in which the plaintiff charges the defendant with unauthorised disclosure of private information concerning the plaintiff, and the defendant seeks to respond by proving that she did not knowingly or intentionally publish without authorisation, but rather acted honestly (if mistakenly) in the belief that authorisation had been given. How might the Bill of Rights be brought to bear on this case, specifically section 14, which guarantees a right of privacy to everyone?

A court might proceed by drawing a standard of care from section 14, and then measuring the defendant’s conduct by that constitutionally prescribed standard, with a view to holding the defendant liable for a delict if her conduct fails the constitutional test. Alternatively, a court might proceed by asking whether the extant common law rule — which absolves publishers of private information from tort liability where they did not intentionally or knowingly publish without authorisation — is consistent with section 14, with a view to possibly revising the common law rule in such a way that the defendant will be found liable for tortious conduct under the common law as revised to make it ‘constitutional’. In the first instance, the court applies the Bill of Rights to conduct; in the second instance, it applies it to law.

17 See Du Plessis v De Klerk 1996 3 SA 850 (CC); 1996 5 BCLR 658 (CC) para 49 (Kentridge AJ) (‘Du Plessis’) (in effect, defining a ‘vertical’ case as one that either involves a state party or turns on a statute); S Woolman ‘Application’ in S Woolman et al (eds) Constitutional law of South Africa (2nd Edition, OS, 2005) ch 31 (Woolman ‘Application’) 18 (similarly describing ‘the traditional, vertical, view’); 4 (reporting general agreement that ‘the heart of the application debate was whether the common law, when relied upon by a private party in a private dispute, was subject to constitutional review.’). (2008) 1 Constitutional Court Review 7

This distinction — between testing conduct by a Bill of Rights standard, and testing law by a Bill of Rights standard — may have been what some early deployments of the direct/indirect distinction had at least partly in mind: application of the constitutional standard to conduct is ‘direct’ application, whereas application of it to law is ‘indirect’ application. If so, that usage of the distinction cannot survive today. As we shall see below, Constitution section 8(3) virtually rules out the immediate application of a Bill of Rights standard to any private person’s conduct, directing rather that all such applications be made, instead, to the legal rules and doctrines on which litigating parties rely to found their claims and defenses.18

Consider, then, another sort of choice that might confront a court in our invasion-of-privacy case, where the court is already committed to test only law, not conduct, under the Bill of Rights. A court disposed to rule in the plaintiff’s favor, despite the defendant’s proven defense of absence of intention to publish without authorisation, might proceed by first holding the extant, delictual law of privacy invalid by reason of inconsistency with a specific right guaranteed by some rights-granting clause in the Bill of Rights — here it would be section 14 (‘Everyone has the right to privacy ...’) — and then, by way of remedy for the unconstitutionality, directing a curative revision or ‘development’ of the common law. As explained further below, the court then would be acting under the aegis of Constitution section 172(1), or of Constitution section 8, or (perhaps it would be best to say) of both of those sections. Alternatively, the court might take its cue from Constitution section 39(2). It might then by-pass the question of the extant law’s possible invalidity by reason of inconsistency with a specific, rights-granting clause, and simply find the extant delictual doctrine unacceptably out of sorts with the overall ‘spirit, purport, and objects’ of the Bill of Rights. We might classify the first sort of judicial action as ‘direct’ application of the Bill of Rights, and the second sort as ‘indirect’.

We then would have severed any special connection of the direct/ indirect distinction to cases involving horizontal application of the Bill of Rights, because the distinction as now defined is equally applicable

18 But the conduct/law distinction does survive in transmogrified form to post- section 8 debate. See, eg, H Cheadle ‘Application’ in H Cheadle, D Davis & N Haysom (eds) South African constitutional law: The Bill of Rights (2002) 19: ‘ ... [B]ecause generally stated rights are not appropriate vehicles for the imposition of standards of conduct, section 8 seeks to avoid application of the right to conduct of private persons by requiring either legislation or a common law rule ‘to give effect to the right.’ This constitutional motif of an intermediate law between the constitutional right and the conduct of the state or of the private actor repeats itself throughout the Bill of Rights.’ Woolman takes Cheadle to be saying, contentiously, that ‘the Bill of Rights rarely, if ever, applies directly to “conduct”. It applies to law.’ Woolman ‘Application’ (n 17 above) 155. 8 Uses of interpretive ‘charity’ to cases that are incontestably vertical in nature.19 Take, for example, a case, which surely must be classed as vertical, in which the state prosecutes a person for alleged commission of the common law crime of sodomy. When that case came before the Constitutional Court, the Court employed the form of application of the Bill of Rights that we are now calling ‘direct’, so as to invalidate the crime of sodomy and erase it from the common law of South Africa.20 Had the Court seen fit to do so, it could have achieve the same result ‘indirectly’, by holding that the common law of crime must be ‘developed’ so as to omit the crime of sodomy, in order to stay in tune with the overall spirit etc of the Bill of Rights.

To be as precise as possible about where we have come, we now have before us the following three, live possibilities:

(a) ‘direct’ = declaration of invalidity followed by remedial development of the common law; ‘indirect’ = development of the common law to keep it attuned to the spirit etc. of the Bill of Rights, with no attendant declaration of invalidity. (b) ‘direct’ = a finding of the extant common law’s inconsistency with norms or standards contained in one or another specific rights-granting clause; ‘indirect’ = no such finding, but only a finding of disharmony with the general spirit etc. of the Bill of Rights taken whole. (c) ‘direct’ = judicial action in terms of Constitution sections 172(1), 8; ‘indirect’ = judicial action in terms of section 39(2).

When Stu Woolman speaks of ‘direct’ versus ‘indirect’ application of the Bill of Rights, he tends to run these oppositions together. No doubt opposition ‘b’ must stand as his core, his official definition of the distinction.21 But Woolman sustains no effort to pry or to hold the three oppositions apart. Rather, he sometimes treats the three as a cluster, in which each opposition in some way entails, insinuates, attracts, or bleeds into both of the other two. Such a treatment flows naturally from Woolman’s strongly held, carefully argued normative view that South African jurists ought to follow a discipline of holding the oppositions in fixed alignment. Woolman’s construction is captured in the table below.22

19 See Woolman ‘Application’ (n 17 above) 5 n 1 (observing that the ‘application’ debate must now be framed in terms of ‘direct’/’indirect’, not ‘vertical’/ ’horizontal’). 20 See National Gay & Lesbian Coalition v Minister of Justice 1996 3 SA 850 (CC); 1996 5 BCLR 658 (CC) paras 73, 90 (Sodomy Case). 21 ‘Direct challenges describe instances in which the prescriptive content of at least one specific substantive provision of the Bill of Rights applies to the law or to the conduct at issue. Indirect challenges describe instances in which the prescriptive content of no specific provision of the Bill of Rights applies to the law or to conduct at issue.’ Woolman ‘Application’ (n 17 above) 5 n 1. 22 See generally Woolman ‘Application’ (n 17 above). (2008) 1 Constitutional Court Review 9

‘direct’ (1) ‘indirect’ (2) form of judicial power judicial-review power inherent power to (a) to declare invalidity develop common law and prescribe remedy in normal course of involving revision of adjudication [2a] common law [1a]

substantive test of is common law is common law inconsistency (b) inconsistent with inconsistent with requirements of a general objects of Bill specific right in the of Rights, even though Bill of Rights? [1b] not inconsistent with requirements of any specific rights? [2b]

controlling section of 172(1) and/or 8 [1c] 39(2) [2c] Constitution (c)

According to what Woolman urges as the best, overall, combined construction of the Constitution’s relevant texts and purposes, whenever a court chooses to proceed — or finds itself proceeding — in terms of any of 1a, 1b, or 1c, it ought to require itself concomitantly to proceed, on that occasion, in terms of other two members of triplet 1; and conversely, of course, for any court that finds itself proceeding in terms of any of the members of triplet 2.

Whether Woolman’s normative/textual arguments in favor of his preferred construction, and of a correspondingly disciplined juristic practice, entirely succeed is not our question here. My question will rather be whether the prism of Woolman’s preferred construction works to the overall advantage of clarity of analysis in approaching the work of the Constitutional Court. If the Constitutional Court should happen to be working with a different (and perhaps itself defensible) construction, then it may not. To be more specific: Suppose that (as I shall suggest the evidence shows) the Constitutional Court is quite comfortable combining 2a and 2c with 1b. Woolman finds that combination wrongheaded, or, at any rate suboptimal from the standpoint of assigning clear, distinct, and jurisprudentially cogent meanings to the several, relevant constitutional clauses. Call that the Court’s ‘construction error’. The construction error — 2a+1b+2c — cannot be counted as contributory towards a flight from meaningful engagement with the substance of the rights-naming provisions of Chapter Two, because one of its components — 1b — is entirely congenial to what Woolman means by engagement. The construction error and the ‘flight error’ (as we may call the latter) are two different things. Woolman sometimes seems to me to be detecting the flight error where only the construction error is clearly in evidence. 10 Uses of interpretive ‘charity’

3 NM v SMITH

3.1 The case

Charlene Smith wrote a book about the public career of Patricia De Lille, an MP and a public figure. The book recounted De Lille’s interventions on behalf of persons living with HIV/AIDS. These included an episode involving controversial drug trials at the University of Pretoria, in which the three women who became the plaintiffs in the case of NM — ‘NM’, ‘SM’, and ‘LH’ — were among the volunteer subjects. The three plaintiffs figure significantly in the book’s account of the controversy, as persons living with HIV. They are in no way otherwise persons of the slightest public note nor would their names be recognisable by members of the public outside their own special circles of acquaintance.

Smith learned the plaintiffs’ names from a copy of an in-house report of a professor’s investigation of the medical-study controversy, conducted on behalf of the University and intended for restricted circulation. (No one complains of the means by which Smith obtained her copy.) The report refers to ‘annexures’ containing forms signed by the three plaintiffs, giving consent to the inclusion of their names in the report. The copy provided to Smith did not include the annexures. Smith’s book, when published, identified the plaintiffs by name as persons living with HIV, as the report had done. Contrary to Smith’s expectation, it emerged that the consents given by the plaintiffs to the report’s author, Professor Strauss, were not general releases to disclose their medical information, with names attached, to the public at large, but only authorised a limited disclosure for purposes related to the University’s investigation.

The plaintiffs brought an action against Smith for damages, claiming that her book’s identification of them by name, without their consent, as persons living with HIV amounted to a violation of their common law rights to privacy, specifically in terms of the actio iniuriarum.23 Under the common law as it then stood (and still stands, in the wake of the Constitutional Court’s decision in NM), the plaintiffs could succeed only if Smith were found to have acted toward them in a manner that was not only objectively wrongful but also intentionally so.24 Smith contended that any wrong she might have committed toward the plaintiffs, by way of publishing their

23 In order to maintain clarity, I focus my examination of NM on the claims for damages against defendant Smith; I do not discuss the claims for other forms of relief, or the claims for damages against defendants Patricia De Lille and New Africa Books Ltd. 24 See NM (n 5 above) para 55 (Madala J); para 151 (O’Regan J). (2008) 1 Constitutional Court Review 11 private information without their consent, was unintentional. The crux of her defense was that she acted in the honest (if mistaken) belief that the plaintiffs’ consents, of which she read in the Strauss Report, covered unrestricted circulation of the information regarding them contained in the report. Apparently persuaded by Smith’s testimony to this effect, the High Court absolved Smith of liability for tortious invasion of privacy.25 The Supreme Court of Appeal refused review, giving no reasons, and the plaintiffs took their case to the Constitutional Court.26

Three members of the Court — Langa CJ, O’Regan J, and Sachs J — concluded that the case against Smith disclosed a need to develop the common law so to remove an incompatibility with the objects of the Bill of Rights in the matter of protecting privacy. These members would have developed the common law so as to impose liability, on ‘media defendants’ only, for negligent (or unreasonable), wrongful disclosure of private information.27 The Court’s majority took a different tack. They joined an opinion by Madala J, holding that Smith’s unauthorised disclosure of the plaintiffs’ private information should be found intentional in terms of the extant common law standard of liability.28 They accordingly ordered a judgment for damages against Smith29 while declining to address any question of developing the common law.30 Of the three justices who thought the common law must be developed, Langa CJ and Sachs J concurred in the order awarding damages to the plaintiffs, on the ground that the record showed negligence on the part of Smith.31 O’Regan J would have absolved Smith of negligence and, hence, of liability.32

3.2 A Critique

Woolman’s objections to the Constitutional Court’s performance in NM are mainly directed to the decision and opinion of the majority. While the objections refer to specific, alleged deficiencies and missteps in this particular case, Woolman presents them as exemplary of broader, worrisome trends in the work of the Constitutional Court. These are most compendiously summed up as a ‘lack of analytical rigour’ suggestive of ‘a penchant for outcome-based decision- making.’33 In Woolman’s view, the decision in NM ‘appears to rest

25 See NM (n 5 above) para 46 (Madala J); para 125 (O’Regan J). 26 See NM (n 5 above) para 19. 27 See NM (n 5 above) paras 92-94 (Langa CJ); paras 170-19 (O’Regan J); paras 203- 04 (Sachs J). 28 See NM (n 5 above) paras 58-65. 29 See NM (n 5 above) para 90. 30 See NM (n 5 above) para 57. 31 See NM (n 5 above) para 92 (Langa CJ); para 207 (Sachs J). 32 See NM (n 5 above) para 189. 33 Woolman ‘Amazing’ (n 2 above) 762. 12 Uses of interpretive ‘charity’ upon a deeply-felt offence to the majority’s moral sensibility about how vulnerable persons in our society ought to be treated.’34

Woolman’s bill of particulars includes the following:

(1) The majority ended up acting ‘as a trier of fact in a run-of-the-mill actio iniuriarum matter’35 — a sort of task, Woolman evidently means to say, that the Constitutional Court was not created to do. (2) The majority’s crucial factual determination regarding Smith’s intention is ‘contrary to the evidentiary record’, which ‘cannot support a factual finding that the respondents had acted intentionally to harm the privacy and the dignity interests of the applicants.’36 (3) The majority failed to provide an adequate explanation for why, once they had determined that the case would be disposed of by applying an unmodified, long-established common law standard of liability, they did not dismiss the appeal as being no proper concern of the Constitutional Court.37 (4) The majority diminished the Constitution by declining to take up for consideration a possible ‘challenge to the actio inuriarum grounded in a specific substantive provision of the Bill of Rights’ or the possible ‘creation of a self-standing constitutional action grounded in the right to privacy or the right to dignity.’38 This objection echoes prior pleas from Woolman, for a shift of the Court’s emphasis away from ‘indirect’ application of the Bill of Rights pursuant to section 39(2) to ‘direct’ application pursuant to section 8.39 The Court’s penchant for section 39(2) is deplorable, Woolman maintains, because it plays into the false seductions of a so-called minimalist approach to constitutional adjudication that is unsuited to the present, early state of development of South African constitutional jurisprudence and culture.40 It invites the Court to issue highly contextualised decisions that may feel right to the justices and many of those looking on, thus avoiding or postponing intracurial division and public controversy, but that remain unexplained on the level of principled, clause-by-clause exposition of the meanings of the rights-guarantees in Chapter 2 and the values underlying them.41 Yet the Court’s presumed, special capacity for that more challenging — in part because potentially divisive — task of principled explication is its raison d’être par excellence. Performance of that task, Woolman perceives, is a service to the country that only the Constitutional Court

34 Woolman ‘Amazing’ (n 2 above) 787. 35 Woolman ‘Amazing’ (n 2 above) 783. 36 Woolman ‘Amazing’ (n 2 above) 781. 37 See Woolman ‘Amazing’ (n 2 above) 782 & n 40; Constitution of the Republic of South Africa, 1996 sec 167(3)(b) (The Constitutional Court ‘may decide only constitutional matters, and issues connected with decisions on constitutional matters’). 38 Woolman ‘Amazing’ (n 2 above) 783. 39 See generally Woolman ‘Application’ (n 17 above). 40 Woolman ‘Amazing’ (n 2 above) 764-65 n 4, 784-87. 41 Woolman ‘Amazing’ (n 2 above) 784-86. (2008) 1 Constitutional Court Review 13

can provide, and that it must perform for the sake of the rule of law in South Africa.42 (5) In sum, the majority’s narrow, non-venturesome treatment of the case shirked a sort of service to the country that the Court was created to provide, to wit: authoritative delineation and explication of the principles animating the several specific substantive rights in Chapter 2, conveyed with definition sufficient to ‘determine the actual validity of the rule being challenged in the instant matter and of similar rules challenged in subsequent matters’; also to enable citizens and government officials to ‘ensure that their behaviour conforms to our Constitution’, and also to equip lower courts and lawyers to ‘identify the law and thereby settle, litigate and adjudicate, with some confidence, Bill of Rights cases.’43 NM, Woolman concludes, ‘shows a court in full flight from any meaningful engagement with Chapter 2 of the Constitution.’44

Woolman’s observations are eminently worth pondering, and by no means without warrant in the majority opinion in NM, as written. I mean to show, however, how the complaints turn on matters of interpretation that can be resolved in the Court’s favour — with, no doubt, varying degrees of effort from the reader. The opinion in NM is indeed thin and chary of explanation, in places where it could and should be thicker and more forthcoming. Contrary to Woolman’s view, however, I think the NM majority got to the right answer, by a proper path (although the path preferred by Langa CJ seems to me superior), and for essentially (and detectably) right reasons. I say so with utmost respect to the opinion in the case that most ambitiously addresses the task of substantive elucidation of constitutional rights that Woolman misses from the Court’s work, the dissenting opinion of O’Regan J. Admirable as that opinion is in many ways, in the end it decides the case wrongly, in my view. Moreover, it perhaps does so for causes not unrelated to O’Regan J’s evident commitment to (and talent for) the work of principled exposition of the contents of fundamental rights.

3.3 A pleading choice

In order to succeed with their case against Smith, the NM plaintiffs had to show that Smith’s conduct not only had harmed them in fact, but had done so in a way that was liability-engendering, according to the terms of some law meant to protect persons in the plaintiffs’ position against the sort of harm at issue. Of course, it would be up to the plaintiffs to say what law they had in mind. Their answer was: the common law of delict, and specifically the actio iniuriarum.45

42 See Woolman ‘Amazing’ (n 2 above) 786-87. 43 Woolman ‘Amazing’ (n 2 above) 763. 44 Woolman ‘Amazing’ (n 2 above) 783. 45 See NM (n 5 above) paras 22, 27-29. 14 Uses of interpretive ‘charity’

A different response was also imaginably available. The plaintiffs might have pleaded as their cause of action Constitution section 14 (‘Everyone has the right to privacy’), as rendered applicable to Smith’s conduct by Constitution section 8(1) (‘The Bill of Rights applies to all law, and binds the legislature, the executive, the judiciary and all organs of state’) and (8(2) (‘A provision of the Bill of Rights binds a natural ... person if, and to the extent that, it is applicable ...’).46 Rather than (or in addition) to claiming damages for a common-law tort, the plaintiffs might have claimed on the basis of a constitutional tort. They did not do so, and it will repay us to ask why they did not. The common-law path selected by the plaintiffs was not, after all, free and clear of evident obstacles. The plaintiffs (or their lawyers) would have known the list of conventionally established requirements for success in a claim for invasion of privacy based on the actio iniuriarum.47 They would have known that among these was a showing of the defendant’s intention (‘animus iniurandi’) to commit, as against the plaintiff, the wrong consisting of a publication of private facts concerning the plaintiff, without the plaintiff’s leave or against her will — where the category of ‘private’ facts includes any and all facts, not yet publicly available, whose disclosure to the public would be expected to ‘cause mental distress and injury to anyone possessed of ordinary feelings in the same circumstances’.48

Consider, then, a defendant who publishes a plaintiff’s entire, anguished medical history (naming her), in the honest but false belief that the plaintiff has given leave for publication, or that the information has been previously (lawfully) published and hence is no longer private. Such a defendant will indeed have committed the wrong in question, ‘objectively’ described as above; her action will, in that sense, have been wrongful under the law. Missing from the case, however, will be the usual basis for an inference of the defendant’s intention to mistreat (‘wrong’) the plaintiff in that way — to wit, the defendant’s actual knowledge of all of the facts that, in combination, would render her action wrongful.

To put the matter simply: Under the common law of delict as it stood when the plaintiffs filed their case against Smith, negligent but non-intentional commission of the objective wrong of unauthorised publication of private facts was not deemed legally actionable, ‘as a

46 See NM (n 5 above) para 132 (O’Regan J) (confirming the applicability of Constitution sec 14 to private actors). 47 See NM (n 5 above) para 55 (Madala J); para 151 (O’Regan J). 48 NM (n 5 above) para 34 (Madala J), citing National Media Ltd v Jooste 1996 3 SA 262 (A); 1996 2 All SA 510 (A). O’Regan J preferred to leave open whether the Constitution might possibly require some expansion of this common-law test for the protected (‘private’) status of the facts disclosed, but she was clear that ‘the publication of otherwise confidential information about a life-threatening illness is likely to cause distress to the person concerned’, and thus falls under the common-law test for liability. NM (n 5 above) para 137. (2008) 1 Constitutional Court Review 15 rule’.49 Plaintiffs might sometimes have succeeded on the basis of a rebuttable presumption of intention, once objectively wrongful publication was shown,50 or by establishing a modified form of intention known as dolus eventualis (or reckless disregard) on the defendant’s part,51 but a defendant who could satisfy a judicial fact- finder that her mistake was bona fide, although careless, would escape liability.

In NM, the available evidence was consistent (at least) with defendant Smith’s claim of an honest belief that the plaintiffs had consented to unrestricted public naming of them as the persons (living with HIV) who had played certain, described roles in the drug-trial controversy. Whether that evidence could further support a claim of absence of negligence on Smith’s part is much more doubtful, as we shall see. Thus there was, from the beginning, reason to calculate that the plaintiffs’ prospect for success might depend on whether the body of law (the ‘cause of action’) they invoked to support their claim would impose liability on a merely negligent defendant.52 At the moment when the plaintiffs sued, the common law of delict evidently was against them on that potentially crucial point. By contrast, for aught anyone could then tell, the law of the putative constitutional tort based on section 14 of the Bill of Rights might have been with them. Even so, the plaintiffs chose to base their claim on the actio iniuriarum, and not on section 14 of the Bill of Rights. That choice requires explanation.

3.4 Two litigation agendas compared

In the view of Madala J, the plaintiffs were forced to their pleading choice by the Constitutional Court’s 1997 decision in Fose.53 Fose certainly does leave a doubt about when, if ever, ‘constitutional damages’ might be awarded in respect of a violation of a provision in the Bill of Rights. It seems to me, however, for reasons I am about to present, that a complete explanation for the NM plaintiffs’ pleading choice would have to include the further point that a claim expressly and immediately invoking the prescriptive content of section 14 of the Bill of Rights, as a test of the legally actionable character of Smith’s conduct, would have brought the plaintiffs to very much the same

49 NM (n 5 above) para 55 (Madala J). 50 See NM (n 5 above) para 65 (Madala J); para 93 (Langa CJ); paras 125, 155, 169 (O’Regan J). 51 See NM (n 5 above) para 155 (O’Regan J). 52 The summons issued by the plaintiffs makes clear they understood that this was so. See NM (n 5 above) para 124 (O’Regan J) (describing the summons). 53 Fose v Minister of Public Safety and Security 1997 3 SA 786 (CC); 1997 7 BCLR 851 (CC); see NM (n 5 above) para 27. 16 Uses of interpretive ‘charity’ pass, regarding the question of that conduct’s intentionality, as resulted from their common-law claim.

This consequence follows inexorably from the mandate of Constitution section 8(3), as construed by the Constitutional Court.54 That section provides:

When applying a provision of the Bill of Rights to a natural or juristic person in terms of subsection (2), a court — (a) in order to give effect to a right in the Bill, must apply, or if necessary develop, the common law to the extent that legislation does not give effect to that right; and (b) may develop rules of the common law to limit the right, provided that the limitation is in accordance with section 36(1).

By this seemingly clear command of constitutional law, any court disposed to impose civil liability, in terms of standards of conduct drawn from one or another of the rights-guarantees in the Bill of Rights, is required to proceed in one of three ways, and not otherwise. Either (1) the court must decide the case in terms of legislation that gives due effect to the constitutional right in question; or (2) in the absence of such legislation, the court must decide the case in terms of applicable common law that gives due effect to the constitutional right in question; or (3) if the court decides that the applicable common law does not currently give due effect to the right, the court must decide the case in terms of the common law as developed by the court so as to cure the deficiency.55

Accordingly, in NM, the result of a claim for a ‘direct’ application of section 14 to Smith’s conduct would have been the Court’s having to decide two principal issues in tandem: whether giving due effect to

54 That is to say, it follows inexorably from the Constitutional Court’s authoritative, ‘black letter’ construction of the three subsecs of sec 8, as announced in Khumalo v Holomisa 2002 3 SA 401 (CC); 2002 8 BCLR 771 (CC) (‘Khumalo’) paras 29-32 and precisely summarised by Woolman ‘Application’ (n 17 above) 6, 45. It would not follow from Woolman’s ‘preferred reading’ of sec 8, summarised at Woolman ‘Application’ (n 17 above) 11, 45-46. But of course the NM plaintiffs’ lawyers would have had to frame their case in the shadow of the Constitutional Court’s reading, not that of any commentator. 55 This directive carries out what logic would seem to require in its absence. We can use the NM case to illustrate. Suppose a court finds that Smith’s conduct has ‘limited’ (ie, violated) NM’s prima facie right to privacy under Constitution sec 14. Smith will seek refuge under Constitution sec 36(1). She will claim that she has acted in terms of law of general application, namely, the extant common law of delict, which — as she will claim — insulates her from liability as long as her conduct has not been knowingly and intentionally wrongful. The court, then, will have to decide whether that rule of the common law, bestowing a legal privilege to commit negligent (although objectively wrongful) publication of private information, is reasonable and justifiable in terms of Constitution sec 36(1). If the answer is ‘no,’ the court will have no option but to develop the common law so as to cure the defect. (2008) 1 Constitutional Court Review 17 section 14 requires that Smith be held liable in this case, given the evidence and what it shows; and whether (and, if so, how) the common law as it currently stands must be ‘developed’ to encompass at least some instances of non-intentionally but objectively wrongful invasion of privacy. The first question for decision surely would have been (a) whether Smith’s conduct was both objectively wrongful and animo iniurandi according to the pre-existing common-law standard. If yes, the plaintiffs would prevail.56 If no, the next question would have been (b) whether the plaintiffs’ case against Smith discloses a need to expand common-law liability to catch at least some instances of non-intentionally but objectively wrongful publication of private facts, in order to vindicate fully the right of privacy guaranteed by Constitution section 14. If no, the defendant would prevail. If yes, the next question would have been (c) precisely what modification of the common law is required in order to give due effect to the constitutional right of privacy, in the light of possibly competing constitutional principles and values such as freedom of expression — a question that would have been decided by judges who were at the same time, inevitably, thinking about (d) whether one or another newly expanded zone of liability that the Court might approve would catch Smith’s conduct.

Compare that agenda with the one that plaintiffs basing their claim on the actio iniuriarum must anticipate. Taking the common law as it stands, they can succeed by establishing that (a) Smith’s conduct was both objectively wrongful and animo iniurandi in terms of the extant common law standard of liability.57 Alternatively, given the court’s inherent power to develop the common law,58 combined with the court’s duty to develop the common law so as to promote the spirit, purport, and objects of the Bill of Rights, imposed by Constitution section 39(2),59 the plaintiffs can succeed by establishing, in the light of their case against Smith, all of the following: (b) that the common law’s current, blanket prescription of non-liability in all instances of non-intentionally but objectively wrongful publication of private facts is not acceptable to the Bill of Rights; plus (c) and (d) that when the common law is properly

56 For completeness, we should note the possibility of the defendant persuading the court that the extant common law standard must undergo development in her favor, either in terms of Constitution sec 8(3) (in order to give due horizontal effect to the right of freedom of expression guaranteed by sec 15), or in terms of sec 39(2) (in order to keep the common law in tune with the objects of the bill of rights in the matter of freedom of expression). 57 And also, if necessary, successfully defending that standard against complaint that it unduly constricts constitutionally valued freedom of expression. 58 ‘The Constitutional Court, Supreme Court of Appeal and High Courts have the inherent power to ... develop the common law, taking into account the interests of justice.’ Constitution sec 173. 59 Carmichele v Minister of Safety and Security 2001 4 SA 938 paras 33-39, makes clear that a court is required to consider whether such development of the common law is required in any case where a party urges the point. 18 Uses of interpretive ‘charity’ modified in order to give due effect to the constitutional right of privacy, in the light of possibly competing constitutional principles and values such as freedom of expression, Smith’s conduct in this case will render her liable to the plaintiffs.

The two agendas look virtually identical.60 Therefore (the bottom line): Since Fose suggests the possibility of special obstacles facing claims for constitutional damages, the plaintiffs’ course of least resistance was to base their demand for damages on a claim of a common law violation of privacy under the actio iniurarum, with section 39(2) hovering overhead. To my mind, Madala J’s opinion adequately (if somewhat elliptically) evinces a clear appreciation of the entire line of thought I have just been unpacking.61 At the very least, Madala J’s opinion is in all respects consistent with this line of thought. If, as Woolman truly says, the opinion never poses the possibility of ‘the [judicial] creation of a self-standing constitutional action grounded in the right of privacy’,62 that is because the applicants, it would seem for sufficient reasons that the majority may fairly be taken to have understood, chose not to start up that particular formulation of their claim.

3.5 Jurisdiction

Although the plaintiffs thus chose to rely on a common law and not a constitutional cause of action, their appeal to the Constitutional Court was quite patently and undeniably one that fell within that Court’s subject-matter jurisdiction — limited though that jurisdiction surely is to ‘constitutional matters’ and issues ‘connected to’ constitutional matters.63 The overarching legal problem presented by the case is whether the common law of delict, in regard to invasion of privacy, requires some modification in order to keep it in line with the Bill of Rights taken as a whole. Involved, as one immediately sees, is not only a question about whether — and if so, how far — the common law must be altered so as to attach liability to at least some instances of non-intentionally wrongful disclosure of private facts, in order to keep faith with the Bill of Rights in regard to privacy and dignity. Equally involved is a question about whether — and if so, how far — the common law may permissibly impose such liability, having in mind the commitment of the Bill of Rights in regard to freedom of expression. These are obviously issues of full-scale constitutional import.

60 Or very closely so. Arguably, there is a difference, which I will address at the close of my discussion of the NM case (see below part 3.10) and in my discussion, to follow, of Masiya. 61 See NM (n 5 above) paras 24, 27, 55-57. 62 Woolman ‘Amazing’ (n 2 above) 783. 63 Constitution sec 167(3)(b). (2008) 1 Constitutional Court Review 19

Woolman sees NM as a case in which the Constitutional Court majority found no occasion to consider these constitutionally fraught issues of common-law development, because it decided the case, instead, by overturning the High Court’s factual conclusion that Smith’s objective violation of the plaintiffs’ privacy was non- intentional and hence non-actionable under the extant common law standard. That might seem to invite a question about why the majority did not order the case dismissed for lack of jurisdiction, once it perceived that the case could and would be decided on standard, common-law grounds (‘run-of-the-mill’ grounds, Woolman calls them).64 And indeed, the majority must be faulted for its own complicity in creating an impression of the irrelevance of the Bill of Rights to its disposition of the case.65

That impression, however, is mistaken. We shall see that it is at least an open question whether the NM majority did not, in effect, ‘develop’ the common law in response to constitutional pressure. (In that regard, the majority may be faulted for misapprehending the implications and ramifications of its own decision in the case.) But suppose we say that it did not. The majority could still be right in affirming the Constitutional Court’s jurisdiction in this case, even after reaching its conclusion against any constitutionally imposed need for adjustment of the common law. Jurisdiction would hold, as long as the case presented a fairly debatable question of such a need — including of a need to weaken the actio iniuriandum in deference to freedom of expression; for surely such a question is itself a constitutional matter. (Ask yourself: Does the Constitutional Court forfeit jurisdiction by deciding, after argument, to uphold as constitutional a plausibly challenged statute?)66

The very presence of the dissents in this case would thus seem to settle in the affirmative the question of jurisdiction. At all events, the very question of how to construe and apply the extant common law to

64 See Woolman ‘Amazing’ (n 2 above) 782 n 40: ‘[T]he court often accepts cases prior to the meaningful application of its collective “mind” to the question of whether or not the case genuinely raises a constitutional issue.’ For an example of a jurisdictional dismissal, upon its becoming apparent to the Court that no constitutionally-pressured development of the common law is in the offing, see Phoebus Apollo Aviation CC v Minister of Safety and Security [2002] ZACC 26; 2003 1 BCLR 14 paras 10-12. 65 See NM (n 5 above) para 69 (excusing the Court from considering the possible impact on freedom of expression from its decision to hold Smith liable, on the ground that ‘this judgment is not extending the common law definition of intention to include negligence in relation to the publication of private medical facts’). 66 If, as I assert, the Constitutional Court is competent to hear any non-frivolous claim of a need (in terms of sec 39(2)) to develop a common-law rule on which a party to the case relies, the effect may be to extend the Court’s jurisdiction too broadly for your comfort. If so, the fault lies not in the Court but in the drafters of the Constitution. See FI Michelman ‘The supremacy of the Constitution and the rule of law’ in S Woolman et al (eds) Constitutional law of South Africa (2nd Edition, OS, 2005) ch 11 7-9 (Michelman ‘Supremacy’). 20 Uses of interpretive ‘charity’ facts of the sort presented by the case of NM, being determinative of whether the common law must undergo development in terms of section 8 or 39, is itself drenched in constitutional-legal considerations. All of the common-law-framed issues in the case — whether the facts concerning the HIV status of the plaintiffs were of ‘private’ quality and whether the private character of those facts had been fatally compromised by previous disclosure for allegedly limited purposes;67 whether Smith’s conduct was intentionally wrongful in the relevant, legal sense;68 whether that conduct was in any degree negligent, and, if so, in what specific respects69 — stood to be decided in the shadow of the Bill of Rights and of the looming possibility that the common law might, depending on how they were decided, have to be developed under the mandate of section 39(2).

Madala J wrote for the majority as follows:

The dispute before us is clearly worthy of constitutional adjudication and it is in the interests of justice that the matter be heard by this Court since it involves a nuanced and sensitive approach to balancing the interests of the media, in advocating freedom of expression, privacy and dignity of the applicants irrespective of whether it is based on the constitutional law or the common law. This Court is in any event mandated to develop and interpret the common law if necessary.70

The last sentence is crucial. In the light of what precedes it, it states why the case does ‘genuinely raise a constitutional issue’71 and it shows that the NM majority in this case necessarily undertook more (and did more) than merely ‘act as a trier of fact in a run-of-the-mill actio iniuriarum matter.’72

I shall make these matters more concrete in discussion to follow below. For now, the point simply is that no one conversant with the established commitment of South African constitutional jurisprudence to a pervasive influence for the Bill of Rights throughout the length and breadth of the country’s law73 could doubt that the knot of common-law controversies in NM deserved — if it did not demand — the attentions of at least one of South Africa’s two pinnacle courts, exercising a responsibility imposed on both of them (given how the plaintiffs chose to frame their case) by section 39(2). Had an appeal to the Supreme Court of Appeal been an open possibility in this case, and the Constitutional Court nevertheless

67 See NM (n 5 above) paras 34-45 (Madala J); paras 135-137; paras 142-43 (O’Regan J). 68 See NM (n 5 above) paras 58-65 (Madala J); paras 155-69 (O’Regan J). 69 See NM (n 5 above) paras 100-11 (Langa CJ); paras 183-89 (O’Regan J). 70 NM (n 5 above) para 31. 71 Woolman ‘Amazing’ (n 2 above) 782 n 40. 72 Woolman ‘Amazing’ (n 2 above) 783. 73 See Michelman ‘Supremacy’ (n 66 above) 37-41. (2008) 1 Constitutional Court Review 21 accepted the case for hearing, I would have expected from the Court some careful explanation for its choice thus to allow circumvention of the SCA. That, however, was not the state of affairs confronted by the Constitutional Court when NM’s application arrived at its doorstep. The SCA had already removed itself from the case by its summary refusal of leave to appeal.74

3.6 A puzzle

The Constitutional Court divided sharply over assessment of the evidence concerning the intentionality of Smith’s conduct. Writing for a majority of seven justices, Madala J concluded that Smith acted intentionally, within the terms of the doctrine then and here surrounding the actio iniuriarum.75 In the views of Langa CJ and Sachs J, Smith’s conduct, although not intentionally wrongful, was negligent or unreasonable (and thus liability-producing under the common law as they would have had it developed).76 In the view of O’Regan J, Smith’s conduct was entirely free of fault.77 There are two divisions worth exploring here: first, the division between Madala J’s majority and the three justices — Langa CJ , Sachs J, and O’Regan J — who rejected the majority’s finding of intentionally wrongful conduct on the part of Smith; and, second, the division among those three over the question of Smith’s negligence.

Smith had directed her trial testimony to explaining how she acted in the honest belief either that the plaintiffs’ names had already been disclosed to the general public or else that the plaintiffs had fully and finally waived objection to any such disclosure. The High Court had absolved Smith of any intention to act contrary to the plaintiffs’ wishes in this regard. As O’Regan J’s extensive and careful summary amply demonstrates, such an inference is, at the very least, reasonably supportable from essentially unchallenged components of Smith’s testimony.78 But then a nagging question arises: How are we to explain the majority’s apparent deviation from the established practice of deference by appellate tribunals to trial court findings on issues of fact? That looks to me like Sherlock Holmes’s dog that did not bark: a puzzle we should probe, if we — I mean we as academics — mean to leave no stone unturned in getting to the innards of the problem presented to the Constitutional Court by the NM case. The question all the more strikingly demands our attention because the

74 See NM (n 5 above) para 26. 75 See NM (n 5 above) paras 58-65. 76 See NM (n 5 above) paras 92-93 (Langa CJ); paras 205-07 (Sachs J). 77 See NM (n 5 above) paras 155-69 (intention); paras 183-89 (negligence). 78 See NM (n 5 above) paras 157, 159-64. 22 Uses of interpretive ‘charity’ majority itself declined to address it, despite its having been pressed by O’Regan J in dissent.79

Academics can speculate — sometimes profitably — about imaginable, realist-style explanations for such puzzling judicial silences. (Was the Court just hell-bent on imposing a result that it likes? Did some intracurial disagreement get in the way of full transparency?) In keeping with my general undertaking for this essay, I choose a different path of speculation — that of assuming (or pretending, if you like that better) that the NM majority thought the legal reasons for its action (here, overturning a lower court’s factual determination) so obvious as not to require explanation. The idea is to see what we might learn about the legal problem presented to the court by the case at hand by giving the court every benefit of the doubt, working to make their judgment, as a follower of Ronald Dworkin might say, the best that it can be.

Approaching the matter in that spirit, we may discover that the majority’s finding of intentionally wrongful action on Smith’s part, disputable though it surely is, is no more flatly ‘contrary’ to the evidentiary record than an opposite finding is.80 We may find that, pace all of Langa CJ, O’Regan J, and Stu Woolman, the record (as described by the opinion-writers in NM) can indeed ‘support a factual finding that [Smith] had acted intentionally to harm the privacy and dignitary and dignitary interests of the applicants.’81 Judgment on this point may depend partly on what you mean by ‘intentionally’, but that feature in the case is, as I have already pointed out and shall explain further below, a part of what makes this case indubitably one of constitutional law for the Constitutional Court to decide.

3.7 ‘Incremental development’ and Smith’s ‘intention’

Therefore, consider Constitution section 39(2). Remember how it goes?

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.

All right, then: Where there is fair room for disagreement over the proper gloss on statutory terminology, and the choice is outcome- determinative for the case sub iudice, courts are instructed by the Constitution to prefer the gloss that they find to be the more truly

79 See NM (n 5 above) paras 125, 169. 80 Woolman ‘Amazing’ (n 2 above) 781. 81 As above (invoking the opinions of Langa CJ and O’Regan J in NM). (2008) 1 Constitutional Court Review 23 promotive of the objects of the Bill of Rights. How goes it, then, when a like question arises with respect to common law terminology?

Please don’t stand there painting me, along with Stu Woolman, into ‘some ... benighted Hartian, positivist view’82 of the common law, or telling me that there is so such thing as ‘common law terminology’ to make a match with statutory terminology. I join Woolman in knowing — and you know, too — that there very obviously is such a thing, which is not to say that Woolman and I do not also fully understand the common law to be, at bottom, an assemblage of always open-ended principles, not of closed-ended rule-sentences. (We can leave aside whether the latter is any the less true of statute law.)83 Where Woolman and I both stand our ground is on the point that lawyers and the rule of law cannot make do with a common law discourse that is not densely populated by rule statements, ‘common law terminology’. We both — and so do you, Reader — ‘work within a tradition ... of which South Africa is most avowedly a part’, which ‘recognises rules as a necessary feature of the legal [including the common law] landscape.’84 Witness:

For the common law action for invasion of privacy based on the actio iniuriarum to succeed, the following must be proved: (a) Impairment of the applicants’ privacy; (b) Wrongfulness; and (c) Intention (animus iniurandi).85 The elements of the actio injuriarum are the intentional and wrongful infringement of a person’s dignitas, fama, or corpus.86

Those are two somewhat variant, but essentially identical, renditions of the same, single instance of a genus that lawyers routinely call ‘rules’ of the common law. As such, they exemplify perfectly what I mean by my jurisprudentially obtuse expression, ‘common law terminology’, coined by way of analogy to ‘statutory terminology’.

Now, cases arise in which statutory terminology is found unacceptably out of sorts with the Bill of Rights, on any reading of the statute falling within the outer limits imposed by a decent respect for grammar and lexicon.87 In those cases, by constitutional command, a judicial declaration of the statute’s ‘invalidity’ — unconstitutionality

82 Woolman ‘Amazing’ (n 2 above) 789. 83 See generally, on these matters, R Dworkin Law’s Empire (1986). 84 Woolman ‘Amazing’ (n 2 above) 791. 85 NM (n 5 above) para 55 (Madala J). 86 NM (n 5 above) para 151 (O’Regan J) (citing both Voet Commentary on the Pandects 47 10 1 and R v Umfaan 1908 TS 62 at 66 (Innes CJ)). 87 See, eg, National Coalition for Gay & Lesbian Equality v Minister of Home Affairs 2000 2 SA 1 (CC), 2000 1 BCLR 39 (CC) paras 25-26 (Home Affairs). 24 Uses of interpretive ‘charity’

— must ensue.88 How goes it with cases in which common law terminology is similarly found to be irreparably out of sorts with the Bill of Rights: say, because it would flatly restrict civil liability for certain sorts of wrongful harming to instances in which the wronging is intentional, when the Bill of Rights would be better satisfied by lifting that flat restriction? You might answer: In such cases, the common law is to be declared invalid as it stands and then remedially ‘developed’ so as to cure the defect. You might offer as an example of such a cure (passing, for now, the absence of any declaration of invalidity) the bit of terminological surgery supported by Langa CJ, O’Regan J, and Sachs J in NM, according to which the term ‘intention (animus iniurandi)’, where it appears in item (c) in Madala J’s above recital of the elements of the civil cause of action for invasion of privacy, would be replaced (in some, if not all, categories of cases) by the term ‘intention (animus iniurandi) or negligence’. We have there, as it were, a remedy of ‘reading in’. (Compare it with a replacement of ‘spouse’, in a statute, by ‘spouse or partner, in a permanent same-sex life partnership in which the partners have undertaken reciprocal duties of support’.)89

Have you begun to glimpse the point I am after? Section 39(2) contains an express instruction — call it a constitution-conforming instruction — to judges engaged in the interpretation of statutory terminology, according to which the judges must interpret with a view to promoting the spirit, purport, and objects of the Bill of Rights, presumably before deciding whether they must declare the statute invalid and resort (perhaps) to remedial reading-in. On its face, section 39(2) entirely omits any parallel, express instruction to judges engaged in the contextualised application (as opposed to any surgical alteration by excision or insertion of words) of common law terminology. Since such an instruction does appear in the Interim Constitution, one might pause for a moment to wonder whether the omission was intentional.90 It is difficult, however, to conceive what the drafters might have intended by it. Are there, after all, no cases in which a due regard for the objects of the Bill of Rights would, or should, crucially affect a judge’s interpretive shading and application

88 See Constitution sec 172(1)(a). 89 Satchwell v President of the Republic of South Africa & Another 2002 6 SA 1 (CC), 2002 9 BCLR 986 (CC) para 37. 90 Compare Constitution of the Republic of South Africa Act 108 of 1994 (Interim Constitution) sec 35(2): No law which limits any of the rights entrenched in this Chapter, shall be constitutionally invalid solely by reason of the fact that the wording used prima facie exceeds the limits imposed in this Chapter, provided such a law is reasonably capable of a more restricted interpretation which does not exceed such limits, in which event such law shall be construed as having a meaning in accordance with the said more restricted interpretation. (The Afrikaans version has ‘reg’ where the English has ‘law’; and so, by the reasoning used by Kentridge J in Du Plessis, it presumably means to cover common law along with statute law. See Du Plessis (n 17 above) para 44.) (2008) 1 Constitutional Court Review 25 of common law terminology as it stands to the facts of a particular case? Thus, without any declaration of invalidity and concomitant surgery on the terminology as received? If so, are not the courts plainly — if implicitly — required by section 39(2) to shade and apply in favor of the Bill of Rights? And is it not a part of the Constitutional Court’s responsibility — and jurisdiction — to speak the last word on whether lower courts have done so properly in a given case?

In a luminous opinion for the Court in K v Minister of Safety and Security (‘NK’),91 O’Regan J answered all of those questions with a resounding ‘yes’. In NK, the plaintiff, having been assaulted and raped by police officers, sued the Minister for damages on a theory of respondeat superior. The plaintiff stood ready to contend for ‘development’ of the common law, but only if her case would not succeed ‘on a proper application’ of the extant common law rule of vicarious liability.92 You can see why she had to argue thus in the alternative. As O’Regan J explained:

the common law [normally] develops incrementally through the rules of precedent ... [It sometimes happens that a court must] determine whether a new set of facts falls within or beyond the scope of an existing rule. The precise ambit of each rule is therefore clarified in relation to each new set of facts. A court faced with a new set of facts, not on all fours with any set of facts previously adjudicated, must decide whether a common-law rule applies to this new factual situation or not. If it holds that the new set of facts falls within the rule, the ambit of the rule is extended. If it holds that it does not, the ambit of the rule is restricted, not extended.93

O’Regan J had no doubt that a court facing such a decision is bound by the mandate of section 39(2) — to heed the Bill of Rights when ‘developing’ the common law — even though no ‘new’ or ‘changed’ common law rule is in the offing:94

The overall purpose of section 39(2) is to ensure that our common law is infused with the values of the Constitution. It is not only in cases where existing rules are clearly inconsistent with the Constitution that such an infusion is required. The normative influence of the Constitution must be felt throughout the common law. Courts making decisions which involve the incremental development of the rules of the common law in cases where the values of the Constitution are relevant are therefore also bound by the terms of section 39(2). The obligation imposed upon courts by section 39(2) of the Constitution is thus extensive, requiring courts to be alert to the normative framework of the Constitution not only when

91 [2005] ZACC 8; 2005 6 SA 419 (CC) (NK). 92 NK (n 91 above) para 14. 93 NK (n 91 above) para 16. 94 As above. 26 Uses of interpretive ‘charity’

some startling new development of the common law is in issue, but in all cases where the incremental development of the rule is in issue.95

By ‘incremental development’, O’Regan J plainly meant the process she had just previously been describing: ‘A court [facing] a new set of facts, not on all fours with any set of facts previously adjudicated, [deciding] whether a [settled] common-law rule applies to this new factual situation or not.’ In NK, O’Regan J held that the judicial work in such cases should count as ‘development’ of the common law for the purposes of section 39(2).96 It seems she felt she had to say so, in order to establish firmly the Constitutional Court’s jurisdiction in the case before her, which she saw as presenting a question about whether certain facts fell properly within the rule of respondeat superior, construing and applying that rule with due regard to the Bill of Rights.

Madala J’s majority in NM found that defendant Smith’s conduct fell under the head of ‘intentionally’ wrongful privacy-invasion — insomuch as (or so the majority found from the evidence) Smith ‘at least foresaw the possibility that the consent had not been given to the disclosure’.97 Might we not well understand them as having followed in the tracks of O’Regan J’s work in NK — deciding, under pressure from the Bill of Rights, ‘whether the common law rule applies to this new factual situation or not’? ‘While the claim falls to be dealt with under the actio iniuriaum,’ the majority had already reminded us, ‘the Constitution must inform the application of the common law’.98 Were they just vamping there, or might we take those words as a cue to understanding that the majority’s application to Smith’s conduct of the common law category of ‘intention’ (which, we must remember, already included dolus eventualis), was shaded or tinctured with a view to bringing or keeping the common law doctrine in full conformity with the objects of Bill of Rights?99

It is not impossible to see why the majority might have chosen such a path to judgment in Smith’s case, in preference to a judicially declared, surgical replacement of the term ‘intention’ in Madala J’s item (c) by ‘intention or negligence’ — even granting that the latter

95 NK (n 91 above) para 17. 96 See NK (n 91 above) paras 16, 17. 97 NK (n 91above) para 64 (Madala J). 98 NM (n 5 above) para 28; see also para 50 (‘If human dignity is regarded as foundational in our Constitution, a corollary thereto must be that it must be jealously guarded and protected.’). 99 I am not alone in thus understanding Madala J’s opinion; I have Sachs J for company. See NM (n 5 above) paras 201-02 (Sachs J) (‘It is in [the] human rights context [addressed by the Bill of Rights] that the competing interests at stake in the present matter must be dealt with. In a fittingly accessible manner, Madala J has indicated how in the particular circumstances of this case competing needs with respect to human dignity, on the one hand, and freedom of expression, on the other, should be reconciled.’) (2008) 1 Constitutional Court Review 27 technique carries with it a more immediately accessible package of instruction to learners of the law.100 Terminological replacement is inevitably — to some irreducible degree — done at wholesale, and it is — to that irreducible degree — inevitably a blunter instrument than pure case-by-case steering. If — to illustrate — we follow the carefully plotted line of thought in O’Regan J’s NM opinion (agreed to by both Langa CJ and Sachs J), a due regard for the Constitution’s privacy right calls for addition of negligence liability to the actio iniuriarum for privacy invasion.101 But hold the phone. A due regard for the Constitution’s freedom-of-expression principle calls for at-least initial or provisional restriction of the to-be-added negligence liability to a category of ‘media’ defendants, with ‘media’ defined expansively enough to include the authors and publishers of books as well as the more strictly journalistic, daily and periodical print and electronic media.102 There is nothing in itself wrong with such a definitional-categorical approach, conducted with the sort of caution and care evinced by O’Regan J’s opinion. But neither can it be thought indefensible for a judge to decide that the process of drawing the distinctions that contextual variations will require is better pursued, at least for the time being, through the inflections of case-by-case application of the extant rule and the operations of stare decisis (as explained by O’Regan J in NK) than through process of word-and- phrase-surgery-and-implant otherwise known as ‘development’ of the common law, even if the choice for pure, case-by-case adjustment entails some sacrifice of transparency to rote learners.

Transparency is not entirely foregone, after all, as long as stare decisis holds in measurable degree. That is why I say that Madala J spoke too hastily in saying that there was no need for him to consider any possible ‘chilling effect’ of his judgment on freedom on expression, given that (as he said) he was merely applying the common law of delict as it then stood, not ‘extending’ it to include liability for negligence.103 As O’Regan J’s analysis in NK makes clear, the NM majority’s quite striking application of the ‘intention’ standard to the facts of the NM case cannot avoid making waves as a precedent, and so the question whether those waves will have a constitutionally significant chilling effect cannot be avoided.

As I am about to show, however, the right answer to that question is pretty plainly ‘no’ in Smith’s case, because the true and entirely

100 ‘Clear and accessible norms’ in the words of Sachs J. NM (n 5 above) para 204. 101 See NM (n 5 above) paras 92, 95 (Langa CJ); paras 171-82 (O’Regan J); paras 203- 04 (Sachs ). Sachs J speaks of the extended standard of liability as one of ‘reasonableness’, but he evidently has in mind approximately what Langa CJ and O’Regan J mean by ‘negligence’. 102 See NM (n 5 above) paras 92-93, 95, 98-99 (Langa CJ); paras 203-05 (Sachs J); paras 177-82 (O’Regan J). 103 NM (n 5 above) para 69. 28 Uses of interpretive ‘charity’ innocuous ‘rule’ of Smith’s case can easily be taken to be this: If you are publishing an account of events and transactions including private facts about real persons, and you attach the true names of those persons to those facts, without having directly in hand and before your eyes proof-positive of their express permission to have you do so, and you cannot adduce any credible reason to think that a public that knows the names is better informed or otherwise more enriched that a public that does not, you are acting in culpable disregard (call it ‘negligent’, call it ‘reckless’) of those persons’ presumptive interest in retaining control over dissemination of that private fact about themselves. Lest the bar and lower bench should be left in the slightest doubt about the lesson to be drawn, Sachs J took pains to point ‘the moral of the story’:

[U]nless overwhelming public interest points the other way, publishers should refrain from circulating information identifying the HIV status of named individuals, unless they have the clearest possible proof to consent to publication having been given.104

That does not seem so hard to learn, even for ‘second-year LL.Bs’.105 I have little doubt of the message by now having found its way into every well-heeled South African publisher’s and writer’s in-box.

Of course, it is very far from a full treatise on the values served by the constitutional rights of privacy and freedom of expression, by comparison, say with the entirely welcome offerings we find in the opinion of O’Regan J — memorable passages on the values underpinning the constitutional rights to privacy and freedom of expression, of which I would not wish the South African constituency (or the world constituency, for that matter) to be deprived.106 For those passages, or anything remotely comparable, the NM majority opinion can claim no credit. And yet it does not necessarily follow that the majority opinion fails to give ‘identifiable content’ to provisions in the Bill of Rights, or that it leaves the bar and lower bench without worthwhile new instruction about ‘how the Bill of Rights might operate’ in an important class of relevantly similar, future matters.107 NM might still be treated as one of those occasions, mentioned by Woolman, of an ‘opportunity’ thrown up by the stream of cases with varying fact-patterns, to ‘expand our understanding of how given rights function in given environments.’108

104 NM (n 5 above) para 209. 105 Woolman ‘Amazing’ (n 2 above) 787 n 47. 106 See NM (n 5 above) paras 126-34 (on the values underpinning the constitutional right to privacy); paras 144-46 (on the values underpinning the constitutional right to freedom of expression). 107 Woolman ‘Amazing’ (n 2 above) 763. 108 Woolman ‘Amazing’ (n 2 above) 765. (2008) 1 Constitutional Court Review 29

My proposed account of the NM majority’s finding of liability- producing ‘intention’ on the part of defendant Smith has the further merit of providing an honorable explanation for the majority’s refusal of deference to the High Court’s contrary finding. When extant common law appears to require development in terms of section 39(2) or section 8(3), the Constitutional Court follows — and for good reason — a general practice of giving the High Courts and the Supreme Court of Appeal a first crack at the work.109 No one, however, would suggest that the Constitutional Court should defer when it comes to deciding — as the Court had to decide, for example, in Khumalo — whether a common law solution reached by the common law judiciary (so to call them) is sufficiently in synch with the Bill of Rights as not to call for further correction. Nor, by the same token, should we expect the Constitutional Court to defer when it judges whether a lower court’s in-context application of a common law rule has been carried out with due regard for the Bill of Rights — see NK — any more than when the Constitutional Court judges whether a lower court’s in-context application of a statute has complied adequately with the command of section 39(2).110

3.8 Smith’s negligence (or ‘intention’?)

I turn now to the division among the three judges who thought the question of Smith’s negligence should properly be decisive of the case. I do so with the expectation that our examination of this division will cast some further light on the majority’s finding of Smith’s tortious ‘intention’.

Having found an absence of intentional wrongdoing by Smith, Langa CJ, Sachs J, and O’Regan J all had to confront the question of possibly developing the common law so as to impose liability for at least some instances of negligently wrongful invasion of privacy. All three agreed, at least tentatively, on the introduction of a negligence standard for ‘media’ defendants only, roughly on the model of the

109 See, eg, Masiya (n 4 above) para 17 (‘Ordinarily, constitutional matters involving the development of the common law should first be taken to the Supreme Court of Appeal before they reach this Court because of the breadth of its jurisdiction and its expertise in the common law.’); Carmichele (n 59 above) paras 55-56: The proper development of the common law under section 39(2) requires close and sensitive interaction between, on the one hand, the High Courts and the Supreme Court of Appeal which have particular expertise and experience in this area of the law and, on the other hand, this Court. Not only must the common law be developed in a way which meets the section 39(2) objectives, but it must be done in a way most appropriate for the development of the common law within its own paradigm. There are notionally different ways to develop the common-law under section 39(2) of the Constitution, all of which might be consistent with its provisions. Not all would necessarily be equally beneficial for the common law. 110 See, eg, Bato Star Fishing (Pty) Ltd v Minister of Environmental Affairs & Tourism [2005] ZACC 8; 2005 6 SA 419 (CC) paras 72-98. 30 Uses of interpretive ‘charity’

SCA’s introduction in Bogoshi111 of a ‘reasonable-to-publish-even-if- false-in-fact’ standard of defamation liability for such defendants,112 approved by the Constitutional Court in Khumalo as an acceptable reconciliation of the potentially conflicting demands of the Bill of Rights regarding the rights to human dignity and to freedom of expression.113 Where the three came apart was over the application of a negligence standard to the evidence in the NM case.

Of the three, O’Regan J was by far the most outspoken on the point — which seems surely correct — that all such applications must be sensitive to all of the affected objects of the Bill of Rights, which in this case very saliently included freedom of expression.114 In order to avoid imposing too great a burden on freedom of expression, said O’Regan J, the law must leave journalists free to ‘to publish information provided to them by reliable [published] sources without rechecking in each case whether the publication was lawful,’ at least where ‘there are no grounds for ... suspicion [of] a risk that the original publication was not lawful.’115 In the view of O’Regan J, grounds for suspicion were lacking in Smith’s case. Therefore, to hold her liable,

one would ... have to find that wherever a reputable source has published [names of persons about whom private facts are disclosed], secondary publication may not take place without the existence of informed consent having been independently verified, so that in each case, the subsequent publisher would have to re-ascertain the facts.116

And such a rule, O’Regan J concluded, would impose an ‘unacceptable burden on the dissemination of information.’117

There is an apparent flaw in this reasoning, and it lies in the word ‘wherever’ (as in ‘wherever a reputable source has published ... ’). Let us grant O’Regan J her premise (strenuously disputed by all the other justices) that grounds for suspicion of a lack of the plaintiffs’ consent to unlimited public identification by name were entirely absent from the totality of the circumstances that Smith confronted.

111 National Media Ltd v Bogoshi 1998 4 SA 1196 (SCA); [1998] 4 All SA 347 (SCA). 112 See NM (n 5 above) paras 172-77 (O’Regan J). 113 See Khumalo (n 54 above) paras 43-45. It is interesting to note how Khumalo and the NM minority approach the ‘reasonableness’/‘negligence’ standard from opposite directions. Khumalo approved a shift of the balance in favor of freedom of expression (and thus against dignitas) by moving away from a prior rule of strict liability for false-in-fact, defamatory publications, whereas as the NM minority would have shifted the balance in favor of dignitas (and thus against freedom of expression) by introducing negligence liability in respect of what had previously been treated as a strictly intentional tort. 114 See Khumalo (n 54 above) paras 1444-48. 115 NM (n 5 above) paras 185, 187. 116 NM (n 5 above) para 186. 117 As above. (2008) 1 Constitutional Court Review 31

(More on this below.) It still does not follow that holding Smith liable for negligence in this case would signal liability ‘wherever’ a journalist or book-writer follows reputable sources in publishing private facts about named persons, and consent from those persons unexpectedly turns out to have been lacking. It only follows that liability will attach to the conduct of the second publisher when she has no sufficient reason for taking the slightest risk of invading privacy by including the names. That is a far cry from ‘wherever’, and moreover it is exactly what is required by a general norm — we call it ‘negligence’ — of acting with due regard for other peoples’ interests at risk, very saliently including (in the South African legal order), their constitutionally recognised interests both in freedom of expression and in privacy.

O’Regan J was prepared to find from Smith’s testimony that Smith ‘simply did not entertain the possibility that either the University of Pretoria or Professor Strauss would have sent a report to a Member of Parliament in circumstances where the consent given was only of a limited variety in a publication that did not draw attention to that fact.’118 I take that to signify an ultimate finding by O’Regan J that, on a preponderance of the evidence, Smith was just sucked in unthinkingly by deceptive appearances that would have prevented the question of the scope of the plaintiffs’ consent from even arising in her mind or anyone’s.119

Yet the evidence on that point, as described by O’Regan J, seems to me quite open to an opposite construction. I have in mind a passage in Smith’s book, flagged by O’Regan J, that appears to say that Smith, while writing the book, decided to include the names after reflecting consciously on a series of events that led her to infer that the ‘consent’ referred to in the Strauss Report must have been ‘a general consent on which she could rely’.120 By what seems to me a quite permissible reading, the passage in question shows Smith consciously deciding that she has a sufficient basis for inferring that a general consent had been given, and the names, therefore, can safely be included. I notice also Smith’s testimony ‘point[ing] to three facts

118 NM (n 5 above) para 184. 119 ‘On a preponderance of the evidence’ because O’Regan J allows that Smith carries the onus of establishing absence of negligence, given proof of the objective fact of her having publishing private facts without consent. See NM (n 5 above) para 179. 120 ‘In the chapter concerned, Ms Smith recounts the story of the clinical trial. She then points to the fact that details of the applicants’ complaints were originally published in the New York Review of Books, though under pseudonyms. She then points to the fact that pseudonyms were not subsequently used in the Strauss Report but that the real names were given and she uses the real names. The chapter itself therefore suggests that she thought that after the New York Review of Books article had been published, the applicants had consented to the publication of their names by Professor Strauss and that that consent was a general consent on which she could rely.’ NM (n 5 above) para 166. 32 Uses of interpretive ‘charity’ which grounded her belief that Professor Strauss had obtained general consent from the applicants for the publication of their names.’121 If we take this testimony to mean that Smith adverted consciously to these facts in forming her belief, at a time when she was still writing her book, then the testimony amounts to Smith’s admission that she knew then that she was choosing on the basis of something less than full and direct knowledge of consent to publication by her. To put the matter in language lawyers will understand, Smith was (her own testimony suggests) consciously choosing on the basis of circumstantial and not ‘real’ evidence.

The testimony as a whole (I mean as described by O’Regan J; I have not read the trial transcript) looks to me quite consistent with — I do not say it compels, but Smith has the onus — a finding that Smith, while writing, adverted consciously to the question of the scope of consent, attempted a judgment about that matter on the evidence she had, and judged erroneously. I do not say she judged that particular question irrationally, or with the slightest bad faith, or even without the very heavy odds in her favor. I do say she nevertheless acted irresponsibly, in an important sense that the legal notion of negligence precisely nails. I say that, under a negligence standard properly understood, Smith’s mistake of judgment cannot be excusable, no matter how unlikely the error, when there was literally nothing worthwhile to be gained by taken any chance at all.

But was there really nothing to be gained? ‘Freedom of expression,’ states O’Regan J,

is important because it is an indispensable element of a democratic society. But it is indispensable not only because it makes democracy possible, but also because of its importance to the development of individuals, for it enables them to form and share opinions and thus enhances human dignity and autonomy. Recognising the role of freedom of expression in asserting the moral autonomy of individuals demonstrates the close links between freedom of expression and other constitutional rights such as human dignity, privacy and freedom. Underlying all these constitutional rights is the constitutional celebration of the possibility of morally autonomous human beings independently able to form opinions and act on them.122

One concurs wholeheartedly, or at any rate I do. The question that stares one in the face, though, is how any of these interests in democracy, autonomy, or opinion-formation is compromised in the slightest (or ‘chilled’, if you like) by holding Smith liable for her borderline-inexplicable choice to publish the real names of three otherwise non-notable women who figured as persons living with HIV

121 NM (n 5 above) para 163 (O’Regan J). 122 NM (n 5 above) para 145 (O’Regan J). (2008) 1 Constitutional Court Review 33 in Smith’s no-doubt highly informative account of Patricia De Lille’s public career. What is the value-added by inclusion of these women’s names? Smith’s own explanation of the choice is that she included the names in order to give her book ‘authenticity’.123 Please! Would the book have been one whit the less authentic, the less informative, the less democracy- or autonomy-supportive, if Smith had fictionalised the names, explaining in a preface or a footnote that she had done so?124 Surely a person acts negligently when she chooses a course of action that anyone can see has the capacity to injure others, and a reasonable member of society would perceive an obvious way to avoid the risk of injury, which is clearly of lesser cost to anyone than the injury might be to any who would suffer it.125

In Smith’s case, avoidance was available at the negligible cost of forbearing to include the names. The cost of avoidance thus would not have been that of the onerous detective work posited by O’Regan J, or of the delays in publishing publicly valuable content that such work might occasion; it would only have been the loss of the smidgin of novelistic coloration that Smith added to her book by inserting the publicly meaningless real names of three women who were otherwise living their lives entirely out of the public eye — a cost bordering on zero, it would seem, in a public perspective.

And now the circle closes. Consider that the effrontery of seeking to excuse the decision to name the plaintiffs by the appeal to ‘authenticity’ may have been instrumental in leading Madala J’s majority to its conclusion that Smith’s choice was ‘intentional’ in the pertinent, legal sense.126 Sachs J, supporting a finding of negligence, remarked that ‘if the slightest doubt existed, there was no need to publish the actual names of the defendants.’127 I so not see why one

123 See NM (n 5 above) paras 24, 61. 124 If ‘authenticity’ really means scooping the competition, or showing what a superior gumshoe you are, it cannot, in the Constitution’s scales, weigh tellingly against a risk to other people’s privacy and dignity. If ‘authenticity’ means providing your readership with persuasive indication that you really have had access to authoritative sources, or to the sources you say you have accessed, it does weigh tellingly in principle, but not in application to this case. Smith had plenty of other concrete particulars to carry that burden, not to mention Patricia De Lille’s endorsement of her book as an ‘authorised’ biography. 125 ‘The traditional test for negligence is axiomatic ... : negligence is established if a reasonable person in the position of the defendant would have foreseen the harm, the reasonable person would have taken steps to prevent it and the defendant did not take those steps.’ NM (n 5 above) para 100 (Langa CJ), citing Krueger v Coetzee 1966 2 SA 428 (A) at 430E. It would seem that Langa CJ meant ‘would have foreseen some risk [not the certainty or near-certainty] of harm’. Later on in his opinion, he wrote that the crucial question in the case was ‘whether the reasonable journalist [in Smith’s position] would have foreseen the possibility of absence of consent.’ NM (n 5 above) para 105. I have suggested that the evidence is compatible with a finding that Smith did in fact foresee that possibility — granted that she, with good reason, regarded it as very slight. 126 See NM (n 5 above) paras 54, 61, 65. 127 NM (n 5 above) para 205. 34 Uses of interpretive ‘charity’ should not go a bit further and say: Since there was nothing of value to be gained by publishing the true names (recall the possibility of openly fictionalising them), it was negligent to publish them without proof positive of consent, expressly directed to the would-be publisher in question.

For this reason, I am led to quibble with the statement of Langa CJ that ‘the’ crucial question to be decided in Smith’s case was ‘whether the reasonable journalist [in Smith’s position] would have foreseen the possibility of absence of consent’.128 That is, of course, a highly pertinent question for any judge approaching this case. To my mind, though, an equally crucial question is whether a responsible author, aware that she lacked proof-positive of the plaintiffs’ consent to the inclusion of their names in her book (as distinguished from Professor Strauss’s report), would have included the names. One can agree with O’Regan J that the reasonable author would not have ‘foreseen’ the possibility of a restricted consent, if by that we mean that the reasonable author would have rated that possibility extremely unlikely, and still find plainly negligent — plainly fraught with an unjustifiable risk of harm to others — the publication of private facts about named persons whom you do not know and with whom you have never met, whose full and clear consents you do not have in hand, and who might just as well have gone nameless in your publication, so far any anyone’s legitimate interests are concerned. Are we perhaps verging here on ‘reckless disregard’? Gaining, then, some further insight into the majority’s finding of ‘intentionally’ wrongful conduct by Smith?129

3.9 A ‘minimalist’ moment?

O’Regan J’s opinion does not address at all this matter of the vanishingly low value for Smith’s book of naming the plaintiffs, even though other justices raised it quite vividly in their opinions, and even though the omission leaves in limbo both O’Regan J’s conclusion that Smith was not negligent and her complaint about the costs to freedom of expression of imposing liability in cases such as Smith’s. If my analysis is sound, those costs are virtually nil. A flat rule of liability for publishing the names of persons attached to private facts, where the names are publicly meaningless and add nothing to the opinion- forming and autonomy- or democracy-supporting value of the published content, could easily be derived as a contextual specification of the negligence principle for cases to which it applies. Such a rule would seemingly be something of a gain for dignitas and

128 NM (n 5 above) para 105. 129 NM (n 5 above) para 125. O’Regan J construes the majority’s finding of intentionality in terms of reckless disregard. (2008) 1 Constitutional Court Review 35 an utterly negligible loss for freedom of expression. It would thus seem to be a rule upon which all members of the Constitutional Court could agree: a triumph, one might say, of ‘minimalism’.130

I am thinking here of the dimension of ‘width’, not ‘depth’,131 for (as I hope I have sufficiently shown) there is substantial depth in the solution I described for the NM case. In the list of possible benefits of a minimalist sensibility marshaled by Iain Currie (following Cass Sunstein),132 there is one that does not appear: Minimalism can be conducive to getting the answer right! I mean right in principle, getting a principle right. Judges whose habit it is to scan for possible narrow grounds of decision, in cases where their court or their country’s legal culture is evidently divided over broader-gauged terms of debate, may sometimes be the likelier to detect some set of relatively contextualised terms of decision that perhaps will not only induce a voting consensus, but also may help to clarify for everyone concerned the true (or at any rate an agreed) meaning for the broader terms of contention (here, ‘intention’ and ‘negligence’) that have been leading their debate into trouble.

Thus, a focus on a relatively contextual point in NM — the point that the inclusion of the plaintiffs’ names was sheerly gratuitous from the standpoint of the public informational and autonomy-supportive value of Smith’s account of Patricia De Lille’s career — might have helped clarify for everyone what the negligence principle really means and is all about. If your lines of reasoning get fixed at too wholesale a level — say, something like ‘negligence liability for publishing and freedom of expression do not mix’ — you not only may overlook a possibly consensus-forming resolution of the case before you, you may fail to see how your generalisations are leading you to a conclusion that traduces the true point and meaning of negligence as a legal concept, and of negligence liability’s value as an instrument for the advancement of the objective normative value system for which the Constitution of South Africa is supposed to speak.

3.10 A case of ‘direct’ or ‘indirect’ application?

A loose end still flaps in the breeze. In part 3.4 above, I compared the litigation agendas respectively implied by a constitution-based and a common law cause of action in NM. The Constitution-based action succeeds (if it does) by way of an exercise of the court’s power of

130 See I Currie ‘Judicious avoidance’ (1999) 15 South African Journal on Human Rights 139 147-50. 131 A judicial decision is ‘narrow’ (as opposed to ‘wide’) when it is ‘minimal in impact on future decisions’. It is ‘shallow’ (as opposed to ‘deep’) when it is ‘minimally theorised’. See Currie (n 130 above) 147. 132 See Currie (n 130 above) 146-50. 36 Uses of interpretive ‘charity’ judicial review: declaration of invalidity and ensuing, remedial modification of the common law.133 The common law action succeeds (if it does) by way of a court’s exercise of its inherent power to develop the common law in the ordinary course of adjudication, under pressure to conform the common law to the norms of the Constitution. Granting this notional difference in the respective judicial powers to be put into play, I concluded that the resulting litigation agendas would be virtually identical, whichever source of power would be chosen. I warned, however, of an arguable difference to be taken up later. Later is now; the arguer for the difference is Stu Woolman, and the difference argued for is this: Arriving at a declaration of invalidity, required for the plaintiffs’ success in a constitution-based action, involves a court in a distinctly more rigorous and exacting mode of constitutional analysis and exposition than does arrival at a decision to exercise the inherent power to develop the common law.

Here it is in a nutshell:

... [T]he drafters intended for there to be two different processes. The first process — direct application — takes the [specifically named] rights and freedoms, and the general rules derived from them, as our point of departure for determining whether [some challenged parcel of common law doctrine] is invalid. The second process — indirect application — allows for a mode of analysis that neither specifies whether a particular right demands vindication nor permits a finding of invalidity. Instead, ... the courts operate under a general injunction to bring all law into line with the ‘spirit, purport and objects’ of the Bill of Rights and the ‘objective, normative value system’ made manifest in the text of the Constitution as a whole.134

The framers, Woolman says, had a particular reason for their provision of these two, distinct processes — direct and indirect application — for measuring the constitutional acceptability of a challenged common law doctrine or rule. As he explains:

Section 8 does not mean that the prescriptive content of the substantive provisions in the Bill of Rights covers each and every legal dispute ... [W]hile the specific provisions in the Bill of Rights cover a large domain of law and conduct, they do not engage all law and conduct. The

133 That is, owing to the command of sec 8(3), it does not succeed by way of damages imposed on Smith for a constitutional tort. 134 Woolman ‘Amazing’ (n 2 above) 769. (2008) 1 Constitutional Court Review 37

independent purpose of s 39(2) is to engage law and conduct not engaged by any of the specific provisions set out in Chapter 2.135

This reason for providing the two, distinct processes of direct and indirect application implies, as Woolman further points out, a definite order of march for courts called upon to consider the constitutional acceptability of one or another rule or doctrine of the common law:

... [O]ne must first ascertain what the ambit is of the allegedly applicable constitutional provisions [he means the several rights- granting clauses in the Bill of Rights]. Only when one has determined that ambit, and found that it does not speak to the issues raised by a [questioned] rule of [the common] law, can one turn to the more open-ended invitation of s 39(2). Analysis of the specific provisions of the Bill of Rights, and the consistency of law or conduct with those provisions, logically must be prior to the analysis of the common law in terms of the general spirit, purpose and objects of the Bill of Rights.136

In other words, the marching order, logically implied by the Constitution’s principle of division of labor between direct and indirect application is: direct application first, then indirect, but only if direct fails to engage the common law rule or doctrine in question.137

But of course Woolman’s point is not merely that the Constitution implies a particular order of priority of resort to direct and indirect application of the Bill of Rights. It is that the obedience of courts to this order of march — direct before indirect — is of crucial importance to the long-run interest in the proper elucidation of South African constitutional law. And that is because of a difference in the sorts and levels of explanatory rigour respectively exacted from courts by direct and indirect application. When applying the Bill of Rights directly in terms of section 8, the Court ‘engages the content of the substantive provisions of the Bill of Rights’ and ‘articulates

135 As above. 136 Woolman ‘Amazing’ (n 2 above) 777. Woolman elaborates: If we reverse the spin, and we first use sec 39(2) to bring the law into line with the general spirit, purport and objects of the Bill of Rights, there is simply nothing left to be done in terms of direct application. The reason is obvious. If the general spirit, purport and objects of Chapter 2, which embraces (at a minimum) the entire value domain reflected by the specific substantive provisions of the Bill of Rights, does not require a change in the law (or a change in conduct brought about by a change in the law), then no narrower set of purposes reflected in a single substantive provision of the Bill of Rights could be expected to do so. 137 See Woolman ‘Amazing’ (n 2 above) 83 (‘The logic ... is one that should require the courts to ... develop the common law in light of the general objects of the Bill of Rights only where no specific right can be relied on by a party challenging a given rule of common law ... ’) (emphasis in original). 38 Uses of interpretive ‘charity’ constitutional rules that amplify that content.’138 When applying the Bill indirectly in term of section 39(2), it does not.

Appropriately does not? Obediently does not? Well, Woolman wants us to read sections 8 and 39 as setting up two cleanly ‘distinct modes of analysis’.139 The suggestion would seem to be that a court, by choosing to frame its inquiry in terms of section 39(2), thereby limits itself to the detection of inconsistencies between the parcel of common law in question and what we might call the normative residue of the Bill of Rights, meaning whatever normative intimations a court can draw from the Bill of Rights as a whole but cannot trace or ascribe to any of the specific rights-naming clauses. If analysis of a rule’s consistency with the specific provisions of the Bill of Rights ‘logically must be prior’ to analysis of its consistency with the normative residue, then the latter process — indirect application — must be reserved for cases in which the former cuts no ice. And conversely, then, if a court, in a private-on-private case turning on common law doctrine, is going to find a ‘direct’ inconsistency — meaning an inconsistency with something contained in one or another the rights-naming clauses — then that court really ought not, on, that occasion, speak the name of 39(2).140

But how, after all, can it be so important that this particular discipline be observed? I do not mean the discipline of searching for clause-based inconsistencies before turning to the question of residual inconsistencies, about which I raise no question here.141

I mean — and mean only — the discipline of not saying ‘39(2)’ when a clause-based inconsistency is what you are after. For this specific practice rule, I can find no reason in Woolman’s magisterial work on application (which is generally rich in jurisprudential and even philosophical reflection) beyond doctrinal tidiness and symmetry. The point is to avoid ‘redundancy’, ‘surplusage’, ‘elision’ in the constitutional text — or, in other words, overlap between the sorts of judicial performances respectively authorised by sections 8

138 Woolman ‘Amazing’ (n 2 above) 768. 139 Woolman ‘Amazing’ (n 2 above) 776; see Part II above. 140 According to Woolman’s preferred construction of secs 8 and 39(2) as a package, sec 39(2) should be understood to stand for the following proposition: Where no specific right can be relied upon by the party challenging a given rule or law or the extant construction of a rule of law the courts are obliged to interpret legislation or to develop the law in light of the general objects of the Bill of Rights. Woolman ‘Application’ (n 17 above) 12. See also at 80 (indicating that, on the preferred construction, ‘s 39(2) is not the intended engine for changes in the common law that flow from the direct application of the Bill of Rights’). 141 A question could, however, be raised about it. See FI Michelman ‘Unenumerated rights under popular constitutionalism’ (2006) University of Pennsylvania Journal of Constitutional Law 121, 135-36, 152-53 (discussing views of Ronald Dworkin and John Rawls). (2008) 1 Constitutional Court Review 39 and 172(1), on the one hand, and 39(2), on the other.142 We would not want two constitutional clauses sharing any of the same work.143

Some might wonder whether Woolman’s interpretive construction of the framers’ design — his strict correlation of clause-bound inspection to sections 8 and 172(1), and residual inspection to section 39(2) — perhaps proceeds from an unrealistically high expectation of formal tidiness and tightness in constitutional drafting. On that score, NK may give a hint. In NK, O’Regan J’s flexing of the respondeat superior standard plainly was driven, not by some normative residue detected in the Bill of Rights, but by attention to the prescriptive contents of several, specific clauses, including those ensuring rights to ‘freedom and security of the person, and in particular, the right to be free from all forms of violence from either public or private sources as well as [the] right to dignity, right to privacy and right to substantive equality.’144 O’Regan J plainly was engaged in the first of the two ‘instances’ of need for common-law development in terms of section 39(2), recalled by her from Thebus: the kind that arises when ‘a rule of the common law is inconsistent with a constitutional provision’.145 Was she, therefore, wrong to file the decision under the head of 39(2) rather than 8?

Might that depend on whether O’Regan J considered herself, after all, to be engaged in ‘development’ of the common law, as contemplated by section 8(3)? Does ‘develop’, there, encompass only an express, terminological re-writing, or does it also take in a court’s choice about how to apply an established standard to an arguably ‘new’ set of facts, as in NK? Section 39(2) speaks to a court’s

142 See Woolman ‘Amazing’ (n 2 above) 771, 777; Woolman ‘Application’ (n 17 above) 10, 11, 12, 55. The ‘two instances’ reading of sec 39(2) in S v Thebus 2003 6 SA 505 (CC); 2003 10 BCLR 1100 (CC) para 28 (‘Thebus’) (where one of the instances is ‘when a rule of the common law is inconsistent with a constitutional provision’) is objectionable to Woolman because it ‘offends the “no surplusage” canon of constitutional interpretation by making parts of FC sec 8 redundant.’ Woolman ‘Application’ (n 17 above) 79. 143 Woolman adduces an incisive and entirely persuasive reason for not allowing sec 39(2) to suck up, exclusively or preemptively, the whole field of constitutionally propelled re-examination of the common law. That reason has to do with the currently imposed, substantial exclusion of the High Courts from such re- examination when conducted under the aegis of sec 39(2). See Afrox Health Care v Strydom 2002 6 SA 21 (SCA); Ex parte Minister of Safety and Security v Walters 2002 4 SA 643 (CC), 2002 2 SACR 105 (CC), 2002 7 BCLR 663 (CC). A like restriction cannot be, and has not been, imposed on High Court re-examination where ‘a party has a colourable claim grounded in the direct application of a substantive provision of the Bill of Rights in terms of section 8'. It is, therefore, of crucial importance, Woolman argues, to resist an ‘elision’ in which sec 39(2) would, in effect, swallow up sec 8. Woolman ‘Application’ 10, 55-56, 64. This argument is, to me, entirely compelling. It does not, however, explain why courts should be barred from ever resorting to sec 39(2) when the detected inconsistency is with a specific provision, which is the issue that my text, above, addresses. 144 NK (n 91 above) para 14. 145 NK (n 91 above) para 16, citing Thebus (n 142 above) para 28. 40 Uses of interpretive ‘charity’

‘interpreting’ law, whereas section 8(3) speaks to a choice between ‘applying’ and ‘developing’ law. Does that difference, perhaps, make section 39(2) more apt to the sort of work accomplished by the Court in NK?

It might seem so, to some; it certainly wouldn’t have to, to all of us. But I do not mean my string of questions to constrain very strongly — much less lead ineluctably — toward a conclusion either way. I only mean it to confirm a truth that Woolman, more than anyone, has recognised and exposed through his meticulous, sterling, searching work on the application conundrum: When the question is that of parsing the exact relations and connections among this Constitution’s several clauses on application and remedies, no fully neat and satisfying answer is to be had. ‘Some surplusage seems inevitable.’146 There are not shells to cover all the peas, and so it is a choice among imperfections — in that sense, a matter, as Woolman says, of ‘interpretation’.147

3.11 The proof of the pudding

But, as Woolman also rightly insists in the same place, it does not follow that grounds are lacking for preferring one interpretation — one construction — over another. One ground of preference might well be the differing ways in which the respective constructions cue and guide, concentrate and dissipate, the judiciary’s deployments of its efforts and energies. Woolman notes that a ‘muscular’ use of section 39(2) ‘could’, in theory, produce the sorts of rights-based, rights-expounding work he finds the Court retreating from.148 But he fears that, in practice, a Court that is allowed to give up too easily on section 8, or to slide too easily to section 39, will be (has been) a Court that too often fails to decide in terms of any of the specific, rights-naming clauses in the Bill of Rights, and accordingly fails too often to engage in the principled elucidation of those clauses that decision in such terms entails. That states a legitimate concern, on its face by no means implausible. Still, the claim is an empirical one, subject to testing as such. NK, I have just suggested, offers a bit of evidence against it. (So, for that matter, does Thebus, where the Constitutional Court, under the sign of section 39(2), considered in great detail whether a prosecutor’s reliance on the common-purpose doctrine ‘trenches on’ or ‘denies’ the constitutional rights of a criminal accused not to be deprived of freedom without just cause and to be presumed innocent.)149

146 Woolman ‘Application’ (n 17 above) 71. 147 See Woolman ‘Application’ (n 17 above) 13 n 1. 148 Woolman ‘Application’ (n 17 above) 766 n 6. 149 Thebus (n 142 above) paras 36, 42. (2008) 1 Constitutional Court Review 41

To what extent, we might ask, does the Constitutional Court’s work in NM tend to bear out Woolman’s empirical concern? Start with O’Regan J. Her NM opinion certainly exemplifies the sort of elucidation that Woolman associates with direct application pursuant to section 8.150 If O’Regan J had given clear indication that she was proceeding in terms of section 39(2), as distinct from section 8, that would have told against Woolman’s empirical claim. However, she did not. What might be a bit worrisome from Woolman’s corner, though, is the further fact that O’Regan J did not ever bother to say under which set of constitutional auspices — 8 or 39(2) — she was undertaking to develop the common law. (But Langa CJ, concurring with O’Regan J’s proposed development, did make clear that he understood himself to be proceeding in terms of section 39(2).)151 Neither did O’Regan J ever pause to declare invalid the extant common law of privacy invasion. All this perhaps does tell a bit against Woolman, insomuch as it might suggest that O’Regan J believes it does not matter which auspices she invokes, which would contradict Woolman’s insistence that it does matter, not just because an order of march has been prescribed, but because marching under section 39(2) has an effect on the justices of releasing them from the very work of clause-by-clause exposition that O’Regan J’s NM opinion so clearly exemplifies.

What about the NM majority? Madala J, denying that he was engaged in any sort of modification of the common law, had no occasion to claim any set of constitutional auspices for modifying it. But suppose he had said directly that his finding of ‘intention’, on Smith’s part, to disclose private facts without authorisation to do so from the persons concerned, resulted from a shading of the common law category of ‘intention’ under pressure of the Bill of Rights, specifically in the matter of privacy protection. What more could Madala J then have said by way of explaining his action, than that the disclosure went to the ‘inviolable core’ of privacy as previously defined in Bernstein,152 and that the privacy-protection principle of section 14 requires exposure to civil liability of anyone who publishes core-private facts when the publisher ‘at least foresaw the possibility that the consent had not been given to the disclosure’?153

Madala J did not say those things in that way, nor could he have, because he did not purport to be judging this case in terms of the Bill of Rights, à la NK. But can we not fairly guess from his opinion that he

150 See NM (n 5 above) paras 126-35 (constitutional right to privacy); paras 144-46 (constitutional right to freedom of expression). 151 See NM (n 5 above) para 120. 152 Bernstein v Bester NO 1996 2 SA 751 (CC); 1996 4 BCLR 449 (CC) para 77. Compare NM (n 5 above) paras 33, 41-42. 153 Compare NM (n 5 above) para 64. 42 Uses of interpretive ‘charity’ would have said them, if he had undertaken to explain himself in such terms? If so, then NM is, to that very modest extent, a sign that section 39(2) does not necessarily, and does not always, function as a lure to a form of judicial decision-making that ignores and by-passes the regulative content of the rights-naming clauses of the Bill of Rights. Masiya, as we are about to see, provides some further evidence to like effect.

4 MASIYA

4.1 The case and a critique

The state charged Masiya with the common law crime of rape. Under the law as it then stood, rape was defined in terms of non-consensual, vaginal penetration (and thus could involve female victims only). The evidence in Masiya’s case showed forcible, anal penetration of a nine- year old girl.154 Anal penetration of a female or a male could support a conviction of indecent assault, but not of rape. Indecent assault is classified as the lesser crime, and it carries a lesser penalty than rape.155

In the course of the state’s prosecution of Masiya, questions arose as to whether the extant common law definition of rape fell to be declared invalid as inconsistent with the Constitution, and whether that law ought to be developed so as to cover Masiya’s proven act of anal penetration.156 The two questions are not unrelated, but neither are they necessarily coupled. In the event of an adjudicative

154 See Masiya (n 4 above) paras 6, 13. 155 See Masiya (n 4 above) para 15 (indicating that ‘indecent assault attracts more lenient sentences than rape’). 156 There was, of course, a further question about retrospective application to Masiya’s case of any extended definition the Constitutional Court might direct. The Court answered ‘no’ to that question, see Masiya (n 4 above) para 56, even as it also answered ‘yes’ to the question of whether the advent of Masiya’s case before the courts offered an apt occasion for developing the definition. See Masiya (n 4 above) para 51: If the definition of rape were to be developed retrospectively it would offend the constitutional principle of legality. However, if it was to be accepted that the principle of legality was a bar to the development of the common law, the courts could never develop the common law of crimes at all. Such a conclusion would undermine the principles of the Constitution which required the courts to ensure that the common law was infused with the spirit, purport and objects of the Constitution. From a US constitutional lawyer’s perspective, major worries remain, not only about prospective-only application of a common law ruling, but also, and relatedly, about South African constitutionalism’s evident relaxation of what we would call a ‘case or controversy’ prerequisite to judicial jurisdiction. See generally R Fallon et al Hart & Wechsler’s the federal courts and the federal system 5 ed (2003) 73-85. Woolman’s critique of the Constitutional Court’s Masiya decision being in no way directed to such worries, I say nothing more about them here. (2008) 1 Constitutional Court Review 43 declaration of the current definition’s constitutional invalidity, corrective development of the common law might well follow on as a part of a just and equitable remedial order, in terms of Constitution s 172(1).157 But even in the absence of such a declaration, a court might arguably still find itself bound by Constitution section 39(2) to develop the definition so as cover Masiya’s case.158

On what grounds, then, might the Constitutional Court, sitting in Masiya’s case, have held constitutionally incompatible the extant common law of rape insofar as it fails to cover cases of anal penetration? Not, in the first place, on any theory of the law’s infringement of Masiya’s constitutional rights or those of other putative offenders covered by the definition. For, obviously,

the current definition of rape criminalises unacceptable social conduct that is in violation of constitutional rights. It ensures that the constitutional right to be free from all forms of violence, whether public or private, as well as the right to dignity and equality are protected.159

In other words, any limiting effect of the law of rape on the liberty of putative rapists is obviously and amply justified in terms of Constitution section 36(1).160 It easily follows that any declaration of the current definition’s unconstitutionality would have to issue in contemplation of constitutionally protected interests, not of those who might find themselves charged with assaultive crime, but of the class consisting of all who might suffer sexual assault — the general public, more or less. Thus, it was with a view to ‘afford[ing] society the full protection of the Constitution’ that the trial court in Masiya’s

157 It seems that, in Masiya’s case, Constitution sec 8 can also serve as the ground of an order to develop the common law, inasmuch as an antecedent declaration of the invalidity of the common law definition of rape might be said to result (in the terms of sec 8(2)) from an application of certain provisions in the Bill of Rights to the conduct (or to common law regulating the conduct) of Fanuel Sitakeni Masiya, a natural person. Since Woolman generally argues in terms of a choice between ss 8 and 39(2), as the constitutional authorisation to be invoked by the Court for some ensuing directive for common law modification, and since nothing seems to depend on whether a remedial order would issue in terms of sec 172(1)(b) or in terms of sec 8, I shall, in what follows, refer to the one or the other as the context most naturally suggests. 158 See Masiya (n 4 above) para 33 (quoting from Carmichele (n 59 above) para 36, and also citing Thebus (n 152 above) paras 28-31): ... ‘[C]ourts must remain vigilant and should not hesitate to ensure that the common law is developed to reflect the spirit, purport and objects of the Bill of Rights ... whether or not the parties in any particular case request the Court to develop the common law under section 39(2).’ Where there is deviation from the spirit, purport and objects of the Bill of Rights, courts are obliged to develop the common law by removing the deviation. 159 Masiya (n 4 above) para 27. 160 See also Constitution sec 12(1)(a): ‘Everyone has the right to freedom and security of the person, which includes the right — (a) not to be deprived of freedom arbitrarily or without just cause’. 44 Uses of interpretive ‘charity’ case took up sua sponte the question of extending the common law of rape to acts of anal penetration.161

For convenience, we might imagine that it would be the Director of Public Prosecutions who presses the claim for some development of the common law that would result in Masiya’s conduct being punishable as rape.162 The DPP, pointing to the state’s constitutionally imposed duty to protect and fulfil the rights in the Bill of Rights,163 might charge the received definition of rape with a constitutional sin of omission, to wit, its omission to make acts of anal penetration punishable as rape. Impermissibly under-protected as a result of that omission, the DPP might say, are the right to freedom from private violence, conferred by section 12(1)(c),164 and the right to have one’s dignity respected and protected, conferred by section 10.165 Alternatively or additionally, the DPP might urge a declaration of invalidity based on a sin of commission by the current law, to wit, commission of an unfair discrimination prohibited by Constitution section 9(3).166 It is unconstitutionally unfair, the DPP might urge, for the state, through its common law, to offer lesser levels of recognition, and of protection-by-deterrence, to male than to female sufferers of forced, sexualised, bodily penetration, or (within the class of female victims) to sufferers of anal than to sufferers of vaginal penetration.

A judicial modification of the common law definition of rape, presented as a remedy for a constitutional violation declared on any of the grounds aforesaid, would be an instance of what Woolman classifies as ‘direct’ application of the Bill of Rights. But a court sitting in Masiya’s case might also arguably be authorised and required to undertake an extension of the common law definition of rape to reach cases of anal penetration, even in the absence of any finding of the classical definition’s inconsistency with any clause in the Bill of Rights. Start by setting aside the Bill of Rights entirely. Perhaps a court might simply decide that considerations of policy and justice call, under present social conditions, for extension of the common law definition of rape so as to take in cases of anal penetration, either of

161 Masiya (n 4 above) paras 8-9. 162 In point of fact the DPP did not urge any modification of the common law definition of rape but rather sought conviction for indecent assault. See Masiya (n 4 above) para 7. I make the opposite assumption for the sake of expository convenience. 163 See Constitution sec 7(2): ‘The state must respect, protect, promote and fulfil the rights in the Bill of Rights.’ 164 ‘Everyone has the right to freedom and security of the person, which includes the right — ... to be free from all forms of violence from either public or private sources’. 165 ‘Everyone has inherent dignity and the right to have their dignity respected and protected.’ 166 ‘The state may not unfairly discriminate directly or indirectly against anyone on one or more grounds, including race, gender, sex ...’. (2008) 1 Constitutional Court Review 45 females only or of males as well. The court might then invoke its ordinary, inherent power to develop the common law as and when occasioned by the interests of justice.167 In doing so, the court might or might not refer, beyond its general authority to keep the common law attuned to the times and to the interests of justice, to its more specific obligation, under section 39(2), to see to the conformation of the common law with the spirit, purport, and objects of the Bill of Rights.

In Masiya, the Constitutional Court, in an opinion by Nkabinde J, concluded by issuing an order directing a revision of the common law definition of rape to include acts of anal penetration of females.168 It did so while (a) expressly declining to issue a declaration of the invalidity (unconstitutionality) of the non-extended definition, and (b) declining to decide whether acts of anal penetration of males would be covered in the extended definition of rape resulting from the adjudication in Masiya’s case — thus, in effect, retaining for the present a definition of rape in which only females can be victims.

Those two features of the Court’s Masiya decision, (a) and (b), are the targets of vigorous complaint from Woolman. Both stand, in his view, as further documentation of the Constitutional Court’s unaccountable habit — and that habit’s doctrinally debilitating consequences — of refusal to engage in direct application of the Bill of Rights or to undertake the rule-like expositions of the contents of the several rights-protecting clauses that direct application is expected to induce.

As with Woolman’s objections to the Court’s performance in NM, these objections to Masiya seem to me in some part answerable. The main idea continues to be this: We want to see what happens if we strain to turn these very features in the judgment — the ones that we may initially find puzzling or disturbing — into keys to discovery of the Court’s own understanding of the legal problem presented by the case; and perhaps, through that, to notice something true — I do not mean necessarily welcome — about the meaning and structure of the law. When we train that sort of effort on the Constitutional Court’s work in Masiya, we may find that Woolman’s objections lose some – though it will not be all — of their sting.

167 See Constitution sec 173 (‘The Constitutional Court, Supreme Court of Appeal and High Courts have the inherent power to protect and regulate their own process, and to develop the common law, taking into account the interests of justice.’). 168 See Masiya (n 4 above) para 74(5). 46 Uses of interpretive ‘charity’

4.2 The decision’s gravamen: Under-protection

In Masiya, the Constitutional Court concluded that the extant common law of rape, although not prone to a finding of invalidity, nevertheless fell to be developed under pressure from the spirit, purport, and objects of the Bill of Rights. By my reading — and this point is central to my examination of Woolman’s critique — the decisive pressure came (in the Court’s view) not from the Bill’s prohibition against unfair discrimination, but rather from its positive commitments to protection of persons against violence and indignity. The Court does mention section 9 — ‘flirts with’ it, if you like169 — but does not finally rest its action on considerations sounding in equality. As I have understood the Court’s opinion, the decisive fault it finds in the common law lies not in any unfair discrimination committed by the law but rather in protection omitted from it: not in the law’s perpetration of inequality, but in its under-protection of constitutionally guaranteed dignity and freedom from violence. A complaint of under-protection is, of course, a matter distinct in principle from a complaint of unequal treatment. By what I presume to offer as the ‘charitable’ reading of Nkabinde J’s opinion, it is the under-protection complaint that carries the day in Masiya.

Oh, yes, to be sure, the ‘issues raised’ in the case included ‘equality’ — specifically citing section 9(1)’s guarantee to everyone of ‘equality before the law’ and ‘equal protection and benefit of the law’.170 And, yes, Nkabinde J made the comparative observation that the trauma associated with anal penetration is ‘just as humiliating, degrading, and physically hurtful’ as that associated with vaginal penetration.171 She said, as well, that inclusion of anal penetration in the definition of rape ‘will increase the extent to which the traditionally vulnerable and disadvantaged group will be protected by and benefit from the law’.172 But all of those remarks are perfectly consistent with the view that the current law is constitutionally deficient in the specific respect that it under-protects the rights to dignity and freedom from violence. They could all easily have been written by a judge applying a bill of rights that contains no equality clause.

As Nkabinde J’s opinion proceeds, we read the following: It is the office of the crime of rape to ‘ensure[] that the constitutional right to be free from all forms of violence, whether public or private, as well as the right to dignity and equality are protected.’173 It is

169 Woolman ‘Amazing’ (n 2 above) 768 n 10. 170 Masiya (n 4 above) para 18 & n 32. 171 Masiya (n 4 above) para 39. 172 As above. 173 Masiya (n 4 above) para 27. (2008) 1 Constitutional Court Review 47 specifically with regard to these objectives that the current definition is fatally ‘under-inclusive’.174 ‘Non-consensual anal penetration of women and young girls such as the complainant in this case constitutes a form of violence against them equal in intensity and impact to that of non-consensual vaginal penetration.’175 Extending the definition to include anal penetration will ‘protect the dignity of survivors’.176 It will ‘increase the extent to which [a] traditionally vulnerable and disadvantaged group will be protected’ against ‘humiliating, degrading and physically hurtful’ treatment, and ‘therefore’ — meaning, in that particular respect — it will ‘harmonise the common law with the spirit, purport and objects of the Bill of Rights.’177

In sum, it is all about a constitutionally required level of protection, not about the distinct injury to the self or to dignity that flows from being subjected to second-class treatment. The Bill of Rights sections in the foreground here are 10 and 12, not 9. The Court convicts the current law of a constitutional sin of omission of due protection; it does not convict it of a constitutional sin of commission of unfair discrimination.

4.3 A case of judicious avoidance?

In Woolman’s view, this short-changing of section 9 counts as unimitigated vice in the decision. In mine, it can claim some support in ordinary judicial prudence. What, after all, is the context — the field — of the differential treatment that troubles Woolman?178 It is the grading of crimes within nests of including and lesser-included offenses. (It is not as if Mr Masiya escapes conviction of crime — serious crime — if he is convicted of indecent assault and not rape.) What is involved is the schedule of penalties devised by the lawmaker for more and less aggravated variations on the same core offense. Plainly, it is not against the grain of the Bill of Rights, or contrary to any clause contained therein, for lawmakers to draw such lines. To the contrary, we should think it monstrous if they did not.

From an equality-clause standpoint, then, an important question presented to the Constitutional Court by the Masiya appeal was one of best judicial practice: whether it would be generally in the interests of justice, and conducive to long-term optimal performance

174 As above. 175 Masiya (n 4 above) para 37. 176 Masiya (n 4 above) para 38. 177 Masiya (n 4 above) para 39. 178 Note that I am not yet dealing with the issue raised by the Court’s postponement of the question of the new rape definition’s application to cases involving male victims. 48 Uses of interpretive ‘charity’ by the judiciary of its role in the over-all governmental order of a democratic South Africa, for the Constitutional Court to swing open its doors to complaints of unconstitutionally unfair discrimination, assertedly wrought by the lawmaker’s choices of where and how to draw the lines required by the gradations of penalties (for more and less aggravated incidents of the same class of crime) that the Bill of Rights and the rule of law almost certainly demand.179 If you think that is an easy question, it must be because you feel sure that answer is ‘no’. If you feel the answer might well be ‘yes’, you must nonetheless grant that answer is not immediately self-evident. The Constitutional Court, I accordingly conclude, had respectable reason for hesitating to tackle the question in a case where the need for a challenged law’s correction was sufficiently clear on constitutionally resonant grounds of substantive under-protection, equality considerations being held at bay.

4.4 A case of direct application?

In her Masiya judgment, Nkabinde J concludes that the extant common law definition of rape ‘is not inconsistent with the Constitution’.180 Accordingly, no declaration of invalidity is to issue, and no development of the common law definition of rape is to ensue in the guise of a remedy for constitutional invalidity.181 Rather, ‘the definition is to be extended ... so as to promote the spirit, purport and objects of the bill of rights.’182 Evidently in play, then, is an exercise of the Court’s inherent authority, in terms of sections 173 and 39(2), to develop the common law, as opposed to its judicial- review power, conferred by section 172(1), to declare invalidity and craft just and equitable remedial orders.183 Just as Woolman says.

Let us next notice what stands immediately behind Nkabinde J’s refusal to declare the extant rape law unconstitutional and hence invalid. Here is the crucial passage:

The current definition of rape criminalises unacceptable social conduct that is in violation of constitutional rights. It ensures that the constitutional right to be free from all forms of violence, whether public or private, as well as the right to dignity and equality are protected.

179 I have in mind here a general standard of appraisal of the Constitutional Court’s performance closely aligned to the one projected by Roux ‘Pragmatism’ (n 16 above). 180 Masiya (n 4 above) para 27. 181 Masiya (n 4 above) para 51. 182 Masiya (n 4 above) para 27. 183 See Masiya (n 4 above) para 45 (‘[T]he extension of the common-law definition of rape to include non-consensual anal penetration of females will be in the interests of justice and will have, as its aim, the proper realisation by the public of the principles, ideals and values underlying the Constitution.’). (2008) 1 Constitutional Court Review 49

Invalidating the definition because it is under-inclusive is to throw the baby out with the bath water. What is required then is for the definition to be extended instead of being eliminated so as to promote the spirit, purport and objects of the Bill of Rights.184

We can all see the conceptual problem to which Nkabinde J is calling attention there. Woolman sees it and chalks it up to ‘fuzzy logic’, a point to which we shall return.185

Nkabinde J suggests a comparison of the rape crime at issue in Masiya with the erstwhile common law crime of sodomy, at issue in the ‘Sodomy’ case.186 Let us follow her lead. Decidedly not by way of ‘a process of developing the common law’, but by way of ‘a direct application of the Bill of Rights to a common-law criminal offence’, the sodomy crime was simply wiped off the books, declared ‘constitutionally invalid in its entirety’.187 If there, why not here?

Because it is not true in Masiya’s case, as it was in the Sodomy case, that the parcel of common law at issue ‘subjected people to criminal penalties for conduct which could not constitute a crime in our constitutional order’,188 and therefore could and should be declared null and void by reason of inconsistency with provisions in the Bill of Rights. Rather, the opposite is true. The rape crime, as historically defined, ‘criminalises unacceptable social conduct that is in violation of constitutional rights’.189 Nullification of the sodomy crime as defined by the common law leaves all as it should be, constitutionally speaking. Nullification of the rape crime as defined by the common law leaves ... what?

Of course, legal insiders know how to spin out an acceptable answer: It leaves, we would say, the common law in need of remedial modification – say, by a remedy of ‘reading in’ — as a cure for its constitutional sin of omission, its defect of inconsistency with sections 10 and 12(1) of the Bill of Rights. That is the gist of Woolman’s complaint against Nkabinde J’s logic.190 The complaint is entirely valid. And yet something also feels right about Nkabinde J’s detection of the insult to ordinary reason that lies in condemning, as inconsistent with the Constitution, a legal prohibition whose presence on the books the Constitution positively demands. A remedial authority that would subtend on such a strange-sounding declaration,

184 Masiya (n 4 above) para 27 (citations omitted). 185 Woolman ‘Amazing’ (n 2 above) 768 n 10. 186 See Masiya (n 4 above) para 27. 187 ‘Sodomy’ Case (n 20 above) para 90. 188 Masiya (n 4 above) para 27. 189 As above. 190 See Woolman ‘Amazing’ (n 2 above) 768 n 10 (‘The court appears to conclude that a finding of invalidity of a rule of common law can result only in a simple declaration of invalidity.’). 50 Uses of interpretive ‘charity’ she quite clearly conveys, ought not to be the first resort of the Court when some other resort is available, some other path of reasoning to a Constitution-satisfying outcome. And indeed another path is available, providentially supplied by the framers. The ram in the thicket (block that metaphor!) is section 39(2).

The choice for section 39(2) is understandable in those terms, if not fully defensible. My reason for drawing that distinction will soon be on the table, but first we should confirm why Woolman is right to criticise the choice. To work the question through, we can start by imagining that the Constitution had been so written as to foreclose recourse in this case to section 39(2); say, it flatly prohibited the courts from performing terminological surgery on the common law, except as a remedy for constitutional invalidity declared. With such a prohibition in place, should the Constitutional Court have been willing to declare invalid the extant common law rape crime, that being the only path left open to them to a judicial cure for what they see as a constitutionally unendurable sin of omission in the law, that is, its failure to impose appropriately severe condemnation and penalty on forcible acts of anal penetration? If you ask me, a Court sitting in a transformative constitutional order would have no choice but to do just that, awkward as the declaration of invalidity might feel to the Court.

The Constitutional Court, it seems, took the view that it did not have to perpetrate that awkwardness in Masiya’s case. They had access to the omission-fixing cure in question via section 39(2), so why should they have engaged in the seeming doubletalk of declaring constitutionally invalid a constitutionally required parcel of the extant law — only in the next breath to restore what they have just condemned as unconstitutional, in the guise of an instantaneously ensuing ‘remedy’? Why mangle straight-line, ordinary reason in that way, when the drafters have provided a fit alternative?

A conclusive answer to that question is prefigured by my imagined case of the constitutional prohibition on judicial rewriting of the common law, except as a remedy for declared constitutional invalidity. Notice that, with respect to statutes, such a constitutional prohibition is not imagined, it is actual! In terms of section 39(2), statutes are open to judicial interpretation but not to judicial ‘development’. By its holding that the old rape definition’s omission of anal penetration is ‘not inconsistent with the Constitution’, the Court has, in principle, disabled itself from curing an identically flawed act of Parliament should one ever happen to issue. And that, surely, is an unacceptable implication.191

191 I am indebted for this point, which I had failed to see, to an anonymous referee. (2008) 1 Constitutional Court Review 51

The Masiya Court thus erred in withholding a declaration of invalidity of the old common law of rape. Its error, however, consists only of resort to section 39(2) where section 172(1) seemed to the Court not to fit the case while section 39(2) seemed fully serviceable. The error is not at all symptomatic of a flight from the burdens of clause-by-clause exposition of constitutional content that attend upon direct application.

The focus of the crime of rape, writes Nkabinde J, ‘ is on the breach of “a more specific right such as the right to bodily integrity” and security of the person and the right to be protected from degradation and abuse.’192 Presumably, there was something, some line of thought, that Nkabinde meant to convey by that direct allusion to O’Regan J’s opinion for the Court in Dawood. What, then? Well, that is the passage in Dawood where O’Regan J exactly differentiates between dignity as ‘a value that informs the interpretation of many, possibly all, other rights’ (or, in other words, dignity acting as a condiment in the spirit-brew of the Bill of Rights) and dignity as ‘a justiciable and enforceable right that must be respected and protected’. Dignity, then, as a target of ‘direct’, not ‘indirect’, application of the Bill of Rights, as Woolman understands those terms.

There, and in other passages I have quoted above, Nkabinde J’s opinion makes sufficiently clear that the Masiya Court’s modification of the common law is designed to respond, not to some spirit summoned from the vasty deeps of the Bill of Rights, but rather to certain, specifically identified, distinctly articulated demands of sections 10 and 12(1)(c) of the Bill — most particularly, the Constitution’s mandate to protect people against private violence. As used (however mistakenly) in the Masiya case, section 39(2) is not serving as an escape-hatch from the Court’s responsibility for clause- by-clause elucidation of the contents of the Bill of Rights. It is rather serving as a handy authorisation for a moderately drastic bit of judicial surgery on the common law, in circumstances where the Court finds inapposite section 172(1)(b)’s license to operate qua remedy for invalidity.

In making this use of section 39(2), Nkabinde J followed precedent set by the Constitutional Court in Carmichele. Just as the Court there chose not to condemn, as constitutionally out of line, whatever affirmative, protective duties were already imposed on the police by the extant common law of delict, so the Court in Masiya chose not to condemn, as constitutionally out of line, whatever liabilities to criminal conviction were already imposed on perpetrators of sexual

192 Masiya (n 4 above) para 25, quoting from Dawood v Minister of Home Affairs 2000 3 SA 936 (CC); 2000 8 BCLR 837 (CC) para 35. 52 Uses of interpretive ‘charity’ violence by the extant common law of crimes. In both cases, the extant impositions of duty and liability were in order as far as they went. In both, the issue was one of broadening those impositions; there was no question of contracting, much less deleting them. Just as the Constitutional Court in Carmichele found that development of the common law under the aegis of section 39(2) was the apt course to take in those circumstances, so did the Court in Masiya. In both cases, moreover, the Court made sufficiently clear its view that a duty- and liability-extending modification of the common law was required not just by some immanent, residual, order of values to be distilled from the Bill of Rights as a whole, but also by an application of the Constitution’s specific guarantees respecting dignity and freedom and security of the person.193

It seems we might extract from the Carmichele/Masiya pair a general guideline of South African constitutional law, as follows: When the constitutional fault detected in some duty- and liability- imposing parcel of the common law lies not at all in the imposition of duty and liability, but only in the failure to define the type of duty- and liability-raising conduct in question with sufficient protective effect, then the authorisation for judicial surgery on the common law is to be found not in the court’s judicial-review/ remedial power granted by section 172(1), but rather in its inherent power to develop the common law in terms of section 39(2). The reason for this result is not any thought that under-protection in duty- defining laws can only offend against some residual ‘spirit’ of the Bill of Rights as a whole, but not against any clear, core, animating principle of any particular rights-naming clause. The reason for the result simply is that the section 172(1)(b) remedial power is expressly conditioned on an antecedent declaration of inconsistency of some law with the Constitution, and, in this class of cases — so thinks the Court — there is nothing to declare inconsistent.

In sum: At stake here is a notional distribution of the source of judicial authority and obligation to develop the common law, as between two clauses in the Constitution. Not at stake is a flight from direct application of the Bill of Rights, as Woolman defines ‘direct’. Rather, the Carmichele/Masiya pair (and let us also recall NK and Thebus) shows the Constitutional Court rejecting Woolman’s preferred construction of the relations among sections 8, 39(2), and 172(1), as outlined in Part 2 of this article.

193 See Carmichele (n 59 above) para 44 (‘[T]here is a duty imposed on the state and all of its organs not to perform any act that infringes these rights. In some circumstances there would also be a positive component which obliges the state and its organs to provide appropriate protection to everyone through laws and structures designed to afford such protection.’); Masiya (n 4 above) paras 25, 27, 37, 38, 39, reviewed above, text accompanying notes 169-177. (2008) 1 Constitutional Court Review 53

Woolman, of course, is fully aware of this gap between his own preferred construction and the Court’s ‘black letter’ construction. He is nevertheless careful not to suggest that the gap must necessarily make a world of difference in practice.194 To the contrary, he explicitly points out that the Constitutional Court, using its own construction, could make a ‘muscular’ use of section 39(2) that ‘could generate the kinds of decisions and constitutional rules’ he is looking for.195 But he sees this possibility as ‘entirely hypothetical’, asserting that the Constitutional Court has not, in fact, ‘deployed s 39(2) in the muscular fashion that would be necessary to rebut my primary arguments.’196 One can assume, though, that were the Court, acting under the aegis of section 39(2), to search specifically for incompatibility between some challenged law and particular provisions of the Bill of Rights, that should count for Woolman as direct application.197 I believe that Masiya — like Carmichele and NK — is most naturally read as just such a case, as opposed to a case of a non-differentiated appeal by the Court to some vaguely conjured ‘spirit’ of the Bill of Rights entire. If Woolman does not see the cases that way, that might possibly be because his normative commitments are getting in the way, by which I mean his insistence that section 8 and 39(2) must, on pain of redundancy, have mutually ‘independent’ purposes,198 entailing distinct tests for the common law’s liability to revision: incompatibility with ‘the specific provisions of Chapter 2’ under section 8, incompatibility with the normative residue under section 39(2).199

4.5 The sex-discrimination issue: Easy or hard?

For several pages, now, I have been reflecting on the Constitutional Court’s designation, in Masiya, of section 39(2) — and not section 8 or 172(1) — as the source of its authority to review the High Court’s order in respect of developing the common law of rape. I turn now to another feature of the Constitutional Court’s Masiya decision that has excited Woolman’s concern, namely, the Court’s postponement of the question whether the definition of rape, once extended to cover anal as well as vaginal penetration, must also be extended to cover penetration of males as well as females. My first suggestion (but then things will grow more complicated) will be that Woolman’s difference with the Court in this regard depends on a substantive disagreement concerning the merits of a particular constitutional claim. More precisely, it has to do with a conflict between Woolman’s and the

194 See generally Woolman ‘Application’ (n 17 above). 195 Woolman ‘Amazing’ (n 2 above) 766 n 6. 196 As above. 197 See as above. 198 Woolman ‘Amazing’ (n 2 above) 769. 199 See Woolman ‘Amazing’ (n 2 above) 769, 771. 54 Uses of interpretive ‘charity’

Court’s respective preliminary judgments regarding the merits of a claim against the rape definition’s consistency with Constitution sections 9(1) and 9(3), if that definition, once extended to cover anal penetration of females, is not immediately extended again to cover anal penetration of males.

Of course, Masiya’s case does not squarely present that question; only a prosecution for rape in a case of anal penetration of a male could do that. Woolman does not — and I doubt he ever would — raise a blanket objection to judicial hesitancy to reach for hard constitutional questions whose resolution is not necessary to a complete disposition of the case sub iudice. What he does clearly think is that the constitutional violation is, in this instance, so exceptionally and blatantly clear that the Constitutional Court could appropriately have resolved it on the spot.200 The hangup is that the Court apparently does not share Woolman’s perception of the obviousness of the answer to the equality clause question thus posed. If the question really is debatable in light of the applicable precedents, then the Court’s forbearance to decide on the spot will seem (to an American constitutional lawyer, at any rate) to be routinely predictable for any court in a constitutional democracy, disposing over an assigned power of judicial review; the forbearance simply cannot then qualify as probative for a general complaint about a judicial addiction to excessive minimalism. And the Court, I am about to suggest, had detectible reason to assess the equality clause question as debatable under its own precedents.

In the view of the trial court in Masiya’s case — as well as in the apparent view of Woolman201 — restricting the newly extended rape law’s protection to women is ‘irrational and totally senseless’, an ‘arbitrary discrimination’202 — and thus, it would seem, a direct violation not only of section 9(3) but also section 9(1), according to the ‘tabulation’ for a section 9 inquiry laid down by Goldstone J in Harksen.203 In the view of the High Court, the restriction results in ‘discriminatory sentencing’, an apparent invocation of section 9(3)’s prohibition of unfair discrimination.204 What is more, the Constitutional Court agrees that ‘non-consensual anal penetration ...

200 Having noted his agreement with Langa CJ’s ‘unassailable’ view that ‘it makes no sense to distinguish between men and women’ with respect to the gravity of the crime of anal rape, Woolman goes on to say that ‘in these circumstances’ the Court was wrong to refrain from addressing the sex-specificity issue in a case whose facts did not require its resolution. Woolman ‘Amazing’ (n 2 above) 767. 201 As above. 202 Masiya (n 4 above) para 9. 203 See Harksen v Lane NO [1997] ZACC 12; 1997 11 BCLR 1489; 1998 1 SA 300 para 53 (item ‘(a)’ in the tabulation). 204 Masiya (n 4 above) paras 2, 15. (2008) 1 Constitutional Court Review 55 is [no] less degrading, humiliating and traumatic’ for males than for females.205 But, again, the hangup is that the Constitutional Court does not allow that it automatically follow from this equivalence that a gender-specific definition of rape must be unconstitutional.206

How could it not? The Court does not spell out its answer, but those conversant with the Court’s equality jurisprudence can see one sitting there. Section 9(3) prohibits unfair discriminations on grounds of sex, but not all sex-based differentiations are unfair (and much less are all of them arbitrary and thus in violation of section 9(1)).207 According to the Court’s carefully developed and precisely articulated doctrine, discrimination is ‘unfair’ when and insofar as it has the effect of impairing dignity or inflicting some comparably serious harm.208 That may not be uncontroversial doctrine or (for aught I know) the doctrine favored by Woolman, but substantive doctrine it surely is, quite carefully and elaborately wrought by the Constitutional Court.

So consider: In the view of the Constitutional Court,

it is ... widely accepted that sexual violence and rape not only offends the privacy and dignity of women but also reflects the unequal power relations between men and women in our society.209

Rape, said the Court, is understood in South Africa as being ‘not simply an act of sexual gratification, but one of physical domination’ — ‘an extreme and flagrant form of manifesting male supremacy over females.’210 Historically, said the Court, ‘rape has been and continues to be a crime of which females are its systematic target.’211 In the prevailing societal view as reported by the Court, not only do females thus bear the major brunt of the degradation and brutality inflicted by sexual violence, but where such violence is consciously or unconsciously directed to the subordination of some sex or gender — as it somewhat systematically is, according to what the Court takes to be the dominant understanding — that sex or gender is the female or feminine, not the male or masculine.

Perhaps that all adds up to a controversial batch of observations about prevailing South African understandings. (I do not know and shall not try to say.) But let us ask: Accepting the observations as correct, do they really provide a grader of penalties for crimes with a

205 Masiya (n 4 above) para 30. 206 As above. (‘That this is so does not mean that it is unconstitutional to have a definition of rape which is gender-specific.’). 207 See Harksen (n 203 above) para 53. 208 See Harksen (n 203 above) paras 46, 50, 51(c). 209 Masiya (n 4 above) para 28. 210 Masiya (n 4 above) para 36. 211 As above. 56 Uses of interpretive ‘charity’ decisive reason to tilt the balance more favorably to liberty and mercy, as against prevention and condemnation, when sexual violence is directed against males? On first look, I find it far from obvious that they do. I cannot, however, deny that the question is a hard one that requires careful thought. The Court’s observations do, then, point to a possible reason, not inherently injurious to the dignity of men or of women — any more than, in Hugo, President Mandela’s reasons for restricting his commutation order to mothers were found injurious to the dignity of men or of women212 — why a lawmaker might decide to treat sexual violence toward women as a more seriously graded crime than sexual violence toward men. If not unfair in Hugo, why unfair here? One can so inquire in good faith, without for a moment doubting that that anal rape is a monstrous affront to the dignity and security of male as well as female victims — any more than any justice sitting in Hugo could have doubted that imprisonment is hard for men as it is for women.213

Granted, ‘strong arguments’ could be made — and were, to the Constitutional Court in Masiya’s case — ‘to the effect that gender- specificity in relation to rape’ must be deemed constitutionally unfair insofar as it undoubtedly reflects ‘patriarchal stereotypes inconsistent with the Constitution.’214 Exactly parallel arguments were made in Hugo, including by justices who found, contrary to the majority of the Constitutional Court, that the President’s gender- specific deployment of his pardon power was indeed an act of unfair, sex-based discrimination.215 In Hugo, the claims of constitutionally unacceptable stereotyping met up with both respectful consideration and formidable rebuttal from the majority side.216 In Masiya, the Court did not reject the parallel arguments, it rather left them for another day.217 Perhaps there are members of the Court engaged in, or open to, reconsideration of the position taken in Hugo. The matter at issue is obviously difficult, deep, and of major import for South African sex-equality jurisprudence. It may well currently divide the justices, just as it divided them in Hugo. If they wait, views might undergo maturation and moderation. Parliament may speak, and that could make a difference. In Hugo, the sex-discrimination question could not wait; if the discrimination were to be found unfair and unjustified, a remedial question would then arise that conceivably could have issued in Mr Hugo obtaining release from prison. In Hugo, the sex-discrimination question could not wait; Mr Hugo's claimed right against unfair, sex-based discrimination was directly and

212 See The President of South Africa v Hugo 1997 4 SA 1 (CC); 1997 6 BCLR 708 (CC). 213 See Woolman ‘Amazing’ (n 2 above) 767. 214 Woolman ‘Amazing’ (n 2 above) para 29. 215 See Hugo (n 212 above) para 80 (Kriegler J); paras 93-94 (Mokgoro J). 216 See Hugo (n 212 above) paras 37-39 (Goldstone J); para 113 (O’Regan J). 217 See Masiya (n 4 above) paras 29, 36. (2008) 1 Constitutional Court Review 57 immediately at stake. In Masiya, by contrast, the question can wait; the prisoner at bar has nothing at stake in the resolution.

4.6 Objective unconstitutionality

But Woolman maintains that the question cannot wait, owing to the Constitutional Court’s adherence to the doctrine of objective unconstitutionality.218 Woolman puts the matter thus: ‘[I]f one cannot imagine the court’s refusing [on some future occasion] to extend the ‘new’ rule to embrace coerced anal intercourse of men, then the new rule is objectively unconstitutional as of the moment it was announced.’219 From this, Woolman apparently deduces that the Court is obliged to decide now (Woolman thinks in the affirmative) whether extension of the new rule to the protection of men is constitutionally required. But how, exactly, would that follow?

Consider this possible response: Under the doctrine of objective unconstitutionality, a court is not free to postpone to some future case a decision on the constitutionality of any law whose application to a pending case would be decisive, even if the putative constitutional defect in that law pertains strictly to rights of persons other than the parties to the pending case. (‘The right to challenge the constitutionality of a statute which affects you directly cannot be made dependent on the finding of some other constitutional right on which to base the challenge. What if there is no such right?’)220 Since the Court could never justify convicting Masiya in terms of an objectively unconstitutional law, it is obliged to face the question of the new rule’s complete constitutionality as a part of its disposition of Masiya’s case.

That cannot be Woolman’s argument, for the simple and obvious reason that the Constitutional Court is not, as matters turn out, proposing to apply the new definition of rape to the pending prosecution of Masiya.221 The only law the Court proposes to apply to him is the old, constitutionally unassailable law defining the crime of indecent assault, and so there is no way that an on-the-spot assessment of the constitutionality of the new rule can affect the prosecution’s outcome. Insofar as the doctrine of objective unconstitutionality means to allow parties to pending cases to claim

218 See, eg, Home Affairs (n 87 above) para 29: On the objective theory of unconstitutionality adopted by this Court a litigant who has standing may properly rely on the objective unconstitutionality of a statute for the relief sought, even though the right unconstitutionally infringed is not that of the litigant in question but of some other person. 219 Woolman ‘Amazing’ (n 2 above) 167 n 7. 220 Ferreira v Levin NO 1996 1 SA 984 (CC), 1996 1 BCLR 1 (CC) para 163 (Chaskalson P) (‘Ferreira’). 221 See Masiya (n 4 above) paras 56-57 (deciding against retrospective application). 58 Uses of interpretive ‘charity’ the benefit of the unconstitutionality (on whatever grounds) of rules and doctrines being then and there invoked against them, it can have no purchase in the pending prosecution of Masiya (nor in any imaginable future prosecution of him, unless in respect of some utterly hypothetical, future criminal act).222

Thus, on what we might call a precise or strict reading of the doctrine of objective unconstitutionality, it cannot have the effect of requiring a decision of the putative section 9(3) claim against the new rule, as a part of the Constitutional Court’s disposition of the case of Masiya v DPP CCT 54/06. But is there, perhaps, some looser version of objective unconstitutionality that might have such an effect? I do not see what it would be. The doctrine of objective unconstitutionality cannot reasonably be taken to commit the Constitutional Court to on- the-spot adjudication of every conceivable claim of constitutional defect that might arguably afflict any of the rules and doctrines that in any way crop up in a case sub iudice. That way surely lies madness.

We can use a variation on NM to illustrate. Suppose the Court decides to take the case up for review, in the expectation that it cleanly presents a question about the consistency of the extant common law of privacy-invasion with Constitution section 14, insomuch as the extant law flatly excludes liability for non- intentional, but negligent, unauthorised disclosures of private information. Upon study of the record, the Court discovers that it contains a binding stipulation from all sides, to the effect that the book-writer was not negligent or in any way at fault for her action — so all that could remain for the Constitutional Court to do would be to affirm the decision below on that ground. It seems Woolman would be among the first to insist (and I would agree) that the Court ought now to dismiss the appeal for want of jurisdiction. But why should that follow, after all? Why should not the Court rather say: ‘There is still a question in the air about whether the extant common law of privacy- invasion is constitutional, insomuch as it flatly excludes liability for negligence, so we really owe it to the country to decide that now, per the doctrine of objective unconstitutionality’. Is that how we want the Court to be conducting its affairs?

222 Thus, this is not a case in which refusal to decide a constitutional question on the spot could possibly involve the Court in ‘recognis[ing] the validity of a [law] in respect of one litigant, only to deny it to another.’ Ferreira (n 220 above) para 26 (Ackermann J). Neither — owing to the non-retrospectivity ruling in Masiya — is it a case in which the postponement compromises ‘legal certainty’ to the disadvantage of any individual contemplating some or other course of action (as above). Of course, the objective-unconstitutionality argument would have bite in Masiya’s case, if the Court had decided in favor of retrospective application of the new rule. But I do not take Woolman to be suggesting that the Court’s ruling against retrospective application is itself symptomatic of a flight from substantive decisionmaking responsibility. (2008) 1 Constitutional Court Review 59

Nothing that I know of in the Constitutional Court’s oeuvre suggests that the Court itself has ever thought so. The Court has rejected a ‘narrow’ approach to standing in constitutional cases.223 Accordingly, the Court will address constitutional challenges by or on behalf of any litigant — whether or not his or her own constitutional rights are at stake — who can show that ‘he or she is directly affected by’ the law in question, and in that sense has an ‘interest in’ the challenge.224 The Court has furthermore indicated that it will be generous in deciding whether such an interest has been shown.225 The Court has also, however, expressly recognised the factors militating against allowance of constitutional challenges by persons claiming only a ‘hypothetical’ or ‘academic’ interest:

The principal reasons for this objection are that in an adversarial system decisions are best made when there is a genuine dispute in which each party has an interest to protect. There is moreover the need to conserve scarce judicial resources and to apply them to real and not hypothetical disputes. The United States courts also have regard to ‘the proper role of the Courts in a democratic society’ which is to settle concrete disputes, and to the need to prevent courts from being drawn into unnecessary conflict with coordinate branches of government.226

These would seem to be reasons why ‘few, if any, countries have at all times allowed all persons to invoke the jurisdiction of Courts to solve all legal problems.’227 South Africa does not draw the line nearly as strictly as some countries do (mine, for example),228 but a line there surely must be, and drawing it at the point where the litigant raising the challenge has something at stake in its resolution — that is, qua litigant, not just qua citizen — seems not a bad place to draw it. That, at any rate, is what I understand the doctrine of objective unconstitutionality to do. Woolman fears the doctrine may be falling into some degree of ‘desuetude’.229 I respond only that Masiya cannot stand as evidence for that proposition.

4.7 The court’s responsibility as lawmaker

Suppose you are persuaded that the doctrine of objective constitutionality, as currently propounded by the Constitutional Court, does not oblige the Court to address a section 9(3) claim

223 Ferreira (n 220 above) para 162 (Chaskalson P). 224 Ferreira (n 220 above) paras 165-66 (Chaskalson P). 225 ‘It is for this Court to decide what is a sufficient interest in the circumstances.’ Ferreira (n 220 above) para 168 (Chaskalson P, rejecting Ackermann J’s view that the Applicants lacked standing to raise a particular constitutional challenge). 226 Ferreira (n 220 above) para 164 (citation omitted). 227 Ferreira (n 220 above) para 31 (Ackermann J). 228 See, eg, n 154 above. 229 Woolman ‘Application’ (n 17 above) 50. 60 Uses of interpretive ‘charity’ against the new rule, as part of its disposition of Masiya’s case. That is certainly not yet to say that nothing so obliges it. What most likely would so oblige it, if anything does, is the simple fact that the Court is the author of the new rule, and we can rightly demand of the Constitutional Court that it — above all other agents in the South African constitutional order — take care, when it gets into the business of writing laws, to write only laws that are fully consistent with the Bill of Rights. I mean, just imagine that Cabinet, having decided to preempt completely the common law of sexually assaultive crimes with a new statutory code, tables before Parliament a draft that defines rape as the vaginal or anal penetration of females. It will most definitely be in order for objectors to demand that Parliament must now, before acting, decide whether, in its advised view, such a statute could pass constitutional muster under section 9(3). If that is so for Parliament, must it not be even more clearly so for the Constitutional Court?

Maybe not. To say so with perfect confidence is to lose sight momentarily of some differences between a court and a parliament in a constitutional democracy. When Parliament passes a bill, it, by necessary implication, clears that bill of constitutional defect so far as it can tell (no doubt subject to possible judicial correction). When the Constitutional Court declines to decide the constitutionality of a parliamentary enactment, in a case where nothing can possibly turn on that decision, it does not do that; it rather says ‘I am a court, I am moderately self-restrained by habit that you [meaning the country] would not really wish me to shake off, and the time for that is not yet ripe.’

‘But’, you will justly reply, ‘the new rule we are concerned with here is not a parliamentary enactment; it is the Court’s own, direct creation; that is just the point! There is no inter-branch dialogue in progress here, no consideration of deference to democratic accountability, no ball in Parliament’s court, to motivate or warrant postponement; there is only the Court refusing to confront itself!’ I am on shaky ground here, owing to deficient understanding of the exact interplay between common law and statute law in the South African law of crimes. (The very idea of common law crime poses a major challenge to the American-trained constitutional-legal temperament!) My impression, though, from the Masiya case itself, is that Parliament has indeed, though the Criminal Procedure Act and its various schedules, quite concretely bought into the judiciary’s historic definitions of the various crimes and, in a quite real sense, (2008) 1 Constitutional Court Review 61 made those definitions its own — including, specifically, the sex- differentiating aspect in the historic definition of the crime of rape.230 If that is correct, then the full of set of normal, inter-branch considerations affecting the degree to which the Court reaches out for constitutional questions is operative here.

It is, to be sure, a set of considerations that can be cashed only through exercises of judgment in particular cases. The cost of waiting always must be among the factors to be considered. That there is such a cost involved in the choice that concerns us here cannot be doubted. If (as Woolman believes) the Court must eventually decide to extend the rape definition to cover cases of male victims, the next rape-like assault on a man still won’t be prosecutable as rape. That is for sure, given the anti-retrospectivity ruling in Masiya; and that could make all the difference. But the odds don’t seem overwhelmingly that way, indecent assault is still a serious crime; and the Court’s public fidelity to its anti-imperialist habit is not in itself a negligible value. Had the Court grasped the nettle in Masiya that would not, in my view, have been over the line (given that there are common law crimes that the courts are responsible to define). But neither, in my view on present information, was the decision to wait an improper one.

How we judge that question may depend on how obviously and inevitably right we judge the section 9(3) claim against the new rule to be. Someone who sees the claim as irresistible is much likelier to fault the Constitutional Court for declining to decide it on the spot, than is someone, like me, who sees the claim as debatable under the precedents as they stand. In presenting the claim in that light, I have gone well beyond anything written in the Constitutional Court’s opinion; no reference to Hugo or to Harksen, for example, appears there. But I have offered nothing that is not entirely consistent with what the Court did write, and perhaps nothing that is not more or less clearly indicated by what the Court did write, at least to those conversant with the body of the Court’s other work. In other words, I have followed the course of interpretive charity. Whether with resulting net profit or loss to the project of the advancement of the benign rule of constitutional law in South Africa must be for the reader to judge.

230 See Masiya (n 4 above) paras 2, 7, 12.

THE CONSTITUTIONAL COURT, COURT WATCHERS AND THE COMMONS: A REPLY TO PROFESSOR MICHELMAN ON CONSTITUTIONAL DIALOGUE, ‘INTERPRETIVE CHARITY’ AND THE CITIZENRY AS SANGOMAS

Tshepo Madlingozi*

1Background

When I think back on the Constitutional Court’s 2007 term, one event stands out for me. On 20 August 2007, the Court was hearing Merafong Demarcation Forum v President of the Republic of South Africa.1 After a long period of struggle in which they had employed both institutional and extra-institutional mechanisms of democracy, the people of Khutsong decided to take their case to the Constitutional Court. Why? ‘The government does not want to listen.’2 On that day more than a 1 000 protesters gathered outside the court room toyi toying and singing liberation songs. During the course of the day, things turned nasty when protesters started burning tyres, brandishing dangerous weapons and allegedly pelting the police with stones.3 For me, this episode vividly demonstrates the fragile state of South Africa’s constitutional culture.

* I would like to thank the editors for inviting me to write this reply. In particular I would like to thank Professor Stu Woolman for regular conversation and extensive comments on the form and the substance of this reply. The paper also benefited from comments made by the participants at the Constitutional Court Review Conference (6-7 August 2008 Somerset West). Needless to say all errors are mine. 1 Merafong Demarcation Forum & Others v President of the Republic of South Africa & Others 2008 5 SA 171 (CC). 2 I Lekota ‘Flames in Khutsong’ Sowetan 25 May 2007 http://www.sowetan.co.za/ Columnists/IdoLekota/Article.aspx?id=473030. 3 See ‘Khutsong battle reaches Constitutional Court’ Mail & Guardian On-line 20 August 2007 http://www.mg.co.za/article/2007-09-20-khutsong-battle-reaches- constitutional-court (accessed 23 October 2008). For dramatic photos capturing these events, visit http://photos.mg.co.za/view_photo.php?pid=2036&gid=124 (accessed 23 October 2008).

63 64 Reply to Michelman

In their opposition to the incorporation of Merafong municipality into the North West province, the people of Khutsong had done everything conceivably possible within the law: They had participated in public consultation forums where they voiced their opposition against the proposal; they sent memoranda to state organs at all three levels of government; they had lobbied political parties and other stakeholders; they had engaged in peaceful marches and protests. Indeed before resorting to the litigation, the people of Khutsong had exhausted all institutional mechanisms available to them in a direct, participatory and representative democracy. In a phrase, they had done all that could be expected of an active citizenry in a constitutional democracy. Khutsong — ‘the place of peace’ — will long be remembered for the violence and mayhem that have gripped the township since December 2005. In scenes very much reminiscent of the 1980s township revolts against the apartheid regime, television cameras captured daily images of young people blockading the streets with burning tyres and stones, preparing ‘molotov cocktails’, torching shops, libraries, schools, government buildings and local councillors’ homes. Between December 2005 and April 2006, violent protests caused damage to private and public property estimated at R70 million.4 The community’s deep level of discontent is further reflected by the successful boycott of the municipal elections in March 2006. Of almost 30 000 registered voters, only 232 cast their ballots.5 From December 2005 until the end of 2007, the township was rendered ‘ungovernable’.

It took a very long time before the people of Khutsong decided to take their struggle to Court. It was not an easy decision. The leaders of the Khutsong struggle were convinced that their struggle could ‘be won in the street’:6

4 Centre for Development and Enterprise ‘Voices of Anger’ 10 CDE Focus (April 2007) 10. 5 As above. 6 N Kolisile ‘The crying of Khutsong’ Amandla Publishers 1 October 2008 http:// www.amandlapublishers.co.za/content/view/71/32/ (accessed 2 November 2008). Similarly, after the Court delivered its judgment, the Young Communist League noted in a press statement that even though it had not yet studied the Court’s decision, it was their ‘contention that matters related to forceful and illegitimate demarcations cannot be mitigated or resolved through the organs of class rule — the courts.’ See ‘YCL Statement on Khutsong’s ruling by the ConCourt’ 13 June 2008 http://groups.google.com/group/yclsa-press/browse_ thread/thread/67c0616a73a6f791 (accessed 20 June 2008). It is also important to note that — with the 2009 election firmly in their sight — the ruling party has recently decided that the Merafong Municipality will be incorporated back into Gauteng province. See ‘Khutsong to be returned to Gauteng’ The Mercury 22 October 2008 http://www.iol.co.za index.php?set_id=1&click_id=13&art_id= vn20081022054258319C323057 (accessed 2 November 2008). (2008) 1 Constitutional Court Review 65

It would be very dangerous to put our faith in legal processes. Courts are seldom neutral and often tend to serve the interests of the ruling class. That is the reason we have delayed taking this matter to the Constitutional Court. For more than a year we have relied on our own mass strength and we still do. In the coming period we will utilise the so- called democratic space provided by the new constitution as well as working class resistance. Key strategies include intensifying mobili- sation, building consciousness and continuing with mass struggle in the form of marches and rallies. This struggle will be won in the street!

Khutsong — during this period of struggle — is both a reminder of the fragility of South Africa’s new political order and an indictment of its ability to respond effectively to those persons and communities in the greatest need. As one of the leaders of the Khutsong anti-demarcation movement wrote: ‘The struggle of the people of Greater Merafong is not just about forceful incorporation into North West, but about fighting for truth from government, people’s rights and democracy.’7

Khutsong is not a singular, if continuous, event in a long but peaceful transition from a fascist apartheid state to a democratic constitutional order. Every week some poor community engages in illegal and often violent actions to back up demands for a more responsive government. Patrick Bond reports that ‘the pace of social protest intensified dramatically during the mid-2000s, soaring from 5,813 recorded protests in 2004-05 to more than 10,000/year’.8 Dale McKinley, one of the leaders of the Anti-Privatisation Forum, has argued that the increased violence employed by these community movements reflects the absence of ‘more inclusive and meaningful forms of direct and participatory democracy, that have little to do with the institutional forms of representation within bourgeois ‘democratic’ society.’9 McKinley claims convincingly that ‘these movements have arisen out of the very failures and betrayals of the “main currents” and the institutional framework that gives them contemporary legitimacy.’ 10

The biggest threat to the consolidation of South Africa’s constitutional culture is not the unprincipled and crass accusations

7 Kolisile (note 6 above). 8 P Bond ‘Rejoinder: Collaborations, co-optation & contestation in praxis-based knowledge production’ (2008) 35 Review of African Political Economy 271 272. 9 D McKinley ‘Democracy and social movements in SA’ (2004) 28 Labour Bulletin 39 40. 10 McKinley (note 9 above) 40. For more in-depth analyses and case studies of these post-apartheid social movements and community organisations, see R Ballard et al (eds) Voices of protest: Social Movements in Post-Apartheid South Africa (2006); N Gibson (ed) Challenging hegemony: Social movements and the quest for a new humanism in South Africa (2006); D McKinley & P Naidoo (eds) Development Update 5 (2004); and T Madlingozi ‘Post-Apartheid social movements and the quest for the elusive “new” South Africa’ (2007) 34 Journal of Law and Society 77. 66 Reply to Michelman lodged by political elites within the African National Congress against South Africa’s ‘independent’ judiciary. The time bomb ticking away inside our polity is the disillusionment of the overwhelming majority of South Africans with the politics and the policies of our post- apartheid democracy. Constitutional culture thrives when mobilised groups feel that they can appeal to constitutional norms and procedures in order to have their dispute resolved. It withers when such appeals go unheard. To be sure, the state’s response to these community movements has been at best to marginalise and vilify these movements by labelling them as ‘ultra-leftists’, ‘counter- revolutionaries’ organised by white ‘racists’, ‘enemies of the new democracy’;11 and at worst to criminalise and to suppress their struggles.12

If one of the objectives of the judiciary in a constitutional democracy is to facilitate ‘constitutional dialogue’, has the Court done enough to facilitate a dialogue between itself and the citizenry? The answer is no. More pointedly, I would caution against the judiciary’s and the academy’s acceptance of Professor Michelman’s proposal that ‘the principle of interpretive charity’ will have a meaningful effect on closing the gap between the Constitutional Court and the people.

Professor Michelman is indeed a giant.13 His breadth of scholarship and his ongoing commitment to South Africa have made him one

11 See, as an example, the response by ANC strategist Michael Sachs to the rise of these movements: M Sachs ‘“We don’t want the fucking vote”: Social movements and demagogues in South Africa’s young democracy’ (2004) 28 Labour Bulletin 23- 27. 12 For extensive reports on the criminalisation of poor people’s struggle as well as its violent repression, see D McKinley & A Veriava Arresting dissent: State repression and post-Apartheid social movements (2005) and M Memeza ‘A critical review of the implementation of the Regulation of Gatherings Act 205 of 1993 — A local government and civil society perspective’ (2006) available at www.fxi.org.za (accessed 14 June 2007). 13 It is a singular honour for me to be afforded this opportunity to engage with Professor Frank Michelman’s excellent foreword to this first issue of the Constitutional Court Review (FI Michelman ‘On the uses of interpretive ‘charity’: Some notes on application, avoidance, equality, and objective unconstitutionality from the 2007 term of the Constitutional Court of South Africa’ (2008) 1 Constitutional Court Review 1 (Michelman ‘Charity’)). Of those non-South African scholars who engage with the work of the South African Constitutional Court and South African Constitutional law theory in general, Professor Michelman is undoubtedly one of the most dedicated and influential. Indeed, as Professors Botha, Andre Van der Walt and Johan Van Der Walt have noted in a preface to a South African book dedicated to the scholarship of Prof Michelman, ‘Frank’s ideas have been — and continue to be — a source of inspiration for many of us who are interested in the possibility — and limits — of ‘transformative constitutionalism’. His writings on the judicial function and the capacity of rights discourse to facilitate and deepen democratic dialogue are among the most original and thoughtful on the topic — as are his reflections on constitutionalism as a tool for protecting a plurality of cultures, thoughtways and lifestyles; redressing inequality; and promoting social justice. It is therefore hardly surprising that South African legal academics and judges would have turned to his ideas in an (2008) 1 Constitutional Court Review 67 of our most important allies in re-thinking the meaning of a truly post- apartheid constitutional democracy. It is therefore most fitting that he should write one of the forewords to the first issue of the Constitutional Court Review. My mandate here, however, is not to pay homage to Professor Michelman. It is, rather, to somehow engage with his response to the equally impressive commentary on the Court’s jurisprudence by Professor Woolman.14 The two interlocutors need to be read together.

Professor Michelman’s piece is a rejoinder to Professor Woolman’s critical analysis of what he describes as the Constitutional Court’s ‘penchant for outcome-based decision-making and a concomitant lack of analytical rigour’.15 This absence of analytical rigour — and its deleterious consequences — are exemplified by, though not limited to, the Court’s tendency, as Woolman sees it, to ‘refuse’ to engage in the direct application of the Bill of Rights: ‘The persistent refusal to give rights identifiable content, by avoiding direct application results in a Bill of Rights increasingly denuded of meaning.’16 Woolman further contends that this refusal amounts to a ‘paradigmatic violation of the rule of law’. This reticence makes it difficult, if not impossible

for lower court judges, lawyers, government officials and citizens to discern, with some degree of certainty, how basic law is going to be applied, and to know, with some degree of certainty, that the basic law is going to be applied equally ...17

Michelman disagrees with Woolman’s conclusion that the thinly reasoned judgments in NM18 and Masiya19 show that the Court is in ‘full flight from any meaningful engagement with Chapter 2 of the Constitution’20 and thus an abdication of the Court’s institutional

13 attempt to break with the formalism and authoritarianism characterising our legal past, and to make sense of the democratic and egalitarian aspirations of the Constitution.’ H Botha et al (eds) Rights and democracy in a transformative constitution (2003) vii. In a paper presented in 2003 at a symposium dedicated to the scholarship of Professor Frank Michelman — Professor Levinson had this to say: ‘There are many reasons to honour Frank Michelman, not least because there are so many different topics — central to the enterprise of thinking about law — about which he has been remarkably illuminating. He is truly a giant on whose shoulders many of us are proud to stand as we try to think our way through the dilemmas presented by trying to take the enterprise of American constitutionalism seriously’. S Levinson ‘Perpetual union, “free love”, and secession: On the limits to the “consent of the governed”’ (2003) 39 Tulsa Law Review 457 457. 14 S Woolman ‘The amazing, vanishing Bill of Rights’ (2007) 124 South African Law Journal 762 (Woolman ‘Amazing’); S Woolman ‘Application’ in S Woolman et al (eds) Constitutional law of South Africa (2nd Edition, OS, 2005) ch 31. 15 Woolman ‘Amazing’ (n 14 above) 762. 16 Woolman ‘Amazing’ (n 14 above) 763. 17 As above. 18 NM v Smith 2007 5 SA 250 (CC) (NM). 19 Masiya v Director of Public Prosecutions 2007 5 SA 30 (CC) (Masiya). 20 Woolman ‘Amazing’ (n 14 above) 783. 68 Reply to Michelman responsibility. Michelman agrees entirely with Woolman’s contention that the majority opinions in NM and Masiya are ‘thinly reasoned’. He demurs from the thesis that these cases ‘have been wrongly or irresponsibly managed, as measured by reasonably discoverable, valid considerations of law and legal administration.’21

The move Michelman then makes may be one of the most powerful rhetorical ploys in analytic philosophy: the use of the Donald Davidson’s principle of interpretive charity. That principle in short requires that if we are genuinely interested in understanding another person or culture, then we must count them ‘right’ in most of what they say and do. To refuse to do so would consign them to the madhouse. Michelman describes the principle of interpretive charity as ‘an approach to understanding a speaker’s statements by interpreting the ... statements to be rational and, in the case of any argument rendering the best, strongest possible interpretation of an argument.’22 The aim of this principle, Professor Michelman tells us23

is to learn. It is aggressively to learn what there is to learn from puzzles the interlocutors pose to us, by assuming there is method in their madness and doing our best to ferret that out, using everything else we know or can guess ... about where they are coming from.

Michelman deploys the technique of ‘interpretive charity’ in order to fill in the gaps in the majority’s opinions in NM and Masiya. After an extensive and careful examination of the majority opinions in NM and Masiya, through a charitable lens, Professor Michelman concludes that the fact that the Constitutional Court seems to be gravitating towards its inherent power to develop common law in terms of section 173 and 39(2) of the Constitution — as opposed to its judicial review power in terms of section 8 and 172(1) when undertaking a modification of common law rules under pressure from the Bill of Rights — does not mean that the court is uncomfortable with engaging with the substance of the Bill of Rights.

Woolman sees the refusal to undertake the direct application of specific substantive provisions of the Bill of Rights as evidence of a court that does not respect its own precedents; as evidence of a court unaccountable to the community of constitutional interpreters and incapable of telling its citizens how to conform their behaviour to the dictates of the basic law; and as evidence that the Court — at the end of it all — is responsible for ‘the amazing, vanishing of the Bill of

21 Michelman ‘Charity’ (n 13 above) 2. 22 Wikipedia http://en.wikipedia.org/wiki/Principle_of_charity as cited by Michelman ‘Charity’ (note 13 above) 4. 23 Michelman ‘Charity’ (n 13 above) 4. (2008) 1 Constitutional Court Review 69

Rights’.24 Michelman sees the choice to go the route of indirect application made possible by section 39(2) — rather than direct application under section 8 — as no more than a matter of ‘doctrinal housekeeping’;25 a ‘filing system’-question26 over which nothing of substance turns. Woolman maintains that form is substance: Direct application ultimately leads to a proper elucidation of South African constitutional law. Michelman remains unmoved. Woolman’s construction — magisterial as it is — proceeds, as Michelman notes, from an ‘unrealistic high expectation of formal tidiness and tightness in constitutional drafting.’27 In Michelman’s opinion, the application conundrum — as brilliantly set out by Woolman — can not be resolved. Both authors agree that some surplusage is inevitable.

2 Engendering ‘constitutional dialogue’ with the Commons

I am not a constitutional lawyer. The main direction of my arguments can, without apology, be said to be political. However, what I will focus on still remains a bone of contention between Woolman and Michelman: the Court’s accountability and responsibility to ensure principled constitutional dialogue and its alleged failure to do so. I am interested in seeing who, beyond Woolman’s ‘chattering classes’,28 influences the Constitutional Court. I am interested in asking whether the Court truly engages in a dialogue with ordinary persons. I would argue — consistent with the positions staked out by Woolman — that one of the ways that the Court can make itself accountable to the public, and thus facilitate constitutional dialogue, is when its judgments are sufficiently reasoned, fully justified and widely published. ‘Thinly reasoned’ judgments that require the reader to enter into a highly sophisticated and technical reconstruction of a judgment — some would say a sangoma-like exercise — do not advance the cause of ensuring that ordinary people are part of the ‘principled’ — or at least transparent — constitutional dialogue to which the Court is ostensibly committed. I must therefore conclude that the principle of ‘interpretive charity’ has limited value in bridging the gap between the Court and the vast majority of its constituents.

24 Woolman argues that the Court’s failure to get application jurisprudence right and that it ‘continues to err in such quite obvious ways reflects, at a minimum three possibilities: (1) the court simply ignores academic interventions; (2) the court believes the academic criticism to be incorrect; and (3) the court ignores obvious doctrinal errors because it holds itself largely unaccountable to the existing community of academic and professional interpreters’ (Woolman ‘Amazing’ (n 14 above) 783). 25 Michelman ‘Charity’ (n 13 above) 3. 26 As above. 27 Michelman ‘Charity’ (n 13 above) 39. 28 Woolman ‘Amazing’ (n 14 above) 762. 70 Reply to Michelman

How did I arrive at this conclusion regarding the limits of the principle of interpretive charity in contemporary South Africa? Again, not on the basis of a nuanced assessment of the highly technical debate between Michelman and Woolman. I have read the exchange between Woolman and Michelman to ultimately be about how the Constitutional Court should conduct itself in relation to various stakeholders. Woolman argues, convincingly, that ‘a principled judicial dialogue’ that the Court claims to stimulate,29 is not possible if the Court, in his opinion, ‘refuses to say more than is necessary to decide a case on its facts.’30 Michelman demands that we try harder, and that if we take the principle of ‘interpretive charity’ seriously then we might just be able also to enter into a ‘principled dialogue’ with the Court.

Woolman does not deny the gravitational pull of Michelman’s charitable approach. Instead, he contends ‘that a court that cannot be held accountable to its own precedent, to the text of the Constitution, or to the community of constitutional interpreters, and that grounds its decisions by mere genuflection in the direction of s 39(2), more closely approximates a “kryptocracy” than a “logocracy”.’31 Whether we agree with this characterisation or not, all of us should be in agreement that trenchant and sustained critique of the Court’s jurisprudence from academics should be encouraged. What should we read into the Court’s sense of imperviousness to being influenced or at least being accountable to a community of ‘constitutional interpreters’? If we agree that the court fails to create the conditions for a ‘principled dialogue’ between itself and what Professor Woolman refers to as ‘a cohort of educated Constitutional Court watchers’;32 through its ‘murky, if not tendentious, lines of reasoning’ then what hope is there for the Commons, the masses of the people that are directly affected by its decisions?33

Writing in 1992, Donald Nicolson noted that although the notion of judicial independence is important in a democratic society, ‘democracy is also threatened by an unaccountable judiciary.’34 I do not wish to enter into debates about the nature and the form that

29 As called for by the Court in S v Mhlungu 1995 3 SA 867 (CC) para 129. 30 Woolman ‘Amazing’ (n 14 above) 785. 31 Woolman ‘Amazing’ (n 14 above) 784. 32 Woolman ‘Amazing’ (n 14 above) 762. 33 Woolman is at pains to point out, however, that he is not simply — or even primarily — interested in what he rather glibly calls the chattering classes. Indeed, he argues in ‘Amazing’ (n 14 above) and elsewhere that a constitutional democracy cannot function as a constitutional democracy when no more than a few hundred people can read and decipher the meaning of the Court’s decisions. See also S Woolman & M Bishop ‘Law’s Autonomy’ in S Woolman & M Bishop (eds) Constitutional conversations (2008) 1. 34 D Nicolson ‘Ideology and the South African judicial process — Lessons from the past’ (1992) 8 South African Journal on Human Rights 50 70. (2008) 1 Constitutional Court Review 71 judicial accountability should take. Nor do I desire to revisit the arguments around how various techniques of judicial accountability can ensure a way out of the ‘counter-majoritarian’ dilemma. I am solely interested in how ‘public accountability’ can contribute to the process of constitutional dialogue between the Court and the Commons. It is also important to note that I use ‘accountability’ in the very narrow sense proposed by Louis Seidman:35

Sometimes we say that persons are accountable when they are required to give an account of themselves — that is, to give reasons or justifications for conduct and to demonstrate that such conduct is not the product of mere whim or caprice.

Seidman argues that opinion writing exposes the judge’s thought process to those who read it, permits criticism of the justifications that the court offers and thereby allows judges to be held accountable.

To state that the ‘culture of justification’36 that flows from the doctrine of the rule of law is just as applicable to the judiciary as it is to the executive and to the legislature, is to state the obvious. In cases where the Court’s decisions run counter to public opinion — assuming that such opinion can be accurately determined — it is vitally important that the Court fully justifies its decisions if it is not to lose public legitimacy.37 Chief Justice Pius Langa has rightly observed that

[T]he relationship which [the judiciary] has with rest of the community is ... important. It should be regarded as an integral part of the community it serves, and it can only function properly if it enjoys the complete trust and confidence of that community.38

Judging from various public statements and op-ed pieces written by members of the judiciary, it would seem that recent public criticism of the judiciary has caught members of the bench by surprise. The judges appear to have assumed that their public legitimacy was secure. However, one simply needs to look at the results of various empirical surveys to see that, for the majority of this country, the

35 LM Seidman ‘Ambivalence and accountability’ (1988) 61 Southern California Law Review 1571 1574. 36 E Mureinik’s term. See E Mureinik ‘A bridge to where? Introducing the interim Bill of Rights’ 1994 (10) South African Journal on Human Rights 31. 37 See T Roux ‘Principle and pragmatism on the Constitutional Court of South Africa’ (2009) 7 International Journal of Constitutional Law 106. See also E van Huyssteen ‘The South African Constitutional Court and the death penalty: Whose values’ (1996) 24 International Journal of Sociology of Law 291. 38 See T Mbeki ANC Today 5(23) http://www.anc.org.za/ancdocs/anctoday/2005/ text/at23.txt (accessed 22 October 2008). 72 Reply to Michelman

Constitutional Court does not enjoy high public support.39 Indeed, recent comments by Anthony Butler are spot on: ‘The idea that the courts are suddenly under siege is a fantasy. Their legitimacy has long been fragile ...’.40

Amongst community movements, activists and ordinary people — as opposed to educated Constitutional Court watchers and members of the chattering class — the judiciary has not been able to shake off the impression that it is an instrument of the elites.41 Take this exchange between the Treatment Action Campaign’s Mark Heywood and another activist. Mark Heywood — an unabashedly activist Constitutional Court watcher — wrote42

The Constitution is at the heart of our democracy. It prescribes accountable and ethical government, openness and a culture of justification by politicians for their actions. But more than this, it should be the touchstone for all public policy, an instrument that can be used to insist on public policy that places equality and social improvement at the heart of all government action.

Now, to a ‘sober-minded’ and indeed ‘educated’ Constitutional Court watcher, this plea might seem quite reasonable. But that was not how it was read by one Commons activist:43

This near spiritual, nay, pathological worship of the handwork of neo- liberals is how the masses and their struggle for full independence, self- reliance and liberation got substituted for ‘the constitution’! All political power and authority somehow got squeezed into the Constitution, and what all need do is learn it, and enforce its provisions, and instantly, all will be solved. What is concealed is that the real social and economic power in this marginal act of giving themselves a liberal constitution automatically moved, and conformed, into the capitalist class. And true, to the extent reasonably possible, the people’s real problems will be attended to! What is also never frequently spoken about are ... the massive racial fears of the white supremacists, who, upon being born again as capitalist class rights fighters through a liberal constitution, abandoned their ‘minority racial rights’, and fought to have the constitution invested with ‘supreme power’ over all else.

39 On such public opinion surveys, see Roux (n 37 above) 107 n 4. (Roux reports that 27.9 per cent ‘attentive’ public support for the CCSA in 1997, ie among citizens who had heard about the Court, and ‘34 per cent public support for the CCSA in 2004’). 40 A Butler ‘Emotional responses overlook good points Mantashe makes’ Business Day 14 July 2008. 41 See T Madlingozi ‘The limits of transformative constitutionalism in the struggle against poverty and marginalisation’ (2009, forthcoming) copy on file with author. 42 M Heywood ‘Political climate dangerous for poor most of all’ Business Day 28 July 2008. 43 A Banda Debate Listserve (28 July 2008) [email protected] (accessed 28 July 2008). (2008) 1 Constitutional Court Review 73

The point here is that criticism of the judiciary does not only come from those who engage in scurrilous attacks on the judiciary in order to advance their narrow political agenda. Criticism of the judiciary also comes from those who are genuinely concerned about the gap between constitutional promises and the lived, impoverished, reality of the majority of South Africans.

The common reaction by judges and legal analysts to criticism of certain court judgments is to argue that the people do not understand the law or that they have not acquainted themselves sufficiently with the facts of the case. The former President of the Supreme Court of Appeal, Justice CT Howie has argued that not enough trouble has been taken ‘to tell our public, and our media, something of what judicial independence is all about’ and that public criticism is often made in the absence of getting to know the ‘vital facts’.44 But as a member of the Commons, I must ask: Whose responsibility is it to make sure that the public not only knows the ‘vital facts’ but also understands the working of the courts? Further, and more significantly, what if the public tries to acquaint itself with the reasons for a particular decision, only to be confronted with a ‘thinly reasoned’ judgment? How reasonable is it to expect members of the public to adopt the principle of ‘interpretive charity’ and engage in the laborious exercise of filling in meanings and passages in the courts’ decisions as Professor Michelman urges us to do?

3Conclusion

During the Court’s 2007 term (and well into its 2008 docket), incidences referred to as ‘threats to the judiciary’ reached a fever pitch. In the weeks and months preceding the appearance of the then deputy president of the African National Congress (ANC) in court to face charges of corruption, robust criticism, emanating largely from members of the tripartite alliance — the ANC, the South African Communist Party and the Congress of South African Trade Unions — became common place. The slogan ‘We will kill for Zuma’ if the Court did not find in his favour became a disturbingly regular part of public discourse in some quarters.

More than ever, the Court needs to court public support in order to protect itself against escalating ‘attacks on the judiciary’ emanating from various political quarters. Theunis Roux has shown that in the past the Court has not needed public support in order to safeguard its institutional security because the ‘ANC political elite has shielded the Court from the political repercussions of its most

44 Justice CT Howie ‘Judicial independence’ (2003) 120 South African Law Journal 679 679 & 682 (emphasis added). 74 Reply to Michelman unpopular decision ...’.45 In light of recent events this protection can no longer be taken for granted. Indeed, these events dictate that the Court change its strategy and work to consolidate its public legitimacy. That is no small task. The Court has not yet engaged in any meaningful outreach programs.46

The Constitutional Court might be based near Hillbrow. It might be located in a non-intimidating, inviting building and it might be geographically accessible. The judges might be very friendly and well meaning people. But the members of the Court continue to operate and to move within the circles of ‘educated Constitutional Court watchers’. I have seen and enjoyed the presence of Constitutional Court judges at various academic conferences. But I am yet to see any of them at various workshops and rallies hosted by social movements. I have seen members of the Court give frank and engaging public lectures at various universities. But I am yet to see any of them address members of social movements.47 I have heard members of the Constitutional Court give interviews on English radio stations. I have never heard any of them offer an interview on an African language radio station.

An ‘educated Constitutional Court watcher’ might tell me that the Court does not need to undertake such a public role. After all, their communications are a matter of public record. But where are the Sesotho or TshiVenda translations of the Court’s judgments?48 And even if they did appear on the Court’s website, I fail to see how that would make any difference at all to the majority of South Africans who remain quaintly removed from the internet and other forms of new media.

As we launch this Review, we need to pause and think of ways that ordinary people, the Commons, can become part of the ‘principled dialogue’ to which the Court and academics such as Woolman and Michelman have committed themselves. Finally, as South Africa enters what has been aptly described as the ‘second transition’ — characterised by, amongst other things, robust engagement with

45 Roux (note 37 above) 138. 46 In its annual strategic/business plan, under a section entitled ‘Outreach programs’ the Director of the Constitutional Court only manages to mention initiatives such as the law researchers program and the setting up of the Court’s choir and soccer team. See ‘Annual Strategic/Business Plan 2008/2009’ http:// www.constitutionalcourt.org.za/site/Admin/ANNUAL%20STRATEGIC%20BUSINESS% 20PLAN%202008-2009.pdf (accessed 21 October 2008). 47 Justice Dennis Davis is a notable exception in this regard. Justice Davis regularly gives key note addresses to social movement conferences and other events. 48 During his interview for position of the Deputy Justice, Justice Dikgang Moseneke noted the practical and cost implications in handing down judgments in other official languages. See http://www.constitutionalcourt.org.za/site/judges/ transcripts/dikgangmoseneke1.html (accessed 1 November 2008). (2008) 1 Constitutional Court Review 75 institutions supporting democracy and the settling of major political disputes in courts49 — the Court’s opinions will more than ever before play a significant role in public discourse and democratic politics. It is, therefore, vitally important that the Court holds itself accountable, not only to ‘educated constitutional court watchers’, but also to the Commons — some 46 million ordinary South Africans. Fully reasoned and sufficiently justified opinions that do not require the public to engage in highly complex exercises in ‘interpretive charity’ would be a first step in that direction. Maybe, just maybe, the events of 20 August 2007 might not be repeated in the future.

49 See S Ndlangisa ‘Political ball is in our courts’ City Press 24 August 2008 24.

NORMATIVE PLURALISM AND ANARCHY: REFLECTIONS ON THE 2007 TERM

AJ van der Walt*

1 The decline of foundation and the allure of anarchy

Waren die klaren Bedeutungen aus dem Haus, tanzte die Sprache auf dem Tisch.1

As a schoolboy, Robin Davies2 was intrigued to hear that, according to the likes of DH Lawrence, if one really wanted something one ‘should have it, ought to have it, must do it or take it if you can’. However, Robin soon learnt that social norms often prevent you from taking what you want and they cause all kinds of trouble when you do. ‘You can have [me and the children], or you can have everyone else. Not both’, his wife Nancy told him, years later, laying down the law about his philandering.3 ‘You can’t do both’, his father ruled when, aged 15 or so, Robin wanted to visit a friend on an afternoon when his parents had already arranged something for that evening.

* B Jur et Art Honns (BA) LLB LLD (Potchefstroom) LLM (Witwatersrand), South African Research Chair in Property Law, Faculty of Law, Stellenbosch University (hosted by Stellenbosch University, funded by the Department of Science and Technology and administered by the National Research Foundation). Thanks to Gustav Muller, Elmien du Plessis, Mikhalien Kellermann and Janke Strydom for research assistance, and to Karl Klare, the participants in the CCR workshop and Geo Quinot, Philip Sutherland, Henk Botha, Irma Kroeze and Wessel le Roux for discussions and comments, to Philip Sutherland for access to a draft of ‘Insuring contractual fairness in consumer contracts’ (Stellenbosch Law Review forthcoming), and to Halton Cheadle for an instructive exchange on application in the context of labour disputes. Remaining shortcomings are my own. My apologies to the reader for indulging a 30-year obsession. 1 T Dorn ‘Ultima ratio’ in D Schwanitz et al Amoklauf im Audimax: Stories (1998) 7- 46 18. 2 In Kingsley Amis’s novel You can’t do both (2004). 3 Patrick Standish received a similar ultimatum in Difficulties with girls (1988), as did Amis in real life.

77 78 Normative pluralism and anarchy: Reflections on the 2007 term

Kingsley Amis, ‘the most clubbable of men’ and a womaniser of note,4 was frustrated by restrictions on his freedom to have things his way, but he was familiar with the power of bourgeois morality — ‘taking it if you can’ wrought havoc when he decided that he should have what one of his characters called ‘marriage, 1960s style’, openly taking his mistress on holiday.5 Like his characters, Amis found that he could have the wife and the children or the girlfriend; not both. Several Amis characters given to ‘take it if they can’ are forced to abandon some of what they wanted or lose everything, including their dignity.6 It is easy to scoff at silly ‘rules’ about not having two social engagements on the same day, but not all social norms are so easily dismissed, even if they appear arbitrary.

The framing discourse for these Amis novels (and my purpose in recalling them) is not about accepting or rejecting social norms — despite his scorn for conventional morality, Amis did not subscribe to moral anarchy7 (and in the 2007 term the Constitutional Court said little about morality).8 Characters in the novels who made a life out of taking it if they could were not very nice persons and, despite flashes of defiance, Amis was not justifying their behaviour.9 In fact, Amis was commenting on what could be described — not in his words — as the complexity of human relationships in a time when social fundamentalism was being replaced by pluralism10 and, consequently, relativism and uncertainty. World War II liberated men and women from their traditional sexual and gendered social roles and

4 On the autobiographical nature of You can’t do both see Z Leader The life of Kingsley Amis (2006) 3 7-9. Martin Amis confirmed that finding parallels between his father’s life and novels is not ‘making the elementary mistake of conflating the man with the work’, since ‘all writers know that the truth is in the fiction’: Experience (2000) 28. 5 Jenny, in Difficulties with girls (1988); the girlfriend was Elizabeth Jane Howard, Amis’s second wife. 6 ‘Better be a bastard than a fool’, Roger Micheldene told himself, without much conviction, towards the end of One fat Englishman (1963). 7 In the Memoirs (1991) Amis remembered that his father put his, Kingsley’s, shortcomings down to his complete lack of religion. ‘And’, he says, ‘I should not be truly his son if I had never felt that he had something there.’ Amis concluded the Memoirs with a poem, entitled ‘Instead of an Epilogue’ and dedicated ‘To H’. In the final verse Amis lists the admirable characteristics of ‘someone’ he met when twenty-four, then proceeds: ‘Well, that was much as women were meant to be, / I thought, and set about looking further. / How can we tell, with nothing to compare?’ 8 But see Schaik v S 2008 2 SA 208 (CC) par 76. 9 Leader (n 4 above) 468: ‘Roger [Micheldene], like Patrick [Standish], knows how horrible he is.’ 10 Like J Dewar ‘The normal chaos of family law’ (1998) 61 Modern Law Review 467- 485 I use the term to indicate a diversity of conflicting norms. See LL Mofokeng ‘The right to freedom of religion: An apparently misunderstood aspect of legal diversity in South Africa’ (2007) 11 Law, Democracy & Development 121-131; S Louw ‘Mysterium inequitatis: Truth, elections, autonomy in the Southern African politic’ (2006) 33 Politikon 221–238 (relying on C Lefort’s notion of ‘the political’ to argue that democracy relies on diversity and pluralism). See further M Koenig & P de Guchteneire (eds) Democracy and human rights in multicultural societies (2007). (2008) 1 Constitutional Court Review 79 from the remnants of Victorian and Christian morality, opening up room for experimentation and free choice,11 but liberation coincided with uncertainty and disruption.12 Freedom from tradition comes at a price: once we give up the certainty of transcendent truth, choices become contingent and dangerous. Abandoning the certainty of tradition created a normative vacuum and raised the spectre of moral anarchy and turpitude. As Amis framed it much later, reflecting on bad choices he had made: ‘how can we tell, with nothing to compare?’13 The problem with postwar pluralism was that, once the certainty of immutable rules is abandoned, there is nothing to compare. When tradition is traded in for pluralism, everything is open for questioning, which could create the impression that everything is up for grabs; that everything goes. In an effort to avoid anarchy, the temptation is to fall back on the formalism of empty rules.

Amis’s depiction of postwar English society illustrates the slippery slope in post-traumatic society from normative pluralism to anarchy or, more accurately, to a facile choice between anarchy and reactionary formalism. The modernist rejection of foundation and acknowledgement of normative pluralism was inevitable, but the resulting uncertainty proved harder than expected. A similar process took place in law: ever since Realist and Critical Legal Studies scholars destroyed the theoretical credibility of fundamentalism and pointed out the indeterminacy and contingency implied by pluralism, lawyers — who usually deny that normative choice plays any role in law — had to account for their normative choices. In the face of nihilism and anarchy a reversion to formalism seems like a small price to pay for clarity and certainty.

11 Social liberalisation after World War II can be traced to the aftermath of World War I, just like the Critical Legal Studies Movement continued a trend initiated by the Legal Realists. M Tushnet ‘Critical legal studies: An introduction to its origins and underpinnings’ (1986) 36 Journal of Legal Education 505-517 argues that the question whether Critical Legal Studies continued Legal Realism is uninteresting; I take continuation between the two world wars for granted. During the wars, women were forced to take men’s places in factories, liberating them from their home-maker role. Similarly, people fighting in the wars realised that the subservient social roles they were used to before the war (because of age, social status or race) no longer suited them. In both cases, liberation from tradition undermined the authority of social hierarchy. 12 The street protests of 1968 that shocked mainstream society from Prague to Paris and from Berkeley to Berlin originated in the same context: the civil rights movement, women’s lib, the hippie movement, anti-war protests and student uprisings during the 1960s were founded upon rejection of traditional authority and mainstream morality. 13 See n 7 above. See IJ Kroeze ‘When worlds collide: An essay on morality’ (2007) 22 SA Public Law 323-335, especially 331-332 333-335. 80 Normative pluralism and anarchy: Reflections on the 2007 term

John Dewar14 argues that normative pluralism, which he describes as a plurality of rules and values that are chaotic, incoherent and even antinomic in that the contradictions between them cannot be resolved in the long run, is normal in so far as contradictions at the level of rules are held at bay by practices not completely determined by text. In Dewar’s view, normative pluralism changes into anarchy when it combines with uncertainty about purposes; as soon as pluralism extends beyond the level of rules, normative anarchy ensues. Dewar identifies normative anarchy in situations where rights are mixed with discretion in a way that defies balancing and where regulatory legislation moves from rules to utility and back without suitable systemic adaptations.

In a way, the early Amis novels described just such a shift from rules to utility and back — like Robin Davies, Amis and thousands of others rebelled against the inflexible and arbitrary post-war dictatorship of surviving pre-war morality and craved the greater freedom and flexibility they experienced during the war. Of course, with flexibility comes the burden of discretion but, according to Dewar and Amis, abandoning foundation does not necessarily end in nihilism: normative anarchy can be avoided as long as contradictions are restricted by common purpose. In the less utilitarian language of Joseph William Singer, the Critical Legal Studies movement ‘brought nihilism to centre stage’ by denying the possibility of determinacy, objectivity and neutrality in legal reasoning, but the absence of foundation ‘does not condemn us to indifference or arbitrariness, nor make it ridiculous to ask, or impossible to answer, the question of what we should do or how we should live’.15 In Singer’s view legal theory, as a form of political activity, opens everything up for questioning without putting everything up for grabs — critical legal theory acknowledges moral and political goals that give direction to ‘what we must do and how we should live’.16

Of course, on one level Dewar and Singer are vulnerable to accusations of fundamentalism — how does shared purpose or political goal that saves pluralism from anarchy differ from Victorian or Christian (or legal Formalist) sources of foundation?17 How is any

14 n 10 above, 468-473. See further J Dewar ‘Reducing discretion in family law’ (1997) 11 Australian Journal of Family Law 309-326; S Parker ‘Rights and utility in Anglo-Australian family law’ (1992) 55 Modern Law Review 311-330; M Maclean Making law for families (2000). 15 ‘The player and the cards: Nihilism and legal theory’ (1984) 94 Yale Law Journal 1-70 8: asking ‘what shall we do and how shall we live?’ opens up dissent rather than consensus. See Tushnet (n 11 above) 509; A Hunt ‘The theory of Critical Legal Studies’ (1986) 6 Oxford Journal of Legal Studies 1-45. 16 Singer (n 15 above) 66-70 identifies cruelty, misery, hierarchy and loneliness as the greatest problems that legal theory faces. 17 On foundation in law see P Fitzpatrick Modernism and the grounds of law (2001) 1-7. (2008) 1 Constitutional Court Review 81 political purpose less arbitrary than Mr Davies senior’s decree that Robin cannot ‘do both’? Once the foundation of traditional rules is rejected, how do we justify the new rules we place in their stead to avoid anarchy?

Developing a satisfactory answer to this challenge, if it can be done at all, will take my analysis too far off course, but a starting point is to accept the argument of Johan van der Walt and Henk Botha that ‘the counter-majoritarian problem and the problem of significant social dissent [both sources of normative pluralism in constitutional discourse] are not technical hiccups’ but aporias that ‘open up the possibility of social deliberation that would exceed technical procedure’ and technical ethics to ‘give politics a chance’.18 If there were a purpose that could avoid both anarchy and fundamentalism in the face of normative pluralism, it could be nothing else than the possibility of open social deliberation and dissent that defies foundation and closure to give politics a chance.19 In the ‘community without community’ that Van der Walt and Botha have in mind, a society of diverse communities, cultures and conflicting normative values, without any justificatory or unifying common will, social deliberation and politics can be possible without anarchy.20

In the ‘community without community’, recognition of the sacrificial nature of decision reveals the essence of law as its very failure to resolve social conflict with reference to a common will or purpose. In other words, politics is not to be found in a foundational, unifying common will that overcomes or mediates plurality; instead, politics is possible only in the presence of unmediated and unresolved diversity, plurality and dissent. Consequently, a purely dialogic or republican politics21 cannot solve conflict or bring closure by

18 J van der Walt & H Botha ‘Democracy and rights in South Africa: Beyond a constitutional culture of justification’ (2000) 7 Constellations 341-362 350. See H Botha ‘Democracy and rights: Constitutional interpretation in a postrealist world’ (2000) 63 Tydskrif vir Hedendaagse Romeins Hollandse Reg 561-581 572-574 581; K van Marle ‘Lives of action, thinking and revolt — A feminist call for politics and becoming in post-apartheid South Africa’ (2004) 19 SA Public Law 605-628 621- 624; F Michelman ‘Property as a constitutional right’ (1981) 38 Washington & Lee Law Review 1097-1114 1110: ‘it is the possibility of partial resolutions that allows us to experience the contradiction [between democracy and rights] as generative tension rather than as dead end.’ 19 Van der Walt & Botha (n 18 above) 350. See Van Marle (n 18 above) 621, citing B Honig (ed) Feminist interpretations of Hannah Arendt (1995) 135. See further W le Roux ‘Bridges, clearings and labyrinths: The architectural framing of post- apartheid constitutionalism’ (2004) 19 SA Public Law 629-665. 20 n 18 above, 350-352. 21 See le Roux (n 19 above) 646-660 on constitutionalism as a dialogic ‘clearing’. 82 Normative pluralism and anarchy: Reflections on the 2007 term mediating dissent or pluralism;22 nor can politics be guaranteed by constitutionalism, rights or law.23 Robert Cover pointed out that an important function of law is to ‘kill’ or suppress plurality, dissent and meaning; not to open it up.24 However, to the extent that law is inevitable, law, rights and constitutionalism are justifiable in so far as it gives politics a chance. A major issue in this article is to ask whether the interpretive and adjudicative approach that the Constitutional Court has apparently elected to follow in deciding cases where the common law tradition and legislation have to be measured against constitutional directives is justifiable in the sense that it gives politics a chance, instead of just closing up gaps and entrenching certainty.

2 Pluralism in post-apartheid constitutional theory

Our Constitution does not sanction a state of normative anarchy which may arise where potentially conflicting principles are juxtaposed.25

South Africans in the post-1994 era understand the anxiety caused by abandoning normative fundamentalism. Having faced up to the wickedness of apartheid and the bankruptcy of its underlying values, the point of constitutional transformation was to replace the exclusivism of segregation with an inclusive democracy that recognises diverse communities, cultures and traditions as well as diverse social, cultural and religious values. The 1996 Constitution’s commitment to diversity is apparent from its emphasis on human dignity, equality, and freedom26 and its recognition of the values of different social, cultural and religious groups,27 but recognition of diversity as a positive value surpasses mere tolerance — social, cultural and religious pluralism (‘more-than-one’) creates and supports the very possibility of politics and democracy.

22 Van der Walt & Botha (n 18 above) 352 argue that judicial decisions do not solve the conflicts that arise from social complexity but suppress them at the cost of one party for the sake of a social goal. R Cover ‘Violence and the word’ (1986) 95 Yale Law Journal 1601-1629 1607: ‘the relationship between legal interpretation and the infliction of pain remains operative even in the most routine of legal acts’. 23 Van Marle (n 18 above) 625-628 warns against ‘constitutional optimism’ based on the pervasiveness of law, constitution and rights, referring to E Christodoulidis ‘Constitutional irresolution: Law and the framing of civil society’ (2003) 9 European Law Journal 401-432. Compare le Roux (n 19 above) 660-665. 24 R Cover ‘The Supreme Court 1982 term. Foreword: Nomos and narrative’ (1983) 97 Harvard Law Review 4-68 41-44. 25 Ngcobo J, Fuel Retailers Association of Southern Africa v Director-General: Environmental Management, Department of Agriculture, Conservation and Environment, Mpumalanga Province, 2007 6 SA 4 (CC) par 93. 26 Compare s 1 of the 1996 Constitution. 27 See ss 9(3) and 9(4), 15, 30, 31; compare ss 39(3), 211-212. (2008) 1 Constitutional Court Review 83

Since 1994 the courts had many opportunities to confirm the Constitution’s commitment to accommodation of diversity and pluralism.28 Such an opportunity again arose in 2007 in Pillay,29 the first equality case to be considered by the Constitutional Court in terms of the Promotion of Equality and Prevention of Unfair Discrimination Act 4 of 2000 (PEPUDA).30 The Court reiterated the confirmation, in the Act, of ‘the Constitution’s commitment to affirm diversity’31 and decided that reasonable accommodation in matters of religion means that

sometimes the community, whether it is the State, an employer or a school, must take positive measures and possibly incur additional hardship or expense in order to allow all people to participate and enjoy all their rights equally. It ensures that we do not relegate people to the margins of society because they do not or cannot conform to certain social norms.32

Accordingly, a school disciplinary code that inflexibly denied an exemption from normal dress code for pupils who wish to wear a nose stud as part of their religious or cultural practice discriminated against them unfairly, in conflict with the Act.

Pillay demonstrates that conflict about normative pluralism in the context of diversity requires more opportunities for politics and difference, not greater consensus. There is no reason why we should agree about the wearing of nose studs, headscarves, yarmulkes or dreadlocks or why one view about any of these cultural and religious symbols should be enforced on everybody, but open deliberation and dissent about the underlying social customs and values and about tolerance is good for the democratic order. However, difference is not tolerated when it contradicts central values of the Constitution or undermines democracy33 — normative anarchy is not permitted. What prevents pluralism from sliding into anarchy in the context of diversity

28 Mostly relating to race, gender, religion and sexual orientation: Amod v Multilateral Motor Vehicle Accidents Fund 1999 4 1319 (SCA) (Islamic marriage); S v Lawrence; S v Negal; S v Solberg 1997 4 SA 1176 (CC) (Christian public holidays); National Coalition for Gay and Lesbian Equality v Minister of Justice 1999 1 SA 6 (CC) (sexual orientation); Du Toit v Minister for Welfare and Population Development 2003 2 SA 198 (CC) (married status); Satchwell v President of Republic of South Africa 2002 6 SA 1 (CC) (sexual orientation); J v Director- General, Department of Home Affairs 2003 5 SA 621 (CC) (sexual orientation); Gory v Kolver NO 2007 4 SA 97 (CC) (sexual orientation). 29 MEC for Education: KwaZulu-Natal v Pillay 2008 1 SA 474 (CC). 30 n 29 above, par 39. See I Currie & J de Waal The bill of rights handbook 5 ed (2005) 267. 31 Pillay (n 29 above) par 65. 32 Pillay (n 29 above) par 73. 33 Courts refused to enforce diversity in Bhe v Magistrate, Khayelitsha 2005 1 SA 580 (CC) (customary law rules of succession); Prince v President, Cape Law Society & others 2001 2 SA 388 (CC) (smoking of cannabis as part of religious observance); Christian Education in South Africa v Minister of Education 1999 2 SA 83 (CC) (corporal punishment as part of Christian education). 84 Normative pluralism and anarchy: Reflections on the 2007 term as a positive value is exactly what gave Amis and formalist opponents of critical legal theory headaches, namely recognising that values are contingent and that everything is open for dissent and discussion, without implying that everything goes.

The new constitutional order brought about a second kind of normative pluralism, where contradiction results not from different values but from the project of constitutional transformation itself. In this arena, the issue for debate is the direction, scope and pace of constitutional reform and, particularly, the effect of the Constitution on existing law. The effect of the Constitution on existing law is a logical locus for deliberation (politics) about social and legal reform, but the notion that a significant part of existing law — the common law — can survive or must be insulated against some or all direct or indirect constitutional influence contradicts constitutional values and threatens constitutional politics. In a word, the preservative impulse in the application debate misrepresents the democratic principle that everything is open for debate and threatens to push the normative pluralism of transformative politics into the anarchy of fractional conflict.

An important premise of my argument is that the Constitution’s commitment to diversity as a positive value is premised upon the reduced authority of the common law.34 The Constitution is founded on a set of values and normative commitments related to democracy, human dignity, equality and freedom and on the replacement of an immoral and discredited social and political system by a new, constitutional and democratic system. The shift from exclusivity to inclusivity that it presupposes is possible only if the existing legal order can be both purified of its exclusivist and authoritarian legacy and adapted to accommodate the normative diversity of its reformist purposes. In a largely uncodified legal system that relies on judicial finding, interpretation and application of old authorities and case law, remnants of the old legal system (characterised by exclusivity and authoritarianism) will survive political change and the establishment of the new constitutional order, for the time being,

34 See s 39(3), read with ss 8(3), 39(3) and 173 of the 1996 Constitution, recognising the continued existence of the common law, to the extent that it is consistent with the Bill of Rights and subject to the power of the higher courts to develop it to promote the spirit, purport and objects of the Bill of Rights. Johan van der Walt has argued that (direct or indirect) horizontal application of the Constitution would promote plurality: JWG van der Walt ‘Perspectives on horizontal application: Du Plessis v De Klerk revisited’ (1997) 12 SA Public Law 1-31; ‘Progressive indirect horizontal application of the bill of rights: Towards a co- operative relation between common-law and constitutional jurisprudence’ (2001) 17 South African Journal on Human Rights 341-363; Van der Walt & Botha (n 18 above) 350-352. (2008) 1 Constitutional Court Review 85 dragging unwanted survivals from the past behind us.35 Every legislative, executive and judicial decision embodies a choice between upholding remnants of the old order and enforcing change brought about by the new; every decision to uphold the existing law implies a sacrifice of constitutional reform, even when it is indubitably correct or unavoidable. Even more importantly, every decision in favour of stability and certainty or vested and acquired rights inevitably comes at the price of suppressing plurality, dissent and change.

To complicate matters the role of the Constitution is ambiguous. On the one hand, any Constitution (with a Bill of Rights) is part of the legal machinery that ensures stability and the protection of rights, particularly against state interference. In this perspective the Constitution is an agent for the protection of established privilege and power. On the other hand, though, the South African Constitution is clearly also an agent for change, including large-scale social, economic and political change that can affect established privilege and power negatively. It could be said that the danger of anarchy is reduced because the Constitution embodies a direction-giving purpose: different value systems are recognised, but we are all working towards an open society built on democracy, social justice and fundamental rights to human dignity, equality and freedom. Even in the face of diversity and pluralism, the Constitution directs us towards certain choices and away from others. Frank Michelman36 argues that the constitutional supremacy principle in s 1 of the Constitution, as interpreted in Pharmaceutical Manufacturers,37 establishes

the value of legal-systemic harmony in the service of the vision of the good society staked out by the entire list of founding values set forth in FC s 1 and instinct in the rest of the Final Constitution. We deal here with the value of the unity of the legal system — meaning the system’s

35 In the lyrics of Roger Waters: ‘they flutter behind you, your possible pasts / some bright eyed and crazy, some frightened and lost / a warning for anyone still in command / of their possible future to take care’: Pink Floyd ‘Your possible pasts’, from The final cut, subtitled A requiem for the post war dream (1983 © Pink Floyd Music Publishers Ltd). 36 FI Michelman ‘The rule of law, legality and the supremacy of the Constitution’ in S Woolman et al (eds) Constitutional law of South Africa (2nd Edition, OS, 2005) 34-44. 37 Ex Parte President of the Republic of South Africa: In re Pharmaceutical Manufacturers Association of South Africa 2000 2 SA 674 (CC) par 44. 86 Normative pluralism and anarchy: Reflections on the 2007 term

normative unity or, as one might say more poetically, its visionary unity.38

The notion of such a normative unity in the goal of constitutional transformation is certainly attractive. For administrative justice, Michelman’s analysis of the rule of law, legality and constitutional supremacy provisions provides a starting point based on the normative unity of the Constitution. Similar arguments can be developed for equality and non-discrimination, where transformation also amounts to wholesale replacing of apartheid law with a fundamentally new set of rules, indicating normative unity in reform. However, some constitutional provisions look less like wholesale, unidirectional transformations and more like contradictory products of difficult and inchoate compromises. In the compromise provisions, Michelman’s ‘visionary unity’ of the Constitution is less obvious or more contested than in abolition and replacement provisions like s 9.

A striking example is the sections relating to the national flag and the national anthem. The national flag described in s 5 (and Schedule 1) is neither based on the apartheid flag, itself the patchwork product of South Africa’s colonial and postcolonial past, nor on the ANC flag, nor is it a patchwork amalgam of the two. The Constitution simply replaced the apartheid flag with a new flag, signalling the clean break style of transformation that illustrates Michelman’s direction-giving ‘visionary unity’ of the Constitution. However, the national anthem illustrates a different approach: under apartheid, Die Stem van Suid- Afrika (The Call of South Africa) was the official national anthem, although Nkosi Sikelel' iAfrika was favoured by political activists and refugees. Under the 1993 Constitution the Republic had two national anthems: Nkosi Sikelel' iAfrika and Die Stem van Suid-Afrika / The Call of South Africa;39 under the 1996 Constitution the single national anthem consists of one verse each from Nkosi Sikelel' iAfrika, Die

38 Michelman (n 36 above) 37, referring to Port Elizabeth Municipality v Various Occupiers 2005 1 SA 217 (CC) par 35 (emphasis added). Michelman (n 36 above) 38, meets the objection that the promise of a shared transformative purpose is another kind of fundamentalism (see nn 18-20 above and accompanying text) by describing legal-systemic unity as a contingent, relative value, arguing that contingency is brought about by the linkage of constitutional supremacy in s 1 with rule of law: the value is contingent on ‘the direction of the pull’. When the direction of the pull is towards rule of justice rather than rule of law, the two values signify the unity of the legal system in the service of transformation by, under, and according to law’. This concern with the ‘pull towards the rule of justice’ points in the same direction as the ‘giving politics a chance’ argument of Van der Walt & Botha (n 18 above and accompanying text). 39 Die Stem van Suid-Afrika (translated as The Call of South Africa) was written by Afrikaans author CJ Langenhoven in 1918; the music was composed by the Reverend ML de Villiers in 1921. The first stanza of Nkosi Sikelel' iAfrika was written in Xhosa and the music composed by Enoch Sontonga in 1897; seven stanzas in Xhoza were added by the poet Samuel Mqhayi. A Sesotho version by Moses Mphahlele was published in 1942. Information: http://www.info.gov.za/ aboutgovt/symbols/anthem.htm (accessed 8 May 2008). (2008) 1 Constitutional Court Review 87

Stem van Suid-Afrika and The Call of South Africa.40 Nothing illustrates the accommodation of post-apartheid pluralism by way of a makeshift, patchwork compromise between opposing value systems better.

However, patchwork compromises are not fatal. Like the national anthem, s 25 embodies a patchwork compromise between constitutional protection of existing property interests (s 25(1)-(3)) and constitutionally sanctioned land reform (s 25(4)-(9)).41 Although it is difficult to imagine how the land-reform goals in s 25(5)-(9) could be promoted without infringing upon existing property interests in conflict with s 25(1)-(3), Michelman’s argument suggests a solution: the ‘legal-systemic unity’ of ‘every site of law’ (including both parts of s 25) is to be found in the constitutional purpose of ‘pulling in the justice direction’, ‘in the service of transformation by, under, and according to law.’42 Reading s 25 in this way suggests that protection of existing property interests and promotion of equitable access to property are guaranteed in so far as those purposes conform to the constitutional purpose of transformation by, under, and according to law (including legislation, common law and customary law). What renders Michelman’s reading compelling is not that it makes the conflict between two seemingly contradictory provisions or normative purposes disappear, nor that it identifies an overarching purpose in terms of which they can be reconciled or balanced, but that it shows that what appears like a contradiction is in fact a tension,43 a locus of significant dissent, that makes social deliberation both inevitable and possible and that therefore ‘gives politics a chance’. The very possibility of politics is founded on the significant dissent embodied in the tension between the two parts of s 25. Constitutional application gives politics a chance in so far as it views this tension as part of the contingent ‘legal-systemic unity’ of ‘pulling in the justice direction of transformation by, under, and according to law’; not in order to deny or resolve the tension, but to explore it in the context of the equally contingent and contradictory justice goals of constitutional transformation itself. The constitutional purpose of transformation by, under and in accordance with law is not uncontested; it is a site for dissent and social deliberation; for politics. Again, in as far as law is inevitable in the process of social and political transformation, its

40 Ss 2, 248(1) of the 1993 Constitution; s 4, Schedule 1 of the 1996 Constitution. The 1997 proclamation was published in Government Gazette 18341. 41 AJ van der Walt Constitutional property law (2005) 12-18. 42 Michelman’s reading finds support in Port Elizabeth Municipality (n 38 above), specifically regarding s 25. 43 Van der Walt & Botha call this an aporia (n 18 above and accompanying text). Michelman explored something like the ‘legal-systemic unity’ (n 36 above) earlier (see n 18 above, 1112): ‘My suggestion is to seek a rapprochement of property and popular sovereignty in the idea that rights under a political constitution, including property rights, are first of all to be regarded as political rights.’ 88 Normative pluralism and anarchy: Reflections on the 2007 term legitimacy must depend on the extent to which it ‘gives politics a chance’ by acknowledging the sacrifice and upholding the dissent inherent in this tension.

My goal here is to consider the Constitutional Court’s response, during the 2007 term, to the compromise-based, patchwork relationship between existing law (particularly the common law) and the Constitution. In earlier constitutional discourse this has mostly been debated on the basis of the so-called application issue, referring to the question how the Constitution should find application in existing law. On the one hand, based on the constitutional aspiration to rectify the injustices and heal the divisions of the past and establish a society based on democratic values, social justice and fundamental human rights, the intention is clearly that the Constitution is the supreme law and that all law should be subject to it44 — in this respect, the Constitution embodies a strong imperative for change in the general direction indicated by the transformation goals set out in the Constitution, as explained by Michelman. On the other hand, the desire for stability and certainty inherent in the rule of law principle implies a constitutional obligation not to disturb settled bodies of existing law45 — and the vested rights protected by them — unnecessarily.46 As I have argued above, this desire for stability and certainty will have a stronger effect in areas where the relevant constitutional provisions do not reveal the ‘visionary unity’ of the Constitution — in the compromise cases, we can expect that tradition and stability will offer stronger resistance against change. What Karl Klare described as ‘professional legal sensibilities’ will also tend to favour the stability of tradition and settled law, preferring to restrict change and uncertainty to instances where it is clearly required or mandated by unambiguous constitutional or statutory prescriptions.47

That brings me to my central premise: tension between the push for constitutional reform and the pull of traditional stability constitutes more than just a minor difference about the pace or the direction of change. Underlying this tension is a larger conflict

44 Ss 1, 2, 7, 8(1) of the Constitution; with regard to ss 1, 2 compare Michelman (n 36 above) 34-44. 45 The phraseology is from S Woolman ‘Application’ in S Woolman et al (eds) Constitutional law of South Africa (2nd Edition, OS, 2005) 10. 46 S 1 (rule of law); s 39(3): the Bill of Rights does not deny the existence of other rights or freedoms recognised or conferred by common law, customary law or legislation, to the extent that they are consistent with the Bill. Compare ss 173, 8(3), 39(2) on the higher courts’ power to develop the common law to promote the spirit, purport and objects of the Bill. Compare T Roux ‘Continuity and change in a transforming legal order: The impact of section 26(3) of the Constitution on South African law’ (2004) 121 South African Law Journal 466-492 466 467. 47 K Klare ‘Legal culture and transformative constitutionalism’ (1998) 14 South African Journal on Human Rights 146-188 166-167. (2008) 1 Constitutional Court Review 89 between two opposing normative values or ideologies, one favouring the stability of existing law, based on the presence and the presumed political neutrality of the status quo; the other relying on the political justifiability of fundamental change indicated by the presumed social construction and injustice of the status quo. Those who think that the existing distribution of wealth, privilege and power and the law that helped create and uphold it are politically neutral and therefore not in need of political reconsideration favour a regime of rights in which the direct effect of the Constitution is preservation of the status quo; those who think that the existing regime and its laws are inherently politically constructed and hence in need of political reconsideration favour the view that neither law nor rights are immune from constitutional effect or democratic redefinition. In this perspective, the relationship between existing law and Constitution is an instance of the tension between two conflicting sets of normative values, respectively favouring stability and change, certainty and justice, rights and democracy, tradition and constitution. In the absence of a ‘visionary unity’ of purpose, conflict between them could precipitate a slide into normative anarchy. Simply ignoring or subjugating one in favour of the other could, however, undermine the diversity and pluralism that is a prerequisite for significant social dissent, politics and democracy. In the terminology of Van der Walt and Botha, the legitimacy of constitutional review depends upon the extent to which constitutional law can ‘give politics a chance’ by acknowledging the aporia inherent in the tension between stability and change, and by recognising the sacrificial inevitability of upholding, in a concrete case, the one at the cost of the other, without the comfort or the certainty of an appeal to foundation.

Michelman recognises the contingency of accepting diversity, but insists that constitutional transformation is possible to the extent that the direction of the pull between conflicting values favours justice rather than law. The constitutional tension between legality, rule of law and constitutional supremacy involves an aporia; so does the tension between stabilising the old and promoting the new, in the sense that rule of law safeguards stability via the back door of legal certainty while constitutional supremacy demands transformation. Consequently, Michelman’s notion of visionary unity under the Constitution applies very usefully to the problem of existing law vs reform, but we should be careful not to read his analysis as superficial pragmatism. The unifying vision he identifies does not indicate an inflexible preference for change and against stability: the visionary unity of the Constitution identifies justice as a normative guideline for making hard decisions in concrete cases, but it does not pre-empt decisions in favour of change and against stability. In the language of Van der Walt and Botha, Michelman’s argument recognises that every choice for or against stability or reform is a sacrificial one: the choice cannot be justified, but making an unavoidable decision is justified in 90 Normative pluralism and anarchy: Reflections on the 2007 term so far as it accommodates social dissent and open democratic deliberation. In every case, the question is whether the direction of the pull favours justice.

Relying on the theoretical groundwork of Van der Walt and Botha and Michelman, my hypothesis is that normative pluralism in the contradictions between constitutional reform and stability could easily result in either normative anarchy or reactionary formalism, unless the normative unity, the general direction of the pull, between these opposing forces favours transformative justice, by which I mean the justice that is implied by renunciation of everything the apartheid legal order represented and espousal of the central transformative values of human dignity, equality and freedom in an open and democratic society. This normative unity does not imply an inflexible preference for change and against stability, but it indicates the direction of the normative pull against which a decision for either change or stability must be justified. The questions that I will pose are whether the Constitutional Court is in the process of developing a direction-giving approach in deciding cases involving the contradiction between constitutional reform and stability and, if it is; whether this approach is justifiable in that it does not simply entrench a new formalism but, instead, opens up space for politics.

3 One legal system

There is only one system of law. It is shaped by the Constitution which is the supreme law, and all law, including the common law, derives its force from the Constitution and is subject to constitutional control.48

The tension between preservation of the common law tradition and constitutional transformation initially emerged in debates about the direct or indirect horizontal application of the Bill of Rights. Commentators and policy makers concerned that private-law enclaves should not be allowed to escape the reach of constitutional reform or to continue ‘privatised’ injustice proposed that constitutional rights should apply directly; others queried the wisdom and the legitimacy of ‘importing’ constitutional values into private law. Clothed as concern about the unnecessary abolition of a well- developed and supposedly politically neutral legal tradition, the latter often concealed anxiety about the continued existence of established privilege and wealth. Conflicting decisions in which the new constitutional order was sometimes welcomed, sometimes accepted with a measure of suspicion, sometimes rejected outright, demonstrated uncertainty about the normative scheme that informs post-1994 legal reasoning.

48 Chaskalson P, Pharmaceutical Manufacturers (n 37 above) par 44. (2008) 1 Constitutional Court Review 91

The constitutional text makes it clear that the Constitution is the supreme law of the land;49 that the Bill of Rights is the cornerstone of democracy and that it enshrines the values of human dignity, equality and freedom;50 that it applies to all law and that it binds all three arms of government and all organs of state;51 that the state must respect, protect, promote, and fulfil the rights in the Bill of Rights;52 that the Bill of Rights binds natural and juristic persons if and to the extent that it is applicable, taking into account the nature of the right;53 that the courts, in order to give effect to a right in the Bill of Rights, must apply or, where necessary, develop the common law to the extent that legislation does not give effect to that right and may develop rules of the common law to limit the right,54 and that courts, tribunals and other forums must, when interpreting legislation and when developing the common law and customary law, promote the spirit, purport and objects of the Bill of Rights.55 However, it was not clear how this should work out in practice — when should the Constitution or new legislation trump the common law; when should the common law be developed to promote the spirit, purport and objects of the Constitution; how much should be changed in the process? Even more importantly, what should the effect of established common law norms and values be when it is unclear whether the Constitution or new legislation trumps the common law?

Some authors and judges share Michelman’s view that transformation and stability can exist together and that an application theory and practice that uphold the tension is obligatory — the very idea of transformative constitutionalism56 is founded upon the notion of a constitutional order that guarantees stability while

49 S 2. 50 S 7(1). 51 S 8(1). 52 S 7(2). 53 S 8(2). 54 S 8(3). 55 S 39(2). 56 The concept, coined by Klare (n 47 above) 146-188, inspired a cottage industry in constitutional theory; see H Botha ‘Metaphoric reasoning and transformative constitutionalism’ 2002 Tydskrif vir Suid-Afrikaanse Reg 612-627; 2003 Tydskrif vir Suid-Afrikaanse Reg 20-36; D Moseneke ‘The fourth Bram Fischer memorial lecture: Transformative adjudication’ (2002) 18 South African Journal on Human Rights 309-319; H Botha ‘Freedom and constraint in constitutional adjudication’ (2004) 20 South African Journal on Human Rights 249-283; le Roux (n 19 above) 629-675; Roux (n 46 above); AJ van der Walt ‘Transformative constitutionalism and the development of South African property law’ 2005 Tydskrif vir Suid- Afrikaanse Reg 655-689; 2006 Tydskrif vir Suid-Afrikaanse Reg 1-31; M Pieterse ‘What do we mean when we talk about transformative constitutionalism?’ (2005) 20 SA Public Law 155-166; S Liebenberg ‘Needs, rights and transformation: Adjudicating social rights’ (2006) 1 Stellenbosch Law Review 5-36; P Langa ‘Transformative constitutionalism’ (2006) 17 Stellenbosch Law Review 351-360. See further the contributions of H Botha, D Davis, J Froneman, J van der Walt and K van Marle in H Botha, AJ van der Walt & JWG van der Walt (eds) Rights and democracy in a transformative constitution (2003). 92 Normative pluralism and anarchy: Reflections on the 2007 term simultaneously requiring (and authorising or even driving) social, economic or political transformation.57

The problem is that the debate about the place and role of the common law in the new dispensation58 has been dominated by the largely undeclared assumption that the common law should be insulated, in some measure, against constitutional amendment, presupposing that a substantial part of the common law might remain unaffected by the Constitution and that development of the rest can take place on its own terms, without raising constitutional issues. The jurisdictional format that this turf war adopted under the 1993 Constitution59 ended in Pharmaceutical Manufacturers when the Constitutional Court rejected the Supreme Court of Appeal’s view that the common law grounds for review continued to exist as a body of norms ‘separate and distinct from the Constitution’:60

There are not two systems of law, ... each operating in its own field with its own highest court. There is only one system of law. It is shaped by the Constitution which is the supreme law, and all law, including the common law, derives its force from the Constitution and is subject to constitutional control.61

57 Michelman and Van der Walt & Botha argue that the tension in the Constitution is central to the possibility of politics; see nn 18-20, 36-38 above. See further Roux (n 46 above) 487. 58 The problem of balancing constitutional supremacy and common law is not unique to South African law. A similar problem exists in German constitutional law and early commentaries on the South African Constitution found inspiration in German theory: D Davis, M Chaskalson & J de Waal ‘Democracy and constitutionalism: The role of constitutional interpretation’ in D van Wyk, J Dugard, B de Villiers & D Davis (eds) Rights and constitutionalism: The new South African legal order (1994) 1-130 89-97. In Carmichele v Minister of Safety and Security (Centre for Applied Legal Studies Intervening) 2001 4 SA 938 (CC) par 54 Ackermann & Goldstone JJ referred to German constitutional law. Michelman (n 36) 40 relies on this passage in Carmichele to show the Constitutional Court’s support for the view that a ‘comprehensive (if doubtless somewhat inchoate) vision of the good society’ is inherent in the Constitution. Since the European Convention was incorporated into English law by the Human Rights Act 1998, English courts have struggled to reconcile common law with human rights standards, eg eviction vs home interest in art 8 European Convention of Human Rights: Harrow London Borough Council v Qazi [2004] 1 AC 983; Kay v London Borough of Lambeth; Leeds City Council v Price [2006] UKHL 10 (HL); Connors v United Kingdom [2004] ECHR 223. 59 Michelman (n 36 above) 22. 60 The SCA decision was Commissioner for Customs and Excise v Container Logistics (Pty) Ltd; Commissioner for Customs and Excise v Rennies Group Ltd t/a Renfreight 1999 3 SA 771 (SCA); the CC decision was Pharmaceutical Manufacturers (n 37 above). Michelman (n 36 above) 15-33 explains Amod v Multilateral Motor Vehicle Accidents Fund 1999 4 1319 (SCA) as a model for judges who want to keep a judgment out of the constitutional arena to reserve for the SCA final control over a category of future cases (18), although Amod might simply have avoided ‘thorny issues’; its cause of action preceded commencement of the 1993 Constitution. 61 Pharmaceutical Manufacturers (n 37 above) par 44. See Michelman (n 36 above) 21-22. (2008) 1 Constitutional Court Review 93

In Pharmaceutical Manufacturers the Constitutional Court ‘established, once and for all, that the SCA cannot insulate a decision on legality from CC review by dressing it as a merely common law (and hence not a constitutional) decision.’62 As far as review of administrative action is concerned, this decision established a unitary vision of the relationship between common law and Constitution: the common law continues to exist in as far as it does not directly conflict with the Constitution. However, it does not exist as an independent body of law — there is just one system of law, shaped by the Constitution as the supreme law. The common law derives its force from the Constitution and it is subject to constitutional control and the tension between the Constitution and the common law is to be understood in terms of the transformative goal, the unitary normative vision, of the Constitution. The application, interpretation and development of the common law must take place in line with this vision. The essence, in the words of Michelman, is that application of the Constitution should promote the shared vision of an open society built on democracy, social justice and fundamental human rights and on human dignity, equality and freedom.

Unfortunately, Pharmaceutical Manufacturers did not bring clarity about the effect of the Constitution on existing law (including the common law) as a whole;63 outside of administrative law, the idea that the common law might escape constitutional influence and that development of the common law can develop on its own terms, without raising constitutional issues, retains some traction. Private law specialists tended simply to establish absence of direct conflict between common law principles and constitutional rights, instead of seeking for normative or visionary unity. This approach presupposes that the common law could coexist alongside the Constitution, more or less unchanged, as long as it is free from the pernicious influence of apartheid. The strong version of this argument, that the law in general is politically neutral and that the abolition of apartheid laws

62 Michelman (n 36 above) 18. This is what L du Plessis ‘“Subsidiarity”: What’s in the name for constitutional interpretation and adjudication?’ (2006) 17 Stellenbosch Law Review 207-231 211 refers to as jurisdictional subsidiarity. 63 Pharmaceutical Manufacturers (n 37 above) was decided prior to commencement of the Promotion of Administrative Justice Act 3 of 2000 (PAJA). Like Fedsure Life Assurance Ltd v Greater Johannesburg Transitional Metropolitan Council 1999 1 SA 374 (CC), Pharmaceutical Manufacturers dealt with the broader notion of legality, which includes executive and legislative action outside of the range of administrative action regulated by PAJA. With regard to the relationship between the common law, PAJA and the Constitution see Bato Star Fishing (Pty) Ltd v Minister of Environmental Affairs 2004 4 SA 490 (CC) par 25: the provisions of s 6 PAJA ‘divulge a clear purpose to codify the grounds of judicial review of administrative action as defined in PAJA. The cause of action for the judicial review of administrative action now ordinarily arises from PAJA, not from the common law as in the past. And the authority of PAJA to ground such causes of action rests squarely on the Constitution ... As PAJA gives effect to s 33 of the Constitution, matters relating to the interpretation and application of PAJA will of course be constitutional matters.’ 94 Normative pluralism and anarchy: Reflections on the 2007 term properly eradicated politics from law, thereby restoring the common law to its former glory, no longer enjoys theoretical credibility64 outside of doctrinally conservative circles.65 The weak version, that the common law is flexible and capable of adaptation that would bring it into line with new demands and requirements in the new constitutional democracy, is shared by many lawyers, including high- ranking judges in the new legal order.66

In the application debate, this argument supported the view that development of the common law, although referred to in the Constitution,67 is something the civil courts are familiar with and that they should be left to do the work as they have always done it.68 This view enjoyed wide support and strengthened the division of work between the Constitutional Court and the Supreme Court of Appeal under the 1993 Constitution.69 The High Courts, the Supreme Court of Appeal and the Constitutional Court now have concurrent jurisdiction to decide on the constitutionality of legislation, but the notion persists that the Supreme Court of Appeal is the specialist court on

64 Botha (n 18 above) 561-581. 65 See J Potgieter ‘The role of the law in a period of political transition: The need for objectivity’ (1991) 54 Tydskrif vir Hedendaagse Romeins-Hollandse Reg 800- 807 802: ‘It must be stressed that the basic assumption that the South African legal system as a whole has become illegitimate, is unfounded. The crisis in South Africa lies primarily in the socio-political rather than the legal sphere’; J Neethling ‘’n Toekomsblik op die Suid-Afrikaanse privaatreg — volwaardige naasbestaan of versoenende sintese?’ (‘A future perspective on South African private law — full coexistence or reconciliatory synthesis?) in A van Aswegen (ed) The future of South African private law (1994) 1-9 at 3: ‘In the first instance, the apartheid era cannot be attributed to Roman-Dutch law — the blame must be placed squarely on the shoulders of the ruling minority who introduced the system of apartheid by way of legislation, and they were primarily enabled to this end by the doctrine of parliamentary sovereignty which derives from English constitutional law.’ See further TJ Scott ‘The future of our Roman-Dutch law: Reflections and a suggestion’ (1993) 26 De Jure 394-400 399. 66 MM Corbett ‘Trust law in the 90s: Challenges and change’ (1993) 56 Tydskrif vir Hedendaagse Romeins-Hollandse Reg 262-270 264. See also I Mahomed ‘The future of Roman-Dutch law in Southern Africa, particularly in Lesotho’ 1985 Lesotho Law Journal 357-365 360; A Sachs ‘The future of Roman-Dutch law’ (chap 8) in A Sachs Protecting human rights in a new South Africa (1990) 90-103 (‘we know it more or less, we have the books and the rules and the procedures available, we might as well use it’). See AJ van der Walt ‘Tradition on trial: A critical analysis of the civil-law tradition in South African property law’ (1995) 11 South Africa Journal on Human Rights 169-206 169-171. 67 Particularly ss 8(3), 39(2) and 173. 68 See Brisley v Drotsky 2002 4 SA 1 (SCA); Afrox Health Care Bpk v Strydom 2002 6 SA 21 (SCA). A Cockrell ‘Private law and the bill of rights: A threshold issue of “horizontality”’ in Bill of rights compendium (1998) paras 3A8, 3A7 argues that ‘the direct application of constitutional rights against private agencies must be mediated by the operation of the common law’. 69 In the 1996 Constitution, as in the 1993 Constitution, the Constitutional Court is a specialist court and not a court of general jurisdiction. In terms of s 167 it is the court of final instance in constitutional matters only. The Constitutional Court must confirm any order of a High Court or the Supreme Court of Appeal to the effect that an Act of Parliament, a provincial Act or conduct of the President is unconstitutional before that order can have effect: s 167(5). See Currie & de Waal (n 30 above) 108-109. (2008) 1 Constitutional Court Review 95 common law matters.70 In Zantsi71 the Constitutional Court stated that, because of the far-reaching implications of constitutional decisions, it would be better if lower courts first heard cases where constitutional issues are raised, because that would allow the law to develop incrementally.72 Incremental development is said to be in line with the inherent logic of the common law and Zantsi therefore reinforced the idea that development of the common law under the Constitution could continue largely as before.73

The notion that the Supreme Court of Appeal is a specialist court in common law matters is not the only reason why cases involving development of the common law are left to the civil courts. The Constitutional Court prefers not to sit as court of first and last instance on any matter and it prefers that the High Court or the Supreme Court of Appeal decide on common law issues before it considers the matter on appeal.74 The outcome in Carmichele75 could be seen as proof that the division of labour between the

70 Amod v Multilateral Motor Vehicle Accidents Fund 1998 4 SA 753 (CC) par 33: the SCA’s view on whether the common law should be developed is important because of ‘the breadth of its jurisdiction and its expertise in the common law’. 71 Zantsi v Council of State, Ciskei 1995 4 SA 615 (CC) par 5. This echoed the decision in S v Mhlungu 1995 3 SA 867 (CC) par 59 that, where it is possible to decide any case without reaching a constitutional issue, that is the course that should be followed (dissenting, Kentridge J). This principle was subsequently confirmed in Zantsi (above) par 3 and followed in Amod (CC) (n 70 above) par 33. Du Plessis (n 62 above) 207-231 calls it the subsidiarity principle. 72 The notion of ‘interstitial’, incremental development that fills up small gaps in received doctrine, case by case, originated in HLA Hart The concept of law (1961) 37-38. The argument is that larger changes in the common law that affect acquired rights should be introduced by legislation. However, judges are well- placed and qualified to make smaller, reasoned ‘elaborations’ or incremental changes to the common law on doctrinal reasoning. The argument that the courts are neither democratically legitimated to bring about major changes nor qualified to judge the wisdom of policy changes is based on the deference approach of the legal process school that responded to Realism by resorting to either deference or craft: A Bickel ‘The Supreme Court 1960 term: The passive virtues’ (1961) 75 Harv LR 40-79; H Wechsler ‘Toward neutral principles of constitutional law’ (1959) 73 Harv LR 1-35. 73 Khumalo v Holomisa 2002 5 SA 401 (CC) par 45 (not shown that the common law ‘as currently developed was inconsistent with the provisions of the Constitution’) can be traced to this argument. The defence of reasonable publication was developed by the SCA in National Media Ltd v Bogoshi 1998 4 SA 1196 (SCA), contradicting earlier case law where publishers were held strictly liable for defamation. Bogoshi argued that this was inspired by the dynamic logic of the law of delict and not by constitutional developments, despite confirmation of the established authority (Pakendorf v De Flamingh 1982 3 SA 146 (A)) in Neethling v Du Preez; Neethling v The Weekly Mail 1994 1 SA 708 (A), the (then still) Appellate Division of the Supreme Court holding that a newspaper could only escape a claim for defamation if it could establish truth. 74 Carmichele (n 58 above) paras 56-59. 75 In Carmichele (n 58 above) par 45 the Constitutional Court referred the matter back to the trial court (paras 81-83), where it was decided that the common law would not have placed a duty upon the state; that the Constitution would place a duty upon the state; that the common law therefore had to be developed to place a duty upon the state; and damages was granted: Carmichele v Minister of Safety and Security 2003 2 SA 656 (C). 96 Normative pluralism and anarchy: Reflections on the 2007 term

Constitutional Court and the civil courts functions well. Given its declared preference for developing the common law incrementally the fact that the Constitutional Court referred the final order back to the High Court could be seen as an effort not to intrude too far into the sphere of the legislature, who should be given maximum scope to amend the common law by legislation.76 At the same time, referring development of the common law to the high courts need not be seen as a sign of deference; the Constitutional Court has indicated that it will not compromise the supremacy of the Constitution and that it will develop the common law if that is what the interests of justice require.77 As Du Plessis points out, the Pharmaceutical Manufacturers decision78 was

the Constitutional Court’s response to a tendency of the Supreme Court of Appeal (at the time) to pretend that constitution-related issues can be disposed of in an ‘enlightened’, rights-friendly manner without reference to (let alone reliance on) the Constitution.79

In some cases the tendency to regard the Supreme Court of Appeal as the specialist court in matters concerning the common law did result in outcomes that restricted the influence of the Constitution on private law. In Brisley80 and Afrox81 it was accepted that public policy (of which open-ended norms such as good faith form part) is informed by fundamental values in the Constitution82 and that the Constitution might ‘spur on the development of new substantive rules of law’,83 but neither good faith nor constitutional values ‘were sufficient to outweigh the traditional bias in favour of the strict enforcement of agreements’84 and the Supreme Court of Appeal was not willing to undertake development of the common law that would amount to judicial interference with agreements willingly entered into by private parties.85

76 Zantsi (n 71 above) par 5. 77 Zantsi (n 71 above) par 4; Harksen v Lane NO 1998 1 SA 300 (CC) par 26. 78 n 37 above. 79 Du Plessis (n 62 above) 217, referring to Bogoshi (n 73 above). Woolman (n 45 above) 10 refers to the same tendency. 80 n 68 above. Roux (n 46 above) 485 argues that the majority in Brisley made a policy choice in favour of legal continuity. 81 n 68 above. 82 In view of Carmichele (n 58 above); see Brisley (n 68 above) par 91; Afrox (n 68 above) par 18. 83 GF Lubbe ‘Taking fundamental rights seriously: The bill of rights and its implications for the development of contract law’ (2004) 121 South African Law Journal 395-423 401. 84 Lubbe (n 83 above) 401 414. 85 Du Plessis (n 62 above) 218-221, referring to Afrox (n 68 above). See Woolman (n 45 above) 95; T Naudé & G Lubbe ‘Exemption clauses — A rethink occasioned by Afrox Healthcare Bpk v Strydom’ (2005) 122 South African Law Journal 441-463; S Woolman & D Brand ‘Is there a constitution in this courtroom? Constitutional jurisdiction after Walters and Afrox’ (2003) 18 SA Public Law 37-82. (2008) 1 Constitutional Court Review 97

During the 2007 term the Constitutional Court decided three cases in which the effect of the Constitution on the common law was in issue: Barkhuizen,86 NM87 and Masiya.88 Stu Woolman concluded that these decisions prove that ‘a penchant for outcome-based decision- making, and a concomitant lack of analytical rigour, has finally caught up with the Constitutional Court’; that ‘the court’s current process of (public) reasoning — its preferred mode of analysis — has genuinely deleterious consequences.’89 In coming to this rather damning conclusion, Woolman attaches particular importance to ‘the court’s persistent refusal to engage in the direct application of the Bill of Rights’ and its insistence upon replacing direct application analysis with what he describes as ‘flaccid analysis in terms of three vaguely defined values’, namely dignity, equality and freedom.90 This strategy of overemphasising the constitutional values and development of the common law (or indirect application) in terms of s 39(2) instead of direct application of the rights provisions in the Bill of Rights, so Woolman argues, ‘has freed the court almost entirely from the text, and thereby grants the court the licence to decide each case as it pleases’, thereby also unwittingly undermining the Bill of Rights and the rule of law.91

These are strong words. If the decisions criticised by Woolman indeed indicate a tendency to free the Constitutional Court from the text of the Constitution so that it can decide each case as it pleases, on the basis of ‘value-speak’, and if that tendency does undermine the Bill of Rights and the rule of law, then we are perhaps indeed seeing the fruits of normative anarchy. What Woolman is suggesting is that the decisions in Barkhuizen, NM and Masiya represent decision making at more or less the same level as Mr Davies senior’s ruling that Robin could not arrange a visit with a friend in the afternoon when he already had another social engagement for that evening, with his parents — the decision is arbitrary because it is based on unilateral pronouncement of vaguely defined values that are neither substantiated by foundation nor open to debate. In the absence of foundation, normative pluralism calls for deliberation and politics, not flaccid but arbitrary and inflexible rules.

86 Barkhuizen v Napier 2007 5 SA 323 (CC) (time-limitation clauses in an insurance contract). 87 NM v Smith (Freedom of Expression Institute as Amicus Curiae) 2007 5 SA 250 (CC) (publication that revealed HIV status of the applicants without consent). 88 Masiya v Director of Public Prosecutions, Pretoria (Centre for Applied Legal Studies and Another, Amici Curiae) 2007 5 SA 30 (CC) (common law definition of rape). 89 S Woolman ‘The amazing, vanishing bill of rights’ (2007) 124 South African Law Journal 762-794 762. 90 Woolman (n 89 above) 763. 91 As above. 98 Normative pluralism and anarchy: Reflections on the 2007 term

In the early stages of the constitutional debate much emphasis was placed on direct horizontal application of the Bill of Rights. In two landmark decisions on direct horizontal application, the Constitutional Court first held that s 7(1) and 7(2) of the 1993 Constitution had to be interpreted restrictively so that direct horizontal application on the law of defamation would be excluded;92 then subsequently decided that direct horizontal application was possible under s 8(2) of the 1996 Constitution.93 In terms of the application debate as it has crystallised in the case law and literature, this does not take the matter much further and we are still wavering between the preservative impulse to uphold the common law tradition unless the Constitution or new legislation explicitly requires us to depart from it and the transformative desire to change the legal infrastructure that inhibits speedy and effective social and economic reform. Most commentators who were initially enthusiastic about direct application have given up on the debate or shifted their focus,94 but Woolman insists that the Constitutional Court’s persistent evasion of direct application ‘has genuinely deleterious effects’.95 In the next two sections I evaluate decisions from the 2007 term, arguing that they hold the key to a different perspective on the application issue. In the following two sections I identify, from the case law of the Constitutional Court, indications of what I describe as a subsidiarity approach that enables the courts to simultaneously promote constitutional reforms and uphold constitutional stability, choosing between the two impulses (and justifying the choice in each case) with reference to constitutional principles of justice and democracy. In the final section of the article I argue that this approach can be justified, even in so far as it creates law that resembles fixed doctrine, in so far as it is not aimed at entrenching shallow formalism but rather at upholding and enabling a constitutionally authorised and required balance between constitution and law, rights and democracy, or stability and change, within the particular historical and constitutional context of the South African Constitution. My underlying assumption, based upon the theoretical analysis in the first two sections of the article above, is that such a balance is justified in so far as it opens up space to give politics a chance, in the sense that it recognises diversity and pluralism and creates room for real social contestation and dissent.

92 Du Plessis (n 62 above) 220, referring to Du Plessis v De Klerk 1996 3 SA 850 (CC). 93 Du Plessis (n 62 above) 221, referring to Khumalo (n 73 above). 94 Eg Van der Walt (n 34 above) 341-363; Roux (n 46 above). 95 n 89 above. (2008) 1 Constitutional Court Review 99

4 Rights and democracy

Legislation enacted by Parliament to give effect to a constitutional right ought not to be ignored. 96

4.1 Introduction

Decisions from the 2007 term indicate the emergence of an ‘angle of approach’97 — an analytic rhythm — according to which courts could decide upon the tension between rights and democracy, while acknowledging what Van der Walt and Botha98 called the irresoluble dissent embedded in the countermajoritarian dilemma. Lourens du Plessis described this kind of thought process as ‘subsidiarity’, defining it as a reading strategy whereby a court refrains from taking a decision that can be taken by a lower court or avoids a constitutional decision if the matter can be decided on a nonconstitutional basis.99 Generally speaking, the notion of subsidiarity poses conceptual and analytical problems for the politics-enhancing approach to the fundamental contradiction set out in the first two sections of the article above. This is particularly true of the way in which the principle of subsidiarity has been enunciated in earlier South African case law. In 1995 the Constitutional Court enunciated a subsidiarity principle in Mhlungu, holding that courts should, whenever possible, decide cases without reaching a constitutional issue;100 in 1998 it formulated another one in Amod,101 stating that the development of the common law should (in terms of the 1993 Constitution) preferably be undertaken by the Supreme Court of Appeal. The latter principle was confirmed (under the 1996 Constitution) in Carmichele.102 Stated this blandly, the principle of subsidiarity looks like a formalist escape

96 Minister of Health NO v New Clicks South Africa (Pty) Ltd (Treatment Action Campaign and Another as Amici Curiae) 2006 2 SA 311 (CC) par 437 (Chaskalson CJ). 97 I am indebted to Henk Botha, who first used this phrase at a workshop in 2007, explaining that it pointedly avoids the pretence of a technique that produces ready or final answers. The phrase originates from NS Ndebele’s novel The cry of Winnie Mandela (2003). 98 See n 19 above and accompanying text. 99 I do not set out or enter into the general subsidiarity debate or literature, where this notion has attracted various understandings and where it remains contested; see Du Plessis (n 62 above) 207-231. Currie & de Waal (n 30 above) 75-78 call this the ‘avoidance principle’, although their interpretation is narrower and more mechanistic. See further L du Plessis ‘The South African constitution as memory and promise’ (2000) 11 Stellenbosch Law Review 385-394 388-389; W le Roux ‘War memorials, the architecture of the Constitutional Court building and counter- monumental constitutionalism’ in W le Roux & K van Marle (eds) Law, memory and the legacy of apartheid. Ten years after Azapo v President of South Africa (2007) 65-90 87. 100 n 71 above, par 59 (dissenting, Kentridge J); confirmed in Zantsi (n 71 above) par 3. 101 n 70 above, par 14. 102 n 58 above, paras 50-55. 100 Normative pluralism and anarchy: Reflections on the 2007 term from politics, but in the historical and constitutional context of the South African Constitution it could also be seen as a politics- confirming and –enhancing device that ensures interplay between constitutional principles and democratic laws, reformist initiatives and vested rights, change and stability. In the former sense, subsidiarity would serve a formalist evasion of politics; in the latter sense it could promote normative pluralism and constitutional change without succumbing to either normative anarchy or reactionary formalism.

In the next two sections I discuss decisions from the 2007 term that could indicate that the Constitutional Court is developing a more general subsidiarity approach.103 In the final section I consider the implications of this approach for the application debate and evaluate the role that subsidiarity could play in promoting normative pluralism and avoiding anarchy.

4.2 The SANDU principle

In SANDU,104 the Court confirmed105 that, once legislation has been enacted to give effect to a right in the Constitution, litigants must rely on the legislation — and may not rely directly on the constitutional provision — when bringing action to protect that right against infringement. The principle establishes a subsidiarity rule because it excludes direct application of the constitutional provision once legislation has been enacted to give effect to that right. Legislation in this category include the Labour Relations Act 66 of 1995 (LRA) (s 23), the Promotion of Administrative Justice Act 3 of 2000 (PAJA) (s 33), the Promotion of Access to Information Act 2 of

103 Aspects of this approach have been decided earlier. In New Clicks (n 96 above) par 97 fn 81 Chaskalson CJ cites with approval a paper in which Cora Hoexter set out the logic of both SANDU and Bato Star. The paper was published as C Hoexter ‘“Administrative action” in the courts’ 2006 Acta Juridica 303-324 (also published as H Corder (ed) Comparing administrative justice across the Commonwealth (2008)). See to the same effect Currie & de Waal (n 30 above) 649. 104 South African National Defence Union v Minister of Defence 2007 5 SA 400 (CC) paras 51-52. See further Pillay (n 29 above) paras 39-40; Chirwa v Transnet Ltd 2008 2 SA 24 (CC) paras 59 (Skweyiya J), 69 (Ngcobo J). In Sidumo v Rustenburg Platinum Mines Ltd 2008 2 SA 24 (CC) par 248 Ngcobo J confirmed the principle; the majority (paras 55-56) did not mention it. In MEC: Department of Agriculture, Conservation and Environment v HTF Developers (Pty) Ltd 2008 2 SA 319 (CC) the majority (par 24) confirmed that the National Environment Management Act 107 of 1998 (NEMA) is legislation contemplated by s 24(b) without mentioning SANDU (but see Ngcobo J par 62). 105 Formulated in NAPTOSA v Minister of Education, Western Cape 2001 2 SA 112 (C) 123B, I-J; mooted in New Clicks (n 96 above) paras 95-96, 434-437 (Chaskalson CJ, Ngcobo J); left undecided in National Education Health and Allied Workers Union v University of Cape Town 2003 3 SA 1 (CC) par 17; Ingledew v Financial Services Board: In re Financial Services Board v Van der Merwe 2003 4 SA 584 (CC) paras 23-24 (Ngcobo J). See IM Rautenbach ‘Overview of Constitutional Court decisions on the bill of rights — 2007’ 2008 Tydskrif vir Suid-Afrikaanse Reg 330-348 330. (2008) 1 Constitutional Court Review 101

2000 (PAIA) (s 32) and the Promotion of Equality and Prevention of Unfair Discrimination Act 4 of 2000 (PEPUDA) (s 9).106 Application of the SANDU principle to PAJA and PEPUDA indicates that the distinction between vertical and horizontal application becomes irrelevant in the context of this principle.

In line with s 39(2), SANDU requires that where ‘the Legislature enacts legislation in the effort to meet its constitutional obligations, and does so within constitutional limits, courts must give full effect to the legislative purpose’.107 In Goedgelegen Tropical Fruits108 the Constitutional Court applied the implications, emphasising that the Restitution of Land Rights Act 22 of 1994 is remedial legislation under the Constitution and that courts, to promote the spirit, purport and objects of the Bill of Rights, must acknowledge the historical background under which labour tenancies were cancelled. The cause of a historical dispossession will often not lie in an isolated moment in time or a single act but in the context and, as the contextual features of this case ‘constituted a grid of integrated repressive laws that were aimed at furthering the government’s policy of racial discrimination’ which ‘materially affected and favoured the ability of the [landowners] to dispossess the applicants of their labour tenancy rights’,109 the former labour tenants could claim restitution.

The SANDU principle indicates that, once legislation has been enacted to give effect to a constitutional right, direct application of the constitutional provision is precluded. However, this applies only when litigants seek to protect rights; when the constitutional validity of legislation is challenged, the constitutional provision applies directly.110 In Islamic Unity111 the Court expanded the qualification: when challenging the constitutional validity of any statute that

106 In NAPTOSA (n 105 above), NEHAWU (n 105 above) and SANDU (n 104 above) the provision was s 23(5); the legislation the LRA. In Sidumo (n 104) and Chirwa (n 104) two provisions (ss 23, 33) and statutes (LRA and PAJA) were in dispute. In Pillay (n 29) the provision was s 9, the legislation PEPUDA. In New Clicks (n 96) the provision was s 33, the legislation PAJA; in Ingledew (n 105) the provision was s 32, the legislation PAIA. In HTF Developers (n 104) the provision was s 24, the legislation NEMA. 107 Sidumo (n 104 above) par 249, citing NEHAWU (n 105 above) par 14. Although Ngcobo J concurred in the outcome in Sidumo, he wrote a separate judgment: see n 104 above. 108 Department of Land Affairs v Goedgelegen Tropical Fruits (Pty) Ltd 2007 6 SA 199 (CC). 109 Goedgelegen Tropical Fruits (n 108 above) paras 53, 66, 69, 71. 110 New Clicks (n 96 above) par 437; SANDU (n 104 above) par 52; Sidumo (n 104 above) par 249; Engelbrecht v Road Accident Fund 2007 6 SA 96 (CC) par 15 (constitutional challenge against legislation involves direct application of s 34). 111 Islamic Unity Convention v Minister of Telecommunications 2008 3 SA 383 (CC) par 59. 102 Normative pluralism and anarchy: Reflections on the 2007 term conflicts with the constitutional principle of administrative justice, s 33 is the review measure.112

SANDU, with its qualification, was originally justified in New Clicks: allowing litigants to rely on s 23 directly instead of on the LRA ‘would encourage the development of two parallel streams of labour law jurisprudence, one under the LRA and the other under s 23(1)’; such a practice would be ‘singularly inappropriate’113 and would contradict the principle of ‘one system of law grounded in the Constitution’.114

Where, as here, the Constitution requires Parliament to enact legislation to give effect to the constitutional rights guaranteed in the Constitution, and Parliament enacts such legislation, it will ordinarily be impermissible for a litigant to found a cause of action directly on the Constitution without alleging that the statute in question is deficient in the remedies that it provides. Legislation enacted by Parliament to give effect to a constitutional right ought not to be ignored. And where a litigant founds a cause of action on such legislation, it is equally impermissible for a court to bypass the legislation and to decide the matter on the basis of the constitutional provision that is being given effect to by the legislation in question.115

The principle expounded by the Court is grounded in the norm that ‘to permit the litigant to ignore the legislation and rely directly on the constitutional provision would be to fail to recognise the important task conferred upon the Legislature by the Constitution to respect, protect, promote and fulfil the rights in the Bill of Rights.’116 The qualification, by contrast, is justified by the norm that the majority is accountable to and can be tested against constitutional limits. SANDU therefore reflects the tension of the countermajoritarian dilemma, with the principle emphasising democracy and the qualification emphasising rights: when the democratically elected legislature meets its constitutional obligations to enact legislation that gives effect to the rights in the Constitution, within constitutional limits, the courts should respect and give full effect to that legislation. To that extent, rights are subject to democracy. However, the courts can review the constitutionality of legislation with a direct appeal to the Constitution. To that extent, democracy is subject to rights. In so far as the SANDU principle promotes this

112 Zondi v MEC for Traditional and Local Government Affairs 2005 3 SA 589 (CC): If a constitutional challenge is brought against legislation for being inconsistent with s 33, the proper approach is to consider whether it can be consistent with s 33: s 39(2) (par 102). The next question is whether it contemplates administrative action for purposes of s 33 (par 103), whether it limits the rights in s 33 (par 106) and whether the limitation is justifiable (s 36(1)). 113 New Clicks (par 96 above) paras 434, 436; NAPTOSA (n 105 above) par 123B, J. 114 Pharmaceutical Manufacturers (n 37 above) par 44. 115 New Clicks (n 96 above) par 437 (Chaskalson CJ). Footnote omitted. 116 SANDU (n 104 above) par 52; Pillay (n 29 above) par 40. (2008) 1 Constitutional Court Review 103 balance between democracy and rights one could say that it could be employed to promote normative pluralism without falling into the trap of either normative anarchy (anything goes) or reactionary formalism (escaping real social contestation and dissent through formalist reasoning). To that extent, the principle could also be said to give politics a chance in that it promotes and enables real social contestation and dissent between the democratic push of the majority and the stabilising pull of constitutional stability.

4.3 The Bato Star principle

In Chirwa117 the Court confirmed a second principle:118 once legislation has been enacted to give effect to a right in the Constitution, and in so far as the legislation was intended to codify the common law, litigants may not rely directly on the common law when seeking to protect that right against infringement. The principle also establishes a subsidiarity rule in that it precludes application or development of the common law;119 once legislation has been enacted to give effect to a right, that right is protected through constitution-conforming interpretation and application of the legislation in accordance with s 39(2).120

Bato Star does not render the common law irrelevant — the common law informs interpretation of the legislation.121 The stabilising role of the common law should be treated with care, though — the common law should inform interpretation of legislation only in so far as it is consistent with the Constitution, consistent with the legislation and able to augment interpretation of the legislation as foreseen in s 39(2), namely to promote the spirit, purport and objects of the Bill of Rights. Du Plessis122 and de Ville123 explain that the relationship between the common law (favouring stability by upholding the status quo) and legislation (favouring change through policy-driven intervention) has to be inverted in view of the centripetal interpretive force of the Constitution.

The areas of existing law identified as affected by this principle are labour disputes (the LRA)124 and just administrative action

117 n 104 above, par 23; see further Fuel Retailers (n 25 above) par 37. 118 Stated earlier in Bato Star (n 63 above) par 25 New Clicks (n 96 above) par 96. 119 In terms of ss 173, 8(3) or 39(2). See to the same effect Currie & de Waal (n 30 above) 650. 120 S 39(2). 121 n 63 above, par 22. 122 L du Plessis Re-interpretation of statutes (2002) 179-181. 123 JR de Ville Constitutional and statutory interpretation (2000) 172 66. 124 Chirwa (n 104 above) par 23. 104 Normative pluralism and anarchy: Reflections on the 2007 term

(PAJA),125 but the Restitution of Land Rights Act 22 of 1994,126 the National Environment Management Act 107 of 1998 (NEMA),127 the Water Services Act 108 of 1997, the Mineral and Petroleum Resources Development Act 28 of 2002128 and the Prevention of Illegal Eviction from and Unlawful Occupation of Land Rights Act 19 of 1998 (PIE)129 could also fit into this category.

Like SANDU, Bato Star is founded on the principle of ‘one system of law grounded in the Constitution’130 and intended to prevent the development of parallel systems of law and jurisprudence, one based on legislation and the other on the common law. Again like SANDU, Bato Star acknowledges the countermajoritarian dilemma in that it prevents the courts from circumventing democratically enacted legislation by reverting to a parallel, alternative system of rules,131 although it leaves room for constitutional review of democratic legislation. Like SANDU, the Bato Star principle enforces normative pluralism by preventing either democratic legislation or constitutional rights from usurping all adjudicative space — by forcefully holding the balance between the two, the principle gives politics a chance in so far as it enables decision-making based on consideration of the contestation between both impulses.

SANDU, Bato Star and the constitutional review qualification, considered together, are signs of a subsidiarity approach: once legislation is enacted to give effect to a constitutional right, litigants must rely on the legislation when bringing action to protect that right; they may not rely directly on the constitutional provision or on the common law. Direct application of the constitutional provision remains possible to attack the constitutional validity of the legislation. Neither constitutional principle nor legislative inter-

125 Fuel Retailers (n 25 above) par 37. 126 Promulgated to give effect to s 121 of the 1993 Constitution, now authorised by s 25(7). See Goedgelegen Tropical Fruits (n 108 above). 127 Enacted to give effect to s 24(b): HTF Developers (n 104 above) par 24. Ss 31A and 32 of the Environment Conservation Act 73 of 1989 must also be interpreted to give effect to s 24: HTF Developers par 19. 128 Arguably promulgated to give effect to s 24(b)(iii) and s 25(5), read with s 25(4)(a). 129 Enacted to give effect to s 26(3): Port Elizabeth Municipality (n 38 above) paras 11-23; Occupiers of 51 Olivia Road, Berea Township, and 197 Main Street, Johannesburg v City of Johannesburg 2008 3 SA 208 (CC) par 16. S 4(1) of PIE explicitly overrides the common law. The CC declined in City of Johannesburg to decide whether PIE applied and what the relationship between PIE and s 26 is, but held that a local authority that evicts residents without meaningfully engaging with them acts at odds with the spirit and purpose of s 26(3). 130 Pharmaceutical Manufacturers (n 37 above) par 44. 131 JW Singer & JM Beermann ‘The social origins of property’ (1003) 6 Canadian Journal of Law & Jurisprudence 217-248 238 point out that the US Supreme Court majority in Lucas v South Carolina Coastal Council 505 US 1003 (1992) characterised ‘traditional common law [property] rules as somehow more worthy than politics advanced through contemporary legislative action’, threatening to derail public policies. (2008) 1 Constitutional Court Review 105 vention or common law tradition is allowed to decide cases automatically by its own force; each case is decided only when and on the strength of contestation between the conflicting impulses of stability and change; rights and constitution; tradition and reform.

In so far as the cases discussed so far reach, these subsidiarity principles are attractive and relatively unproblematic. It is also reasonably simple to justify their application and outcome in the straightforward cases foreseen in these decisions. However, the majority of cases will not be nearly as simple, in that they will not fall within the rather narrow ambit foreseen in the decisions discussed so far. The difficulty is to decide how far the subsidiarity approach reaches; particularly, whether the subsidiarity principles apply only where legislation was enacted specifically to give effect to a constitutional right and intended to codify existing law; whether they also to apply to other legislation; and how they affect the common law.

In the sections that follow I embroider upon the logic of the subsidiarity principles that I have identified in the SANDU and Bato Star decisions, without thereby pretending either that the embroidered or expanded subsidiarity principles reflect the view (or would meet with the approval) of the Constitutional Court or that they constitute doctrinal necessity. The line of argument in what follows is simply to trace what the extent and outlines of the SANDU and Bato Star subsidiarity principles could be if their logic is developed further, into areas not so far foreseen or discussed by the Constitutional Court. I do not present this analysis as a proposal of some kind of subsidiarity or deference principle either — in the final sections of the article, I still want to subject the outcome of this logical exploration to critical scrutiny in view of the normative framework set out earlier: if the logic of subsidiarity is embroidered upon to present a fuller picture than can at present be gleaned from case law, and should this logic be adopted and followed by the courts, and in so far as it might eventually establish legal or doctrinal principles, are or would they be justified? In asking this question, my point of departure will once again be that law (including subsidiarity principles), in so far as it is inevitable, will only be justified to the extent that it promotes normative pluralism but does not descend into either anarchy or reactionary formalism. In allowing room for normative pluralism but simultaneously avoiding anarchy, it must give politics a chance by allowing for real social contention and dissent.

In this spirit, I rely on the perceived logic of subsidiarity in SANDU and Bato Star to determine whether, and how, the subsidiarity principles could apply with regard to other legislation (that it, pre- 1994 legislation and partial or competing legislation) or to cases where there is no applicable legislation at all. 106 Normative pluralism and anarchy: Reflections on the 2007 term

4.4 Application to other legislation

SANDU was formulated for legislation particularly enacted to give effect to a right in the Bill of Rights; Bato Star was aimed at legislation that ‘covers the field’.132 This raises the question whether these principles also apply to legislation that protects the right in question but that was neither enacted specifically with the constitutional provision in mind nor intended to codify existing law.133

The first indication that these principles could apply more widely is the justification for the subsidiarity principles: if the underlying norm is that legislation should not be sidelined in protecting a constitutional right out of respect for the transformative role of the democratic legislature, the principles should apply as widely as possible to legislation that gives effect to a right, including legislation not specifically promulgated with the constitutional provision in mind or not intended to codify existing law on the protection of that right.134

The second reason why flexible application of Bato Star makes sense is that it is often difficult to decide when legislation was intended to codify existing law.135 On the one hand, some gaps are left on purpose: constitutional review of legislative, executive and judicial action is covered by neither s 33 nor PAJA and was made possible by a direct appeal to the constitutional legality principle.136 On the other hand, even legislation intended to ‘cover the field’ might leave gaps on purpose or by mistake: PAJA, to which SANDU and Bato Star clearly apply, covers the grounds for review of administrative action only imperfectly and the common law probably

132 Stated in New Clicks (n 96 above) par 95 with reference to PAJA; see Sidumo (n 104 above) par 90. See I Currie ‘What difference does the Promotion of Administrative Justice Act make to administrative law?’ 2006 Acta Juridica 325- 351 328-334 (whether PAJA is codifying legislation). 133 The Rental Housing Act 50 of 1999 gives effect to s 25(6) (insecure tenure), but is not intended to codify landlord-tenant law. 134 Currie (n 132 above) 328 critiques the tendency to treat the notion that PAJA codified the law on administrative review as if it meant that PAJA ‘made no difference to the existing law’; in other words, we can apply existing law as if it was legitimised by PAJA. Extending Bato Star to partial legislation could force interpreters to acknowledge that the common law survives only in so far as it is consistent with and augment legislation. 135 It is normally easier to identify legislation specifically intended to give effect to a provision, but even there some cases might be difficult. 136 See C Hoexter Administrative law in South Africa (2007) 169-175. Review of these actions is possible on the basis of the legality principle in the Constitution: Fedsure (n 63 above) paras 33-45 (legislative action); President of the Republic of South Africa v South African Rugby Football Union 2000 1 SA 1 (CC) par 142 (executive action); Nel v Le Roux NO 1996 3 SA 562 (CC) par 24 (judicial action). See C Hoexter ‘The principle of legality in South African administrative law’ (2004) 4 Macquarie Law Journal 165-185. (2008) 1 Constitutional Court Review 107 still applies in the gaps.137 The subsidiarity principles should therefore not be restricted to implementing or codifying legislation and it should be accepted that gaps will occur in legislation. Judging on the justification argument, a gap in the field covered by legislation should mean that the subsidiarity principles do not apply in that specific area, allowing direct reference, on that specific point, to either the constitutional provision (SANDU) or the common law (Bato Star). I return to this point later.

The third argument that supports wider application of the subsidiarity principles is the common law presumption that legislation does not amend the common law more than is necessary;138 legislation does amend the common law in so far as the intention to amend is either stated explicitly or implied clearly.139 Considering that pre-constitutional law survives only to the extent that it is consistent with the Constitution140 and that all legislation must be interpreted to promote the spirit, purport and objects of the Bill of Rights,141 it must now be the case that all legislation, whether specifically intended to codify existing law or not, either gives effect to the rights in the Bill of Rights and promotes the spirit, purport and objects of the Bill of Rights or is unconstitutional.142 It thus becomes axiomatic that all legislation must give effect to the rights in and promote the spirit, purport and objects of the Bill of Rights to a certain extent; the intention to amend the common law must therefore be implicit in all legislation and SANDU and Bato Star should apply to it. Lourens du Plessis143 and Jacques de Ville144 argue in the same direction when they state that the common law presumption either cannot apply or should be inverted in the context of the new Constitution.145 Existing law must be interpreted to fit in with constitutional rights and therefore all legislation must be assumed to amend the common law in so far as is necessary to give effect to the Constitution, which is why the common law no longer exists as a

137 Prior to PAJA the courts reviewed decisions taken by private bodies: Theron v Ring van Wellington van die NG Sendingkerk in Suid-Afrika 1976 2 SA 1 (A); Dawnlaan Beleggings (Edms) Bpk v Johannesburg Stock Exchange 1983 3 SA 344 (W); Johannesburg Stock Exchange v Witwatersrand Nigel Ltd 1988 3 SA 132 (A). It has been argued that the common law still applies in this situation: Hoexter Administrative law (n 136 above) 180-18; Currie & de Waal (n 30 above) 645. Other gaps are left by excluding grounds for review from s 6 PAJA, e g the common law rule against vagueness. 138 Du Plessis (n 122 above) 177-181; de Ville (n 123 above) 170-176. 139 Hoexter (n 103 above) 308: the principle that PAJA cannot be sidestepped by resorting directly to the common law is logical, ‘since statutes inevitably displace the common law’ (emphasis in original). Compare Rautenbach (n 105 above) 330. 140 S 39(3). 141 S 39(2). 142 S 172. 143 n 122 above, 179-181. 144 n 123 above, 172, 66. 145 Compare Pharmaceutical Manufacturers (n 37 above) par 45. 108 Normative pluralism and anarchy: Reflections on the 2007 term parallel, alternative system of rules and cannot be reverted to directly (Bato Star).

To put it simply, direct application of the Constitution and the application and development of the common law should only come up in the absence of legislation. Some legislation will give effect to rights in the Bill of Rights more directly and some will affect existing law more explicitly and extensively, but in line with SANDU and Bato Star all legislation either fails constitutional scrutiny or triggers a subsidiarity principle according to which the right must primarily be protected via the legislation and not via direct application of the constitutional provision or the common law. In litigation aimed at protection of a constitutional right, the subsidiarity principles in SANDU and Bato Star postpone direct application of the constitutional provision and application or development of the common law until it is clear that legislation does not cover the case.

Once a flexible reading of the subsidiarity principles is accepted, the next question is how these principles should apply to partial legislation; in other words, legislation that gives partial effect to a constitutional right but does not attempt or manage to ‘cover the field’ as far as protection of the right is concerned. This category is difficult to define because it is impossible to say how partial legislation must be to disqualify it from being comprehensive, as the example of PAJA in the previous section proves. If the subsidiarity principles must apply to legislation that ‘covers the field’ as imperfectly as PAJA does they should arguably apply to all partial legislation when and in so far as the legislation gives effect to a constitutional right. An example illustrates the point.

In Magudu Game Co146 the KwaZulu-Natal High Court argued that the common law principles regarding ownership of wild game should be developed, in line with s 39(2), to provide for the exigencies of commercial game farming. However, the state had already enacted the Game Theft Act 105 of 1991 to protect the rights of commercial game farmers. The Act does not codify the law on wild game — outside of commercial game farming the common law remains unchanged. However, the Act does regulate commercial game farming, and the game farmer in this case could not rely on it purely because he failed to comply with registration requirements. According to the expanded version of Bato Star, development of the common law should not be an option in litigation to protect the right in so far as the legislature has enacted legislation to give effect to it, which in this case it had.

146 Magudu Game Co (Pty) Ltd v Mathenjwa and others NNO 2008 2 All SA 338 (N). (2008) 1 Constitutional Court Review 109

SANDU and Bato Star must therefore be applied flexibly; even supposedly codifying legislation will leave gaps and, if the Constitution or the common law provides for such a gap, the next step could be to turn to the Constitution or the common law to fill that gap rather than challenge the legislation. Shifting the problem up to the Constitution or down to the common law before reverting to a constitutional challenge makes sense if a gap in the legislation means that the specific aspect is not covered by the legislative scheme, which means that the subsidiarity principles do not apply and the gap can be filled by application of constitutional provisions or the common law, as illustrated by constitutional review of legislative, executive or judicial acts147 and by judicial review of administrative action in cases where private bodies exercise public power.148

It should have become clear by now that application of the subsidiarity principles in SANDU and in Bato Star respectively to partial legislation would work out differently. Applying SANDU to partial legislation involves the question whether litigants should be allowed to appeal directly to the constitutional provision to protect a right if legislation that does provide for its protection but was not made specifically to give effect to the right does not provide the required remedy in a particular case. The extension would probably work better with pre-constitutional than with new legislation, but generally there is no reason why such a relaxation of SANDU should be impossible in principle. Two qualifications should probably apply, though. Firstly, the actual purpose and scope of partial legislation should be decisive in deciding whether a direct appeal to the constitutional provision would be justified if the legislation fails to provide a remedy. Secondly, to prevent the development of parallel systems of law, this relaxation of SANDU should arguably only be possible if the gap cannot be filled by applying or developing the common law. In other words, as long as a gap in legislation can be filled by application of the common law, a direct appeal to the Constitution should not be possible in an action to protect the right — this is why the Supreme Court of Appeal’s decision in Tswelopele149 to leave the common law ‘untouched’ and to create a new constitutional remedy was arguably a mistake. The result of Tswelopele is two parallel remedies for ante omnia restoration of spoliated possession, based on the Constitution and the common law, the only difference being that the constitutional remedy permits a restoration order that includes the use of replacement materials. Developing the common law, for a clearly circumscribed range of cases, would have vindicated and celebrated judicial victories over

147 n 136 above. 148 n 137 above. 149 Tswelopele Non-Profit Organisation v City of Tshwane Metropolitan Municipality 2007 6 SA 511 (SCA). 110 Normative pluralism and anarchy: Reflections on the 2007 term apartheid injustice in decisions such as Fredericks150 and could have made a contribution to the rehabilitation of the common law of property, which did not cover itself in glory during apartheid, by demonstrating how it can support social justice.151

Since Bato Star purely prevents direct appeals to the common law by circumventing legislation, it is easier to argue that the principle should be qualified so that the common law still applies (and can be developed) in cases where partial legislation does not in fact give complete, covering effect to a constitutional right. However, a qualification is once again necessary in line with s 39: the common law can only fill gaps in legislation if and to the extent that it is not only consistent with the Bill of Rights and with the legislative scheme but, in so far as the legislation leaves a gap, offers the possibility (where necessary through development) of augmenting the legislation, keeping in mind the constitution-enhancing interpretive rule in s 39(2). The governing principle should not be that the common law survives where it remains unaffected by constitutional or legislative provisions, but rather that the common law survives only if and in so far as it is consistent with the Bill of Rights, consistent with existing legislation, and capable of complementing the legislation in giving effect to constitutional rights, either as it stands or through being developed for the purpose.

The Magudu Game Co152 example suggests that SANDU and Bato Star could also apply to pre-constitutional legislation. According to s 39, pre-constitutional law only survives in so far as it does not conflict with constitutional provisions and can be interpreted in conformity with constitutional rights and values. Accordingly, the 1991 Act protects ownership of game to give effect to the property clause (s 25) in a specific area, namely commercial game farming. A similar argument applies to all pre-constitutional legislation that can survive the new Constitution and therefore SANDU and Bato Star should apply to it. However, although it has to be interpreted in line with the Constitution, pre-constitutional legislation cannot be assumed to reflect the legislature’s effort to comply with constitutional obligations. The Expropriation Act 63 of 1975, being the only general expropriation legislation currently in force, predates the Constitution and must therefore be interpreted in line with s 25(2) and (3).153 Legislation is required to give effect to the expropriation and compensation provisions in s 25 and since there is nothing else, the

150 Fredericks v Stellenbosch Divisional Council 1977 3 SA 113 (C). 151 See AJ van der Walt ‘Developing the law on unlawful spoliation and squatting’ (2008) 125 South African Law Journal 24-36. 152 n 146 above. 153 A Draft Expropriation Bill 2008 was tabled in the Portfolio Committee on Public Works on 26 March 2008: Parliamentary Monitoring Group’s website at www.pmg.org.za. This Bill was recalled for further consideration in August 2008. (2008) 1 Constitutional Court Review 111

1975 Act applies to all expropriation, but it was drafted in another era and not intended for application in the remedial and transformative context of s 25, and therefore the Act must be read through the corrective lens of s 39(2). SANDU and Bato Star should apply to the 1975 Act, but in applying it the courts must keep in mind that it embodies the policies of another time and a different government and therefore it requires conscious and comprehensive re-interpretation to ensure that it gives effect to current constitutional rights and values. The majoritarian consideration that justifies SANDU, namely that the legislature’s role in the transformation process must be respected and that it must be given full effect, applies only indirectly to pre-constitutional legislation: it can only be assumed to embody the legislature’s effort to give effect to constitutional rights in so far as it allows constitution-promoting interpretation according to s 39(2), but there is always the possibility that parts of the old legislation still reflect the policies of a disgraced pre-constitutional legislature and then a constitutional challenge might be required. Up to that point, the subsidiarity principles apply in the form of the obligation to read the legislation in conformity with the Constitution, as s 39(2) requires.

4.5 Competing and complementary legislation

A further problem created by flexible application of SANDU and Bato Star is that several pieces of legislation, including supposedly codifying legislation like PAJA and the LRA and partial, even pre- constitutional legislation, sometimes apply to the same dispute simultaneously.154 If more than one statute compete with or complement each other, the subsidiarity principle should be that competing or complementary legislation must be applied to optimally give effect to the Bill of Rights and to promote the spirit, purport and objects of the Bill.155 As a starting point this will require purposive interpretation that aligns all the applicable laws with the

154 Eg SANDU (n 104 above); Sidumo (n 104 above); Chirwa (n 104 above); Fuel Retailers (n 25 above). 155 In Nakin v MEC, Department of Education, Eastern Cape Province 2008 2 All SA 599 (EC) paras 49-51 Froneman J argues that courts should not be forced to ‘compartmentalise’, by conceptual exclusion, the possibilities for adjudicating labour disputes, namely common law of contract, labour relations and administrative justice, since a conceptual selection between the alternatives could rob the applicant of an effective remedy because the route selected (e g PAJA) might not offer the remedy (e g payment of money) the applicant asks for. If this suggests that SANDU / Bato Star should not apply and that litigants (or the courts) should have a free choice, it conflicts with normative arguments in favour of subsidiarity. However, in so far as it points out that a meaningful remedy is important the problem could be overcome by applying subsidiarity (e g PAJA applies) but then, if PAJA does not offer a remedy, cascade the issue on remedy down the subsidiarity hierarchy, ie by either reverting to the common law, if that offers a suitable remedy, or to constitutional review, possibly opening up space for reading in a remedy. 112 Normative pluralism and anarchy: Reflections on the 2007 term constitutional scheme. Some remaining problems could be solved by focusing on the separate fields of application of each law156 or by identifying their position in a subsidiarity grid. It has, for example, been proposed that PAJA embodies general administrative law while other statutes grant particular administrative powers, thus indicating their relative status: the LRA is particular legislation with regard to labour disputes and therefore its provisions with regard to administrative action supersede but are ‘suffused’ by the general reasonableness principles of PAJA, which is general legislation.157 A similar hierarchical relationship was construed between NEMA and the Environment Conservation Act 73 of 1989.158

When legislation limits one right in the process of giving effect to another, the affected person can protect her right with a direct attack on the constitutional validity of the offending legislation; SANDU allows a direct appeal to the constitutional provision. When constitutional rights provisions come into conflict, the Constitutional Court supports an optimising interpretation strategy.159 In other cases s 39(2) requires the courts to select the interpretation that avoids the conflict or promotes the spirit, purport and objects of the Bill of Rights. In these cases there does not seem to be a reason to depart from the subsidiarity principle that protection of a constitutional right should primarily be based on legislation giving effect to that right and not on the constitutional right itself or on the common law, except in so far as the constitutional validity of the legislation or the common law is attacked (in which case the constitutional provision applies directly) or in so far as the legislation does not cover the field (in which case a direct appeal to the Constitution or the common law could be possible).

Several decisions from the 2007 term concerned regulatory legislation that limits a constitutional right, without the legislation being intended to give effect to that or any other right. Mohunram160 is the latest decision161 in which the proportionality of civil forfeiture of instrumentalities of crime in terms of the Prevention of Organised Crime Act 121 of 1998 (POCA) was contested, without attacking the

156 Land reform laws enacted under s 25(6) to improve security of tenure, e g the Land Reform (Labour Tenants) Act 3 of 1996, the Extension of Security of Tenure Act 62 of 1997 and the Rental Housing Act 50 of 1999, each has its own field of application. 157 Chirwa (n 104 above) par 23 confirmed Bato Star (n 63 above) paras 59, 149; I Currie & J Klaaren The Promotion of Administrative Justice Act benchbook (2001) par 1.24; Currie & de Waal (n 30 above) 643 fn 4. Sidumo (n 104 above) paras 104, 110 states that the LRA is suffused by the constitutional reasonableness principle but in fact reads it via PAJA. 158 HTF Developers (n 104) par 24; Fuel Retailers (n 25). 159 See eg Christian Education (n 33 above) par 42. 160 Mohunram v National Director of Public Prosecutions (Law Review Project as Amicus Curiae) 2007 4 SA 222 (CC). 161 See Prophet v National Director of Public Prosecutions 2007 6 SA 169 (CC). (2008) 1 Constitutional Court Review 113 constitutional validity of the Act. It has been accepted that civil forfeiture (particularly beyond the ambit of organised crime) could result in disproportionate and arbitrary deprivation of property; hence the Constitutional Court devised a definition of instrumentalities of crime162 and a proportionality test to determine whether forfeiture is justified.163 Both tests are aligned with the non- arbitrariness test in FNB164 to ensure that civil forfeiture orders under POCA would not offend the protection of property against arbitrary deprivation in s 25(1). The proportionality test employed in these cases to interpret the legislation in a constitution-complying way confirms SANDU.

In two other forfeiture cases, Armbruster165 and Van der Merwe,166 the facts were similar — the affected persons had large sums of foreign currency in their possession when boarding an international flight; the currency was seized by customs officials and declared forfeit; the affected persons attempted to reclaim the money. In Armbruster the applicant launched a constitutional attack against the regulations with an appeal to s 34; in Van der Merwe he appealed to the common law. SANDU allows a direct appeal to a constitutional provision in an attack on the validity of legislation, but according to Bato Star a direct appeal to a common law remedy should only be available if legislation does not ‘cover’ that particular aspect and the common law is capable of interpretation that conforms with the Constitution and complements the legislation.

In Armbruster the Constitutional Court held that the connection between the purpose of the forfeiture, the property and the person deprived ‘could hardly be closer’; that there is sufficient reason for the deprivation and that forfeiture would not be arbitrary.167 Forfeiture could produce ‘arbitrarily harsh consequences’ but the

162 Prophet (n 161 above) was applied in Mohunram (n 160 above) par 44: instrumentalities had to be given a wide meaning, but the property must play a reasonably direct role in commission of the offence. This test is based on the non- arbitrariness standard in First National Bank of SA Ltd t/a Wesbank v Commissioner, South African Revenue Service; First National Bank of SA Ltd t/a Wesbank v Minister of Finance 2002 4 SA 768 (CC) paras 51, 100. 163 Prophet (n 161 above) par 65; Mohunram (n 160 above) paras 64, 145-147. In Mohunram the majority decided that the forfeiture would be disproportionate because part of the immovable property was used for legal purposes and because the owner had already been prosecuted, convicted and fined. 164 The non-arbitrariness standard was set out in First National Bank (n 162 above) paras 51, 100. 165 Armbruster v Minister of Finance 2007 6 SA 550 (CC). The Treasury declared the foreign currency forfeit in terms of the Exchange Control Regulations under s 9 of the Currency and Exchanges Act 9 of 1933. 166 Van der Merwe v Inspector Taylor 2008 1 SA 1 (CC). Various grounds were raised, ranging from the Criminal Procedure Act 51 of 1977 to the Income Tax Act 58 of 1962; before the CC consensus was that the seizure was authorised by s 20 of the Criminal Procedure Act. 167 Armbruster (n 165 above) par 71. The nexus remarks are inspired by First National Bank (n 162 above) paras 51, 100. 114 Normative pluralism and anarchy: Reflections on the 2007 term regulations grant an opportunity for representations and a discretion to return the currency ‘to ameliorate undue hardship or injustice that might be perpetrated on the person affected’; therefore the regulations do not allow for arbitrary deprivation.168 The Court also rejected the attack based on s 34, pointing out that the reasons why legislation was struck down in Zondi and Chief Lesapo, namely a particular historical and social context and complete lack of judicial control, were absent in this case.169

In Van der Merwe, the applicant elected to claim the money on the basis of the common law rei vindicatio, arguing that foreign currency bought with funds from his banking account was his property. The High Court decided that some of the currency did not belong to him because he held it for someone else. The Constitutional Court was divided; the majority held that he proved ownership of the amount bought and held for himself, but O’Regan J would have dismissed leave to appeal because the issues have not been properly ventilated.170 One has to agree that the Court had not seen enough evidence or heard enough argument on whether the rei vindicatio is at all feasible when cash is reclaimed;171 when ownership vests in cash withdrawn from a bank; and whether the applicant’s state of mind with regard to ownership of currency could have had any effect. In terms of Bato Star the primary question in Van der Merwe should have been whether reclamation of the money was regulated by the authorising legislation — if it was, a direct appeal to the common law remedy should have been precluded.

When there is no legislation to trigger SANDU or Bato Star, the tug of war between the Constitution and the common law is most intense. The question is then whether the civil courts should be allowed to withdraw into the comfort zone of the common law, where the reach of democratic reform is weak and where the influence of the Constitution could be reduced to that of an ally in entrenching vested rights, privilege and power against democratic interference. The most difficult issues of application feature in this area, with the focus ultimately on the question whether (and how far) the independence of private law is affected by the supremacy of the Constitution.

168 Armbruster (n 165 above) par 80, relying on Dawood v Minister of Home Affairs; Shalabi v Minister of Home Affairs; Thomas v Minister of Home Affairs 2000 3 SA 936 (CC) para 53. 169 Armbruster (n 165 above) paras 59-60, referring to Chief Lesapo v North West Agricultural Bank 2000 1 SA 409 (CC), Zondi (n 112 above), Metcash Trading Ltd v Commissioner, South African Revenue Service 2001 1 SA 1109 (CC). 170 Van der Merwe (n 166 above) par 103. 171 Woodhead Plant & Co v Gunn (1894) 11 SC 4: the currency status of money is incompatible with the right of anyone to follow stolen money into the hands of an honest receiver for value, whether it had been mixed up with other money or not. (2008) 1 Constitutional Court Review 115

5 Constitutional supremacy vs independence of private law172

Mir scheint es sich [bei der Konzeption einer ‘Grundrechtsfreiheit des Privatrechts’] um eine Art Münchhausentheorem zu handeln, das den Privatrechtlern die Möglichkeit verschaffen soll, sich am eigenen Schopf aus dem Sumpf des Verfassungsrechts herauszuziehen.173

5.1 SANDU and Bato Star in the absence of legislation

When no legislation applies to the protection of a constitutional right, the logic of the SANDU subsidiarity principle would imply that a litigant who wants to protect that right must rely on the common law and may not rely directly on the constitutional right, unless she challenges the constitutional validity of the common law. Mhlungu and Zantsi were no doubt intended to justify exactly such a subsidiarity principle to prevent every private quarrel being blown up into a constitutional issue when it could be dealt with in terms of common law. In terms of s 39(2), the obligation to have first resort to the common law includes development of the common law to align it with the spirit, purport and objects of the Bill of Rights. However, once it is clear that the constitutional right cannot be given effect to, according to the spirit, purport and objects of the Bill of Rights, by either application or development of the common law a litigant can resort directly to the constitutional provision,174 either to launch a constitutional attack against a rule or institution of the common law that limits or is in conflict with the constitutional right175 or to craft a special constitutional remedy.176

It can be argued that the legislature, in so far as it leaves the common law unamended by legislation, accepts that the common law, for the moment being, sufficiently gives effect (or can be developed to give effect) to the rights in the Bill of Rights — if the legislature is dissatisfied with the common law it can change it. Out of respect for the legislature’s democratic role, protection of the constitutional rights should, therefore, in the absence of legislation begin with the common law — a direct appeal to the Constitution should be reserved for cases where it is impossible to interpret the common law in a constitution-compliant way. This is possibly what

172 Based on the title of M Ruffert Vorrang der Verfassung und Eigenständigkeit des Privatrechts (2001). I am indebted to Friedrich Schoch for bringing it to my attention and to him and Michael Sachs for explaining German theory to me. 173 CW Canaris Grundrechte und Privatrecht (1999) 18. 174 Masiya (n 88 above) par 27. 175 The CC declared the common law crime of sodomy unconstitutional: Gay and Lesbian (n 28 above). 176 In line with s 172. 116 Normative pluralism and anarchy: Reflections on the 2007 term inspired the unfortunately phrased remark that the common law should be allowed to develop incrementally so as not to interfere with legislative authority.177

The justification for this residuarity principle is similar to that of SANDU: if litigants are allowed to choose freely between constitutional provisions and common law remedies as if they were independent, parallel systems of law, the common law will never be ‘brought into’ the new constitutional dispensation. This is what happened in Tswelopele:178 crafting a constitutional remedy and leaving the common law ‘untouched’ created two parallel remedies instead of developing the common law in line with constitutional values. The goal of ensuring one system of law under the guidance of the Constitution is promoted not only by preventing the common law to develop as a parallel system (Bato Star), but also by preventing development of a parallel system of unnecessary constitutional remedies. Leaving some aspect of the common law intact and crafting a new constitutional remedy should be justifiable only on constitutional grounds.

In was argued in the previous section that SANDU and Bato Star apply to legislation regardless of horizontality;179 the same should hold in the absence of legislation. Direct application and development of the common law are broader than and should not be confused with or collapsed into horizontal application; the former two will come up in vertical and horizontal relationships.180

5.2 Direct vs indirect application

The most intractable question when no legislation exists is whether the constitutional provisions apply directly or whether they should be mediated through the common law. This issue came up in three 2007 decisions.

177 Zantsi (n 71 above) par 5; Masiya (n 88 above) par 31. Roux (n 46 above) 487 notes, with reference to RG Teitel Transitional justice (2000) 24, that judges may have to do more work in transitional or transforming societies and leave less for the legislature. 178 Tswelopele (n 149 above). 179 The principle finds support in s 8(3)(a), which applies particularly to horizontal relationships: in applying the provisions of the Bill of Rights to natural and juristic persons in terms of subs 8(2) a court, ‘in order to give effect to a right in the Bill, must apply, or where necessary, develop, the common law to the extent that legislation does not give effect to that rights’ (emphasis added). I pay no direct attention to development of the customary law because none of the 2007 decisions dealt with it. 180 Tswelopele (n 149 above); Masiya (n 88 above). (2008) 1 Constitutional Court Review 117

In Masiya181 the question was whether the common law definition of rape should be developed to include anal penetration.182 The Constitutional Court decided that the common law definition was not unconstitutional because it criminalised morally and socially unacceptable behaviour. However, the definition fell short of the spirit, purport and objects of the Bill of Rights and had to be expanded.183 The Court distinguished between its roles in testing whether legislation is consistent with the Constitution and developing the common law, adding that the courts should not ‘appropriate the Legislature’s role in law reform’ by developing the common law in a way ‘closer to codification than incremental, fact-driven development’.184 This supports the subsidiarity principle that direct application (validity challenge) should only follow once it is clear that the common law cannot be developed appropriately. Accordingly, the case was decided as an opportunity to develop the common law rather than a constitutional validity challenge. The majority decided to develop the common law only as required by the facts of this case, to include non-consensual anal penetration of a female,185 but in a dissenting judgment Langa CJ opined that the development should have included anal rape of men.186

Woolman187 criticises Masiya because it ‘never truly considers the direct application of the substantive provisions of the Bill of Rights to the challenged common-law definition of rape’; instead, the ‘entire analysis of the common-law rule takes place within the rubric of s 39(2) and in terms of indirect application.’ Woolman argues that, in

181 n 88 above. 182 The regional court decided that the common law definition was underinclusive and extended it to non-consensual anal penetration of any person, but it did not have the power to develop the common law: Masiya (n 88 above) par 66. Magistrates’ courts are prevented from pronouncing on the validity of any law: s 110 of the Magistrates’ Courts Act 32 of 1944. The High Court endorsed the a quo decision and it came before the CC for confirmation under s 172(2)(a). 183 Masiya (n 88 above) paras 27, 32, 70. Masiya paras 66, 69 underlines the point that s 39(2) has a restricted purpose: the power to develop the common law is granted to the CC, the SCA and the High Courts by s 173 and not in s 39(2), which merely instructs the courts, when developing the common law or customary law, to do it so as to promote the spirit, purport and objects of the Bill of Rights — the section concerns the how of constitutional development of the common law but not the when. For similar reasons, s 8(3) should not be seen as the authorising provision for a general obligation or power to develop the common law either; the subsection pertains to cases of horizontal application, as is evident from the introductory phrase ‘[i]n applying the provisions of the Bill of Rights to natural and juristic persons in terms of subsection (2), ...’ (emphasis added). In spelling out the direction of development s 39(2) indicates when the common law might require development, but it is not the authorising provision. 184 Masiya (n 88) par 31. 185 Masiya (n 88) par 45. 186 Masiya (n 88) par 75. Sachs J concurred in the dissenting judgment. 187 n 89 above, 768. 118 Normative pluralism and anarchy: Reflections on the 2007 term view of the language of ss 8(1)188 and the decision in Khumalo,189 the substantive provisions in the Bill of Rights should have been applied directly instead of via the development of the common law. The point in which Woolman differs from Masiya is not easily identifiable; his statement that direct application ‘takes the rights and freedoms, and the general rules derived from them, as our point of departure for determining whether law or conduct is invalid’ echoes the SANDU principle almost exactly,190 as does his view that indirect application (protection of the right) does not permit a finding of invalidity.191 The problem is that he describes indirect application as ‘a mode of analysis that neither specifies whether a particular right demands vindication nor permits a finding of invalidity.’192 Woolman’s view that indirect application does not tell us whether the right requires vindication is based on the narrow assumption that vindication of constitutional rights primarily takes place via direct application. In terms of SANDU logic, constitutional rights can also be vindicated indirectly, via constitution-compliant interpretation of the common law. According to Masiya, indirect protection should precede direct constitutional challenges as a matter of subsidiarity, out of respect for democratic processes, while SANDU declares direct appeals to the constitutional provisions impermissible when protection of the right is in issue.193

Woolman therefore differs from the principle laid down in SANDU in that he claims the major part of constitutional litigation for direct application and assigns indirect application to a vague and apparently insignificant area of ‘law and conduct not engaged by any of the specific provisions set out in Chapter 2’, ensuring that this part of all law is also brought ‘into line with the “spirit, purport and objects of

188 Although s 8(1) provides that the Bill of Rights applies to all law and that it binds the legislature, the executive, the judiciary and all organs of state, it does not say that the Bill of Rights applies directly or horizontally to all law. See to much the same effect Khumalo (n 73 above) paras 31-33. Woolman (n 45 above) 44 disagrees, arguing that s 8(1) alone makes all law subject to the direct application of the Bill of Rights. 189 In Khumalo (n 73 above) the CC indicated that direct application was possible under the 1996 Constitution. 190 Woolman (n 89 above) 769. As far as legislative, executive and judicial conduct is concerned Woolman reflects the approach of the courts: these forms of conduct are excluded from PAJA by the definition of ‘administrative action’ but review is possible on the basis of legality: n 136 above. However, as far as conduct by organs of state generally is concerned, at least some of it is administrative action covered by PAJA, in which case Woolman’s formulation does not reflect SANDU in that review of this conduct takes place via PAJA and not directly. 191 All law, including legislation and the common law, should only be declared unconstitutional once it has been established that it is impossible to find a constitutionally consistent interpretation: Masiya (n 88 above) par 27. However, rules of the common law can be declared unconstitutional when it is impossible to interpret or develop it according to values of the Constitution: Gay and Lesbian (n 28 above). 192 Woolman (n 89 above) 769 (emphasis added). 193 The legality principle case law is explained by this formulation; n 136. (2008) 1 Constitutional Court Review 119 the Bill of Rights” and the “objective, normative value system” made manifest in the text of the Constitution as a whole’.194 By contrast, the residuarity principles establish a rational link between direct application and constitutional remedies or validity challenges on the one hand and indirect application and protective litigation on the other. As was argued earlier, this link is justifiable in view of the normative roles it assigns to democratic processes and judicial review respectively. Moreover, it makes sense if considered from the perspective of remedies: direct application is aimed at reviewing the constitutional validity of legislation or the common law and will at least sometimes result in declarations of invalidity, which might not be what a litigant was after when seeking to protect her right. In some cases direct application can admittedly produce a special constitutional remedy and therefore the subsidiarity principles keep the possibility of direct application open for that purpose, but based on the respect-for-democracy norm the subsidiarity approach restricts this to a last resort solution rather than an open option that exists parallel to statutory or common law protection.

By shifting the focus from constitutional validity challenge to protective litigation, where the point is to develop the common law in line with s 39(2), Masiya is consistent with SANDU. The Constitutional Court decided that the correct approach was to expand the common law definition by developing it in line with s 39(2), in this case to include at least non-consensual anal penetration of females.195 The minority decision shows that one can disagree with this narrow development of the definition and still agree that indirect development of the common law was the right approach. The cautious, restricted development undertaken by the majority cannot be blamed on their choice not to apply the constitutional provisions directly but to develop the common law; much rather, it results from earlier decisions in which it was said that the avoidance principle (subsidiarity) implies deciding just what is necessary for the facts of the case and no more.196 This is an exaggerated interpretation of subsidiarity and hardly defensible in view of constitutional provisions and decisions to the effect that all law is governed by the Constitution. The subsidiarity principle in Mhlungu and Zantsi could be read less restrictively (in view of SANDU) as saying that a court should not protect a constitutional right by way of a direct validity attack or by way of a direct constitutional remedy before considering whether the legislation or common law in question could be interpreted in a constitution-conforming and –confirming way.

194 Woolman (n 89 above) 769. 195 Masiya (n 88 above) par 27. 196 Especially Zantsi (n 71 above) par 8. 120 Normative pluralism and anarchy: Reflections on the 2007 term

Apart from an unnecessarily narrow view of subsidiarity or the avoidance principle, the Constitutional Court should also not have allowed itself to be pulled into common law talk about incremental development of the common law. Constitutional transformation will sometimes require large developments and changes. The warning in Mhlungu and Zantsi against undermining the separation of powers is important, but the subsidiarity principles in SANDU and Bato Star promote respect for the role of the legislature, without confusing that with respect for the common law logic of incremental development. Respect for democratic processes should not prevent the courts from rooting out remnants of tradition such as discrimination or inequality in conflict with the new constitutional dispensation; anal rape will be unconstitutional regardless of the sex or gender of the victim and hence the courts need not shy away from larger development of the common law out of respect for the legislature.

The decision in NM197 also turned on a choice whether to develop the common law; Woolman198 again criticised it for failure to apply the constitutional provisions directly. The applicants claimed that the respondents had violated their rights to dignity and privacy by publishing their names and HIV status without permission. The case could support direct horizontal application of ss 10 and 14, but the applicants claimed on the basis of the common law actio iniuriarum and accordingly, in terms of residuarity, their claim must be considered on the basis of s 39(2) first. Initially, the question is therefore whether the common law provides a remedy as it stands and, if it doesn’t, whether it can be developed. The applicants brought the matter in the form of a constitutional case because they argued that the common law would have to be developed; the Constitutional Court decided that the case ‘involves a nuanced and sensitive approach to balancing the interests of the media, in advocating freedom of expression, privacy and dignity of the applicants irrespective of whether it is based on the constitutional law or the common law.’199 In other words, even though the claim is brought under the common law and the application and development of the common law are the main issues, balancing of the constitutional rights raises a constitutional issue. This reflects the meaning of s 8(1) according to SANDU subsidiarity logic: the Constitution applies to all law, but not necessarily directly.

The decision eventually turned on the majority’s factual finding that the respondents were aware that the applicants had not given permission for the publication of their names, followed by the legal decision that this was sufficient to establish intent for the actio

197 n 87 above. 198 n 89 above, 781-783. 199 NM (n 87 above) par 31. (2008) 1 Constitutional Court Review 121 iniuriarum and that the requirements for the action were thus met. On the factual finding it was unnecessary to develop the common law. The minority, who disagreed with the majority’s factual finding, would nevertheless have developed the common law so that negligence could also satisfy the requirements.200 In either case, direct application of the constitutional provisions remains uncalled for under the subsidiarity principles.

Woolman201 criticises Barkhuizen202 for the same reason as Masiya and NM, namely that the Constitutional Court should have applied the constitutional provisions directly. Although it is difficult to tell whether Barkhuizen was decided in line with SANDU, the case was clearly brought as a direct constitutional challenge but not decided as one. The constitutional challenge was brought against a time-limitation clause in a short-term insurance contract; the applicant argued that the clause violated his s 34 right of access to court. The High Court upheld the claim but the Supreme Court of Appeal dismissed it, arguing that the Constitution did not prevent time-limitation clauses in contracts entered into freely and voluntarily and that there was inadequate evidence in this case to indicate that the contract was not entered into freely and voluntarily. The Constitutional Court found that the 90 day time limitation was not manifestly unreasonable on the face of it, nor was it unfair, because there was no evidence that the contract was not concluded freely between parties in equal bargaining positions. Since the applicant did not furnish reasons for non-compliance it would not be unjust to enforce the time bar against him.203

Testing a contractual provision directly against a provision in the Bill of Rights (s 34) amounts to horizontal application under s 8(2)- (3).204 The Constitutional Court was uncomfortable with horizontal application, noting that the contractual time limitation reveals no law of general application. The Supreme Court of Appeal identified the underlying principle pacta sunt servanda as the law of general application, but the Constitutional Court dismissed this option, pointing out that the contractual term was challenged, not the general principle.205 Enforcement of an allegedly unfair contractual term admittedly is not law of general application, but it is conduct that can be declared unconstitutional in terms of s 172(2). The fact

200 NM (n 87) paras 44-47 57. Langa CJ and O’Regan J agreed that the journalist and publisher might have been negligent, but not that intent had been proved: paras 111, 125. See Woolman (n 89) 781. 201 n 89 above, 772-781. 202 n 86 above. 203 Barkhuizen (n 86 above) paras 63 66-67 84 86. 204 Barkhuizen (n 86 above) par 23. It is untrue that the Court yet had to consider this issue: Khumalo (n 73 above). 205 Barkhuizen (n 86 above) paras 24 26. 122 Normative pluralism and anarchy: Reflections on the 2007 term that conduct does not establish law of general application merely means that it cannot limit a constitutional right and cannot be saved by the general limitation clause, which is not surprising — the limitation clause is intended to give the state the opportunity to save otherwise unconstitutional law. However, conduct can infringe upon a right and can therefore be the focus of a constitutional validity attack.206 It was therefore conceivable that this case could have been decided as a direct constitutional attack on allegedly unconstitutional private conduct. If the court should find, in such a case, that enforcement of the contract would indeed infringe s 34 or any other constitutional right, it would have to declare the conduct invalid. At the same time, private conduct that infringes a private right should, in most cases, be regulated by private law unless there is a constitutional reason for deciding the conflict on a constitutional basis.

Barkhuizen could also have been decided purely on the basis of private law — time limitation clauses in private contracts are generally invalid if they are against public policy. The question is therefore whether there was a reason to decide the case on constitutional grounds. One such a reason could have been the desire to set a constitutional standard for contractual time limitation clauses that limit the right of access to courts. In that case the format of a direct constitutional attack on validity of the conduct would have been suitable. Another possibility could have been the need to develop the common law principle that gives credence to time limitation clauses, namely that contracts entered into freely between equal parties should be enforced — pacta sunt servanda. Such an approach would arguably have been justified if the Constitutional Court thought that the common law gave too much weight to the sanctity of contract principle and too little to justice, and that the common law should be developed in terms of s 39(2) to rectify the balance. However, this would have required the case to be transformed from a direct validity challenge to an indirect development of the common law protection case, as happened in NM and in Masiya. In the latter instance one would expect the court to test the relevant common law principle against central constitutional values and explain how it is to be amended in view of those values.

In the event, the Constitutional Court did neither of the above, preferring to decide the case in another way altogether, namely to determine whether the contractual clause was contrary to the values underlying the constitutional democracy and contrary to public policy.207 The Court pointed out that contractual terms denying

206 Woolman (n 89 above) 774-775 makes much the same point. 207 Barkhuizen (n 86 above) par 36. (2008) 1 Constitutional Court Review 123 access to the courts were contrary to public policy and therefore invalid in common law208 — in principle, the time limitation clause could have been tested for fairness and validity purely on common law grounds, without any input from the Constitution. It then proceeded to analyse, in great detail but without much clarity, the relations between the substantive constitutional rights provisions such as s 34, the values of the Constitution and public policy. In effect, the Constitution and its underlying values are described as a source of values that inform the public policy that plays a yardstick role in the common law. Woolman correctly describes this as deciding a direct constitutional challenge indirectly by ‘speaking in values’,209 that is, not with reference to the relevant rights provisions in the Bill of Rights but mediated through a veil of values and policy vaguely infused by the substantive content of the constitutional provisions. In deciding the case this way, the Court did not set a constitutional standard for contractual time limitation clauses, but it also failed to either test or develop the underlying common law principle — in fact, the Court confirmed that the principle pacta sunt servanda ‘gives effect to the central constitutional values of freedom and dignity’, of which ‘self- autonomy, or the ability to regulate one’s own affairs, even to one’s own detriment’ is the very essence.210 In other words, the Court not only implied that value analysis, rights analysis and policy analysis is the same thing211 but also accepted that the central private-law principle pacta sunt servanda operates on the same level and enjoys the same status as constitutional values, without giving much indication how this should work out in the larger constitutional project.

In a society torn apart by historical injustices and inequality, transformation under the aegis of a democratic and supreme Constitution means that the seemingly neutral but in fact highly contested principal tenets of private law, with pacta sunt servanda and absolute ownership top of the list, have to be treated with contextual sensitivity in the reconstruction of a new democratic legal order. By reducing concrete constitutional rights provisions to a veil of values that inform public policy, and by allowing private law principles like pacta sunt servanda to join those values in the shaping of public policy, the Constitutional Court might appear to subscribe to the highly problematic view, in decisions like Brisley212 and Afrox,213 that there are no significant differences between constitutional and common law values and that the two systems could co-exist in a way

208 Barkhuizen (n 86 above) par 34. 209 n 89 above, 763. 210 Barkhuizen (n 86 above) par 57. 211 Woolman (n 89 above) 779. 212 n 68 above. The critical comments of Roux (n 46 above) are highly relevant here. 213 n 68 above. 124 Normative pluralism and anarchy: Reflections on the 2007 term that does not demand substantive changes of the common law. Instead of adapting the common law, it looks as if the constitutional values have been integrated into the common law.

One could argue that this result is justifiable in view of the importance of the principle involved, namely sanctity of contract, but even the Supreme Court of Appeal is apparently not that set on pacta sunt servanda. In Linvestment214 the Court overturned long-standing precedent and decided that the owner of a servient tenement can insist upon relocation of a defined and registered servitude of right of way and that the dominant owner would be obliged to accept the relocation, against her will, if non-relocation would cause significant harm for the applicant and relocation would not cause significant harm for the respondent. The reasons for the decision and the result are justifiable on policy grounds, but it is puzzling that the Court would be so willing to make what is really a dramatic change to the common law on policy grounds, while being so unwilling, despite equally strong or stronger policy reasons, to change the common law very slightly in Tswelopele. Judging from Linvestment the best solution in Barkhuizen was the approach of the Supreme Court of Appeal, namely to focus on the general common law principle that supports enforcement of time limitations, pacta sunt servanda. Even when subjected to a healthy dose of constitutional skepticism, this principle embodies significant social and moral values and therefore it is highly unlikely that the courts would be willing to declare it unconstitutional just because it sometimes allows enforcement of unfair contracts. Instead, the approach in Masiya indicates that the courts would most likely, if this principle were challenged, attempt to rectify any shortcomings by developing it in view of s 39(2).

Deciding Barkhuizen on this basis would have confirmed the subsidiarity approach that the Court seems to be developing, and it would have given a welcome indication that the Court is working towards a truly substantive, critical analysis of the common law in view of s 39(2). The point should not be whether the common law protects values like freedom and autonomy that are also important in constitutional law, but where and how common law principles need to be developed to give effect to the spirit, purport and objects of the Bill of Rights. More concretely, it was necessary in Barkhuizen to ask where and how the principles of contract need to be developed to give effect to the right of access to the courts in s 34, given the historical and constitutional context within which inequalities in privilege and power developed and within which the right of access to courts was denied the poor and weak. It is in this context that the mere possibility that contractual time limitations could be unfair and

214 Linvestment CC v Hammersley 2008 3 SA 283 (SCA). (2008) 1 Constitutional Court Review 125 against public policy in current private law doctrine might not pass constitutional muster — the test that was ultimately applied by the Constitutional Court majority in Barkhuizen is little more than the private law test and it lacks the critical sensitivity for context that one would expect of a full-on constitutional challenge. Moreover, as in NM, the Court’s analysis of the facts in Barkhuizen was disappointing; the factual analysis of Sachs J is a much better example of the kind of analysis that was required and that is to be expected.

As it stands, Barkhuizen neither tested the parties’ conduct for constitutional validity, nor did it consider either development of or a challenge to the supporting common law principle; instead, common law principle was elevated to the sphere of constitutional values that indicate whether conduct is against public policy. In the broader context of the 2007 decisions, Barkhuizen looks like a constitutional non sequitur.

6Conclusions

The subsidiarity principles set out in SANDU and Bato Star could, according to the analysis above, reasonably be expanded so as to include pre-1994 legislation, partial and competing legislation, and instances where no legislation is applicable at all. As has been argued in the sections above, the subsidiarity logic of these cases could be developed in a way that would make it possible simultaneously to promote constitutional reform goals and to respect democratic legislative interventions that reflect reformist or stabilising policies. Obviously the wider implications of subsidiarity have not been worked out by the courts and therefore the analysis above is purely speculative. Furthermore, the more comprehensively worked out subsidiarity logic set out in the previous sections is not presented as a proposal to develop or entrench a judicial doctrine of subsidiarity or deference — it is nothing more than an exercise in the logic of the SANDU and Bato Star decisions, and it is presented here purely in order to have an opportunity to test whether such a logic, if it were in fact adopted by the courts, would be justifiable in view of the broader normative and constitutional principles discussed in the first sections of the article.

With the exception of Barkhuizen (and perhaps Sidumo), the Constitutional Court’s decisions in the 2007 term display a growing and reasonably consistent tendency to favour a particular subsidiarity approach, always present in our constitutional setup but apparently now emerging with increasing clarity, according to which direct application of the constitutional rights provisions is restricted to constitutional challenges against the validity of legislation. In the absence of such a challenge, the Court prefers to treat cases for the 126 Normative pluralism and anarchy: Reflections on the 2007 term protection of constitutional rights as indirect application cases, involving either the interpretation of legislation or development of the common law in terms of the constitution-compliance and —confirming prescriptions of s 39(2). By extension, taking into account earlier jurisprudence, direct application could also be possible in the absence of a direct validity challenge, if the goal is to craft a special constitutional remedy. However, again extrapolating from earlier jurisprudence, the subsidiarity approach indicates that direct application (to craft a constitutional remedy or to challenge a law) is considered only when indirect protection (interpretation of legislation or development of the common law) proves impossible for some reason. Furthermore, the subsidiarity hierarchy in the category of indirect protection cases prescribes that application and development of the common law should only be considered once and in so far as there is no legislation to apply.

The subsidiarity approach of the 2007 cases offers a fresh perspective on Mhlungu, according to which courts should, whenever possible, decide cases without reaching a constitutional issue.215 Woolman articulated five objections against the implication that the salutary ‘principle of avoidance’ — or subsidiarity — in Mhlungu should turn into a ‘full-blown doctrine in which a court must never “formulate a rule of constitutional law broader than is required by the precise facts to which it is to be applied”’;216 some of these objections are relevant to my analysis. Firstly, a too narrow interpretation of the avoidance rule (as in Zantsi) contradicts the project that constitutional interpretation should evolve into a principled dialogue between the Constitutional court, other courts, the legal profession, law schools, Parliament and, indirectly, the public at large.217 In addition, a too narrow interpretation would undermine rational political discourse and the possibility of a coherent jurisprudence. These are convincing reasons why the subsidiarity principle should not be understood too narrowly, as it arguably was in at least Zantsi — it should not become a directive to avoid deciding constitutional issues per se. Subsidiarity should be understood in terms of its constitutional purpose and justification, as that has been spelled out in 2007: to preserve the constitutional power and obligation of the courts to control the constitutional validity of legislation, while at the same time paying due respect to the democratic power and legitimacy of policy makers and legislatures in giving effect to their reform obligations under the Constitution.

215 n 71 above, par 59 (dissenting, Kentridge J); Zantsi (n 71 above) par 3. 216 Woolman (n 89 above) 784-785, citing Zantsi (n 71 above) par 8. 217 Mhlungu (n 71 above) par 129. (2008) 1 Constitutional Court Review 127

The purpose of subsidiarity can therefore not be pure avoidance. Much rather, it should be acknowledgement and respect for the fact that courts face the aporia of the countermajoritarian dilemma every time they hear a constitutional case. On the one hand they must respect and give full effect to the legitimate efforts of the democratically elected branches to honour their constitutional obligations in bringing about the reforms legitimised by the Constitution. Doing that requires the courts to consider the interplay between Constitution and legislation and to give the fullest possible purposive effect to legislation enacted by the legislature. At the same time the courts must apply pre-constitutional legislation and the common law in a way that acknowledges the undemocratic origin and history of those sources of law and the dilemmas caused by their continued validity, in both instances by reading them through the corrective lens of s 39(2). In addition to its obligations concerning the interpretation of legislation and development of the common law, the courts have the constitutional duty and power to control the legitimacy of new legislation made by democratically elected legislatures, testing it against Constitutional requirements and limits. The only way to juggle these tasks is to acknowledge, upfront, that these duties of the courts imply contradictory and conflicting normative values and that every decision sacrifices one for the sake of the other.

In the introduction I identified the push for constitutional transformation and the preservative pull of the common law tradition as conflicting normative impulses. I pointed out that this instance of normative pluralism differs from normative diversity as a positive constitutional value in that, in the case of the conflict between tradition and change, the complete picture is not enriched by the simultaneous presence of the more-than-one; these two impulses are locked in a zero-sum game for control over the balance between stability and reform. Either one wants and gets the upper hand at the direct cost of the other. Stability is attainable only at the cost of reform; change is possible only by sacrificing certainty. You can’t do both. And yet, if you are a judge in the new South African constitutional order, you have to.

Normative anarchy, always a threat in a situation like this, is only to be avoided, I argued in the introduction, by giving politics a chance, which starts with realising that all law is political, but politics cannot be enclosed in law. It proceeds with acknowledging that law does not resolve the aporia of stability and change; it merely sacrifices one for the other, for now. It requires making the sacrifice, as far as possible, with a view to open up rather than close down room for social deliberation and significant dissent. Obviously, any choice for an approach that would avoid anarchy could easily end up by entrenching reactionary formalism; in the face of uncertainty and 128 Normative pluralism and anarchy: Reflections on the 2007 term anarchy it is only too easy to settle for the false security of shallow and cheap certainty.

Giving effect to constitutional rights in a fragmented legal system consisting of legislation that was purposely enacted to protect and support those rights, legislation that pre-dates the political transformation and common law that was traditionally applied as if it embodied a-contextual neutral principles is a complex task. This task becomes more difficult when the courts are empowered both to protect constitutional rights and to control the constitutional validity and fittingness of all law, finding a justifiable balance between democratic will and constitutional legitimacy. Above all, the courts exercise this awesome responsibility as just one small part of the large project of social and political transformation, within a dynamic context that reflects both the ills of the past and the aspirations of the future.

The Constitutional Court appears to have chosen the analytic rhythm of a set of subsidiarity principles to direct the choices it must make in this arduous process; in 2007 it followed these principles with a measure of consistency and some sense of purpose and justification. Developing a subsidiarity approach to give effect to its contradictory and conflicting powers and responsibilities is dangerous because it could easily disintegrate into shallow formalism, but it also holds the possibility of evolving into a more reflective, democracy-confirming recognition of the tension between the Constitution and existing law than has been evident from earlier judicial explorations into application issues. In that respect the direction of the pull of 2007 must be welcomed. LEGAL SUBSIDIARITY & CONSTITUTIONAL RIGHTS: A REPLY TO AJ VAN DER WALT

Karl Klare*

1Introduction

The Constitutional Court of South Africa is one of the most respected legal institutions in the world. Its work excites interest among lawyers and others far beyond South Africa’s borders who are inspired by the Constitution’s democratic and egalitarian aspirations. The Constitution1 and the Court2 are precious accomplishments in the human striving to discover and realise the institutional foundations of freedom and self-determination. South Africans are guardians of these achievements for all of the world’s people.

It was therefore unnerving when inane and thuggish attacks were lobbed at the judiciary in the run-up to Nicholson J’s judgment in Zuma’s Case.3 Prominent public figures branded the judiciary ‘counter-revolutionaries’. Adolescent poseurs threatened to ‘kill’ for their heroes and ‘eliminate’ their enemies. The objects of these attacks included the Court’s great jurists such as Pius Langa CJ, who has devoted his lifetime to the cause of social justice under law, and Dikgang Moseneke DCJ, who at age 15 was the youngest political prisoner sent to Robben Island.

* George J & Kathleen Waters Matthews Distinguished University Professor, School of Law, Northeastern University, Boston (USA). Thanks to Lucy Williams and Sandra Liebenberg for comments and suggestions, and to André van der Walt for the friendship and the intellectual challenge. 1 ‘Constitution’ or ‘FC’ refers to the Constitution of the Republic of South Africa, 1996. (In the context of discussing US law, ‘Constitution’ is used in reference to the US Constitution.) 2 The ‘Court’ (with an upper case ‘C’) or ‘the CC’ refers to the Constitutional Court. In the context of the concluding discussion of US law, ‘the Court’ refers, as is customary, to the United States Supreme Court. References to all other courts use a lower case ‘c’. 3 Zuma v National Director of Public Prosecutions [2009] 1 All SA 54 (N) (setting aside decision to prosecute, reversed on appeal in National Director of Public Prosecutions v Zuma 2009 2 SA 277 (SCA)).

129 130 Reply to vd Walt

In this fraught context, friends of the CC must avoid the temptation to rally around the Court by showering it with congratulatory platitudes. We best respect the Court by continuing as before to study its work meticulously and to provide the most rigorous and critical analysis of which we are capable. In that spirit, I have the honour, as a non-South African, to contribute this reply to Professor van der Walt.

Professor van der Walt gifts us with a densely layered essay in which he tries to invent a new grammar for legal analysis.4 He argues that we must become comfortable with gaps rather than seeking refuge in supposedly stable and all-inclusive rules because, first, this brings us closer to understanding what actually goes on in legal reasoning, and, second, because gaps open space for conversation, diversity, dissent, contestation, and judgment. That is, legal gaps open space for politics and the emergence of moral communities. Van der Walt believes, and I agree, that a legal discourse that embraces uncertainty and conflict fits better with a transformative constitution than arid syllogisms. Surely he is correct that South African jurists must work out new methods for addressing legal problems consonant with the Constitution’s transformative project. It is highly implausible that legal reasoning methods developed in earlier times, including apartheid times, are adequate for this task.

It is important when reading van der Walt to appreciate that ‘politics’ carries a positive valence. In the mainstream view, something has gone badly wrong when ‘politics’ enters legal decision making. By contrast, Van der Walt does not regard politics as an enemy of the rule of law. A reception of politics into legal discourse can be a welcome contribution to democracy,5 provided that (1) adjudicators acknowledge the discretion that permits and compels them to make value-laden (‘ideological’) choices in the routine course of their work, (2) they become deeply self-conscious, self-critical, and candid about the ‘inarticulate premises, culturally and historically ingrained’6 that inform their thinking and choices, and (3)

4 AJ van der Walt ‘Normative pluralism and anarchy: Reflections on the 2007 term’ (2008) 1 Constitutional Court Review 77. 5 From this perspective, the problem with the recent attacks on the judiciary is not that they were political. The problem is that they were uninformed, fatuous, and threatening, and they were aimed to achieve power for a particular political leader then facing serious criminal charges. In short, their purpose and their effect were to close down, rather than to open up, space for discussion and dissent. 6 Du Plessis v De Klerk 1996 3 SA 850 (CC), 1996 5 BCLR 658 (CC) para 119 (Kriegler J). (2008) 1 Constitutional Court Review 131 we understand law making processes as a conversation between adjudicators, the other branches, and the public.7

Van der Walt focuses on the proper relationship between three bodies or sources of law — the Constitution, legislation (pre- and post- 1994), and the common law. He is concerned about the ‘preservative pull of the common law tradition’ in light of the common law’s ‘undemocratic origin and history’.8 He fears that the continued validity and prestige of the common law threaten to short-circuit the transformative project, and he wants to minimise that risk:

An important premise of my argument is that the Constitution’s commitment to [political and ethical, as well as social] diversity as a positive value is premised upon the reduced authority of the common law.9

He warns that uncritical obeisance to common law thinking will immobilise the transformative project, and I could not agree more.10

2 From normative pluralism to legal subsidiarity

In quick succession, van der Walt courses through a philosophical problem, a jurisprudential problem, and a question of constitutional interpretation. Subsidiarity comes in later as an instantiation of the approach van der Walt crafts to address more abstract considerations. The background philosophical problem is that modernism and postmodernism have undermined the foundations of epistemological fundamentalism: there is no knowledge-platform from which to identify trans-historical and trans-cultural (‘universal’) norms. The jurisprudential problem is that Legal Realism and Critical Legal Studies ‘destroyed the theoretical credibility of fundamentalism’ by demonstrating ‘the indeterminacy and contingency’ of legal norms.11 Superficially this would appear to leave lawyers with a choice between retreating to empty formalism, on the one hand; or, on the

7 In this regard, see the comment of Sachs J in S v Mhlungu 1995 3 SA 897 (CC), 1995 7 BCLR 793 (CC) para 129: ‘[T]he question of interpretation [is] one to which there can never be an absolute and definitive answer and ... in particular, the search of where to locate ourselves on the literal/purposive continuum or how to balance out competing provisions, will always take the form of a principled judicial dialogue, in the first place between members of this court, then between our court and other courts, the legal profession, law schools, Parliament, and, indirectly, with the public at large’. 8 Van der Walt (n 4 above) 127. 9 Van der Walt (n 4 above) 84 (footnote omitted). 10 On this theme, see generally D Davis & K Klare, ‘Transformative constitutionalism and common and customary Law’ (forthcoming). 11 Van der Walt (n 4 above) 79. 132 Reply to vd Walt other, sliding into normative anarchy (the idea that ‘anything goes’).12

Van der Walt refuses this binary, insisting that we can ‘recognis[e] that values are contingent and that everything is open for dissent and discussion, without implying that everything goes’.13 He holds out for ‘normative pluralism’ in the belief that rational discussion about values is possible and meaningful even in our post-modern condition. We must conceive legal practices and discourses as an opening to endless political debate and dissent.14 This means that every legal decision involves a matter of judgment with a decisionist element.15 One value-set is chosen over and at the expense of others, but only ‘for the time being’. All legal decisions are ‘provisional’ and may be qualified or superseded in the next case or next generation. Van der Walt craves ‘fundamentality without fundamentalism’, to use Morton Horwitz’s phrase.16

The constitutional problem is the built-in ‘tension between the push for constitutional reform and the pull of traditional stability’.17 Van der Walt’s ‘central premise’18 is that this tension reflects ‘a larger conflict between two opposing normative values or ideologies’:

Those who think that the existing distribution of wealth, privilege and power and the law that helped create and uphold it are politically neutral and therefore not in need of political reconsideration favour a regime of rights in which the direct effect of the Constitution is preservation of the status quo; those who think that the existing regime and its laws are inherently politically constructed and hence in need of political reconsideration favour the view that neither law nor rights are immune from constitutional effect or democratic redefinition ...19

How are jurists to negotiate the push and pull of these competing values without rejecting either in favour of the other — that is,

12 Van der Walt (n 4 above) 103. 13 Van der Walt (n 4 above) 84. 14 Van der Walt joins a growing number of commentators who think the time has come to replace the evocative and influential metaphor of the Constitution as a ‘bridge’. See Interim Constitution, Postamble, ‘National unity and reconciliation’ para 1 (‘[t]his Constitution provides a historic bridge ... ’); see also E Mureinik’s memorable ‘A bridge to where? Introducing the interim Bill of Rights’ 1994 (10) South African Journal on Human Rights 31. A new metaphor is needed for the deliberative process launched by the Constitution, perhaps seeing it as an ‘endless highway’. 15 ‘Decisionist’ refers to Van der Walt’s insistence upon ‘the sacrificial nature’ of legal decisions and their inability ever ‘to resolve social conflict with reference to a common will or purpose’ (Van der Walt (n 4 above) 6). 16 M Horwitz ‘The Supreme Court 1992 term — Foreword — The constitution of change: Legal fundamentality without fundamentalism’ (1993) 107 Harvard Law Review 32. 17 Van der Walt (n 4 above) 88. 18 As above. 19 Van der Walt (n 4 above) 89. (2008) 1 Constitutional Court Review 133 without retreating to a conservative formalism that sweeps the question under the rug, or to a moral anarchy that simply disregards it? The answer is to find an idiom of Constitutional adjudication that ‘uphold[s] the tension’20 thereby ‘giving politics a chance’.21 While no particular decision can be justified by reference to foundational principles, making decisions can be justified, provided that we are able to extract a direction-giving purpose from the Constitution.22 A direction-giving purpose is not a ‘foundational principle’. It is a rudder to steer us between the Scylla of formalism and the Charybdis of anarchy.

Borrowing in part from Frank Michelman,23 van der Walt makes the crucial, albeit contestable, assumption that the Constitution does indeed provide a direction-giving purpose, viz, to ‘favour transformation’ — to renounce everything that the apartheid legal order stood for and to embrace the central transformative values of human dignity, equality, freedom, social justice, democracy, and an open society, values which are the reverse of everything apartheid was. Constitutional analysis must ‘yield to the Constitution’s direction-giving (transformative) purpose’ while ‘holding open’ the tension between stability and transformation, between the law as it is and the law as it should be in the free society toward which South Africans are transitioning.24

At this point Van der Walt proposes subsidiarity theory as a transformation-oriented and closure-refusing approach to sorting out

20 Van der Walt (n 4 above) 91. 21 Van der Walt (n 4 above) 127. See also 105: ‘[L]aw ... in so far as it is inevitable, will only be justified to the extent that it promotes normative pluralism but does not descend into either anarchy or reactionary formalism. In allowing room for normative pluralism but simultaneously avoiding anarchy, it must give politics a chance by allowing for real social contention and dissent’. 22 Van der Walt (n 4 above) 87. 23 Michelman argues that constitutional supremacy means that all sites of legal practice and norm-creation must pull in the same direction, and that the principle of constitutional supremacy is a value within the meaning of sec 1 of the Constitution (not just a trumping rule) when that direction is toward transformation of South African society in light of the ‘vision’ (such as it is) contained in the founding values (s 1) and elsewhere instinct in the Constitution. See ‘The rule of law, legality and the supremacy of the Constitution’ in S Woolman et al (eds) Constitutional law of South Africa (2nd Edition, OS, 2005) 11 - 36 — 11 - 38. 24 Van der Walt (n 4 above) 82 - 90. I agree that, on its best reading, the Constitution contains the direction-giving purpose to ‘favour transformation’. But more work is needed on the content and historical provenance of the direction- giving purpose. As formulated by Van der Walt, the direction-giving purpose has greatest analytical power with respect to unravelling apartheid and abolishing its legal incidents. ‘Favour transformation’ provides less guidance with respect to the construction of the new society of democracy, freedom, and social justice. For example, what forms and degree of redistribution of social and political power and wealth does this ‘direction’ imply? Could a critic accept Van der Walt’s ‘direction’ for purposes of discussion and yet plausibly claim that it is silent on such questions? 134 Reply to vd Walt the relationships between common law, legislation, and a transformative Constitution. Van der Walt’s subsidiarity refers to the hierarchy of sources of law — not, as in European law, to the hierarchy of decision making levels.25 The core tenet is that courts should ‘avoid[ ] a constitutional decision if the matter can be decided on a nonconstitutional basis’.26 Van der Walt assumes that legislation embodies a forward-looking, transformative ‘pull’, whereas the common law tugs in a preservative direction.27 Accordingly he seeks to prioritise legislation over the common law and to shield Parliament’s work from judicial encroachment. In this respect, subsidiarity restates a familiar norm of almost all modern, democratic legal systems that courts should defer to the legislature (except perhaps in special situations, for example, when the constitutional rights of politically excluded and subordinated groups are at stake). A second goal is to encourage courts to avoid any tendency to proliferate separate tracks or sub-systems of law such as, for example, parallel systems of remedies respectively grounded on the Constitution itself and on legislation or the common law.

Van der Walt finds traces of subsidiarity thinking in the Court’s recent work, which appropriately balances stability and transformation. He ‘embroider[s] upon the logic of the subsidiarity principles that [he has] identified in [these] decisions’.28 The provisional verdict is that the Court’s emerging subsidiarity perspective will permit normative pluralism to flourish by avoiding premature resolution of the stability/transformation tension. In van der Walt’s argot, the Court’s cases embrace aporia so as to give politics a chance on a continuing basis.

25 In EU law, ‘subsidiarity’ denotes a mandate for downward devolution of decision making to the lowest level at which a particular decision may appropriately be taken. The idea is to harvest local knowledge, facilitate public participation and empowerment, and respect local concerns. The principle was formally introduced into EU law by the Treaty of Maastricht (entered into force 1 November 1993). The present formulation is found in Article 5 of the Treaty Establishing the European Community (consolidated version following the Treaty of Nice (entered into force 1 February 2003)), which provides in relevant part: ‘In areas which do not fall within its exclusive competence, the Community shall take action, in accordance with the principle of subsidiarity, only if and in so far as the objectives of the proposed action cannot be sufficiently achieved by the Member States and can therefore, by reason of the scale or effects of the proposed action, be better achieved by the Community’. Strictly speaking, this type of subsidiarity is not Van der Walt’s concern. At one point he mentions a precept that courts should refrain from taking a decision that can be taken by a lower court, but very little discussion is devoted to the hierarchy of courts. 26 Van der Walt (n 4 above) 99 (footnote omitted). See also (95, 99, and 119 - 120) discussing S v Mhlungu 1995 3 SA 867 (CC), 7 BCLR 793 (CC) para 59 (giving rise to the principle, as somewhat reformulated by Van der Walt, that ‘a court should not protect a constitutional right by way of a direct validity attack or by way of a direct constitutional remedy before considering whether the legislation or common law in question could be interpreted in a constitution-conforming and -confirming way’) (n 4 above) 119. 27 Van der Walt (n 4 above) 104 - 106. 28 Van der Walt (n 4 above) 105. (2008) 1 Constitutional Court Review 135

The orchestration might differ, but my thinking is largely in tune with van der Walt’s philosophy and jurisprudence. My misgivings concern his concluding discussion of subsidiarity. I am not persuaded that subsidiarity theory can perform the functions van der Walt asks of it or that it is a reliable starting point for developing the transformative constitutional methodology he seeks.

3 Subsidiarity through a critical lens

3.1 Introduction

The raison d’être of subsidiarity principles is to strike an authoritative balance between the conflicting values of judicial deference and constitutional supremacy, so that courts are not at large weighing the conflict on an ad hoc, case-by-case basis. The central weakness of the theory is that, when called upon to resolve difficult cases, subsidiarity does little more than instruct courts to reopen the question and conduct the ad hoc balancing exercise it counselled them to avoid. Moreover, at least as presently formulated, subsidiarity imports conventional separation-of-powers discourse without rigorous critical examination; it relies upon the deceptively simple but under- examined and ambiguous notion of a statute ‘giving effect’ to a constitutional right; and, as a result, it tends to skew analysis in the direction of stability rather than transformation.

3.2 Imports

The subsidiarity approach is not a critical theory. Like most conventional discourse of separation-of-powers, subsidiarity rests on timeworn and reified conceptions of the various branches and organs of government and their inter-relationships. This imagery relies as much on wishful thinking as on evidence or analysis. Van der Walt knows this of course, but when he recruits subsidiarity to the cause of transformation, he cannot shed all of its baggage. Because this is a general problem in legal theory, I simply register my objection pro forma and leave detailed examination of these issues to another day.

The trite assumption is that the legislature is the democratic and representative branch, to which non-elected judges owe deference. We twist ourselves into knots trying to explain why and under what circumstances counter-majoritarian judicial review might be legitimate in a democratic society. Jurists spin out a catechism of legal fictions and maxims from the simplistic notion of legislative 136 Reply to vd Walt primacy, which on occasion van der Walt incongruously deploys.29 Van der Walt no doubt takes a critical view of all this, but the discourse is implicit in the subsidiarity theory he imports. He even goes a step farther, consistently portraying the legislature as the forward- looking, progressive branch and a repository of transformative intentions and energy, whereas he treats the courts as necessarily backward-looking, reactionary, and obstructionist.

These assumptions fall to be interrogated in the perspective of transformative constitutionalism. They make a fine starting point for discussion but ought not substitute for investigation and analysis. We must move outside the standard frame of the ‘counter majoritarian’ dilemma,30 for at least two reasons.

First, political institutions (parliaments, courts, agencies) do not have an intrinsic nature or an inherent leaning toward or against transformation. Whether, for example, the legislature is representative and/or plays a democratic, let alone a progressive, role are matters of contextual judgment on the historical record. ‘Process failures’ happen.31 Legislatures have been known to ally with reaction, to be captured by wealthy interests, and/or to be so dominated by an entrenched oligarchy as to lose touch with the grass roots. I do not say any such thing has occurred in South Africa since 1994, although Parliament’s lock-step ratification of the arms deal provides food for thought.32 The point is simply that any separation- of-powers theory worth having in the 21st century cannot assume that, but must inquire whether, legislatures as we find them actually function in service to the popular will.

Second, it is much too late in the day to assume that judicial review of legislation under a supreme and rights-rich constitution is inherently anti-democratic. Yes, judicial review may derogate from majority rule. But at the same time, judicial enforcement of rights and other constitutional provisions can deepen and enhance

29 For example, he cites the fiction that the legislature approves that portion of the common law which it has not amended by statute (Van der Walt (n 4 above) 115). 30 It is worth noting that Van der Walt makes this precise point in an illuminating, earlier essay. See AJ van der Walt ‘Transformative constitutionalism and the development of South African property law’ (part 1) 2005 (4) Tydskrif vir die Suid-Afrikaanse Reg 655 (contrasting the ‘counter majoritarian dilemma’ — the familiar problem of the legitimacy of judicial review of legislation — with a new, ‘transformative dilemma’). The Constitution ‘created a new adjudicative dilemma of simultaneously upholding the supremacy of the new constitution and the integrity of a well-developed and established system of private law’ (657 (footnote omitted)). 31 By analogy to the terms ‘market failure’ and ‘regulatory failure’, I use ‘process failure’ to refer to a systemic breakdown in the idealised relationships between branches of government. 32 See P Govender Love and Courage: A Story of Insubordination (2007) 197 - 205 (former ANC parliamentarian’s account of process of voting on arms deal). (2008) 1 Constitutional Court Review 137 democracy by making legislative bodies more responsive, legislative processes more transparent, or elections more competitive, or by opening communication channels between the populace and government officials. The net result in context might be that the democracy-enlarging consequences of the court’s decision outweigh its counter-majoritarian form. Treatment Action Campaign33 comes to mind as an example.

Moreover, South Africa’s Constitution embodies a conception of democracy that goes well beyond representation, elections, and majority rule. South African constitutional democracy contemplates not only robust political rights, but also that the resources and social circumstances necessary for meaningful political participation and authentic experiences of self-determination will gradually be extended to all South Africans. As Theunis Roux meticulously demonstrates:

The principle of democracy in South African constitutional law is not that collective decisions shall be taken by majority vote, but something far deeper than this, including, at the very least, the notion that the people’s will may be trumped by individual rights where this serves the democratic values of ‘human dignity, equality and freedom’.34

Depending on what they do with their power, a more assertive role for courts than is contemplated by the conventional wisdom may be consistent with or possibly even required by democracy as envisioned by this Constitution. Recognising that judicial enforcement of the Bill of Rights can be democracy-enhancing disrupts traditional separation- of-powers assumptions and calls for a re-examination of the old shibboleths.35

33 Minister of Health v Treatment Action Campaign (No 2) 2002 5 SA 721 (CC), 2002 10 BCLR 1033 (CC). 34 T Roux ‘Democracy’ in S Woolman et al eds (n 23 above) 10 - 66 (quoting FC sec 7(1)). 35 At one point, Van der Walt discards the conventional discursive framework and reverts to his accustomed, critical self. His insight is so powerful that it deserves separate mention. The CC and other courts, he says, conflate the common law insistence that courts restrict themselves to incremental development of the law with the separation-of-powers imperative that courts defer to the legislature. He convincingly argues that courts are perfectly capable of respecting the superior authority of the legislature while at the same time taking bold steps as necessary and appropriate: ‘Respect for democratic processes should not prevent the courts from rooting out remnants of tradition such as discrimination or inequality in conflict with the new constitutional dispensation ... hence the courts need not shy away from larger development of the common law out of respect for the legislature’ (Van der Walt (n 4 above) 120). In fact, Van der Walt hopes that subsidiarity will provide courts with a decisional framework authorising and encouraging them to take big (transformative) steps when Parliament is silent and to challenge Parliament when its efforts to give effect to the Bill of Rights are too grudging. 138 Reply to vd Walt

3.3 Problems of application

Subsidiarity performs a ‘gate-keeping function’. It instructs courts to answer certain threshold questions before entertaining suits seeking to vindicate a constitutional right or to obtain constitutional damages or other special relief on the Constitution itself (‘Constitution-based remedies’). In principle, subsidiarity analysis tells a court whether to proceed with an action styled as a claim on the Constitution itself (or for Constitution-based remedies); or whether, in the alternative, to limit the claimant to such rights and entitlements as are available within the compass of a pertinent statute giving effect to the constitutional right, if any, or the common law, if not. At stake in this inquiry is whether a court may heed its own judgment on how to give effect to the constitutional right or whether it must leave that task, and the prudential and fiscal choices involved, to the legislature. Van der Walt explains well enough why in a given type of case courts should take the approach prescribed by subsidiarity. Unfortunately, subsidiarity often cannot tell us which cases have the specified characteristics, notwithstanding van der Walt’s assiduous efforts to ‘embroider’ the theory. As a practical matter, courts can only answer the threshold, gate-keeping questions by examining and balancing the pertinent constitutional values and principles. Courts cannot apply subsidiarity theory without addressing precisely the questions, and making precisely the value judgments, that the theory means them to avoid.

Under SANDU, for example, if Parliament enacts legislation to give effect to a constitutional right (‘effect-giving statute’), a claimant seeking to protect or enforce that right must rely on the legislation and is precluded from bringing a claim directly on the Constitution. However SANDU preclusion does not apply when a claimant challenges the constitutionality of the statute.36 Therefore, in order to apply SANDU, a court must be able to distinguish rights- enforcement claims from constitutional challenges.37 Drawing that distinction will be easy in many cases, but not in others. Important cases fall into middle-ground, creating difficulties for the theory that limit its usefulness.

A claimant eligible under an effect-giving statute for the relief sought will sue on the statute. She will sue on the Constitution or for Constitution-based remedies only if she cannot obtain what she wants on the statute. It is reasonable to assume, therefore, that after

36 See Van der Walt (n 4 above) 100 - 103, discussing South African National Defence Union v Minister of Defence (SANDU) 2007 5 SA 400 (CC), 2007 8 BCLR 863 (CC) paras 51 - 52. 37 Obviously the claimant’s own description of her claim cannot be determinative. If it were, the SANDU principle could be nullified by strategic pleading. (2008) 1 Constitutional Court Review 139

SANDU, whenever a claimant is ineligible for relief on an effect-giving statute, she will pursue her claim by alleging that the enforcement provisions of the statutory scheme are constitutionally inadequate. If the statute is constitutionally adequate as applied, and the claimant is remitted to her statutory remedies, she is out of luck. If the statute is not constitutionally adequate, the claimant may proceed on the Constitution, and the court must hear her case. Thus, every non- frivolous case in which the effect-giving statute denies relief, that is, virtually every serious case likely to arise post-SANDU, raises a preclusion-avoiding, constitutional question. In all cases seeking constitutional relief beyond that provided in an effect-giving statute, the courts must make a pre-threshold determination as to whether the plaintiff has a legitimate claim of constitutional inadequacy before it can make the supposedly threshold, subsidiarity-prescribed determination whether the case should be decided on the statute alone.

This renders the SANDU principle a nullity, unless courts are able to make the pre-threshold determination on the face of the pleadings. But how is a court to know whether it is in the presence of a bona fide constitutional question, short of airing and taking a view on the matter? Inevitably, the court must consider the full panoply of relevant constitutional values, including separation-of-powers values. In addition to substantive constitutional concerns, the court must inquire whether the question of remedies is one peculiarly within the competence of the legislature, or whether the principle of constitutional supremacy authorises or obliges the court to consult its own views of the matter. By itself, the SANDU principle provides little guidance in determining whether a valid claim on the Constitution has been pleaded; that determination must be made before SANDU kicks in.

In sum, unless a claim is patently frivolous, a court cannot answer the pre-threshold question — whether the remedies provided by an effect-giving statute are constitutionally adequate — without fully considering and reaching at least a tentative conclusion on that point. But the purpose of subsidiarity was to prevent claimants from precipitating full-dress adjudication of constitutional issues when the legislature has given effect to a constitutional right. The process of applying subsidiarity principles reinstates the problem it is intended to solve.

This dilemma, if it is one, derives ultimately from an ambiguity in the constitutional provisions that authorise or enjoin Parliament to 140 Reply to vd Walt

‘give effect’ to an enumerated constitutional right.38 The bare words of the text can bear at least two polar meanings and perhaps a range of others in between. To ‘give effect’ to a constitutional right might mean that the right is free-standing, with a content of its own, but that Parliament is invited or mandated to implement and give concrete, practical significance to the right. In this interpretation, the content of the right is not exhausted by the effect-giving statute. Alternatively, and in contrast to textual references to the ‘progressive realisation’ of constitutional rights through legislation,39 the phrase ‘give effect’ might mean that the right has the meaning and effects that Parliament gives it. The case for the former interpretation seems overwhelming to me, but, followed to its logical end-point (a course which van der Walt does not advocate), subsidiarity pushes the law toward the latter approach.

I should pause to say that, in my view, constitutional rights are in part constituted by the extent and scope of relief awarded for their violation. Generations ago, the Legal Realists argued that remedies define legal rights and entitlements, rather than the other way around. On the traditional view, courts fashion remedies based on their understanding of the nature of the legally protected interests infringed and the character of the wrong to those interests. According to the Legal Realists, we learn the identity of the legally protected interests and the nature of wrongs by examining the remedies courts dispense. A wrongful injury to a legally protected interest is that for which and to the extent that the courts provide a remedy.40 One need not press the point to an extreme to accept that to some significant extent, enforcement and remedies determine what a right means in practical effect in the lives of the parties. When Parliament ‘gives effect’ to a constitutional right, it may task itself with giving the right an enforceable floor of protections and implementations. In practice, it may also erect a ceiling and walls around the right. At a certain point, ‘giving effect’ to a constitutional right slides into defining the right by setting out its metes and bounds. The question we deal with here is whether and to what extent the courts are confined within the houses that Parliament builds.

The constitutional adequacy of the relief afforded by an effect- giving statute is not, strictly speaking, a question subsidiarity theory addresses — it is a substantive problem of constitutional law that must be decided by courts. However, the issue of constitutional adequacy

38 See, eg, FC sec 33(3) (just administrative action) (‘[n]ational legislation must be enacted to give effect to these rights’). 39 See, eg, FC secs 26(2) (housing) and 27 (health care, food, water & social security). 40 See, eg, WH Hamilton & I Till Property (1933) 12 Encyclopedia of Social Science 528 536 (‘[i]t is incorrect to say that the judiciary protect[s] property; rather they call[ ] property that to which they accord[ ] protection’). (2008) 1 Constitutional Court Review 141 does implicate core subsidiarity concerns in the sense that how one approaches it may be strongly influenced by one’s attitude regarding the appropriate roles of courts and legislatures. Too eager a judicial willingness to find constitutional fault with effect-giving statutes might deprive the legislature of the discretion to which it is properly entitled in a democratic society. But an overly cautious attitude, comfortable with directing most claimants to their statutory remedies, might abdicate the courts’ obligation to protect and promote the Bill of Rights.

Particularly in the field of socio-economic rights, when Parliament enacts an effect-giving statute providing a range of benefits, rights, and entitlements to qualified applicants and imposing corresponding obligations and prohibitions on the government, it will trade off limitations on the coverage, administration, and enforcement of the right in return for greater generosity of benefits and ease of access to them. Such tradeoffs are entirely legitimate means to make economical use of scarce resources needed to fulfil other constitutional rights and to provide for orderly conduct of executive and judicial business. Legislatures are supposed to make such tradeoffs, and the Constitution requires South African courts to respect the considered judgment of Parliament.41 But must courts give it dispositive weight on practical questions of enforcement and remedies? Leave aside crudely under-protective and otherwise patently inadequate remedial schemes. Assume the legislature’s scheme falls within a range of reasonable decision making. In a democracy, are there any circumstances under which a court may properly substitute its own thinking for that of the legislature? What if a court concludes that a different choice of remedies would do a significantly better job of protecting the constitutional right in question, with a net gain to democracy? Is it nevertheless precluded by the legislative judgment (except in extreme cases)?

Assume that in a statute giving effect to a constitutional right, Parliament provides relief for denial or infringement of the right by lawsuit, and that the statute specifies a four-year time-limitation period for instituting such proceedings. X sues on the statute against an appropriate defendant. Her claim is meritorious in all respects save that she waited five years to bring the claim. She was tardy for a combination of reasons: she was not well-schooled in her rights, she faced practical and financial difficulties in obtaining representation, and she pursued the claim with insufficient vigour and attention. X pleads her cause both on the statute and on the Constitution itself. She argues that the four-year cut-off is unconstitutional, and that the court should either grant Constitution-based remedies or, alter-

41 See FC sec 41(1)(e). 142 Reply to vd Walt natively, read a longer limitations-period into the statute and then grant relief on the statute so revised.

If Parliament had specified a 24-hour limitations period, most everyone would agree that the statute is unconstitutional to that extent—both for failing to give meaningful effect to the substantive right and in terms of the section 34 right of access to courts.42 Assume for purposes of discussion, however, that the Constitutional Court had previously ruled that Parliament does not offend the Constitution merely by imposing some reasonable limitations-periods in effect- giving statutes, the constitutionally acceptable minimum length to be determined in light of the nature of the right, the nature of potential violations, and any other relevant considerations. If the court in X’s case were to conclude that a four-year prescription period plainly, unjustifiably, and unreasonably limits enjoyment of the constitutional right, presumably it would be common cause that the court may properly, indeed must, hold the statute invalid to that extent. Suppose instead that the court concludes (1) a four-year time-frame is within the range that could reasonably be regarded as appropriate and constitutional, but nevertheless (2) a seven-year period would better safeguard the right in question. The court arrives at this conclusion because it deems the substantive right to be very fundamental and in need of generous protection; because of its concern about the practical barriers to rights-enforcement faced by millions of poor South Africans; and because it feels the more forgiving limitations period better comports with ‘the values that underlie an open and democratic society based on human dignity, equality, and freedom’.43 May the court impose its preference for seven years, thereby trumping Parliament’s choice among a range of reasonably acceptable approaches to time-limitation? Are courts, or at least the Constitutional Court, charged with the authority to optimise protection of fundamental rights? Or would such a decision violate a proper understanding of separation-of-powers?

42 In Moise v Transitional Local Council of Greater Germiston 2001 4 SA 491 (CC), 2001 8 BCLR 765 (CC), the CC confirmed an order of the Witwatersrand High Court finding unconstitutional and invalid sec 2(1)(a) of the Limitation of Legal Proceedings (Provincial and Local Authorities) Act 94 of 1970, which barred suits against local authorities unless the claimant gave written notice of a claim within 90 days of the date on which it arose. (Plaintiffs had 24 months within which to file suit.) The short notice period was held to impair the FC sec 34 right of access to courts in a manner that was not reasonably justifiable within FC sec 36(1). In Engelbrecht v Road Accident Fund 2007 6 SA 96 (CC), 2007 5 BCLR 457 (CC), the Court held unconstitutional regulation 2(1)(c), GN 17939, 25 April 1997, promulgated in terms of the Road Accident Fund 56 of 1996. With some qualifications, the regulation barred compensation to a victim of a hit-and-run accident unless the claimand filed a police affidavit within 14 days of the accident. 43 FC sec 39(1)(a) (concerning interpretation of the Bill of Rights). (2008) 1 Constitutional Court Review 143

Subsidiarity demands that the court dismiss X’s case even though it is in part styled as a constitutional challenge. If we rule out ultra- short and patently unfair limitations periods, setting the precise time- limit on claims is quintessentially a question of ‘policy’ as to which courts should refrain from substituting their own judgment for Parliament’s. This is so even if the legislature, in reaching its decision, predictably and properly took into account such nonconstitutional factors as the government’s need for efficient and frugal management of its legal and material resources, its desire to remove impediments to economic development, and/or its unadorned interest in avoiding litigation. It is surely defensible to argue that courts have no business critically scrutinising, let alone second- guessing, Parliament. It is worth noting that there is nothing particularly ‘transformative’ about this position. Most jurists who administer democratic constitutions providing for judicial review would probably counsel deference in this situation. Is that the best approach for South Africa or one that the Constitution obliges South African courts to take?

It is not obvious, at least to me, that the familiar separation-of- powers considerations supporting judicial deference are equally compelling when invoked in a case the outcome of which will define the operative bounds of a constitutional right. The legislature may have superior institutional competence to make what are conventionally called ‘policy decisions’, but it is not necessarily better situated than the Constitutional Court to strike the appropriate balance between majoritarian interest and constitutional principle. Is it not paradoxical to entrench judicially enforceable rights in a supreme constitution — a constitution meant to constrain the legislature — and then leave it to the legislature effectively to determine how those rights are to be protected and enforced, subject only to highly deferential judicial review? Given the paramount importance of constitutional supremacy in South Africa’s evolving constitutional understanding, the courts might legitimately be entitled to greater latitude in cases involving the effective definition and enforcement of constitutional rights than they may rightly claim in reviewing social and economic legislation that does not derogate from important rights and freedoms. In a transformative context, a bit more active role for the courts (or at least the Constitutional Court) and a bit more back-and-forth between the branches might be in order.44

44 Courts — particularly the peak level courts (CC and SCA) — could evolve practices designed to reduce the institutional tension between the branches and establish a more ‘dialogic’ relationship between them. A reviewing court could, for example, announce its conclusions but withhold judgment, pending further representations from the government or Parliament itself. A court could declare the four-year 44 time limitation constitutionally deficient, but suspend the effective date of its order to give the legislature time to reconsider and, perhaps, devise an 144 Reply to vd Walt

If the reader will grant that these questions are at least worthy of debate, subsidiarity provides us no new tools with which to approach them. The theory instructs courts to leave to legislation (or to the common law, developing under the mandate of section 39(2)45) cases in which it is appropriate in light of separation-of-powers theory to do so, but not in cases where that course is inappropriate. But the theory lacks the analytical traction necessary to perform any real work in assigning actual claims to their appropriate category. Any application of subsidiarity in a borderline case pushes the decision maker back to the general debate about separation-of-powers. Subsidiarity recites some of the countervailing considerations of which account ought to be taken, but it adds no new insight as to how they should be balanced.

Likely van der Walt would say that this is exactly his point — there are no easy answers to such questions. Precisely what he wants is a general acknowledgment that resolving such conundrums calls for ‘politics’ — discussion, debate, dissent, and the taking of judgments about how best to serve competing imperatives (‘for the time being’). Discuss, debate, and balance; decide provisionally; hold tensions open — I could not agree more. But van der Walt told us that before he brought subsidiarity into the picture. Subsidiarity cannot help us in pursuing the discussion.

The reader may doubt that anything of constitutional substance can turn on the length of a limitations period. If so, she will find sobering a US case in which the Supreme Court sanctioned a man’s execution without affording a hearing on his constitutional claims because his lawyer was three days late in filing a notice of appeal in a lower court.46 Nevertheless, I turn to examples in which the substantive constitutional stakes are more apparent.

intermediate solution or approach the problem from a different angle. A court could invite, ask, or order Parliament to reconsider, standing ready to defer to the legislature or review its work with less exacting scrutiny if, upon reconsideration, it reaffirms the choice of a four-year limitations period. It could invite, ask, or order Parliament to propose a new solution within stipulated guidelines. The invention of intermediate review practices between the extremes of deference and trumping is one of the great challenges to constitutional law in our time. 45 FC 39(2) (development of the common law and customary law ‘must promote the spirit, purport and objects of the Bill of Rights’). 46 Coleman v Thompson 501 US 722 (1991) (O’Connor J). Coleman was convicted of crimes and sentenced to death. He contended that his trial was infected by violations of his constitutional rights. The Court held that the attorney’s late- filing was a procedural default rendering Coleman’s conviction final under of the state law of Virginia, and therefore unreviewable in the US Supreme Court. Coleman was condemned by the equivalent of a ‘subsidiarity rule’ in the federalism context. The rule was based on two legal foundations — limitations on federal jurisdiction imposed by Article III of the US Constitution, and principles of federal/state comity. (The ‘procedural default’ doctrine has exceptions, but the Court did not find them applicable in Coleman’s case.) (2008) 1 Constitutional Court Review 145

Imagine that in response to Carmichele47 and its progeny, Parliament enacted a comprehensive programme (the ‘domestic violence programme’ or DVP) intended to give effect to women’s equality rights (FC section 9), children’s rights (FC section 28), the right to security of the person (FC section 12), and the obligation of government to protect against violence (FC section 12(1)(c) read with FC section 7(2)) in the context of violence against women and other domestic violence. DVP commits the national government to massive expenditures for increased policing, a well-designed scheme of initiatives for police and other officials to receive training on violence against women and children, a victim advocate programme in the courts and in the community, and secure shelters and temporary financial assistance to victims and potential victims. DVP also provides scheduled compensation, obtainable through simple and prompt administrative procedures, for injuries caused by negligent acts or omissions of the police, court authorities, and other public officials. In response to concerns about costs, the statute declares the specified administrative compensation to be the exclusive remedy in such cases, and moreover it grants the government and police absolute immunity from suit for negligence in any case involving violence against women or children. Assume that all interested groups in civil society and leading experts from numerous disciplines were consulted and had substantial input into the design of DVP. Many objections were raised in the consultation process, but in the end a wide consensus emerged that DVP is a state-of-the-art programme promising great benefits to vulnerable or victimised persons and to the community at large. The vote in Parliament in favour of DVP was unanimous.

H attacks and shoots his wife W with an illegally possessed firearm, so seriously injuring and disabling her that she will need care- taking assistance, social work, and other services for life. Her young children witnessed the attack and suffered deep emotional trauma; they will require specialised services for the foreseeable future. H’s attack occurs under circumstances reminiscent of Carmichele and related cases — H had a history of violence known to local authorities; the police failed, for no legitimate reason, to impound H’s guns; and no warning or protection was given to W after police officers observed H intoxicated and threatening to kill her. H is arrested shortly after the attack and is eventually sentenced to a long prison term.

W claims administrative compensation under DVP, to which she is unquestionably entitled. However, the DVP benefits do not provide redress for the emotional injury to her or the children, and they do

47 Carmichele v Minister of Safety and Security (Centre for Applied Legal Studies Intervening) 2001 4 SA 938 (CC), 2001 10 BCLR 995 (CC). 146 Reply to vd Walt not come close to the estimated life-time cost of her and the children’s caring needs. W therefore sues the local police and the Minister of Safety and Security for unlawful omission to protect her. W brings the claim in delict, urging that, if necessary, the common law be developed so as to promote the spirit and objects of the Bill of Rights. Alternatively, she sues on the Constitution itself seeking Constitution-based damages. The defence concedes that the police were guilty of wrongful omissions, that these omissions proximately caused W’s injuries, and that the police’s omissions violated common law and constitutional duties owed to W. Nevertheless the police and Minister seek dismissal of both branches of W’s claim based on the DVP immunity provision. W in turn concedes that the defendants are immune in terms of DVP, but she argues that in application the statutory scheme is unconstitutional to that extent.

May or must a court hear either her delictual and/or her constitutional claim? Should the court remit her to her remedies under DVP (with the result that a large portion of her injuries will go uncompensated and her constitutional challenge will go unheard)? At first glance, the Bato Star principle48 suggests that W’s common law claim is precluded, but she escapes that tenet because she has challenged the constitutionality of DVP. If bypass is so easy, can subsidiarity effectively constrain courts from second-guessing the legislature and from undermining its carefully engineered policy compromises? If the court hears W’s claim on the Constitution, must it defer to the legislature’s policy judgment, and if so, how much deference must it show? On the other hand, if the court may not independently examine the constitutionality of the immunity provision — if, for example, subsidiarity directs the court to abstain from hearing W’s claim on the Constitution — have we not vested the legislature with effective power to define and limit the constitutional rights at stake, in violation of the fundamental principle of constitutional supremacy?

I do not mean to suggest that there are easy or clear answers to these questions. My point is simply that subsidiarity brings us to their threshold, and then its analytic power runs out. It leaves us without moorings in attempting to figure out the appropriate roles and inter- relationships of legislation and judge-made law.

48 See Van der Walt (n 4 above) 103 - 105 discussing Bato Star Fishing (Pty) Ltd v Minister of Environmental Affairs 2004 4 SA 490 (CC), 2004 7 BCLR 687 (CC) para 25 and Chirwa v Transnet Ltd 2008 2 SA 24 (CC), 2008 3 BCLR 251 (CC) para 23. According to Van der Walt (103), these cases state the principle that ‘once legislation has been enacted to give effect to a right in the Constitution, and in so far as the legislation was intended to codify the common law, litigants may not rely directly on the common law when seeking to protect that right against infringement’ (italics in original). (2008) 1 Constitutional Court Review 147

3.4 A cautionary tale — of Schweiker v Chilicky49

In this case, the US Supreme Court denied claimants the right to pursue constitutional claims because Congress had enacted an administrative scheme to redress their grievances. At first glance, Chilicky is SANDU with an American accent. But subsidiarity thinking runs wild in the Court’s opinion, which would make Kafka proud. To be clear, I do not believe that subsidiarity theory entails the Chilicky result, and I am confident that van der Walt would regard the case as wrongly decided. I offer it as a cautionary tale for South Africans — to show how much damage can be done when jurists who do not possess van der Walt’s sophistication and intellectual integrity use subsidiarity-talk to resolve a case.

Comparative discussion of constitutional torts requires caution because the US system is shaped by several factors that do not play a role South Africa. The US states are conceived as semi-autonomous sovereigns with a much more significant and independent legal identity than is associated with South African provinces. US federal courts are courts of constitutionally limited, not general, jurisdiction. US law seems to be more fastidious in distinguishing between rights, sources of law, and causes-of-action for redress of violations. Notably, the US Constitution does not contain a general ‘standing’ provision such as found in FC section 38,50 and the case-law is generally more restrictive than what FC section 38 permits.51 The key point for present purposes is that the US Constitution does not expressly create any causes-of-action under which individuals may seek redress for constitutional violations. Congress long ago enacted statutes expressly creating causes-of-action against persons acting under the colour of state law who deprive individuals of federal constitutional rights.52 There is no comparably generic cause-of- action for violation of constitutional rights by the federal government or others acting under the colour of federal law. As a historical matter, it was generally assumed, but not entirely certain, that an individual could sue a federal actor for redress of constitutional rights

49 Schweiker v Chilicky 487 US 412 (1988) (‘Chilicky’) (O’Connor J). Brennan J, joined by Marshall and Blackmun JJ, dissented. 50 FC sec 38 (enforcement of rights) both lists the parties who may approach a court when a right in the Bill of Rights ‘has been infringed or threatened’ and gives them a constitutional right to do so. 51 The picture is further complicated because South African constitutional law has numerous distinctive features missing or largely absent in the US. These include affirmative governmental duties, social and economic rights, and direct horizontal application. The latter exists in the US — eg, the Thirteenth Amendment prohibition of slavery has direct horizontal application — but US constitutional jurisprudence is dominated by a powerful public/private distinction. 52 The most salient of these is the provision of the Civil Rights Act of 1871 now codified at 42 USC § 1983. 148 Reply to vd Walt violations. In the modern Bivens case,53 the Court held that the Constitution impliedly gives rise of its own force to a cause-of-action in favour of an individual deprived of federal constitutional rights by a person acting under colour of federal law.

Bivens remains good law, but the Court has narrowed its scope in recent years. In my view, the Court moved in this direction partly because of hostility to the Bill of Rights and an unarticulated antipathy to some of the kinds of people (eg, poor people) that attempt to bring Bivens claims. The Court’s stated reasons are based on subsidiarity principles. Specifically, the Court has held that a claimant may not pursue a Bivens tort where Congress has enacted an enforcement scheme providing meaningful remedies for unconstitutional action. In a leading case, Bush v Lucas,54 a federal government employee claimed that his employer had violated his constitutional right of freedom of speech.55 He was precluded from bringing a Bivens action and was instead confined to the administrative relief available under statutes governing the employment of federal civil servants, even though he alleged that these remedies were inadequate to deter unconstitutional wrongdoing and did not allow compensation for certain of his injuries (eg, emotional harms).56

The Bush Court assumed that the civil service procedure was constitutionally adequate.57 One might think that, had the Court treated Bush’s claim as a direct challenge to the constitutional sufficiency of the administrative remedies,58 it would have permitted his Bivens claim to go forward. This would bring the US approach into

53 Bivens v Six Unknown Named Agents of Federal Bureau of Narcotics 403 US 388 (1971). Actions pursuant to the rule of this case are called ‘Bivens torts’. 54 462 US 367 (1983). 55 See US Constitution, 1st amend. 56 Mr Bush is not related to either of the former Presidents Bush. The Supreme Court affirmed that federal courts have the power to establish remedies needed and appropriate to vindicate constitutional rights, even if not expressly authorised by Congress (Bush (n 54 above) 374). However, it declined to exercise that power in Bush’s case because, after careful consideration, Congress had struck a balance between ‘the conflicting interests involved in providing job security, protecting the right to speak freely, and maintaining discipline and efficiency in the federal work force’ (385). Congress had developed ‘an elaborate, comprehensive scheme ... by which improper [personnel] action may be redressed’, (385), and which ‘provides meaningful remedies for employees who may have been unfairly disciplined for making critical comments about their agencies’ (386) (footnote omitted). The Court saw as the central question in the case whether the remedial system established by Congress ‘should be augmented by the creation of a new judicial remedy for the constitutional violation at issue’ (388). That question requires a ‘policy judgment’, including a cost-benefit analysis of the value of additional remedies and their possible impact on the efficiency of the civil service, a policy judgment that Congress is in a far better position than the Court to make (388 - 389). 57 Bush (n 54 above) 378 n 14. 58 Bush’s pleadings were unclear on this point, although on a generous reading his suit did advance a constitutional challenge to the administrative scheme. (2008) 1 Constitutional Court Review 149 conformity with SANDU. If one thought that, one would be mistaken. Chilicky, to which I now turn, disallows a Bivens tort even where the claimant raises an explicit constitutional challenge to the statute (as applied), and even though the alternative, statutory scheme to which she is remitted excludes recovery for unconstitutional conduct.

The Federal Government (in cooperation with the states) operates several programs dispensing monetary benefits to persons who suffer specified forms of income interruption (due, for example, to unemployment, retirement, spouse’s death). Chilicky concerned a benefits programme for certain categories of people unable to perform paid work due to disability. Many recipients rely entirely on these benefits for their livelihood and that of their families. At the time, eligibility for disability (or other welfare) benefits was in many cases a requirement of access to Medicaid, the programme of healthcare coverage for poor people.

Former President Ronald Reagan’s Administration was hostile to all forms of welfare and in the early 1980s imposed directives and informal pressures on programme managers designed to get them drastically to reduce the number of social welfare recipients. As a result, the disability-benefit programme alone terminated hundreds of thousands of benefit-recipients over a period of years using questionable, and often sleazy, methods. The Chilicky claimants argued that the procedures the Government used to clear the welfare rolls violated the right to Due Process guaranteed by the Bill of Rights.59 Congress eventually brought a halt to the wholesale terminations and introduced procedural safeguards. By then, approximately 200,000 recipients had been wrongfully terminated, by the Government’s own reckoning. Many were eventually reinstated, but the episode caused heart-breaking and traumatic suffering. Many thousands of recipients and their families lost their livelihood, became homeless, and/or scraped by on the kindness of friends and relatives for long periods. Regrettably, South Africa is also familiar with large-scale delivery-failures in social welfare programmes.

The disability-benefits programme was enacted under the Social Security Act,60 which incorporates an administrative appeals procedure for correcting errors. Terminated recipients who were in fact eligible invoked these procedures. They were entitled to, and most eventually received, reinstatement for future benefits and

59 See US Constitution, 5th amend. (‘No person shall ... be deprived of life, liberty, or property [by the Federal Government], without due process of law’). The Fourteenth Amendment guarantees Due Process in deprivations undertaken by state governments. 60 Social Security Act, Title II (1935), principally codified at 42 USC §§ 420 - 425. 150 Reply to vd Walt retroactive payment of past benefits wrongfully withheld. The problem was that these were the only remedies available under the administrative system. In other words, a fully successful claimant received at the end of the administrative process only what she was entitled to in the first place. The procedure made no provision for mandatory or injunctive relief, it afforded no compensation for the turmoil and dislocation caused by wrongful termination, and it did not entertain, let alone remedy, constitutional claims.

The claimants in Chilicky were eligible but wrongfully terminated benefits-recipients who sought monetary damages for ‘emotional distress and for loss of food, shelter and other necessities proximately caused by [the Government’s] denial of benefits without due process’.61 Consequential damages and the other remedies requested were not available through the administrative appeals system. Lacking any other vehicle, plaintiffs lodged their claim as a Bivens tort.

Chilicky differs from the cases van der Walt discusses in that the Social Security Act was not enacted in order to ‘give effect’ to a constitutional right. The US Constitution contains no guarantee of social security such as appears in FC section 27(1)(c). This should not greatly alter the analysis of Chilicky, however, because the claim was grounded on a constitutional right, the right to Due Process. Indisputably, the remedial scheme attached to the disability benefits programme must provide constitutional Due Process — in that sense, it is a statute designed in relevant part to ‘give effect’ to a constitutional right.

In a 1970s case called Arnett, Justice Rehnquist floated a theory that recipients of government largesse or public employment must ‘take the bitter with the sweet’.62 By this he meant that any nonconstitutional benefit created or extended by law is qualified and limited by the procedures specified in the legislation for removal or discontinuance of the benefit, even if such procedures do not measure up to the minimum requirements of constitutional Due Process. Rehnquist J reasoned that no deprivation of property occurs if a benefit is terminated or removed according to the procedures for removal specified in the law creating the benefit. In these circumstances, the claimant is deprived of nothing to which she has a

61 Chilicky v Schweiker 796 F 2d 1131 1134 (9th Cir. 1986) (quoting plaintiffs’ complaint), reversed by Schweiker v Chilicky 487 US 412 (1988). Other remedies were sought, notably an interdict against the benefit cut-offs. The additional remedial requests dropped out of the case partly because it meandered through the courts for about six years, by which time Congress had halted the Reagan campaign to slash the rolls and had provided enhanced procedural safeguards. 62 See Arnett v Kennedy 416 US 134 151 - 54 (Rehnquist J for the plurality), rehearing denied 417 US 977 (1974). (2008) 1 Constitutional Court Review 151 legal claim of entitlement. Therefore, Rehnquist J argued, the strictures of Due Process are inapplicable.63

A majority of the Court emphatically rejected Justice Rehnquist’s thesis, both in the Arnett decision itself and on subsequent occasions.64 Under settled doctrine, even though the government is not constitutionally obliged to confer a certain benefit, once it elects to do so, it may not remove or terminate the benefit except by procedures consistent with constitutional Due Process. The Constitution, not the statute, sets the procedural floor.

Evidently the Court forgot this when it decided Chilicky, which shows legislatures how to immunise their work from constitutional review. Therein lies a warning for South Africans. The Chilicky Court held that ‘[w]hen the design of a Government program suggests that Congress has provided what it considers adequate remedial mechanisms for constitutional violations that may occur in the course of its administration, we have not created additional Bivens remedies’.65 In the result, the Court precluded the plaintiffs from proceeding with an action on the Constitution. Under the banner of subsidiarity, they were remitted to statutory remedies which, it was common cause, did not even consider, let alone remedy, claims of constitutional injury. Under American-style subsidiarity, constitutional claimants were channelled to and compelled exclusively to utilise a forum that did not entertain constitutional claims. This meant that the claimants could not obtain a hearing for their constitutional claims in any forum. In practical effect, the legislation simply erased their constitutional right to Due Process.

One might have thought that a self-respecting court would be embarrassed to decree something so absurd. But no, the US Supreme Court made its conclusion official: ‘the relief sought by the [claimaints] is unavailable as a matter of law ... ’.66 In sum, when enacting a benefits programme, the legislature may render the Bill of Rights inoperative by attaching remedial procedures that claimants are exclusively constrained to utilise when challenging the

63 See Arnett (n 62 above) (under constitutional text, a ‘deprivation’ is required to trigger Due Process guarantees). 64 See, eg, Cleveland Board of Education v Loudermill 470 US 532 (1985). 65 Chilicky (n 49 above) 423. The Court could find no legal distinction between Chilicky and Bush: ‘[W]e declined in Bush to create a new substantive legal liability [on the Constitution] ... because we [were] convinced that Congress is in a better position [than the judiciary] to decide whether or not the public interest would be served by creating it. That reasoning applies as much, or more, in this case ... In neither case ... does the presence of alleged unconstitutional conduct that is not separately remedied under the statutory scheme imply that the statute has provided ‘no remedy’ for the constitutional wrong at issue (426 – 428) (citations and internal quotation marks omitted; italics in original). 66 Chilicky (n 49 above) 429. 152 Reply to vd Walt constitutionality of the programme’s administration, even though those procedures expressly disallow constitutional challenges of any kind. Chilicky allows Congress to stamp ‘void’ on the Bill of Rights.67

The Court’s deference to the statutory remedies fully accords with subsidiarity thinking. True, the Court’s articulated rationale invoked the theory of institutional competence (the legislature is better situated than the courts to study complicated policy matters and to engineer comprehensive solutions) rather than classical separation-of-powers doctrine (the legislature is representative and acts democratically). The distinction might be highly significant in some contexts, but here the two theories seem closely intertwined. Subsidiarity draws on both. Substitute ‘Parliament’ for ‘Congress’, and ‘government agencies’ for ‘state agencies’, and one can easily imagine a South African court committed to the SANDU principle denying an action on the Constitution with words similar to the following:

Congress ... has addressed the problems created by state agencies’ wrongful termination of disability benefits. Whether or not we believe that its response was the best response, Congress is the body charged with making the inevitable compromises required by the design of a massive and complex welfare benefits program ... Congress has discharged that responsibility to the extent that it affects the case before us, and we see no legal basis that would allow us to revise its decision.68

I am confident that van der Walt would say that Chilicky goes too far. His version of subsidiarity is more flexible than its US counterpart and does not rule out constitutional challenges to effect-giving statutes. But these subtleties can easily be lost in less able hands than his.

67 Chilicky is in a line of disastrous Nixon-Reagan era decisions that eviscerated the Due Process clauses. In Bishop v Wood 426 US 341 (1976), the Court held that a public employer may discharge an employee for cause without providing a hearing, an opportunity to respond to the charges against him, or any other minimum component of constitutional Due Process, if the government structures its relationship with its employees on an at-will basis. In Paul v Davis 424 US 693 (1976), a local police department distributed circulars to merchants containing the names and photographs of individuals identified by the police as known to be actively engaged in shoplifting. The names were selected ex parte, without notice to the individuals involved, hearing, or any other opportunity for the accused to respond. The Supreme Court found no constitutional deficiency in this procedure, holding that those who were ‘merely’ publicly branded as criminals without trial suffered no impairment of any interest protected by the Due Process clauses. Justice Brennan remarked in dissent that ‘[t]he potential of today’s decision is frightening for a free people’ (721). 68 Chilicky (n 49 above) 429 (citations omitted). The legal basis the Court had difficulty seeing is, of course, the Constitution’s Fifth Amendment and the Court’s own case-law. (2008) 1 Constitutional Court Review 153

5Conclusion

I conclude with a hypothetical problem aimed to identify interpretive questions that await South Africa’s courts. My purpose is to show the limitations of subsidiarity analysis, but also to offer a platform for discussion of issues that may well arise as the jurisprudence of SANDU and Bato Star evolves. In fact, the hypothetical bears more than a passing resemblance to matters that have already concerned South African courts.69

Imagine that Parliament enacts legislation to give effect to the right to social security guaranteed by FC section 27(1)(c). In addition to setting forth the substantive content of the programme, the legislation contains a remedial scheme intended to address and remedy all wrongs committed by the government in connection with the benefits programme, including any constitutional violations that may occur. Successful claimants are entitled to retroactive payment of benefits wrongfully withheld, access to the programme if it was wrongfully denied, and restoration to the programme in the event of wrongful termination. In addition, a claimant may receive monetary compensation for consequential damages, including damages proximately caused by a constitutional violation, up to a maximum of R100. The statutory procedure makes no provision for mandatory relief and does not allow class actions. These restrictions were included after careful study, consultation with experts and affected stakeholders, and plenary discussion in Parliament. The drafters are in good faith convinced that it is necessary to limit litigation and remedies in connection with delivery problems in order to preserve scarce resources for the beneficial purposes of the programme. Parliament declares the statutory remedial procedure to be the exclusive avenue of redress for any grievance in connection with the programme. Neither the Promotion of Administrative Justice Act (PAJA)70 nor any other statute applies, and any claim that might arise under the common law is expressly pre-empted.

69 See, eg, Permanent Secretary, Department of Welfare v Ngxuza 2001 4 SA 1184 (SCA), 2001 10 BCLR 1039 (SCA); MEC, Department of Welfare v Nontembiso Norah Kate 2006 2 All SA 455 (SCA). In Kate, the SCA sustained an award of constitutional damages in a case of excessive delay in the processing of a social grant application, as against the view that the claimant should be restricted to statutory or delictual remedies. See (para 27) stating that the relief permitted by FC sec 38 (enforcement of Bill of Rights): ‘is not a remedy of last resort, to be looked to only where there is no alternative — and indirect — means of asserting and vindicating constitutional rights. While that possibility is a consideration to be borne in mind in determining whether to grant or to withhold a direct section 38 remedy, it is by no means decisive, for there will be cases in which the direct assertion and vindication of constitutional rights is required ... [T]he endemic breach of the rights that are now in issue justifies — indeed, it calls out for — the clear assertion of their independent existence’. 70 Act 3 of 2000. 154 Reply to vd Walt

An eligible recipient of welfare benefits is arbitrarily terminated from the program for no good reason and without notice or an opportunity to be heard. She suffers consequential damages in an amount greater than R100 and asserts that the cap on remedies renders the statutory remedial procedure constitutionally inadequate, citing FC section 27(1)(c) itself and perhaps FC section 33 (just administrative action), FC section 34 (access to courts), and other provisions of the Bill of Rights. May she prosecute a claim on the Constitution for the full amount of her damages? Suppose an additional 100,000 eligible recipients experience similar problems due to the incompetence or malfeasance of the authorities. May they sue on the Constitution as a class to remedy their own injuries and interdict the authorities from future, administrative misbehaviour?

If we conclude that these claimants may not proceed on the Constitution and are limited to working within the statutory procedure, we vindicate the pro-legislature, pro-representation spirit of subsidiarity at the expense of constitutional supremacy and substantive constitutional rights. We put the legislature in a position effectively to define rights entrenched by a Constitution intended to constrain it. And because an interdict is precluded by the statute, we effectively prevent the courts from granting a timeous remedy that might induce the authorities to straighten out the mess. But suppose we conclude that these suits may go forward. Would we not be saying, in practical effect, that we will observe the tenets of subsidiarity except in cases where we prefer not to? And if this is what it comes down to, what has subsidiarity taught us? ‘OH WHAT A TANGLED WEB WE WEAVE ...’* HEGEMONY, FREEDOM OF CONTRACT, GOOD FAITH AND TRANSFORMATION - TOWARDS A POLITICS OF FRIENDSHIP IN THE POLITICS OF CONTRACT

AJ Barnard-Naude**

‘Q6: “ ... and I wondered how you think your concept of friendship, your non-canonical concept of friendship, can address economic exclusion in its most extreme form which is the exclusion by those who own capital of everyone else. You can’t ask those who own capital to be hospitable ...”. JD: “I ask them nevertheless [laughter].” Q6: “It’s naïve to ask them, it’s a naïve request.” JD: “Perhaps, but I still do.”’1 ‘One must speak for a struggle for a new culture, that is, for a new moral life that cannot but be intimately connected to a new intuition of life, until it becomes a new way of feeling and seeing reality.’2

1Introduction

The pivot on which this contribution hinges is hegemony in the context3 of the South African law of contract. I must point out at the

* ‘... when first we practice to deceive.’ Sir Walter Scott Marmion (1825) Canto VI Stanza XVII, 349.  BCom(Law) LLB LLD(Pretoria). Senior lecturer, Department of Private Law, University of Cape Town. The author would like to thank the following people for valuable discussions and comments on earlier drafts of this essay: Karin van Marle, Danie Brand, Drucilla Cornell, Dennis Davis and Karl Klare. 1 From ‘Politics and friendship: A discussion with Jacques Derrida’ available at http://www.hydra.umn.edu/derrida/pol+fr.html (accessed 29 May 2008). 2 A Gramsci Selections from cultural writings (1985) 98. 3 And it will of course be unavoidable when considering the lexicography of hegemony, that the concept of conning (in both its senses) — and the conning of/ by the text; the context — will raise its head.

155 156 Hegemony, freedom of contract, good faith and transformation outset that the purpose or aim of this ‘hinging’ on hegemony and/in the South African law of contract, is not necessarily to take issue with, underscore or re-illustrate the claim that contract holds hegemony — in the sense that it ‘occupies a privileged position’ — in the South African legal system.4 Nevertheless, it might very well be the case that Alfred Cockrell’s admirable illustration of the hegemony of contract in the South African legal system can be explained also by the particular focus on hegemony in the South African law of contract on which I embark in this contribution. In other words, the hegemony of contract might well be attributable to the hegemony within contract. This is the case because the hegemony within might very well sustain the hegemony of and perhaps also vice versa. (A symbiosis comes to mind.) This hegemony within contract relates to a particular understanding of freedom of contract. Thus, the focus and point of departure in this contribution will be the hegemony — or hegemonic order(ing)5 — of/in a particular understanding of (freedom of) contract. This understanding of freedom of contract — and indeed of contract itself — is the individualist, rule-committed6 — liberal7 — understanding of freedom of contract.

The enactment of the Constitution,8 the transformative hopes it disseminates and the view that its normative framework is explicitly post-liberal9 occasions an opportunity for a re-evaluation of and a challenge to the individualist (hegemonic) understanding of freedom of contract, its concomitant commitments and, as we shall see, the tangled web it weaves in order to sustain the false consciousness on

4 A Cockrell ‘The hegemony of contract’ (1998) 115 South African Law Journal 286 286. 5 I want to insist on this marking according to which both senses of the word ‘order’ remain constantly in play, namely order as socio-political concept of hierarchy and order as sequential arrangement. 6 D Kennedy ‘Form and substance in private law adjudication’ (1976) 89 Harvard Law Review 1685 1685: ‘individualism seems to harmonise with an insistence on rigid rules rigidly applied.’ 7 See the description of the liberal self — grounded in individualist morality — offered by F Michelman ‘The subject of liberalism’ (1994) 46 Stanford Law Review 1807 1812. Also see van K Marle & D Brand ‘Enkele opmerkings oor formele geregtigheid, substantiewe oordeel en horisontaliteit in Jooste v Botha’ (2001) 12(3) Stellenbosch Law Review 408 415. 8 Constitution of the Republic of South Africa, 1996. 9 See H Botha ‘Democracy and rights: Constitutional interpretation in a postrealist world’ 2000 (63) Tydskrif vir die Hedendaagse Romeins-Hollandse Reg 561 574. Botha notes three reasons why the Constitution requires more than a classical liberal interpretation. First, he points out that the Constitution contains a commitment to an open, value-orientated, participatory democracy. This is a commitment that cannot be reconciled with the reduced concept of democracy which pervades liberal theory. Secondly, Botha argues that the Constitution does not support a liberal conception of rights as boundaries between the individual and the collective; the rights in the Bill of Rights have a contingent and non- absolute meaning and to that extent they do not operate as a shield against government intervention or as trumps over collective interests. Thirdly, the Constitution is structured such that it requires a far more activist stance of the judiciary than what would be acceptable under a liberal interpretation (574-576). (2008) 1 Constitutional Court Review 157 which its legitimacy turns. However, in order to embark upon such a re-evaluation it will be necessary to begin by revisiting the concept of hegemony as encountered in the work of Antonio Gramsci as well as the expansions and reformulations of Gramsci’s concept as it relates to law. Second, I will articulate more fully what is meant with and what are the commitments of the hegemonic liberal understanding of contract and freedom of contract. Third, I will track the rise of the hegemonic order(ing) in the South African law of contract. I will argue that the story of its rise can and must be considered as the story of the fall (that is, the marginalisation) of values and concepts in the law of contract that came to be understood as somehow threatening the hegemonic order(ing) (notably in this regard, good faith).10 An ongoing moment in this contribution interrogates the way in which the hegemony confronts the Constitution (both as an interruption,11 as a value system and as law) in order continuously to pose the question whether the hegemonic order(ing) within the South African law of contract (and the adjudication that keeps this hegemony in place) should be resisted in the name of the spirit, purport and objects of the Constitution.

A primary contention in this article will be that the hegemony of a radically individualist understanding of freedom of contract is as firmly in place under the newly constituted legal order as it was under the apartheid legal order in South Africa. The law of contract is still seen as an institution of separation rather than relation, of apartness, rather than togetherness. That ‘nothing changed again’12 is evident in a series of cases decided after the reconstitution of the South African legal order.13 What will also become evident, however, is the series of contradictions and deconstructions that emerge in the course of

10 See in this regard G du Toit ‘The significance of postmodern theories of interpretation for contractual interpretation: A critical analysis’ unpublished LLD thesis, University of Stellenbosch, 2006 234-235. 11 See, for instance, J Gilbert-Walsh ‘Deconstruction as narrative interruption’ (2007) 38(4) Interchange 317. 12 For a variety of reasons I want to draw explicit attention here to Karin van Marle’s discussion of the inscription ‘and nothing changed again’ on one of the etchings in Dianne Victor’s celebrated series Disasters of Peace. Van Marle remarks as follows: ‘“And nothing changed again”, portrays the state of mind of many South Africans who do not experience easy living in post-apartheid South Africa. For them, nothing has changed again.’ See K van Marle ‘Art democracy and resistance: A response to Professor Heyns’ (2005) Pulp Fictions: Disasters of Peace: An Exchange 15 26. 13 The most important of these cases are De Beer v Keyser 2002 1 SA 827 (SCA) (De Beer), Brisley v Drotsky 2002 4 SA 1 (SCA) (Brisley), Afrox Healthcare Bpk v Strydom 2002 6 SA 21 (SCA) (Afrox Healthcare), Johannesburg Country Club v Scott and Another 2004 5 SA 511 (SCA) (Johannesburg Country Club), Juglal NO v Shoprite Checkers t/a OK Franchise Division 2004 5 SA 248 (SCA) (Juglal N0), South African Forestry Co Ltd v York Timbers 2005 3 SA 323 (SCA) (South African Forestry Co), Napier v Barkhuizen 2006 4 SA 1 (SCA) (Napier) and Barkhuizen v Napier 2007 5 SA 323 (CC) (Barkhuizen). 158 Hegemony, freedom of contract, good faith and transformation the marginalisation of good faith in the South African law of contract, which of course mirrors the rise of the hegemonic order(ing).

In the last part of the article I will argue why the hegemonic order(ing) of contract, mirrored by the marginalisation of good faith can and should be questioned in light of the transformative commitments inherent in the spirit of the Constitution. I will argue that good faith as the ethical element of contract directly involves the constitutional ideal of civic friendship in the South African law of contract. Civic friendship as an aspirational ideal of the new legal order then enjoins us to transform the hegemonic order(ing) in the law of contract by way of a direct involvement of good faith.14 I see the starting point of this transformation as a call for the politics of friendship in the politics of contract. In this context, I will argue that the provisions of the Draft Consumer Protection Bill stand in stark contrast to the commitments of the hegemonic understanding of freedom of contract and, although inadequate, these provisions point to a future law of contract that will have to be far more concerned with the ethical element of contract than it has been up to now. These provisions require a transformed understanding of freedom of contract in which good faith as the ethical element of contract comes to again play a constitutive role in the understanding of freedom of contract. This has obvious transformative implications for adjudication in the South African law of contract. In this context I will take issue with the Constitutional Court’s 2007 judgement in Barkhuizen in which it followed a hesitant approach to the transformative role of good faith in the South African law of contract. Ultimately my contention will be that good faith interpreted in line with a transformative reading of the law of contract mandates a relational interpretation of freedom of contract that is far removed from its standard libertarian version.

2 Gramsci, hegemony and the law

In order to come to terms with the present focus on the hegemonic understanding of (freedom of) contract, it is necessary — as a point of departure — to review the development of hegemony as a conceptual tool of critique. For the benefit of the argument that will follow, it will also be necessary to enquire into contemporary views on the relationship between hegemony and law.

14 I realise and concede that the theoretical and methodological framework developed here is contested. Apart from the justifications offered in the text, the primary focus of this article (adjudication on the interaction between the common law of contract and the Constitution in 2007) does not allow me to fully defend this approach. In this regard, I intend to embark more fully on such an explanation in future work. (2008) 1 Constitutional Court Review 159

Hegemony as a critical concept was originally developed in the work of Antonio Gramsci15 who employed the word to explain how capitalism maintains control ‘not just through violence and political and economic coercion, but also ideologically, through a culture in which the values of the bourgeoisie became the “common sense” values of all.’16 In Gramsci ideology operates in such a way that the dominant class in fact instills a false sense in the dominated class that the good of the dominant class is also the good of the dominated class. This condition serves conformity and perpetuates a status quo.17 John and Jean Comaroff aptly describe the cultural effect of hegemony:

[W]e take hegemony to refer to that order of signs and practices, relations and distinctions, images and epistemologies — drawn from a historically situated cultural field — that have come to be taken-for- granted as the natural and received shape of the world and everything that inhabits it.18

Gramsci insisted that hegemony is only transformed once the dominated class realised the importance of creating its own culture which depends on the realisation that it had unwittingly adopted the values of the bourgeoisie and that its own values in fact opposed the values of the bourgeoisie.19 Breaking hegemony thus depends fundamentally on the contestation of values and ideals and perhaps more importantly on the contestation of the relationships between and the content of these values and ideals.20

When it comes to the relationship between law and hegemony, Litowitz points out that law fulfils a dual function in relation to hegemony.21 On the one hand, the law represses (through the State’s monopoly on physical force) any disturbance that might challenge the hegemony. On the other hand, the law authorises and legitimates the status quo and so produces and perpetuates the hegemony without the need to revert to physical force.22 This contribution will focus primarily on this second dimension of hegemony, as I shall interrogate

15 A Gramsci Selections from the prison notebooks trans Q Hoare & GN Smith (1971). 16 JM Russel Philosophical classics (2007) 94. 17 As above. Also see TR Bates ‘Gramsci and the theory of hegemony’ (1975) 36(2) Journal of the History of Ideas 351 352: ‘The concept of hegemony is really a very simple one. It means political leadership based on the consent of the led, a consent which is secured by the diffusion and popularisation of the world view of the ruling class.’ 18 J Comaroff & J Comaroff Of revelation and revolution: Christianity, colonialism and consciousness in South Africa (1991) 23. 19 Russell (n 16 above) 95. 20 Bates (n 17 above) makes the point that the theory of hegemony recognises that man is ruled not just by force but also by ideas. The author also points to the Marxist mantra that ‘the ruling ideas of each age have ever been the ideas of its ruling class.’ 21 D Litowitz ‘Gramsci, hegemony and the law’ (2000) Brigham Young University Law Review 515 531. 22 As above. 160 Hegemony, freedom of contract, good faith and transformation specifically how law — a particular version of contract law — endorses and thus normalises or legitimates the individualist or libertarian worldview. For Gramsci as for Litowitz this second dimension of hegemony is the more dangerous one as it carries within it the potential to paralyse all resistance and silence all questioning. Litowitz argues that the concept of hegemony requires the attention of students of law, precisely because it possesses this profound ‘ability to induce submission to a dominant worldview.’23

Whereas Gramsci originally maintained that law is a hegemonic tool of the dominant class, Litowitz argues that when it comes to contemporary legal analysis a more nuanced notion of hegemony is required. Litowitz contends that under conditions of post-modernity we need an understanding of hegemony that avoids ‘Gramsci’s reliance on orthodox Marxist categories that are no longer tenable.’24 Litowitz argues that postmodern theory brought about a profound but gradual shift in the understanding of hegemony away from essentialist Marxist discourse rooted in rigid class distinction.25 Postmodern theorists understood that oppression does not exclusively exist as the imposition of a dominant class and recognised that oppression is often invisible and can exist merely in the silencing of alternative perspectives.26 Postmodernism thus changed the operative terminology in discourse about hegemony from ‘“class/exploitation” to “discourse/marginalisation.”’27

Litowitz explains that this development in hegemony discourse can help legal scholars to recognise that law is hegemonic at the meta-level (in that it always already represents and legitimises dominant interests) while contesting hegemony at the micro-level. For Litowitz law participates in the establishment of hegemony through the dissemination of a ‘dominant code or map’28 that relies on ‘unchallenged background assumptions that undergird the law.’29 Thus law always already is hegemonic in that it ‘induces people to comply with a dominant set of practices and institutions without the threat of physical force’30 because people are lulled into a false consciousness about the legitimacy of law which causes them uncritically to accept legal practices and institutions that might in fact be harmful and exploitative. In other words the meta-level at which law is hegemony sustains and keeps in place law’s hegemonies at micro-level — and vice versa.

23 Litowitz (n 21 above) 516. 24 Litowitz (n 21 above) 518. 25 Litowitz (note 21 above) 533. 26 Litowitz (n 21 above) 534. 27 As above. 28 n 21 above, 517. 29 As above. 30 As above. (2008) 1 Constitutional Court Review 161

Under the force of hegemony the assertion that ‘law ... is also the site of everyday resistance and struggle’31 comes to be either dismissed and ignored or actively undermined and suppressed. However, contesting the hegemonies upon which the law as hegemony is founded can be the first step in raising consciousness upon which refusal and resistance of things as they are and supposedly always have been, can be built.32 This is the central motivation behind the present focus on the hegemonic order(ing) in the South African law of contract. I do not want to suggest that the Marxist roots of hegemony discourse can or should be simply or finally denied or that escape from these roots should be relentlessly pursued.33 As we shall see, the South African law of contract vividly reflects the Gramscian insight that bourgeois best interests come to be uncritically accepted as the best interests of the working classes and the poor through the phenomenon of false consciousness which generates false consent. However, what postmodern theory contributes to hegemony discourse is the notion of discourse marginalisation. I will argue in what follows that the rise of the hegemonic order(ing) in the law of contract occurs through an explicit judicial marginalisation of discourses on equity, fairness and good faith.34 In the context of the reconstitution of the South African legal order this raises important and difficult questions about the potential of the South African law of contract (in its current guise) to contribute to the project of transformative constitutionalism.35

31 K Calavita ‘Book review: M Lazarus-Black et al Contested states: Law, hegemony and resistance’ (1996) 25(1) Contemporary Sociology 94 94-96. 32 Litowitz (note 21 above) 518. Also see D Kennedy ‘Antonio Gramsci and the legal system’ (1982) VI(1) ALSA Forum 32 37: ‘What the Gramscian analysis suggests is that one of the ways out of the reform/revolution problem in all these legal activities is to try to develop, at the level of conscious communication with other people, the extent to which they are letting their goals be perverted by the hegemonic false consciousness generated by law.’ 33 See in this regard E Laclau & C Mouffe Hegemony and socialist strategy (2ed, 2001) viii. 34 A detailed discussion of the link between, on the one hand, the ideas that predominate in judicial decisions in the law of contract and, on the other, the beliefs about social organisation that are dominant in elites and the general populace, falls beyond the scope of this article. But see, for instance, E Mensch ‘Freedom of contract as ideology’ (1981) 33 Stanford Law Review 753 757 who points out that the rise of the classical view of contract in the courts can be attributed to the popularity of Victorian moralism amongst the elites of the time. There are, in addition, several places in this article where I do not use the term ‘hegemony’ in any purist Gramscian sense of the word but rather in its more general sense to denote predominance. Where the word is primarily used in this way there is nevertheless a secondary allusion to the Gramscian notion operating in the background. 35 See K Klare ‘Legal culture and transformative constitutionalism’ (1998) 14 South African Journal on Human Rights 146 146. 162 Hegemony, freedom of contract, good faith and transformation

3 The rise of the hegemonic order(ing) in contract

3.1 The rise of the market economy, the subjectivity of value and the will theory

The present hegemonic order(ing) in contract begins, of course, in the late 18th century. At this point in history, freedom of contract still had an explicit and intricate connection with principles of morality such as fairness, conscience and justice.36 The maxim pacta servanda sunt, in this understanding of freedom of contract, was still regarded as ‘the imperative of good faith in contracts’.37 As Atiyah points out, this meant that: ‘the Courts were, at that time, still more interested in seeing that parties to a contract made a fair exchange, than they were in enforcing bare promises.’38 On this approach, contracts were not enforced meticulously. Instead, they were meticulously subjected to enquiries into the fairness of the exchange.

The problem with this approach to freedom of contract was that by the nineteenth century it became regarded as anachronistic, because it did not serve the interests of the rising commercial classes of the emerging market economies.39 In these emerging economies value became regarded as entirely subjective. Horwitz indicates that, for this very reason, the late 18th century businessman became interested in a freedom of contract that would guarantee the express value of the agreed performance, regardless of the substantive equality of the exchange.40 This meant that, as the market economies and the commercial classes became dominant, the law of contract as it existed up to that point came to be considered unsuited for the purposes of the new economies. In short, the eighteenth century equitable understanding of freedom of contract was far too invested in the concept of objective value. The general view of the eighteenth

36 MJ Horwitz ‘The historical foundations of modern contract law’ (1974) 87(5) Harvard Law Review 917 920. The eighteenth century approach is stated succinctly in Evans v Llewellyn [1787] 29 ER 1191 (Evans) 1191: ‘if the party is in a situation in which he is not a free agent and is not equal to protecting himself, this Court will protect him.’ 37 Van der Merwe et al Contract general principles (3 ed, 2007) 21 n 7 quoting H Eichler Die rechtslere vom vertrauen (1950) 8 (emphasis added). C Visser ‘The principle pacta servanda sunt in Roman and Roman-Dutch law’ (1984) 101 South African Law Journal 641 647 points out that originally pacta servanda sunt is integrally connected to good faith: ‘it would therefore appear that both canon law and the lex mercatoria from very early on accepted a general principle of sanctity of contract, based upon a moral conception of good faith ...’ (emphasis added.) 38 PS Atiyah The rise and fall of freedom of contract (1979) 438. 39 Horwitz (n 36 above) 945-946. 40 MJ Horwitz ‘The triumph of contract’ in AC Hutchinson (ed) Critical legal studies (1989) 104 106. (2008) 1 Constitutional Court Review 163 century was that substantive justice existed to ensure that legal subjects do not abuse the legal system to exploit each other.41 In the nineteenth century the role and function of contract was no longer regarded as guaranteeing the justice of the exchange but to enforce so-called ‘willed’ transactions inter partes.42 So arises the so-called ‘modern’ or ‘classical’ law of contract in which the source of contractual obligation shifted from equitable exchange to consensus.43

The pervasive political view of the nineteenth century — that of unlimited liberal freedom — informed and provided the justification for a conception of ‘freedom’ of contract as the shibboleth for the privileging of the parties’ wills in conceptualising their relationships.44 Once the concept of value became perceived as entirely subjective, the only basis for ascribing value could be concurrent, individual will. Principles of substantive justice came to be regarded as necessarily arbitrary and an uncertain standard of value.45 At the point where intrinsic value could no longer be ascribed to anything, so the argument went, no substantive measure could exist by reference to which it could be determined whether one party was exploiting the other.46 On this view, the appearance of consensus came to represent the ‘evidence’ that the contractual exchange itself was a fair one.

However, the loss incurred as a result of this shift in emphasis was particularly acute. Gordley indicates that nineteenth century jurists in fact entirely eliminated the concept of virtue from their discussion of contract and were left with the will alone.47 The motto of modern contract thus became this: ‘a man is obliged in conscience to perform a contract which he has entered into freely, although it be a hard one ...’.48 As Hahlo so aptly described it: Darwinian survival of the fittest, the law of nature, also became the law of the marketplace.49

41 Mensch (n 34 above) 753, 756. 42 As above. 43 Horwitz (n 36 above) 946. 44 Horwitz (n 36 above) 917. Also see J Gordley The philosophical origins of modern contract doctrine (1991) 161 where the point is made that consensus was applied almost exclusively to explain the basis of contractual obligations. 45 Horwitz (n 36 above) 917. 46 As above. 47 Gordley (n 44 above) 162. 48 IJ Powell ‘Essay upon the law of contracts and agreements’ (1790) as quoted in Horwitz (n 36 above) 917. 49 HR Hahlo ‘Unfair contract terms in civil-law systems’ (1981) 98 South African Law Journal 70. 164 Hegemony, freedom of contract, good faith and transformation

3.2 The role of individualism

The above formulation evinces the understanding of freedom of contract of advancing and advanced liberal capitalism underpinned by what is euphemistically called the ‘morality’ of individualism.50 Individualism accepts as given a world of independent individuals who are encouraged to prefer the pursuit of self-interest rigorously. A consideration or sensitivity for the interests of others falls outside of the aims of this way of life, although one should be prepared to obey the rules that make it possible to co-exist with other self-interested individuals.51 Other than this, the individual is entirely self-reliant. His conduct conforms to the belief that other individuals in the community are themselves motivated only by pure self-interest.52 As regards individualism’s economic manifestation, Cockrell puts it succinctly: ‘[I]ndividualism assumes a world of traders who meet briefly on the market floor, where they engage in discrete and furtive transactions.’53 This view is often described as market-individualism and reveals that the capitalist idea of the market and the ‘moral’ idea of individualism are heavily invested in each other.54

Individualist beliefs regarding the role of law in society follow from the above convictions. Individualism holds that the law cannot impose upon legal subjects as a group the liability of shared profits or loss.55 The law merely fixes the boundaries of individual freedom by defining and enforcing rights.56 In the context of the law of contract, individualism believes that the parties create their own law through agreement (consensus) which is itself a manifestation of the individual’s autonomy. Mensch refers to this phenomenon as a ‘magic moment of formation, when individual wills created a right whose enforcement was necessary for the protection of free will itself.’57 Contractual liability is thus only determined by the formal agreement (consensus) of the parties. The law of contract in an individualistic world, to borrow from Macaulay, provides the glue that binds

50 Kennedy (n 6 above) 1774. Kennedy goes so far as to argue that ‘[t]he “freedom” of individualism is negative, alienated and arbitrary. It consists in the absence of restraint on the individual's choice of ends, and has no moral content whatever.’ 51 JM Feinman ‘Critical approaches to contract law’ (1983) 30 UCLA Law Review 829 839; Kennedy (n 6 above) 1713. 52 Kennedy points out that the individualist ethic should be distinguished from the egotistical ethic in the sense that the individualist ethic has a strong positive ‘moral’ content whereas the egotist believes that it is entirely impossible and undesirable to place any limits on the perusal of self-interest. See Kennedy (n 6 above) 1714–1715. 53 A Cockrell ‘Substance and form in the South African law of contract’ (1992) 109 South African Law Journal 40 41. 54 See CJ Pretorius ‘Individualism, collectivism and the limits of good faith’ (2003) 66 Tydskrif vir die Hedendaagse Romeins-Hollandse Reg 638 639. 55 Kennedy (n 6 above) 1713. 56 Cockrell (n 53 above) 42. 57 Mensch (n 34 above) 760. (2008) 1 Constitutional Court Review 165 individuals to their agreements.58 This emphasis on the creation of an own law through the agreement causes individualism to take the view that courts should enforce agreements rather than intervene in relation to their terms on grounds of fairness or equity.59

With the rise of the market economy in the nineteenth century, the entire conceptual apparatus of the law of contract thus developed around this understanding of the relationship between the individual and society, the individual and the State and the individual and the law.60 As Feinman indicates:

Classical legal thought imagined a world of independent individuals, each of whom acts within a broad sphere of legal autonomy to pursue her own self-interest. The market was the model of social organisation and the acquisitive capitalist the paragon of personal behavior.61

In the law of contract this thought established freedom of contract as the hegemony (predominance) of the will theory. Duncan Kennedy famously added that the hegemony of individualism on the level of substance, in turn established the hegemony of rules on the level of contract law’s formal commitments. 62

But because of the fact that domination is implied in the concept of hegemony it would be inaccurate to contend that hegemony manages to liquidate its other entirely. This means that hegemony always implies a relation — a relation of domination and privileging, yes, but nevertheless, a relation. The trace of the Other thus always haunts any hegemony. The reason why this particular privileging or domination is hegemonic is because, as Cockrell indicates, ‘this privileging invariably proceeds on the basis that the preference for individualism and the rule-form is an axiomatic truth rather than a controversial premise in an ongoing argument.’63 The danger is that the hegemonic version of contract can lead us to think that this is the only order(ing) available in the law of contract — that freedom of contract grounded in individualism and the commitment to the rule- form, is the only version of contract that exists.64

58 S Macaulay ‘An empirical view of contract’ (1985) Wisconsin Law Review 465 466. 59 Pretorius (n 54 above) 640. 60 As Dalton points out, the will theory of contract flows from the political doctrine of contractualism, which holds that ‘all restraint is evil and that the government that governs least is best.’ See C Dalton ‘An essay in the deconstruction of contract doctrine’ (1985) 94 The Yale Law Journal 997 1013. 61 J Feinman ‘The conservatives, the market and the common Law’ (2005) 37 Review of Radical Political Economics 288 290. 62 Kennedy (n 6 above) 1685: ‘[t]he substantive and formal dimensions are related because the same moral, economic and political arguments appear in each.’ 63 Cockrell (n 53 above) 46. 64 Pretorius (n 54 above) 644: ‘There is no difficulty in recognising that individualism is the primary ideology underlying the law of contract. The problem lies in suggesting that it is the only one.’ 166 Hegemony, freedom of contract, good faith and transformation

The individualistic or liberal worldview in the understanding of contract did not and could not manage to altogether absent the law of contract from alternative worldviews. Applied to our current context this insight would imply that the traces of the pre-nineteenth century understanding of freedom of contract could not be erased by either the domination of the commercial classes of the nineteenth century or by the courts that represented their interests. The old underlying moral conceptions on which the law of contract was founded prior to the domination of individualism could thus not be abandoned outright.65 For one, the courts still naively wanted to believe that parties to the contract were in fact reaching consensus on its terms as honest, just, fair and non-exploitative persons (that is, persons who are in good faith). To this extent the courts still acknowledged external standards of justice. But the critical legal issue had shifted from whether the contract was fair to whether there objectively appeared to have been a ‘meeting of the minds’ between the contracting parties.66

Horwitz points out that although nineteenth century courts could not succeed in negating the ancient relation between natural law and contract law, they did succeed in setting up a system in which the courts could effectively ‘pick and choose’ which groups within the broad society they wanted to benefit in a given case.67 The discourse of the nineteenth century managed to set-up an intellectual divide between the system of formal rules (associated with the ‘rule of law’ or law as rules)68 and the ancient perceptions of morality and equity (which was seen as necessarily undermining the rule of law or perhaps, then, the law of rules). As already mentioned, the argument in Critical Legal Studies has been that this preference for rules was part of the classical laissez-faire individualist morality grounded in non-interventionism.69 It is thus no coincidence that most of modern general contract law consists of rules by way of which one determines whether the will of the parties coincided and if not, rules that regulate the undoing of the exchange.70 These rules are ordered around the doctrines of misrepresentation, duress, undue influence and so forth — theories that allegedly turn on the negation of the will — and do not visit explicitly — or explicitly do not visit — matters such

65 Horwitz (n 36 above) 953. 66 Horwitz (n 36 above) 955. 67 As above. 68 C Douzinas & A Geary Critical jurisprudence (2005) 7. 69 Kennedy (n 6 above) 1741. 70 See Van der Merwe et al (n 37 above) 19-32. (2008) 1 Constitutional Court Review 167 as the fairness of the bargain or the good faith of the parties.71 The theory of reasonable reliance on the appearance of consensus — as an alternative basis for the existence of a contract — is itself simply seen as a qualification of the will theory and not as a negation or in any way a complete alternative to it.72 Yet, as Cockrell indicates, the origin of the defect in the defences based on the negation of the will does not in fact reside in the will of one of the parties, but rather in the improper conduct. The same reasoning applies to the theory of reasonable reliance. These defences, as Cockrell argues, are all concerned with the legitimacy of conduct, and one way to link them would be to say that they all amount to instances of bad faith on the part of one of the parties.73 On the hegemonic understanding, however, the open-ended norms or standards of the law of contract cannot do this work. It cannot be directly employed or appealed to in the adjudication of contractual disputes. In this regard, the point that has repeatedly been made by our own courts in the recent past is that good faith, reasonableness and fairness, although subjacent to, or underpinning the law of contract, can only be used in the resolution of contractual disputes to the extent that they have become embodied in so-called crystallised rules of the law of contract.74 We will return to this problematic assertion below but for now it is important to state the general contention that these assertions serve only (in very real political ways) to keep the hegemony firmly in place.

4 The hegemonic order(ing) in the South African law of contract

4.1 Public policy as handmaiden and adversary of freedom of contract

Colonialism and imperialism of course carried the individualist understanding of freedom of contract grounded in the will to the numerous colonies. It is thus not surprising that the South African law of contract from its very inception reflects the hegemony of the will

71 CFC Van der Walt ‘Die huidige posisie in die Suid-Afrikaanse reg met betrekking tot onbillike kontraksbedinge’ (1986) 103 South African Law Journal 646 658. Also see Dalton (n 60 above) 1001 who makes the point that dealing with unfairness via those constructs that affect the will of the parties (duress, misrepresentation, undue influence etc) effectively constitutes a privatisation of the public enquiry into contract when ‘the undoing of a defective deal [is] presented as depending upon the absence of will or intent rather than on mere inequivalence of exchange.’ 72 Van der Merwe et al (n 37 above) 22. 73 Cockrell (n 53 above) 56. 74 See Brisley (n 13 above) 16B-C; Afrox Healthcare (n 13 above) 40H-41A; South African Forestry (n 13 above) 338J-339B; and Napier (n 13 above) para 7. 168 Hegemony, freedom of contract, good faith and transformation theory of contract. One of the most frequently quoted passages — ‘the most privileged statement of all’75 — justifying the privileging of (the ‘will theory version’ of) freedom of contract as the basis of contractual obligation in the South African law of contract, is to be found in a late nineteenth century case from the English law, Printing and Numerical Registering Co v Sampson:76

If there is one thing more than another which public policy requires, it is that men of full age and competent understanding shall have the utmost liberty in contracting, and their contracts, when entered into freely and voluntarily, shall be held sacred and shall be enforced by Courts of Justice. Therefore you have this paramount public policy to consider — that you are not lightly to interfere with this freedom of contract.77

In Burger v Central South African Railways78 Innes CJ ‘developed’ this basic premise and held in no uncertain terms that the South African law of contract does not recognise the right of a court to release a party to a contract from his obligations on considerations of fairness.79 And this is precisely the point, (as will hopefully become clearer below): when the focus is on the South African law of contract particularly, the rise of the hegemonic order(ing) of (freedom of) contract should be considered in the context of the fall of numerous values, principles and rights that came to be understood as freedom of contract’s adversaries and so became regarded as opposed to it.

A doctrinal aspect that has remained uncontroversial in the grand narrative of the South African law of contract has been the recognition that contracts concluded contra bonos mores and/or contrary to public policy fall foul of the validity requirement of legality and are therefore either void or unenforceable.80 The extent of this power to declare void or unenforceable contracts or terms

75 Cockrell (n 53 above) 46. See, for example, Wells v South African Alumenite Company 1927 AD 69 73 (Wells); Roffey v Catterall, Edwards & Goudré (Pty) Ltd 1977 4 SA 494 (N) (Roffey) 504G-H; Sasfin (Pty) Ltd v Beukes 1989 1 SA 1 (A) (Sasfin) 9F; Edouard v Administrator, Natal 1989 2 SA 368 (D) (Edouard) 379A; Baart v Malan 1990 2 SA 862 (E) (Baart) 868A; De Klerk v Old Mutual Insurance Co Ltd 1990 3 SA 34 (E) (De Klerk) 44; Benlou Properties (Pty) Ltd v Vector Graphics (Pty) Ltd 1993 1 SA 179 (A) (Benlou Properties) 187H; Standard Bank of SA Ltd v Wilkinson 1993 3 SA 822 (C) (Standard Bank) 830D; and Basson v Chilwan and Others 1993 3 SA 742 (A) (Basson) 761G. See also V Terblanche (2002) ‘The Constitution and general equitable jurisdiction in South African contract law’ unpublished LLD thesis, University of Pretoria, 2002 153 who refers to the ‘privileged position of 19th century contract law theory in South Africa.’ 76 1875 LR 19 (Printing and Numerical Registering Co) Eq 462 per Jessel MR. 77 Printing and Numerical Registering Co (n 76 above) Eq 465. The mood is also expressed in more general terms by Kotze CJ in Brown v Leyds (1897) 4 OR 17 (Brown) 31 who held that ‘no Court of Justice is competent to inquire into the internal value, in the sense of the policy, of the law, but only in the sense of the meaning or matter of the law’. 78 1903 TS 571 (Burger). 79 Burger case (n 78 above) 576. 80 Van der Merwe et al (n 37 above) 192-193, 200-203. (2008) 1 Constitutional Court Review 169 contrary to public policy was expressed as follows in Morrison v Angelo Deep Gold Mines Ltd:81

[I]t is a general principle that a man contracting without duress, without fraud and understanding what he does, may freely waive any of his rights. There are certain exceptions to that rule, and certainly the law will not recognise any arrangement which is contrary to public policy.

The dialectic that emerges here is that, on the one hand, public policy generally favours the utmost freedom of contract but, on the other, contracts concluded in violation of public policy are unenforceable or void. Public policy is thus regarded (from the perspective of freedom of contract) as both a legitimating and controlling device in our law of contract.82 The resolution/synthesis of this dialectic is, however, not impossible when one understands that the exercise of alleged freedom of contract is precisely not freedom of contract — that is, it is not freedom that can legitimate a contractual arrangement — when it is exercised contrary to public policy. This understanding also points to a deeper problem with the notion of autonomy as freedom of contract, which is said to form the cornerstone or foundational principle of the law of contract. The public policy requirement communicates that some contracts that are freely entered into (that is, in which the wills of legally competent parties overlap) will not be enforced for want of legality (in the broad sense). This means that, as a matter of a contract’s validity, there is a constitutive limit on freedom of contract that inevitably denies absolute party autonomy. And since legality is determined externally (not by the parties but with reference to ‘the interests and convictions’ of the society in which the contract is concluded)83 it means that the cornerstone of the law of contract is both autonomous and heteronymous.

Our courts, however, do not approach the matter in this way. Freedom of contract is consistently opposed to other ideals, values and rights, instead of integrally connected to and thus determined by those other values and rights. When public policy is, for instance, considered in relation to restraint of trade clauses the position is taken that ‘two values or freedoms come into conflict ... namely freedom of contract and freedom of trade.’84 In our law the position (since the decision in Magna Alloys)85 has been that restraints of trade are in principle enforceable in accordance with freedom of contract,

81 1905 TS 775 (Morrison) 779. 82 See Van der Merwe et al (n 37 above) 17. 83 Van der Merwe et al (n 37 above) 191. 84 Van der Merwe et al (n 37 above) 213. Also see Roffey (n 75 above) 505F (per Didcott J): ‘I am satisfied that South African law prefers the sanctity of contracts ... Freedom of trade does not vibrate nearly as strongly through our jurisprudence ... it is intrinsically the less commanding of the two ideas.’ 85 Magna Alloys & Research (SA) (Pty) Ltd v Ellis 1984 4 SA 874 (A) (Magna Alloys). 170 Hegemony, freedom of contract, good faith and transformation unless the restraint is contrary to public policy, the measure of public policy in relation to restraints of trade being reasonableness.86 On this formulation, freedom of contract is said to be the ‘preferred value’.87

Recently the constitutionality of the Magna Alloys formulation was questioned in the case of Reddy v Siemens Telecommunications (Pty) Ltd88 where it was argued that restraints of trade constitute limitations on the right to freely choose a trade, occupation or profession — a right expressly protected in section 22 of the Constitution.89 This means, so the argument went, that a restraint of trade is enforceable only when it can be proved by the employer who wants to enforce it to be a reasonable and justifiable limitation (as required by section 36 of the Constitution) on the section 22 right.90 The Supreme Court of Appeal (per Malan AJA) held that this argument effectively constituted an argument regarding onus and did not contest the earlier findings of the courts

that a restraint that is found to be reasonably required for the protection of the party who seeks to enforce it, in accordance with the test that has been laid down in the cases, is constitutionally permitted.91

The Court declined to adjudicate the issue of onus, holding that, ‘whether the onus was reversed or not, the result in the present case would be the same.’92 It continued to hold that what is called for in

86 In the context of the rules/standards discourse it should be pointed out that our courts do not seem to experience difficulty with the standard of ‘reasonableness’ both in the context of the public policy requirement and outside of it. In addition to its role in the area of restraints of trade provisions and public policy, reasonableness also plays a decisive role in the determination of contractual liability on the basis of reliance or the doctrine of quasi-mutual assent. In this regard see, for instance, Sonap Petroleum (SA) (Pty) Ltd v Pappadogianis 1992 3 SA 234 (A) (Sonap Petroleum) and Constantia Insurance v Compusource (Pty) Ltd 2005 4 SA 345 (SCA) (Constantia Insurance). 87 Van der Merwe et al (n 37 above) 215. 88 2007 2 SA 486 (SCA) (Reddy). 89 In Canon KwaZulu-Natal (Pty) Ltd t/a Canon Office Automation v Booth 2005 3 SA 205 (N) (Canon) 208-209 the court endorsed this reasoning. 90 Reddy (n 88 above) 495C-E. 91 Reddy (n 88 above) 495D. See Waltons Stationery Co (Edms) Bpk v Fourie en 'n Ander 1994 4 SA 507 (O) (Waltons Stationary) 510I-511F (SA); Kotze & Genis (Edms) Bpk en 'n Ander v Potgieter en Andere 1995 3 SA 783 (C) (Kotze & Genis) 786E-I; Knox D'Arcy Ltd and Another v Shaw and Another 1996 2 SA 651 (W) (Knox) 657Hff and 661D-F, CTP Ltd and Others v Independent Newspaper Holdings Ltd 1999 1 SA 452 (W) (CTP) 468G-H; Fidelity Guards Holdings (Pty) Ltd t/a Fidelity Guards v Pearmain 2001 2 SA 853 (SE) (Fidelity Guards) 861F-862G; and Oasis Group Holdings (Pty) Ltd and Another v Bray [2006] 4 All SA 183 (C) (Oasis Group) paras 23-25. 92 Reddy (n 88 above) 496A-B. (2008) 1 Constitutional Court Review 171 the case of restraints of trade is a value judgment

with two principal policy considerations in mind in determining the reasonableness of a restraint. The first is that the public interest requires that parties should comply with their contractual obligations, a notion expressed by the maxim pacta servanda sunt. The second is that all persons should in the interests of society be productive and be permitted to engage in trade and commerce or the professions.93

It concluded that the common law approach to the determination of the reasonableness or not of the restraint ‘gives effect to the precepts of section 36(1) of the Constitution’.94 Thus the enquiry into the reasonableness of the restraint simultaneously answers the question whether the restraint is a reasonable and justifiable limitation of the right in question.

Clearly, the Court here views the right to choose a trade, occupation or profession freely as coming into conflict with freedom of contract as pacta servanda sunt. However, an alternative approach was followed in the later judgment of Davis J in Advtech Resourcing t/a Communicate Personnel Group v Kuhn.95 In this judgment Davis J held that if contractual autonomy is part of the constitutional value of freedom and informs the ideal of dignity,96 then under a transformative Constitution such as ours, contractual autonomy must be read as internally limited by the rights in the Bill of Rights and ‘the spirit of the Constitution read as a whole’.97 In the context of restraints of trade this means that section 22 of the Constitution places an internal limit on freedom of contract to the extent that agreements in restraint of trade that constitute unreasonable and unjustifiable limitations on the section 22 right, will not be enforced. And since a restraint of trade constitutes a prima facie violation of the section 22 right, an employer who wants to enforce the restraint must prove that it is a reasonable and justifiable limitation of the right.98

When the Constitution states that the Bill of Rights does not deny the existence of any other rights or freedoms that are recognised or conferred by common law ‘to the extent that they are consistent with the Bill’ of Rights99 it places the common law notion of freedom of contract on an equal footing with all the rights in the Bill of Rights. In so doing, it contests the a priori elevation of a liberal notion of freedom of contract in the common law in the sense that it calls for

93 Reddy (n 88 above) 496D. 94 Reddy (n 88 above) 497 F/G. 95 2008 2 SA 375 (C) (Advtech). 96 Advtech (n 95 above) 387I. 97 Advtech (n 95 above) 388A. 98 Advtech (n 95 above) 387A. 99 Section 39(3) of the Constitution. 172 Hegemony, freedom of contract, good faith and transformation a synchronisation of freedom of contract with the ideals, values and rights in the Bill of Rights. Drucilla Cornell distinguishes the concept of synchronisation from both Dworkinian coherence and liberal conceptions of balancing: ‘Synchronisation recognises that there are competing rights situations and real conflicts, which may not be able to yield a “coherent” whole.’100 For Cornell synchronisation refuses closure and so calls for an ongoing, never-ended (re-)thinking of the whole even as it marks the failure of constructive coherence. For Cornell ‘the tension between the promise of synchronisation and the failure of its achievement’ opens the law to its transformative potential.101

In this context and on this approach, freedom of contract as pacta servanda sunt, becomes internally defined by the right in section 22 and, for that matter, any other right in the Bill of Rights as well as the founding values and ideals of the Constitution. But such an approach is only possible once one accepts that section 22 (and indeed all the rights in the Bill of Rights) is on an equal footing with freedom of contract (as the Constitution clearly says it is). Simply put, there cannot be talk of freedom of contract where the restraint of trade (or whatever contractual provision) is found to unreasonably and unjustifiably limit a right in the Bill of Rights or can be said to be contrary to the spirit, purport and objects of the Constitution. It goes without saying that where the limitation of a constitutional right is alleged in the context of restraints of trade, such an approach would — in this context — require the prioritisation of direct horizontal application of the Bill of Rights, even though indirect horizontal application might lead to the same result.102

The oppositional/dualistic reading of freedom of contract — by way of which its hegemonic understanding is maintained — is, in addition, also clearly illustrated through the unwillingness to read a fairness jurisdiction into the public policy requirement.103 As Glover argues, the defence that the enforcement of a contract is in specific circumstances unfair or unconscionable, has only been approved of in

100 D Cornell Transformations (1993) 36. 101 n 100 above, 35. 102 See in this regard S Woolman ‘The amazing, vanishing bill of rights’ (2007) 3 South African Law Journal 762 777. To reach directly for the baton of indirect horizontal application by way of sec 39(2) when a constitutional right is alleged to have been infringed is to render sec 8 of the Constitution redundant. As Woolman indicates, such an approach is inconsistent with the Constitutional Court’s decision in Khumalo v Holomisa 2002 5 SA 401 (CC) (Khumalo) para 32 where O’Regan J made it clear ‘that we should not attribute a meaning to one section of the Constitution that renders another section, quite literally, senseless.’ 103 Van der Merwe et al (n 37 above) 199. P Aronstam Consumer protection, freedom of contract and the law (1979) 43. (2008) 1 Constitutional Court Review 173 the restraint of trade context where, as we have seen, the restraint of trade is subjected to a reasonableness enquiry.104

However, in Sasfin v Beukes105 the Appellate Division considered the extent to which ‘simple justice between man and man’106 can trump the public interest in the strict enforcement of contracts generally. Primarily, the case centred around the validity of a deed of cession in which a customer of a bank (a doctor) ceded all his future debtors to the bank regardless of whether he owed the bank money or not.107 The cession effectively rendered the doctor the slave of the bank. The majority of the court was of the opinion that these aspects of the cession could not be severed from the rest of the agreement and held that the entire transaction was unenforceable.108

In this case the Appellate Division was willing to evaluate the substantive fairness of the disputed deed of cession to come to a conclusion that the cession was clearly unreasonable and irreconcilable with the public interest. Lewis describes the decision as ‘the one decision which yields a ray of light in the field of contractual policy, where the court was both bold and innovative in escaping the shackles of formalism’.109 Be that as it may, one again sees the opposition between freedom of contract and other values in Smalberger JA’s judgment where it is held that unlawfulness comes into play where the public interest in the strict enforcement of contracts in accordance with the principle of freedom of contract, is trumped by other relevant factors.110 These relevant factors are expressed by the Court to ensure that ‘public policy ... properly take into account the doing of simple justice between man and man.’111

The uneasiness with which the Appellate Division approaches the issue of unfair contracts is illustrated in the following famous dictum:

No court should therefore shrink from the duty of declaring a contract contrary to public policy when the occasion so demands. The power to declare contracts contrary to public policy should however be exercised sparingly and only in the clearest of cases, lest uncertainty as to the validity of contracts result from an arbitrary and indiscriminate use of the power. One must be careful not to conclude that a contract is

104 G Glover ‘Lazarus in the Constitutional Court: An exhumation of the exceptio doli generalis?’ (2007) 124 South African Law Journal 449 451. 105 1989 1 SA 1 (A) (Sasfin). 106 As above. 107 Sasfin (n 105 above) 6A-D. 108 Sasfin (n 105 above) 17-19. 109 C Lewis ‘Towards an equitable theory of contract: The contribution of Mr Justice EL Jansen to the South African law of contract’ (1991) 108 South African Law Journal 249 264 n76. 110 Sasfin (n 105 above) 9D-G. 111 Sasfin (n 105 above) 13J. 174 Hegemony, freedom of contract, good faith and transformation

contrary to public policy merely because its terms (or some of them) offend one's individual sense of propriety and fairness.112

Hawthorne has remarked that the dicta above merely emphasises the South African judiciary’s narrow interpretation of the relevance of equity considerations in the public interest.113 By holding that it is only in the ‘clearest of cases’ that a court may use its power to refuse to enforce an unfair term, and that the power to do this must be used ‘sparingly’, the court suggests that unconscionability in and of itself cannot invalidate a contract.114 The promise that we can correct for clear or extreme cases only, simply suggests, to paraphrase Dalton, that the worst features of the system can be held in check, without tinkering with its regular operation.115 Van der Merwe et al confirm the matter:

Sasfin’s decision must not be interpreted to mean that the unreasonably harsh or unconscionable effect of a term between the parties is by itself a sufficient ground for making an agreement illegal.116

In Donelly117 Kriegler J stressed this point and placed a great deal of emphasis on the fact that the Sasfin judgment should not be regarded ‘as a free pardon to recalcitrant and otherwise defenceless debtors’118 because it is ‘decidedly not that’.119 Kriegler J also did not let the opportunity pass to emphasise that ‘pacta servanda sunt is still a cornerstone of our law of contract’120 and that ‘nothing said or implied’121 in the Sasfin principle can be said to derogate from this important fact.122 In Botha (now Griessel) v Finanscredit (Pty) Ltd123 the Appellate Division confirmed that

a Court’s power to declare contracts contrary to public policy should be exercised sparingly and only in cases in which the impropriety of the transaction and the element of public harm are manifest.124

112 Sasfin (n 105 above) 9A-B. (emphasis added) 113 L Hawthorne ‘Public policy and micro lending — has the unruly horse died?’ (2003) 66 THRHR 116 118. 114 These sentiments are also reflected in cases decided by the lower courts in the aftermath of Sasfin. See in this regard Standard Bank (n 75 above) 825C-827A and Pangbourne Properties Ltd v Nitor Construction (Pty) Ltd 1993 4 SA 206 (W) (Pangbourne Properties) 210G-214F. 115 Dalton (n 60 above) 1037. 116 Van der Merwe et al (n 37 above) 219. 117 Donelly v Barclays National Bank Ltd 1990 1 SA 375 (W) (Donelly). 118 n 117 above, 381F. 119 As above. 120 Donelly (n 117 above) 381H. 121 As above. 122 As above. 123 1989 3 SA 773 (A) (Botha). 124 Botha (n 123 above) 782I-783C. (2008) 1 Constitutional Court Review 175

In this case the court refused to declare a deed of suretyship void on the grounds of public policy, holding that, although ‘somewhat rigorous’, the surety was not left ‘helpless in the clutches of the plaintiff’.125 This prompts one to ask whether one party should necessarily be left helpless in the clutches of the other for the court to exercise its Sasfin power? If so, very little of the work of reform and transformation of the law of contract can be expected to be done through the utilisation of the Sasfin principle.

But the point that I would like to stress going forward is that the courts’ concern with maintaining the hegemony of the will theory of contract causes them to assume that there is an area in which freedom of contract can be exercised independent of public policy, fairness, reasonableness and good faith. From a political point of view, the paradox which inscribes itself in this opposition between freedom of contract and the so-called constraints of public policy, fairness, reasonableness and good faith exists in that the claim that the reliance on freedom of contract and its concomitant rules is somehow a-political and thus unproblematic, generates increasing and proliferating politics in the law of contract. The more the courts attempt to depoliticise the law of contract by way of limiting the open-ended norms in favour of liberal legalism grounded in a rule jurisprudence, the more it politicises it by way of these oppositions and dualisms.

Below I will argue that the unwillingness to deal with unfair contracts in terms of the public policy requirement is symptomatic of a deeper illness, namely the courts’ refusal to accept that the ethical element of contract (good faith or the bona fides) — as a foundational value of the law of contract — forms part and parcel of not only the public policy requirement126 but also of the very maxim pacta servanda sunt.127 In other words, the opposition between the foundational norms of freedom of contract and good faith serves — as we shall see — to maintain a certain instrumental politics in the law of contract weighted in favour of commercial interests. The opposition itself thus furthers the privileging or hegemony of a liberal version of freedom of contract, which closes off the law of contract from the question of the ethical embodied in the good faith requirement.

125 Botha (n 123 above) 783J. 126 Van der Merwe et al (n 37 above) 200: ‘in the context of legality there is a close relationship between what is fair and reasonable and what is in accordance with good faith’. 127 See n 37 above. 176 Hegemony, freedom of contract, good faith and transformation

4.2 Good faith and the exceptio doli generalis in the South African law of contract

The South African law of contract has long considered all contracts as acts of good faith.128 This is the case because we inherited substantial parts of the Roman-Dutch law of contract in which contracts were considered iudicia bonae fidei.129 In Roman law, however, a distinction existed between negotia stricti iuris and the negotia bonae fidei.130 Contracts from the ius stricti bound the debtor to perform strictly in accordance with what he promised in the formula and not in accordance with what the bona fides could expect of him, unless the formula itself referred to the bona fides.131 In the case of the negotia bonae fidei the bona fides were conclusive and the absence thereof, whether during negotiations, conclusion or institution of the action, gave rise to a defence.132 The bona fides thus operated as an evaluative yardstick to determine the enforceability of the negotia bonae fidei.

To curb possible injustices or unconscionable conduct as a result of the enforcement of the negotia stricti iuris, the praetor introduced the exceptio doli generalis.133 Here the defendant was allowed to submit facts that he would otherwise not have been able to submit because of the operation of the strict ius civile.134 The exceptio therefore functioned to curb the abuse of rights in appropriate circumstances and became the instrument with which more equitable

128 Meskin v Anglo-American Corporation of SA Ltd (1968) 4 SA 793 (W) (Meskin) 802; Paddock Motors (Pty) Ltd v Ingesund 1976 3 SA 16 (AD) (Paddock Motors) 28; Magna Alloys (n 85 above) 893C; Tuckers Land & Development Corporation (Pty) Ltd v Hovis 1980 1 SA 645 (A) (Tuckers Land) 652; Mutual & Federal Insurance Co Ltd v Oudtshoorn Municipality 1985 1 SA 419 (A) (Mutual & Federal) 433; Savage & Lovemore Mining (Pty) Ltd v International Shipping Co (Pty) Ltd 1987 2 SA 149 (W) (Savage & Lovemore Mining)198A, Sasfin (n 75 above) 8C-D; Botha (n 123 above) 782J; LTA Construction Bpk v Administrateur Transvaal 1992 1 SA 473 (A) (LTA Construction) 480D; Eerste Nasionale Bank van Suidelike Afrika Bpk v Saayman NO 1997 4 SA 302 (SCA) (Eerste Nasionale Bank) 321; Janse van Rensburg v Grieve Trust 2000 1 SA 315 (C) (Janse van Rensburg) 325-326; Afrox Healthcare (n 13 above) 40I-J; Van der Westhuizen v Arnold 2002 6 SA 453 (SCA) (Van der Westhuizen) 469E-F. 129 Van der Merwe et al (n 37 above) 319. Aronstam (n 103 above) 173: ‘the sole question facing a court in Roman-Dutch law was whether the conduct of a party to a contract, either in any negotiations that preceded the contract, or in any action that was based upon it, had complied with the requirements of good faith.’ 130 See CFC van der Walt (n 71 above) 648 and NJ Grové ‘Kontraktuele gebondenheid, die vereistes van die goeie trou, redelikheid en billikheid’ (1998) 61 Tydskrif vir Hedendaagse Romeins-Hollandse Reg 687 688 who show that the distinction was closely related to the Roman procedural law and the specific defences it allowed for in different cases. 131 Van der Walt (n 71 above) 648. 132 As above. 133 As above. 134 As above. (2008) 1 Constitutional Court Review 177 principles were introduced in the law of contract by the praetorian law.135

The question regarding the reception of the exceptio doli generalis into the South African law of contract has been contentious. The question is complicated by the fact that all contracts became negotia bonae fidei in Roman-Dutch Law, thus appearing to do away with the need for such a defence. However, Aronstam argues that the exceptio doli generalis became part of the Roman-Dutch law in that it was implied by good faith:136

if all contracts were bonae fidei they had to be performed in accordance with the dictates of good faith, and any absence of bona fides could be taken into account by a judge in any action based on such contracts.137

Aronstam’s argument thus is that the exceptio doli generalis formed part of the requirement of good faith and as such was received into the South African law of contract.

Six months before the decision in Sasfin (discussed above), the Appellate Division had occasion to consider the question regarding the reception of the exceptio doli generalis into the South African law of contract in the case of Bank of Lisbon and South Africa v De Ornelas & Others.138 In this case the court primarily concerned itself with the question whether the exceptio survived the reception of Roman Law into Roman-Dutch Law and of Roman-Dutch Law into the South African law. The court held that the exceptio doli generalis never formed part of the Roman-Dutch law,139 could therefore not have been received into the South African law140 and that its occasional appearances on the scene of the South African law141 should finally be prohibited by burying it ‘as a superfluous, defunct anachronism.’142

This decision is also generally regarded as the decision in which the Appellate Division did away with the perception that a contract could be adjudicated and declared unenforceable by a court on the

135 G Lubbe ‘Bona fides, billikheid en die openbare belang in die Suid-Afrikaanse kontraktereg’ (1990) 1 Stellenbosch Law Review 9. Also see C Lewis ‘The demise of the exceptio doli: Another route to contractual equity?’ (1990) 107 South African Law Journal 26 31-32. 136 n 103 above, 173. 137 n 103 above, 172. 138 1988 3 SA 580 (A) (Bank of Lisbon). 139 Bank of Lisbon (n 138 above) 605H. 140 Bank of Lisbon (n 138 above) 607A. 141 For a discussion of the various interpretations of the exceptio doli generalis see SWJ Van der Merwe et al ‘The exceptio doli generalis: Requiescat in pace — vivat aequitas’ (1989) 106 South African Law Journal 235. Also see Weinerlein v Goch Buildings Ltd 1925 AD 282 (Weinerlein); Zuurbekom Ltd v Union Corporation Ltd 1947 1 SA 514 (AD) (Zuurbekom); Paddock Motors (n 128 above); and Arprint Ltd v Gerber Goldschmidt Group South Africa (Pty) Ltd 1983 1 SA 254 (A) (Arprint). 142 Bank of Lisbon (n 138 above) 607B. Also see Lubbe (n 135 above) 9. 178 Hegemony, freedom of contract, good faith and transformation basis of good faith.143 The conflation of these two issues (the abolition of the exceptio doli generalis on the one hand and the abolition of an explicit good faith jurisdiction, on the other) finds its origin in the following dictum of Joubert JA:

Nor can I find any evidence of the existence of a general substantive defence based on equity. This is not surprising inasmuch as the Dutch Courts, unlike the English Courts until the Judicature Act 1873 became operative in 1875, did not administer a system of equity as distinct from a system of law. Roman-Dutch law is itself an inherently equitable legal system. In administering the law the Dutch Courts paid due regard to considerations of equity but only where equity was not inconsistent with the principles of law. Equity could not override a clear rule of law. That is also the position of our Courts as regards their equitable jurisdiction ... Moreover, I cannot find any support in Roman-Dutch law for the proposition that in the law of contract an equitable exception or defence, similar in effect to the exceptio doli generalis, was utilised under the aegis of bona fides.144

In his discussion of the majority decision, Cockrell shows that the decision that the exceptio doli constitutes ‘a superfluous, defunct anachronism’,145 is founded upon an extreme form of individualism which denies that the law may ‘legitimately superimpose an overriding duty to act in good faith’146 upon the voluntary agreements of legal subjects with full capacity.147 As Cockrell points out, the Court’s problem with accepting the exceptio was that it enjoins a judge to employ a standard that cannot be cast in ‘a clear rule of law’.148 As the Court held, quoting Van der Linden’s account of Voet 1.1.16: ‘... [j]udges and jurists ought to look to nothing more carefully than this, that they do not forsake the written law for some headstrong equity ...’.149 Joubert JA’s rejection of the exceptio doli

143 See Eerste Nasionale Bank (n 128 above) 323B. 144 Bank of Lisbon (n 138 above) 605I-606D. 145 Bank of Lisbon (n 138 above) 607B. 146 n 53 above, 44. 147 As above. 148 Also see the judgment in Tjollo Ateljees (Eins) Bpk v Small 1949 1 SA 856 (A) (Tjollo Ateljees). In this decision the Appellate Division recommended the abolition by the legislature of the doctrine of laesio enormis. According to laesio enormis a party who had either agreed to pay a price of more than double the value of the thing or agreed to sell for less than half of the value of the thing could either cancel the agreement or claim price adjustment (see Van der Merwe et al (n 37 above) 132). Cockrell (n 53 above) 45 illustrates clearly that the basis of the decision was, on the one hand, that laesio enormis constituted an altruist intrusion on the terrain of individualism and, on the other, that laesio enormis could not be cast in the rule form. 149 Bank of Lisbon (n 138 above) 610D-E. (2008) 1 Constitutional Court Review 179 is thus clearly linked to his disapproval of the discretion which will be afforded to judges by its acceptance.150

The attempt to rid our law of the exceptio doli generalis should then, given these observations, be understood as part and parcel of the attempt to keep the hegemony of freedom of contract as the will theory and the maxim pacta servanda sunt, firmly in place. With this decision, the Appellate Division attempted finally to seal the hegemony of the will theory as freedom of contract. This is particularly evident when Bank of Lisbon is considered in the light of the Sasfin decision six months later in which possible interference with liberal freedom of contract by way of public policy was further curtailed.151 The abolition of the exceptio doli generalis confirmed that the enquiry in the South African law of contract when it comes to enforceability is conducted with reference to so-called clear rules of law that are grounded in individualism. Consequently as we shall see, the Bank of Lisbon decision forms the focal point when it comes to the marginalisation of good faith as the ethical element of contract in South Africa.

This marginalisation, however, was not achieved without some strife. In the first place — regarding the exceptio doli generalis as a technical remedy — the Appellate Division contradicted itself when it held in the later case of Van der Merwe v Meades152 that the replicatio doli still formed part of our law. As Kerr argued at the time, the same court had previously held in Bank of Lisbon that the conclusions about the non-reception of the exceptio doli generalis ‘equally hold for the replicatio doli generalis’.153 Kerr submits that the linking was correct.154 If, as Kerr points out, the replicatio doli, on the later decision of the court, survived the reception, and the same court linked the replicatio with the exceptio and has not departed from that position, then the only conclusion that can follow is that the exceptio doli generalis must also have survived the reception.155

150 The majority decision clearly does not take account of, or acknowledge, the subjectivity/discretion which is at play in considering whether ‘a clear rule of law’ is in fact that and, even more problematic, the subjectivity that is at play in any decision as regards the question whether the ‘clear rule of law’ is applicable or not. 151 The mantra that public policy should properly take into account the doing of simple justice between man and man and the way in which that mantra was employed in Sasfin does not, without more, lead to a conclusion that Sasfin somehow provided an escape route from strict enforcement of contract. This is evident from the cases that followed Sasfin, as discussed earlier. 152 1991 2 SA 1 (A) (Van der Merwe) 2F-3A. 153 Bank of Lisbon (n 138 above) 608F-G. 154 AJ Kerr ‘The replicatio doli reaffirmed. The exceptio doli available in our law’ (1991) 106 South African Law Journal 583 584. 155 n 154 above, 585. 180 Hegemony, freedom of contract, good faith and transformation

Regarding the question whether the Bank of Lisbon judgment — in doing away with the exceptio doli — in fact did away with the power to declare a contract or a provision thereof as void or unenforceable on the basis of good faith, Olivier JA argued strongly in his famous (perhaps notorious) minority judgment in Eerste Nasionale Bank van Suidelike Afrika Bpk v Saayman NO156 that Bank of Lisbon did not and could not have done away with this power. Olivier JA argued that the reception of all contracts as iudicia bonae fidei into the South African law of contract made the exceptio doli generalis (as a technical remedy that had to be pleaded), superfluous in that it was implied in the idea of iudicia bonae fidei that a court would have discretion to examine and pronounce on the bona fides. Olivier JA continued that this would be the position from the perspective of those who accept the broad, dynamic working of the bona fides and the discretion of judges to apply the bona fides.157 Olivier JA argued bravely that the bona fides principle forms part of the public policy requirement158 and could be directly applied to the facts before him. This was a certain attempt to radicalise the public policy principle as enunciated in Sasfin. To quote Olivier JA:

[e]k hou dit as my oortuiging na dat die beginsels van die goeie trou, gegrond op openbare beleid, steeds in ons kontraktereg 'n belangrike rol speel en moet speel, soos in enige regstelsel wat gevoelig is vir die opvattinge van die gemeenskap, wat die uiteindelike skepper en gebruiker van die reg is, met betrekking tot die morele en sedelike waardes van regverdigheid, billikheid en behoorlikheid.159

(I believe that the principles of good faith, based on public policy, still play and should continue to play an important role in our law of contract as it does in any legal system which is sensitive to the convictions of the community — the ultimate creator and user of the law — with respect to the moral and ethical values of justice, fairness and propriety.)

In his discussion of unfair contracts in the fourth edition of The law of contract in South Africa, Richard Christie remarked that

[t]here is every reason to hope that when the opportunity arises the Supreme Court of Appeal will apply Olivier JA’s reasoning, harnessed to the concept of public policy, in the context of the unfair enforcement of a contract.160

In NBS Boland Bank Ltd v One Berg River Drive CC and Others; Deeb and Another v ABSA Bank Ltd; Friedman v Standard Bank of SA Ltd161

156 n 128 above. 157 Eerste Nasionale Bank (n 128 above) 323. 158 Eerste Nasionale Bank (n 128 above) 323-324 159 Eerste Nasionale Bank (n 128 above) 326 F-H. 160 RH Christie The law of contract in South Africa (4ed, 2001) 19. 161 1999 4 SA 928 (SCA) (Friedman). (2008) 1 Constitutional Court Review 181 the Supreme Court of Appeal (SCA) held obiter that in the matter before it

[a]n analogous conclusion may well be reached if one applies the modern concept of the role of public policy, bona fides and contractual equity to the question in issue (see for example, Eerste Nasionale Bank van Suidelike Afrika Bpk v Saayman NO [reference omitted] per Olivier JA.162

This appeared at the time to have constituted at least an implicit endorsement of what Olivier JA said in that case.

In Miller163 Ntsebeza AJ interpreted the judgment of Olivier JA in Saayman to mean that a court can indeed refuse to uphold a so-called non-variation or Shifren clause164 on the basis of considerations of good faith.165 In turn, Van Zyl J argued — relying explicitly on Olivier JA’s judgment in Saayman — in Janse van Rensburg v Grieve Trust166 that ‘the principles of justice, equity, reasonableness and good faith’ required the development of the common law in the case before him.167 Davis J, in obiter remarks in Mort NO v Henry Shields-Chiat168 also invoked Olivier JA’s judgment and added that

it is clear that if the doctrine [of bona fides] is to be taken seriously then the primary importance accorded to the private autonomy of contracting parties must be reconsidered as must the hegemony of the will theory of contract which survives even in the context of dicta which nod in the direction of social responsibility.169

Davis J implicitly accepted Olivier JA’s reasoning in Eerste Nasionale Bank that good faith forms part of the public policy requirement when he stated that Olivier JA’s approach builds on the judgment of Smalberger JA in Sasfin and particularly on the dictum from that case as quoted earlier.170 Similar to Olivier JA, Davis J attempted to radicalise Sasfin by the attempt to infuse it with the bona fides.

But what sets Davis J’s judgment in Mort apart is the express link made in the judgment between, on the one hand, the argument in favour of the bona fides and, on the other, the ethical implications

162 Friedman (n 161 above) 937F. Also see Pretorius (n 54 above) 643. 163 Miller and Another NNO v Dannecker 2001 1 SA 928 (C) (Miller). 164 This name for a non-variation clause (which generally stipulates that no amendments to a written agreement shall be valid unless reduced to writing) is derived from the decision in SA Sentrale Ko-op Graanmaatskappy Beperk v Shifren en Andere 1964 4 SA 760 (A) (Shifren). 165 Miller (n 163 above) 938 para 19 166 n 128 above. 167 n 128 above, 325D-F. 168 2001 1 SA 464 (C) (Mort). 169 n 168 above, 474I (Emphasis added). 170 See note 112 above. 182 Hegemony, freedom of contract, good faith and transformation for adjudication in the law of contract of the reconstitution of the legal order occasioned by the enactment of the Constitution. To quote Davis J:

Like the concept of boni mores in our law of delict, the concept of good faith is shaped by the legal convictions of the community. While Roman- Dutch law may well supply the conceptual apparatus for our law, the content with which concepts are filled depends on an examination of the legal conviction of the community — a far more difficult task. This task requires that careful account be taken of the existence of our constitutional community, based as it is upon principles of freedom, equality and dignity. The principle of freedom does, to an extent, support the view that the contractual autonomy of the parties should be respected and that failure to recognise such autonomy could cause contractual litigation to mushroom and the expectations of contractual parties to be frustrated. [Reference omitted.] But the principles of equality and dignity direct attention in another direction. Parties to a contract must adhere to a minimum threshold of mutual respect in which ‘the unreasonable and one-sided promotion of one's own interest at the expense of the other infringes the principle of good faith to such a degree as to outweigh the public interest in the sanctity of contracts.’ [Reference omitted.] The task is not to disguise equity or principle but to develop contractual principles in the image of the Constitution.171

However, in Brisley v Drotsky172 the majority of the SCA rejected — in no uncertain terms — both Olivier JA’s Eerste Nasional Bank judgment as well as the judgments that built on it. In the first instance, the majority criticised Olivier JA’s judgment for attempting indirectly to resurrect the exceptio doli generalis.173 With this statement the majority seems to have uncritically accepted that the exceptio doli generalis is not a part of our law. But as we have seen, on the Court’s own jurisprudence it is not at all clear whether the exceptio doli generalis has in fact been abolished. The terse statement in Brisley regarding the attempt to resurrect the exceptio doli generalis dismisses this controversy/contradiction by pretending that the exceptio is in fact dead and buried. This statement in Brisley can of course be seen as implicitly overriding the court’s Van der Merwe decision (in which it, ironically, resurrected the exceptio doli generalis), but the Court explicitly stated in a footnote that the question regarding the reconsideration of the exceptio did not arise in Brisley.174 This left as uncertain as ever the position in our law regarding the status of the exceptio.

171 Mort (n 168 above) 474I-475D 172 n 13 above. The dispute in this case also concerned the question regarding the enforceability of contractual terms (including a Shifren clause) in circumstances where oral amendments to the agreement were made but not reduced to writing. 173 Brisley (n 13 above) 14 para 17. 174 Brisley (n 13 above) n 10. (2008) 1 Constitutional Court Review 183

As regards Davis J’s view in Mort that the bona fides (like the boni mores in the law of delict) are shaped by the legal convictions of the constitutional community (as an ideal community), the majority held that if this means that the enforcement of contractual terms depends on the convictions of the community (gemeenskapsgevoel), then it cannot agree because there are material policy differences between the approach to the law of delict and the approach to the law of contract.175 According to the majority the unqualified acceptance of Davis J’s judgment will create a state of unacceptable chaos and disorder in the law of contract.176 A court cannot seek shelter in the shadow of the Constitution, writes the Brisley court, in order to attack and overthrow principles.177 In this regard it should be emphasised that Davis J did not appeal to an unqualified, actually existing community in his judgment. He explicitly referred to the constitutional community — which can only be an ideal community — and that is why Davis J explicitly invoked the constitutional ideals of dignity, equality and freedom. The point Davis J was making was that the very reconstitution of the legal order required that good faith should be given content with reference to the normative (and transformative) commitments of the Constitution. He was not hiding in the shadow of the Constitution in order to attack and overthrow principles from that vantage point.

Be that as it may, the majority in Brisley accordingly rejected the idea that good faith can be utilised as an independent, free-floating basis for the invalidation or unenforceability of contractual provisions178 on the basis that good faith as ‘the legal convictions of the community’ would prove to be too nebulous a concept to determine the enforceability of contractual terms. The Court stated, however, that good faith is a fundamental principle which in general underlies the law of contract and becomes expressed in its particular rules and principles.179

At this point in the judgment the Brisley court introduced what it called another underlying value of the law of contract: the principle

175 Brisley (n 13 above) para 21. 176 Brisley (n 13 above) 15 para 21. ‘Om eensklaps aan regters 'n diskresie te verleen om kontraktuele beginsels te verontagsaam wanneer hulle dit as onredelik of onbillik beskou, is in stryd met hierdie werkswyse. Die gevolg sal immers wees dat die beginsel van pacta sunt servanda grotendeels verontagsaam sal word omdat die afdwingbaarheid van kontraktuele bepalings sal afhang van wat 'n bepaalde regter in die omstandighede as redelik en billik beskou.’ This statement prompted L Hawthorne ‘The end of bona fides’ (2003) South African Mercantile Law Journal 271 276 to remark as follows: ‘This cri de coeur has two deplorable aspects: the tacit belief that the majority of contractual terms are not reasonable or equitable, and the belief that judges are not to be trusted with what is reasonable and equitable.’ 177 Brisley (n 13 above) 16D para 24. 178 Brisley (n 13 above) 16D para 22 179 Brisley (as above). 184 Hegemony, freedom of contract, good faith and transformation that public policy requires that contracts that have been entered into freely and in all earnest by parties with full capacity should be enforced.180 The Court continued to hold that these two underlying values (good faith and freedom of contract) sometimes come into conflict with each other and that it is the task of the Courts and the SCA principally to weigh these values against each other and to make gradual and slow changes on occasions when they appear to be necessary.181

Again, this judgment makes it clear that good faith and pacta servanda sunt are positioned as opposing values, that is to say an instrumental politics is set up between the two underlying values. But it is also the case that it is at this point that the Brisley judgment deconstructs itself.182 It does so in the following way: When the court discusses the idea of an underlying value in relation to good faith, it makes the point that it is in the very nature of an underlying value that courts do not use it directly to adjudicate contracts. Rather, they wait for the underlying value to become embodied in the particular rules of the law of contract. For this reason the Court concludes that good faith is not an independent or free-floating basis for the invalidation of contractual terms. But when it comes to the discussion of the other underlying value — the enforcement of contracts in the public interest — the Court does not realise the implications of its own opinion regarding underlying values. Certainly, the consequences of identifying freedom of contract as an underlying value should be the same — courts should wait for the underlying value to become embodied in the rules of the law of contract. Yet, the history of the South African law of contract is the history of the direct reliance on the so-called underlying principle of freedom of contract. If there were to be no direct reliance on so-called underlying values, then it would be hard to imagine how the rules of the law of contract were formulated in the first place. The fact that freedom of contract is an underlying value has never prevented a court from directly relying on it. Why then should the ‘underlying value’-character of good faith prevent direct reliance on it? Yet this is exactly what the Brisley majority wants us to believe: good faith cannot be directly relied upon because it is an underlying value. This was stated even more explicitly in Afrox Healthcare Bpk v Strydom:183

180 Brisley (n 13 above) 16D para 23 quoting Magna Alloys (n 85 above) 893I-894A. 181 Brisley (n 13 above) 16D para 24. 182 J Derrida & M Ferraris A Taste for the Secret (2001) 80: ‘I ought to have specified that what happens deconstructs itself in the process. It is not I who deconstruct; rather something I called “deconstruction” happens to the experience of a world, a culture, a philosophic tradition: “it” deconstructs, ça ne va pas, there is something that budges, that is in the process of being dislocated, disjointed, disadjoined, and of which I begin to be aware. Something is “deconstructing” and it has to be answered for.’ 183 Afrox Healthcare (n 13 above). (2008) 1 Constitutional Court Review 185

Concerning the place and role of abstract ideas such as good faith, reasonableness, fairness and justice, the majority in the Brisley case held that, although these considerations are subjacent to our law of contract, they do not constitute an independent or ‘free-floating’ basis for the setting aside or the non-enforcement of contractual provisions; put differently, although these abstract considerations represent the foundation and very right of existence of rules of law and can also lead to the shaping and transformation of these rules, they are not self- contained rules of law. When it comes to the enforcement of a contractual provision, the Court has no discretion and does not act on the basis of abstract ideas, but precisely on the basis of crystallised and established rules of law.184

One clearly sees in this dictum the consistent political privileging of the individualism/rules position by way of which the hegemonic understanding of freedom of contract is kept firmly in place (which in turn keeps the hegemonic order(ing) of the law of contract in place). The Court’s unease with good faith (described by Cockrell as ‘the epitome of a legal standard that embodies communitarian values of altruism, care and concern’185 (the opposite thus of individualism)) is that it cannot be regarded as a rule of law. On the other hand, freedom of contract is simply, unproblematically understood as being both an underlying value and a rule which could be relied upon to hold the defendant in Afrox for instance, to an agreement that exempted a hospital from liability for everything but its intentional infliction of harm.186

A number of deductions regarding good faith and freedom of contract in the South African law of contract can be made at this point. The first is that good faith is currently not considered in the South African law of contract to be a part of the public policy requirement (which, from a transformative point of view, presents its own set of problems, primarily in that public policy becomes closed off from the ethical question generally when good faith does not form part of it). Second, good faith is not considered to be an independent or free-floating basis for the invalidation or non-enforcement of contracts or contractual terms. Third, the latter is the case because good faith is an underlying value of the South African law of contract and is/becomes embodied in the particular rules of the law of

184 Afrox Healthcare (n 13 above) 40H-41A (author’s translation from the original Afrikaans). The point was restated in South African Forestry (note 13 above) 338J-339B as follows: ‘although abstract values such as good faith, reasonableness and fairness are fundamental to our law of contract, they do not constitute independent substantive rules that courts can employ to intervene in contractual relationships. These abstract values perform creative, informative and controlling functions through established rules of the law of contract. They cannot be acted upon by the courts directly ... After all, it has been said that fairness and justice, like beauty, often lie in the eye of the beholder.’ 185 Cockrell (n 53 above) 55. 186 Afrox Healthcare (n 13 above). 186 Hegemony, freedom of contract, good faith and transformation contract. (How this happens is uncertain given that courts are not allowed to touch good faith.) Fourth, freedom of contract understood as the maxim pacta servanda sunt, although itself an underlying value of the law of contract, is and can be used in a direct way.187 Fifth, there appears to be no reason other than an ideological or political one which would explain why one underlying value — freedom of contract — can be used to develop rules by way of which a party can be held to or released from her agreements, whereas another underlying value — good faith — cannot be used in this way.

In addition, a matter that is repeatedly ignored in the debate regarding the role of good faith in the South African law of contract is the implications for the law of contract of the newly constituted legal order. In the Brisley case, for instance, the Court faulted Davis J’s and other judges’ appeal to this order as putting in place an ideal community to which can be appealed in the law of contract by way of good faith (as part of the public policy requirement). The stance of the Brisley court represents a static, closed view of a legal system in which good faith has somehow been worked out in the past and incorporated into the rules of the law of contract. As such, this view questions the dynamic role of the bona fides in relation to the transformation of the South African law of contract and, as Olivier JA suggested in Eerste Nasionale Bank, it is out of step both with the spirit of our law and with the needs of the community.188

In 2007 the Constitutional Court delivered two decisions that involved the tangled web weaved with freedom of contract, good faith, the exceptio doli generalis and public policy in the South African law of contract. Below I shall critically evaluate these decisions after which I shall argue for a re-emphasis on good faith as the ethical element of contract. I believe that this re-emphasis is mandated by an understanding of the new legal order as aspiring to the ideal of civic friendship. By deploying the word ‘re-emphasis’ I want to signal my belief that the normative character of the Constitution requires us to consider good faith as part of the public policy requirement in two ways — both in that contracts concluded contrary to public policy are unenforceable but also in that public policy requires that parties who enter into agreements freely

187 In this context, D Hutchison ‘Non-variation clauses in contract: any escape from the Shifren straitjacket?’ (2001) 118 South African Law Journal 720 745 has remarked that ‘to reach directly for the baton of good faith would be to confess to a want of technical expertise or creativity. Palm-tree justice no doubt has its virtues, but as lawyers we should adhere to the ideal of justice according to law.’ Hutchison does not explain why it is acceptable and in accordance with the ‘ideal of justice according to law’ to reach directly for the baton of freedom of contract. Surely, another form of ‘palm-tree justice’ is in play when one reaches directly for the baton of freedom of contract as itself an underlying value of the law of contract? 188 Eerste Nasionale Bank (n 128 above) 319H-320B. (2008) 1 Constitutional Court Review 187

(ethically, in good faith) should be held to them. The meaning of freedom of contract in this way becomes constituted by a transformative understanding of good faith.

5 Good faith and the Constitution: Towards transforming189 the hegemony

5.1 Introduction

In the Preface to the fifth edition of The Law of Contract in South Africa, Richard Christie wrote as follows:

In the preface to the fourth edition I was rash enough to express the hope that the judges would further develop the concepts of good faith and public policy. This they have done, but not in the way I expected. Since Brisley v Drotsky, Afrox Healthcare Bpk v Strydom and South African Forestry Co Ltd v York Timbers good faith as an abstract value cannot be used to intervene in contractual relationships, but public policy still can. To achieve a just result there are advantages in using the concept of public policy which our law has developed over the centuries and has linked to the Constitution, rather than the less familiar concept of good faith. But it will be a mistake to regard the door as forever closed ...’190

Prof Christie has also suggested that the courts are likely to find that the Sasfin principle (namely that a court will not enforce a contract if its enforcement would be contrary to public policy) is the most serviceable instrument for developing the common law of contract to give effect to a provision of the Bill of Rights.191 Below we shall see that this has indeed become the preferred route when it comes to the enforceability (and of course also constitutionality) of contractual terms. But we shall also see that the Constitutional Court did indeed leave the door wide open in its 2007 decision in Barkhuizen.192 The question thus currently looming large in the South African law of contract is whether the ‘objective, normative value system’193 of the Constitution requires the introduction of a good faith jurisdiction in the law of contract as part of the Sasfin principle. The judgments that

189 The way in which I employ transformation in relation to hegemony closely relates to Drucilla Cornell’s description of transformation. For her transformation means that ‘a system can so alter itself that it not only no longer confirms its identity, but disconfirms it and, indeed, through its very iterability, generates new meanings which can be further pursued and enhanced by the sociosymbolic practice of the political contestant within its milieu.’ Cornell (n 100 above) 2. 190 RH Christie The law of contract in South Africa (5ed, 006) v. 191 n 190 above, 348. 192 Barkhuizen (n 13 above). 193 Carmichele v Minister of Safety and Security and Another (Centre for Applied Legal Studies Intervening) 2001 4 SA 938 (CC) (Carmichele) para 54. 188 Hegemony, freedom of contract, good faith and transformation supported Olivier JA’s opinion in Eerste Nasionale Bank certainly argue that it does. It is implicit in the judgments of the SCA that it does not. A related question is whether the above question requires reconsideration of the exceptio doli generalis as a technical remedy. The latter question was also before the Constitutional Court in 2007.

5.2 Crown Restaurant CC v Gold Reef City Theme Park (Pty) Ltd194

The facts in this case were substantially similar to the facts in Brisley save for the fact that the parties were both commercial entities. Crown Restaurant fell into arrears in terms of a written lease agreement with Gold Reef City Theme Park, which subsequently cancelled the lease and applied for an ejectment order. Crown Restaurant, however, contended that Gold Reef City ‘had verbally agreed to grant it an indulgence and allow it time within which to make proposals for settling the rental arrears.’195 This verbal indulgence, it was argued on behalf of Crown Restaurant, meant that the company implicitly waived its right to cancel the lease. Gold Reef City, however, initially denied the verbal agreement, then conceded that it did ‘allow’ Crown Restaurant to make proposals for settlement within a stipulated time period196 but ultimately contended that it was in any event entitled to rely on the non-variation clause in the written agreement. Gold Reef City further relied on a clause that stipulated that no indulgence granted precluded either party from enforcing any of its rights in terms of the agreement.197

Gold Reef City was successful in the Court a quo. An application for leave to appeal to the SCA was unsuccessful. Crown Restaurant subsequently applied for leave to appeal to the Constitutional Court.198 In the application for leave to appeal to the Constitutional Court, the Court was asked — for the first time in the proceedings - to consider the reintroduction of the exceptio doli generalis by way of a development of the common law under section 39(2) of the Constitution. The applicant contended that the remedy as an ‘equitable remedy is in line with constitutional values’199 and that on these particular facts it would — given the verbal agreement — be unconscionable for Gold Reef City to rely on the written agreement. There were thus two distinct issues before the court: first, whether the common law should be developed so as to reintroduce the exceptio doli generalis as an ‘equitable remedy in line with

194 2008 4 SA 16 (CC) (Crown Restaurant). 195 n 194 above, para 1 196 See n 194 above, Applicant’s Founding Affidavit para 23.2.(d). 197 n 194 above, para 2. 198 n 194 above, para 3. 199 As above. (2008) 1 Constitutional Court Review 189 constitutional values’ and second, if the answer to the first question is in the affirmative, whether on the facts the applicant could succeed with such a defence.

In its judgment the Constitutional Court pointed out that the trial court judge ‘was called upon to deal only with the waiver defence and did so. He was not invited to develop the common law of contract to promote the spirit, purport and objects of the Bill of Rights, nor to address any of the other constitutional issues now raised by the applicant.’200 After this statement the Court elaborated on the undesirability for it to sit as a court of first and last instance. It stated that the High Court and the SCA have a vital role to play when it comes to the development of the common law of contract and concluded that there are no compelling reasons for the Court to deal (as a court of first and last instance) with the issues raised by the applicant. It again warned litigants that ‘care should be taken to identify properly at the time of the institution of the proceedings which constitutional issue they wish to have addressed.’201 Accordingly, the application for leave to appeal was dismissed.

In order to evaluate critically the apparently unproblematic nature of the Constitutional Court’s dismissal of the application for leave to appeal, it is necessary that we recall here what was said in the decision of the Constitutional Court in Carmichele. In this case the applicant (plaintiff in the Court a quo) also did not invite either the trial court or the SCA to develop the common law in accordance with the spirit, purport or objects of the Bill of Rights.202 She argued the issue relating to the development of the common law for the first time in the Constitutional Court and contended that had the High Court and the SCA applied the relevant provisions of the Constitution in determining whether a legal duty (in the context of the law of delict) was owed to her, it would have developed the common law and would have found that such a legal duty exists.203 To quote from the judgment of Ackermann and Goldstone JJ:

Despite the failure by the applicant to rely directly upon the provisions of either section 35(3) of the [Interim Constitution] or section 39(2) of the Constitution in the High Court and SCA, counsel for the respondent did not object to this issue being raised in this Court. If covered by the pleadings, and in the absence of unfairness, parties are ordinarily not

200 Crown Restaurant (n 194 above) para 4 (emphasis added). 201 Crown Restaurant (n 194 above) para 6. 202 Carmichele (n 193 above) para 41. 203 Carmichele (n 193 above) para 28. 190 Hegemony, freedom of contract, good faith and transformation

precluded from raising new legal arguments on appeal. [Reference omitted.]204

Furthermore, in the course of the Carmichele judgment the Court stated the following:

In South Africa, the [Interim Constitution] brought into operation, in one fell swoop, a completely new and different set of legal norms. In these circumstances the courts must remain vigilant and should not hesitate to ensure that the common law is developed to reflect the spirit, purport and objects of the Bill of Rights. We would add, too, that this duty upon judges arises in respect both of the civil and criminal law, whether or not the parties in any particular case request the court to develop the common law under section 39(2).205

The Carmichele judgment made it clear that there is a general obligation on courts to develop the common law in accordance with the spirit, purport and objects of the Bill of Rights. The judgment also makes it clear that the duty exists whether or not the parties in any particular case explicitly invite the court to develop the common law. Furthermore, it is clear that the Carmichele judgment envisages that there might be circumstances where a court is obliged to raise the matter on its own and require full argument. This would be required presumably when the legal issue ‘can hardly be described as an insignificant one, lying at an exotic periphery’206 of the law in question.

In the light of these findings in Carmichele regarding the general obligation to develop the common law, it is not clear why the Constitutional Court considered it material (and thus fatal to the applicant’s application for leave to appeal) in Crown Restaurant that the trial court was not invited to develop the common law. Carmichele tells us that judges have a general obligation to consider the development of the common law in light of the spirit, purport and objects of the Bill of Rights, even when they are not invited by parties before them to do so. This is particularly important when a significant legal issue calls for constitutional resolution. By the time that this case came before the trial court it was clear that much dispute existed over, inter alia, the status of the exceptio doli generalis in

204 Carmichele (n 193 above) para 31. Also see para 38: ‘It does not appear to have been suggested that there was any obligation on the High Court or the SCA to develop the common law of delict in terms of section 39(2) of the Constitution.’ 205 Carmichele (n 193 above) para 36. Also see para 39: ‘we do not mean to suggest that a court must, in each and every case where the common law is involved, embark on an independent exercise as to whether the common law is in need of development and, if so, how it is to be developed under section 39(2). At the same time there might be circumstances where a court is obliged to raise the matter on its own and require full argument from the parties.’ (emphasis added.) 206 Carmichele (n 193 above) para 59. (2008) 1 Constitutional Court Review 191 our law, the role of good faith and the enforceability of a Shifren clause and/or a no indulgence clause in circumstances of subsequent verbal amendments/indulgences (that is, in the context of the waiver question).207 Moreover, it was (and still is) by no means clear how these matters are affected by the normative transformation of the legal order occasioned by the Constitution. Given the fact that the Constitution ‘brought into operation, in one fell swoop, a completely new and different set of legal norms’ it could certainly be asked whether the question of the development of the common law in this area should not have been raised by the trial court judge — as mandated by Carmichele — even though he was not explicitly invited to consider the development of the common law. It is thus not clear, under the circumstances, why further exploration by the trial court of the issues that were eventually raised in the Constitutional Court, was ‘understandably not undertaken’.208

In Carmichele the court held that the trial court and the SCA should indeed have considered the question regarding the effect of section 39(2) even though the parties did not raise the matter regarding development of the common law initially. The way in which the court dealt with this difficulty in Carmichele is very different from the way in which the Crown Restaurant matter was handled. Crown Restaurant was dealt with in terms of Constitutional Court rule 19(6)(b) which states that ‘[a]pplications for leave to appeal may be dealt with summarily, without receiving oral or written argument other than that contained in the application itself.’ In Carmichele the court heard argument on the merits of the dispute in accordance with the principle enunciated in what is now Constitutional Court Rule 19(6)(c). Inter alia, this move allowed the court to deal with the question regarding the prospects of success in considering whether to grant the application for special leave to appeal. In Crown Restaurant the court did not utilise rule 19(6)(c) and did not explicitly deal in its judgment with the question of the prospects of success. The refusal to make an order in terms of rule 19(6)(c) leaves one to assume that in the view of the court the prospects of success were improbable on the papers. Yet this is not the reason the court gives for denying the application for leave to appeal. The reason it gives for the failure of the application is that the applicant did not invite the trial court to develop the common law, that is, it did not raise the constitutional argument timeously. In Carmichele the remedy decided on in circumstances where the constitutional argument was not considered by the trial court or the SCA was to refer the matter back to the trial

207 The trial judge in fact referred to Brisley v Drotsky (n 13 above). See Gold Reef City Theme Park (Pty) Ltd v The Crown Restaurant CC unreported, case no: 2006/ 595 WLD, 7 June 2006 para 23 — where the controversy regarding good faith and the exceptio doli generalis arose in the context of waiver. 208 Crown Restaurant (n 194 above) para 6. 192 Hegemony, freedom of contract, good faith and transformation court. Admittedly, a difference between the Crown Restaurant matter and the Carmichele matter is that in Carmichele absolution was granted whereas in Crown Restaurant a full trial took place. In Carmichele, the fact that absolution was granted by the trial court provided a reason for referring the matter back to the trial court in addition to the referral on the grounds that the trial court had to be given the opportunity to consider section 39(2) in relation to the matter before it.209 Even though a full trial had taken place in Crown Restaurant, there is in principle no reason why the matter could not be referred back to the trial court in that instance for a consideration of the relevance of section 39(2). Given the controversy and uncertainty in relation to the contract law issues raised in this matter and given that many of these contentious issues are raised in the context of waiver, a case could well be made that it would have been in the interests of justice for the Constitutional Court to have granted leave to appeal, upheld the appeal, set aside the order of the trial court and to have referred the matter back to the trial court for a proper consideration of the section 39(2) obligation in relation to the exceptio doli generalis and its philosophical underpinning — good faith.210 This may be a costly exercise — as was the case in Carmichele — but it allows opportunity for the constitutional resolution of an issue that is crucially in need of it. However, this did not happen because the Constitutional Court was clearly of the view that on the papers there were not sufficient prospects of success even if the section 39(2) enquiry required development of the common law in such a way that the exceptio doli generalis be reintroduced.

From the point of view that even if there were prospects of success on appeal, a full trial had already occurred in the High Court and the cost to have the matter referred back to the trial court would have been prohibitive and unfair to the respondent, the Constitutional Court could itself have granted leave to appeal and could have decided the issue of the development of the common law on appeal. However, this was again not considered a viable option both presumably on the basis that there were not sufficient prospects of success on appeal and on the basis that it would have been extraordinary for the Constitutional Court to sit as court of first and last instance, because, as was indicated in Carmichele (referring to Bequinot)211 the Court would be put at a ‘grave disadvantage’ were it not to have ‘the benefit of a fully considered judgment from either the SCA or the High Court’212 on the point of the development of the common law. In the Crown Restaurant matter it is a pity that a point of law in such urgent need of constitutional resolution fell through the

209 Carmichele (n 193 above) para 81. 210 See Glover (n 104 above) 451. 211 S v Bequinot 1997 2 SA 887 (CC) (Bequinot). 212 Carmichele (n 193 above) para 58. (2008) 1 Constitutional Court Review 193 cracks as it were and led to a judicial impasse which, seemingly, could only be normalised/justified by invoking the idea that the common law can only be developed when parties invite the courts to develop the common law. And it is this idea that sits uncomfortably with the ‘general obligation’ to develop the common law in line with constitutional values enunciated in Carmichele.213

However, it is quite possible that there might be another reason for the Court’s decision not to grant leave to appeal in the Crown Restaurant matter. This reason might relate particularly to the controversy regarding the status of the exceptio doli generalis and good faith in our law of contract. In other words, the Constitutional Court’s refusal to hear argument on the merits and thus also its refusal to grant leave to appeal in Crown Restaurant might relate to the apprehensive way in which the substantive legal issue is generally treated in our law of contract. Simply put, my contention is that the Constitutional Court made a policy decision to deal with this case by not dealing with it. Implicit in the Court’s reasoning is that it would be inappropriate for it to sit as Court of first and last instance (that is without the benefit of judgments from the lower courts) on a matter as contentious and as vexed as the status of the exceptio doli generalis and the good faith approach in our law of contract. In and of itself, this refusal might well have amounted to a silent endorsement of the hegemonic order(ing) in our law of contract.

But when considered in the light of the other contract law case before the Constitutional Court in 2007, the judgment of the Constitutional Court in Crown Restaurant might have been influenced by the fact that a more appropriate contract law case for the Constitutional Court (that is, a case where the constitutional issue had already passed through the trial court and the SCA) was before it at the time of Crown Restaurant that would, in any event, provide the Court with the opportunity to pronounce on good faith (as the philosophical underpinning of the exceptio doli generalis), albeit by not pronouncing on it. In other words, the silence we see in Crown Restaurant regarding the exceptio (and for that matter, good faith as its philosophical underpinning) foreshadows the silence about good faith that we shall see in Barkhuizen. The silence in Barkhuizen, however, is different in that it is a silence that opens Kafka’s proverbial door of the law, whereas Crown Restaurant closed it.214

213 Carmichele (n 193 above) para 39. 214 See, in general, F Kafka The trial trans B Mitchell (1998). 194 Hegemony, freedom of contract, good faith and transformation

5.3 Barkhuizen v Napier

In this case the Constitutional Court had to consider the constitutionality of a time limitation clause in a short-term insurance policy.215 The case came before the Constitutional Court on appeal from the SCA which held that there was no evidence before it that indicated that the contract/policy containing the time-bar clause was not entered into freely and voluntarily in the exercise of the constitutional rights to dignity, equality and freedom. This led the court to the conclusion that constitutional norms and values could not operate to invalidate the bargain concluded.216

Before proceeding to the judgment of the Constitutional Court, a few important points should be made regarding the SCA’s Barkhuizen judgment not the least of which is that this judgment constitutes the latest in the series of judgments that attempt to provide a constitutional fit and justification for the hegemonic order(ing) of the law of contract in South Africa.217 This legitimation of the hegemonic order(ing) was most prominent in the Brisley majority judgment and separate concurring judgment of Cameron JA. Of his separate concurring judgment in Brisley Cameron JA says the following in his judgment in Barkhuizen:

Brisley rejected the notion that the Constitution and its value system confer on Judges a general jurisdiction ... or power to decide that contractual terms cannot be enforced on the basis of imprecise notions of good faith.218

Cameron JA then added that the judgments in Brisley and Afrox ‘affirmed that the common law of contract is subject to the Constitution’219 and that public policy, as rooted in the founding values of the Constitution — human dignity, the achievement of equality and the advancement of human rights and freedoms, non- racialism and non-sexism — is the proper instrument by way of which the courts can invalidate agreements that are offensive to the values of the new constitutional order.220

Cameron JA’s judgment openly maintains an explicit opposition between public policy and good faith. However, Cameron JA offers no

215 Barkhuizen (n 13 above) para 1. 216 Napier (n 13 above) para 28. The court held that this lack of evidence was attributable to the way in which the matter came before it, namely by way of ‘a terse statement of facts’ (para 9). 217 In a critique of the judgment of the SCA, Deeksha Bhana argues that the SCA reaffirmed the classical liberal position. See D Bhana ‘The law of contract and the Constitution: Napier v Barkhuizen (SCA)’ (2007) 2 South African Law Journal 269. 218 Napier (n 13 above) para 7 (emphasis added). 219 Napier (n 13 above) para 6. 220 Napier (n 13 above) para 7. (2008) 1 Constitutional Court Review 195 justification as to why good faith should be regarded as more of an imprecise and vague notion than the notions in which public policy are supposedly rooted: dignity, the achievement of equality, the advancement of human rights and freedoms, non-racialism and non- sexism. Dignity particularly has in the past been repeatedly criticised for being, an imprecise notion.221 Some have even gone so far as to describe equality as an empty concept.222 Are the concepts of non- racialism and non-sexism not also open concepts? And what about ‘the advancement of human rights and freedoms’? Moreover, can it not be said that in a constitutional democracy founded on these values, good faith — as the ethical element of contract — encompasses, in fact embraces, all these founding values at once? Hector MacQueen and Alfred Cockrell have argued that the South African courts have long conceded that public policy is a dynamic concept in the South African law of contract. It nevertheless, argues MacQueen and Cockrell, falls to be distinguished from good faith.223 Whereas good faith applies community standards ‘much more directly’ to the relationship between the contracting parties, public policy enunciates moral and social standards that are not necessarily dependent on the relationship between the parties.224 The SCA appears to have, at least implicitly, accepted a dynamic conception of public policy.225 At the same time it appears to have adopted the view that public policy should be distinguished from good faith, but not along the lines suggested by MacQueen and Cockrell.226 The SCA rejects any direct reliance on good faith on the basis that it is too imprecise a notion. Yet, it accepts direct reliance on public policy — a concept as dynamic and as imprecise as public policy. Such a basis for the opposition between good faith and public policy is incoherent and arbitrary. The point is that both good faith and public policy are open-ended norms, imprecise notions, text that calls for interpretation and justification much like the entire Bill of Rights and the founding values and ideals of the Constitution. Both good faith and public policy amount to devices by way of which freedom of contract (as embodying the value of autonomy) is limited (and thus constituted) in the South African law of contract.227 To privilege public policy on the basis that good

221 See for instance D Davis ‘Equality: The majesty of Legoland jurisprudence’ (1999) 116 South African Law Journal 398 413 and C McCrudden ‘Human dignity’ University of Oxford Faculty of Law Legal Studies Research Paper Series Working Paper 10/2006, April 2006. 222 P Westen ‘The empty idea of equality’ (1982) 95 Harvard Law Review 537. 223 H MacQueen & A Cockrell ‘Illegal contracts’ in R Zimmerman et al (eds) Mixed legal systems in comparative perspective: Property and obligations in Scotland and South Africa (2004) 143 173. 224 As above. 225 See Magna Alloys (n 85 above) 891 and Sasfin (n 75 above) 7I-J. 226 See South African Forestry (n 13 above) para 32. 227 Perhaps the privileging of public policy can be explained by reference to the privileging of the mantra that public policy generally favours the strict enforcement of contracts, whereas good faith does not, without more, favour such a privileging of liberal ideology. 196 Hegemony, freedom of contract, good faith and transformation faith is too imprecise a notion (and thereby suggesting that public policy is somehow more precise) sits uncomfortably with the context of — and openness to — an entire legal order that is founded on more or less ‘imprecise’ notions namely dignity, equality and freedom, notions to which meaning is given in a long process over time at the same time as their meaning always depends — at least to a certain extent — on context.

A further aspect of the SCA’s Barkhuizen judgment to which I want to draw attention is the uncritical — and alarming — assumption in the judgment that the hegemonic order(ing) of the law of contract in South Africa is already constitutionally entrenched. I use the word ‘alarming’ here given the vast economic disempowerment in South Africa and the grossly unequal, undignified and in fact unfree bargaining relations and situations in South Africa that the law of contract normalises and on a continuing basis sustains. From this perspective, it is hard to believe the SCA when it argues that the law of contract — in its current permutation — approximates the ideals of a transformative legal order in the best possible way, even when it concedes that public policy is a dynamic concept the meaning of which is to be determined by reference to the values and rights in the Constitution. Of course, some argue that the achievement of socio- economic equality is not — and should not be — something which a general law of contract can do much — if anything substantial — about. Yet, this approach fails to take into account what Woolman calls ‘the structured silence of [contractual] disputes that never make it to court.’228 Moreover, when the Constitution requires in section 39(2) that when the common law is developed it must be developed taking into account the spirit, purport and object of the Bill of Rights, the Constitution explicitly involves all law in the project of transformative constitutionalism. It thus explicitly requires that one takes a view of the common law as something other than the neutral backdrop against which interactions between legal subjects occur. Judgments such as those delivered in Brisley, Afrox and the SCA’s Barkhuizen judgment are attempts to provide the authoritative statements that the Constitution did not significantly change/affect — that is, it attempts to legitimise — the hegemonic order(ing) in the South African law of contract. The positions taken in these judgments of course offer a particular ideological reading of the Constitution. The ideology that prevails reads the Constitution in a classical liberal way against which it is easy to legitimise the hegemonic order(ing) of a classical liberal law of contract as in accordance with it. Cameron JA’s judgment in Brisley is exemplary of such a classical liberal ideology when it comes to the relationship between the Constitution

228 See S Woolman ‘Category mistakes and the waiver of constitutional rights: A response to Deeksha Bhana on Barkhuizen’ (2008) 1 South African Law Journal 10 11. (2008) 1 Constitutional Court Review 197 and the law of contract. To recall, Cameron JA held that ‘[t]he constitutional values of dignity and equality and freedom require that courts approach their task of striking down contracts or declining to enforce them with perceptive restraint ... contractual autonomy is part of freedom’229 and ‘shorn of its obscene excesses’230 it ‘informs also the constitutional value of dignity.’231 This judgment (like the judgment of the majority) is predicated upon an individualist ethic that is not only difficult to square with the altruistic ideals inherent in the spirit of a Constitution that aspires to a post-apartheid society; it also uncritically assumes that the strict enforcement of contracts (that is, a principled — and thus political — commitment to non- intervention by the courts in contractual matters) is in service of the constitutional values of dignity, equality and freedom.

The way in which Cameron JA’s judgment in Brisley was later used in Afrox was criticised along the lines suggested above.232 Lubbe for instance argued that when the foundational concepts of the Constitution are marshalled in the under-nuanced way in which they were in Afrox, the founding values of the Constitution serve to do only one thing and that is to dissipate pressure on the legitimacy of this hegemony.233 As Wagener argued, the linking of freedom of contract with dignity does not, without more, favour the strict enforcement of contracts.234 On the contrary, dignity might well be invoked to justify more active judicial control. Bhana and Pieterse also argued in this regard that ‘the [constitutional] value of freedom does not equate with complete individual liberty and does not found an independent right to unlimited contractual liberty.’235

Turning now to the Constitutional Court’s decision in Barkhuizen, it must be said that this decision of the Constitutional Court is, in a very real sense, not possible without the decisions and positions taken

229 Brisley (n 13 above) 35E. 230 Brisley (n 13 above) 35F. (I am also interested in understanding in which circumstances the court will regard freedom of contract as sufficiently ‘shorn of its obscene excesses’.) 231 Brisley (as above). 232 D Bhana & M Pieterse ‘Towards a reconciliation of contract law and constitutional values: Brisley and Afrox revisited’(2005) 122 South African Law Journal 865 874; G Lubbe ‘Taking fundamental rights seriously: The Bill of Rights and its implications for the development of contract law’ (2004) 121 South African Law Journal 395. 233 See Lubbe (n 232 above) 420. 234 S Wagener ‘A proposal for the reform of the law of contract brought about by the judicial development of novel naturalia’ (2005) Responsa Meridiana 19 39-40. Also see K Hopkins ‘The influence of the Bill of Rights on the enforcement of contract’ (2003) 425 De Rebus 25. 235 n 232 above, 879. Bhana and Pieterse also argue that the value of freedom ‘is reined in significantly by its interaction with the constitutional values of equality and dignity’. Again, I do not believe that freedom is somehow reined in by its interaction with equality and dignity. Rather, freedom is constituted by its interaction with equality and dignity. 198 Hegemony, freedom of contract, good faith and transformation in Brisley and Afrox and of course also in the SCA decision in Barkhuizen. Here, in the first case before the Constitutional Court on the law of contract, it had the benefit of the judgments of the lower courts. Things were as they should be.

In Barkhuizen the appellant argued that the time limitation constituted an unjustifiable limitation on the constitutional right to approach a court for redress.236 In the High Court the appellant argued on the basis of the direct horizontal application of the constitutional right and succeeded in his claim.237 However, the Constitutional Court confirmed in this case the approach followed in the decision by the SCA. The proper approach when dealing with the constitutionality of contractual provisions is to subject the term to the public policy test ‘as evidenced by the constitutional values, in particular, those found in the Bill of Rights.’238 This means, of course, that the route of indirect horizontal application is the preferred route when considering the constitutionality of a contractual arrangement.239 In determining whether a particular clause offends public policy as animated by the Constitution, the court held that the first question is whether the clause itself is reasonable.240 The second question, held the Court, is whether the clause should be enforced in light of the circumstances that prevented compliance with it.241 In this leg of the enquiry unequal bargaining power is a factor, but not the only factor, that plays a role in the determination of public policy.242 On the first question, the court held that the reasonableness of the provision is determined by weighing up the public interest in compliance with freely and voluntarily concluded terms (or, otherwise put, freedom of contract) against the public interest in the right of access to courts.243 The second question effectively involves an enquiry into the reasons why non-compliance is manifest. The onus to prove that it was impossible or unreasonable in the circumstances to comply with the clause rests on the party who alleges it.244 Thus, where the clause has been found not to be in violation of public policy (objectively speaking), it can still be escaped where the defendant proves good reason why there was failure to comply.

But there are two aspects regarding the Court’s statements on public policy on which I want to dwell for the moment because they

236 Barkhuizen (n 13 above) para 1. 237 Barkhuizen (n 13 above) para 9. 238 Barkhuizen (n 13 above) para 30. 239 For criticism see Woolman (note 102 above) 774-781. 240 Barkhuizen (n 13 above) para 59. 241 As above. 242 As above. 243 Barkhuizen (n 13 above) para 57. 244 Barkhuizen (n 13 above) para 58. (2008) 1 Constitutional Court Review 199 have implications for good faith. First, I want to emphasise that the Court reconfirmed the trite notion that the validity or otherwise of contractual terms is determined by asking whether the clause is in accordance with public policy. Second, the Court emphatically stated that ‘public policy represents the legal convictions of the community; it represents those values held most dear by the society.’245 It also held that ‘[p]ublic policy imports the notions of fairness, justice and reasonableness.’246 According to the Constitutional Court then, the enforceability or not of contractual terms are determined by the legal convictions of the community as represented by ‘the values that underlie our constitutional democracy’.247 This statement flies in the face of the Brisley judgment in which the Court explicitly and in the strongest terms overruled Davis J’s judgment in Mort because it attempted to make the enforceability of contractual terms dependent on the legal convictions of the community.

Recall, however, that Davis J referred to the legal convictions of the community in relation not to public policy but to good faith. It thus appears that, to the extent that Davis J did indicate that the enforceability of contractual terms depends on the legal convictions of the constitutional community — as that term is understood as the aspirational ‘community of the “ought to be”’248 — the Constitutional Court is in agreement with him and disagrees with the SCA. The only disagreement between the Constitutional Court and Davis J is the doctrinal name of these legal convictions. The Constitutional Court calls it public policy, Davis J calls it good faith. Can it be that the Constitutional Court in this case in fact merged public policy with good faith? Can it be that good faith after all forms part of public policy as Olivier JA argued in his much-maligned judgment in Eerste Nasionale Bank? The answer to this question is by no means clear given Ngcobo J’s emphatic statement that the facts of the case did not require the Court to answer the question whether, under the Constitution, the limited role for good faith expounded since Brisley, is appropriate.249 Yet, the Court attributes exactly the same meaning to public policy as the meanings that have been attributed to good faith. A striking example is the similarity between the definition by the Constitutional Court of public policy as ‘importing the notions of fairness, justice and reasonableness’250 and Olivier JA’s definition of

245 Barkhuizen (n 13 above) para 28. 246 Barkhuizen (n 13 above) para 73. 247 Barkhuizen (n 13 above) para 29. 248 See D Cornell Moral images of freedom (2007) 22. 249 Barkhuizen (n 13 above) para 82. This point is emphasised in the judgment of O’Regan J (Barkhuizen (n 13 above) para 120): ‘there is in my view no need for this court to consider in what circumstances a court may, in terms of the principles of contract, decline to enforce a time-limitation clause against a particular applicant based on the defences of impossibility or good faith. That difficult question can stand over for decision in an appropriate matter.’ 250 Barkhuizen (n 13 above) para 73. 200 Hegemony, freedom of contract, good faith and transformation good faith as realising the community’s legal convictions regarding propriety, reasonableness and fairness.251

On the other hand, the Court raised the good faith issue as it was dealing with the second leg of its test to determine the enforceability of the term on the ground of public policy: whether the terms are contrary to public policy in the light of the relative situation of the contracting parties. The court held that the applicability of the principle of good faith would depend on the reason advanced for non- compliance. However, the applicant did not disclose his reason for non-compliance. This made it impossible to say whether ‘enforcement of the clause against the applicant would be unfair and thus contrary to public policy.’252 In other words, the absence of a reason for non-compliance blocked, it seems, the consideration of the good faith requirement within the public policy test. In the absence of pleaded reasons for non-compliance, it would, according to the Court, be unfair to the respondent and contrary to pacta servanda sunt to excuse Barkhuizen from non-compliance with the time-bar clause.253 For these reasons, the majority accordingly dismissed the appeal.

On my reading it seems then at least probable that the court would have been willing to consider good faith (in the context of its formulated public policy test) had reasons been advanced for non- compliance with the term. In this way the court seems to both accept and reject the category distinction that MacQueen and Cockrell make between public policy and good faith. The Court clearly views good faith as the measure that would curb the enforcement of the term if enforcement ‘would be unfair or unjust to the applicant.’254 Given the Court’s statement regarding the role of unequal bargaining power in this leg of its public policy test, the suggestion seems to be that part of the good faith test will involve enquiring into the relative bargaining position of the parties and the way in which this impacted on the contractual arrangement. In this way it appears that the Court left the door wide open as regards good faith’s potential role as part of the public policy test. Moreover, this means that the exceptio doli generalis, albeit in drag, has risen from the grave255 — a grave in which it was in any event never interred. Another implication of the judgment is that the Constitutional Court has confirmed that the alleged exercise of freedom of contract in violation of both the rights in and the spirit, purport and objects of the Constitution, will not be enforced on grounds of public policy.

251 Eerste Nasionale Bank (n 128 above) 319B. 252 Barkhuizen (n 13 above) para 84. 253 n 13 above, para 85. 254 As above. 255 Glover (n 104 above) 455. (2008) 1 Constitutional Court Review 201

To what extent this confirmation will constitute internal limitations on freedom of contract — and thus amount to a synchronisation of freedom of contract with the other constitutional rights and freedoms in furtherance of a transformative understanding — depends on whether there will be a departure from/transformation of the hegemonic order(ing) we have interrogated. Moreover, by leaving the door of public policy open to good faith, the court has created the possibility that good faith may in the future again become constitutive of freedom of contract, rather than marginalised and seen as opposed to it. Again, everything here depends on the content that will be afforded to good faith. It goes without saying that it is always possible (and indeed likely) that good faith can be afforded weak content that would just further legitimise the hegemonic understanding of freedom of contract. To put it perhaps more dramatically, it is of course always an imminent danger that good faith will simply become a private and privatising device.256 For this reason it becomes necessary to defend a progressive and transformative version of good faith.

6 Negotiating otherwise — good faith, civic friendship and freedom of contract

When considering the justifiable potential content of good faith — as the ethical ideal that undergirds the law of contract — I believe it is necessary to take as one’s point of departure the nature and the aspirations of the reconstituted legal order within which the law of contract operates as an institution of the common law subordinated to the Constitution. If it is the case that the ‘proper’ approach (post- Barkhuizen) to the influence of the Constitution on the common law of contract is the public policy enquiry, then good faith — as the ethical element of contract, as ‘the law of the responsible transaction’257 — can and must come to play a role in the determination of the public policy requirement in order for the spirit of the Constitution as a whole to have a better chance of a transformative effect in the law of contract. In other words if there is to be talk of and hope for what Johan van der Walt has called progressive indirect horizontal application258 then we cannot afford to lose sight of good faith and the transformative potential it holds for the South African law of contract. Good faith holds this

256 See JWG van der Walt ‘Die toekoms van die onderskeid tussen die publiekreg en die privaatreg in die lig van die horisontale werking van die grondwet (deel 2)’ 2000 Tydskrif vir die Suid-Afrikaanse Reg 605 612. 257 J Derrida ‘On forgiveness’ in J Derrida On cosmopolitanism and forgiveness (2001) 56. 258 See in general JWG van der Walt ‘Progressive indirect horizontal application: Towards a co-operative relation between common-law and constitutional jurisprudence’ (2001) South African Journal on Human Rights 341. 202 Hegemony, freedom of contract, good faith and transformation transformative potential because, more than any other normative contractual device, it is good faith that invokes, as an ethical notion, respect for the dignity of contracting parties, precisely because, as MacQueen and Cockrell suggest, good faith focuses the enquiry explicitly on the relationship between the particular contractual parties. This respect for dignity when it comes to legal relations is a core aspect of the newly constituted legal order and its aspirations to civic friendship.

I would like to suggest, in conclusion then, that the Constitution aspires to an ideal of civic friendship (a ‘politics’ of friendship if you will) which requires, mandates and demands negotiation (contracting) otherwise — that is, with respect for and consideration of the other contracting party. I believe that the Constitution aspires to the post-liberal ideal of civic friendship precisely because of its foundational injunction to respect the dignity of all others.259 This, I would argue, requires that we afford content to good faith in the law of contract that would maintain it as a post-liberal concept. I want to emphasise here that I do not consider the dignity of which the Constitution speaks (and on which the instant conception of civic friendship and good faith turns) to be inevitably fated as a radically individualistic concept that, moreover, forms the basis of exclusively negative freedom. This argument is possible because of the links drawn in our early dignity jurisprudence between dignity and ubuntu.260 Drucilla Cornell draws the further link between ubuntu, (Kantian) dignity and friendship as follows:

In Kant, I am a friend to myself because of the dignity of my humanity. Under ubuntu, I am a friend to myself because others in my community have already been friends to me, making me someone who could survive at all, and therefore be in the community. It is only because I have always been together with others and they with me that I am gathered together as a person and sustained in that self-gathering.261

It was Hannah Arendt who, in modern political theory, resuscitated and defended a ‘politics of friendship’ relying heavily on the work of Aristotle.262 Arendt’s theory of civic friendship offers an account of politics that is both post-individualist and post-liberal. In fact, Arendt’s theory of civic friendship invokes an understanding of citizenship as a radically ‘horizontal allegiance to friends.’263 She sees in this understanding of political life the possibility of ‘words that

259 Constitution sec 10. 260 See, for instance, S v Makwanyane 1995 3 SA 391 (CC) paras 224, 308 & 311. 261 D Cornell ‘A call for a nuanced constitutional jurisprudence: Ubuntu, dignity and reconciliation’ (2004) 19 SA Public Law 666 669. 262 H Arendt The human condition ed M Canovan (1998) 243. 263 ME Stortz ‘Beyond justice: friendship in the city’ (1994) XIV(4) Word & World 409 414 (emphasis added). (2008) 1 Constitutional Court Review 203 are not empty and deeds that are not brutal.’264 As Martha Stortz comments on Arendt’s concept of civic friendship:

Civic friendship of all sorts emphasises the interdependence of citizens in public life. It articulates a horizontal understanding of citizenship, which prizes the relationship to another citizen and places that relationship at the centre of civic life.265

Most importantly, though, this form of friendship is not at all about the association of the self and the same. At its heart, Arendt’s understanding of civic friendship turns on plurality — that is, difference — because it is plurality in Arendt that is constitutive of the political.266 And plurality conditions us in the sense that our very individuality only takes shape through our recognition that we share the world with others.

This immediately implicates the idea of respect, for without respect there can be no durability for plurality. And as Arendt puts it in The human condition: ‘what love is in its own, narrowly circumscribed sphere, respect is in the larger domain of human affairs.’ ‘... [R]espect’, so she continues,

is a kind of ‘friendship’ ... it is a regard for the person from the distance which the space of the world puts between us, and this regard is independent of qualities which we may admire or of achievements which we may highly esteem.267

For Arendt then there can be no talk of civic friendship without this respect, which, to emphasise, does not depend on conditions of the self and the same. However, Arendt conceded that the modern age is characterised by the loss of respect and this ‘constitutes a clear symptom of the increasing depersonalisation of public and social life.’268 It is also because of Arendt’s acute awareness of the implications of this depersonalisation due to the loss of respect that she insists famously at the end of the preface to The origins of totalitarianism that, in the aftermath of the disasters of modernity ‘human dignity needs a new guarantee which can be found only in a new law on earth.’269

264 As above. 265 As above. 266 Arendt (n 262 above) 7. 267 n 262 above, 243. 268 As above. 269 H Arendt The origins of totalitarianism (1973) ix. 204 Hegemony, freedom of contract, good faith and transformation

If we recall Derrida’s assertion in Politics of Friendship that there is no friendship without respect of the other, that ‘the co-implication of responsibility and respect can be felt at the heart of friendship’270 we will begin to see the trembling overlap between the ethical, the political and the juridical in the South African Constitution in which respect for the dignity of all others comes to play a constitutive role. Justice O’Regan has referred explicitly to respect and responsibility as ‘the building blocks of the Constitution.’271 And this respect is owed to the equal human dignity of everyone.272 On my reading, this unconditional horizontality of the South African Constitution appeals to a prioritisation of Arendt’s, but also Derrida’s, vision of friendship in which the Greek form of love as philia is instituted as a socio- political concept but here in holding out the promise of a beyond to any condition of fraternity or ‘confraternisation’.273 Shin Chiba argues that it is this form of love that permeates Arendt’s search for a new and durable vinculum or social bond, which, let us recall, is the original name for law.274 The implication of this institutionalisation is that it is in fact respect, as phronesis and praxis for the Other275 — respect that resides in love of the Other — that becomes constitutive of the legal order.

Judge Dennis Davis has been foremost amongst those who have argued that the notion of respect for equal dignity (civic friendship) is translated into the law of contract by way of good faith.276 He has also opposed this concept to South Africa’s political past. As Davis J puts it:

this concept of good faith is congruent with the underlying vision of our Constitution ... to the extent that our Constitution seeks to transform our society from its past, it is self-evident that apartheid represented the very opposite of good faith ... Our Constitution seeks to develop a community where each will have respect for the other ... Whatever the uncertainty, the principle of good faith must require that the parties act honestly in their commercial dealings. Where one party promotes its own interests at the expense of another in so unreasonable a manner as to destroy the very basis of consensus between the two parties, the

270 J Derrida Politics of friendship trans G Collins (1997) 252. Of course, Derrida is problematising Kant’s formulations with regard to friendship here, notably Kant’s separation of love from friendship. Time and space does not permit a discussion of this deconstruction. 271 C O’Regan ‘The three r's of the Constitution: Responsibility, respect and rights’ (2004) Acta Juridica 86. 272 As above. 273 See S Chiba ‘Hannah Arendt on love and the political: Love, friendship, and citizenship’ (1995) 57 The Review of Politics 505 512-513. Also see Derrida (n 267 above) viii. 274 n 273 above, 509. 275 Douzinas & Geary (n 68 above) 40. 276 But also see Grové (n 130 above) 689. (2008) 1 Constitutional Court Review 205

principle of good faith can be employed to trump the public interest inherent in the principle of the enforcement of a contract.277

In his judgment in Mort Davis J quoted the words of Reinhard Zimmermann278 in relation to this aspect:

the principles of equality and dignity require that the parties to a contract do none other but ‘adhere to a minimum threshold of mutual respect in which the unreasonable and one-sided promotion of one’s own interest at the expense of the other infringes the principle of good faith to such a degree as to outweigh the public interest in the sanctity of contracts’.279

This is the formulation for which Barkhuizen leaves the door open. Ultimately in our context, respect for the Other resides, as Davis J has also indicated, in the acknowledgement that:

[a] transformative constitution needs to engage with concepts of power and community ... the concept of contractual autonomy within the concept of a community of contracting persons must mean something distinct from a libertarian connotation, particularly if the concept of ubuntu is to play any role in our law.280

These statements about the law of contract as law subordinated under the new Constitution are explicit interpretations of the constitutional mandate as requiring a re-emphasis on the ethical element of contract in the furtherance of a post-liberal or positive freedom of contract. A freedom of contract that comes to understand that conduct cannot be characterised as free when it disrespects/ violates dignity,281 when it pretends that contract is a relation between things and not between persons, when it does not proceed according to respect for whoever is on the other side of the negotiation. To quote Colombo: ‘Good faith implies a developed sense

277 Shoprite Checkers (Pty) Ltd v Bumpers Schwarmas CC and Others 2002 6 SA 202 (C) (Shoprite) 215G-216A. 278 R Zimmermann ‘Good faith and equity’ in R Zimmerman and D Visser (eds) Southern cross: civil and common law in South Africa (1996) 259-260. 279 n 168 above, 475B-C. 280 Advtech (n 95 above) 388A. 281 The mantra in Ackermann J’s celebrated minority judgment in Ferreira v Levin NO and Others; Vryenhoek and Others v Powell NO and Others 1996 1 SA 984 (CC) (Ferreira) para 49 ‘to deny people their freedom is to deny them their dignity’ should perhaps then be reconsidered in this light because, on this understanding in the law of contract, to deny people their dignity is to deny one’s own freedom (of contract). 206 Hegemony, freedom of contract, good faith and transformation of community and a high level of awareness of personal responsibilities towards society.’282

When Lubbe argues that our understanding of dignity in contract should be informed by Kant’s precept that people should not

act contrary to the equally necessary self-esteem of others, as human beings, that is, [they are] under obligation to acknowledge, in a practical way, the dignity of humanity in every other human being283 he already acknowledges the relational aspect of contract that would amount to a Hegelian version of contract as ‘reciprocal recognition.’284 Rosenfeld explains that

under this paradigm contract is not only compatible with the reconciliation of individual autonomy and communal values, but it is also suited to foster the mutual determination of the individual and — or, perhaps more precisely, against — the communal.285

In turn this understanding of contract in terms of relationality and mutual recognition bears marked resemblances with Nancy’s analysis of existence as ‘being singular plural.’286 Once we understand that both good faith and freedom of contract are essentially relational concepts it becomes possible to think a law of contract beyond continuing and continuous exploitation, subservience and injustice — beyond ‘our proud and deadly dualisms.’287

The legislature has made substantial inroads in this regard and, at the time of writing the Consumer Protection Bill had been approved by Parliament and is awaiting the President’s signature. Amongst various other consumer protection mechanisms the Bill explicitly prohibits ‘unfair, unreasonable or unjust’ contracting.288 In addition, courts will be required in terms of clause 52(3) to decide whether a contract is unconscionable, unjust, unreasonable or unfair. Although guidelines for the determination of the unreasonableness, injustice or unfairness are set out in clause 48(2) they are explicitly intended not

282 S Colombo ‘Fascism, community and the paradox of good faith’ (1994) 113 South African Law Journal 482 482. 283 Lubbe (n 232 above) 422, quoting I Kant The Metaphysics of Morals M Gregor ed (1996) 209. 284 M Rosenfeld ‘Hegel and the dialectics of contract’ (1989) 10 Cardozo Law Review 1199 1200. 285 As above. 286 JL Nancy Being singular plural (2000). In this work Nancy denounces the liberal idea of an individual atomistic self as the subject of ontology. As Nancy puts it: ‘That which exists, ... coexists because it exists. The co-implication of existing [l’exister] is the sharing of the world.’ 287 G Rose Mourning becomes the law (1996) 76. 288 Consumer Protection Bill 2008 clause 48(1) available at http:// www.thedti.gov.za/ccrdlawreview/DraftConsumerProtectionBill.htm (accessed 20 June 2008). (2008) 1 Constitutional Court Review 207 to limit the generality of the section prohibiting unreasonable, unfair or unjust contracting.289 This Bill thus already points to a future law of contract that will be required to be far more concerned than it has been up to now, with good faith as the ethical element of contract and thus with the spirit of the ideal of civic friendship in South Africa.

7Conclusion

At the heart of contract lies the idea that I have an interest in something of yours and that you have an interest in something of mine. The hegemonic capitalist over-emphasis on the things and the utter neglect of the persons who have these things, has provided an extremely distorted version of what the word ‘interest’ in the above formulation originally entails. Levinas takes his cue in his re-emphasis on the original meaning of the word ‘interest’ from Hegel, for whom contract exists as reciprocal recognition and freedom of contract exists in the free willing of one’s duty on the road to self- actualisation.290 In his essay ‘Sociality and money’,291 Levinas reminds us that the word interest is a composite of ‘inter’ that is, of ‘between’ and of ‘esse’, that is ‘being’. For Levinas, essence is always interestedness or differently put being is always inter-estedness. Thus, to have a contractual interest always already implies

289 For incisive criticism and important suggestions see TJ Naudé ‘Unfair contract terms legislation: The implications of why we need it for its formulation and application’ (2006) 17 Stellenbosch Law Review 361 374. Also see TJ Naudé ‘The use of black and grey lists in unfair contract terms legislation in comparative perspective’ (2007) 124 South African Law Journal 128. In the latter article (135- 136) Prof Naudé criticises this author for — in her opinion, exclusively — supporting an open-ended version (based on good faith) of the Unfair Contractual Terms Bill proposed by the Working Committee of the South African Law Commission (as it then was) in SA Law Commission Discussion Paper 65 Unreasonable Stipulations in Contracts and the Rectification of Contracts (July 1996). See AJ Barnard ‘A critical legal argument for contractual justice in the South African law of contract’ unpublished LLD dissertation, University of Pretoria, 2005 241. Apart from the fact that there is no disagreement between myself and Prof Naudé that, to quote Prof Naudé, ‘the use of non-exhaustive guidelines and a grey list by a court cannot be equated with the “mechanical application of precedent” and does not preclude a decision guided by “the relational, the collective and the transformative” at all’, I wish to make it clear that my doctoral study focused on the rhetorical history and lexicon of adjudication in the South African law of contract (which was and still is primarily based on the application of precedent). This mode, I argued, was, and still is in danger of perpetuating an individualistic interpretation of both future guidelines or lists of suspect terms and an individualistic reification of open-ended norms. This is an issue Prof Naudé herself raises — but then dismisses — as a potential problem with so-called ‘grey lists’. Where Prof Naudé and I part company as I see it, is that whereas I support Kennedy’s belief that individualism prefers the rule form and the commitment to formal realisability (which in turn fuels the false belief in neutrality), Prof Naudé believes that the rule form does not necessarily follow individualist political morality. 290 See GWF Hegel Elements of the philosophy of right ed A Wood (1991) 104. 291 E Levinas ‘Sociality and money’ (2007) 16 Business Ethics: A European Review 203 206-207. 208 Hegemony, freedom of contract, good faith and transformation togetherness, being between and with. Levinas admits that ‘in the concrete of everyday life, human interestedness always already conceals itself’292 but at the same time this concealment points to a beyond to it. This is the astonishing possibility of the human being between, or amongst, beings: the possibility of unconcealment, of giving his place, of sacrificing himself for the other, of the goodness of giving, ‘the positivity of an attachment to being as being for the Other.’293

What is needed in the South African law of contract is the best possible negotiation between these two dimensions of the human condition — inter-est and dis-inter-estedness. The hegemonic order(ing) of the South African law of contract at present does not allow for the best possible negotiation between human interestedness and human dis-interest. This is the case because ‘individualism is the structure of the status quo.’294 In this sense it does not allow for real transformation. This is the case because the obsession with individualism and the rule form does not allow for that which is inherent in transformation: the turn towards the Other. Yet, as Cornell reminds us: ‘[t]ransformation is demanded of us precisely because there is no self-enclosed subject who can truly cut herself off from the Other. We are constantly being challenged by otherness’.295 It is only once the (im)possibility of the ethical relation is explicitly inscribed and described in the South African law of contract by way of at least the explicit incorporation and exploration of good faith in the public policy/legality requirement that the best negotiation/the negotiation otherwise, becomes possible. The Constitutional Court’s decision in Barkhuizen and the Consumer Protection Bill allow for this possibility. To that extent the possibility of the impossible ethical relation can be part of the law’s work.

In the second of the films in the famous Matrix trilogy, the lead character, Neo, consults a present day (or perhaps futuristic) version of the ancient Greek oracle, which in the film is represented as a wise older woman. Let me end with the counsel the Oracle gives to Neo. Emphatically she states: ‘I am interested in one thing Neo — the future. And believe me, the only way to get there is together.’296

292 n 290 above, 204. 293 Levinas (n 290 above) 205. 294 Kennedy (n 6 above) 1775. 295 Cornell (n 100 above) 41. 296 A & L Wachowski (dirs) The Matrix reloaded (2003). CLEARING THE INTERSECTION? ADMINISTRATIVE LAW AND LABOUR LAW IN THE CONSTITUTIONAL COURT

Cora Hoexter*

The dismissal of employees in the public sector is as much a matter of administrative law as it is one of labour law — or so we all thought, encouraged by the judgment of the Appellate Division in Administrator, Transvaal v Zenzile.1 But that decision comes from an era in which labour law was notable for its failure to protect employees in the public sector and administrative law had to be called on to perform this role.2 Much has changed since then, and in recent years there has been debate about the continuing relevance of administrative law in cases of dismissal and other employment- related conduct.

In 2007 the Constitutional Court was asked to settle the debate in Chirwa v Transnet Ltd & Others,3 a case of public-sector dismissal that called attention to Zenzile once more. The Court also dealt with two other cases of dismissal that touched on questions pertinent to administrative law. Masetlha v President of the Republic of South Africa & Another4 raised the issue of procedural fairness in the dismissal of a high-ranking public official. Sidumo & Another v Rustenburg Platinum Mines Ltd & Others5 arose out of a dismissal in the private rather than the public sector, but administrative law was again relevant by virtue of an arbitration award by the Commission for

* Professor of Law, University of the Witwatersrand, Johannesburg. I am grateful to those who read and commented on the first draft of this article: Clive Plasket, the participants in the CCR Conference of 6-7 August 2008 and the CCR’s anonymous reviewers. Their perceptive comments were of great assistance to me, and they are not to blame for any deficiencies that remain. 1 1991 1 SA 21 (A) (Zenzile). 2 The legislation of the time, the Labour Relations Act 28 of 1956, excluded state employees from its ambit. 3 2008 29 ILJ 73 (CC), handed down on 28 November 2007 (Chirwa). 4 2008 1 SA 566 (CC), handed down on 3 October 2007 (Masetlha). 5 2008 2 SA 24 (CC), handed down on 5 October 2007 (Sidumo).

209 210 Clearing the intersection? Administrative law and labour law in the CC

Conciliation, Mediation and Arbitration (CCMA). In all, the Court had three opportunities in 2007 to explore the intersection of labour law and administrative law and to elucidate the relationship between sections 23 and 33 of the Constitution — and between the statutes that are supposed to give effect to those rights, the Labour Relations Act 66 of 1995 (LRA) and the Promotion of Administrative Justice Act 3 of 2000 (PAJA).

As I shall show in this article, the results are interesting; but they are also rather disappointing, at least from the perspective of an administrative lawyer. In my commentary on these three cases (in the order Sidumo, Chirwa and Masetlha) that is the perspective I adopt. My aim is to elucidate aspects of these cases as they touch on administrative law and, in the process, to make three main points.

First, it seems to me that from the various judgments of the Constitutional Court no coherent picture emerges of this intersection or of the relationship between the rights concerned. Indeed, in their constitutional approach or ethos the Sidumo and Chirwa majority judgments seem quite different, notwithstanding that the two cases were decided by almost identical benches and within weeks of each other. A theme emerging strongly from Sidumo, and to a lesser extent from Masetlha, is that fundamental rights cannot be hermetically sealed from one other. In Sidumo labour law and administrative law intermingle and converge, and in Masetlha administrative law in the broad sense is permitted to solve a problem that labour law apparently fails to address. In Chirwa, by contrast, the majority insists on the strict compartmentalisation of fundamental rights and on the pre-eminence of one right to the exclusion of the other. Sections 23 and 33 of the Bill of Rights are placed in separate jurisdictional boxes, and the attitude seems to be that labour law and administrative law must have nothing to do with one other.

Secondly, in Chirwa the various approaches to the ‘administrative action’ issue are all problematic to some extent, and the majority view on the question of jurisdiction is wholly unconvincing. While the Court certainly managed to tidy up the busy intersection of labour law and administrative law in this case, it did so at the expense of the Constitutional Court’s own precedent — and apparently without much faith in the success of its efforts.

Thirdly, whatever labour lawyers may think of the majority judgment in Masetlha, administrative lawyers are likely to be perturbed by it. For us it is a decision that appears to set the law of procedural fairness back twenty years. (2008) 1 Constitutional Court Review 211

1 Sidumo

The appellant, Mr Sidumo, had been employed by Rustenburg Platinum Mines Ltd (the mine) for almost fifteen years when he was dismissed for negligent failure to apply established search procedures at a redressing station. He contested his dismissal and referred the dispute to the CCMA in terms of section 191 of the LRA. Conciliation failed and the matter went for compulsory arbitration. The commissioner found that while Sidumo was indeed guilty of misconduct, dismissal was not a fair or appropriate sanction; and he was reinstated with three months’ compensation.

The mine applied for review of the award under s 145 of the LRA. The Labour Court dismissed its application, and the Labour Appeal Court dismissed an appeal to it. The mine was more successful on subsequent appeal to the Supreme Court of Appeal, which overturned these two decisions and found the dismissal to have been fair. In Rustenburg Platinum Mines Ltd (Rustenburg Section) v Commission for Conciliation, Mediation and Arbitration6 it held that the commissioner’s decision to reinstate Sidumo had not been rationally connected to the information before him.

The reasoning behind this conclusion is significant, for section 145 of the LRA lists only four grounds for the review of arbitration awards and irrationality does not feature explicitly amongst them.7 However, Cameron JA found for the Court that the PAJA had extended the grounds of review available to parties to CCMA arbitrations — there being no doubt that CCMA arbitrations qualified as administrative action.8 In this regard he drew on Carephone (Pty) Ltd v Marcus NO,9 a pre-PAJA case in which the Labour Appeal Court held that section 145 was suffused with the standard of justifiability in section 24 of the interim Constitution10 (the predecessor to section 33). Cameron JA held that as general legislation relating to administrative action the PAJA had ‘superseded the LRA’s specialised enactment within the [labour] field’,11 and it thus subsumed and overrode the more restrictive grounds of review in section 145 of the LRA.12 An unrelated but equally significant holding of the Supreme Court of Appeal in

6 2007 1 SA 576 (SCA) (Rustenburg). 7 Section 145(2)(a) allows review to be obtained for misconduct on the part of the commissioner; for a gross irregularity in the conduct of the arbitration proceedings; and where the commissioner exceeded his powers. Section 145(2)(b) provides for review where ‘an award has been improperly obtained’. 8 Rustenburg (n 6 above) para 25, though without detailed analysis of the question. 9 1999 3 SA 304 (LAC) (Carephone) paras 15 and 30 et seq. 10 Constitution of the Republic of South Africa, 1993. 11 Rustenburg (n 6 above) para 24. 12 Rustenburg (n 6 above) para 25. 212 Clearing the intersection? Administrative law and labour law in the CC

Rustenburg was that CCMA commissioners should give some deference to the employer’s choice of sanction.13

An appeal to the Constitutional Court was unanimously upheld and the award of the commissioner restored. The Court considered two main issues. On the first it held that a commissioner is not required to defer to the decision of the employer but rather to consider all the relevant circumstances.14 On the second issue, the applicability of the PAJA, the Court agreed on the result but was divided in its approach.

Writing for a narrow majority, Navsa AJ held that arbitral decisions of CCMA commissioners were indeed administrative action in terms of section 33 of the Constitution,15 but that the PAJA was not applicable to such decisions. In his judgment the LRA had to be regarded as specialised legislation alongside the PAJA — that is, as legislation dealing with ‘administrative action’ in the labour sphere.16 As such, the LRA had to be interpreted in accordance with section 33.17 Thus section 145 of the LRA was suffused with the content of section 33, including the constitutional standard of reasonableness18 — just as in Carephone the Labour Appeal Court had found section 145 to be suffused with the former standard of justifiability. Applying the new standard, Navsa AJ found that the commissioner’s decision was a reasonable one: it was not a decision ‘that a reasonable decision- maker could not reach’.19 Four justices concurred in his judgment.20

Ngcobo J, who had the support of three colleagues,21 arrived at a similar conclusion by means of different reasoning. In his judgment CCMA arbitrations were not administrative but judicial action, which meant that section 33 had no role to play.22 However, CCMA commissioners exercised public power and were thus constrained by other constitutional requirements, including those found in the doctrine of legality — an aspect of the rule of law — and in sections 23 and 34 of the Bill of Rights.23 These constraints all informed the interpretation of section 145(2)(a) of the LRA and considerably

13 Rustenburg (n 6 above) para 42 et seq, and see para 48(d). 14 Sidumo (n 5 above) para 79 in the judgment of Navsa JA, and see para 161 in the judgment of Ngcobo J. 15 Sidumo (n 5 above) para 88. 16 Sidumo (n 5 above) para 89. 17 Sidumo (n 5 above) paras 89 and 91. 18 Sidumo (n 5 above) paras 106 and 110. 19 Sidumo (n 5 above) para 119, using the words of O’ Regan J in Bato Star Fishing (Pty) Ltd v Minister of Environmental Affairs 2004 4 SA 490 (CC) (Bato Star) para 44. 20 Moseneke DCJ and Madala, O’Regan and Van der Westhuizen JJ. 21 Mokgoro, Nkabinde and Skweyiya JJ. 22 Sidumo (n 5 above) paras 207-40. 23 Sidumo (n 5 above) para 260. (2008) 1 Constitutional Court Review 213 expanded the three grounds of review listed there.24 Ultimately, however, the facts did not establish any of those grounds of review.25

O’Regan J, who concurred in the judgment of Navsa AJ, wrote a separate judgment in which she responded to Ngcobo J and gave additional reasons for supporting the view that CCMA arbitrations are administrative action.26

Sachs J, too, produced a separate judgment in which he found himself ‘in the pleasant but awkward position of agreeing with colleagues who disagree with each other’.27 He saw the judgments of Navsa AJ and Ngcobo J as being animated by the same goal — to determine what standard of conduct the Constitution expects of a CCMA commissioner — and as substantially in agreement on the interests and values involved.28 The key to the case, as he saw it, was to make those interests and values explicit. This he proceeded to do, identifying values such as fair dealing and rationality in sections 23, 33 and 34. His application of these standards to the commissioner’s conduct then led him to the same conclusion as his colleagues.

Of the many points of interest raised by Sidumo, I wish to highlight three features for the attention of constitutional lawyers in general and administrative lawyers in particular.

The first is the slender majority achieved on the ‘administrative action’ issue — an improvement on the bewildering inconclusiveness of New Clicks,29 but surely further evidence of the intrinsic difficulty of deciding what is and what is not administrative action. That difficulty exists even when, as here, the Court is relying on the general conception of administrative action developed judicially under section 33 rather than the more nit-picking definition in the PAJA. I think the majority gets it right. Navsa AJ reasons convincingly on the basis of significant differences between a court of law and a tribunal such as the CCMA, including the relative informality of its processes, the absence of a system of binding precedent and the less secure tenure of its presiding officers,30 and O’Regan J adds that ‘it is entirely consistent with our constitutional order that the

24 Sidumo (n 5 above) paras 256-77, and for the grounds of review see n 7 above. 25 Sidumo (n 5 above) para 289. 26 Sidumo (n 5 above) paras 122-41. 27 Sidumo (n 5 above) para 146. 28 As above. 29 Minister of Health v New Clicks South Africa (Pty) Ltd 2006 2 SA 311 (CC) (New Clicks), a case concerning pharmaceutical regulations. Two members of the Court found that regulation-making in general amounted to administrative action under the PAJA, while three others regarded the Act as applicable but on a narrower basis. Five members found it unnecessary to decide the question, and one found that subordinate legislation was generally governed by the principle of legality rather than the PAJA. 30 Sidumo (n 5 above) para 85. 214 Clearing the intersection? Administrative law and labour law in the CC procedures and decisions of the CCMA should be lawful, reasonable and procedurally fair’.31 But the merits are not my main concern here. I wish merely to note that Sidumo bears out (yet again) the Court’s gloomy prognostication in the SARFU case32 about the difficult boundaries that would have to be drawn in the administrative action inquiry.

The second feature is the further development, or complication, of administrative law by the official recognition of yet another pathway to review. The pre-democratic era knew only two such routes: ‘ordinary’ review of the decisions of public bodies at common law33 and the special statutory review jurisdiction created occasionally by the legislature (of which section 145 of the LRA was a prime example). With this second type of review the grounds specified by the legislature, and indeed the remedies available, could be wider or narrower than those associated with ordinary review.34 The democratic constitution brought with it a new constitutional pathway to administrative-law review in the form of section 33 (previously section 24 of the interim Constitution), duly informed by the common law; but this third route was reserved for exercises of ‘administrative action’. A few years later the Constitutional Court identified a fourth pathway, the principle of legality, for the residual review of all those exercises of public power that did not qualify as administrative action,35 and proceeded to develop the principle so that it mimics the content of section 33 to a considerable extent.36 Then came the PAJA, the national legislation enacted to give effect to section 33 and supposedly the default pathway to review37 — but one that often cannot be used because the conduct being reviewed does not qualify as administrative action in terms of the Act’s very demanding definition. That makes five pathways, as we still have common-law review (though it is used only for exercises of private power these days)38 and presumably we still have instances of special statutory review in the traditional sense.

31 Sidumo (n 5 above) para 140. 32 President of the Republic of South Africa v South African Rugby Football Union 2000 1 SA 1 (CC) (SARFU) para 143. 33 Johannesburg Consolidated Investment Co Ltd v Johannesburg Town Council 1903 TS 111 115. 34 See Nel NO v The Master 2005 1 SA 276 (SCA) para 23 and the discussion by Froneman J in the court a quo: Nel NO v Master of the High Court ECD case no 1633/2001 (undated) 8-9, unreported. 35 Fedsure Life Assurance Ltd v Greater Johannesburg Transitional Metropolitan Council 1999 1 SA 374 (CC) (Fedsure) paras 56-9. 36 SARFU (n 32 above) and in Pharmaceutical Manufacturers Association of SA: In re Ex parte President of the Republic of South Africa 2000 2 SA 674 (CC) (Pharmaceutical Manufacturers Association). 37 Bato Star (n 19 above) para 25. 38 See C Hoexter Administrative law in South Africa (2007) 119ff. (2008) 1 Constitutional Court Review 215

The majority judgment in Sidumo would seem authoritatively to have opened up a sixth route: special statutory review under the LRA for labour-related action that also qualifies as administrative action under section 33. This, then, is not special statutory review in the traditional sense — review, wide or narrow, on whatever grounds the legislature specifies within a particular statutory regime. In this new type of (extra-special?) statutory review the grounds specified in the LRA are suffused with the content of the rights to administrative justice in section 33, thus as it were achieving full administrative-law review via labour law. Presumably this will apply to other statutory regimes too, such as section 131 of the Liquor Act 27 of 198939 and section 151 of the Insolvency Act 24 of 1936: whenever the conduct in question qualifies as administrative action under section 33, the grounds of review in such statutes will also be suffused with the content of section 33.

If this is correct, an interesting situation will arise in relation to the Promotion of Access to Information Act 2 of 2000. Decisions taken by information officers are specially reviewable under that statute40 and would almost certainly qualify as administrative action for the purposes of section 33 — but they are expressly excluded from the more detailed definition of administrative action in the PAJA.41 This piquant scenario is no doubt further evidence of the constitutional untenability of having two different and sometimes incompatible meanings for ‘administrative action’.42 (It remains one of life’s mysteries that the constitutionality of the PAJA definition has not yet been challenged in court.)

That brings me to a related point, which is that the approach of the Constitutional Court has a distinct practical advantage over that of Cameron JA in the Supreme Court of Appeal: it means not having to engage with the detail of the PAJA. In Rustenburg Cameron JA is faced with conflicting time limits within which review must be sought — six weeks in terms of section 145 of the LRA and six months in terms of the PAJA — and his choice of the shorter time limit43 sits rather awkwardly with the proposition that the PAJA overrides section 145 of the LRA. For Navsa AJ the difficulty does not arise: the PAJA and section 145 of the LRA exist alongside one another as separate and

39 Cf the approach in Bulk Deals Six CC v Chairperson, Western Cape Liquor Board 2002 2 SA 99 (C), where the court applied the grounds in the PAJA to a review under s 131 of the Liquor Act. 40 Secs 78-82. 41 Sec 1(hh). 42 On the various problems arising from the PAJA definition and the disparity between the two meanings, see C Hoexter ‘“Administrative action” in the courts’ 2006 Acta Juridica 303 and Hoexter (n 38 above) 216ff. 43 Rustenburg (n 6 above) para 27. 216 Clearing the intersection? Administrative law and labour law in the CC equally valid manifestations of section 33, and there can thus be no conflict between them.

The third feature to note about Sidumo is the ethos or vision of the Constitution that underlies the judgments handed down by the Constitutional Court. The majority judgment of Navsa AJ, in particular, supports a view of constitutional rights as interconnected and overlapping. Nor is this view merely implicit in the judgment. There is strong and explicit affirmation of it when Navsa AJ is rejecting an argument that the rights implicated in CCMA arbitrations are those in sections 23 and 34, and not section 33, of the Constitution:

This submission is based on the misconception that the rights in ss 23, 33 and 34 are necessarily exclusive and have to be dealt with in sealed compartments. The right to fair labour practices, in the present context, is consonant with the right to administrative action that is lawful, reasonable and procedurally fair. Everyone has the right to have these rights enforced before the CCMA acting as an impartial tribunal. In the present context, these rights in part overlap and are interconnected.44

Although Ngcobo J may not share this view of things,45 it is given particularly strong expression in the judgment of Sachs J. He sees the function of the commissioner as a hybrid one, ‘composed of an amalgam of three separate but intermingling constitutional rights’46 whose underlying values ‘resist compartmentalisation’.47 The Bill of Rights ‘should not always be seen as establishing independent normative regimes operating in isolation from each other, each with exclusive sway over a defined realm of public and private activity’,48 and indeed the relationship of the various rights should be regarded as ‘osmotic rather than hermetic’.49 In this judgment ‘seepage’, ‘permeability’, ‘interpenetration’ and ‘hybridity’ are all desirable qualities that help ensure the full achievement of constitutional justice, while an attempt to establish the primacy of one right or another could actually defeat the constitutional objectives to be realised.50

To say the least, the constitutional vision outlined here seems to be absent from the two main judgments in Chirwa. Rather, these judgments insist on the primacy of section 23 and the LRA in a labour matter, and indeed do their utmost to exclude section 33 and the PAJA from the picture. It is to Chirwa that I now turn.

44 Sidumo (n 5 above) para 112. 45 There are hints of it, however: see eg Sidumo (n 5 above) para 266. 46 Sidumo (n 5 above) para 147. 47 Sidumo (n 5 above) para 149. 48 Sidumo (n 5 above) para 151. 49 As above. 50 Sidumo (n 5 above) paras 151-7. (2008) 1 Constitutional Court Review 217

2 Chirwa

Ms Chirwa, the appellant, had been dismissed by the Transnet Pension Fund pursuant to a disciplinary hearing. She alleged procedural unfairness in the hearing and referred the dispute to the CCMA for conciliation. When that failed she sought administrative-law review of the dismissal in the Witwatersrand Local Division. There, relying on the common law rather than the PAJA, Brassey AJ set aside the decision to dismiss Ms Chirwa and ordered her reinstatement.51 He followed the reasoning of the Appellate Division in Zenzile,52 holding that the termination of an employment contract in the public sector is an exercise of public power and not (as the respondent argued) a matter of pure contract falling beyond the reach of administrative law. Zenzile-type reasoning had indeed become the dominant approach53 in such cases, and here I may mention an important and more recent example of it: the judgment of Plasket J in POPCRU v Minister of Correctional Services.54 But, as evidenced by cases such as South African Police Union v National Commissioner of the South African Police Service,55 a contrary approach had begun to gain momentum — one that challenged the post-1994 status of Zenzile and denied administrative law a role in employment matters.

In Transnet’s appeal against the decision of Brassey AJ two main questions were raised: whether this was a matter over which the High Court had jurisdiction, and whether the dismissal qualified as administrative action under the PAJA.56 Though the Court agreed that the appeal had to be upheld, its members adopted three quite different approaches. Cameron JA, Mpati DP concurring, endorsed both Zenzile and POPCRU.57 These judges held that this was a case of administrative action and that the High Court did have jurisdiction, concurrently with the Labour Court, to hear it in terms of section 157

51 Chirwa v Transnet Ltd WLD case no 03/01052 of 4 January 2005, unreported. 52 Zenzile (n 1 above). 53 Before 1994 the approach in Zenzile was applied and extended beyond dismissal in cases including Administrator, Natal v Sibiya 1992 2 SA 532 (A), Bula v Minister of Education 1992 4 SA 716 (Tk) and Hlongwa v Minister of Justice, KwaZulu- Natal 1993 2 SA 269 (D); and after 1994 it was applied in cases such as Mbayeka v MEC for Welfare, Eastern Cape 2001 1 All SA 567 (Tk) and Simela v MEC for Education, Eastern Cape 2001 9 BLLR 1085 (LC). 54 2006 4 BLLR 385 (E) (POPCRU). See also Nxele v Chief Deputy Commissioner, Corporate Services, Department of Correctional Services 2006 10 BLLR 960 (LC) (Nxele). 55 2005 26 ILJ 2403 (LC) (SAPU). Other examples of this approach are Public Servants’ Association obo Haschke v MEC for Agriculture 2004 25 ILJ 1750 (LC), Louw v SA Rail Commuter Corporation Ltd 2005 26 ILJ 1960 (W), Western Cape Workers Association v Minister of Labour 2006 1 BLLR 79 (LC) and Hlope v Minister of Safety and Security 2006 3 BLLR 297 (LC). 56 Transnet Ltd v Chirwa 2006 27 ILJ 2294 (SCA) (Chirwa SCA). 57 n 54 above. 218 Clearing the intersection? Administrative law and labour law in the CC of the LRA.58 The relevant parts of the provision read as follows:

(1) Subject to the Constitution and section 173, and except where this Act provides otherwise, the Labour Court has exclusive jurisdiction in respect of all matters that elsewhere in terms of this Act or in terms of any other law are to be determined by the Labour Court. (2) The Labour Court has concurrent jurisdiction with the High Court in respect of any alleged or threatened violation of any fundamental right entrenched in Chapter 2 of the Constitution of the Republic of South Africa, 1996, arising from – (a) employment and from labour relations; (b) any dispute over the constitutionality of any executive or administrative act or conduct, or any threatened executive or administrative conduct, by the State in its capacity as an employer; (c) the application of any law for the administration of which the Minister is responsible.

Mthiyane and Jafta JJA agreed that the High Court had jurisdiction concurrently with the Labour Court but held that this was not an instance of administrative action. These judges relied on SAPU59 and on Supreme Court of Appeal authority, Cape Metropolitan Council v Metro Inspection Services (Western Cape) CC,60 in holding that the dismissal was based on contract and so did not satisfy one of the most important elements of the PAJA definition: it was not an exercise of public power.61

Conradie JA was prepared to accept that the dismissal amounted to administrative action but held that the High Court did not have jurisdiction over what was quintessentially a labour matter. The legislature, he reasoned, had set its face against labour matters being litigated in other courts.62 Thus, even if Chirwa had a cause of action under the PAJA she was required to pursue it under the LRA in the Labour Court.63

The two issues were aired again on further appeal to the Constitutional Court.64 Skweyiya J gave judgment for the majority with the concurrence of seven justices.65 Essentially he supported the line taken by Conradie JA, and held that the High Court did not have

58 Chirwa SCA (n 56 above) paras 57-65. Similar reasoning was subsequently adopted by the Supreme Court of Appeal in Old Mutual Life Assurance Co SA Ltd v Gumbi 2007 5 SA 552 (SCA) (Gumbi) and Boxer Superstores Mthatha v Mbenya 2007 5 SA 450 (SCA). 59 n 55 above. 60 2001 3 SA 1013 (SCA) (Cape Metropolitan). 61 Chirwa SCA (n 56 above) para 15. 62 Chirwa SCA (n 56 above) para 32. 63 Chirwa SCA (n 56 above) para 33. 64 Chirwa (n 3 above). 65 Moseneke DCJ, Navsa AJ and Madala, Ngcobo, Nkabinde, Sachs and Van der Westhuizen JJ. (2008) 1 Constitutional Court Review 219 jurisdiction. The appellant should therefore have followed to its end the route laid down by the LRA for resolving her dispute. Ngcobo J produced a separate concurring judgment which dealt in addition with the ‘administrative action’ issue and was supported by six justices.66 Langa CJ dissented with the support of Mokgoro and O’Regan JJ.

2.1 Jurisdiction

The majority conclusion that the Labour Court had exclusive jurisdiction in the matter was evidently inspired by considerations of policy, and by the legislature’s failure to heed various calls for the amendment of section 157. Both Skweyiya and Ngcobo JJ reasoned that employees ought not to be in a preferential position merely by virtue of their employment in the public sector. Public and private employees ought rather to be treated equally, and parity was indeed one of the main aims of the current labour regime.67 That regime included a ‘one-stop shop’,68 a specialised and purpose-built system for resolving labour disputes, which should be pursued to its end.69 Furthermore, to allow the High Court concurrent jurisdiction with the Labour Court in employment matters would be to encourage forum- shopping and the development of a dual70 and possibly less coherent71 system of law. In the light of these considerations, Ncgobo J thought the use of the word concurrent in section 157(2) was ‘unfortunate’.72

The problem, however, is that policy arguments cannot alter the express terms of section 157 of the LRA — however unfortunate those terms may seem. Leaving aside for a moment the cogency of the arguments, the majority’s interpretation of the provision is in conflict not only with the wording of section 157 but also with the Court’s own previous interpretation of that wording in Fredericks & Others v MEC for Education and Training, Eastern Cape, & Others.73

In Fredericks, teachers who had been refused voluntary retrenchment sought review of this action in a High Court, alleging a violation of their constitutional rights to equality and just administrative action. White J found that his Court had no jurisdiction

66 Moseneke DCJ, Navsa AJ and Madala, Nkabinde, Sachs and Van der Westhuizen JJ. 67 Chirwa (n 3 above) para 66 (Skweyiya J) and paras 99-102 and 149 (Ncgobo J). 68 Chirwa (n 3 above) para 47 (Skweyiya J). 69 Chirwa (n 3 above) paras 65-7 (Skweyiya J), and see paras 105, 110-13 and 117-20 (Ngcobo J). 70 Chirwa (n 3 above) paras 65 (Skweyiya J) and 121, 124 (Ngcobo J). 71 Chirwa (n 3 above) para 118 (Ngcobo J). 72 Chirwa (n 3 above) para 121. 73 2002 2 SA 693 (CC) (Fredericks). See also Fedlife Assurance Ltd v Wolfaardt 2002 1 SA 49 (SCA); United Public Servants Association of South Africa v Digomo NO 2005 26 ILJ 1957 (SCA). 220 Clearing the intersection? Administrative law and labour law in the CC in the matter. He saw it in essence as a labour case that had to be dealt with under section 24 of the LRA, a provision that envisages the involvement of the CCMA. On appeal, however, a unanimous Constitutional Court corrected this misapprehension. O’Regan J pointed out that section 157(2) is the sole source of the Labour Court’s jurisdiction to determine disputes arising from the alleged infringement of constitutional rights, and that it clearly gives the Labour Court and the High Court concurrent jurisdiction over such matters.74 She emphasised that the LRA gives the Labour Court no general jurisdiction over labour matters, so that the High Court’s jurisdiction is not ousted by the mere fact that something happens to be a labour matter.75 Nor could section 158(2)(h) be read with section 157(1) of the LRA to oust the High Court’s jurisdiction, since section 158(1)(h) ‘does not expressly confer on the Labour Court constitutional jurisdiction to determine disputes arising out of alleged infringements of the Constitution by the state acting as an employer’.76 O’Regan J also made it clear that in terms of section 169 of the Constitution, such ousting of the High Court’s jurisdiction is permissible only where the matter is assigned to a court of similar status — which the CCMA clearly is not.77

In Chirwa, notwithstanding the way the matter was framed and the many detailed references to the PAJA in the papers, Skweyiya J chose to fix on the appellant’s brief reference to provisions in the LRA (evidently included for the purpose of establishing grounds of review in the PAJA).78 He held that this was a dispute envisaged by section 191 of the LRA, which provides a procedure for the resolution of disputes about unfair dismissals. It was thus a matter that ‘must, under the LRA, be determined exclusively by the Labour Court’.79 While Fredericks clearly stood in the way of such an interpretation of section 157, Skweyiya J distinguished it on the basis that there had been no reliance in the earlier case on section 23 or on the LRA.80

That, with respect, seems an unconvincing point of distinction. After all, Fredericks was so obviously a ‘labour’ matter that the High Court believed it lacked jurisdiction to hear it — despite the applicants’ own characterisation of the case as one relating to sections 9 and 33 of the Bill of Rights. Conversely, what about the inconvenient fact that the appellant in Chirwa evidently abandoned

74 Fredericks (n 73 above) para 41. 75 Fredericks (n 73 above) paras 38 and 40. 76 Fredericks (n 73 above) para 43. Section 158(1)(h) allows the Labour Court to review ‘any decision taken or any act performed by the State in its capacity as employer on such grounds as are permissible in law’. 77 Fredericks (n 73 above) paras 29-31. 78 Chirwa (n 3 above) para 157 in the judgment of Langa CJ. 79 Chirwa (n 3 above) para 63. 80 Chirwa (n 3 above) para 58. (2008) 1 Constitutional Court Review 221 her initial reliance on section 23 and the LRA81 and deliberately framed her case as one in administrative law?82

As Nugent JA has remarked, it is difficult to find a ‘clear legal — as opposed to policy — reason for the outcome in Chirwa’.83 The majority was apparently bent on characterising the case purely as a labour matter whatever the cost. It thus depicted the appellant’s High Court application as an instance of asking an ordinary court to decide a specialist dispute.84 Langa CJ bluntly described this manoeuvre as a mischaracterisation,85 for in truth Ms Chirwa was ‘asking a High Court to consider an administrative law issue’.86 Langa CJ pointed out that in most cases unfair dismissal claims are decided not by the Labour Court but by the CCMA, so that in terms of section 169 of the Constitution exclusive jurisdiction could not lawfully be conferred on the Labour Court in any event.87 The Chief Justice also indicated the irrelevance of the various policy considerations to the issue of jurisdiction, which in this instance was a matter of legislative choice. Ultimately, as he said, ‘[w]hile we may question that intention and may have preferred a legislative scheme that more neatly divided responsibilities between the various courts, that is not the path the legislature has chosen.’88

Skweyiya J seems indeed to have been aware of this, and not entirely convinced by his own strained interpretation of section 157. Otherwise it is hard to see why he ends his judgment by underscoring the ‘urgent need for the legislature to revisit the provisions of s 157(2) of the LRA’89 — for his account of the provision suggests no such need.

At any rate, wishful thinking cannot change the fact that in practice labour law and administrative law are not neatly divided, and that in many cases they do overlap. I doubt that neatness will ever be achieved at this intersection, legislatively or otherwise, for it seems to me that some untidiness is both natural and unavoidable. Nor, I think, should neatness be regarded as a particularly worthy aim in this

81 Chirwa (n 3 above); see para 67 in the judgment of Skweyiya J and paras 84-5 in the judgment of Ngcobo J. 82 Chirwa (n 3 above); see paras 157-9 in the judgment of Langa CJ. 83 In a separate concurring judgment in Makambi v MEC for Education, Eastern Cape 2008 5 SA 449 (SCA) (Makambi) para 21. 84 Chirwa (n 3 above) para 61 (Skweyiya J) and paras 124-5 (Ngcobo J). 85 Chirwa (n 3 above) para 159. 86 Chirwa (n 3 above) para 173. Not only was the claim couched in the language of administrative law, but it was ‘based squarely on the PAJA’. Chirwa (n 3 above) para 157. 87 Chirwa (n 3 above) para 170. Furthermore, s 157(5) of the LRA states: ‘Except as provided for in section 158(2), the Labour Court does not have jurisdiction to adjudicate an unresolved dispute if this Act requires the dispute to be resolved through arbitration.’ 88 Chirwa (n 3 above) para 174. 89 Chirwa (n 3 above) para 71. 222 Clearing the intersection? Administrative law and labour law in the CC area of the law. Both of these propositions find some support in the facts and reasoning applied in a post-Chirwa case, Nakin v MEC, Department of Education, Eastern Cape & Another.90

The applicant in Nakin was a school principal who had been transferred to a lower-level post through no fault of his own. This injustice was officially rectified several years later when the department agreed to reinstate him and to correct the accumulated shortfall in salary and benefits. The department neglected to effect the correction, however, and the applicant approached a High Court for an order compelling it to do so in terms of the PAJA. In argument the respondents challenged the Court’s jurisdiction to deal with what they saw as an unfair labour practice within the meaning of section 186(2) of the LRA. But Froneman J upheld the High Court’s jurisdiction to determine the case as one of unlawful administrative action.91 Fredericks, which had not been overruled in Chirwa, was still good law, all the more so in a matter not involving dismissal.92

The facts of Nakin nicely illustrate the natural overlap just referred to. More importantly, however, Froneman J challenges the policy behind the tidy-minded majority view in Chirwa. In relation to the concerns about forum-shopping and the growth of parallel systems of law, he points out that the coherence of labour jurisprudence depends on its giving proper expression to section 23 of the Constitution rather than its development in a single forum.93 Froneman J goes on to show that in practice, labour law has gained and not lost from administrative-law insights — an example being the recent development of the common-law contract of employment so as to include a right to a pre-dismissal hearing.94 His judgment also suggests that the policy of equality may be a misdirected one. Froneman J points out that in public employment fairness ‘may conceivably have a different content to that in the private sector, for reasons relating to constitutional demands of responsiveness, public accountability, democracy and efficiency in the public service’.95

This is a point of fundamental importance. Public and private employers are not the same, for under our Constitution they are not under the same public duties. To put it simply, employers in the private sector have the luxury (to the extent that they are not

90 2008 6 SA 320 (Ck) (Nakin). 91 Nakin (n 90 above) para 39. 92 Nakin (n 90 above); see paras 27-39. Cf Makambi (n 83 above) para 17, where Farlam JA found it was ‘not possible to hold that this case falls on the Fredericks side of the line of distinction drawn in the Chirwa case’. 93 Nakin (n 90 above) para 30. 94 Nakin (n 90 above) para 30, with reference to Gumbi (n 58 above); and see also paras 31 et seq. 95 Nakin (n 90 above) para 34. (2008) 1 Constitutional Court Review 223 thwarted by regulation) of acting out of pure self-interest. Organs of state do not have the same freedom, for they are generally constrained by their duty to act in the public interest — even when there is no legislation explicitly saying so.96 Thus, while the aim of achieving equality between all employees certainly appeals to one’s sense of justice, it may not be an attainable ideal under our present constitutional dispensation; not, at any rate, if one conceives of that equality as a matter of paring down the rights of public-sector employees to match those of other employees.

That does not, however, rule out the possibility of equalising up rather than down by allowing the insights of administrative law (or some of them)97 to benefit employees more generally. To borrow the words of Froneman J, ‘the substantive coherence and development of employment law can only gain’ from administrative law.98 While such reform would optimally be achieved by methodical legislative means, one must not underestimate the potential of section 39(2) of the Constitution. As I have suggested elsewhere, section 39(2) may be regarded as constitutional authority for the courts to set public-law standards for ‘private administrators’ to follow, especially in relationships characterised by inequality (as the employment relationship tends to be) or where coercive power is being exercised.99

2.2 Administrative action

Though the Court in Chirwa agreed that the dismissal was not an instance of administrative action, its members displayed three different approaches to the issue. As I suggest in what follows, none of these approaches is unproblematic.

Skweyiya J found that it was unnecessary to decide the administrative action question since the case could be resolved on the jurisdictional issue (though he indicated his agreement, obiter, with the judgment of Ngcobo J on the administrative action diagnosis).100

96 This most basic principle of administrative law was expressed by Schreiner JA in his dissenting judgment in Mustapha v Receiver of Revenue, Lichtenburg 1958 3 SA 343 (A) at 347E-G — a judgment vindicated by the Supreme Court of Appeal in Logbro Properties CC v Bedderson NO 2003 1 SA 460 (SCA) (Logbro) para 13. It is in fact unusual for enabling legislation to specify that action has to be taken ‘in the public interest’, but an example appears in Nxele supra note 54 para 59. 97 The requirements implied by ‘reasonable’ and ‘procedurally fair’ administrative action could certainly be extended in this way, but imposing the various requirements of ‘lawfulness’ on private employers would make less sense. 98 Nakin (n 90 above) para 34. 99 Hoexter (n 38 above) 124. Unlike the direct horizontal application of rights under s 8(2) of the Constitution, indirect application under s 39(2) does not depend on the ‘nature’ of the right sought to be applied. 100 Chirwa (n 3 above) para 73. 224 Clearing the intersection? Administrative law and labour law in the CC

This approach is clearly mistaken, however. As the Chief Justice pointed out, the administrative action inquiry could not legitimately be avoided: it was indeed the ‘primary question’101 in the case, for the Court’s jurisdiction depended on whether the dismissal was ‘an administrative act or conduct ... by the State in its capacity as an employer’ for the purposes of section 157 of the LRA.

Ngcobo J did not seek to avoid the inquiry. Relying on the section 33 meaning of administrative action, he found (rightly, I believe) that the dismissal clearly amounted to an exercise of public power.102 However, it lacked one of the hallmarks of administrative action listed in the SARFU case:103 it did not involve the implementation of legislation but had rather been effected in terms of the contract between the parties.104 Ngcobo J thus treated a single factor as decisive in a manner arguably not contemplated by the Court in SARFU. But the real issue for him seems to have been that the dismissal was ‘more concerned with labour and employment relations’ than with administration,105 and he went on to emphasise the formal division in the Constitution between labour relations and administrative conduct.106

By contrast, Langa CJ relied on the definition of administrative action in the PAJA, which covers decisions made by an organ of state when ‘exercising a public power or performing a public function in terms of any legislation’.107 He, too, held that in the absence of any particular statutory authority the dismissal had taken place in terms of the contract itself;108 but unlike Ngcobo J, he went on to find that the dismissal was not an exercise of public power. In this regard Langa CJ noted the absence of features that tend to point to the existence of a public power: any imbalance in the employment relationship was not attributable to the respondent’s status as a public body; the dismissal had little or no impact on the public; and the source of the power to dismiss was contractual.109 Furthermore, he could see none of the ‘strengthening factors’ that had been present in the POPCRU case.110 The Transnet Pension Fund did not have the same obviously public character as the Department of Correctional Services, the respondent in POPCRU; unlike that department it pursued no obviously public goals; and, unlike in POPCRU, the public interest in

101 Chirwa (n 3 above) para 154. 102 Chirwa (n 3 above) para 138, quoting with approval from the judgment of Cameron JA in the court below. 103 n 32, para 143. 104 Chirwa (n 3 above) para 142. 105 As above. 106 Chirwa (n 3 above) paras 143 et seq. 107 Section 1 of the PAJA, my emphasis. 108 Chirwa (n 3 above) para 185. 109 Chirwa (n 3 above) paras 186-94. 110 Chirwa (n 3 above) para 192, referring to POPCRU (n 54 above) para 54. (2008) 1 Constitutional Court Review 225 the administration of the Transnet Pension Fund was not pre- eminent.111

The different approaches of Ngcobo J and Langa CJ illustrate, once again, the problem of having two meanings for ‘administrative action’. Given that reality, however, Langa CJ is surely correct to rely on the PAJA definition. This approach accords with the constitutional principle that general norms should be resorted to only when norms of greater specificity have run out,112 or when testing the constitutionality of a specific norm (the PAJA or other original legislation) against the more general norm (section 33).113

Ngcobo J, who has preferred the section 33 meaning in other cases too,114 does so on the basis that ‘PAJA only comes into the picture once it is determined that the conduct in question constitutes administrative action within the meaning of section 33.’115 This approach is rather cumbersome as it calls for two separate inquiries: Is the conduct administrative action for the purposes of s 33? If so, does the PAJA definition nevertheless exclude it? (Stopping at the first inquiry would of course make the PAJA definition completely redundant.)116 More importantly, this approach seems to be inconsistent with the principle mentioned above and with the Constitutional Court’s jurisprudence on the subject of constitutionally mandated legislation. As the statute giving effect to the rights in section 33, the PAJA should be resorted to directly — a point that has indeed been stressed by Ngcobo J himself.117 Even if it does not have the effect of bypassing the PAJA altogether, the mediation through section 33 he insists on seems quite unwarranted. It means, too, that in this instance we are deprived of his views on the application of the various elements of the PAJA definition.

As to the merits of the ‘administrative action’ diagnosis in Chirwa, it depended largely on a curiosity. Like appointment, dismissal in the public sector is almost always governed by statute. It certainly was in Zenzile, where ‘the element of public service injected by statute’118 played a considerable part in the reasoning of Hoexter JA. A quirk of

111 As above. 112 See I Currie & J de Waal The bill of rights handbook (5 ed 2005) 13. 113 See Hoexter (n 38 above) 115. 114 See eg New Clicks (n 29 above) para 446. 115 Chirwa (n 3 above) para 139. 116 Construing the PAJA consistently with s 33 is constitutionally appropriate (as indicated by Nugent JA in Grey’s Marine Hout Bay (Pty) Ltd v Minister of Public Works 2005 6 SA 313 (SCA) para 22), but it is quite another thing to behave as if the statute and its detailed definition of administrative action did not exist. See further Hoexter (n 42 above) and I Currie ‘What difference does the Promotion of Administrative Justice Act make to administrative law?’ 2006 Acta Juridica 325. 117 See eg New Clicks (n 29 above) para 436, and in relation to s 23 of the Constitution see eg Sidumo (n 5 above) paras 248-9. 118 n 1 above, 34C. 226 Clearing the intersection? Administrative law and labour law in the CC the Chirwa case was the absence of governing legislation after the repeal of the South African Transport Services Conditions of Service Act 41 of 1988. Technically, then, the dismissal really did seem to be a purely contractual matter in this instance — if not in others.119

But the judgment of Langa CJ (and that of Mthiyane JA before him) is nevertheless open to criticism on the contractual issue. As Stacey has pointed out in a perceptive article,120 it is difficult to square with the approach of the Constitutional Court in AAA Investments (Pty) Ltd v Micro Finance Regulatory Council.121 There the Court firmly rejected the idea that a decision loses its public character simply because the most immediate source of the power happens to be a contract. In that case the contractual element led the Supreme Court of Appeal to regard the Micro Finance Regulatory Council as ‘a private regulator of lenders who choose to submit to its authority by agreement’122 — and was rebuked by the Constitutional Court for putting form above substance.123 Is the Chief Justice not open to the same charge here?

In Transnet Ltd v Chirwa, it is interesting to note, the absence of governing legislation does not trouble Cameron JA at all. As he sees it,

Transnet is a public entity created by legislation and operating under statutory authority. It would not exist without statute. Its every act derives from its public, statutory character, including the dismissal at issue here.124

On this approach public entities can never be on exactly the same footing as private ones, contract or no contract; and even when there is a contract, the principles of administrative justice (in one form or another) still frame the contractual relationship.125 In the Constitutional Court Ngcobo J supports this line of reasoning:126

In my view, what makes the power in question a public power is the fact that it has been vested in a public functionary, who is required to exercise the power in the public interest. When a public official performs a function in relation to his or her duties, the public official

119 Cf for instance SAPU (n 55 above), where the ‘purely contractual’ label is far less convincing, and see further C Hoexter ‘Contracts in administrative law: Life after formalism?’ (2004) 121 South African Law Journal 595. 120 R Stacey ‘Administrative law in public-sector employment relationships’ (2008) 125 South African Law Journal 307 315. 121 2006 11 BCLR 1255 (CC) (AAA Investments). 122 Micro Finance Regulatory Council v AAA Investments (Pty) Ltd 2006 2 SA 27 (SCA) para 24. 123 AAA Investments (n 121 above) para 45. 124 Chirwa SCA (n 56 above) para 52. 125 Cameron JA for a unanimous court in Logbro (n 96 above) para 8, writing an important gloss on the Cape Metropolitan case (n 60 above) 126 Chirwa (n 3 above) para 138. (2008) 1 Constitutional Court Review 227

exercises public power. I agree with Cameron JA that Transnet is a creature of statute. It is a public entity created by statute and it operates under statutory authority. As a public authority, its decision to dismiss necessarily involves the exercise of public power, and ‘(t)hat power is always sourced in statutory provision, whether general or specific, and, behind it, in the Constitution’.

By contrast, Langa CJ finds that the dismissal is not an exercise of public power. In arriving at this conclusion he relies on a number of factors that have been used by the courts for determining whether a power or function is public. However, his paradigm seems to be quite different from that of Cameron JA and Ngcobo J, for in his analysis the public and intrinsically statutory nature of the entity apparently carries no independent weight. The factors seem to be applied just as they would be to a completely private entity such as a club or a business. Because of this, and notwithstanding the detailed and nuanced reasoning of the Chief Justice and the cumulative force of the various features listed by him, one is left with questions.

Here are some of them. Given that Transnet would not exist without statute, does it really matter that a contract happened to intervene in this instance? (Langa CJ is clear that where it is achieved ‘in terms of a specific legislative provision’ a dismissal may amount to administrative action.)127 Given the undoubted public status of Transnet, does it really matter that the Transnet Pension Fund is not as distinctively public as the Department of Correctional Services was in POPCRU, or that it does not pursue its public goals in so obvious a manner? Given that Transnet is generally supposed to serve the public interest, does it really matter that this particular dismissal was of scant interest to the public or that it had no real impact on the public? (As Plasket J pointed out in POPCRU, very many actions acknowledged to be administrative have little or no such impact.)128 And is it realistic to say — is it ever entirely realistic to say — that no power imbalance derived from the employer’s public status?

Similarly, there is something troubling about the Chief Justice’s statement that ‘[w]hilst Transnet is certainly subservient to the Constitution, so are all business entities in South Africa’.129 This fails to acknowledge that as an organ of state Transnet is under special duties (such as the obligation in section 7(2) of the Constitution) that simply do not apply to private employers.

In short, I wonder whether an entity like Transnet is ever entirely free to do as it pleases, and whether it is capable of shedding its

127 Chirwa (n 3 above) para 194. 128 n 54 above, para 53. 129 Chirwa (n 3 above) para 192. 228 Clearing the intersection? Administrative law and labour law in the CC public, statutory nature so conveniently and completely. I doubt that public entities are capable of acting ‘simply in [their] capacity as employer’, as Mthiyane JA would have it;130 and with respect, nothing said in Chirwa eliminates that doubt.

2.3 The interconnectedness of rights

In Sidumo,131 as we have seen, the majority fully acknowledges the interconnectedness of constitutional rights and their natural tendency to overlap. This view of things is also clearly apparent in the dissenting judgment in Chirwa, in which Langa CJ more than once recognises the overlap of the LRA and the PAJA and their parent rights.132 But there is little hint of this understanding in the majority judgments in Chirwa. Instead, emphasis is laid on the entrenchment of the respective rights in two separate provisions in the Constitution, ‘each with its own aims and specialised legislation ... that seeks to give effect to its own distinct objectives’.133 While Skweyiya J concedes that the dismissals of public sector employees ‘appear to implicate’ not only labour rights but also those of administrative justice (or that this ‘is at least what Ms Chirwa is asserting’),134 and though he briefly admits a possible overlap of the LRA and the PAJA,135 he is clear about the ‘pre-eminence of the LRA’ in labour disputes, and about the absence of any legislative intention that the PAJA should detract from it.136 In short, he refuses to see Chirwa as a case involving administrative law at all. Ngcobo J achieves this view of the case too, and he does so by relying heavily on the Constitution’s ‘clear distinction between administrative action on the one hand and employment and labour relations on the other’.137

This categorical view of things is a pity — not just because the vision of the majority is so completely at variance with the ethos of Sidumo, but also because it ignores an important constitutional principle. In POPCRU Plasket J warns against formalistic attempts to pigeonhole administrative law and labour law, which are after all mere classifications of convenience.138 Under our Constitution, he observes,

[t]here is nothing incongruous about individuals having more legal protection rather than less, or of more than one fundamental right

130 Chirwa SCA (n 56 above) para 15. 131 n 5 above. 132 Chirwa (n 3 above) paras 167 and 176. 133 Chirwa (n 3 above) para 46 (Skweyiya J). 134 Chirwa (n 3 above) para 46. 135 Chirwa (n 3 above) para 71. 136 Chirwa (n 3 above) para 50. 137 Chirwa (n 3 above) paras 143-4. 138 n 54 above, para 61. (2008) 1 Constitutional Court Review 229

applying to one act, or of more than one branch of law applying to the same set of facts.139

In Transnet Ltd v Chirwa Cameron JA finds in the Constitution ‘no suggestion that, where more than one right may be in issue, its beneficiaries should be confined to a single legislatively created scheme’.140 In Chirwa the Chief Justice expresses the same principle in the following way:141

Both PAJA and the LRA protect important constitutional rights and we should not presume that one should be protected before another or presume to determine that the ‘essence’ of a claim engages one right more than another. A litigant is entitled to the full protection of both rights, even when they seem to cover the same ground.

It is unfortunate that the majority is impervious to this wisdom.

3 Masetlha

Masetlha,142 chronologically the first case in the series, arose out of an atypical case of ‘executive’ dismissal featuring the President of the Republic and the head of the National Intelligence Agency (NIA). The ‘one-stop shop’ of the LRA was not open to the appellant in this case, and nor was the PAJA of application; but it is nevertheless a matter that touches both labour law and administrative law more broadly. For administrative lawyers its main interest lies in what the majority judgment says, or fails to say, about procedural fairness.

The facts are widely known. Briefly, in December 2004 the President appointed the appellant, Mr Billy Masetlha, head and Director-General of the NIA for a period of three years. In October 2005 Masetlha was suspended from this position as a result of the ‘Macozoma affair’, wherein a prominent businessman was placed under surveillance and subsequently lodged a complaint about it with the Minister. The suspension followed an investigation by the Inspector-General of Intelligence, and was challenged by Masetlha in an application for setting aside. In March 2006, before that had been decided, the President unilaterally amended Masetlha’s term of office so that it expired two days later — thus effectively dismissing him 21 months before his term was due to end. The President’s justification was that the relationship of trust between him and the NIA head had

139 POPCRU (n 54 above) para 60. See further Stacey (n 120 above) 323ff, where he demonstrates the potentially absurd consequences of the proposition that the remedies offered by one branch of the law oust those of another. 140 Chirwa SCA (n 56 above) para 65. 141 Chirwa (n 3 above) para 175. 142 n 4 above. 230 Clearing the intersection? Administrative law and labour law in the CC broken down irreparably. Masetlha was offered his full salary and benefits for the unexpired period. He declined the offer and sought review of the President’s decisions and reinstatement in his position.

Du Plessis J held for the Pretoria High Court that the dispute about the suspension had been rendered moot by the dismissal, and that the dismissal itself amounted to lawful executive action.143 This decision was upheld on appeal to the Constitutional Court (which did not doubt that the President’s action amounted to a dismissal).144 Moseneke DCJ gave judgment for the majority with the concurrence of six colleagues;145 Ngcobo J dissented with the support of Madala J; and Sachs J produced a separate judgment.

The first main issue related to lawfulness. Neither the Constitution nor the relevant legislation mentioned dismissal of the head of the NIA but, notwithstanding some disagreement about where the authority resided, all the justices agreed that the President had the implied authority to do what he did.146 More interesting was the second issue: what constraints there were on the exercise of the power to dismiss, and whether they included a requirement of procedural fairness — for the President had acted without giving the appellant notice or a hearing on the issue of his dismissal.147

The usual constraints were soon ruled out. They clearly did not include the LRA, since it does not apply to members of the NIA.148 The majority did not pause to consider whether the appellant still had the benefit of the section 23 right to fair labour practices — which was the solution favoured by Sachs J in his separate judgment — and so there was no hint here of the Chirwa insistence on the primacy of labour law in a case of dismissal. In fact administrative law ultimately came to the rescue, though not in the form of the PAJA. The latter did not apply either: the dismissal was an exercise of executive powers that are specifically excluded from the definition of administrative action in the PAJA (and would not count as administrative action for the purposes of section 33).149 As for the common-law right to a pre- dismissal hearing recognised by the Appellate Division in Zenzile,150 the majority reasoned that this case was distinguishable from the

143 See Masetlha (n 4 above) paras 22-4 in the judgment of Moseneke DCJ. 144 Masetlha (n 4 above) para 52. 145 Langa CJ, Navsa AJ and Nkabinde, O’Regan, Skweyiya and Van der Westhuizen JJ. 146 The majority found that the necessary power lay in s 209(2) of the Constitution and s 3(3) of the Intelligence Services Act 65 of 2002 (see Masetlha n 4 above, para 73), whereas Ngcobo J found that provisions of the Public Service Act 103 of 1994 were equally relevant (Masetlha n 4 above, para 157). 147 Masetlha (n 4 above) para 84, and see para 197 in the judgment of Ngcobo J. 148 See s 2(b) of the LRA. 149 Masetlha (n 4 above) para 76n39, though Moseneke DCJ does not deal fully with the s 33 meaning. 150 n 1 above. (2008) 1 Constitutional Court Review 231 earlier one: it concerned a ‘special legal relationship’ between the President and the head of the NIA, and the dismissal involved the exercise of an executive power deriving from the Constitution and national legislation.151 In the view of the majority the dismissal attracted only the grounds of review implied by that residual pathway to administrative-law review, the principle of legality.152

As already noted, the content of the principle has been developed considerably over the years: it has been held to imply that those who wield public power must act lawfully153 and in good faith,154 must not misconstrue their power,155 and must act rationally in relation to the purpose for which the power was given.156 These requirements effectively cover many of the grounds of review in ‘regular’ administrative law (the PAJA), and they explain why the rule of law in general, and the principle of legality in particular, have been described as ‘administrative law applied under another name’.157 But while procedural fairness may be a standard component of the rule of law,158 the narrower principle of legality has not yet been held to require such fairness. In Masetlha the majority showed no inclination to change the position. Moseneke DCJ observed that powers to appoint and dismiss this type of official were conferred specially on the President for the effective business of government and the pursuit of national security, and he concluded that ‘it would not be appropriate to constrain executive powers to requirements of procedural fairness’.159

When one considers how well established the audi alteram partem principle is in our law generally, and how well established it is in the context of dismissal (at least in administrative law),160 the majority’s conclusion seems sadly retrogressive. First, in what way would national security or effective government be jeopardised by requiring the President to hear the appellant? Secondly, as Ngcobo J indicates in his dissenting judgment, it seems inconceivable that the content of the rule of law should be less today than it was under the

151 Masetlha (n 4 above) para 75. 152 Masetlha (n 4 above) para 81, where it is referred to as ‘the principle of legality and rationality’. 153 Fedsure (n 35 above) paras 56-9. 154 The SARFU case (n 32 above) para 148. 155 As above. 156 Pharmaceutical Manufacturers Association (n 36 above) para 85. 157 C Plasket ‘The fundamental right to just administrative action: Judicial review of administrative action in the democratic South Africa’ unpublished doctoral thesis, Rhodes University, 2002 164. 158 See C Hoexter ‘The principle of legality in South African administrative law’ (2004) 4 Macquarie Law Journal 165. 159 Masetlha (n 4 above) para 77; and see also para 78. 160 See the cases cited in n 53 above. 232 Clearing the intersection? Administrative law and labour law in the CC old regime.161 In Zenzile, a judgment handed down during the apartheid era, the lowliest employees, temporary hospital workers, were held to have the right to be heard before being dismissed. Now, in the democratic era — under a constitution committed to accountability, responsiveness and openness162 — a member of the public service163 is held to have no entitlement at all to procedural fairness. The majority apparently made nothing of the differences between this appointment and ‘purely political appointees’ such as Cabinet Ministers;164 though I remain unconvinced that under our Constitution it is justifiable to dismiss anyone, however political their appointment, without the benefit of a hearing.

In his dissenting judgment Ngcobo J held that the rule of law required the appellant to be heard before being dismissed. He linked this procedural requirement to the existing substantive requirement of rationality, pointing out that fairness provides insurance against irrationality or arbitrariness — it ensures that the decision-maker has all the facts before making a decision.165 Ngcobo J also reminded us of the importance of the audi principle by quoting from the Court’s own judgment in Zondi v MEC for Traditional and Local Government Affairs:166

A hearing can convert a case that was considered to be open and shut to be open to some doubt, and a case that was considered to be inexplicable to be fully explained.

It was not necessary for Ngcobo J to point out that if the President had acted fairly before dismissing the appellant, misunderstandings between them might conceivably have been cleared up.167

Judicial disagreement about the content of the rule of law is not all there was to the case, however. Moseneke DCJ went on to say that if procedural fairness were indeed a requirement, it had been satisfied in this instance. The appellant had had ‘ample occasion’168 to respond to the allegations made against him in relation to the Macozoma affair: he had had at least two meetings with the Minister at which he was called upon to explain the surveillance and his role in

161 Masetlha (n 4 above) para 188. 162 Section 1 of the Constitution. 163 See Masetlha (n 4 above) para 37. 164 Masetlha (n 4 above) paras 288-9, where Sachs J records the differences: Ministers ‘know they are hired and can be fired at the will of the President; and if fired, they can mobilise politically’, whereas the head of the NIA has ‘one foot in government and one in the public administration’ and has no equivalent political remedies. 165 Masetlha (n 4 above) paras 184-7. 166 2005 3 SA 589 (CC) para 112, quoted in para 204 of his dissenting judgment. 167 Whether they would have been is of course immaterial: see Masetlha (n 4 above) para 204. 168 Masetlha (n 4 above) para 83. (2008) 1 Constitutional Court Review 233 it, and he had made submissions to the Inspector-General’s investigation. There had also been a meeting with the President at which the appellant had expressed his dissatisfaction with the Inspector-General’s findings and the recommendation that disciplinary action should be taken against him. So,

[a]lthough the President did not ask the applicant for his views at the point of dismissing him, he had the benefit of the views of the applicant on all material issues that led to the dismissal.169

With respect, this reasoning gives cause for concern. To say that there is no right to procedural fairness at all in this unique executive setting is one thing. It is quite another to say that if there was such a right, it was satisfied in the circumstances. The potential effect is to dilute the content of the audi alteram partem principle in relation to dismissal — which is just where it deserves to be applied rigorously. As Ngcobo J points out in his dissenting judgment, fairness surely demands that any person being dismissed be heard on the proposed dismissal itself.170 It is certainly not enough, at least not in administrative law, for one party to have had the benefit of the other party’s views in general. The essence of a fair hearing is that the person adversely affected has a chance to comment specifically on the proposed adverse action.171 It is a pity, then, that the majority did not make it clearer that this was an exception to the general principles of procedural fairness, and it is to be hoped that its proposition does not infect those general principles.

The judgment of Sachs J, too, is worrying in this regard. While he found that the appellant was entitled to be treated fairly under section 23 of the Constitution and took a broad view of what substantive fairness demanded, Sachs J seemed to take a strangely constrained view of what procedural fairness required in the circumstances. In his judgment the President ought to have consulted the appellant on the manner in which the termination was to be publicly communicated;172 but, in the absence of a charge of misconduct or other form of breach, Sachs J apparently saw no need for a hearing on the dismissal itself.173 This seems a somewhat formalistic approach, too, since misconduct was surely implied: the Court apparently accepted that the breakdown of trust between the parties was a direct result of the appellant’s conduct in the Macozoma affair — for which the Inspector-General had recommended disciplinary action against him.

169 Masetlha (n 4 above) para 84. 170 Masetlha (n 4 above) para 205. 171 As to the common law, see L Baxter Administrative law (1984) 546; and more generally, see Hoexter (n 38 above) 332ff. 172 Masetlha (n 4 above) para 236. 173 Masetlha (n 4 above) para 234. 234 Clearing the intersection? Administrative law and labour law in the CC

4Conclusion

The cases of Masetlha, Sidumo and Chirwa presented the Constitutional Court with opportunities to pronounce on the intersection of labour law and administrative law in employment matters, and to expound on the relationship between the constitutional rights concerned and their associated pieces of legislation. As I have tried to show in this article, the judgments of the Constitutional Court are full of interest. Sidumo in particular develops the connection between the relevant rights in a creative manner, while Chirwa raises fundamental questions about the concept of administrative action and the nature of public power.

But the judgments of the Constitutional Court in these cases also leave one with a sense of wasted opportunity. Masetlha seems a retrogressive decision, at least to an administrative lawyer, for allowing an unfortunate exception to the established principles of procedural fairness. And Sidumo and Chirwa appear to be far apart in the constitutional vision or ethos expressed by the majority in each case, for the former encourages the interconnectedness of rights while the latter disavows it. As I have argued above, Chirwa is unsatisfying in other respects as well: none of the Court’s approaches to the ‘administrative action’ issue is unproblematic, and the majority’s reasoning on the question of jurisdiction is not only unconvincing but also contradicts the Court’s own jurisprudence. The decision in this case certainly clears up the intersection between labour law and administrative law, but the cost of that tidiness may be thought unacceptably high. SUSTAINABLE DEVELOPMENT IN PRACTICE: FUEL RETAILERS ASSOCIATION OF SOUTHERN AFRICA V DIRECTOR-GENERAL ENVIRONMENTAL MANAGEMENT, DEPARTMENT OF AGRICULTURE, CONSERVATION AND ENVIRONMENT, MPUMALANGA PROVINCE

Loretta Feris*

1Introduction

In the opening salvo of his dissenting judgment in Fuel Retailers1 Justice Sachs notes the irony behind the fact that the first environmental rights case before the Constitutional Court came from ‘an organised section of an industry frequently lambasted both for establishing world-wide reliance on non-renewable energy sources and for spawning pollution’.2 However, this fact does not come as a total surprise. The fuel sector and filling stations in particular have been involved in several cases concerning environmental matters. In some of these cases the issue of sustainable development, which is guaranteed in the South African environmental right, has taken centre stage.3 As a result, even before Fuel Retailers, we have seen the emergence of a South African jurisprudence on sustainable development. Fuel Retailers distinguishes itself by having made it all the way to the Constitutional Court and so having provided an opportunity for this court to express itself on this matter.

* Associate Professor, Department of Public Law, University of Pretoria. 1 2007 6 SA 4 (CC) (Fuel Retailers). 2 Fuel Retailers (n 1 above) para 109. 3 The so-called filling station or fuel related cases include All the Best Trading CC t/a Parkville Motors, and Others v S N Nayagar Property Development and Construction CC and Others 2005 3 SA 396 (TPD); BP Southern Africa (Pty) Ltd v MEC for Agriculture, Conservation, Environment and Land Affairs 2004 5 SA 124 (W); Capital Park Motors CC and Fuel Retailers Association of SA (Pty) Ltd v Shell SA Marketing (Pty) Ltd unreported, case no 3016/05, 18 March 2005 (TPD); Sasol Oil (Pty) Ltd & another v Metcalf 2004 5 SA 161 (W); and MEC for Agriculture, Conservation, Environment and Land Affairs, Gauteng v Sasol Oil and Another (368/2004) (2005) SCA 76.

235 236 Sustainable development in practice

However, it has to be made clear that this case is not primarily about section 24, the environmental right4 of the Bill of Rights. It does not focus on the substance and normative content of the right. Rather, the cause of action lies in administrative law and governance and it is through this prism that one of the concepts central to section 24, ‘sustainable development’, is explored, analysed and evaluated. Much of the case is in fact devoted to sustainable development and Ngcobo J, writing on behalf of the majority, spent a significant part of the judgment exploring the development of the concept in international law, its application in South African law and finally its import and consequence for the matter at hand. In addition, Sachs J in his dissent provides a contrary, and to my mind a more conceptually sound analysis of sustainable development.

As a result of the majority’s expansive overview of sustainable development and Sach’s cogent construction thereof, the case contributes significantly to the conceptualisation of sustainable development in environmental law, a process that is by its very nature ongoing. It underscores and buttresses the development of the concept through earlier case law and as such stimulates the debate on how exactly we should apply sustainable development.

In this note I view Fuel Retailers through the lens of sustainable development and explore the way in which the Constitutional Court has approached the concept. I refrain from engaging in a comprehensive analysis of the historical development of the concept or an in-depth analysis of the normative content thereof. I do, however, consider one of the core elements of sustainable development — integration, an element that took centre stage in Fuel Retailers and is at the heart of sustainable development decision- making. In particular, I investigate the value choices inherent in sustainable development decision-making and consider the failure of the Court to interrogate the practical and normative application of integration thoroughly. I furthermore assess the dissenting judgement of Sachs J and conclude that his interpretation provides some guidance on how to interpret instruments such as legislation and policy that require sustainable development.

4 Section 24 reads: ‘Everyone has the right — (a) to an environment that is not harmful to their health or well-being; and (b) to have the environment protected, for the benefit of present and future generations, through reasonable legislative and other measures that — (i) prevent pollution and ecological degradation (ii) promote conservation; and (iii) secure ecologically sustainable development and use of natural resources while promoting justifiable economic and social development.’ (2008) 1 Constitutional Court Review 237

2The facts

The applicants in this matter objected to an authorisation that was granted by the Mpumalanga provincial environmental authorities for the establishment of a filling station in White River in Mpumalanga. The filling station was to be established in an area that was already the site of six filling stations within a five kilometre radius. An environmental impact assessment (EIA) was submitted to the Mpumalanga Department of Agriculture, Conservation and Environ- ment (DACE) in terms of GN R1182 of 19975 issued under the Environment Conservation Act (ECA).6 A record of decision (ROD) was signed on 9 January 2002 for the installation of three tanks for petrol and diesel, the erection of a convenience store, and ablution facilities. The ROD was issued on the basis of a scoping report, a geotechnical report and a geohydrological report.

The scoping report dealt with the potential impact of noise, visual factors and traffic and the effect on municipal services, safety and crime, and cultural and historical sites; as well as the feasibility of the proposed station. However, it did not address the potential economic impact on other filling stations in the area. With regard to environmental impacts it found that no plant or animal species were threatened. The geotechnical report suggested, however, that a subterranean aquifer would need protection from pollution and recommended that — if required by the Department of Water Affairs and Forestry (DWAF) — an impermeable base layer should be installed to prevent contamination, and that water quality be tested bi- annually.7

Fuel Retailers lodged an appeal to the MEC of DACE on the basis, inter alia, that the need, desirability and sustainability of the proposed filling station had not been considered and that the geophysical report regarding prevention of fuel leaks and possible contamination of the aquifer is inadequate. The appeal was considered and dismissed. In response, the Director argued that need, desirability and sustainability need not be considered, as these factors had been considered by the local authority when it had earlier approved the rezoning of the area from ‘special’ to ‘Business 1’,8

5 GG 18261 of 5 September 1997. 6 Act 73 of 1989. In terms of sec 22(1) read with sec 21(1) no activity which may have a substantial detrimental impact on the environment may be undertaken without written authorisation from the competent authority. In terms of sec 22(2) such authorisation may only be granted ‘after consideration of reports concerning the impact of the proposed activity and of alternative proposed activities on the environment’; and after the furnishing of a report concerning the impact of the proposed development on the environment. 7 At no point did DWAF require any preventive measures. 8 As required by the relevant Ordinance 15 of 1986, which governs rezoning. 238 Sustainable development in practice which allowed for the construction of a filling station at the identified site.

The applicants unsuccessfully applied to the High Court9 and to the Supreme Court of Appeal10 to set aside the ROD in terms of the ECA, the common law and the Promotion of Administrative Justice Act.11 The applicants’ arguments were based on a number of issues, but on application for leave to appeal to the Constitutional Court, only one issue was pursued, namely that the environmental authorities failed to consider the socio-economic impact of constructing the proposed filling station. The applicants argued that this obligation is wider than the assessment of need and desirability considered in the rezoning process and that it required DACE to assess, amongst other things, the cumulative impact on the environment produced by the proposed filling station and all existing filling stations in close proximity to the proposed one. In sum, the proper analysis required an assessment of demand or necessity and desirability, not the feasibility, of the proposed station and its impact on the sustainability of existing filling stations. In support, the applicant relied upon the provisions of section 24(b)(iii)12 of the Constitution, as well as sections 2(4)(a), 2(3), 2(4)(g), 2(4)(i), and 23 of NEMA.13

9 2005 JDR 0915 (T). 10 2007 2 SA 163 (SCA). 11 Act 3 of 2000. 12 It provides: ‘Everyone has the right ... to have the environment protected, for the benefit of present and future generations, through reasonable legislative and other measures that ... secure ecologically sustainable development and use of natural resources while promoting justifiable economic and social development’. 13 2(4)(a): Sustainable development requires the consideration of all relevant factors including the following: (i) That the disturbance of ecosystems and loss of biological diversity are avoided, or, where they cannot be altogether avoided, are minimised and remedied; (ii) that pollution and degradation of the environment are avoided, or, where they cannot be altogether avoided, are minimised and remedied; (iii) that the disturbance of landscapes and sites that constitute the nation's cultural heritage is avoided, or where it cannot be altogether avoided, is minimised and remedied; (iv) that waste is avoided, or where it cannot be altogether avoided, minimised and re-used or recycled where possible and otherwise disposed of in a responsible manner; (v) that the use and exploitation of non-renewable natural resources is responsible and equitable, and takes into account the consequences of the depletion of the resource; (vi) that the development, use and exploitation of renewable resources and the ecosystems of which they are part do not exceed the level beyond which their integrity is jeopardised; (vii) that a risk-averse and cautious approach is applied, which takes into account the limits of current knowledge about the consequences of decisions and actions; and (viii)that negative impacts on the environment and on people's environmental rights be anticipated and prevented, and where they cannot be altogether prevented, are minimised and remedied. (2008) 1 Constitutional Court Review 239

The environmental authorities accepted that the socio-economic impact of the proposed filling station had to be considered, but argued that they were not required to consider need, desirability and sustainability of the proposed filling station as it was taken into account by the local authority during the rezoning of the property and that ‘rezoning forms part and parcel of the process of an application for authorisation in terms of section 22 of ECA’.14 The applicant, in turn, argued that the two processes are distinct and separate. The local authority considers a rezoning application from a town-planning perspective and focuses on what land uses it would allow, whilst the environmental authorities are required to consider the impacts of the proposed development on the environment and on socio-economic conditions.15 The applicant also noted that the rezoning had taken place approximately eight years prior to the approval of the application for authorisation in terms of the ECA and that significant environmental changes had occurred in the area since then.16

13 2(3): Development must be socially, environmentally and economically sustainable. 2(4)(g): Decisions must take into account the interests, needs and values of all interested and affected parties, and this includes recognising all forms of knowledge, including traditional and ordinary knowledge. 2(4)(i): The social, economic and environmental impacts of activities, including disadvantages and benefits, must be considered, assessed and evaluated, and decisions must be appropriate in the light of such consideration and assessment. 23: The purpose of this Chapter is to promote the application of appropriate environmental management tools in order to ensure the integrated environmental management of activities. (2) The general objective of integrated environmental management is to — (a) promote the integration of the principles of environmental management set out in section 2 into the making of all decisions which may have a significant effect on the environment; (b) identify, predict and evaluate the actual and potential impact on the environment, socio-economic conditions and cultural heritage, the risks and consequences and alternatives and options for mitigation of activities, with a view to minimising negative impacts, maximising benefits, and promoting compliance with the principles of environmental management set out in section 2; (c) ensure that the effects of activities on the environment receive adequate consideration before actions are taken in connection with them; (d) ensure adequate and appropriate opportunity for public participation in decisions that may affect the environment; (e) ensure the consideration of environmental attributes in management and decision-making which may have a significant effect on the environment; and (f) identify and employ the modes of environmental management best suited to ensuring that a particular activity is pursued in accordance with the principles of environmental management set out in section 2. (3) The Director-General must coordinate the activities of organs of state referred to in section 24(1) and assist them in giving effect to the objectives of this section and such assistance may include training, the publication of manuals and guidelines and the co-ordination of procedures. 14 Fuel Retailers (n 1 above) para 31. 15 Fuel Retailers (n 1 above) para 32. 16 Fuel Retailers (n 1 above) para 33. 240 Sustainable development in practice

3The judgment

The cause of action for judicial review centred on whether or not the environmental authorities had failed to take into consideration matters that they were required to consider before granting the authorisation under section 22(1) of the ECA.17 Ngcobo J framed the issue as one that concerns the ‘nature and scope of the obligations of environmental authorities when they make decisions that may have a substantial detrimental impact on the environment’; and, in particular, ‘the interaction between social and economic development and the protection of the environment’.18 It is this latter aspect — that is, the interaction between social and economic development and environmental protection — that provides the impetus for the lengthy interrogation of sustainable development by the court.19

3.1 Sustainable development

Ngcobo J’s departure point is the inherent challenge that gave rise to the concept of sustainable development: namely that the promotion of development requires protection of the environment, whilst the environment cannot be protected if development does not pay attention to the costs of environmental destruction.

This challenge has long been recognised and the Court details how this challenge gave rise internationally to the development of sustainable development and it traces this development to important events such as the Brundtland Report,20 the 1972 United Nations Conference on the Human Environment or so-called ‘Stockholm Conference’21 (which stressed the relationship between development and the protection of the environment), the 1992 United Nations Conference on Environment and Development or so-called ‘Rio Conference’ (which generated principles on sustainable development and provided a framework for the development of the law of sustainable development)22 and, finally, the 2002 Johannesburg

17 Fuel Retailers (n 1 above) para 39. 18 Fuel Retailers (n 1 above) para 1. 19 In Fuel Retailers (n 1 above) para 40 Ngcobo J confirms that in light of the dictates of sec 24 of the Constitution, the case does, indeed, raise a constitutional issue. What is disconcerting though is the way in which he frames sec 24. He states that sec 24 ‘guarantees to everyone the right to a healthy environment’. This is a positive framing of the right, whilst the right is in fact negatively framed, thus lessening the impact and extent of the right. One can only hope that this careless use of language was a mere oversight on the side of the Court. 20 Fuel Retailers (n 1 above) para 44. 21 Fuel Retailers (n 1 above) para 46. 22 Fuel Retailers (n 1 above) para 49-50. (2008) 1 Constitutional Court Review 241

World Summit on Sustainable Development (WSSD) which, according to the Court ‘reaffirm[ed] that sustainable development is a world priority’.23

In delineating the concept the Court refers to the Brundtland Commission’s definition of sustainable development as ‘development that meets the needs of the present without compromising the ability of future generations to meet their own needs,’24 but noted that recent commentators have refrained from defining sustainable development and have instead chosen to identify ‘the evolving elements of the concept of sustainable development’. These include:

the integration of environmental protection and economic development (the principle of integration); sustainable utilisation of natural resources (the principle of sustainable use and exploitation of natural resources); the right to development; the pursuit of equity in the use and allocation of natural resources (the principle of intra-generational equity); the need to preserve natural resources for the benefit of present and future generations (the principle of inter-generational and intra-generational equity); and the need to interpret and apply rules of international law in an integrated systematic manner.25

In highlighting the first principle — that of integration as fundamental to the concept of sustainable development — the Court notes that the formal application thereof ‘requires the collection and dissemination of environmental information, and the conduct of environmental impact assessments’ and argues that the ‘practical significance of the integration of the environmental and developmental considerations is that environmental considerations will now increasingly be a feature of economic and development policy’.26

With regard to South African law, the Court is of the opinion that sustainable development offers a principle for the resolution of tensions between ‘the need to protect the environment on the one hand, and the need for socio-economic development on the other hand’. The concept thus, according to the Court, provides a framework for reconciling socio-economic development and environmental protection.27 Ngcobo J states: ‘sustainable development does not require the cessation of socio-economic development but seeks to regulate the manner in which it takes place’.28 According to the Court the National Environmental Management Act (NEMA), which was enacted to give effect to section

23 Fuel Retailers (n 1 above) para 46. 24 Fuel Retailers (n 1 above) para 47. 25 Fuel Retailers (n 1 above) para 51. 26 Fuel Retailers (n 1 above) para 52 (references omitted). 27 Fuel Retailers (n 1 above) para 57. 28 Fuel Retailers (n 1 above) para 58. 242 Sustainable development in practice

24 of the Bill of Rights, defines sustainable development as being the ‘integration of social, economic and environmental factors into planning, implementation and decision-making for the benefit of present and future generations’.29 Ngcobo J argues that ‘whenever a development which may have a significant impact on the environment is planned, it envisages that there will always be a need to weigh considerations of development’, with these considerations being ‘underpinned by the right to socio-economic development’, against ‘environmental considerations, as underpinned by the right to environmental protection. In this sense it contemplates that environmental decisions will achieve a balance between environmental and socio-economic developmental considerations through the concept of sustainable development’.30

3.2 Sustainable development and the EIA process

The Court argues that the

nature and scope of the obligation to consider the impact of the proposed development on socio-economic conditions must be determined in the light of the concept of sustainable development and the principle of integration of socio-economic development and the protection of the environment.31

If this relationship is accepted, so Ngcobo J argues, it follows that socio-economic conditions have an impact on the environment. He illustrates this point by stating that the proposed filling station may have implications for the job security of existing filling stations, but may equally have environmental consequences, such as that if existing stations close down that will require rehabilitation of the environment after closure.32

He furthermore cautions that the proliferation of filling stations close to each other may increase the likelihood of adverse impacts on the environment. In this respect the Court reminds us that NEMA ‘requires that the cumulative impact of a proposed development, together with the existing developments on the environment, socio- economic conditions and cultural heritage must be assessed’. As such ‘a consideration of socio-economic conditions ... includes the consideration of the impact of the proposed development not only in

29 Fuel Retailers (n 1 above) para 59. The reference indicated is sec 1(1)(xxix). It should be noted, however, that this section does not exist and this ‘definition’ is in fact from the preamble to NEMA. 30 Fuel Retailers (n 1 above) para 61. 31 Fuel Retailers (n 1 above) para 71. 32 As above. (2008) 1 Constitutional Court Review 243 combination with the existing developments, but also its impact on existing ones’.33

With respect to the EIA process, the Court notes that NEMA requires the consideration, assessment and evaluation of the social, economic and environmental impact of proposed activities and this ‘enjoins the environmental authorities to consider and assess the impact of a proposed activity on existing socio-economic conditions, which must of necessity include existing developments’.34 The Court explains that the objective of this exercise is not to stamp out competition or protect existing developments at the expense of future developments. Rather it requires a risk-averse and cautious approach when decisions are made.35 It concludes that what was demanded of the environmental authorities was to consider ‘the impact on the environment of the proliferation of filling stations as well as the impact of the proposed filling station on existing ones’, an obligation which is ‘wider than the requirement to assess need and desirability under the Ordinance’.36

It was common cause that this assessment never happened. The Court, however, contends that this duty fell on environmental authorities because the two organs of state have different roles and functions. Local authorities have a town planning role and as such a proposed development may satisfy the need and desirability criteria from a town-planning perspective, given that the local authority is not required to consider the social, economic and environmental impact of a proposed development. Environmental authorities, however, are required to do so by the provisions of NEMA.37 The assumption that the duty under the Ordinance to consider need and desirability imposed the same obligation as is required by the duty under NEMA to consider the social, economic and environmental impacts of a proposed development, is therefore wrong.38

The Court, as a result sets aside the decision of the environmental authorities on a number of bases:

(i) They misconstrued the nature of their obligations under NEMA and consequently failed to apply their minds to the socio-economic impacts of the proposed filling station in the instant case.39

33 Fuel Retailers (n 1 above) para 72. See also para 74. 34 Fuel Retailers (n 1 above) para 77. 35 Fuel Retailers (n 1 above) paras 78-81. 36 Fuel Retailers (n 1 above) para 82. 37 Fuel Retailers (n 1 above) para 85. 38 Fuel Retailers (n 1 above) para 86. 39 As above. 244 Sustainable development in practice

(ii) They left the consideration of need and desirability to the local authority and consequently failed to properly discharge of their statutory duty.40 (iii) They failed to take into account socio-economic conditions as required by NEMA and as such they did not comply with a mandatory and material condition set by NEMA.41 (iv) Given the lack of consideration to the social and economic impact of the proposed filling station, including its cumulative impact on the environment, the environmental authorities failed to apply their minds to the matter.42

4 Discussion

4.1 Sustainable development through the cases

As mentioned above Fuel Retailers focuses on the concept of sustainable development and more specifically how it is to be interpreted, contextualised and applied in South African law. The Court thoroughly interrogated the concept’s history and status in international law and concluded that it is now an established concept in international law. With regard to South African law the concept is still in its infancy.43 A number of cases have, however, addressed sustainable development and over the last four years a South African jurisprudence has begun to see the light.

The first case to explore this concept was one very similar to Fuel Retailers: Sasol Oil (Pty) Ltd & another v Metcalf.44 The applicants

40 Fuel Retailers (n 1 above) paras 87-88. 41 Fuel Retailers (n 1 above) para 89. 42 Fuel Retailers (n 1 above) para 90–97. The Court emphasises the fact that the rezoning decision by the local authorities, on which the environmental authorities relied, had probably occurred in 1995, some eight years prior to the decision by the environmental authorities. This was done prior to the coming into effect of NEMA and it argues that it was necessary for the environmental authorities to consider the matter afresh in the light of the provisions of NEMA, given that a significant change in the environment could have taken place in the intervening period. 43 Note, however, that as early as 1971 our neighbouring (then) Rhodesian court recognised that the principle of inter-generational equity can have the effect of limiting other rights, such as property rights. Thus in King v Dykes 1971 3 SA 540 (RA) 545G-H MacDonald ACJ stated: ‘The idea which prevailed in the past that ownership of land conferred the right on the owner to use his land as he pleased is rapidly giving way in the modern world to the more responsible conception that an owner must not use his land in a way which may prejudice his neighbours or the community in which he lives, and that he holds his land in trust for future generations. Legislation dealing with such matters as town and country planning, the conservation of natural resources, and the prevention of pollution, and regulations designed to ensure that proper farming practices are followed, all bear eloquent testimony of the existence of this more civilised and enlightened attitude towards the rights conferred by ownership of land.’ 44 Sasol (n 3 above). (2008) 1 Constitutional Court Review 245 sought the respondent’s, the Gauteng Department of Agriculture, Conservation, Environment and Land Affairs’, authorisation in terms of section 22 of the ECA for the construction of a filling station and convenience store. This authorisation was refused and in its refusal GDACE was guided by departmental guidelines that it had developed to assist in decision-making. The guidelines stipulated that development must be socially, environmentally and economically sustainable and required that filling stations in an urban, residential or built-up area should not be situated within three kilometres of an existing filling station. The applicants thus approached the High Court for a declaration that certain sections of these guidelines issued by the respondent were ultra vires the ECA and were, therefore, invalid and unenforceable. They furthermore argued that whilst GDACE had the power to authorise the construction of structures for the storage and handling of hazardous substances at filling stations, this power did not include the erection of filling stations itself. As such the guidelines did not apply. Finally they argued that even if the power did include the erection of filing stations, GDACE was not entitled to apply the guidelines as they were predominantly based on socio- economic as opposed to environmental considerations. The High Court dismissed the ultra vires argument, but upheld the applicants’ argument that GDACE did not have the power to require authorisation for the construction of filing stations per se.45 It furthermore held that the sustainable development principles contained in section 2 of the NEMA did not extend the mandate of GDACE to take socio- economic (and not just environmental) factors into account.46

This decision was a set-back for sustainable development jurisprudence. It ignored the core element of integration as part of sustainable development.47 The Brundtland Report, referred to by Ngcobo J in Fuel Retailers, contextualises the principle of integration and explains that sustainable development not only prioritises the needs of the poor, but that it also captures the limitations to development imposed by the present state of technology and social organisation on the environment’s ability to meet present and future needs.48 The Report suggests an inherent link between social and environmental needs and the need for technological advancement and development. An imbalance amongst these elements, where global patterns of development put the environment under pressure, places the earth in crisis. This principle of integration between three pillars — environmental protection, economic development and social

45 Sasol (n 3 above) 170A-E. 46 Sasol (n 3 above) 171E-172B. 47 For a more in-depth critique of this case see Patterson A ‘Fuelling the sustainable development debate in South Africa’ (2006) 1 South African Law Journal 53. 48 Report of the World Commission on Environment and Development: Our Common Future (1987). 246 Sustainable development in practice needs — is thus widely recognised as a core element of sustainable development.49

This decision was eventually overturned by the SCA.50 The SCA’s reversal was, some ways, due to a conflicting judgement by the same court in yet another filling station case, BP Southern Africa (Pty) Ltd v MEC for Agriculture, Conservation, Environment and Land Affairs.51 The applicant in this case sought the review and setting aside of a decision by the Gauteng Provincial Department of Agriculture, Conservation, Environment and Land Affairs (GDACE) to refuse its application in terms of section 22(1) of the ECA for authorisation to develop a filling station on one of its properties. The Department based its refusal, inter alia, on environmental concerns. The applicant contended, however, that its application was refused not because the new filling station itself posed a danger to the environment, but rather, because of the fact that there were already two other filling stations within three kilometres of applicant's site and the Department regarded it as unacceptable to allow proliferation of filling stations where existing filling stations were economically vulnerable to more competition. It argued that under the guise of ‘environmental concerns’, the department was instead seeking to regulate the economy on the basis of what were essentially economic considerations unrelated to the environment.

The court explored the concept of sustainable development in somewhat more detail than in Sasol and in so doing considered the wide definition of ‘environment’ employed by ECA which defines it in section 1 as: ‘the aggregate of surrounding objects, conditions and influences that influence the life and habits of man or any other organism or collection of organisms’. According to the court the broad definition of ‘environment’ would include all conditions and influences affecting the life and habits of man, which would also include socio-economic conditions and influences.52 With reference to the state’s obligation under section 24(b) with regard to sustainable development, the Court held that ecologically sustainable development and the use of natural resources must be promoted jointly with justifiable economic and social development.53 It stated that:

49 See for example P Sands Principles of international law (2003) 153. See also D Tladi Sustainable development in international law: An analysis of key enviro- economic instruments (2007) 58. 50 MEC for Agriculture, Conservation, Environment and Land Affairs (Pty) Ltd & another (n 3 above). 51 2004 5 SA 124 (W). 52 BP (n 3 above) 145E. 53 BP (n 3 above) 143C-D. (2008) 1 Constitutional Court Review 247

... sustainable development constitutes an integral part of modern international law and will balance the competing demands of development and environmental protection. The concept of ‘sustainable development’ is the fundamental building block around which environmental legal norms have been fashioned, both internationally and in South Africa ... pure economic principles will no longer determine, in an unbridled fashion, whether a development is acceptable. Development, which may be regarded as economically and financially sound, will, in future, be balanced by its environmental impact, taking coherent cognisance of the principle of inter- generational equity and sustainable use of resources in order to arrive at an integrated management of the environment, sustainable development and socio-economic concerns. By elevating the environment to a fundamental justiciable human right, South Africa has irreversibly embarked on a road, which will lead to the goal of attaining a protected environment by an integrated approach, which takes into consideration, inter alia, socio-economic concerns and principles.54

The court thus concluded that the department was obliged to develop an integrated environmental management programme, which took cognisance of a wide spectrum of considerations, including international conventions and approaches as a result of the broad and extensive definition of ‘environment’ in ECA and the principles of NEMA which, inter alia, include the consideration of socio-economic conditions.55

With the more detailed attention to sustainable development provided by the Constitutional Court in Fuel Retailers we now have significant judicial guidance on interpreting and implementing sustainable development. In essence we now know that sustainable development is central to the environmental right and to environmental regulation. We also know that our jurisprudence acknowledges the inter-connected nature of environmental, social and economic considerations of sustainable development and that the principle of sustainable development aims (albeit, as argued further on, mostly unsuccessfully) to serve the purpose of facilitating the achievement of this balance. One could, therefore argue, that a central tenet of a ‘South African jurisprudence’ on sustainable development is the principle of integration.

4.2 Sustainable development applied

As noted by Ngcobo J in Fuel Retailers, the overarching definition of sustainable development is the one as espoused by the Brundtland Report: ‘development which meets the needs of the present

54 BP (n 3 above) 144A-144D. 55 BP (n 3 above) 150D-E. 248 Sustainable development in practice generation without compromising the ability of future generations to meet their own needs’.56 In attempting to find the practical application of sustainable development it may be helpful to view this definition of the Brundlandt Report as the aim of sustainable development — that is, that which we want to achieve. This aim operates in acknowledgment of the fact that whilst human beings are driven by their developmental needs to use, exploit and even exhaust natural resources, this can and may not happen in a limitless way. Thus, as noted by Field, sustainable development could be described as the ‘conceptual vehicle chosen by a diverse range of actors to negotiate the tensions arising from the need for social and economic development on a planet with finite resources’.57

This earlier definition has been elaborated upon by more recent authors through the identification of ‘the evolving elements of the concept of sustainable development’.58 I would suggest that these elements can, in turn, be viewed as the ‘means to achieve the aim’. These means would therefore include: sustainable utilisation of natural resources, the pursuit of equity in the use and allocation of natural resources, and the integration of environmental protection and economic development.59 These elements attempt to give concrete existence to a concept that may be viewed as elusive and impractical, largely because the concept involves competing considerations or normative tensions.

Out of the three elements, the integration principle has been identified as the key principle.60 It is also the element that the Court focused on when it suggested that sustainable development requires integrating the often contesting demands of economic development, social development and environmental protection. This element remains, however, open to criticism. It has been suggested, not without justification, that its contents remain rather opaque.61

What, for example does ‘integration’ really mean and how is it practically achieved? If a waste site is situated close to a residential area, where that site generates an income not only for the managing company, but also for the residents that live nearby, should that site

56 Report of the World Commission on Environment and Development: Our Common Future (1987) 8. 57 Field T ‘Sustainable development versus environmentalism: Competing paradigms for the South African EIA regime’ (2006) 123 South African Law Journal 409 411. 58 Fuel Retailers (n 1 above) para 51. 59 See Sands P Principles of international environmental law (2003) 253. Ngcobo refers to these elements, but also includes a wider range of elements (para 51). These three elements are, however, the most widely recognised elements of sustainable development. 60 D Tladi Sustainable development in international law: An analysis of key enviro- economic instruments (2007) 58. 61 Tladi (n 60 above) 75. (2008) 1 Constitutional Court Review 249 be closed down so as accommodate environmental health considerations? Or should it be allowed to remain open in order to accommodate social and economic considerations? How does one integrate, if at all, these three contesting considerations? One could argue that integration is the ‘happy medium’ where one tightly regulates the operations of the waste site so as to minimise the exposure of the nearby residents, while still ensuring that the site contribute to the economy and provides a source of income for its workers. In reality this happy medium is a value choice. In this instance the decision-making is primarily driven by socio-economic considerations. Requiring strict operating conditions, however, means that the third pillar, the environment, remains part of the overall decision-making process.

The practical and normative application of integration was only marginally addressed by the Court in Fuel Retailers as the ‘collection and dissemination of environmental information’, the ‘conduct of environmental impact assessments’ and the fact that environmental considerations ‘will now increasingly be a feature of economic and development policy’.62 While EIAs and policy tools may be useful in the practical application of the integration element of sustainable development, they clearly do not address other forms of decision- making, such as decisions regarding activities where no EIAs were conducted, nor do they explain how these decisions are to be made even when there is an EIA.

A current burning issue for example is that of land restitution and its relationship vis-à-vis sustainable development.63 A number of current land claims include claims to land that have been declared protected areas. This includes both private and state-owned land, including for example claims in the Kruger National Park. In making a decision on whether to award such claims decision-makers would have to take into account the possibility that claimants may not utilise the land for conservation purposes, but rather to engage in other commercial ventures such as farming. This would clearly promote an economic and social goal as opposed to an environmental goal. However, whilst the environmental aim of preserving our natural heritage may weigh very heavy, equally so the would the idea of restoring land to people who were unjustly deprived thereof in the past.

The judgment also does not address the notion that the integration principle could be used equally effectively by diverse

62 Fuel Retailers (n 1 above) para 52. 63 For a more detailed discussion see A Du Plessis ‘Land restitution through the lens of environmental law: Some comments on the South African vista’ (2006) 1 Potchefstroom Electronic Law Journal 1 1. 250 Sustainable development in practice groups with conflicting aims, such as environmentalists versus those pursuing economic development aims.64 Tladi argues that sustainable development is inherently a flexible concept, which would have the effect that for those advocating economic growth, the emphasis would fall on the economic growth value of sustainable development. As such sustainable development could mean ‘lasting economic growth’, with the aim being to sustain economic growth.65 Such an understanding effectively dilutes and detracts from the original aim of sustainable development. The other two elements, sustainable use of natural resources and equitable utilisation of natural resources, attempt to safeguard the environment against unbridled economic development.

Decisions motivated by socio-economic considerations can therefore potentially be disguised as decisions prompted by environmental concerns. This was, in fact, the argument made by the applicants in BP, namely that the decision to refuse authorisation for the proposed activity was motivated not by environmental considerations, but rather by the socio-economic considerations as set out in the province’s guidelines on EIAs.

True integration is often depicted as three circles that intersect with sustainable development situated in the intersection of these circles as shown in the diagram below.

However, whilst the diagram suggests optimal overlap it is always possible that reality indicates otherwise and there may be cases where there is very little overlap and where the greater part is in one

64 As above. 65 As above. (2008) 1 Constitutional Court Review 251 of the circles. In other words, the three elements, environmental sustainability, economic sustainability and social sustainability do not always carry equal weight in decision-making. When a decision- maker, whether it is an administrative official or a judicial officer, takes into account sustainable development in the decision-making process, he or she makes a value-based judgement. This judgment is informed by the values of environmental sustainability, and economic sustainability as part of the integration process. However, one (or sometimes two) of these values may inevitably triumph.

Tladi therefore suggests a more nuanced approach in the application of sustainable development, one that provides three variations of integration based on the value that is the preferred one in cases of conflict. In the economic growth-centred variation, economic growth takes centre stage, whilst in the environment- centred variation, the natural environment triumphs. Finally in the human needs-centred (or social needs centred) variation the social needs of humans are placed at the forefront.66 He argues that such a variation approach allows decision-makers to decide which variation best serves the aims of sustainable development.67 In essence his analysis suggests that sustainable development decisions are inevitably value driven. It means that decision-makers decide in advance which of the values they prefer to advance, and whilst still taking into account the other two values, base the decision on the preferred value.

If we then apply this ‘variation approach’ to integration, it would lead us to the conclusion that the decision in Fuel Retailers may fall within the economic growth or human centred (or a combination of both) variation since the Court has emphasised the failure of the environmental authorities to take socio-economic considerations into account? I would suggest not. Whilst the integration process is a value- driven process, the preferred value cannot be without a legitimate basis. In other words, a decision maker’s decision should be grounded in law and there should be some justifiable base in law for the preferred value. Such a basis may be found in a legal or policy instrument and may provide an indication of the preferred variation of sustainable development.

In this respect Sachs J’s dissent is instructive. In essence he provides us with the application of this ‘variation’ approach to the integration element of sustainable development and takes NEMA as his ‘legitimising base’. With regard to the application of the preamble and principles of NEMA, he notes that ‘economic sustainability is not

66 Tladi (n 60 above) 80. His idea is not that placing one value centre stage would obliterate the other, but rather that it would re-enforce the other two. 67 Tladi (n 60 above) 82. 252 Sustainable development in practice treated as an independent factor to be evaluated as a discrete element in its own terms’, but rather that the focus is on the inter- relationship between economic sustainability and environmental protection.68 Accordingly, he argues, NEMA does not envisage that social, environment and economic sustainability should proceed along separate tracks, with each being assessed separately and only considered together at the end of the decision-making process. It is his contention that economic sustainability takes on significance only to the extent that it implicates the environment. As such, it is only ‘when economic development potentially threatens the environment that it becomes relevant to NEMA’ and it is only at this point that it should be considered within the context of the sustainable development requirements of NEMA.69 Sachs bases this conclusion on the thesis that ‘all environmental controls were in place and that any potential deleterious effect of over-trading was speculative and remote’.70

Sachs situates his position within the dictates of sustainable development as required by NEMA. The overall aim of NEMA is to first and foremost ensure environmental protection. NEMA thus grants priority to the environment-centred variation of sustainable development. In situations of conflict between economic, social and environmental considerations, the latter must be preferred. Given that NEMA operates within this model Sachs’s argument that social and economic considerations are only ‘triggered’ once the environment is implicated makes sense.

Sachs could equally have used section 24 of the Constitution as his ‘legitimising base’. Section 24(b)(iii) of the Constitution refers to the need to ‘secure ecologically sustainable development’. [own emphasis]. It can be argued that ‘ecologically’ qualifies the type of sustainable development that is envisioned by the Constitution. It therefore clearly places an emphasis on environmental considerations and as such it places the environmental value centre-stage. Section 24 of the Constitution therefore mandates the environment-centred variation of sustainable development. Any decision-making regarding sustainable development that is mandated by section 24 should, therefore, be situated within this model.

The Sachs approach could be seen as ‘applied variation’ as it gives us some guidance on how to interpret instruments such as legislation and policy that requires sustainable development. In applying the model not only to the decision in the Fuel Retailers case but also the decision in the BP case, it exposes both for their inadequate and

68 Fuel Retailers (n 1 above) para 113. 69 As above. 70 Fuel Retailers (n 1 above) para 112. (2008) 1 Constitutional Court Review 253 ultimately unsatisfying application of sustainable development. Whilst both decisions were at first glance ‘good for the environment’, they were really motivated by socio-economic considerations, as such applying economic centred variations of integration when the Constitution and NEMA really required environment centred variations. If the latter were applied, different outcomes would have followed in both cases.

5Conclusion

Fuel Retailers may have been hailed as a victory for the environment and for environmental rights, but was it the correct decision from a normative point of view? Sustainable development is a truly laudable concept, but as pointed out it raises a number of challenges in negotiating the increasingly contested demands of development and environmental protection. It embodies competing values and requires other conceptual and normative framework to address these tensions and to provide practical solutions. On its own, the concept fails to address practical and normative considerations.

We therefore need a more principled, value-based approach to sustainable development that identifies the value that is being prioritised in particular circumstances. The variation approach to integration takes into account that certain norms and values will most often be paramount in sustainable development decision-making.

Whilst Fuel Retailers has highlighted the ongoing debate on sustainable development in the South African context, it ought not have been the final word. The majority judgement failed to interrogate the normative nature of sustainable development comprehensively and in the process provided us with an inherently flawed and incomplete application of the concept. Sachs’ dissent provides an alternative, and more nuanced perspective on this complex norm.

FUEL RETAILERS, SUSTAINABLE DEVELOPMENT & INTEGRATION: A RESPONSE TO FERIS

* Dire Tladi

1Introduction

Fuel Retailers is, without question, a most interesting and significant decision: (1) Interesting because the decision tackles head on the difficult task of balancing apparently conflicting interests — the right to a healthy environment and economic imperatives of development; (2) Significant because, as the first decision in which the Constitutional Court gives content to section 24 of the Constitution, one expects the decision to play a major role in the development of environmental law and sustainable development.

Loretta Feris’ comment1 provides a fairly detailed (and accurate) overview of Fuel Retailers.2 The exploration of the nuances of sustainable development in Professor Feris’ contribution, rooted as it is in pragmatism, makes the contribution particularly useful to practitioners — especially those who evaluate applications for approval for activities with a potential impact on the environment.

As someone concerned more with the theoretical and normative aspects of sustainable development, I thought the paper important for engaging the complexity of the process of integration in sustainable development discourse — a complexity that is often

* Legal Counsellor to the South African Permanent Mission to the United Nations in New York. Extraordinary Professor in the Department of Public Law at the University of Stellenbosch. 1 L Feris ‘Sustainable development in practice: Fuel Retailers Associatioin of Southern Africa v Director-General Environmental Management, Department of Agriculture, Conservation and Environment, Mpumalanga Province’ (2008) 1 Constitutional Court Review 235. 2 Fuel Retailers Association of Southern Africa v Director-General Environmental Management, Department of Agriculture, Conservation and Environment, Mpumalanga Province, and Others 2007 6 SA 4 (CC).

255 256 A response to Feris missed because of our collective reverence for the concept of sustainable development.3 Feris and I are in accord over a large range of issues, in particular the view that Sachs J’s dissent provides the most ‘conceptually sound analysis of sustainable development’.4 Our common ground established, I would now like to offer my slightly different take on what the Fuel Retailers Court misses.

2 The Court’s treatment of sustainable development

Feris states upfront that she will avoid a ‘comprehensive analysis of the historical development of the concept or the normative content’ of sustainable development.5 For me, the historical development of the concept is critical for understanding the normative content, which, in turn, influences its practical application. Ngcobo J’s majority judgment does, in fact, survey the historical evolution of sustainable development — he traces it back to the 1972 Stockholm Conference on the Human Environment.6 The historical exposition of sustainable development by the Court suffers from the same ailment as many similar expositions. First, the Court erroneously states that the term sustainable development was ‘coined’ by the Brundtland Commission in 1987.7 Second, the Court does not explain how this evolution influences the conceptualisation that follows in the rest of the judgement. The Court offers, so to speak, a recollection of history for history’s sake. While time and space do not allow for a discussion of how the historical evolution of sustainable development ought to influence its conceptualisation, the Court’s history is notably flawed in this regard.8

The problem with the Court’s analysis, in my view, begins when the Court, in connection with section 24 of the Constitution, makes reference to the ‘explicit recognition of the obligation to promote justifiable “economic and social development”’ and links this notion with the ‘well-being of human beings’ and ‘socio-economic rights’.9 Precisely what that link is, the Court never explains or explores. But the result of this linkage is that, throughout the judgement, the terms

3 Someone, I cannot recall who or in what context, once remarked that sustainable development, like human rights, is a concept that is not to be questioned in polite company. 4 Feris (n 1 above) 236. 5 Feris (n 1 above) 236. 6 Fuel Retailers (n 2 above) para 46. 7 Fuel Retailers (n 2 above) para 47. Already in 1980, the IUCN World Conservation Strategy contained several references to sustainable development. 8 For my views, see D Tladi Sustainable development and international law: An analysis of key enviro-economic instruments (2007) 34 et seq. 9 Fuel Retailers (n 2 above) para 44. (2008) 1 Constitutional Court Review 257

‘socio-economic rights’, ‘development’ and ‘economic development’ are used interchangeably as the values that most often oppose the right to a clean and a healthy environment.10 At one point the Court, for example, refers to the integration of environmental protection and economic development.11 Elsewhere, the Court states that as a result of sustainable development ‘environmental considerations will now increasingly be a feature of economic and development policy’.12 Further on, the Court states that ‘economic development, social development and the protection of the environment’ are considered to be the three pillars of sustainable development.13 Finally, the Court asserts that sustainable development ‘provides a framework for reconciling socio-economic development and environmental protection’.14

By lumping these concepts together the Court misses an opportunity to develop a sound understanding of sustainable development. The result of treating these concepts as interchangeable is that the Court never stops to ask whether the factors that the Fuel Retailers Association requested that the environmental authorities consider are socio-economic or purely economic. To use language from the common definition of sustainable development, the Court does not ask whether these factors are social or economic. The Court’s judgment implies — incorrectly — that economic considerations are the same as social considerations.

A similar conflation of these concepts is evident in Feris’ analysis. After applying the model of three variations of sustainable development that I have proposed, she asks whether Fuel Retailers would ‘fall within the economic growth or human-centred (or a combination of both) variation’ of sustainable development because of the failure of the authorities to consider ‘socio-economic considerations’.15 Although Feris answers the question in the negative, it is the manner in which she treats economic growth centred variation and human well-being centred variation as similar that is of some concern.

One can readily understand the intuitive lumping together of these two variations of sustainable development (and the values they represent). After all, at the centre of both of these variations are human-related concerns, namely economic concerns and social concerns. However, in my view these variations (and their values) tend to pull in different directions. Factors relevant for economic

10 Fuel Retailers (n 2 above) paras 51, 52, 53 & 55. 11 Fuel Retailers (n 2 above) para 51. 12 Fuel Retailers (n 2 above) para 52. 13 Fuel Retailers (n 2 above) para 53. 14 Fuel Retailers (n 2 above) para 55. 15 Feris (n 1 above) 249. 258 A response to Feris growth variation are, for example, trade related concerns. Access to clean drinking water and food reflect a human well-being variation of sustainable development. They are not the same.16 The approach of the Court in this instance — and reflected in some of Feris’ comments — takes us back in time to the old definition of sustainable development. On this outré account, sustainable development balanced environmental needs and development needs — the latter notion represented both social and economic concerns.

Sustainable development — and this is where a proper historical sketch is important — was born out of a realisation that the existing paradigm (in which economic concerns trumped all other concerns (social and environment) — could not continue. By blurring the distinction between social and economic concerns, our jurisprudence flirts with the undesirable outcome of preserving the status quo: namely, paying lip service to sustainable development and integration. The failure to distinguish more carefully between these values facilitates the instrumentalisation of sustainable development for economic ends. Fuel Retailers is a case in point.

3 Concluding remarks

The treatment of sustainable development by the Constitutional Court was long overdue. For that reason alone Fuel Retailers is a welcome addition to South African jurisprudence. Given the shortcomings of the judgment delineated in Professor Feris’ comment and my reply, we can only hope that the Court will, over time, develop a more nuanced approach to sustainable development that does justice to its history, makes subtle but important distinctions between economic concerns and social concerns, and does not allow this concept to be captured by those parties with purely pecuniary motives.

16 For an in-depth discussion of how these variations and their factors, see Tladi (n 8 above) 81-90. CULTURAL AND RELIGIOUS ACCOMMODATIONS TO SCHOOL UNIFORM REGULATIONS

Patrick Lenta*

1Introduction

Laws and regulations in modern liberal democracies rarely discriminate deliberately against members of religious and cultural groups or target religious and cultural practices. In one relatively recent example, Church of Lukumi Babalu Aye, Inc v City of Hialeah (Church of Lukumi Babalu Aye), the US Supreme Court invalidated municipal ordinances adopted by the city of Hialeah for the specific purpose of proscribing animal sacrifice practised by the Santeria religion.1 Since these ordinances did not constitute a neutral law of general applicability, but deliberately targeted a religious practice, the Supreme Court determined that they were invalid unless they served a compelling state interest. Since the state could not show such an interest, the ordinances were declared, relatively uncontroversially, to be in violation of the Free Exercise clause (the US equivalent of South Africa’s right to freedom of religion and conscience)2 and so invalid.

Laws and regulations that are facially neutral in the sense that they do not deliberately target members of particular cultural or religious groups are more common than those which are overtly discriminatory. In some instances, however, these facially neutral laws may have non-neutral effects and impose disparate burdens on group members. These individuals sometimes claim from the courts

* MA LLB (Natal) PhD (Cantab). Senior Lecturer, School of Philosophy and Ethics, University of KwaZulu-Natal. 1 Church of Lukumi Babalu Aye Inc v City of Hialeah 508 US 520 (1993). 2 Section 15(1) of the Constitution of the Republic of South Africa, 1996 (‘the Constitution’) provides that ‘everyone has the right to freedom of conscience, religion, thought, belief and opinion’.

259 260 Cultural and religious accommodations to school uniform regulations exemptions to generally applicable laws and regulations on the grounds that particular laws or regulations impose burdens on them that are not imposed on others. For example, Muslim female pupils may seek exemptions from school uniform regulations that forbid the wearing of headscarves on the grounds that their religious beliefs obligate them to wear a headscarf or that wearing a headscarf is a genuine expression of their faith.

Should Muslim girls be required to relinquish this cherished religious practice as a condition of availing themselves of the opportunity of education, or are schools obligated to meet their demands for accommodation? The most compelling argument that claims for exemptions in such cases should be granted begins with the claim that cultural group membership (including membership of religious groups) is intrinsically valuable for individual group members. Will Kymlicka argues that cultural membership is

the context within which we choose our ends, and come to see their value, and this is a precondition for self-respect, of the sense that one’s ends are worth pursuing.3

Membership of groups provides individuals with personal and social identities, on which their dignity and prosperity may depend.4

3 W Kymlicka Liberalism, community and culture (1989) 193. 4 For other accounts of the significance of group membership, see J Raz ‘Multiculturalism: a liberal perspective’ in his Ethics in the public domain (1994), B Parekh Rethinking multiculturalism (2000), J Tully Strange multiplicity (1995) and C Taylor Multiculturalism and the politics of recognition (1992). For a recent acknowledgment by the Constitutional Court concerning the value of group membership to members, see Minister of Home Affairs & Another v Fourie & Another 2006 1 SA 524 para 89. The existence of cultural groups has social benefits too. Characterising religious groups in the United States as a source of ‘social capital’, R Putnam Bowling alone: The collapse and revival of the American community (2000) 79, has observed that ‘faith-based organisations serve civic life both directly, by providing social support to their members, and indirectly, by nurturing civic skills, inculcating moral values [and] encouraging altruism’. It is not fanciful to imagine that the moral values to which Putnam refers could potentially provide a counterweight to what B Barry Why social justice matters (2005) 236, calls ‘the recrudescent ideology of individualism, with its concomitant implications that members of society are nothing to one another’, in contemporary capitalist society. Also important are the ‘benefits to society of having religious groups operating as vital associations intermediate between individuals and government and creating a barrier to government domination of social life’ (K Greenawalt Religion and the constitution (2006) 439). The Constitutional Court recognised certain of the benefits of religion to society in Fourie (above) para 90. It must nevertheless be conceded that ‘there is nothing automatically good about religion, which can after all lead to war, insanity, mutilation, suicide, soul-murder’. (JB White ‘Talking about religion in the language of the law: Impossible but necessary’ (1999) 81 Marquette Law Review 177 193). (2008) 1 Constitutional Court Review 261

The argument that Muslim headscarves should be accommodated by school authorities proceeds as follows.5 According to the liberal principle of equality of opportunity, individuals with differing aims should be afforded an equal chance to realise their ambitions and the costs that people should have to bear to do so should be, as far as possible, equal. But since cultural membership is bound up with individual autonomy, dignity, prosperity and self-respect (all of which liberals have a reason to value), in assessing whether people have equal opportunities, their cultural and religious commitments must be taken into account. A prospect only represents an opportunity for an individual if she can avail herself of it without incurring excessive costs. But in the case of a Muslim pupil, the cost of complying with the school uniform regulation may be sufficiently high that the opportunity is effectively removed. Genuine equality of opportunity, on the liberal multiculturalist conception, requires an exemption to permit Muslim female pupils to wear headscarves in school to eliminate the excessive costs attached to compliance with the uniform regulations.

A common response to the contention that granting an exemption in such cases would amount to unfairly preferential treatment of members of the particular group — one that recommended itself to the Constitutional Court of South Africa in MEC for Education, KwaZulu-Natal and Others v Pillay (Pillay),6 the most recent occasion on which it was confronted with a claim for an accommodation7 — is that when a cultural or religious group constitutes a minority, and so is less powerful than majority groups, as Muslims are in South Africa, then its members are disadvantaged compared to members of majority (or more powerful) groups. Laws and regulations are often framed in a way that is consistent with the beliefs and values of the dominant, mainstream cultural groups, but not with those of vulnerable, minority groups. As a result, members of minority groups bear costs in pursuing opportunities with which members of mainstream and more powerful groups are not confronted. Since on this argument membership of a cultural or religious group is like a physical handicap, inasmuch as it is usually an unchosen feature of individuals which is in certain respects disadvantaging, exemptions

5 I have presented this argument in greater detail in P Lenta ‘Muslim headscarves in schools and in the workplace’ (2007) 124 South African Law Journal 296 298-299. 6 2008 1 SA 474 (CC) (Pillay). 7 See (n 6 above) paras 44 (‘The norm embodied by the Code is not neutral, but enforces mainstream and historically privileged forms of adornment, such as ear studs, which also involve the piercing of a body part, at the expense of minority and historically excluded forms’), and para 83 (‘many individual communities still retain historically unequal power relations or historically skewed population groups which may make it more likely that local decisions will infringe on the rights of disfavoured groups’). 262 Cultural and religious accommodations to school uniform regulations should be granted to create equality of opportunity for all persons regardless of power and influence.8

In this paper, I shall comment on Pillay, which concerns the duty of school authorities to accommodate religiously and culturally expressive clothing and accoutrements — in this case, a nose stud, which, it was claimed, expressed the pupil’s Hindu faith and South Indian culture, but which contravened school uniform regulations.9 The issue of whether a pupil should be permitted to wear a tiny nose stud might seem trivial. In fact, however, the politics of dress in schools is highly emotive, both in South Africa and elsewhere, because items of clothing and adornment often serve as visible markers of religious and cultural identity.10

It might be imagined that this issue could easily be disposed of with reference to the question of whether exemptions should be granted to accommodate Muslim headscarves. I have argued elsewhere that exemptions from school uniforms should be granted to permit Muslim pupils to wear headscarves: in some cases, Muslim pupils may sincerely believe that they are under a religious obligation

8 A version of this argument is made in C Eisgruber & L Sager ‘Equal Regard’ in S Feldman (ed) Law and religion (2000) 200. For a similar argument in support of religious exemptions, according to which, since members of a majority religion will usually protect their own religious practices when enacting laws, but ignore, often unintentionally, the burden that otherwise valid laws impose on minority religions, exemptions are justified in order to remedy this flaw in the political process, see D Laycock ‘Formal, substantive, and disaggregated neutrality toward religion’ (1990) 39 DePaul Law Review 993 1014; M McConnell ‘Free exercise revisionism and the Smith decision’ (1990) 57 University of Chicago Law Review 1109 1130-36 and M Nussbaum Liberty of conscience: In defense of America’s tradition of religious equality (2008) 20-21 116-119. In Pillay, Langa CJ drew an analogy between membership of a minority religious or cultural group and disability (n 6 above, 74). 9 For useful discussions of the right to freedom of religion and the ways in which it has been applied in South Africa, see P Farlam ‘Freedom of religion, conscience, thought and belief’ in S Woolman et al (eds) Constitutional law of South Africa (2nd Edition, OS, 2003) ch 41, S Woolman ‘Community rights: Language, culture and religion’ in S Woolman et al (eds) Constitutional law of South Africa (2nd Edition, OS, 2003) ch 58, D Meyerson ‘Religion and the South African constitution’ in P Radan, D Meyerson & R Croucher (eds) Law and religion (2005) ch 5, G J Pienaar ‘The effect of equality and human dignity on the right to religious freedom’ (2003) 66 Tydskrif vir Hedendaagse Romeins-Hollandse Reg 579, I Currie & J de Waal The bill of rights handbook (5th ed 2005) ch 15, N Smith ‘Freedom of religion: The right to manifest our beliefs’ (2002) 119 South African Law Journal 690. Discussions of cultural expression are few, but see, for example, Woolman (above), especially at 31-43, S Woolman ‘Defending discrimination: On the constitutionality of independent schools that promote a particular, if not comprehensive, vision of the good life’ (2007) 22 Stellenbosch Law Review 31, B Fleisch & S Woolman ‘On the constitutionality of single-medium public schools’ 23 South African Journal on Human Rights 34 and I Currie & J de Waal The bill of rights handbook (5th ed 2005) ch 28. 10 As Frantz Fanon observes, ‘immense cultural regions can be grouped together on the basis of original, specific techniques of men’s and women’s dress ... [t]he fact of belonging to a given cultural group is usually revealed by clothing traditions’ (F Fanon, ‘Algeria unveiled’ in Studies in a dying colonialism (1962) 35). (2008) 1 Constitutional Court Review 263 to wear a headscarf and the objectives furthered by having uniforms may not be significantly undermined by granting this exemption.11 But if headscarves should be exempted, then so too, one might think, should Hindu nose studs.

My purpose in this essay is to show that, although South African Hindus of South Indian extraction are, like Muslims, a minority group, the facts of Pillay are not as closely analogous as they might first appear to the case of a Muslim pupil who seeks an accommodation to permit her to wear a headscarf which she sincerely believes to be a matter of religious obligation. Furthermore, the ways in which these cases differ provide grounds for concern about the Constitutional Court’s decision to accommodate nose studs in Pillay. I shall begin by providing, as background, an account of the Constitutional Court’s jurisprudence in the two cases in which it has previously considered claims for accommodations. I shall then focus on the Constitutional Court’s reasoning in Pillay. Throughout, I shall contrast the approach of the Constitutional Court to claims for accommodation with similar claims that have been adjudicated by courts of ultimate jurisdiction in three other jurisdictions: the United States, the United Kingdom and Canada.

2 Corporal punishment and cannabis

In Christian Education South Africa v Minister of Justice (Christian Education),12 the first occasion on which the Constitutional Court was confronted with a claim for an accommodation, it unanimously affirmed that accommodations are, in principle, justifiable. Sachs J aligned the Court with liberal multiculturalist proponents of exemptions by stating that the granting of religious exemptions ‘would not be unfair to anyone else who did not hold those views’13 and that

the State should, wherever reasonably possible, seek to avoid putting believers to extremely painful and intensely burdensome choices of either being true to their faith or else respectful of the law14 provided that religious practices do not violate the rights included in

11 This is the argument of Lenta (n 5 above). 12 2000 4 SA 757 (CC). 13 Christian Education (n 12 above) para 42. 14 Christian Education (n 12 above) para 35. 264 Cultural and religious accommodations to school uniform regulations the Bill of Rights.15

The Constitutional Court nevertheless declined to grant an exemption to legislation proscribing corporal punishment in schools to permit teachers to inflict corporal punishment in Christian schools. The exemption was sought on the grounds that verses in chapters 9, 22 and 23 of the Book of Proverbs command the corporal punishment of children at school. Sachs J argued that the right to dignity, coupled with the right not to be subjected to cruel and unusual, inhuman or degrading punishment, placed a duty on the state to ‘reduce violence in public and private life’,16 particularly in view of the state’s resort to violence to counter protests by children during apartheid and the high incidence of child abuse in contemporary South Africa.

The Constitutional Court’s decision in Christian Education is, I think, correct.17 In a liberal democracy, the state is the ultimate guardian and arbiter of children’s interests; it has a duty to protect

15 The granting of religious accommodations is, of course, far from uncontroversial. J Locke argues forcefully in favour of state neutrality, inasmuch as the state must not prefer particular religions over others, yet he does not endorse the practice of granting exemptions, insisting instead that ‘the private judgement of any person conceiving a law enacted in political matters, for the public good, does not take away the obligation of that law, nor deserve a dispensation’ (J Locke ‘A letter concerning toleration’ in I Shapiro (ed) Two treatises of government and a letter concerning toleration (2003) [1689] 243). Locke’s position is that all people, regardless of their affiliations, must comply with generally applicable laws. If individuals’ religious beliefs require them to disobey the law, they should accept the punishment prescribed by law. Thus, for Locke, if generally applicable laws are neutral in the sense that they do not deliberately target religion, or a particular religion, they should be construed as fair. Nowadays, the principal objection to granting religious exemptions is that to do so is unfair to those who might also wish to be excused from otherwise legitimate laws that burden practices motivated by secular moral beliefs. See, for example, FM Gedicks ‘An unfirm foundation: The regrettable indefensibility of religious exemptions’ (1998) 20 University of Arkansas Little Rock Law Journal 555. Some defenders of the approach taken in J Rawls Political liberalism (1993) may be tempted to argue that the principle of state neutrality, according to which the state should not prefer one religious viewpoint over another, nor religious beliefs over secular beliefs, renders the granting of religious exemptions unjustifiable. On this view, exempting religious believers immunises religion in a way that advances it in comparison to equally heartfelt secular reasons for claiming exemptions. Accommodations for religious beliefs, some Rawlsians may argue, amount to favourable treatment to which religious believers are not entitled. A common response to this objection to mandatory accommodations — one that is implicit in Sachs J’s judgment — is that, if a generally applicable law puts members of a religious or cultural group (especially a minority group) at a disadvantage relative to others, refusing the exemption cannot be justified on the basis that the law provides formal equality: justice requires that exemptions be granted in appropriate cases to eliminate or reduce the burden as far as possible. 16 Christian Education (n 12 above) para 47. 17 For a more detailed commentary, see P Lenta ‘Religious liberty and cultural accommodation’ (2005) 122 South African Law Journal 352 363-371. (2008) 1 Constitutional Court Review 265 children against what it deems to be abusive practices.18 One of the most fundamental liberal rights is to be free from physical assault. Most people think that children have this right.19 Since there is considerable evidence that corporal punishment is psychologically harmful to pupils, then, even though there may be countervailing evidence indicating that mild and infrequent corporal punishment has an improving effect on children, the state is entitled to proscribe it in schools. Once it does so, there is no room for an exemption for Christian schools, since corporal punishment inflicted at these schools will not be rendered non-abusive because it is a practice underpinned by religious beliefs. To grant an exemption would be to frustrate the purpose behind the legislation, which is not to reduce the incidence of corporal punishment in school, but to prevent every instance of an abusive practice.

The Court’s decision in Christian Education is consistent with that of the UK House of Lords in R (Williamson) v Secretary of State for Education and Employment (Williamson),20 the facts of which were virtually identical, except that whereas Sachs J emphasised the duty on the state to reduce the incidence of physical violence because of South Africa’s circumstances, Lord Nicholls of Birkenhead took the view that deference to parliament on an issue of ‘broad social policy’ was appropriate.21 Baroness Hale of Richmond agreed in substance with Lord Nicholls, adding that to ban only corporal punishment that infringed against children’s right not to suffer abusive punishment would present ‘difficult problems of definition, demarcation and enforcement’.22

18 Although there is a presumption in favour of permitting parents to raise their children in accordance with their religious beliefs, the ultimate responsibility of the state for the welfare of children may require it to overrule the religious beliefs of parents. See, for example, Hay v B & Others 2003 3 SA 492 (WLD), in which parents of a minor opposed a doctor’s application to conduct an emergency blood transfusion to save the life of their child principally on the grounds that receipt of a blood transfusion was contrary to their religious beliefs. Jajbhay J held that the parents could not insist that their child should be denied a blood transfusion if the child’s survival, in the opinion of a doctor, depended on it. Jajbhay J is absolutely correct: adults may reject life-saving medical treatment for themselves, but they may not condemn their children to death on the basis of their religious beliefs. A similar position was adopted in the UK in Re O (A Minor) (Medical Treatment) (1993) 2 FLR 149; Re R (A Minor) (Blood Transfusion) (1993) 2 FLR 757. The Canadian Supreme Court arrived at this conclusion in B(R) v Children’s Aid Society of Metropolitan Toronto (1995) 122 DLR (4th) 1 (SCC), in which it upheld a lower court order to remove a baby from the custody of its Jehovah’s Witness parents temporarily so that it could receive life-saving medical treatment, including a blood transfusion. In the US, judges have readily ordered treatment over the parents’ wishes. See Matter of Hamilton, 657 SW 425, 429 (Tenn Ct App 1983); Custody of a Minor, 393 NE2d 836 (Mass 1979). 19 See C McKinnon Toleration: A critical introduction (2006) 107. 20 2005 2 AC 246 (HL). 21 n 20 above, para 51. 22 n 20 above, para 86. 266 Cultural and religious accommodations to school uniform regulations

In Prince v President of the Law Society of the Cape of Good Hope (Prince),23 a narrow majority of the Constitutional Court refused to grant to Rastafarians an exemption from legislation forbidding the possession and use of cannabis, despite an existing permit system for administering medical exemptions. The majority denied the claim principally on the grounds of the practical difficulty of policing the exemption, and because South Africa’s obligations under international law required a ban to be uniformly enforced.24

The applicant, a Rastafarian attorney, applied for an exemption to statutes which proscribe the possession and use of cannabis, on the grounds that these Acts prevent an activity central to the sacramental activities of his religion. The majority of the Court decided that although medical exemptions from the Act could be effectively controlled and administered, the same would not be true of a religious exemption for Rastafarians. Since Rastafarian use of cannabis was not restricted to small quantities at religious ceremonies, but was consumed at home and on social occasions, there would be no way for enforcement officials to distinguish between authentic religious consumption and consumption for recreational reasons. Moreover, the majority decided, South Africa’s blanket ban on the use and possession of cannabis was intended to meet its obligations under international law. There would, in addition, be financial and administrative difficulties in establishing a permit system and this coupled with the fact that the Rastafarian religious community is not easily delineated, since its structures are informal, would render a permit system incapable of preventing cannabis being distributed to non-Rastafarians. The majority, despite aligning itself with the approach of the minority of the US Supreme Court in Employment Division v Smith (Smith),25 according to whom religious practices could only be subordinated to a general governmental interest if the state could show a ‘compelling state interest’ and that the law was the ‘least restrictive means’ of serving that interest,26 held that a uniform ban on a drug such a cannabis, which is widely used for recreational purposes and in which there is significant trade, was justified.

The dissenting judgements of Ngcobo and Sachs JJ took the view that a partial rather than full exemption (as sought by the applicant) should be granted to enable Rastafarians to use small quantities of cannabis for sacramental purposes on religious occasions. A limited exemption, Sachs J held, would secure for the Rastafari ‘a modest but

23 2002 2 SA 794 (CC). 24 For a more detailed analysis, see Lenta (n 17 above) 371-375. 25 Employment Division, Department of Human Resources of Oregon v Smith 494 US 872 (1990). 26 Prince (n 23 above) para 122. (2008) 1 Constitutional Court Review 267 meaningful measure of dignity and recognition’ and was required because ‘the Constitution requires the state to walk the extra mile’.27 Against the majority, Ngcobo and Sachs JJ contended that a permit system, coupled with administrative guidelines, could be instituted to ensure that the terms of a limited exemption would be effectively enforced. Ngcobo J argued that an exemption would not violate South Africa’s obligations under international law since provisions to the effect that such measures were subject to each party’s constitutional principles and limitations were included in international protocols and conventions binding on South Africa. If the South African Constitution is interpreted by the Court as requiring an exemption, it would not fall foul of international conventions. Sachs J similarly claimed that South Africa’s international obligations do not rule out an exemption for religious purposes.28

The majority’s failure to take seriously a limited exemption for an act of worship central to faith that would use administrative procedures similar to medical exemptions is, I think, regrettable. The burden imposed by the relevant legislation on Rastafarians is severe: sacramental and liturgical practices are at the core of religion and forbidding these practices forces Rastafarians either to violate the tenets of their religion or become outlaws. This appears discriminatory since it is unthinkable that major faiths should have to face this choice. The use of wine in the Catholic Mass is unlikely ever to be forbidden.

On the issue of enforcement, screening devices might have been employed to separate genuine Rastafarians wishing to use cannabis for purposes of worship from others seeking to smoke cannabis for recreational purposes. The terms of the exemption proposed by the minority were consistent with dispensing cannabis in small quantities directly to Rastafarian priests for distribution restricted to individuals participating in church services, the whole to be overseen by state officials.29 People masquerading as Rastafarian worshippers would have to go to great lengths to pass themselves off as believers before other believers and state officials. They would be required, at considerable inconvenience to themselves — which would in most cases, one imagines, outweigh the pleasure they would otherwise derive from smoking — to participate in the ceremony. It is implausible to think that a limited exemption such as this would promote the illegal trafficking of drugs any more than would a medical exemption. I concede, however, that justifying the limited

27 Prince (n 23 above) paras 148-9. 28 Prince (n 23 above) para 164. 29 The granting of this limited exemption might have had an economic cost. It might have been appropriate to ask Rastafarian worshippers to bear at least part of this cost. 268 Cultural and religious accommodations to school uniform regulations exemption might depend on the plausibility of the majority’s interpretation of South Africa’s obligations under international conventions.

By aligning itself with the ‘compelling interest test’ established by the US Supreme Court in Sherbert v Verner30 and Wisconsin v Yoder (Yoder),31 the Constitutional Court affirmed its willingness to grant exemptions in principle.32 In doing so, it offered greater protection for religious liberty than US Supreme Court did in Smith, in which the majority refused to grant an exemption from a criminal prohibition to permit Native American Indians to ingest as an element of their worship, peyote, a psychotropic drug. The majority in Smith, abandoning the ‘compelling interest’ standard (except for unemployment compensation cases and ‘hybrid’ cases involving two constitutional rights), held that if a law is generally valid, the government may apply it against religious claimants without having to show more than that the law has a rational basis. Provided the state can show this, there is no constitutional requirement to grant religious groups exemptions from facially neutral laws which do not have the purpose (though they may have the incidental effect) of burdening the practice of religion. As Scalia J put it, the right to freedom of religion ‘does not relieve an individual of the obligation to comply with a “valid and neutral law of general applicability”’.33

In response to Smith, the US Congress, at the instigation of a large number of religious groups, enacted the Religious Freedom Restoration Act (‘RFRA’). The RFRA was intended to restore the position prior to Smith relating to the granting of religious exemptions, according to which, as the Act states

Government may not substantially burden a person’s exercise of religion even if the burden results from a rule of general applicability … [unless] it demonstrates that the application of the burden to the person (1) is in furtherance of a compelling government interest; and (2) is the least restrictive means of furthering that ... interest.34

In City of Boerne v Flores, Archbishop of San Antonio et al,35 the US Supreme Court declared the RFRA invalid as it applies to states and

30 374 US 398 (1963). 31 406 US 205 (1972). 32 Prince (n 23 above) para 128. 33 Smith (n 25 above) p 879. 34 Religious Freedom Restoration Act, 42 USC §§ 2000bb-1-2000bb-4 (2004). 35 521 US 507 (1997). (2008) 1 Constitutional Court Review 269 localities, but not as it applies to the federal government, for which purposes it remains valid.36

In a recent US Supreme Court decision, Gonzalez, Attorney General et al v O Centro Espírita Beneficente União Do Vegetal et al (Gonzalez),37 members of a church, the Christian Spiritists, wished to import hoasca, a tea containing a federally proscribed hallucinogen, which, it claimed, facilitated communion. The Supreme Court dismissed the Government’s submission that it had a compelling interest in the uniform application of the law to the extent that no exemption could be granted to accommodate the sect’s sacramental use of hoasca. The Court held that the government’s failure to grant an exemption to the claimants was inconsistent with their rights under the RFRA.

Although the Supreme Court in Gonzalez adopted an approach similar to that employed in Prince, there are two important respects in which the approaches of the two courts differ.38 First, the majority in Prince gave greater weight to a need for uniform enforcement if the ban on the use and possession was to be effectively policed. Second, the US Supreme Court ruled that the fact that hoasca was covered by the Convention on Psychotropic Substances, did ‘not automatically mean’ that the federal law furthered a compelling interest that outweighed the claim for an exemption under the RFRA. Despite acknowledging the importance of ‘honouring international obligations and of maintaining the leadership position of the United States in the war on drugs’, the invocation of these general interests alone, the Supreme Court decided, was ‘not enough’ in the absence of evidence addressing the international consequences of granting an exemption.39 By contrast, the majority of the Constitutional Court attached greater significance to South Africa’s obligations under international law, holding that ‘[t]he use made of cannabis by Rastafari cannot in the circumstances be sanctioned without

36 The American Indian Religious Freedom Act Amendments were enacted by the US Congress in 1994 to permit peyote use by Native Americans for the purposes of religious worship and many states enacted legislation to exempt Native American Indians from existing prohibitions against drug use. This legislation does not protect the sacramental use of cannabis by Rastafarians, however, raising the issue of whether the 1994 Amendments constitute unfair discrimination. This issue arose in the Kansas Court of Appeals in State v McBride 955 P2d 133 (Kan Ct App 1998), in which the court distinguished between the religious use of the two substances on the following grounds: peyote is used in small quantities and is restricted to religious ceremonies by Native American Indians; recreational use and abuse of peyote is rarer than is the case with cannabis, and the state in the US has a special duty to respect the political and cultural integrity of Native Americans. See B Taylor ‘Kansas denies religion-based defense to Rastafarians on marijuana charges’ (1998) 38 Washburn Law Journal 307. 37 No 04-1084 (2006). 38 See P Edge ‘Religious drug use in England, South Africa and the United States of America’ (2006) 1 Religion and Human Rights 165 176. 39 Gonzalez (n 37 above) 17-18. 270 Cultural and religious accommodations to school uniform regulations impairing the state’s ability ... to honour its international obligation to do so’.40

Do these differences in approach indicate that the US Supreme Court in Gonzalez provided greater protection to freedom of religion than the Constitutional Court did in Prince? Before concluding that it did, it is important to recognise that the facts of the two cases are different. Gonzalez deals with a drug that is not widely used: the market for hoasca is small and there had not in the past been practical difficulties in policing the ban on recreational use of hoasca. By contrast, as the majority in Prince noted, ‘cannabis, unlike peyote, is a drug in which there is substantial illicit trade’.41 Unlike hoasca, cannabis is widely used beyond the confines of religious worship for recreational purposes. For these reasons, an approach such as that employed in Gonzalez may well have denied an exemption for religious use of cannabis, just as Blackmun J in his dissenting judgement in Smith held that although an exemption should be granted in the case of peyote, the state could legitimately deny claims for religious exemptions involving drugs such as marijuana and heroin, the use of which is not restricted to a ceremonial context and in which there is ‘significant illegal trade’.42 Would the majority’s approach in Prince commit it to denying an exemption in a case in material respects the same as Gonzalez? Few of the difficulties of policing and administering an exemption for cannabis are likely to occur in the case of an exemption for hoasca. Nevertheless, if the Constitutional Court determined that South Africa’s obligations under international law required a ban to be uniformly enforced, the view of the majority in Prince suggests that the Court might reject a claim for an exemption for hoasca too.

The refusal of the majority of the Constitutional Court in Prince to grant an exemption to enable Rastafarians to use cannabis on the grounds of international law obligations is consistent with the approach of the UK judiciary. In Taylor,43 later endorsed in Andrews,44 a Rastafarian had been observed by police officers approaching a Rastafarian temple. On being searched, he was found to be in possession of cannabis, in violation of a criminal prohibition. The Rastafarian’s defence consisted in the contention that the cannabis was intended to be consumed as part of an act of worship at the temple. Accordingly his possession was a manifestation of his religion, protected under the right to freedom of religion (Article 9 of the European Convention on Human Rights). The Court of Appeal held,

40 Prince (n 23 above) para 139. 41 n 23 above, para 129. 42 Gonzalez (n 37 above) 917-8. 43 [2001] EWCA Crim 2263; [2002] 1 Cr App R 37. 44 [2004] EWCA Crim 947. (2008) 1 Constitutional Court Review 271 on the basis of several international conventions whose purpose was to restrict the use, possession and trafficking of psychotropic narcotics, that whatever the interference with the Rastafarian’s freedom of religion, it was justified in order to prevent public health and safety dangers arising from cannabis use, possession and supply.

3 Nose studs, headscarves and daggers

Pillay concerned a claim by the parent of a Hindu pupil at a state school that the failure of the school’s authorities to grant her daughter an exemption from the school’s Code of Conduct to permit her to wear a nose stud constituted unfair discrimination on the grounds of religion and culture. The claim was brought under the Promotion of Equality and Unfair Discrimination Act,45 which prohibits unfair discrimination on the grounds of religion and culture. Although the claim was not brought under the rights to freedom of religion and culture, the Court held that there may be, as in this case, an

overlap [between the Promotion of Equality and Unfair Discrimination Act and the rights to freedom of religion, belief and opinion and to participate in cultural life46] … where the discrimination in question flows from an interference with a person’s religious or cultural practices’.47

Pillay clearly represents a development of the Court’s jurisprudence on religious liberty and cultural accommodation.

The Court ruled that the school’s failure to grant the exemption was unfairly discriminatory and that the school was required to exempt the pupil concerned, Langa CJ, on behalf of the majority, deciding that the wearing of a nose stud represents an expression of the Hindu religion and of South Indian Tamil culture. The majority found that the failure of the school to grant the accommodation was unfairly discriminatory against the pupil concerned relative to other pupils whose religious beliefs and cultural practices did not bring them into conflict with the Code. Langa CJ stated that although the school uniform served ‘admirable purposes,’ these objectives would not be undermined by granting religious and cultural exemptions.48

In a separate judgement in which she agreed with the majority’s finding of unfair discrimination, but disagreed in part with its

45 Act 4 of 2000 (‘the Equality Act’). 46 Section 30 of the Constitution reads: ‘Everyone has the right to use the language and participate in the cultural life of their choice, but no one exercising these rights may do so in a manner inconsistent with any provision of the Bill of Rights.’ 47 Pillay (n 6 above) para 46. 48 Pillay (n 6 above) para 101. 272 Cultural and religious accommodations to school uniform regulations reasoning, O’Regan J found that there is no evidence that the wearing of a nose stud has religious significance for the claimant, although she accepted that it is a form of cultural expression. She agreed with the majority that the failure of the school to grant an exemption was unfairly discriminatory, but for a different reason: the applicant was discriminated against, not relative to those pupils on whom the Code imposed no burden of religious or cultural denial, but relative to other pupils whom the school had in the past exempted: Hindu pupils had been permitted to wear ‘Lakshmi strings’ in honour of the Goddess Lakshmi and other pupils had been allowed to wear hide bracelets as a mark of respect on the death of a relative.49

Both Langa CJ and O’Regan J recognised that exemptions should not be made in response to every claim. Claims for exemptions call for a proportionality exercise. The strength of the claim must be determined. This involves an enquiry into whether, if a religious belief is relied upon, it is genuinely a religious belief, or, in the case of a cultural practice, that it is genuinely a cultural practice; the sincerity of the claimant, and the nature and severity of the burden on religious exercise. The strength of the competing interest of the state or other relevant institution in the uniform application of the rule must equally be considered. In cases where the law in question protects fundamental rights of individuals, or serves a compelling state objective that would be significantly frustrated by granting an exemption, or where the exemption would present insurmountable practical difficulties such that it could not be administered effectively, it may be that an exemption is ruled out. In the rest of this section, I consider the way in which the Constitutional Court negotiated each step in this enquiry.

3.1 Religious belief or cultural practice?

The claimant was required to show that the practice for which the accommodation is claimed is cultural in nature or that it is motivated by religious beliefs. The task of determining what beliefs and practices are religious — as opposed to secular, moral or personal — is a difficult matter, since religion is a highly complex concept. By what criteria can religion be defined and the distinction between it and the non-religious drawn, given the ineffable nature of religious experience and the fact that individuals have shifting relationships with religious communities which may themselves be riven with profound theological differences? Defining culture, to the extent that

49 Pillay (n 6 above) para 130. (2008) 1 Constitutional Court Review 273 it differs from religion, is no easier, since ‘cultural forms possess inherent indeterminacy’.50

On the issue of the cultural significance of the nose stud, Langa CJ asserted two claims. First, ‘cultural convictions or practices may be as strongly held and as important to those who hold them as religious beliefs are [for religious believers]’ since individual identity and dignity is bound up with culture. Second, cultural practices ‘will differ from person to person within a culture’, since ‘[c]ultures are living and contested formations’.51 Noting the regrettable absence of testimony from the pupil herself, Langa CJ found that an admission by the school authorities that the nose stud has ‘cultural significance’52 for the pupil was dispositive of the question of whether the practice should be considered to be a cultural practice.

On the issue of whether the nose stud has religious significance, Langa CJ referred to the testimony of Dr Rambilass, an expert on the Hindu religion, who had given evidence on behalf of the school. Dr Rambilass had conceded that the wearing of a nose stud is a cultural practice, but denied that the nose stud has independent religious significance, being ‘part of the Shringaar which is concerned with love, beauty and adornment’.53 Langa CJ contested this expert witness’s conclusion that the wearing of a nose stud is devoid of religious significance on three grounds. First, he argued that it is difficult to separate Hindu culture and Hindu religion. Secondly, ‘there are many different sects of Hinduism with different beliefs and practices’.54 Thirdly, Dr Rambillas’s testimony

self-consciously focused on defining Hindu religion according to the specific wording of the Vedic texts rather than on a broader view of religion as being informed and even defined by culture, tradition and practice.55

On the grounds that ‘the borders between culture and religion are malleable and that religious belief informs cultural practice and cultural practice attains religious significance’, Langa CJ concluded that Sunali’s (the pupil in question’s) wearing of a nose stud was ‘an expression of both culture and religion’.56

Langa CJ’s reasoning in determining that the wearing of a nose stud has religious significance may be questioned from several

50 L Rosen, quoted in W Fallers Sullivan ‘Judging religion' (1998) 81 Marquette Law Review 441 456. 51 Pillay (n 6 above) paras 53-54. 52 Pillay (n 6 above) para 58. 53 Pillay (n 6 above) para 59. 54 As above. 55 As above. 56 Pillay (n 6 above) para 60. 274 Cultural and religious accommodations to school uniform regulations perspectives. There was no evidence to suggest that it bore religious significance for Sunali beyond the claimant’s assertion that it did. Expert testimony indicated otherwise, and Langa CJ’s rejection of Dr Rambilass’ testimony concerning the content of the Hindu religion, about which he and not Langa CJ is an expert, is illegitimate. As for Langa CJ’s contention that it is often difficult to distinguish between culture and religion, it is one thing to say that many religious practices are in some sense cultural and that many cultural practices are religious — both of which propositions are true, since culture and religion are inter-imbricated. It is quite another thing to aver that all cultural practices are religious, which is false. There is a subset of cultural practices that do not carry religious significance, which the evidence of Dr Rambilass indicated included the wearing of a nose stud.

That the borders of religion and culture are porous and that cultural practices may attain religious significance does not, contra Langa CJ, constitute evidence that the nose stud has religious significance. It would do so only if all cultural practices had religious significance, which is surely not the case. Langa CJ failed to cite credible evidence for thinking that nose studs have religious significance. Accordingly, O’Regan J’s assessment is more plausible:

Although the applicant argues that the nose stud was part of religious practice, it is clear that its primary significance to her family arises from its associative meaning as part of their cultural identity, rather than personal religious beliefs. This is consistent with Dr Rambilass’s evidence that wearing a nose-stud is not part of Hindu religion’.57

O’Regan J’s approach to determining whether a practice is cultural is to be preferred to Langa CJ’s. She contended that since cultural practices are associative and that the right to cultural life is to be practiced as a member of a community and not primarily a question of sincere, but personal belief, Langa CJ’s ‘individualised and subjective approach to what constitutes culture is faulty’. Rather ‘in probing whether a particular practice is a cultural practice, some understanding of what the cultural community considers to be a cultural practice is important’.58 O’Regan J elaborated:

57 Pillay (n 6 above) para 62. O’Regan J’s conclusion — ‘that the applicant has established that the wearing of a nose stud is a matter of associative cultural significance, which was a matter of personal choice at least for the learner in this case, but that it is not part of a religious or personal belief of the applicant that it is necessary to wear the stud as part of her religious beliefs’ — is open to the interpretation that she accepted that the practice has non-mandatory religious significance for Sunali. But this would be inconsistent with O’Regan J’s failure to challenge, and apparent acceptance of, Dr Rambilass’s statement that wearing a nose stud is not a religious practice. 58 Pillay (n 6 above) para 154. (2008) 1 Constitutional Court Review 275

It will not ordinarily be sufficient for a person who needs to establish that he or she has been discriminated against on the grounds of culture to establish that it is his or her sincerely held belief that it is a cultural practice, or that his or her family has a tradition of pursuing this practice. The person will need to show that the practice that has been affected relates to a practice that is shared in a broader community of which he or she is a member and from which he or she draws meaning.59

O’Regan J’s distinction between religion and culture has merit. Although most religions are also cultures, a religion may be constituted by the beliefs and practices of a single individual. A religion need not be constituted by shared practices and beliefs (even though it will usually be). By contrast, culture signifies the range of beliefs and assumptions held in common by members of a group, as well as the ensemble of customs and practices (fluid and mobile to be sure) that reflect the way members do things in common: modes of dress and decoration, types of circulation and social activity, ways of working and so on. Thus, where it is claimed that a practice is a cultural practice, we can examine the culture concerned to determine whether members engage in it in common with other members. We are not necessarily in a position to do this when it is claimed that a particular practice is religious in nature, since the beliefs and practices of a single individual may be sufficient to constitute a religion.

O’Regan J noted that there are a number of women of South Indian heritage who wear a nose stud because it identifies them as members of that community and connects them with their community and with former generations. She rightly considered this sufficient evidence to conclude that the wearing of a nose stud by women of South Indian extraction is a matter of ‘associative cultural significance’.60 She nevertheless noted, again correctly, that the practice is optional (that there is no cultural norm or obligation putting pressure on Sunali to wear a nose stud is evidenced by the fact that her sisters opted not to do so).

No evidence was adduced by the applicant to show that the wearing of the nose stud by her daughter was religious; indeed, it is clear that the wearing of a nose stud is, for Hindu women, an expression of cultural identity that is entirely optional and is not a religious practice. Nevertheless, that the practice has cultural significance is sufficient to bring it within the ambit of the Equality Act.

59 Pillay (n 6 above) para 159. 60 Pillay (n 6 above) para 162. 276 Cultural and religious accommodations to school uniform regulations

3.2 Sincerity

In order to be successful, the claimant must be sincere in her assertion that her opposition to conforming to the standard requirements rests on religious or cultural grounds: she must demonstrate that her beliefs are sincerely held (‘neither fictitious, nor capricious ... [nor] an artifice’)61 or that she is sincerely expressing her cultural identity. Although the enquiry into sincerity precludes determining the truth or legitimacy of religious claims, it is important because charlatans may be tempted, for tactical purposes, to misrepresent as religiously or culturally motivated their insistence on acting contrary to the rules. The sincerity enquiry

attempts to ensure that exemptions are not granted to individuals who frame their opposition ... in religious [or cultural] terms in order to be afforded the legal remedy desired when in fact the religious beliefs are not sincerely held because the opposition stems from secular convictions.62

A claimant is not acting on sincerely held religious beliefs or expressing her cultural identity simply because she says she is. The claimant must demonstrate to the satisfaction of the court that the asserted beliefs are sincerely held or that the practice in question is an expression of cultural identity. For Langa CJ, the fact that Sunali refused to remove the nose stud after initially agreeing to do so — equivocation that is most plausibly explained, he stated, ‘as a young woman uncertain about the consequences of standing up against the imposing authority of the School’s headmistress’ — ‘points to the conclusion that Sunali held a sincere belief that the nose stud was part of her religion and culture’.63 In fact, however, Sunali’s conduct was also consistent with the actions of a pupil who was initially prepared to forgo wearing the nose stud, but whose parent resolved to defy the school’s authorities in order that her daughter could continue to engage in a practice expressive of cultural identity only. The sincerity of the claimant’s averment that the practice has religious significance is questionable, particularly in view of Dr Rambilass’s evidence that Hindu women do not regard it as a religiously significant practice. Even allowing that different believers within the same faith may hold different beliefs and attach different significance to the same

61 Iacobucci J in Syndicat Northcrest v Amselem (2004) 2 SCR 551 para 52 (Amselem). 62 C Kraus ‘Religious exemptions — applicability to vegetarian beliefs’ (2001) 30 Hofstra Law Review 197 215. As the US District Court observed in US v Kuch 288 F Supp 439 443 claimants ‘must not be permitted the special freedoms this sanctuary may provide merely by adopting religious nomenclature and cynically using it as a shield to protect them when participating in antisocial conduct that otherwise stands condemned’. 63 Pillay (n 6 above) para 58. (2008) 1 Constitutional Court Review 277 practice, the applicant provided no convincing evidence that Sunali’s wearing of a nose stud was motivated by sincerely held religious belief. In the absence of any evidence that the wearing of a nose stud is motivated by a religious belief (such as the scriptural evidence provided by the claimants in Christian Education or the verses of the Quran referred to by Muslim women claiming exemptions to permit them to wear headscarves) the impression that remains is that Sunali’s wearing of a nose stud only masqueraded as a religious practice for the purposes of securing an exemption.

This is not to insist on the semiotic neutrality of a nose stud for a pupil of South Indian descent, nor is it to dispute the sincerity of the claimant’s assertion that Sunali wished to wear the nose stud to express her South Indian culture. Far from it: as both Langa CJ and O’Regan J rightly observed, it is a practice that is clearly invested with cultural significance: an expression of the pupil’s cultural identity. Although the nose stud is an adornment worn by women out of personal choice within South Indian culture, there is no reason to doubt that Sunali wore it as a fashion accessory that expressed her cultural identity.

3.3 Claimant’s burden versus state interest

Langa CJ observed that the constitution places a duty of ‘reasonable accommodation’ on the institution from which an exemption is claimed. How should reasonableness be construed in this context? According to Langa CJ, the institution must make more than mere negligible effort ‘to enable those outside the “mainstream” to swim freely in its waters’, but determining the extent of the hardship it must suffer to do so will involve a ‘contextual’ determination, ‘an exercise in proportionality that will depend intimately on the facts’.64 In practice, courts must assess the burden imposed on the claimant’s religious freedom or cultural practice and whatever interest is served by refusing the exemption, and strike an appropriate balance between the two. The claimant should only succeed if she suffers a substantial burden on the exercise of her religion or culture and if the state or institution’s interest in denying her an exemption is not strong. Even if the burden on the claimant is substantial, the claim should not succeed if the government’s interest in denying the exemption is strong: human sacrifice may be central to a claimant’s religion, but it should not be accepted as a defence against a charge

64 Pillay (n 6 above) para 76. 278 Cultural and religious accommodations to school uniform regulations of murder. Likewise the claim should fail if it affects religious or cultural activities only marginally.65

The claimant was required to show that, as a result of the school’s code, Sunali was placed under a substantial burden. A showing of this kind is easier where the religious or cultural practice for which the claimant is seeking protection is mandatory rather than optional: where there is ‘a Hobson’s choice between observance of their faith and adherence to the law’66 as there was in Christian Education. In Pillay, by contrast, the applicant conceded that her wearing a nose stud was not required by the tenets of her faith, but was instead a voluntary practice. The Constitutional Court held — consistently with the Canadian Supreme Court’s ruling in Amselem67 (affirmed, though the Constitutional Court does not say so, by the House of Lords in Williamson68) — that it is not essential to the success of a claim for a religious exemption that the religious practice be required by the claimant’s religious or cultural beliefs, but only that it be an expression of those beliefs. That is so, stated Langa CJ, because the Constitution is committed to ‘affirming diversity’ and ‘differentiating between mandatory and voluntary practices does not celebrate or affirm diversity but merely tolerates it’.69

The Constitutional Court is clearly correct that in order to be afforded protection under the right to freedom of religion the belief or practice need not be mandatory, although this has nothing to do with celebrating rather than tolerating diversity. To be successful, the claimant must demonstrate not that the practice for which she seeks protection is compelled by her religion, but that the regulation imposes a substantial burden on the practice of her religion. A trivial impact imposed by a law on an individual would not be sufficient to qualify for an exemption. Even if religious practices are not mandatory, being prevented from engaging in such practices may be substantially burdensome. Kent Greenawalt illustrates this point with an example: individual A might not consider wearing a cross to be mandatory, yet could regard wearing a cross as ‘an important symbol of witness and commitment and a great aid to devotion’.70 Were a school to forbid all jewellery, including crosses, A’s convictions should

65 See Lyng v Northwest Indian Cemetery Protection Association 374 US 398 (1963) at 475, in which Brennan J, dissenting, argued that a trivial interference with religious practice was insufficient; the claimants were required to show the ‘centrality’ of the affected practices and the presence of a ‘substantial and realistic threat of frustrating their religious practice’. 66 Pillay (n 6 above) para 62. 67 n 61 above, paras 67-8. 68 n 20 above, para 33. 69 Pillay (n 6 above) para 65. 70 Greenawalt (n 4 above) 210. (2008) 1 Constitutional Court Review 279 be sufficient to vault her over the threshold of substantial burden.71 Of course, not all voluntary practices are similarly justifiable. Consider individual B, who wears a cross without attaching great religious significance to it, though aware that it represents for Christians the crucifixion of Christ. A rule forbidding B from wearing jewellery would not constitute a substantial burden.

Even allowing that proscribing voluntary practices may be burdensome, interfering with obligatory practices is likely to be more burdensome. That is so, because forgoing a practice mandated by the tenets of religion (as wearing a headscarf is for some Muslim women) will be intensely injurious to them, because what is being demanded of them is that they violate transcendental religious commands, on which the favour of their deity or sanctions, salvation or damnation, may depend. A clash between religious commandments and civil laws or regulations situates believers between competing sources of authority, a predicament particularly grave given ‘the widespread sense that one’s religious obligations are more ultimate than those of the social order and should take priority if the two come into conflict’.72 Voluntary practices, precisely because they are not commanded, are less damagingly relinquished, even if forbidding such obstacles may be substantially burdensome.

Langa CJ appeared not to appreciate this when he asserted, ‘that we choose voluntarily rather than through a feeling of obligation only enhances the significance of a practice to our autonomy, our identity and our dignity,’ to which he added that ‘it may be even more vital to protect non-obligatory cultural practices’.73 But this obscures the fact that the costs to a believer of forgoing an optional practice will almost always be lower than the costs that a believer pays in deciding not to engage in a practice that the tenets of his faith mandates, because in the case of an optional practice, the individual is not required to make the agonising choice between competing sources of authority.

71 See, by way of illustration, Sasnett v Sullivan 908 F Supp 1429 (WD Wis 1995), aff’d, 91 F 3d 1018 (7th Cir 1996), vacated and remanded, 521 US 1114 (1997), in which a prisoner challenged under the RFRA regulations limiting the wearing of all religious jewellery except wedding bands. Prison officials argued that the exercise of a prisoner’s religion was substantially burdened only if the practice was required by their religion, which the wearing of crucifixes was not. Posner J held that the question was whether the burdened practice was motivated by a sincere religious belief, not whether it was ecclesiastically mandated. He found that the regulation forbidding the wearing of crosses substantially burdened the plaintiff’s religious liberty. Winnifred Fallers Sullivan quotes one of the plaintiffs: ‘There’s nothing saying you have to wear a cross, but it brings a person closer to God’ (Sullivan (n 50 above) 451). 72 Greenawalt (n 4 above) 439. 73 Pillay (n 6 above) paras 64 & 66. 280 Cultural and religious accommodations to school uniform regulations

How should the magnitude of the burden imposed on the claimant be measured? Langa CJ decided that the failure of the school to grant an exemption constituted a ‘significant infringement’ of the claimant’s religious and cultural identity on the basis of what he referred to as a ‘subjective investigation’.74 The issue, he stated, is whether the pupil considered the nose stud central to her religion or cultural identity, not whether other members of the South Indian Tamil Hindu community do so (since the status of a practice within a particular group may be disputed). Langa CJ acknowledged that ‘the Court can properly look at a range of evidence including evidence of the objective centrality of the practice to the community at large’.75 Yet it is striking that he, unlike O’Regan J, did not consider the way in which the practice is viewed by other members of the South Indian Hindu community. Instead he considered the pupil’s beliefs and attitudes concerning the practice, which he inferred from her conduct. Where Langa CJ’s reasoning went awry was in assuming that behaviour is an accurate indicator of beliefs. In fact, conduct is not reliable evidence of beliefs: it is usually the case that multiple beliefs and motivations are consistent with any individual’s activity.

Langa CJ appears to have accepted uncritically the applicant’s assertion that she was placed under a substantial burden and refers to the pupil’s insistence on wearing the nose stud under the pressure of threatened disciplinary action as evidence of this.76 But in fact, all that the pupil’s conduct evidences is a strong commitment to wearing the nose stud, which could reflect a substantial burden but could equally be motivated by her obedience to her mother, who appears to have engaged in a power struggle with the school authorities. The difficulty with Langa CJ’s enquiry into burden is that he seemed to treat the conclusory allegation on the part of the claimant regarding burden as dispositive. He was too ready to interpret the pupil’s conduct as indicative of the strength of her religious and cultural commitment to the practice, without allowing that a strong commitment to persisting with the practice is not necessarily generated by a substantial burden.

The school claimed that the burden, if any, was slight since the pupil could wear the nose stud out of school hours. Langa J rejected this claim on the grounds that ‘[t]he practice to which Sunali adheres is that once she inserts the nose stud, she must never remove it’.77 However, he referred to no evidence in support of the claim that the practice of wearing a nose stud adopted by certain members of the

74 Pillay (n 6 above) paras 85 & 88. 75 Pillay (n 6 above) para 88. 76 Pillay (n 6 above) para 90. 77 Pillay (n 6 above) para 85. (2008) 1 Constitutional Court Review 281

South Indian community obliges them to wear the nose stud at all times.

The school claimed that the burden placed on the pupil was slight for the additional reason that she could have transferred to another school that would have allowed her to wear a nose stud. Langa CJ rejected this argument on the grounds that ‘permitting it only when no other option remains’78 was inconsistent with South Africa’s ‘constitutional project which not only affirms diversity, but promotes and celebrates it’. The effect of requiring the pupil to relocate to another school to express her cultural identity, Langa CJ asserted, ‘would be to marginalise religions and cultures’.79

If in Pillay the burden on the pupil did not, contra Langa CJ, appear to be particularly substantial, neither was the school’s interest in enforcing the school uniform without exception strong. It is often argued that uniforms promote crucial educational interests: minimising external differences between pupils of different social classes; discouraging competitive fashions; promoting school spirit and encouraging discipline.80 At issue was the empirical question of whether the accommodation sought by the claimant would significantly impair the achievement of these educational objectives. The school did not offer evidence on this point and it is difficult to see how it could have done so. In fact, as Langa CJ noted, it is difficult to see how granting an exemption to permit the wearing of a nose stud would interfere substantially with the effective running of a school and with the purposes the school uniform was designed to further.81

This conclusion is consistent with that of the UK House of Lords in Mandla v Dowell Lee,82 decided under the Race Relations Act of 1976, in which a headmaster had refused to admit a Sikh pupil to a private school on the grounds that the pupil insisted on wearing a turban, mandated by his faith, in violation of the school uniform. Lord Fraser

78 Pillay (n 6 above) para 65. 79 Pillay (n 6 above) para 92. 80 There is some debate about whether school uniforms are successful in fostering these educational objectives, although (contested) studies conducted by American researchers have found school uniforms increase discipline, foster school morale, reduce unhealthy competition amongst pupils and so forth. For a critical review of the literature, see D Brunsma School uniforms: A critical review of the literature (2002). The United States Department of Education’s Manual on school uniforms, (http://www.ed.gov/updates/uniforms.html (accessed: 6 June 2007)) begins with the sentence, ‘A safe and disciplined learning environment is the first requirement of a good school’ and under the heading ‘Treat school uniforms as part of an overall safety program’ the Manual states as follows: ‘Uniforms by themselves cannot solve all of the problems of school discipline, but they can be one positive contributing factor to discipline and safety’. 81 Pillay (n 6 above) para 101. 82 [1983] 2 AC 548. 282 Cultural and religious accommodations to school uniform regulations likewise dismissed the headmaster’s argument that an exemption would frustrate the objectives furthered by the uniform.

In Pillay, the available evidence suggested that the burden imposed on the pupil and the interest of the school in denying her an exemption were roughly in equipoise. The political philosophy underlying the Court’s holding that the severity of the burden placed on the pupil outweighed the burden placed on the school is revealed in its commitment not only to tolerating difference but to ‘affirm[ing] it as one of the primary treasures of our nation’.83 The Court appears to be committed to a ‘politics of difference’ that in certain respects resembles that outlined by Iris Marion Young in Justice and the Politics of Difference.84

Young’s contrast between what she considers the traditional liberal approach and the politics of difference is instructive:

The vision of liberalism as the transcendence of group difference seeks to abolish the public and political significance of group difference, while retaining and promoting both individual and group diversity in private, or nonpolitical, social contexts.85

For Young, ‘[g]roups cannot be socially equal unless their specific experience, culture and social contributions are publicly affirmed and recognised’.86 She deprecates the ‘the typical liberal approach ... which tolerates any behaviour so long as it is kept in private’,87 arguing that ‘the private, as traditionally conceived, is what should be

83 Pillay (n 6 above) para 92. 84 IM Young Justice and the politics of difference (1990). It may be useful to deal briefly with the change in the way in which the idea of respect for difference has been viewed in South Africa. In a perversion of the contemporary politics of difference, the apartheid government justified its policy of creating ‘homelands’, an order of ‘plural nations’ in semi-autonomous polities with no economic viability, with reference to something like a multicultural respect for difference. As a result, opponents of apartheid became suspicious of the idea of respect for difference. As R Nixon Homelands, Harlem and Hollywood: South African culture and the world beyond (1994) 206 asserts, ‘Since the late 1950s, difference (as opposed to unity) has been perceived — for sound historical reasons — as a government term. This is so because the South African brand of racial supremacy has been couched as a form of sensitivity to the special needs and cultural particularities of diverse “peoples”. Attentiveness to difference is thus perceived as “apartheid” business, a way of coating state racism with a democratic gloss’. Nevertheless, as John and Jean Comaroff ‘Criminal justice, cultural justice: The limits of liberalism and the pragmatics of difference in the new South Africa’ (1994) 31 American Ethnologist 188 195 observe, there has been in South Africa ‘a growing recognition of the gravitas of difference’. In the jurisprudence of the Constitutional Court, attentiveness to difference has emerged, particularly in its jurisprudence dealing with sexual orientation and religious and cultural accommodation, as an integral part of constitutional justice. 85 n 84 above, 168 (emphasis added). 86 n 84 above, 174 (emphasis added). 87 Young (n 84 above) 161 (emphasis added). (2008) 1 Constitutional Court Review 283 hidden from view, or what cannot be brought out to view. It is connected with shame and incompleteness’.88

The refusal to permit members of cultural groups to manifest their culture publicly is considered by Young and the Constitutional Court to devalue members by relegating cultural practices to the private sphere. It is from this perspective that we should understand Langa CJ’s assertion that ‘the symbolic effect of denying her the right to wear it ... sends a message that Sunali, her religion and her culture are not welcome’.89 From the perspective of the politics of difference, the school’s refusal to permit Sunali to express her cultural identity at school (that is, in public) constitutes an injury to her. Young’s depreciation of liberal toleration chimes with O’Regan J’s statement that ‘treating people as worthy of equal respect in relation to their cultural practices requires more than mere tolerance’.90

3.4 Discrimination, neutrality and fairness

Under the Equality Act, the claimant was required to show that the school’s Code unfairly discriminated against her on the basis of religion or culture. The school argued that in this case there was no comparator, no group relative to which Sunali received inferior treatment. The location of the appropriate comparator was the subject of disagreement between the judgements of Langa CJ and O’Regan J. Langa CJ identified the appropriate comparator as ‘those learners whose sincere religious or cultural beliefs or practices are not compromised by the Code, as compared to those whose beliefs are uncompromised’. He found that ‘the Code has a disparate impact’91 on certain religions and cultures, the members of which are disproportionately burdened or incommoded relative to others who are not members. He took to view that the school’s Code

is not neutral, but enforces mainstream and historically privileged forms of adornment, such as ear studs which also involve the piercing of a body part, at the expense of minority and historically excluded forms.92

This again is consistent with Young’s ‘politics of difference’, according to which

88 Young (n 84 above) 119. 89 Prince (n 6 above) para 85. Langa CJ here paraphrases the Canadian Supreme Court in Multani v Commission Scolaire Marguerite-Bourgeoys [2006] SCC 6 para 79 (Multani): ‘A total prohibition against wearing a kirpan to school undermines the value of this religious symbol and sends students the message that some religious practices do not merit the same protection as others’. 90 n 84 above, 156. 91 Pillay (n 6 above) para 44. 92 As above. 284 Cultural and religious accommodations to school uniform regulations

blindness to difference disadvantages groups whose experience, culture and socialised capacities differ from those of privileged groups [who] ... implicitly define the standards according to which all will be measured. Because their privilege involves not recognising those standards as culturally and experientially specific, the ideal of common humanity in which all can participate without regard to race, gender, religion or sexuality poses as neutral and universal.93

For O’Regan J, by contrast,

the correct comparator is those learners who have been afforded an exemption to allow them to pursue their cultural or religious practices, as against those learners who are denied exemption, like the learner in this case.94

As indicated above, the school had in the past granted exemptions to permit the wearing of red ‘Lakshmi strings’ and of hide bracelets to mark respect after a funeral,95 both of which were ‘associative cultural or religious practices’. According to O’Regan J, the unfairness lay in the school’s inconsistency in exempting certain practices but not others without adequate justification of the differential treatment. Whereas Langa CJ asserted that the Code was non- neutral, O’Regan J appears to have accepted that the Code was neutral: she stated that ‘the Code is entitled to establish neutral rules to govern the school uniform’.96

Which group is the correct comparator? The answer, I think, is that both groups are. Langa CJ was certainly correct that the unfairness that justifies the granting of an exemption is present when a rule imposes a burden on members of a minority or vulnerable cultural or religious group to which others are not subject. On the other hand, O’Regan J is right that the granting of exemptions to permit certain religious and cultural practices allows those who have been denied exemptions to claim that they have been unfairly treated. Since both judges found that there had been discrimination, does it matter which comparator is to be considered the appropriate one? It does: on O’Regan J’s approach, had the school refused to grant any exemptions there would have been no discrimination and the claim would have failed. By contrast, Langa CJ would have still found the rule to be discriminatory and the claim would still have been successful. The correct approach would be to hold that both groups are appropriate comparators and both can give rise to claims for exemptions.

93 n 84 above, 164. 94 Pillay (n 6 above) para 164. 95 Pillay (n 6 above) para 170. 96 Pillay (n 6 above) para 165. (2008) 1 Constitutional Court Review 285

Given the disagreement between Langa CJ and O’Regan J about whether the Code is neutral, who is right? Langa CJ’s determination that the code is non-neutral is preferable, comporting as it substantially does with the approach of the minority of the US Supreme Court in Goldman v Weinberger (Goldman).97 In that case, an orthodox Jewish rabbi serving as a psychologist in the military claimed an exemption from an air force regulation that proscribed the wearing of headgear indoors. The military’s uniform dress regulation permitted the wearing of rings and the wearing of religious symbols underneath military uniforms. This allowed latitude for Christians to wear articles with religious significance. The minority in Goldman were prepared to grant the exemption in part because ‘the accommodation of items Christians were likely to wear contrasted disturbingly with the rigidity of the rule that disfavoured Orthodox Jews and other minorities’.98

The School argued that if an exemption were to be granted, then ‘some of the girls might feel that it is unfair’.99 Langa CJ, with whom O’Regan J agreed on this point, determined, correctly and consistently with the Canadian Supreme Court in its recent Multani decision,100 that once the rule is found to be discriminatory the appropriate response to complaints from those not covered by the exemption is to explain to them that the Code does not impose a burden of cultural denial that is comparable to the burden imposed on them in requiring them to forgo a fashion item.101

3.5 Deference

The school raised the argument that deference should be accorded to the professional judgements of the school authorities regarding school uniforms. Langa CJ responded by conceding that ‘the Court must give due weight to the opinion of experts, including school authorities, who are particularly knowledgeable in their area’. He insisted, however, that courts ‘are best qualified and constitutionally mandated’ to answer the question ‘whether the fundamental right to equality has been violated, which in turn requires the Court to determine what obligations the school bears to accommodate diversity reasonably’.102 ‘The Court cannot abdicate its duty by deferring to the school’s view on the requirements of fairness’ since

97 475 US 503 (1986). 98 Greenawalt (n 4 above) 165. 99 Pillay (n 6 above) para 103. 100 n 89 above, para 104. 101 As O’Regan J observed, ‘A school is an ideal place to educate other learners about the difference between fashion and cultural practices, and should an exemption for nose-studs be granted, a school would be obliged to furnish such education to its learners’ (Pillay (n 6 above) para 172). 102 Pillay (n 6 above) para 81. 286 Cultural and religious accommodations to school uniform regulations to do so would be to make the school a judge in its own case.103 Langa CJ referred approvingly to the fact that the Code had been ‘devised after extensive consultation with parents, educators, staff and learners’, but decided that the fact of consultation ‘does not immunise the resultant decisions ... from constitutional scrutiny and review’ since

many individual communities still retain historically unequal power relations or historically skewed population groups which may make it more likely that local decisions will infringe on the rights of disfavoured groups.104

Langa CJ’s view of judicial deference is correct.105 A moderate degree of deference should be accorded school authorities because they possess specialist expertise. The Court should allow ample scope for the development of policy and the promulgation of regulations by the school administration, recognising that rights can sometimes properly give way to contrary public needs, such as the creation of conditions for effective education. Yet the Court, as ultimate authority on questions of right, should arrive at its own independent evaluation, rather than accepting the view of the school authorities. Brennan J took this view in Goldman, criticising the majority for deferring unduly to the judgement of the military.106

3.6 A slippery slope? The end of school uniforms?

Langa CJ rejected an objection that to grant an exemption in this case would result in a ‘slippery slope scenario’, an opening of the floodgates to endless claims for accommodation for, amongst other things ‘dreadlocks, body piercings, tattoos and loincloths’.107 On this objection, once nose studs have been accommodated, fairness requires that other claims for exemptions be granted, resulting in the erosion of the uniformity of school uniforms to the point that there would be a uniform in name only. This argument has no merit, Langa CJ claimed, since, first, exemptions must only be considered for ‘bona fide religious and cultural practices’.108 Second, claims for exemptions may be refused on the grounds that ‘a practice may be so insignificant to the person concerned that it does not require a departure from the ordinary uniform’.109 Third, schools could refuse

103 Pillay (n 6 above) para 81. 104 Pillay (n 6 above) para 83. 105 For a discussion of what constitutes appropriate deference in the context of rights adjudication, see P Lenta ‘Judicial deference and rights’ (2006) Tydskrif vir Suid-Afrikaanse Reg 456. 106 n 97 above, 515. 107 Pillay (n 6 above) para 107. 108 As above. 109 Pillay (n 6 above) para 114. (2008) 1 Constitutional Court Review 287 to accommodate a particular practice ‘if accommodating [it] would impose an unreasonable burden on the School’:110 if the practice will ‘create a real possibility of disruption ... threaten[ing] academic standards or discipline’.111 Finally, as to the extent that other cultural and religious practices are exempted from the Code, ‘that is something to be celebrated, not feared’ since ‘the display of religion and public is ... a pageant of diversity which will enrich our schools and in turn our country’.112

Langa CJ’s response accords with that of Brennan J in Goldman. Brennan J stated that each claim would have to be evaluated against the reasons for refusing it. He added that the court could defer to ‘dress and grooming rules that have a reasoned basis in, for example, functional utility, health and safety considerations and the goal of polished, professional appearance’.113 Claims for accommodation could justifiably be refused in the case of departures from the uniform that infringed these requirements.

Despite the US Supreme Court’s recent characterisation of the ‘slippery slope’ argument as ‘the classic rejoinder of bureaucrats throughout history’,114 the raising of this concern is not always an indication of bad faith. It may instead reflect ‘the reasonable belief that the proliferating recognition of difference might itself generate an accelerating and potentially destabilising realignment of institutional practices’.115 Langa CJ was adamant that the decision in Pillay ‘does not abolish school uniforms’.116 Yet, given the broadness of culture, or as O’Regan J put it, the fact that ‘all human beings have a culture’,117 the potential number of claims for accommodation that may have to be granted consistently with the jurisprudence in Pillay may be high. Langa J mentions turbans, yarmulkes and headscarves,118 all of which schools would have to accommodate. What about dreadlocks? Provided they were tied up, they may have to be accommodated as an expression of Rastafarian culture or religion. Body piercings and tattoos? Perhaps: tattoos, for example, may express not only the faith of Coptic Christians but also Maori culture.

110 Pillay (n 6 above) para 107. 111 Pillay (n 6 above) para 114. 112 Pillay (n 6 above) para 107. 113 Goldman (n 97 above) 519. 114 Gonzalez (n 37 above) 15. 115 A Sarat & R Berkowitz ‘Disorderly differences: Recognition, accommodation, and American law’ (1994) 6 Yale Journal of Law and the Humanities 285 288. 116 Pillay (n 6 above) para 114. 117 Pillay (n 6 above) para 150. 118 Pillay (n 6 above) para 106. 288 Cultural and religious accommodations to school uniform regulations

Kilts? A pupil of Scottish extraction may insist on wearing a kilt at school and it is not clear, on the approach outlined in Pillay, that schools could refuse to grant this exemption.119 It is difficult to predict how many claims for religious and cultural accommodation are likely to be pressed,120 but were all the claims that are required to be accommodated in accordance with the approach in Pillay to be brought, school uniforms, although they might not be abolished, would be far less uniform. In that case, the educational objectives promoted by uniforms may be to some extent undermined. This is the price that may have to be paid for extending the obligation to accommodate to include optional cultural practices.

A claim that could be refused on the grounds of the approach in Pillay is a claim for an exemption to permit a pupil to carry a weapon. In Multani, the Canadian Supreme Court — adopting an approach to religious exemptions strongly resembling the approach of the Constitutional Court in Pillay to religious and cultural exemptions — decided that the ‘duty to make reasonable accommodation’121 obliged a school to accommodate a Sikh pupil’s demand for an exemption to permit him to wear a kirpan, a metal dagger, which Sikh men believe themselves to be religiously obligated to wear at all times. Despite accepting that a kirpan is a bladed weapon, the Supreme Court held that the increased risk of violence introduced by granting the exemption to the school’s code de vie was acceptably low, provided the pupil wore the kirpan sealed and sewn inside his clothing. To refuse an accommodation in this case, the Court (consistently with the view of the Constitutional Court in Pillay) added, would be to ‘stifle the promotion of values such as multiculturalism, diversity and the development of an educational culture respectful of the rights of others’.122

The Constitutional Court would be unlikely to grant an exemption for a kirpan (and mistaken were it to do so), since South Africa is ‘an

119 It might be contended in response that South Indian/Hindu culture is a more authentically South African culture than Scots Presbyterian culture, but that would be mistaken. While Hindu South Indians arrived in South Africa as indentured labourers from the 1860s onwards, the Scots began to arrive before that. Two aspects of Scots culture are relevant here. First, it is strongly religious. During the mid nineteenth century, a group of Scots who were ministers of the Presbyterian Church of Scotland helped to revitalise the Dutch Reformed Church. Secondly, it is difficult to imagine a more forceful and nationalistic assertion of culture than that characteristic of the Scots. 120 The granting of an exemption by the US Supreme Court in Yoder did not result in an avalanche of claims. See W Galston Liberal pluralism (2002) 121: ‘The limited public education accommodation for the Old Order Amish endorsed by the Supreme Court in Wisconsin v Yoder a quarter of a century ago has not lead to an escalation of faith-based demands. Indeed few other groups have sought similar treatment for themselves’. 121 Multani (n 89 above) para 53. 122 Multani (n 89 above) para 78. (2008) 1 Constitutional Court Review 289 exceptionally, possibly uniquely, violent society’,123 whereas Canada is one of the least violent countries in the world. Yet its probable refusal to do so would not necessarily reflect a difference in approach between the Constitutional Court and the Canadian Supreme Court: the social contexts of Canada and South Africa are sufficiently different as to result in the same approach to accommodations leading in the case of weapons to different outcomes.

4 Accommodating dress and adornment: A comparative analysis

The jurisprudence of the UK House of Lords indicates that it may be prepared, in certain circumstances, to grant an exemption to permit the wearing of religious clothing and adornments, although its approach suggests that it is less solicitous towards claims for exemptions to school uniforms for religious dress than its South African and Canadian counterparts. In Begum,124 the House of Lords rejected a claim for a religious exemption from a Muslim pupil, who wished to wear not only a headscarf (hijab), but also a jilbab, a long garment the purpose of which is to hide the contours of the female body. Despite accepting that the application was motivated by a sincere belief on the part of the pupil that she was under a religious obligation to wear a jilbab, the Court denied the claim, for the following reasons: first, the current school uniform’s dispensation for Muslim pupils was designed in consultation with Muslim authorities to meet their religious requirements; second, the pupil could have transferred to another school in the area that permitted the wearing of a jilbab;125 third, deference — a ‘margin of appreciation’126 — should be accorded to the decision of Parliament to allow schools to make their own decisions about uniforms127 and, fourth, to grant this exemption might encourage undesirable religious extremism and might place undue pressure on Muslim female pupils who did not wish to wear the jilbab to do so.128

Faced with a case with similar facts to those in Begum, the jurisprudence of the Constitutional Court would seem to commit it to

123 A Altbeker A country at war with itself: South Africa’s crisis of crime (2007) 12. 124 R (on the application of Begum) v Headteacher and Governors of Denbigh High School [2006] UKHL 15 (Begum). 125 Begum (n 124 above) paras 25 & 89. 126 Begum (n 124 above) para 62. 127 Begum (n 124 above) para 34. 128 Begum (n 124 above) para 65. 290 Cultural and religious accommodations to school uniform regulations granting an exemption to permit the pupil to wear a jilbab.129 Langa CJ rejected the argument that the Court should substantially defer to the judgement of the school authorities about whether an exemption would interfere with the effective running of the school. He also disputed the contention that the pupil could have relocated to another school that permitted the practice, since this would fail to celebrate diversity, and ‘would be to marginalise religions and cultures’.130 The Constitutional Court is likely to take the view, as far as claims for accommodation are concerned, that there is only one sensible stopping point: the point at which religious and cultural practices violate the rights of others or interfere with the business of the school. So, for example, the Constitutional Court should not only refuse to accommodate weapons in schools; it should also agree with Silber J in Headteachers and Governors of Y School131 that it is not reasonable to expect a school to accommodate a niqab, a veil which covers the entire face and head save for the eyes, which would frustrate educational objectives by impeding identification of and communication with pupils.132

129 Begum was followed in Playfoot (a minor), R (on the application of) v Millais School [2007] EWHC 1698 (Admin) (16 July 2007), in which the High Court refused to grant an exemption to permit a female pupil to wear a silver ring that expressed her commitment, as a Christian, to remaining sexually abstinent prior to marriage. The Court rejected the claim primarily on the grounds that the wearing of the ring at school was not required by her religion unlike other practices that the school had accommodated, such as the wearing of headscarves by Muslim pupils and the wearing of Kara bangles by Sikh girls. Since the Constitutional Court has emphasised that the duty of reasonable accommodation includes optional as well as mandatory practices, it may well be that it would grant an exemption in these circumstances. The reasoning in Playfoot is anyway inconsistent with that of the House of Lords in Williamson, in which it was stated that a perceived obligation is not a ‘prerequisite to manifestation of a belief in practice’ (n 20 above, para 33). 130 Pillay (n 6 above) para 92. 131 R (on the application of X) v Headteachers and Governors of Y School [2007] EWCH 298 (Admin); [2007] HRLR 20 (QBD (Admin)). 132 Two points about this. First, although the Court would probably agree with Silber J’s conclusion, it would decline to follow the reasoning of Silber J, who, adopting the approach of the House of Lords in Begum, declined to grant the exemption on the grounds that deference is due to school authorities and because the pupil concerned could have transferred to another school which permits the wearing of a niqab. Secondly, if the school is acting reasonably by declining a pupil’s request that she be permitted to wear a niqab, a school’s decision to refuse to permit a teacher to wear a niqab is equally reasonable, since a teacher’s obscuring of her face and mouth makes non-verbal communication more difficult and, as a result, may impede effective communication between the teacher and pupils. The UK Employment Appeal Tribunal took this view in Azmi v Kirklees Metropolitan Borough Council [2007] UKEAT/0009/07 (30 March 2007). Of course, non-verbal communication might be less important with older learners. See, for example, Nussbaum (n 8 above) 350. (2008) 1 Constitutional Court Review 291

The US Supreme Court has yet to be confronted with a claim for an exemption from school uniform regulations. Would it be likely to grant a claim for an exemption to permit religious clothing and adornments?133 In Smith,134 the US Supreme Court held that, other than in exceptional circumstances, there is no constitutional requirement to grant religious exemptions from facially neutral laws which do not have the purpose, but may have the incidental effect, of burdening the practice of religion. Post-Smith, a facially neutral school uniform policy, if fairly administered, would be constitutionally valid. Nevertheless, as the Supreme Court stated in a subsequent decision, Church of Lukumi Babalu Aye,135 the regulation would have to be applied in a religiously neutral manner. If exemptions are made for some religious groups but not for others, as they were in Pillay, they may be held unconstitutional. Moreover, the regulation might be challenged, as in Pillay, under a ‘hybrid’ claim that it infringes not only religious liberty but also freedom of expression. The US Supreme Court in Smith indicated that hybrid claims might be treated differently from free exercise claims standing alone.136 The claim that the wearing of clothing and adornments constitutes expressive conduct appears strong. In Pillay, Langa J found that the ban on nose studs under the school uniform infringed the pupil’s right to freedom of expression.137 And in her concurring judgement in Begum, Baroness Hale of Richmond contended that a woman wearing a hijab ‘may have chosen the garment as a mark of her defiant political identity’ and may be engaging ‘in a highly complex autonomous act intended to use the resources of the

133 See J Mikhail ‘The free exercise of religion: An American perspective’ in M Mahlman & H Rottleuthner (eds) Ein Neuer Kampf Der Religionen? Staat, Recht und Religïose Toleranz (2006) 271. 134 n 25 above. 135 n 1 above. 136 n 25 above, 881 - 882. The category of the ‘hybrid’ claim, though difficult to interpret, has resulted in some successful claims, particularly in areas that link free speech to religious liberty. See, for example, Chalifoux v New Caney Ind School Dist, 976 F Supp 659 (SC Tex 1997). 137 n 6 above, para 94. 292 Cultural and religious accommodations to school uniform regulations tradition both to change and to preserve it’.138

5Conclusion

The relative levels of protection for religious liberty afforded by the South African Constitutional Court, the UK House of Lords, the Canadian Supreme Court and the US Supreme Court in response to claims for exemptions from school uniforms may now be assessed. The highest level of protection is provided by the South African Constitutional Court and the Canadian Supreme Court. These courts are most accommodating of clothing and adornment practices of religious (and in the case of the South African Constitutional Court, cultural) significance, emphasising the values of multiculturalism and diversity.

The UK House of Lords offers a lower level of protection for religious claimants. It is more deferential to the legislature, on the grounds of a Dworkinian distinction between policy and principle (as in Williamson), and to school authorities, on the basis of their specialist expertise (as in Begum). It takes the view that pupils should as far as possible accommodate themselves to the relevant regulation by, for example, transferring where reasonably possible to another school that would permit the practice for which protection is being sought. By contrast, the South African Constitutional Court and the Canadian Supreme Court’s emphasis on affirming culturally and religiously differentiated identities suggests a stance in accordance with which pupils are to a lesser extent required to accommodate themselves to the regulation. The approach of the House of Lords in Begum reflects an anxiety about religious extremism which is pervasive in post 9/11 and 7/7 UK but not in the different

138 n 124 above, para 94. The US Supreme Court has not yet addressed the issue of whether school uniforms violate the right to freedom of expression, but lower courts have decided that school uniforms do not violate the right to free speech. The US Court of Appeals upheld the constitutionality of a mandatory public school uniform policy in a Louisiana school district (Canady v Bossier Parish School Board, 240 F 3d 437, US Ct App 6th Circuit (2001)). The court found that ‘improving the educational process’ was an important and substantial government interest. In upholding the imposition of mandatory uniforms, the court noted that the school’s policy was ‘viewpoint-neutral’. Most importantly of all for the present case, the Court went on to find that [t[he School Board's purpose for enacting the uniform policy is to increase test scores and reduce disciplinary problems throughout the school system. This purpose is in no way related to the suppression of student speech. Although students are restricted from wearing clothing of their choice at school, students remain free to wear what they want after school hours. Students may still express their views through other mediums during the school day. The uniform requirement does not bar the important ‘personal intercommunication among students’ necessary to an effective educational process. (2008) 1 Constitutional Court Review 293 environment of South Africa:139 the House of Lords’s concern about the prospect of undesirable extremism and the potential for moderate Muslims to become radicalised is not even engaged with in Pillay, despite being raised in the school’s Heads of Argument. The lowest level of protection is that provided by the US Supreme Court, which will in most cases decline to grant exemptions to facially neutral state laws and local regulations, including school uniform regulations.

139 JM Coetzee Diary of a bad year (2007) 122 describes South Africa as a country in which ‘Islamist extremism still takes a lowly place on the list of public concerns’.

THE CASE FOR RELIGIOUS INCLUSIVISM AND THE JUDICIAL RECOGNITION OF RELIGIOUS ASSOCIATIONAL RIGHTS: A RESPONSE TO LENTA

Iain T Benson*

The idea of culture derived from anthropology, a discipline which studied the encapsulated exotic, is no longer appropriate. There are no longer (if there ever were) single cultures in any country, polity or legal system, but many. Cultures are complex conversations within any social formation. These conversations have many voices.1

1Introduction

Patrick Lenta, in his comment on Pillay and in earlier work,2 has done South African jurisprudence a great service by forcing us to attend to the complexity of cultures and the problems associated with the

* Barrister & Solicitor, Executive Director, Centre for Cultural Renewal (Ottawa, Canada). Research Associate, SAIFAC, Johannesburg. Research Fellow, Faculty of Law, University of the Free State, Bloemfontein. I would like to acknowledge the contributions and suggestions made to this article from those who attended the Constitutional Court Review Symposium — sponsored by the South African Institute for Advanced Constitutional, International and Human Rights Law, Constitutional Law of South Africa and the Konrad Adenauer Stiftung — in August 2008. I would be remiss if I did not single out the Executive Director of SAIFAC, Theunis Roux, for his encouragement. The expert, substantive editorial interventions of University of Pretoria Professor, and SAIFAC Senior Research Fellow, Stuart Woolman, have significantly improved this work. Justice Laurie Ackermann, as always, has been an unstinting source of support and helpful insights. In addition, I have benefited from discussions with Shaun de Freitas (Associate Professor) Department of Constitutional Law and Philosophy of Law, Faculty of Law, University of the Free State: he generously shared with me his own unpublished paper on Pillay. Errors and omissions remain the author’s alone. 1 M Chanock ‘Human rights and cultural branding: Who speaks and how’ in A An- Na’im (ed) Cultural transformation and human rights in Africa (2002) 41. See also S Benhabib The claims of culture: Equality and diversity in the global era (2002) 3-9. 2 See P Lenta ‘Cultural and religious accommodations to school uniform regulations’ (‘School uniforms’) (2008) 1 Constitutional Court Review 259 ‘Muslim headscarves in schools and in the workplace’ (2007) 124 South African Law Journal 296 (‘Headscarves’) and ‘Religious liberty and cultural accommodation’ (2005) 122 South African Law Journal 352 363-371 (‘Accommodation’).

295 296 A response to Lenta accommodation of religious belief and practice in constitutional democracies. The usefulness of his work extends beyond South Africa’s borders. Questions of religious (and cultural) accommodation are on the front burner in many other constitutional democracies. This reply focuses on a narrow set of questions that fall within the framework Lenta and others have constructed for discussions about the relationship between religion and the State. First, I want to suggest that: (a) commentators and courts alike reflexively invoke a rather reductive conception of ‘the secular’; (b) that this reductive conception is generally ‘anti-religious’; and (c) the South African Constitution ought to be read so as to recognise that religions and religious practice occupy an important role in the formation of public discourse and the support of public institutions.3 Religions are not merely to be tolerated: In societies such as South Africa, they ought to be embraced. Second, I hope to demonstrate that the failure to take seriously religions and cultural communities with comprehensive understandings of the good, results in a body of jurisprudence — in South Africa and Canada — that privileges individual religious autonomy over the communal dimension of rights to religious freedom and practice.

2 The problem of single cultures or ‘convergence liberalism’

John Gray and others have of late suggested that the concept of a ‘single culture’ or a ‘one-size fits all’ conception of the public sphere hides illiberal strains of liberalism implicitly committed to an endorsement of ‘convergence’. Gray has described the two main approaches to liberalism in relation to pluralism as follows:

Liberalism contains two philosophies. In one, toleration is justified as a means to truth. In this view, toleration is an instrument of rational consensus, and a diversity of ways of life is endured in the faith that it is destined to disappear. In the other, toleration is valued as a condition of peace, and divergent ways of living are welcomed as marks of diversity in the good life. The first conception supports an ideal of ultimate convergence on values, the latter an ideal of modus vivendi. Liberalism's future lies in turning its face away from the ideal of rational consensus and looking instead to modus vivendi.

3 These readings of ‘secularism’ may appear to some to be variant readings. However, they are consistent with the meaning of the neologism coined by George Jacob Holyoake. While others may use the term in a way that suggests that secularism is religion-friendly and consistent with the contours of every liberal democracy, I think it important to emphasise the often aggressive strategy of religious exclusion and marginalisation that occurs in many ‘secular’ liberal democracies (eg, France). (2008) 1 Constitutional Court Review 297

The predominant liberal view of toleration sees it as a means to a universal civilisation. If we give up this view, and welcome a world that contains many ways of life and regimes, we will have to think afresh about human rights and democratic government. We will refashion these inheritances to serve a different liberal philosophy. We will come to think of human rights as convenient articles of peace, whereby individuals and communities with conflicting values and interest may consent to coexist.4

Such a characterisation of liberalism would seem to fit Lenta’s notion that modern liberal states often assume or propound a false view of the public sphere as a ‘neutral’ space. However, Lenta does not discuss variant interpretations of the ‘secular’. And when he tends to use the term at all, he employs it in a Rawlsian sense5 to mean the opposite of ‘religious.’ This disjunction is odd (and flawed). For the Rawlsian characterisation6 of the secular and the religious leads us toward religious exclusivism — a terminal point largely at odds with the religious inclusivism of Lenta’s analysis.7

In religiously exclusive liberal discourse, ‘secular’ often serves as a synonym for ‘non-religious’. While the two terms are commonly conflated in this manner, this terminology does not sit comfortably with Lenta’s call for inclusivity of religious believers and what he describes as a constitutional ‘presumption in favour of the government’s being required to grant an exemption [for religious belief].’8 Such an approach, according to Lenta, means ‘departing from the principle of uniformity.’9 On this matter, Lenta and I are in accord. Further evidence of a more nuanced position can be gleaned from Lenta’s work as a whole. In ‘Headscarf’, Lenta describes the differences between the French approach to laïcité and the Canadian approach to multi-cultural inclusion — and then notes that the Canadian commitment to inclusivism — even with its limitations —

4 J Gray Two faces of liberalism (2000) 105. See also PD Lauwers ‘Religion and the ambiguities of liberal pluralism: A Canadian perspective’ (2007) 37 The Supreme Court Law Review 1. 5 See Lenta ‘School uniforms’ (n 2 above) 264. 6 See J Rawls Political liberalism (1993). For a useful criticism of the approach taken by John Rawls, see PF Campos ‘Secular fundamentalism’ (1994) 94 Columbia Law Review 1814 1825. Campos would have been better served had he noted that Rawls’ argument is not so much ‘secular’ as ‘secularist’. That is, a secularist is someone who advocates secularism in the construction of the political domain. See further CL Niles ‘Epistemological nonsense?: The secular/ religious distinction’ (2003) 17 Notre Dame Journal of Law, Ethics & Public Policy 561 575 n 50 (criticising Rawls' A theory of justice on these same grounds.) 7 Religiously inclusive secularism, which we might better call a ‘religiously inclusive public sphere’, has been adopted by the Supreme Court of Canada. The Canadian jurisprudence shall be discussed in greater detail below. See Chamberlain v Surrey Sch Dist No 36, [2002] 4 SCR 710 749 (Can) (‘Chamberlain’) 8 Lenta ‘Headscarves’ (n 2 above) 307. 9 As above. 298 A response to Lenta more closely approximates both the South African position and his preferred reading of South Africa’s basic law.10

And yet, Lenta remains committed to the proposition that the Western liberal political tradition is committed to freedom of religion largely as a response to historical religious strife. Of course, there’s more than a grain of truth to that proposition. For one can mark liberalism and constitutionalism’s true beginning with Locke’s Letter on Toleration. But if one does mark ‘religious freedom’ as beginning with this modest, but still revolutionary, tract, then, upon a closer reading, it is hard to view — from a Lockean perspective — state neutrality as liberal exclusivism regarding religion, as anything but an anti-religious ideology.11

For many liberals, the term ‘secular’ means only ‘non-religious’. But that mistakes the term’s origins — the saecularum — as a particular way of marking periods of time. It is important, therefore, to stop and to examine what we mean when we use terms such as ‘secular’ or ‘secularism’. The failure to do so can lead, I contend, to fundamentally anti-religious outcomes.12 A decision from the Canadian courts provides an illuminating and instructive account of why getting this term right is of such great import.

Mr Justice McKenzie, in the unanimous decision of the British Columbia Court of Appeal in Chamberlain, stated that:

In my opinion, ‘strictly secular’ in the School Act can only mean pluralist in the sense that moral positions are to be accorded standing in the public square irrespective of whether the position flows out of a

10 The South African approach might well be said to accept a ‘co-operation’ of Church and State — rather than any American notion of ‘strict separation’. Of course, anyone au fait with recent American constitutional law knows that the US Constitution is no longer read as endorsing an absolutely strict separation of Church and State. The principles of such co-operation and useful distinctions about avoidance of both atheistic and religious theocracy are neatly set out in A Sachs Protecting human rights in a new South Africa (1990) 43–49. See also S Woolman ‘Community rights: Language, culture and religion’ in S Woolman et al (eds) Constitutional law of South Africa (2nd Edition, OS, 2008) ch 58. 11 Something of this sort is occurring now in Canada. A Commission made up of two academics — including noted philosopher Charles Taylor — have endorsed ‘open secularism’ and defined secularism at variance with its history. For comment, see IT Benson & TL Nguyen ‘The need to re-evaluate the language of the secular and secularism in the quest for fair treatment of minorities and belief in Quebec and Canada today’ 18 December 2007, http://www.culturalrenewal.ca/downloads/ sb_culturalrenewal/BriefTaylorBouchardCommissionDecember2007Fin a.pdf (accessed 12 September 2008). 12 For more on the history of the terms ‘secularism’ and ‘secular’, see IT Benson ‘Considering secularism’ in D Farrow (ed) Recognising religion in a secular society: Essays in pluralism, religion and public policy (2004) 83–93 and ‘Notes towards a (re)definition of the secular’ (2000) 33 University of British Columbia Law Review 519 520–538. (2008) 1 Constitutional Court Review 299

conscience that is religiously informed or not. That meaning of strictly secular is thus pluralist or inclusive in the widest sense ... No society can be said to be truly free where only those whose morals are uninfluenced by religion are entitled to participate in deliberations related to moral issues of education in public schools. In my respectful view ‘strictly secular’ so interpreted could not survive scrutiny in light of the freedom of conscience and religion guaranteed by section 2 of the Charter [conscience and religion] and equality rights guaranteed by section 15.13

What is said here about moral positions applies equally to religious and cultural beliefs in a public school setting such as Pillay. Simply put, convictions emanating from religious beliefs ought to be at no disadvantage in terms of public respect by comparison to belief sets that emanate from non-religious convictions. When the case reached the Supreme Court of Canada, all nine judges agreed with the reasoning of McKenzie J. ‘Secular’ in Canadian constitutional jurisprudence embraces a religiously inclusive understanding of the term.14

Justice Gonthier provides further support for this proposition when he writes:

In my view, Saunders J [the trial judge] below erred in her assumption that ‘secular’ effectively meant ‘non-religious’. This is incorrect since nothing in the Charter, political or democratic theory, or a proper understanding of pluralism demands that atheistically based moral positions trump religiously based moral positions on matters of public policy. I note that the preamble to the Charter itself establishes that ‘ ... Canada is founded upon principles that recognise the supremacy of God and the rule of law’. According to the reasoning espoused by Saunders J, if one's moral view manifests from a religiously grounded faith, it is not to be heard in the public square, but if it does not, then it is publicly acceptable. The problem with this approach is that everyone has ‘belief’ or ‘faith’ in something, be it atheistic, agnostic or religious. To construe the ‘secular’ as the realm of the ‘unbelief’ is therefore erroneous. Given this, why, then, should the religiously informed conscience be placed at a public disadvantage or disqualification? To do so would be to distort liberal principles in an illiberal fashion and would provide only a feeble notion of pluralism. The key is that people will disagree about important issues, and such disagreement, where it does

13 Chamberlain v Surrey School Board (2000) 80 BCLR (3d) 181 (CA) (‘Chamberlain’) at paras 33 & 34, reversing Chamberlain v Surrey School Board (1998) 60 BCLR (3d) 311 (SC) (emphasis added). For a detailed analysis of this decision, see IT Benson & B Miller ‘Court corrects erroneous understanding of the secular and respects parental rights’ Lex View No. 40.0, available at http:// www.culturalrenewal.ca/qry/page.taf?id=64 (accessed 12 December 2008). 14 Chamberlain (n 7 above) 749. Madam Justice McLachlin, who wrote the decision of the Chamberlain majority, endorsed the reasoning of Justice Gonthier on the correct interpretation of ‘secular’. 300 A response to Lenta

not imperil community living, must be capable of being accommodated at the core of a modern pluralism [emphasis added].15

The approach of the Supreme Court of Canada and the implicit recognition in Pillay that a public school must accommodate a variety of beliefs (religious or cultural) are at stark variance with the approaches taken by countries such as France. In France, the public sphere has generally been stripped of all religious accommodation under the mistaken belief that the removal of all religious signifiers is ‘neutral.’ In Pillay, the Constitutional Court of South Africa arrived at a similar conclusion about the place of religion in the public sphere as the Canadian Supreme Court did in Chamberlain. (I would have preferred, however, a somewhat different characterisation of the relationship between the ‘sacred’ and the ‘secular’.) In Fourie, Sachs J offers the kind of careful and nuanced understanding of the public realm as a sphere of ‘co-existence’ that is largely on all fours with the notion of religious inclusivism defended here and set out as above in Chamberlain:

In the open and democratic society contemplated by the Constitution there must be mutually respectful co-existence between the secular and the sacred. The function of the Court is to recognise the sphere which each inhabits, not to force the one into the sphere of the other ... The hallmark of an open and democratic society is its capacity to accommodate and manage difference of intensely-held world views and lifestyles in a reasonable and fair manner. The objective of the Constitution is to allow different concepts about the nature of human existence to inhabit the same public realm, and to do so in a manner that is not mutually destructive and that at the same time enables government to function in a way that shows equal concern and respect for all ... It is clear from the above that acknowledgment by the State of the right of same-sex couples to enjoy the same status, entitlements and responsibilities as marriage law accords to heterosexual couples is in no way inconsistent with the rights of religious organisations to continue to refuse to celebrate same-sex marriages. The constitutional claims of same-sex couples can accordingly not be negated by invoking the rights of believers to have their religious freedom respected. The two sets of interests involved do not collide; they co-exist in a constitutional realm based on accommodation of diversity.16

A more coherent and accurate way of describing what is at issue is to recognise that religious beliefs and believers are within the public sphere shared by all. Utilising the secular/sacred dichotomy confuses the principles at issue. What should be emphasised is the co-operative relationship — between politics and religion — in the same manner as

15 Chamberlain (n 7 above) para 137 (emphasis added). 16 Minister of Home Affairs & Another v Fourie & (Doctors for Life Intenational & Others, amici curiae); Lesbian and Gay Equality Project & Others v Minister of Home Affairs 2006 1 SA 524 (CC) paras 94–98. (2008) 1 Constitutional Court Review 301 the passage correctly notes the co-operation of the religious and the non-religious within ‘the same public realm.’ Unfortunately, the distinction between the ‘secular’ and the ‘sacred’ at the outset does little to assist what I take to be South Africa’s religiously inclusive position. For a religious citizen, the public order of the State, too, has its own sacred dimension. Why? Because everything within creation flows, in a some sense, from ‘grace’ or ‘the holy’ or ‘the divine’. The Fourie Court would have done better to describe the public realm as encompassing believers of all sorts — whether atheist, agnostic or religious — and identifying the role of the law, when certain types of conflicts emerge, as ordering all relationships according to principles of justice. When most people use the term ‘secular’ they mean ‘public’ — and it would clarify matters greatly if the Constitutional Court (and courts elsewhere) said so in the future.

Lenta’s use of the work of Iris Marion Young,17 which is critical of ‘the liberal tendency’ (the singular conception of ‘liberal’ here is telling) is good as far as it goes in suggesting a tendency towards the privatisation and consequent public irrelevance of religion within one strand of liberal thought. Any monofocal view of ‘liberal’ fails, however, to look at different conceptions of liberalism and the conceptions of ‘the secular’ themselves that play into such an implicit or explicit privatisation and marginalisation of the religious and the sacred. Secularism (as an anti-religious movement) goes unanalysed in Lenta’s work. But the same, it must said, is true of Pillay and Fourie and the Supreme Court of Canada’s reasoning in Chamberlain.

In discussing Pillay, Lenta does not address how we will try to accommodate (if we should) non-religious or non-cultural claims regarding belief sets and practices. Unfair discrimination, by both the State and private parties, including on the grounds of both religion and culture, is specifically prohibited by sections 9(3) and (4) of the Constitution. Section 9(3) reads:

The state may not unfairly discriminate directly or indirectly against anyone on one or more grounds, including race, gender, sex, pregnancy, marital status, ethnic or social origin, colour, sexual orientation, age, disability, religion, conscience, belief, culture, language and birth.

Pillay focuses exclusively on ‘religion’ and ‘culture’ and concludes that cultural practices are entitled to the same degree of constitutional solicitude as religious practices. One would do well to heed without necessarily agreeing with Justice O’Regan’s concerns about the extent to which culture and religion differ in terms of the

17 See Lenta ‘School uniforms’ (n 2 above) 282. 302 A response to Lenta community and the individual.18 Two important questions, raised by Pillay, but not engaged by Lenta, are: (1) can there be ‘a culture of one?’ and (2) does the accommodation approach apply if the objection to wearing of a uniform is based simply on ‘belief’ or ‘opinion’?

3 Equating ‘culture’ with religion: Can there be a ‘religion or culture of one?’

If it is true, as John Donne famously observed, that ‘no man is an island’,19 then it might be fair to ask if one person can be a culture or one person’s beliefs a religion. What are the requirements in order for something to constitute a ‘culture’ or ‘religion’? The judgment in Pillay does not help us here. Chief Justice Langa’s judgment begins and ends its analysis with the following ‘brief introduction’:

Without attempting to provide any form of definition, religion is ordinarily concerned with personal faith and belief, while culture generally relates to traditions and beliefs developed by a community. However, there will often be a great deal of overlap between the two; religious practices are frequently informed not only by faith but also by custom, while cultural beliefs do not develop in a vacuum and may be based on the community’s underlying religious or spiritual beliefs. Therefore, while it is possible for a belief or practice to be purely religious or purely cultural, it is equally possible for it to be both religious and cultural.20

A tension always exists between understanding religion in its personal dimensions and broader communal dimensions. Too great a focus on the individual’s belief runs the risk of trivialising the communal foundation to which any individual belief is invariably related. It is not too strong a statement to say that Pillay gets things back to front: The meaning that an individual draws from a religious belief or a religious practice is contingent upon the existence of a pre-existing religious community with well-developed tenets of belief and practice. Put differently, in a very important sense, the religious community creates and nurtures the religious believer.21 Unfortunately, Pillay fails to recognise this relationship.

18 KwaZulu-Natal MEC for Education v Pillay 2008 1 SA 474 (CC) (Pillay) paras 141–148. As I expand upon below, Justice O’Regan’s reasons view religion individualistically and fail to accord religion an appropriate associational dimension. 19 J Donne ‘Meditation XVII’ in MH Abrams et al (eds) The Norton anthology of English literature Vol. 1 (3rd ed, 1974) 1215. 20 Pillay (n 18 above) para 47. 21 S Woolman ‘Freedom of association’ in Woolman et al (eds) (n 10 above) ch 44. See also M Walzer ‘On involuntary association’ in A Gutmann (ed) Freedom of association (1998) 64. (2008) 1 Constitutional Court Review 303

Consider this statement from another Constitutional Court decision on religious faith and culture, a decision referred to in the most recent Supreme Court of Canada decision touching on religious rights:

For many believers, their relationship with God or creation is central to all their activities. It concerns their capacity to relate in an intensely meaningful fashion to their sense of themselves, their community and their universe. For millions in all walks of life, religion provides support and nurture and a framework for individual and social stability and growth. Religious belief has the capacity to awake concepts of self- worth and human dignity which form the cornerstone of human rights. It affects the believer’s view of society and founds the distinction between right and wrong.22

The reference both to ‘community’ and ‘social stability’ in this passage points towards a better framework for locating the right to religion (and also to culture). For it is precisely in the relationship between the person and the community that the importance of the freedom of religion is best viewed. Social stability, after all, is not something achieved by the individual. Its achievement is a function of the co-ordination of action between and among individuals within a community. Neither Pillay nor the Canadian Supreme Court in Amselem23 seem comfortable with affirming this proposition.

But should they be? For example, we recognise many religiously or culturally sponsored charitable endeavours as public goods. Religious institutions that promote ‘health care’ or ‘education’ are only possible because various communities commit themselves to their realisation and their ongoing maintenance.

In another setting, that of whether a privilege should extend to religious communications in the criminal law area, two justices of the Supreme Court of Canada recognised the wider context and referred to the following passage with approval:

In a very real sense, then, the value of religious confidentiality is the value to society of religion and religious organisations generally. Even from a purely utilitarian perspective, that value cannot be overstated. Religious organisations based on claims to unchanging truths are a stabilising influence in an increasingly fast-paced and atomised society where bonds of community are scarce and worth preserving. Moreover, many provide

22 Christian Education South Africa v Minister of Education 2000 4 SA 757 (CC) para 36. See, generally, I Currie & J de Waal The Bill of Rights handbook (5th ed 2005) 336–357; P Farlam ‘Freedom of religion’ in Woolman et al (eds) (n 10 above) ch 41. See further Bruker v Marcovitz 2007 SCC 54; K Boonstra & IT Benson ‘When should the courts enforce religious obligations?’ Lex View No 63 (2008), available at http://www.culturalrenewal.ca/qry/page.taf?id=150 (accessed December 12, 2008). 23 Syndicat Northcrest v Amselem 2004 SCC 47. See discussion in n 28 below. 304 A response to Lenta

needed social services that government is unwilling or unable to provide in a cost-efficient and humane manner.24

Religious organisations and their special character (including their internal rules, employment contracts etc.) would be ignored or threatened were religious or cultural accommodation limited to the protection of only individual practices or belief sets or if courts adopt too ready a willingness to scrutinise internal arrangements. Moreover, the undeniable public benefits of these privately funded institutions are what make these projects of interest to the ‘public’.25 Yet what is clearly understood in charities law gets overlooked all too frequently in constitutional dicta. Consider the errant approach of the Supreme Court of Canada in the following passage:

While some provisions in the Constitution involve groups, such as section 93 of the Constitution Act 1867 protecting denominational schools and section 25 of the Charter referring to existing aboriginal rights, the remaining rights and freedoms are individual rights; they are not concerned with the group as distinct from its members. The group or organisation is simply a device adopted by individuals to achieve a fuller realisation of individual rights and aspirations. People by merely combining together, cannot create an entity which has greater constitutional rights and freedoms than they, as individuals, possess.26

This position — a standard liberal individualistic approach to communities — is both epistemologically incorrect and politically unsound: it gets the ontological priority of meaning back to front. Individuals only come to participate in these institutions because a community has — in most instances — already produced them and the meaning that flows from their ongoing reproduction. It is because communities exist that individuals have anything to choose at all. Indeed, sections 15, 30 and 31 of the South African Constitution expressly recognise the communal nature of these religious and cultural institutions. When the Constitutional Court is ultimately faced, again, with questions of religious and cultural accommodation it would do well to recall that these charitable institutions are not run by an individual but built and supported over time by identifiable communities. The express legal recognition of the ‘public benefits’ of many religious and cultural charities points us in the right direction.

A real danger — and not a mere oversight — exists when the ‘individualising tendency’ of one strand of liberalism is ascendant.

24 R v Gruenke [1991] 3 SCR 263 299-300 per L’Heureux-Dubé J for herself and Gonthier J concurring with the majority (emphasis added). 25 On the ‘public benefit’ of religious charities, see H Picard (ed) The law and practice relating to charities (3rd ed 1999) 84-116; J Warburton (ed) Tudor on charities (9th ed 2003) 78-81 and DWM Waters (ed) Waters' law of trusts in Canada (3d ed 2005) 716-745. 26 Public Service Employee Relations Act (Alberta) Ref [1987] 1 SCR 313. (2008) 1 Constitutional Court Review 305

The individualising tendency, I would argue, poses a genuine threat to the proper recognition of religious liberties. Holding the individual and communitarian dimensions together in theory and in practice is essential for the very social stability that the Constitutional Court in Christian Education refers to above. As William Galston notes:

In some measure, religion and liberal policies need each other. Religion can undergird key liberal values and practices; liberal politics can protect — and substantially accommodate — the free exercise of religion. But this relationship of mutual support dissolves if the respective proponents lose touch with what unites them. Pushed to the limit, the juridical principles and practices of a liberal society tend inevitably to corrode moralities that rest either on traditional forms of social organisation or on the stern requirements of revealed religion ... [L]iberal theorists (and activists) who deny the very existence of legitimate public involvement in matters such as family stability, moral education, and religion are unwittingly undermining the values and institutions they seek to support.27

Yet once one recognises the importance religious and cultural rights have for a society, we still need to define what count as religions and cultures and ask what are the rules or requirements of entrance, exit and membership (to name but three) for a ‘culture?’28 Is something worthy of constitutional protection simply because I say or believe it is? If, like religion, protection and recognition extend to those things that are voluntarily chosen as well as obligatory (and what would be the obligatory aspect of ‘a culture of one’ except mere assertion of the will?), should we be concerned that the important category of culture could be thinned out to nothingness, deflated to the point of implosion, by lack of meaningful definition? If this concern is accurate, as I think it is, then too broad a conception of culture can have the paradoxical result that in trying to give content to the

27 W Galston Liberal purposes: Goods, virtues, and diversity in the liberal state (1991) 279. 28 On entrance, membership and exit see Woolman ‘Freedom of association’ (note 21 above). See also A Hischmann Exit, voice and loyalty (1970); E Brody ‘Entrance, voice and exit: The constitutional bounds of freedom of association’ (2002) 35 UC Davis Law Review 821. Though not dealing with associational rights and focused more on some of the other tensions common to Pillay and Amselem, see a helpful exchange between two Canadian scholars: R Moon ‘Religious identity and commitment: Syndicat Northcrest v Amselem’ (2005) 29 Supreme Court Law Review (2d) 201 and DM Brown ‘Neutrality or privilege?: A comment on religious freedom’ (2005) 29 Supreme Court Law Review (2d) 221. 306 A response to Lenta concept we actually denude it.29

I have already contended that turning ‘religion’ into a merely individually assessed matter can have the damaging effect of reducing and weakening the category. Benjamin Berger has observed that one of the most profound implications of the relationship between religious commitment and assessing a contemporary liberal order is that:

... there is a fundamental, though eminently explicable, shortfall at the core of liberal legal discourse about religious liberties. Religion is not only what law imagines it to be. Law is blind to critical aspects of religion as culture. That being so, even if successful at accommodating or tolerating what it understands to be religion, aspects of religion as culture remain entirely unattended to and, therefore, unresolved in their tension with the constitutional rule of law. And with this insight we come to one important part of the explanation for why the story we tell about law and religion has proven so unsatisfactory: law — in whose

29 Thus Pillay could be said to give either insufficient attention to the relationship between religion and culture and communities (Chief Justice Langa’s judgment) or too much attention to religion and culture as individual practices (as in Justice O’Regan’s reasons). A richer examination of the communal prerequisites for religious practice can be found in Taylor v Kurtstag No and Others 2005 1 SA 362 (W). Here, in Judge Malan’s reasons for his judgment, the rules of entrance, membership, exit and excommunication in the context of Orthodox Judaism are taken seriously by the Court. The Canadian Supreme Court has, to date, focused more on the individual side of the religious rights ledger. See Amselem (n 23 above). In this case the Supreme Court of Canada’s majority stated that religion is about ‘self-definition and spiritual fulfillment’ (para 39). Amselem collapses the individual and communitarian dimensions of religious practice. Amselem focused on what might be termed a ‘fringe religious belief’ and held expressly that it was not necessary to show the practice in question was required by the religion. A more nuanced reading of the judgement, however, shows that since the concern about delving too far into religious beliefs would involve the courts in an assessment of ‘dogma’ and that would constitute ‘unwarranted intrusions into the religious affairs of the synagogues, churches, mosques, temples and religious facilities of the nation with value-judgment indictments of those beliefs that may be unconventional or not mainstream’ (para 55) the court, as in Pillay, gave respect to the communitarian dimension but in a round-about and implicit manner which should have been addressed expressly.. So while the communitarian aspects of Charter protection were insufficiently elucidated, they clearly informed the decision. See also K Boonstra & I Benson ‘Religion is in the eye of the beholder’ Lex View No 64 [comment on Amselem] available at http:// www.culturalrenewal.ca/qry/page.taf?id=166 (accessed February 23, 2009). (2008) 1 Constitutional Court Review 307

capacity to tolerate, accommodate, and ‘make space’ for cultural claims we place so much faith — fails to appreciate religion as culture. 30

Pillay and Amselem underscore the need for courts to make express the proposition that religion acquires meaning through communities and that religion matters to individuals precisely because they are (usually) born into communities that give their (religious) lives meaning. Belief, in the context of religion, cannot be rightly understood as an individual dispositional state. The communal character of religious believers poses strong, but not ultimately incompatible, challenges to liberal theories grounded in predominantly individualistic notions of autonomy. Liberal judges ought, in fact, to recognise that communal religious freedom is what enables many individuals in our society — religious and not — to flourish.

4 Are belief and opinion to be analysed in the same way as religion and culture?

What becomes of the analysis of accommodation if a student’s objection to wearing a uniform or part of a uniform is based on conscience, belief or opinion simpliciter? Since a term such as ‘belief’ is also a protected category, what analytical tools are we to bring to bear on assessing the importance and bona fides of a belief claim that some aspect of uniform wearing is offensive to a student? Are religious or cultural beliefs elevated forms of belief in terms of constitutional protection? Should religion and culture be considered more weighty rights than idiosyncratic or individualistic personal beliefs? Given what I and others have argued about the ‘public benefits’ of religious charities and projects generally, the answer should be ‘yes.’

Lenta mentions the Scottishness of kilt wearing as a cultural practice that might have to be accommodated in light of the reasoning in Pillay. One wonders how and whether a belief sheltered by, say, iconoclastic or eccentric expressions of individual will, but nonetheless a ‘belief’, would be analysed? What weight is placed on

30 B Berger ‘Law’s religion: Rendering culture’ in R Moon (ed) Law and religious pluralism in Canada (2008) 264 288 available in an earlier version of the article at http://ohlj.ca/english/documents/45-2_02_Berger_postFR2_July10.pdf (accessed 20 September 2008). David Brown’s critique of a certain overly expansive view of law, see n 28, above, at 230 - 233 would apply to Berger’s concept that law somehow constitutes a ‘culture’ rather than, as another view would have it, a set of rules that must order relationships between cultures. Law understood as a ‘culture’ or ‘community’ in its own right would seem to call into question the ability of the law to adjudicate in a neutral fashion between competing claims amongst sub-cultures and communities. The laws are the equal property of all and owned especially by none; if this is so then it is probably wiser to avoid understanding constitutional law as constituting a ‘culture’ or a ‘community’ given how important constitutional principles are to citizenship shared by all. 308 A response to Lenta the balancing scales where uniforms are weighed not against religious or cultural beliefs and membership (which the cases show can marshal significant social arguments for their respect) but against what may seem merely idiosyncratic whims or expressions of pique, adolescent rebellion or ego? Say, a clerk at the Constitutional Court who sports a Mohawk and goes barefoot with or without wearing a kilt. If the test here is largely subjective (as it would appear to be with both religion and now culture after Pillay), why should we treat the non-religious, non-cultural individual claim for respect differently than claims that are at bottom communal, but which the Court treats as individual forms of expressive conduct? Examining the associational dimension and the goods consequent upon such associations might provide some way around this difficulty.

Courts — and commentators — have tried to avoid rank-ordering rights.31 Is this agnosticism about which rights are more weighty than others sustainable? In the most recent decision from the Canadian Supreme Court touching upon the freedom of religion in relation to civil laws, the majority judgment began by affirming that multiculturalism and pluralism must be protected, but then goes on to employ a fact-specific application of Canadian Charter rights that denies ‘bright-line application but does employ rank-ordered choosing without naming it as such’:

Canada rightly prides itself on its evolutionary tolerance for diversity and pluralism. This journey has included a growing appreciation for multiculturalism, including the recognition that ethnic, religious or cultural differences will be acknowledged and respected. Endorsed in legal instruments ranging from the statutory protections found in human rights codes to their constitutional enshrinement in the Canadian Charter of Rights and Freedoms, the right to integrate into Canada’s mainstream based on and notwithstanding these differences has become a defining part of our national character. The right to have differences protected, however, does not mean that those differences are always hegemonic. Not all differences are compatible with Canada’s

31 Dagenais v Canadian Broadcasting Corp [1994] 3 SCR 835 para 72. Chief Justice Lamer stated that ‘[w]hen the protected rights of two individuals come into conflict ... Charter principles require a balance to be achieved that fully respects the importance of both sets of rights.’ See also Trinity Western University v British Columbia College of Teachers [2001] 1 SCR 772 para 29. Justice Iacobbucci wrote that ‘[n]either freedom of religion nor the guarantee against discrimination based on sexual orientation is absolute.’ For a useful critique of balancing in hard cases and where matters of principle are at stake, see S Woolman & H Botha ‘Limitations’ in Woolman et al (eds) (n 10 above) ch 34 citing New Jersey v TLO (1985) 469 US 325 369 (Brennan J describes balancing as ‘doctrinally destructive nihilism’). See also the Supreme Court of Canada’s decision allowing a student to wear a kirpan (Sikh ceremonial dagger worn for religious reasons) in a public school as long as it was sealed within his clothing; Multani v Commission Scolaire Marguerite Bourgois [2006] 1 S.C.R. 256 and case comment P Lauwers and IT Benson ‘Allowing Kirpans in Public Schools’ Lex View No. 57 http:// www.culturalrenewal.ca/qry/page.taf?id=119 (last accessed February 20, 2009). (2008) 1 Constitutional Court Review 309

fundamental values and, accordingly, not all barriers to their expression are arbitrary. Determining when the assertion of a right based on difference must yield to a more pressing public interest is a complex, nuanced, fact-specific exercise that defies bright-line application. It is, at the same time, a delicate necessity for protecting the evolutionary integrity of both multi-culturalism and public confidence in its importance.32

The French administrative court’s decision that the wearing of a niqab (concealing the face of an individual except for the eyes) is a ground for denial of citizenship is, itself, grounded in the underlying notion that such concealment is ‘inconsistent with Republican values’. Perhaps that is true for France.33 But what should religiously inclusive jurisdictions such as South Africa and Canada do in the face of such a dispute and what are the grounds for deciding matters differently? The relatively empty appeal to ‘balancing’ articulated above is of little assistance here.

In his earlier writings, Lenta hints, without giving a definitive viewpoint on the matter, that a niqab might be distinguished from a headscarf on the basis that in certain contexts facial covering could lead to ‘making identification and communication difficult’. He thereby avoids the religiously inclusive requirement to accommodate such a face-covering in public settings where such identification is essential or important.34 I think Lenta is correct — if I have understood him properly — and look forward to his further analysis on this point. The debate in Canada regarding whether Muslim women should be allowed to vote while wearing face-coverings suggests that the issue may find practical means of resolution which allow for accommodation.35

To what extent headscarves or face-coverings should or should not be accommodated in public settings such as public school classrooms

32 Bruker v Marcovitz 2007 SCC 54 paras 1-2 per Abella J. 33 K Bennhold ‘A veil closes France’s door to citizenship’ The New York Times 19 July 2008 available at http://www.nytimes.com/2008/07/19/world/europe/19france. html?_r=1&pagewanted=1&ei=5070&en=2b01d3cac5c8b204&ex=1217304000&emc =eta1&oref=slogin (accessed 22 July 2008). 34 See Lenta ‘Headscarves’ (n 2 above) 319. 35 A Canadian report has noted that there are very few women who wear full head coverings that obscure the face. One estimate puts the number in Quebec at 50 Muslim women out of a total number of around 200 000. According to this report, Muslim women in Canada are used to showing their faces for identification purposes at banks, airports and at the United States border. There has been some debate around the question of whether full facial coverings should be allowed at election booths and how identity can be assured. One commonly suggested solution has been for election officials to make provision for Muslim women to uncover to women election officers for the purpose of establishing identity. Another solution is to bring with them an adult who can swear to their identity. See http://www.cbc.ca/canada/story/2007/09/09/harper-veil.html (accessed 19 November 2008). 310 A response to Lenta or government offices must be based upon nuanced and careful line- drawing. I do not believe that in a Canadian or South African context a blanket gendered approach or ‘Republican values’ conclusion of the sort that the French have embraced is likely to be the path ahead. A proper understanding of the principles of accommodation and the religiously inclusive public sphere of the kind adopted by the Canadian Supreme Court in Chamberlain and implicitly endorsed in Pillay and Fourie, suggests that civil function issues (such as voting identification) may be creatively accommodated.

What is certain, however, is that the courts have yet to address sufficiently the communal nature of religious belief and the importance that this communal nature (legally comprehended as associational rights) plays in serving the goods that religious beliefs can offer to free and democratic societies. An individualistic rights approach will tend to dissolve this good for the public and, ironically, for individuals as well.36

36 Thus D Schneiderman ‘Associational rights, religion and the Charter’ in R Moon (ed) (n 30 above) 65 notes that: ‘ ... pluralists will call upon state actors to take care that they do not impair associational rights more than is necessary. Courts have often not been so careful. They have assumed a unity of purpose between state and society that should not so readily be presumed’ (80). Compare in the same volume (239) L Weinrib ‘Ontario’s Sharia law debate: Law and politics under the Charter’, who focuses on the individual aspect of rights and asserts: ‘These [Charter] provisions make clear that the relationship between the individual and the state is primary and direct: that is, undiminished by personal characteristics and unmediated by given or chosen social affiliations’ (247) (emphasis added). To argue, as does Weinrib, that relationships between the individual and his or her association (church, organisation etc) are not or cannot be, in constitutionally significant ways, ‘mediated’ by what a person has chosen (ie contractually agreeing to Membership Rules or a Code of Conduct etc) overstates the importance that individual rights may have in a constitutional order and contributes to the sort of threats to associational rights I have criticised throughout this article. The volume in which these papers appears shows this tension between approaches and the emerging awareness of associational rights as against what might be termed the earlier but incomplete focus upon more individualistic approaches. It seems clear the law can and should move towards greater associational rights recognition. MEDIA FREEDOM AND THE LAW OF PRIVACY: NM AND OTHERS V SMITH AND OTHERS (FREEDOM OF EXPRESSION INSTITUTE AS AMICUS CURIAE) 2007 (5) SA 250 (CC)

Glenn Penfold & Dario Milo*

1 Introduction

Patricia de Lille, a biography of the well-known Member of Parliament, was published in March 2002.1 The book included a chapter on Ms De Lille’s work in campaigning for the rights of persons living with HIV/AIDS. In a chapter recounting Ms De Lille’s involvement in a dispute relating to a controversial HIV drugs trial at Kalafong Hospital,2 the names of three trial participants, and their HIV-positive status, were published without their consent. After the reference to their names was drawn to their attention, the three women instituted action against the author (Charlene Smith), Ms De Lille and the publisher for breach of their rights to privacy, dignity and psychological integrity. The plaintiffs alleged that the defendants intentionally published their HIV status without the plaintiffs’ consent or, alternatively, that they acted negligently in doing so. While the former did not require an extension of the actio injuriarum (the delictual action for infringement of personality rights), the latter did.

The defendants argued that the women’s HIV status was not private at the time of the book’s publication, in that their status had been placed in the public domain. They also denied that the publication was either intentional or negligent.

* Directors, Webber Wentzel; Lecturers, University of the Witwatersrand School of Law. 1 C Smith Patricia de Lille (2002). 2 The clinical trial was conducted by the head of the Immunology Clinic of the Medical Faculty of the University of Pretoria, and was aimed at determining the efficacy of a combination of drugs on patients’ HIV levels.

311 312 Media freedom and the law of privacy

After a trial that lasted eight days in the High Court, Schwartzman J held that the defendants lacked both intention and negligence in respect of the publication of the plaintiffs’ HIV status.3 With regard to the latter, Schwartzman J found that the defendants reasonably believed, at the time of publication, that the plaintiffs had consented to the disclosure of their HIV status.4 The High Court, however, held the publisher liable for continuing to distribute copies of the book after it became aware that the women had not in fact consented to the general disclosure of their HIV status.5 As a result, the publisher was ordered to pay R15 000 in damages to each plaintiff. The plaintiffs appealed against the High Court’s exoneration of the defendants in respect of the initial publication of the book.

After the Supreme Court of Appeal denied leave to appeal, the matter came before the Constitutional Court, which (by nine judges to one) held the defendants liable for the initial publication of the plaintiffs’ HIV status.6 The judges were further divided as to how they came to this finding. The majority, in a judgment penned by Madala J,7 did not deviate from the conventional approach to the actio injuriarum, holding that, contrary to the factual finding of the High Court, the defendants intentionally violated the plaintiffs’ privacy. Madala J was satisfied that the defendants ‘were certainly aware that the [plaintiffs] had not given their consent or at least foresaw the possibility that the consent had not been given to the disclosure’.8 This rendered it unnecessary for the majority to consider whether the actio injuriarum should be extended to cover harm brought about by negligence (as opposed to the traditional mental element of intention).9

The remaining three judges, in separate judgments, took the legal road less travelled, effectively holding that liability for breach of privacy should be extended to include liability for media defendants based on negligence. On the facts, Langa CJ and Sachs J found that the ‘media defendants’ (ie the author and the publisher) fell short of

3 NM and Others v Smith and Others [2005] 3 All SA 457 (W) (NM High Court) paras 40-41. 4 The principal reasons for this were that the women had been identified by name as being HIV-positive in a report on various allegations relating to the manner in which the clinical trials were conducted, prepared by an outside expert, Prof SA Strauss, for the University of Pretoria (‘The Strauss Report’) (para 40 of the High Court judgment). The Strauss Report did not suggest that the contents of the Report were confidential (para 40.2). 5 Para 44.2 of the High Court judgment. 6 NM and Others v Smith and Others (Freedom of Expression Institute as Amicus Curiae) 2007 5 SA 250 (CC) (NM or NM v Smith). 7 In which six judges, Moseneke DCJ, Mokgoro, Nkabinde, Skweyiya, Van der Westhuizen and Yacoob JJ, concurred. 8 NM (n 6 above) para 64. 9 NM (n 6 above) para 57. (2008) 1 Constitutional Court Review 313 the standard of negligence (or reasonableness),10 while O’Regan J held that they did not.11

Against the background of this variety of factual findings by the Court, and to some extent because of them, the judgment in NM v Smith raises various issues of significant interest, some of which are discussed in this note.12 The first issue that interests us is the majority’s cursory treatment of the ‘constitutional matter’ requirement (the threshold requirement for the Constitutional Court’s jurisdiction) in a judgment that proceeds to decide the factual question of the existence of intention for purposes of the actio injuriarum. This is a ‘garden-variety judicial task’13 that one would expect ordinary appellate courts to undertake, but which is not generally the domain of a specialist court whose jurisdiction is confined to ‘constitutional matters’. The second, related issue that we briefly deal with is the questionable merits of the majority’s factual finding that the defendants acted intentionally. This is followed by a discussion of the most significant legal development that arises from this case: the approach of the minority judgments to the role of negligence and unreasonableness in an action for breach of privacy. The issues that arise in this context include the distinction between media and non-media defendants, the transferability of the negligence standard that applies in defamation law to privacy law, and the application of the negligence test endorsed by the minority judges to the facts of the case.

2 The ‘constitutional matter’ requirement

The Constitutional Court is a specialist court that does not have unlimited jurisdiction. While section 167(3)(a) of the Constitution proclaims that the Constitutional Court is the highest court on all ‘constitutional matters’, the next paragraph stipulates that the Court ‘may decide only constitutional matters, and issues connected with

10 NM (n 6 above) paras 111 & 207. Although Sachs J relies on a case dealing with media liability (the decision of the Supreme Court of Appeal in National Media Limited and Others v Bogoshi 1998 4 SA 1196 (SCA) (Bogoshi)) for this approach, it is unclear whether he limits liability based on reasonableness to the media defendants. For example, he states that he supports the ‘reasons and conclusions’ in Madala J’s judgment and finds that Ms De Lille (who, on the approach of Langa CJ and O’Regan J, is not a media defendant) did not meet the standard of reasonableness (paras 202 & 207). 11 NM (n 6 above) para 189. 12 Other issues that the case addresses but which are outside of the scope of this note include the Court's approach to the privacy of a person's HIV/AIDS status, the relevance of dignity and reputation claims in this context (especially given the stigma attached to HIV/AIDS in society) and the proper approach to the issue of costs. 13 This phrase is taken from FI Michelman ‘The rule of law, legality and the supremacy of the Constitution’ in S Woolman et al (eds) Constitutional law of South Africa (2nd Edition, OS, 2005) (Michelman ‘The rule of law’) 11-10. 314 Media freedom and the law of privacy decisions on constitutional matters’.14 The threshold of a ‘constitutional matter’ (and connected issues) thus determines the line beyond which the Constitutional Court’s authority does not extend. While the Constitutional Court is the highest court on constitutional matters, the Supreme Court of Appeal has the last word on non-constitutional matters.15

A ‘constitutional matter’ is defined in section 167(7) as including ‘any issue involving the interpretation, protection or enforcement of the Constitution’. The Constitutional Court has clarified this definition by identifying certain categories of constitutional matters.16 The category that most commonly applies when one is dealing with the common law (like the actio injuriarum) is the obligation, under section 39(2) of the Constitution, to develop the common law in light of the spirit, purport and objects of the Bill of Rights.17 Nevertheless, much uncertainty remains as to the meaning of a ‘constitutional matter’.18

Given the fact that classification of an issue as a ‘constitutional matter’ determines whether the Constitutional Court is lawfully empowered to deal with it, and given the lack of clarity as to the meaning of this requirement, it is both surprising and disappointing

14 Sec 167(3)(b). 15 Sec 168(3). 16 Frazer v ABSA Bank Limited (National Director of Public Prosecutions as amicus curiae) 2007 3 484 (CC) (Frazer v ABSA) para 38; Boesak v S 2001 1 SA 912 (CC) (Boesak) para 14. See also Michelman ‘The rule of law’ (n 13 above) 11-7 - 11-8. 17 See, for example, Khumalo and others v Holomisa 2002 5 SA 401 (CC) (Khumalo v Holomisa); Carmichele v Minister of Safety and Security and Another (Centre for Applied Legal Studies intervening) 2001 4 SA 938 (CC); and K v Minister of Safety and Security 2005 6 SA 419 (CC) (K v Minister of Safety and Security). 18 Part of the reason for this uncertainty is the breadth of our Constitution. As Van der Westhuizen J remarked in Frazer v ABSA (n 16 above) para 36: ‘[p]hilosophically and conceptually it is difficult to conceive of any legal issue that it not a constitutional matter within a system of constitutional supremacy’. See also Ngcobo J in Van der Walt v Metcash Trading Limited 2002 4 SA 317 (CC) (Metcash) para 32; and C Lewis ‘Reaching the pinnacle: principle, policies and people for a single apex Court in South Africa’ (2005) 21 South African Journal on Human Rights 509 519. Michelman ‘The rule of law’ (n 13 above) 11.2(b) suggests that the Constitutional Court’s acceptance of the legality principle (ie the exercise of all public power must be in accordance with law) logically means that all appeals are constitutional matters. This is because, as Michelman points out, a court’s decision amounts to the exercise of public power and must therefore be correct as a matter of law. Michelman notes that this is apparently in conflict with the Court’s endorsement of its ‘less-than-plenary subject-matter competence’ (11-12). Although engagement with Michelman’s thought-provoking analysis falls beyond the scope of this case note, we note that it may be based on an overly broad conception of the principle of legality (an argument that is apparently acknowledged by Michelman at 11-14). In any event, this is not the approach taken by the Constitutional Court, which has remarked that ‘[a] contention that a lower Court reached an incorrect decision is not, without more, a constitutional matter’ (Frazer v ABSA (n 16 above) para 40). See also the cases cited below at ns 24, 26 & 27. (2008) 1 Constitutional Court Review 315 that the majority’s judgment in NM v Smith contains only three short paragraphs under the heading ‘Is this a constitutional issue?’:

The applicants approached this Court with a view to vindicate their constitutional rights to privacy, dignity and psychological integrity which, they allege, have been violated by the respondents. Their claim is, however, based on the actio iniuriarum and, therefore, falls to be determined in terms of the actio iniuriarum. It is important to recognise that even if a case does raise a constitutional matter, the assessment of whether the case should be heard by this Court rests instead on the additional requirement that access to this Court must be in the interests of justice and not every matter will raise a constitutional issue worthy of attention. The dispute before us is clearly worthy of constitutional adjudication and it is in the interests of justice that the matter be heard by this Court since it involves a nuanced and sensitive approach to balancing the interest of the media, in advocating freedom of expression, privacy and dignity of the applicants irrespective of whether it is based on the constitutional law or the common law. The Court is in any event mandated to develop and interpret the common law if necessary.19

It is apparent that only the first paragraph quoted above deals with the mandatory requirement of a constitutional matter. The remaining paragraphs deal with whether granting leave to appeal is ‘in the interests of justice’.20 All that one gleans from this first paragraph is that the aim of the plaintiffs was to vindicate their constitutional rights through the actio injuriarum. Nevertheless, as Woolman points out, the desire of litigants cannot be decisive of whether a constitutional matter arises.21 The third paragraph perhaps provides an indication of the majority’s thinking on this issue, stating that the case ‘involves a nuanced and sensitive approach to balancing the interests of the media, in advocating freedom of expression, privacy and dignity of the applicants’. While this may be true of other portions of the judgment,22 it is difficult to envisage how this balancing comes into play in the most significant aspect of the

19 NM (n 6 above) paras 29-31. 20 While the existence of a constitutional matter is a necessary requirement for the Constitutional Court to entertain a matter, it is not in itself sufficient. The Court has a discretion as to whether or not to hear a particular constitutional matter. This discretion is exercised on the basis of the ‘interests of justice’, taking into account a range of factors. See, for example, Boesak (n 16 above) para 12. In relation to direct appeals to the Constitutional Court, the ‘interests of justice’ test is expressly included in sec 167(6) of the Constitution. 21 S Woolman ‘The amazing, vanishing Bill of Rights’ (2007) 124 South African Law Journal 762 782. See also Frazer v ABSA (n 16 above) para 40. 22 For example, the majority’s reasoning on the quantum of damages (which balances various constitutional rights and values) or, more clearly, the minority’s extension of the actio injuriarum to negligence in relation to media defendants. The latter is a clear constitutional matter, in that the judges sought to develop the common law in light of the values of the Constitution (as contemplated in sec 39(2) of the Constitution). 316 Media freedom and the law of privacy majority’s judgment, namely, the assessment of intention. This is a subjective assessment of the question whether the defendants knew that they were disclosing private information without the plaintiffs’ consent or actually foresaw that this may be the case and proceeded recklessly to disclose the information. It seems to us that this is purely a factual enquiry which leaves no room for balancing of interests.

The paucity of the majority’s explanation as to why the issues in its judgment amount to ‘constitutional matters’ is even more surprising given the fact that, irrespective of whether it was right or wrong to exercise jurisdiction on this matter, it is difficult to reconcile the Court’s consideration of this matter with its previous jurisprudence on the meaning of ‘constitutional matters’.

Earlier decisions of the Constitutional Court emphasise that, while the jurisdiction of the Constitutional Court is extensive and the term ‘constitutional matter’ should be given a wide meaning,23 it is not all- embracing and some proper meaning must be given to it.24 Most strikingly for our purposes, some of the Court’s decisions prior to NM v Smith indicate that purely factual disputes do not amount to constitutional matters.25 We briefly discuss two of them.

The first is Boesak, in which the applicant alleged that his constitutional rights to be presumed innocent and to freedom and security had been infringed by a decision of the Supreme Court of Appeal upholding fraud and theft convictions against him. According to the applicant, the Court erred in finding him guilty beyond reasonable doubt. During the course of his judgment, on behalf of a unanimous Constitutional Court, Langa DP identified certain broad principles that apply to the classification of criminal cases, including that a challenge to a decision of the SCA on the basis only that it is wrong on the facts is not a constitutional matter; and ‘[u]nless there is some separate constitutional issue ... no constitutional right is engaged when an applicant merely disputes the findings of fact made

23 S v Basson 2005 1 SA 171 (CC) (Basson) para 90-1; and Boesak (n 16 above) para 14. See also the later decision of the Constitutional Court in Frazer v ABSA (n 16 above) para 37 & 39. 24 See Basson (n 23 above) para 91; Metcash (n 18 above) para 32; and Dikoko v Mokhatla 2006 6 SA 235 (CC) para 123. Van der Westhuizen J, on behalf of a unanimous Court in the more recent decision in Frazer v ABSA (n 16 above) para 39 stated as follows: ‘While the conception of a constitutional matter is broad, the term is of course not completely open. The fact that sec 167(3)(b) of the Constitution limits this Court’s jurisdiction to constitutional matters presupposes that a meaningful line must be drawn between constitutional and nonconstitutional matters and it is the responsibility of this Court to do so’. See also Wary Holdings (Pty) Ltd v Stalwo (Pty) Ltd and another, Trustees of the Hoogekraal Highlands and another v Minister of Agriculture and Land Affairs [2008] ZACC 12 para 39. 25 I Currie & J De Waal The Bill of Rights handbook (5th ed, 2005) 104 state that: ‘[m]atters that turn purely on questions of fact are not constitutional matters’. (2008) 1 Constitutional Court Review 317 by the SCA’.26 The second decision is Phoebus Apollo Aviation CC v Minister of Safety and Security,27 in which the Court held that it did not have jurisdiction in relation to a dispute about whether the State should, on the facts of the particular case, be held vicariously liable for the dishonest acts of three off-duty police officers. Kriegler J stated as follows:

It is not suggested that in determining the question of vicarious liability the SCA applied any principle which is inconsistent with the Constitution. Nor is there any suggestion that any such principle needs to be adapted or evolved to bring it into harmony with the spirit, purport or objects of the Bill of Rights. On the contrary, counsel for the appellant expressly conceded that the common-law test for vicarious liability, as it stands, is consistent with the Constitution. It has long been accepted that the application of this test to the facts of a particular case is not a question of law but one of fact, pure and simple. The thrust of the argument presented on behalf of the appellant was essentially that though the SCA has set the correct test, it had applied that test incorrectly — which is of course not ordinarily a constitutional issue.28

It may be that these cases are not, on a proper reading, authority for a general principle that a dispute of fact can never be a constitutional matter, particularly if one has regard to the fact that in both cases the Court found that no constitutional matter arose only after stating that no fundamental right was implicated by the factual dispute. In Boesak, for example, Langa DP remarked that upholding a conviction when the SCA ought to have had reasonable doubt as to the applicant’s guilt ‘is no violation of the applicant’s right to be presumed innocent’ before he stated that the appeal did not involve a constitutional matter.29 Similarly, in Phoebus Apollo the Court found that the right to property clearly did not apply, before stating that no constitutional matter arose.30

These cases may therefore leave room for an argument that a dispute of fact can amount to a constitutional matter where a plausible argument can be made that the matter is relevant to the

26 Boesak (n 16 above) para 15(a). 27 2003 1 BCLR 14 (CC) (Phoebus Apollo). 28 Phoebus Apollo (n 27 above) para 9. 29 n 16 above, para 16. See also para 35. This approach is consistent with the decisions of the Constitutional Court to the effect that an incorrect decision by a court does not infringe a constitutional right. While one has the right to a fair trial (s 35(3) of the Constitution) or a fair public hearing (s 34), one does not have a right to the correct outcome in judicial proceedings. See Lane and Fey NNO v Dabelstein and others 2001 2 SA 1187 (CC) para 4; Metcash (n 18 above) paras 14 & 19; and Basson (n 23 above) paras 91, 94 & 99. 30 n 27 above, paras 4-6 & 9. See also Metcash (n 18 above) para 14. 318 Media freedom and the law of privacy vindication of a constitutional right (whether directly or indirectly).31 This reading of Boesak and Phoebus Apollo finds some support in the following dictum of Chaskalson CJ in Basson:

Where no constitutional right is engaged, a challenge to a decision of the SCA or any other court, on the basis only that it is wrong on the facts, is not a constitutional matter. And a dispute that does not impact upon or give effect to an entrenched right or other constitutional provision, will not ordinarily be a constitutional matter.32

There may thus have been scope for distinguishing Boesak and Phoebus Apollo, and arguing that because the claim in NM Smith was a plausible one that was clearly aimed at indirectly vindicating constitutional rights, any factual dispute that is relevant to vindicating those rights (including one as to intention on the part of the defendants) amounts to a constitutional matter. However, the difficulty which such an argument would face is the unanimous decision of the Constitutional Court in K v Minister of Safety and Security. This case involved a claim based on vicarious liability for rape committed by uniformed, on-duty police officers. The claim sought to vindicate a number of the victim’s constitutional rights, including her rights to freedom and security of the person and dignity. The Court, in assessing whether a constitutional matter arose, considered the meaning of the ‘development’ of the common law for purposes of section 39(2) of the Constitution. O’Regan J, writing for a unanimous Court, noted that the common law is most clearly developed when a common law rule is changed or a new rule is introduced.33 She pointed out, however, that courts more commonly decide cases within the framework of an existing rule. O’Regan J identified two types of cases where this occurs. The first is where a court merely applies the rule to ‘a set of facts which it is clear fall within the terms of the rule or existing authority’, in which case the rule is not developed.34 The second instance is where a court determines ‘whether a new set of facts falls within or beyond the

31 Michelman ‘The rule of law’ (n 13 above) 11-10 puts it as follows: ‘what the [Constitutional Court] is really telling us in Pheobus Apollo Aviation is that it sometimes will decline to hear argument on a claim of unconstitutionality because of the extreme prima facie implausibility of the claim ...’. An example of the indirect vindication of a constitutional right would be an action, such as that in NM, for breach of privacy (a constitutional right entrenched in sec14 of the Constitution) through the actio iniuriarum. 32 n 23 above, para 91, emphasis added. See also Rail Commuters Action Group and others v Transnet Limited t/a Metrorail and others 2005 2 SA 359 (CC) (Rail Commuters) para 52, where O’Regan J, after quoting from Boesak, stated as follows: ‘This reasoning does not imply that disputes of fact may not be resolved by this Court. It states merely that where the only issue in a criminal appeal is dissatisfaction with the factual findings made by the SCA, and no other constitutional issue is raised, no constitutional right is engaged by such a challenge’. 33 n 17 above, para 16. 34 As above. (2008) 1 Constitutional Court Review 319 scope of an existing rule’ so that the ‘precise ambit of each rule is ... clarified in relation to each new set of facts’.35 This falls within the concept of the development of the common law contemplated in section 39(2), which is aimed at ensuring ‘that our common law is infused with the values of the Constitution’.36

Turning to NM v Smith, it seems to us that a finding as to whether the defendants, as a matter of fact, knew that the information was private or that the plaintiffs had not consented to its disclosure or that the defendants actually foresaw that possibility, does not amount to the development of the common law in the manner contemplated in K v Minister of Safety and Security. It is difficult to see in what way the values of the Constitution are relevant to this enquiry.37 Moreover, if this does amount to the development of the common law, it is difficult to reconcile this with the approach of the Constitutional Court in Minister of Safety and Security v Luiters,38 which was argued after, and decided before, NM v Smith. In this case, dealing with vicarious liability in respect of the actions of an off-duty police officer who allegedly placed himself on duty, the Court unanimously held (with reliance on K v Minister of Safety and Security) that the question as to whether the officer subjectively intended to act within the course and scope of his employment was not a constitutional matter. Langa CJ remarked as follows:

The Minister queried the finding of the Supreme Court of Appeal that Constable Siljeur subjectively intended to act as a policeman at the time of the shooting. A number of reasons were given in support of the Minister’s criticism of this finding by the Supreme Court of Appeal. The thrust of the Minister’s submission, however, was to urge this Court to reconsider the facts as found by the High Court and the Supreme Court of Appeal. This submission does not raise a constitutional issue for, as it was made clear in K, the question whether a police officer has subjectively acted as a police officer is purely factual.

35 As above. 36 K v Minister of Safety and Security (n 17 above) para 17. O’Regan J referred to this as ‘the incremental development of the rule’ in light of the normative framework of the Constitution. 37 As Woolman (n 21 above) 783 comments: ‘the majority in NM acts as a trier of fact in a run-of-the-mill actio iniuriarum matter’. It may be suggested that the approach of the majority in NM falls within the second instance in K v Minister of Safety and Security, ie determining whether a particular set of facts falls within or beyond the ambit of an existing rule, in that the majority determined that the facts fell within the ambit of dolus eventualis in circumstances which would, in the absence of the Constitution, amount only to negligence. For a variation on this argument, see FI Michelman ‘On the uses of interpretive ‘charity’: Some notes on application, avoidance, equality, and objective unconstitutionality from the 2007 term of the Constitutional Court of South Africa’ (2008) 1 Constitutional Court Review (Michelman ‘Interpretive charity’) 30 - 32. For the reasons set out in this article, we do not find this argument persuasive. 38 2007 2 SA 106 (CC) (Luiters). 320 Media freedom and the law of privacy

The Constitutional Court’s own case law therefore suggests that a good argument can be made that the factual question as to the defendants’ state of mind in NM v Smith was not a constitutional matter, notwithstanding the fact that the plaintiffs were seeking to vindicate their constitutional rights.

Elsewhere in this volume, Michelman, adopting what he describes as a ‘charitable’ reading of the majority judgment in NM v Smith, suggests that the majority may have developed the common law ‘in response to constitutional pressure’ and that all of the issues in this case

stood to be decided in the shadow of the Bill of Rights and of the looming possibility that the common law might, depending on how they were to be decided, have to be developed under the mandate of section 39(2).39

Michelman goes on to assert that the majority, in finding that the conduct of the defendants amounted to intention, may be said to have been considering whether the common law rule applies to the new factual situation as envisaged in K v Minister of Safety and Security.40 While Michelman’s thoughtful argument merits a more detailed response than this case note permits, it seems to us that there are at least three possible responses. The first is that one cannot get away from the fact that a finding of intention (dolus) involves a finding of subjective state of mind (ie actual knowledge or foresight) on the part of the defendant. This is a purely factual finding which, it seems to us, leaves no space for a consideration of constitutional rights or values. Second, if the majority intended, in light of the Bill of Rights, to extend the scope of dolus eventualis to conduct which would otherwise not have constituted this form of intention, one would expect them to have said so expressly. On the contrary, the approach of the majority was simply to apply the dolus eventualis test to the facts of the case before them.41 Third, even if one is able to bring the majority’s approach within the wording of O’Regan J’s judgment in K v Minister of Safety and Security, it seems to us more difficult to do so in relation to Luiters.

Irrespective of whether or not the majority was correct in holding that the defendants’ intention (or lack thereof) properly fell within the Court’s jurisdiction, our primary difficulty with the judgment of

39 Michelman ‘Interpretive charity’ (n 37 above) 20; see also 21 and 28. 40 Michelman ‘Interpretive charity’ (n 37 above) 28. 41 As Michelman ‘Interpretive charity’ (n 37 above) points out at 19, Madala J expressly states that, in light of the fact that his judgment ‘is not extending the common-law definition of intention to include negligence in relation to the publication of private medical facts, there will be no “chilling effect” on freedom of expression in South Africa ...’ (para 69). (2008) 1 Constitutional Court Review 321 the majority in NM v Smith is that none of these issues were canvassed in the judgment and no attempt was made to reconcile the majority’s approach with the previous case law. The Court thus dealt with a matter that could conceivably fall beyond its jurisdiction — an outcome that would have serious implications for the rule of law — without providing a meaningful explanation of its reasons for doing so.

Moreover, if the majority’s reason for regarding the defendants’ intention as a constitutional matter is that it was an issue that needed to be decided in order to determine whether the plaintiffs’ constitutional rights should be vindicated, the effect of this is that a wide range of ordinary factual disputes may well occupy the Court’s resources in the future.42 These matters could, for example, include an extensive factual dispute as to whether an allegation in a defamation action is true (which implicates the right to dignity) or a dispute as to whether a person who caused physical harm to another did so with intention or negligence (which implicates the right to security of the person and bodily integrity).

Finally on this issue, we reiterate that the Constitution provides that the Constitutional Court’s jurisdiction extends to ‘issues connected with decisions on constitutional matters’. The Court has held that this phrase should be given a wide interpretation so as to extend the Court’s jurisdiction to matters that ‘stand in a logical relationship’ to constitutional matters.43 It includes ‘any anterior matter [that], logically or otherwise, is capable of throwing light on or affecting the decision by this Court on the primary constitutional matter’.44 It is unfortunate that the Court did not consider whether the question as to the defendants’ intention fell within the scope of this concept.45

42 Subject to the ‘interests of justice’ filter discussed above. 43 Alexcor Limited and another v Richtersveld Community and others 2003 12 BCLR 1301 (CC) para 30. 44 As above. The Court went on to state it more formally: ‘when any factum probandum of a disputed issue is a constitutional matter, then any factum probans, bearing logically on the existence or otherwise of such factum probandum, is itself an issue “connected with a decision on a constitutional matter”.’ See also Rail Commuters (n 32 above) para 52; and Basson (n 23 above) para 22. 45 The judgment in Luiters (n 38 above) suggests that this may well be the case: ‘[i]t was argued on behalf of the Minister that once the Court assumes jurisdiction on one basis, it has the power to alter the findings of both the High Court and the Supreme Court of Appeal on factual issues even if we do not find it necessary to change the test. It may well be that the factual issues are issues that are connected to a decision on a constitutional matter in such circumstances’ (para 29). 322 Media freedom and the law of privacy

3 The majority’s finding of intention

The most extraordinary aspect of the judgment in NM v Smith is the extent to which the Court was divided on the factual question as to the defendants’ state of mind.46 The majority held that the defendants had intended wrongfully to infringe the plaintiffs’ privacy, finding that they ‘knew well of the wrongfulness of their conduct and that the disclosure of private facts was likely to invade the privacy rights of the applicants’.47 The majority’s criticism of the defendants is, at times, strident, stating at one point that Ms Smith

did a half-hearted check [for the consents] but soon became tired of the exercise and so decided to go ahead and produce the book without having obtained the consents of the applicants.48

These are very serious findings for the majority of the Constitutional Court to make in relation to persons (the defendants) who had testified that they honestly assumed that suitable consents had been given and did not consider the possibility that this was not the case,49 and in circumstances where the Court states, ‘[t]hat they are good activists in the field of HIV/AIDS admits of no doubt’.50

Not only are the findings of the majority severe, they are also at odds with the decision of Schwartzman J in the trial court who, having heard several days of evidence at the trial, held that the defendants had lacked both intention and negligence. In relation to the former, he stated that Ms Smith and Ms De Lille had

by their long standing involvement with people infected with HIV, demonstrated that they are two of the most unlikely people to intentionally invade the privacy of a person infected with HIV.51

The findings of the majority are also dramatically at odds with O’Regan J’s dissenting judgment. It is remarkable that the majority finds that the defendants knew that the plaintiffs had not consented

46 See also Woolman (n 21 above) 781; and J Steinberg ‘Generous judgment instills stigma’ in Notes from a fractured country (2007) 221. 47 NM (n 6 above) para 64. 48 NM (n 6 above) para 88. 49 NM (n 6 above) paras 159-168. 50 NM (n 6 above) para 59. 51 NM High Court (n 3 above) para 40. Michelman ‘Interpretive charity’ (n 37 above) states that the majority's refusal to defer to the High Court's finding on the facts is explicable if it viewed the High Court's application of the common law rule of intention as not being compliant with the Bill of Rights (at current page 37: Editors to insert new page number). The difficulty we have with this approach is that it assumes that the majority’s discussion of whether intentional conduct was present on the facts, constitutes a constitutional matter with which the Court ought to have concerned itself. We have argued above that the Court itself does not identify this as a constitutional matter raised by the case. (2008) 1 Constitutional Court Review 323 to the disclosure of their HIV status (or had actually foreseen this possibility and recklessly disregarded it) where O’Regan J finds, on considering the same record, that the reasonable journalist would not have been obliged to make further enquiries in this regard.52

Given this context, one would expect that the majority would have been careful to set out their reasons for differing so dramatically from the factual findings of the trial court and O’Regan J. We agree with Langa CJ that, in light of the defendants’ heartfelt denials of intention, the Court ‘would need a great deal of evidence to find that these activists would intentionally infringe the rights of the very people whom they are committed to protect’.53

The reasons that the majority advances in this regard are, with respect, not convincing. While we do not intend to engage in a detailed factual analysis for purposes of this note, we mention two aspects of concern. The first is that the majority did not engage in any detail with the reasons offered by O’Regan J as to why she found that the defendants did not act with intention or negligence. The second is that a number of the majority’s statements in the portion of its judgment dealing with intention indicate that the defendants did not have intention and were, at best for the plaintiffs, negligent. For example, Madala J states that: ‘[b]oth respondents assumed, without any enquiry, that the information contained in the Strauss Report was not confidential’ and ‘[Ms Smith] assumed that the applicants had consented to the public disclosure of their names and HIV status because the source of the publication came from a reputable institution’.54 The Court also notes that ‘[Ms De Lille] failed to take sufficient steps to ascertain whether the [plaintiffs] had in fact given unlimited consent’.55 It is difficult to reconcile these statements, which suggest that the defendants actually made certain (incorrect) assumptions, and ought to have made further investigations, with a finding of intention, even in the form of dolus eventualis.56

52 NM (n 6 above) paras 185-187. 53 NM (n 6 above) para 93. 54 NM (n 6 above) paras 60-61, emphasis added. 55 NM (n 6 above) para 60. 56 Michelman ‘Interpretative charity’ (n 37 above) suggests that the fact that the inclusion of the names of the three plaintiffs in the book could add no value to the publication (and in particular, despite protestations by Ms Smith to the contrary, certainly was not required to give authenticity to the publication) ‘may have been instrumental in leading Madala J's majority to its conclusion that Smith's choice was “intentional” in the pertinent, legal sense’ (at current page 43: editors to insert new page no). We submit, however, that the utility of including the plaintiffs' names in the book can at most be a factor to consider in determining whether the state of mind of Ms Smith was such that she acted with intention (either in the sense that she knew that she was acting unlawfully or foresaw that she may be doing so and acted in reckless disregard of that foresight). This factor cannot elevate otherwise negligent but non-intentional conduct to the status of dolus. 324 Media freedom and the law of privacy

4 The Court's approach to the development of the common law of privacy

The law of defamation underwent radical reform with regard to media defendants in the Bogoshi case.57 Although the decision has been rightly criticised for blurring the doctrinal boundaries between fault and unlawfulness in the law of delict,58 it now appears to be generally accepted by our courts that the effect of Bogoshi is that media defendants sued for liability in defamation will be entitled to rebut the presumption of unlawfulness by proving that their conduct was in all the circumstances objectively reasonable (a position that is favourable to the media relative to non-media defendants). In addition, if the defence of absence of knowledge of unlawfulnness, and hence lack of fault, is pursued by a media defendant, the media defendant must establish that it was not negligent in this regard (in this respect, the media is in a less favourable position relative to non- media defendants).59

Despite these developments in the law of defamation, until the NM v Smith decision, the common law principles that governed the

57 n 10 above. The reform in fact started at the High Court level with the decision of Cameron J (as he then was) in Holomisa v Argus Newspapers Ltd 1996 2 SA 588 (W) (Holomisa v Argus Newspapers). 58 The Court's discussion of lawfulness in the form of reasonable publication overlaps significantly with its discussion of lack of fault in the form of a reasonable belief that the publication was lawful. For instance, Hefer JA at 1214 stated that ‘the indicated approach is intended to cater for ignorance and mistake at the level of lawfulness; and in a given case negligence on the defendant's part may well be determinative of the legality of the publication’. For discussion of the blurring between the elements of fault and unlawfulness by the Court, see J Burchell Personality rights and freedom of expression: The modern actio injuriarum (1998) 226; JR Midgley ‘Media liability for defamation’ (1999) 116 South African Law Journal 211; D Milo ‘The cabinet minister, the Mail & Guardian, and the report card: The Supreme Court of Appeal's decision in the Mthembi-Mahanyele case’ (2005) 122 South African Law Journal 28 38-39; J Neethling ‘The protection of false defamatory publications by the mass media: Recent developments in South Africa against the background of Australian, New Zealand and English law’ (2007) 40 Comparative and International Law Journal of South Africa 103 123-124. For further discussion on how courts in the area of delict continue to grapple with the distinction between fault and unlawfulness, see eg A Fagan ‘Rethinking wrongfulness in the law of delict’ (2005) 122 South African Law Journal 90; J Neethling ‘The conflation of wrongfulness and negligence: is it always a bad thing for the law of delict?’ (2006) 123 South African Law Journal 204; RW Nugent ‘Yes, it is always a bad thing for the law — a reply to Professor Neethling’ (2006) 123 South African Law Journal 557; J Neethling & JM Potgieter ‘Wrongfulness and negligence in the law of delict: A Babylonian confusion?’ (2007) 70 Tydskrif vir Hedendaagse Romeins-Hollandse Reg 120; FDJ Brand ‘Reflections on wrongfulness in the law of delict’ (2007) 124 South African Law Journal 74; A Fagan ‘Blind Faith: A Response to Professors Neethling and Potgieter’ (2007) 124 South African Law Journal 285. 59 The leading judicial recognition of these principles is the Constitutional Court's unanimous decision in Khumalo v Holomisa (n 17 above) paras 19-20. See also Mthembi-Mahanyele v Mail & Guardian Ltd and Another 2004 6 SA 329 (SCA) (Mthembi-Mahanyele) paras 45-46. (2008) 1 Constitutional Court Review 325 law of privacy — more particularly liability based on intention — had not been challenged under the Constitution.60 The facts of NM v Smith raised squarely the issue of the fault element of liability in privacy cases: assuming that Ms Smith, Ms De Lille and the publisher held the view that consent had been provided to reveal the plaintiffs’ identities and that this view was mistaken, was it necessary for them to prove that such a mistake was reasonable? As stated, the majority of the Court did not reach this issue; it found that the defendants published with knowledge that the plaintiffs had not provided the requisite consent (or at least with reckless disregard as to whether consent had been granted).61 On the other hand, the minority judges grappled with the issue at some length. Their judgments raise a number of issues of profound significance to media liability in privacy cases, and we deal with each in turn: the distinction between media and non-media defendants for purposes of privacy (and defamation) liability and the desirability of a negligence test in this context; and how negligence on the part of media defendants should be assessed where journalists rely on official reports.

5The distinction between media and non-media defendants and the negligence test

In Bogoshi, the Court expressly perpetuated the distinction between media and non-media defendants that was created in Suid-Afrikaanse Uitsaaikorporasie v O'Malley,62 where the Appellate Division opined that strict liability applied in defamation cases to members of the mass media.63 Although the Court in Bogoshi rightly jettisoned the restrictive principle of strict liability, it held that it would not be just to allow media defendants to escape liability on the same basis as non-media defendants, that is, by relying on the absence of animus injuriandi.64 Hefer JA relied for this distinction on the Australian High Court's judgment in Lange v Australian Broadcasting Corporation:65 'the damage that can be done when there are thousands of recipients of a communication is obviously greater than when there are a few

60 A number of academic commentators had, however, expressed the view (albeit without much analysis) that the developments in defamation law had the effect that equivalent principles now applied in the law of privacy. See eg D McQuoid- Mason ‘Privacy’ in S Woolman et al (eds) Constitutional law of South Africa (2nd Edition, OS, 2003) 38-19; Burchell (n 58 above) at 429. 61 See text accompanying n 8 & 9 above. Madala J, for the majority, stated at para 57 that this was not an appropriate case to depart from ‘the age-old approach to the actio injuriarum. I do not, by any means, wish to be understood to say the common law should or could never be developed in this regard’. 62 1977 3 SA 394 (A). 63 O’Malley (n 62 above) 403. This was confirmed by the Appellate Division in Pakendorf en Andere v De Flamingh 1982 3 SA 146 (A). 64 Bogoshi (n 10 above) 1214. 65 (1997) 189 CLR 520 (Lange). 326 Media freedom and the law of privacy recipients';66 this made the 'additional burden' of proving lack of negligence 'entirely reasonable'.67

These principles were developed by the minority in NM v Smith in judgments that are now the leading discussions in South African law with respect to the question of how to distinguish between media and non-media defendants, and why such a distinction is necessary. All the minority judges were clear that, as in defamation law, a higher level of care is required in privacy law for media defendants as opposed to non-media defendants. As O'Regan J stated:

For purposes of this case, I accept that the legal principles developed in Bogoshi should apply not only in the law of defamation but also to the infringement of privacy rights by the media ... Modern electronic, print and broadcast media are immensely, and indeed, increasingly powerful. Publications often reach hundreds of thousands of readers, viewers and listeners. It is accordingly appropriate, given the scale of damage to an individual that can be caused by such widespread publication, to confer special obligations upon the media in respect of publication. In so doing, we recognise that the media are not only bearers of rights under our constitutional order, but also bearers of obligations.68

Given the immense power wielded by the media,69 we submit that, while a negligence test for breaches of privacy by the media clearly limits their rights to media freedom, such a limitation is justifiable. We agree with Sachs J in this context that ‘the principles developed in [the Bogoshi] case are eminently transportable to the law of privacy’,70 at least in the

66 Lange (n 65 above) 572. 67 Bogoshi (n 10 above) 1214. 68 NM (n 6 above) para 177. See also Langa CJ at para 94: 'It is ... constitutionally appropriate that the media should be held to a higher standard than the average person'. 69 For foreign studies in this regard, see generally J Curran & J Seaton Power without responsibility: The press, broadcasting, and new media in Britain (6th ed, 2003). For an interesting South African perspective, see R Calland Anatomy of South Africa: Who holds the power? (2006) 194-204. 70 NM (n 6 above) para 203. See also J Neethling ‘The right to privacy, HIV/AIDS and media defendants’ (2008) 125 South African Law Journal 36 45. For a different view, see H Scott 'Liability for the mass publication of private information in South African law: NM v Smith (Freedom of Expression Institute as Amicus Curiae)' (2007) 18 Stellenbosch Law Review 387 396-398. Scott argues that the analogy between privacy and defamation law in this context is flawed, amongst other things, because in Bogoshi the result was to extend much greater protection to the media than was previously thought to be the case, whereas the impact in relation to breach of privacy is to significantly limit freedom of the media (398). However, we submit that this was only the case because of the unconstitutional imposition of strict liability on the media; the point is that media defendants, post-Bogoshi, are not treated as generously as non-media defendants in the law of defamation. We consider that there is no principled objection to utilising defences that have developed in defamation law in the privacy context, in clearly analogous contexts, as the common law of privacy itself recognises. See eg Jansen van Vuuren and another NNO v Kruger 1993 4 SA 842 (A) 850. (2008) 1 Constitutional Court Review 327 context of fault liability.71 The Constitutional Court in Khumalo v Holomisa stated that the defence of reasonable publication — which is in many respects akin to a negligence test — strikes the correct balance between freedom of expression and reputation and thus limits each right as little as possible. O'Regan J stated in that case for a unanimous court:72

The defence of reasonable publication avoids ... a winner-takes-all result and establishes a proper balance between freedom of expression and the value of human dignity. Moreover, the defence of reasonable publication will encourage editors and journalists to act with due care and respect for the individual interest in human dignity prior to publishing defamatory material, without precluding them from publishing such material when it is reasonable to do so.73

We submit that, in principle, assessing the media’s conduct through the prism of a negligence test achieves a reasonable balance between the competing considerations of privacy and dignity on the one hand, and freedom of the media on the other. On this approach, unless a defence negating unlawfulness (such as consent) applies, the media would have to establish lack of negligence in order to escape liability, such as, on the facts of NM v Smith, that its mistake as to the existence of consent was not negligent.74 Requiring the media to establish lack of negligence does not in our view constitute a disproportionate restriction on media freedom; indeed, the standards expected of reasonable journalists are often reflected in codes of conduct, such as the Press Council’s Code, which media houses

71 Another respect in which the Bogoshi principles seem to us to be capable of application in privacy actions is in the context of the defence of truth and public interest. While the ‘truth’ aspect of this defence is seldom in issue in privacy cases (which typically turn on information that is true but invasive: see eg E Barendt ‘Privacy and the press’ 1995 Yearbook of Media and Entertainment Law 22 26), the truth of an allegation that implicates the privacy of the plaintiff may nevertheless need to be established by the media in order to rely upon the defence of public interest (see also Tshabalala-Msimang and others v Makhanya and others [2008] 1 All SA 509 (W) para 48). Thus a plaintiff is entitled in principle to vindicate her right to privacy and dignity where false but non-defamatory facts are published, where the matter ‘is by its nature such as to attract the law of [privacy]’: Ash v McKennitt and others [2007] 3 WLR 194 (CA) para 80. We submit that the application of Bogoshi to the law of privacy in this context has the result that facts which may not be true but which have been published reasonably, should also benefit from the defence of public interest. See also the approach to defences to breach of confidence actions in England: Attorney-General v Guardian Newspapers Ltd (No 2) [1990] 1 AC 109 at 283. See also the decision of the English High Court in Mosley v News Group Newspapers Ltd [2008] EWHC 1777 (QB) para 143 and what follows (testing the invasion of privacy in issue against the test of whether the journalist acted responsibly based on information then available to him). 72 Khumalo v Holomisa (n 17 above) para 43. 73 Compare also the decision of Cameron J in Holomisa v Argus Newspapers (n 57 above) 617. 74 We submit that there is also no reason in principle why the qualification that a mistake need be reasonable should not extend to mistakes about other elements of the delict, such as whether the information constituted a private fact. 328 Media freedom and the law of privacy themselves espouse,75 and negligence is a concept with which our law is familiar.76 Moreover, a negligence standard represents a reasonable compromise between the extremes of strict liability, on the one hand, and a purely subjective test of intention, on the other. While the former provides no protection to media freedom, the latter may have the deleterious result that privacy concerns are effectively eviscerated.77 A standard based on negligence also finds support in the analogous context of defamation law in such jurisdictions as England78 and Australia,79 and in the jurisprudence of the Privy Council80 and the European Court of Human Rights.81

Accepting that negligence should apply to ‘media defendants’, the difficulty that remains is a definitional one: how should the line be drawn between media and non-media defendants? While traditional print and electronic media entities are easily recognisable, the digital revolution has the result that ‘[l]ines between different forms of media are blurring, making it difficult to apply medium- specific definitions’.82 O’Regan J’s guidance in NM v Smith in regard to the definitional category of media defendants is instructive:

Professional and commercial purveyors of information are well placed to ensure that appropriate systems prevent the unreasonable disclosure of private facts and the negligent disclosure of those facts. This is not the case for ordinary citizens. Moreover, generally, disclosure by ordinary

75 As Lord Hoffmann recently stated in an English defamation case, ‘the standard of responsible journalism is made more specific by the Code of Practice which has been adopted by the newspapers and ratified by the Press Complaints Commission’ (Jameel v Wall Street Journal Europe Sprl [2007] 1 AC 359 (HL) (Jameel) para 55). Expert evidence as to reasonable conduct on the part of journalists could also be led, as it was in the NM, where Professor Anton Harber led evidence as to what is expected of a journalist reporting the identity of persons living with HIV/AIDS. 76 See Sachs J in NM (n 6 above) para 204. 77 We submit that the experience of the US courts with the actual malice standard that applies in defamation and false light privacy cases is instructive in this regard. This rule — first articulated by the US Supreme Court in New York Times v Sullivan 376 US 254 (1964) — has had a disproportionate impact on the rights of the plaintiff and has also adversely affected the quality of public discourse. See generally D Milo Defamation and freedom of speech (2008) 203-206. This was, however, not the outcome on the approach of the majority in NM v Smith, who found intention on the part of the defendants. 78 See Reynolds v Times Newspapers Ltd [2001] 2 AC 127 (HL) at 202, 205. See also Jameel (n 75 above) paras 32, 55, 149. 79 Lange (n 65 above) 573. 80 See eg Bonnick v Morris [2003] 1 AC 300 (PC) paras 22, 23. 81 See eg Bladet Tromso and Stensaas v Norway (2000) 29 EHRR 125 (Bladet) para 65. For a discussion of the respects in which these jurisdictions articulate a negligence test for media liability, see Milo Defamation and freedom of speech (n 77 above) 188-99. 82 DA Anderson ‘Freedom of the press’ (2002) 80 Texas Law Review 429 438. (2008) 1 Constitutional Court Review 329

citizens will not be as widespread as disclosure by those involved in professional or commercial dissemination of information.83

On the basis of this test, the minority judges rightly regarded the publisher and Ms Smith as media defendants, but viewed Ms De Lille as a non-media defendant.84 As we have discussed, our courts in defamation cases have adopted the proposition that the primary rationale for imposing a negligence standard on the media is the increased capacity to cause harm inter alia because of the widespread manner in which allegations may be published. We note that it is not only the traditional media that has the capacity to cause severe harm to a person’s reputation and privacy because of the extent of the publication; the wide circulation of a leaflet containing private information may also have this effect.85 It remains an open question whether the negligence standard should apply beyond the traditional media categories, as is the case in Australian law.86 This will require an assessment of various factors, including the capacity to cause harm; the resources available to the publisher to guard against such harm occurring;87 whether professional codes of conduct — such as the Press Council's Code and the Broadcasting Complaints Commission Code — govern the publishers;88 the credibility attached to the medium of communication in issue;89 and the commercial motive of the publication.90 We caution that the imposition of a negligence standard on publishers other than the traditional media

83 NM (n 6 above) para 181. See also Langa CJ at para 94: ‘It makes sense that media defendants, who are experts in the field and who routinely distribute facts to vast numbers of people, with a particular air of authority and for commercial gain, should be held liable for disclosures which they should reasonably have foreseen would cause harm’. 84 NM (n 6 above) per Langa CJ paras 98-9; O’Regan J para 182. It is more difficult to understand Sachs J's decision in this regard. Sachs J concludes that all three defendants ‘did not meet the standard of reasonableness’, albeit that he accepted that the Bogoshi test applied to media defendants. It is not clear whether Sachs J regards Ms De Lille as a media defendant, thus attracting the higher standard of care, or whether he means to impose the higher standard of care beyond the traditional categories of media defendants. 85 In one High Court case, Marais v Groenewald en ’n ander 2001 1 SA 634 (T), it was precisely such reasoning that led to the Court imposing defamation liability based on gross negligence on the defendant, who had widely disseminated a letter to various political party structures. Van Dijkhorst J held that ‘[s]kriftelike laster kan anders as deur die media — soos hier deur partystrukture — tog ook wyd versprei word en ’n goeie naam skaad. Publikasie kan soveel wyer wees as in ’n plaaslike koerantjie’ (646). See further our comments in the text in regard to imposing negligence liability on non-media defendants. 86 See also Lange (n 65 above) 572. O’Regan J in NM (n 6 above) para 181 left open the issue of whether harmful targeted disclosure of information, e.g. to a small community, should also trigger a negligence standard of fault. 87 See also O'Regan J in NM (n 6 above) para 181. 88 See the dictum of Lord Hoffmann in Jameel (n 75 above). 89 See Bogoshi (n 10 above) 1213. 90 See NM (n 6 above) para 98 (Langa CJ) & para 181 (O’Regan J). 330 Media freedom and the law of privacy should be carefully considered on a case-by-case basis, in view of the serious consequences for freedom of expression that are involved.91

6 The negligence test and reliance on official reports

The final aspect of NM v Smith that we address in this note is the application by the minority judges of the negligence standard to the facts. It is here that O’Regan J parts ways with Langa CJ and Sachs J; while O’Regan J held that the media defendants did not act negligently, Langa CJ and Sachs J reached the opposite conclusion. The crisp issue for resolution was whether a reasonable journalist and publisher would have assumed from the Strauss Report that the plaintiffs had consented to their identities and HIV-status being revealed.

Langa CJ held that the Strauss Report was not a public document and was not widely distributed; in these circumstances:

The reasonable media defendant would ... not have relied on the Strauss Report as a document that removed their duty to ensure informed consent had been obtained ...

The inescapable conclusion is that a reasonable journalist or a reasonable publisher would have foreseen the possibility that there was not consent. Because the possible harm was great, the effort necessary to avoid the harm minimal and the benefit of publishing the names negligible, a reasonable journalist or publisher would have taken steps to avoid that harm.92

A diametrically opposed factual conclusion was reached by O’Regan J, who would have exonerated Ms Smith and the publisher from any negligence liability.93 The interesting aspect of O’Regan J’s judgment

91 As Langa CJ stated in NM (n 6 above) para 94: ‘However, to extend that standard to ordinary people, and thus to everyday relationships, would be to extend the law too far into intensely personal space ... [This] is not a matter that is appropriate for the law to regulate’. 92 NM (n 6 above) paras 110-111. Langa CJ also attached significance to the expert evidence of Professor Harber, that a journalist cannot assume that consent has been given in relation to HIV/AIDS reporting. Compare Sachs J at para 205, who held that Ms Smith ‘should have left no stone unturned in her pursuit of verification’. We submit that this dictum, which Langa CJ also adopted (para 111) sets the bar too high for media defendants. Even allowing for the nature of the private information in issue, it is, we submit, not correct as a matter of principle to postulate that reasonableness required that every possible step available to verify whether consent had been given, should have been followed. Rather, what is required of Ms Smith is that, within the context of a book publication (as opposed to, for instance, the reportage of breaking news in a daily newspaper), she ought to have taken reasonable steps to verify the issue of consent. 93 A conclusion also reached by the High Court (NM High Court (n 3 above) para 41). (2008) 1 Constitutional Court Review 331 for media freedom is her treatment of whether a reasonable journalist and publisher would be entitled to rely on the contents of the Strauss Report, which had disclosed the identities of the plaintiffs. O’Regan J reasoned as follows:

To hold that in the circumstances … [Ms Smith and the publisher] were under a further duty to contact either the University [of Pretoria] or the [plaintiffs] to ensure that they had in fact consented to publication of their names would impose a significant burden on freedom of expression. The result of such a finding would be that where personal private facts have been published already by a reputable organisation, another organisation may not rely on that publication as having been done lawfully and without infringement of privacy.94

While we agree with much of O’Regan J’s reasoning, it is difficult to fault Langa CJ’s analysis on the facts. We consider that it may well be that O’Regan J did not attach enough significance to the expert evidence of Professor Anton Harber and the nature of the privacy interests at stake, which would seem to us to require that a reasonable author and publisher of a book — presumably not under significant time pressures of the kind experienced by daily newspapers and broadcasters where breaking news is reported — ought to have taken additional steps to ascertain whether consent had been given for widespread disclosure to the public at large.95 This is more particularly so given that the inclusion of the names of the plaintiffs could not, in the circumstances, have added much of value to the discussion in the biography of Ms De Lille's role in the saga.96

What is ground-breaking about O’Regan J’s judgment from the perspective of media freedom is her view that, as a general rule, a reasonable journalist is entitled to rely on the contents of official documents:

Journalists must be entitled to publish information provided to them by reliable sources without rechecking in each case whether the publication was lawful, unless there is some material basis upon which to conclude that there is a risk that the original publication was not lawful. If there is a reasonable basis for suspecting that the publication of private information was without consent, a journalist will, of course, bear an

94 NM (n 6 above) para 185. 95 Langa CJ in NM (n 6 above) para 111 suggests that the steps that a reasonable journalist would have taken on the facts included finding the annexures to the Strauss Report containing the terms of the consent by the plaintiffs or contacting the plaintiffs directly. 96 Michelman ‘Interpretive charity’ (n 37 above) [current page 45 of article; editors to insert] of this volume.] attaches great significance to this point in rejecting O'Regan J's analysis on the facts. 332 Media freedom and the law of privacy

obligation to check. If there are no grounds for such suspicion, it cannot be said that a journalist acts negligently in not checking.97

This observation takes our reasonable publication law a significant step further than the Bogoshi progeny to date. In Bogoshi, Hefer JA set out a number of relevant factors to determine reasonableness of media conduct, including the ‘character and known provenance’ of the information, ‘the nature, extent and tone of the allegations’, ‘the reliability of their source as well as the steps taken to verify the information’, and ‘the opportunity given to the person concerned to respond, and the need to publish before establishing the truth in a positive manner’.98 However, it has been disappointing that, in the ten years since Bogoshi was handed down, very little guidance has emerged from the Courts as to how the reasonable publication test should be applied in a concrete set of circumstances.99 The leading decision on the application of the Bogoshi case, Mthembi-Mahanyele v Mail & Guardian Ltd,100 muddied the waters, with the Supreme Court of Appeal split evenly on the issue of whether the newspaper could benefit from a reasonable publication defence. Whereas Lewis JA — with whom Howie P concurred — regarded the journalists' conduct as reasonable, inter alia on the basis that the tone of the article was irreverent and that the newspapers had relied on information in the public domain,101 Mthiyane JA — with whom Mpati DP concurred — ruled that none of the guidelines suggested in Bogoshi was followed by the newspaper.102 Mthembi-Mahanyele did little to illustrate how to concretise some of factors that Hefer JA had enunciated in the Bogoshi case.

O’Regan J’s approach in NM v Smith that a journalist is, all things being equal, entitled to publish information derived from a reliable source, and particularly to repeat allegations made by a reputable organisation, therefore represents a desirable development in our jurisprudence in favour of media freedom. We submit that O’Regan J

97 NM (n 6 above) para 187. Langa CJ endorsed this suggestion; he held that ‘journalists should not be forced to verify disclosures made by reputable organisations’, but this would depend on ‘the nature of the document, the nature of the institution that produced the document, the importance of the interests involved and the relevant circumstances of the case’ (para 102). 98 Bogoshi (n 10 above) 1212-1213. 99 There have been only a handful of reported decisions applying Bogoshi, and not much clarity as to its application has emerged. For instance, in Sayed v Editor, Cape Times 2004 1 SA 58 (C), Davis J held that the defence was available for a story that had been produced with ‘considerable care’, whereas the defence failed in Lady Agasim-Pereira v Johnnic Publishing Eastern Cape (Pty) Ltd [2003] 2 All SA 416 (SE), inter alia on the basis that obvious sources had not been contacted to verify the allegations. 100 n 59 above. 101 Mthembi-Mahanyele (n 59 above) 74. 102 Mthembi-Mahanyele (n 59 above) para 110. The fifth judge, Ponnan AJA, held that the publication was not defamatory to begin with; he therefore did not need to address defences available to the media (para 85). (2008) 1 Constitutional Court Review 333 is correct to observe that, in general, it would set the standard of reasonableness too high to require journalists to verify the contents of official reports or documents, and that to require this would have a chilling effect on freedom of the media. This position is in harmony with the developing reasonable publication jurisprudence in England and in the European Court of Human Rights.

In England, the responsible publication defence has been held to entitle publishers to argue that their reporting of an important issue of public interest qualifies for protection under the ‘neutral reportage’ doctrine. In one leading case, Charman v Orion Group Publishing Group Ltd and others,103 the Court of Appeal confirmed that a responsible publication defence will be available where 'judging the thrust of the report as a whole, the effect of the report is not to adopt the truth of what is being said, but to record the fact that the statements which were defamatory were made'.104 In such cases, journalists will not be required to verify the truth of the underlying allegations.105

The European Court of Human Rights has expressly adopted the proposition contended for by O’Regan J. In Bladet Troms and Stensaas v Norway,106 the facts were that defamation proceedings were brought against a newspaper and its editor in relation to an article alleging breaches of seal hunting regulations. The Court held that an interference with the article 10 rights107 of the press was not justifiable. The newspaper had relied on the official report prepared by a seal hunting inspector for the Ministry of Fisheries. The Court held as follows:

In the view of the Court, the press should normally be entitled, when contributing to public debate on matters of legitimate concern, to rely on the contents of official reports without having to undertake independent research. Otherwise, the vital public-watchdog role of the press may be undermined ...

[T]he Court considers that the paper could reasonably rely on the official ... report, without being required to carry out its own research into the accuracy of the facts reported. It sees no reason to doubt that the newspaper acted in good faith in this respect.108

The sanctioning in this jurisprudence of reliance on official reports and other documents of solid providence has the potential to grant

103 [2008] 1 All ER 750 (CA) (Charman). 104 Charman (n 103 above) para 48. 105 See also Roberts v Gable [2008] 2 WLR 129 (CA) para 53. 106 n 81 above. 107 Article 10 of the European Convention on Human Rights protects freedom of expression. 108 Bladet (n 81 above) paras 68, 72. 334 Media freedom and the law of privacy the media significant breathing space to pursue important stories on matters of public interest. For instance, where a media publisher repeats allegations made in an official report, it should not be required of the media publisher to take extensive steps to verify the contents of the report (although the subject of the allegation should generally be provided with a right of reply). Stated differently, the media defendant's conduct in not verifying the truth of the allegations emanating from such a source should not count against it when assessing its conduct, both for purposes of the reasonableness defence (which negates unlawfulness) and for the defence of lack of negligence (which negates fault).

7Conclusion

NM v Smith is a fascinating — and disappointing — decision of the Constitutional Court on many levels. We have discussed some aspects of the decision in this note. From the perspective of constitutional jurisprudence, it is unfortunate that the majority of the Court treated the factual enquiry as to whether the defendants acted with intention in publishing the plaintiffs’ identities as a constitutional matter, without undertaking any substantive analysis of this difficult issue. This is particularly the case where the majority’s approach is difficult to reconcile with previous decisions of the Constitutional Court on what is and what is not a ‘constitutional matter’. The majority of the Court in our view also appeared to reach the incorrect conclusion on the facts, finding that the defendants had intentionally violated the plaintiffs’ privacy. We prefer the judgments of the minority judges, who reached a conclusion on the major constitutional issue involved in the case — that the media need to establish absence of negligence in publishing private information. The decision of Langa CJ on the facts — visiting negligence liability on the media defendants — appears to us to be defensible. Further, O’Regan J’s endorsement of a general rule that journalists are entitled to rely on reliable sources and information from reputable organisations, represents a significant advance in our media freedom jurisprudence. WANTED: A PRINCIPLED APPROACH TO THE BALANCING OF PUBLIC POLICY CONSIDERATIONS. STEENKAMP NO V PROVINCIAL TENDER BOARD, EASTERN CAPE

Sanele Sibanda*

It matters how judges decide cases1

1Introduction

Steenkamp NO v Provincial Tender Board, Eastern Cape2 is unlikely to generate too much controversy in academic circles or in the profession. The majority decision entrenches the status quo in so far as the delictual liability of tender boards is concerned. And apart from the majority judgment’s rather confounding comments that appeared to conceive the subject of remedies in rather rigid and dichotomous terms, nothing in the written text appears outwardly contentious. Even the disagreement between the majority and the minority concerns a point of law upon which reasonable people could disagree.

However, that is not to say that the judgment does not raise any issues of interest and concern. My aim in this note is to reflect upon, and where appropriate critique, the manner in which the majority and minority arrived at their respective positions with respect to the

* Senior Lecturer, University of Witwatersrand. BA, LLB (Cape) LLM (Wits). I wish to thank the editors of the Constitutional Court Review for inviting me to be a contributor to the inaugural edition and for affording me the opportunity to present an earlier version of this note at the Constitutional Court Review Conference in Somerset West. I am thankful for the comments and questions raised by the participants at the conference, particularly those of Cora Hoexter. I would especially like to thank Stu Woolman for his editorial input, as well as for his continuing support, encouragement and patience. In the spirit of Harambee, I must express my heartfelt thanks to Mtendeweka Mhango for his insightful comments and valuable engagement with an earlier version of this paper. Finally, I must convey my thanks to Jonathan Klaaren for his critical and insightful comments. Any mistakes or omissions remain my own. 1 See R Dworkin Law’s empire (1986) 1. 2 2007 3 BCLR 300 (CC).

335 336 Wanted: A principled approach to the balancing of public policy considerations determination of wrongfulness. In particular, I will consider the manner in which the both judgments dealt with the balancing of the public policy considerations raised. I critique the rather thinly substantiated, individualised and subjectivised approach of the majority to the balancing of the policy considerations. In particular, I take issue with the fact that the majority’s approach is focused on justifying the outcome of the matter as and between the parties, thus missing an important opportunity to guide us on how matters involving similar considerations should be approached in future. I then contrast this approach with the minority’s clearer identification of the values and interests represented by the policy considerations raised.

The minority’s approach to the balancing of policy considerations — grounded in the principle of proportionality — is to be preferred. Proportionality doe not simply require that courts clearly identify the competing values or interests at stake in a dispute. Properly conceived, proportionality demands that courts elaborate in some detail the reasons that support their conclusions.

2 Background and facts

Although judgment in Steenkamp was handed down in 2006 and reported in 2007, the facts giving rise to this matter preceded the Constitutional Court’s decision by over a decade. The matter has its origins in disputed tender proceedings in the Eastern Cape. In 1996, Balraz Technologies (Pty) Ltd (‘Balraz’) was awarded a tender by the provincial tender board. However, that award was subsequently set aside in review proceedings instituted by one of the unsuccessful tenderers, Cash Paymaster Services (Pty) Ltd.3 In Cash Paymaster Services (Pty) Ltd v Eastern Cape Province, the court held that given the irregularities in the tender adjudication process, the tender board had acted in an administratively unfair manner.

Unfortunately, Balraz did not survive to tender another day. It was placed under final liquidation before it could retender. As a result, all future claims relating to the original tender were instituted by Balraz’s liquidator. At the trial court, the applicant, Balraz’s liquidator, was unsuccessful in its claim for the recovery of out-of- pocket expenses incurred as a result of its having commenced to deliver services as per the contract arising out of the original tender award. The court held that the applicant’s status as an unregistered entity at the time of submitting its tender documentation rendered the tender award void. No relationship giving rise to a duty of care could have arisen as and between itself and the tender board. Put

3 For the review decision, see Cash Paymaster Services (Pty) Ltd v Eastern Cape Province 1999 1 SA 324 (Ck). (2008) 1 Constitutional Court Review 337 differently, the conduct of the tender board vis-à-vis the applicant could not be properly conceived as being wrongful.

Undeterred by this outcome, the applicant appealed to the Supreme Court of Appeal. That appeal was dismissed. The Supreme Court of Appeal’s decision focused on the delictual liability of the tender board as a public body. It held, on grounds of public policy, that neither statute nor common law ought to be extended to cover alleged delictual damages caused by the tender board (to the applicant) that were of a purely economic nature. The Supreme Court of Appeal held so despite the fact that it had found that the tender board had acted negligently (albeit in good faith).

3The majority decision4

In the Constitutional Court, the issue was crystallised as follows: Can a once successful tenderer recover its out-of-pocket expenses incurred after the contract was awarded, but before the said award was set aside on review?5 The majority judgment, delivered by Moseneke DCJ, turned on whether or not the respondent’s decision to award the tender to the applicant, whilst in breach of its constitutional and statutory mandate to exercise its powers in an administratively fair manner, was wrongful in a delictual sense.6 The respondent had conceded that it had acted in an administratively unfair manner, albeit in good faith, when it assessed and awarded the tender to Balraz.7

In considering the import of the tender board’s failure to uphold the principles of administrative justice, the majority rejected the idea that the breach of a constitutional or a statutory duty should as a matter of course be regarded as being delictually wrongful. According to the majority, the determination of wrongfulness giving rise to the recognition of a duty of care and liability on the part of a defendant could not be made without having regard to public policy considerations of what was fair and reasonable in the particular circumstances.8

4 I do not consider the concurring decision delivered by Sachs J because the issues raised therein are not germane to those presently under consideration. 5 Two other issues were raised in the Constitutional Court: firstly, whether or not there was a constitutional issue raised; and secondly, whether or not it was in the interests of justice to consider the matter. These issues will not be discussed in this note. For present purposes it suffices to say that in both instances the decision of the court was in the affirmative. 6 See sec 187 of the Constitution of the Republic of South Africa, Act 200 of 1993 (Interim Constitution) and sec 217(1) of the Constitution of the Republic of South Africa, 1996 (‘Final’ Constitution); and the State Tender Board Act 2 of 1994 (Repealed by Provincial Tender Board Repeal Act (Eastern Cape) 6 of 2004). 7 Steenkamp (n 2 above) para 36. 8 Steenkamp (n 2 above) para 37. 338 Wanted: A principled approach to the balancing of public policy considerations

In holding that the applicant’s claim must fail, the majority held that the applicable constitutional and statutory provisions in terms of which the tender board acted did not envision a tenderer in the position of Balraz having a claim to damages.9 Instead the majority took the view that had Balraz been prudent in the circumstances, it would have utilised other avenues of redress available to it to mitigate its risk of loss.10 For example, according to the majority, Balraz could have tendered again after the adverse outcome of the review proceedings.11 Alternatively it could have negotiated for a right to reimbursement for out-of-pocket expenses at the time of contracting in order to safeguard it from loss in the event of the tender award being overturned on review.12

In concluding that Balraz was not owed a duty of care by the tender board, the majority held that no rights or values in the Constitution would justify an extension of the ambit of the law of delict in the manner sought. In reaching this conclusion, the majority found that compelling public policy considerations required that it provide some level of immunity to tender boards that made innocent but negligent decisions.13 One such policy consideration that weighed heavily on the majority was a deep-seated concern that an extension of liability in the manner sought would open the floodgates of litigation by unsuccessful tenderers. This prospect, the majority feared, would have the undesirable consequence of eroding the efficacy of the tendering process and tarnishing the image of tender boards in the public’s eyes.14 Given that the majority had disposed of the matter on the basis of wrongfulness, the majority declined to answer questions relating to the validity of Balraz’s tender or the tender board’s negligence.15

4The minority decision

Langa CJ and O’Regan J, in delivering the dissenting minority judgment, identified one especially critical ground upon which they

9 Steenkamp (n 2 above) para 47. 10 Moseneke DCJ, clearly unmoved by the applicant’s contentions, makes this point when he writes that ‘[o]n the facts, Balraz wasted no moment to accept the tender award. But once the order to supply goods and services was made by the department, Balraz should have curbed its commercial enthusiasm as it was well within its right to require that its initial expense not lead to its financial ruin should the award be nullified. Balraz unnecessarily chose the more hazardous course which is to incur mainly salary expenses of directors without fashioning out an appropriate safeguard. Its loss could have been easily curbed by prudent conduct and precaution.’ Steenkamp (n 2 above) para 52. 11 Steenkamp (n 2 above) para 49. 12 Steenkamp (n 2 above) para 50. 13 Steenkamp (n 2 above) para 55(a). 14 Steenkamp (n 2 above) para 55(b). 15 Steenkamp (n 2 above) para 61. (2008) 1 Constitutional Court Review 339 differed with the majority, namely the determination of wrongfulness on the part of the tender board. Langa CJ and O’Regan J deemed it important to characterise the applicant’s claim as one of pure economic loss. Such a characterisation had historically played a part in the courts’ disinclination to recognise a duty of care.16

According to the minority, the nature of the wrongfulness inquiry in the post 1994-era must, in essence, accord with the norms and the values of Constitution. In a matter such as the one before the Court, the minority held that the following norms or principles had to be taken into account when determining wrongfulness:

(1) The need to guarantee that successful tenderers have an enduring faith in the tender process and the resultant contracts. Such a guarantee ensures that tenderers commence to deliver on their contracts timeously without having to labour under the fear of facing financial ruin in the event that a tender is subsequently set aside through no fault of their own. Further, a failure to provide such a guarantee posed a greater risk of financial ruin to smaller or new business enterprises that government procurement policies sought to promote.17 (2) The need to properly differentiate between claims for damages arising out of loss of profit instituted by an unsuccessful tenderer on the one hand; and damages resulting from out-of-pocket expenditure incurred by an initially successful tenderer who had commenced to deliver services in accordance with its contractual obligations. The former would constitute a windfall claim and the latter a claim for reimbursement for services rendered.18 (3) The need to promote accountability in the entire tender process in order to ensure that that once successful tenderers who incur out-of- pocket expenditure spent in the good faith fulfilment of a tender contract were not unduly burdened with unrecoverable expenditure. Such a burden would inevitably have a detrimental impact on the future delivery of tender obligations.

Having considered the factors set out above, the minority concluded that in the circumstances of the case, the tender board acted wrongfully, owed the applicant a duty of care, and thus created the basis for a claim for delictual damages.

16 Steenkamp (n 2 above) para 67. 17 Steenkamp (n 2 above) para 82. 18 Steenkamp (n 2 above) para 83. 340 Wanted: A principled approach to the balancing of public policy considerations

5 Comparing the approach of the majority and the minority to the balancing of public policy considerations

My focus in this section will be on what I take to be the main point of disagreement between the two opinions: namely the approach to the balancing of public policy considerations. I will not, therefore, concern myself with the actual decision reached on the nonconstitutional point of delictual wrongfulness. For purposes of this note I am satisfied that the Court could conceivably have gone either way.

5.1 Wrongfulness as a requirement for delictual liability

Before I turn squarely to the central concern of this note, namely the balancing of public policy considerations I want to consider briefly the requirement of wrongfulness as an element of delictual liability. The requirement of wrongfulness is a sine qua non for the attachment of delictual liability. Wrongfulness, in a delictual sense, is said to exist where particular conduct either infringes upon a person’s legally recognised right or where it results in the breach of a legal duty owed to a person by another.19 According to Neethling et al, whether particular conduct is in fact wrongful is determined with reference to the legal convictions of the community or the community’s boni mores. This inquiry, according to the authors, entails an objective test based on the criterion of reasonableness. The pertinent question to ask is whether or not the conduct in question was justifiable or unjustifiable.20

In the context of the wrongfulness inquiry, the determination of reasonableness requires (a) that a court must take into consideration all the circumstances of the matter pertaining to the parties and (b) that a court must consider issues of public and legal policy in making a value judgment as to how the alleged wrongdoer’s conduct should be viewed by society.21 The second proviso requires that the

19 JR Midgley & JC Van der Walt ‘Delict’ LAWSA (2nd ed, 2005) Volume 8(1) para 60. See also J Neethling et al Law of delict transl & ed J Knobel (5th ed, 2005) 31. 20 Neethling et al (n 19 above) 34. 21 See Olitzki Property Holdings v State Tender Board and Another 2001 8 BCLR 779 (SCA) para 12 (Cameron JA commenting on how reasonableness is determined puts it as follows: ‘The determination of reasonableness here in turn depends on whether affording the plaintiff a remedy is congruent with the court’s appreciation of the sense of justice of the community. This appreciation must unavoidably include the application of broad considerations of public policy determined also in the light of the Constitution and the impact upon them that the grant or refusal of the remedy the plaintiff seeks will entail.’ (footnotes omitted).) See also Minister of Safety and Security v Van Duivenboden 2002 6 SA (2008) 1 Constitutional Court Review 341 reasonableness of particular conduct be determined in terms of the legal convictions of the community.22 Historically, the test for reasonableness in this context has been described as being open- ended and flexible. It necessarily takes cognisance of the ever- evolving nature of the legal convictions of the community.23

5.2 The important role of public policy considerations in determining wrongfulness in novel situations

The role played by public policy considerations in the determination of reasonableness in areas where liability has not previously been established cannot be overstated.24 Van Aswegen, writing in the pre- constitutional era, pointed out that the most important role played by policy considerations has been in ‘the new developments, expansions and adaptations to the settled body of delict law.’25 The overriding utility of public policy considerations is that they afford the court an opportunity to consider novel problems for which there is no precedent or authority. In other words, despite the clearly open- ended nature of the concept of ‘public policy’, the concept, at the very least, requires judges to provide some level of principled reasoning grounded in social policy for the decisions that they make.

The importance of public policy has been entrenched further in the current constitutional dispensation. Significant developments have occurred in extending the limits of delictual liability through the recognition of new duties of care. These new duties have often been

21 431 (SCA) para 16; Premier of the Province of the Western Cape v Fair Cape Property Developers (Pty) Ltd [2003] 2 All SA 465 (SCA) para 41. 22 Neethling et al (n 19 above) 34. See also A Fagan ‘Rethinking wrongfulness in the law of delict’ (2005) 122 South African Law Journal 90 93-94. Although, strictly speaking, it is outside the scope of this paper, it is worth mentioning that the Fagan’s robust analysis of wrongfulness as an element of delict in the aforementioned article has ignited a debate with respect to the manner in which our courts engage in the wrongfulness inquiry (and its relationship to negligence). See J Neethling ‘The conflation of wrongflness and negligence: Is it always such a bad thing for the law of delict’ (2006) 123 South African Law Journal 204; RW Nugent ‘Yes, it is always a bad thing for the law: A reply to Professor Neethling’ (2006) 123 South African Law Journal 557; J Neethling & JM Potgieter ‘Wrongfulness and negligence in the law of delict: A Babylonian confusion?’ (2007) 70 Tydskrif vir Hedendaagse Romeins-Hollandse Reg (2007) 120; J Neethling & JM Potgieter ‘In (self)defence of the distinction between wrongfulness and negligence’ (2007) 124 South African Law Journal 280; and A Fagan ‘Blind Faith: A response to Professors Neethling and Potgieter’ (2007) 124 South African Law Journal 285. 23 See Midgley et al (n 19) para 60. See also Amod v Multilateral Motor Vehicle Fund (Commission for Gender Equality Intervening) 1999 2 All SA (SCA) para 23. 24 MM Corbett ‘Aspects of the role of policy in the evolution of our common law’ (1987) 104 South African Law Journal 67. 25 A van Aswegen ‘Policy considerations in the law of delict’ 1993 (56) Tydskrif vir Hedendaagse Romeins-Hollandse Reg 171; See also Midgley et al (n 19 above) para 60. 342 Wanted: A principled approach to the balancing of public policy considerations rooted in the ‘constitutionalisation’ of public policy. In Carmichele,26 the Constitutional Court recognised that in making the type of value judgments required in order to determine wrongfulness in novel situations, the constitutional injunction contained in s 39(2) requires all courts to promote the spirit, object and purport of the bill of rights.27 In deciding to extend the delictual liability of police officers and prosecutors to instances wherein the wrongful conduct constituted an actionable omission, the Carmichele Court accepted that the determination of wrongfulness demands that some regard be paid to public policy considerations. Ackermann J and Goldstone J wrote:

Before the advent of the [Interim Constitution], the refashioning of the common law in this area entailed ‘policy decisions and value judgments’ which had to ‘reflect the wishes often unspoken, and the perceptions, often dimly discerned, of the people’. A balance had to be struck between the interests of the parties and the conflicting interests of the community according to what ‘the (c)ourt conceives to be society’s notions of what justice demands’. Under s 39(2) of the Constitution concepts such as policy decisions and value judgments reflecting ‘the wishes ... and the perceptions ... of the people’ and ‘society’s notions of what justice demands’ might well have to be replaced, or supplemented and enriched by the appropriate norms of the objective value system embodied in the Constitution.28 (emphasis added)

The Court’s acknowledgment of the fact that the advent of the Constitution has established ‘an objective normative value system’ has influenced the development of notions of public policy and, concomitantly, the understanding of what constitutes the legal convictions of the community.29 The influence of this ‘phrase’ has led

26 See Carmichele v Minister of Safety and Security 2001 4 SA 938 (CC) (hailed as a groundbreaking judgment that saw the extension of the delictual liability for omissions to instances in which the police and public prosecutors failed to act in circumstances where they could have prevented the plaintiff from being raped). Section 39 (2) of the Constitution reads as follows: ‘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’. 27 Woolman has made the point that the effect of sec 39(2) is to place a mandatory injunction on all courts to interpret all laws in a way that promotes the spirit, purport and objects of the Bill of Rights. Put differently, sec 39(2) establishes an interpretive canon that dispels any notions of judicial discretion. In accordance with this view, a failure by the lower courts to fulfil the obligation imposed by sec 39(2) renders their decisions susceptible to constitutional review by the Constitutional Court. See S Woolman ‘Application’ in S Woolman et al (eds) Constitutional law of South Africa (2nd Edition, OS, 2005) ch 31. See also F Michelman ‘The rule of law, legality and the supremacy of the Constitution’ in Woolman et al (eds) (above) ch 11. 28 Carmichele (n 26 above) para 56 (footnotes omitted). 29 See Woolman ‘Application’ in Woolman et al (eds) (n 27 above) 31-93, for a discussion on the origins, application and problems with use of the phrase ‘an objective normative value system’ in South African constitutional jurisprudence. (2008) 1 Constitutional Court Review 343 to positive developments in the lower courts. Olitzki,30 Premier, Western Cape v Fair Cape Property Developers,31 and the Supreme Court of Appeal decision of Steenkamp32 all turn on the grounding of fairness and reasonableness in terms of constitutional rights and constitutional norms.33

However, in spite of these positive developments, a couple of important questions are left unanswered. How does a court determine what public policy considerations to take into account when engaging with the wrongfulness inquiry in a novel situation? And further, is there a principled basis upon which courts go about balancing public policy considerations; or is it a purely a matter of judicial discretion? In fact, what becomes evident from a review of the South African academic literature and case law is that although the determination of wrongfulness is ultimately a legal question, no explicit legal rules or principles guide the selection or the deployment of public policy considerations.34

So despite the apparent mandatory injunction of s 39(2), the courts enjoy and exercise a wide discretion with regard to the type of public policy considerations that they deem appropriate and relevant in any particular situation. At a minimum, the litigants’ pleadings will influence the choice of policy considerations. However the courts, as was demonstrated in Carmichele, must consider, mero motu, other relevant policy considerations including those that flow from our basic law. While sound reasons exist for this type of judicial discretion — the inevitability of novel disputes — the same considerations do not

30 Olitzki (n 21 above) para 31. 31 Premier of the Province of the Western Cape v Fair Cape Property Developers (Pty) Ltd [2003] 2 All SA 465 (SCA) para 33. 32 Steenkamp NO v Provincial Tender Board, Eastern Cape [2006] 1 All SA 478 (SCA) para 25 (Steenkamp (SCA)). 33 See Pharmaceutical Manufacturers Association of South Africa: In re Ex parte President of the Republic of South Africa 2000 2 SA 674 (CC) para 44. In this matter, Chaskalson CJ emphatically pronounced that ‘there is only one system of law. It is shaped by the Constitution which is the supreme law, and all law, including the common law, derives its force from the Constitution and is subject to constitutional control.’ Although the context and issues under consideration when Chaskalson CJ made this pronouncement were quite different from the issues under consideration presently, the potency of this holding can equally be extended to the present context. This is to say that the courts have, in my opinion, correctly recognised the overarching influence of the Constitution such that even in the determination of matters of public policy, such determination must be made within the rubric of constitutional norms and principles. See also B Leinius & JR Midgley ‘The impact of the Constitution on the law of delict’ (2002) 119 South African Law Journal 20; Neethling et al (n 19 above) 36. 34 See van Aswegen (n 25 above) 174. She considers the philosophical underpinnings of the nature of policy considerations and whether or not they can appropriately be considered to constitute legal rules when applied in judicial decision-making. Although Van Aswegen argues convincingly that policy considerations when invoked by judges are equally part of the law as are other legal sources, she does not deal with the question as to how judges should decide on what these considerations are to be, nor how they should be applied in a particular instance. 344 Wanted: A principled approach to the balancing of public policy considerations come into play when it comes to the determination of how competing public policy considerations should be assessed and reconciled.

5.3 A comparative evaluation of the approach taken by the majority and the minority to the balancing of policy considerations

In Steenkamp, both the majority and minority judgments took into account similar policy considerations with respect to the impact that an extension of delictual liability would have on the functioning of tender boards.35 However, the two judgments came to vastly different conclusions as far as the weight to be attached to these considerations is concerned and whether the tender board acted wrongfully. The majority was clearly moved by arguments advanced against the extension of liability, whilst the minority was less persuaded. The arrival at different conclusions is not something that in itself is notable. Different judges will disagree regarding the construction of the facts and proper application of the law. I would even go so far as to contend that such differences of opinion amongst judges are a desirable feature of any legal system that is attuned to the contested views of the role of law from different sectors of a heterogeneous polity. However, for present purposes my concern is not with the different conclusions reached by judges of the majority and the minority. It remains, rather simply, focused on the manner in which they engage competing policy considerations.

One major problem with the majority’s approach to wrongfulness is that it reveals a decidedly individualised inquiry that places undue emphasis on Balraz’s particular circumstances and fails to focus on the public policy considerations underlying its claim.36 It would appear from the majority decision that although the majority does take cognisance of the public policy grounds raised by the plaintiff, it did not seriously engage them.37 By way of contrast, the considerations of public policy in support of not extending liability to

35 See Steenkamp (n 2 above) paras 49-50 & 55 (majority judgment) and paras 87-92 (minority judgment). 36 See Steenkamp (n 2 above) paras 43-46 (majority judgment). 37 In fact Moseneke DCJ, Steenkamp (n 2 above) para 47, states: ‘I must at the outset say that the submissions of the applicant are attractive but not sustainable’. (2008) 1 Constitutional Court Review 345 tender boards, although also not engaged with at great length,38 are considered in a decidedly more objective fashion that clearly identifies the public interests that they seek to promote without subjectivising the inquiry to focus on the conduct of this particular tender board. Ultimately, the competing public policy considerations pit the individualised considerations of the applicant against the potentially negative social consequences and public interest concerns that militate against the extension of delictually liability to tender boards. The entire exercise ineluctably results in the emasculation of the applicant’s claim. This diminishment of the applicant’s (public policy) interests not only creates an extra hurdle for the applicant, but unjustifiably shifts the focus away from the open and candid consideration of competing interests that a proportionality exercise demands when determining the question of wrongfulness.

Put somewhat differently, the primary problem with the approach taken by the majority is its failure to identify and elaborate upon the values and interests underlying the applicant’s claim for a delictual remedy and properly weighing them up against the competing values and interests of the tender board as a public body. The applicant had an underlying right to administrative justice39 and a right to appropriate relief.40 And yet neither the right nor the desired remedy is engaged by the majority’s opinion. Instead the majority’s reasoning is predicated on the belief that the applicant should have shown that it had done everything within its power to mitigate its loss before it could allege wrongfulness.

The minority opinion takes a markedly different approach to the identification, elaboration and assessment of competing policy considerations when determining the question of wrongfulness. The minority’s approach, although also clearly grounded in the factual matrix of the matter before the court, is more deliberate in the manner in which it engages with these competing considerations. By this I simply mean that the minority’s approach displays a more deliberate effort to identify the societal and public interests that

38 See Steenkamp (n 2 above) para 55: Moseneke DCJ holds that the majority concur with the ‘significant findings and conclusions’ of the Supreme Court of Appeal. It is interesting to note that Moseneke DCJ finds it adequate to concur in this fashion and simply leave the matter there. What makes this fact all the more peculiar is the fact that earlier on in the decision the Deputy Chief Justice goes to some length to justify the constitutional importance of the matter at hand; and its potential to influence the future development of the law in respect of appeals to private law remedies in matters that are clearly founded in public law. See Steenkamp (n 2 above) paras 17-23. 39 See sec 33(1) of the Constitution: ‘Everyone has the right to for the right to administrative action that is lawful, reasonable and procedurally fair.’ 40 Section 38 of the Constitution affords a person a right to approach a court seeking ‘appropriate relief’ where that person alleges that their rights have been infringed. 346 Wanted: A principled approach to the balancing of public policy considerations underpinned the applicant’s claim.41 Although not expressly articulated in these terms, the minority approach to the assessment of public policy considerations would appear to be grounded in the principle of proportionality.42

41 The best way to illustrate this point is by way of example. The approach of the majority and that of the minority differ on the point of whether or not the law should be developed to accommodate a claim for out-of-pocket expenses instituted by a once successful tenderer. The majority position on this point is set out as follows: The residual question is whether there is justification to develop the common law to embrace this narrow claim for damages based on out-of- pocket expenses in favour of an initially successful tenderer where the award is subsequently set aside by the court and the tenderer retains the right to participate in the subsequent tender process. I think not. First, there is no magic in characterising financial loss as out-of-pocket. If public policy is slow to recompense financial loss of disappointed tenderers it should not change simply because of the name the financial loss bears. Second, even if there may not be a public law remedy such as an interdict, review or appeal this is no reason for resorting to damages as a remedy for out-of-pocket loss. This is so because first, as I found earlier, the loss may be avoided and second it is not justified to discriminate between tenderers only on the basis that they are either disappointed tenderers or initially successful tenderers. To do so is to allot different legal rights to parties to the same tender process. There is no justification for this distinction particularly because ordinarily both classes of tenderers are free to tender again should the initial tender be set aside. Steenkamp (n 2 above) para 54. The minority position on this point is set out as follows: A claim for out-of-pocket expenses, however, is a far more modest claim. Moreover, the inability to recover out-of-pocket expenses may well render smaller and less financially viable tenderers at risk of liquidation. Indeed, such was the case here. In our country, government procurement is one of the key mechanisms for ensuring that those previously locked out of economic opportunity by the policies of apartheid, are given an opportunity to participate. By definition such companies and individuals are often new, small and not financially robust. In our view, this is an additional and important factor which supports our conclusion that both Moseneke DCJ and the Supreme Court of Appeal are incorrect to conclude that aquilian liability cannot arise on the circumstances of this case. Steenkamp (note 2 above) para 80. 42 See Steenkamp (n 2 above) para 69. Although Langa CJ and O’Regan make no direct reference to the principle of proportionality, it is, in my view, implicit in their approach to the balancing of policy considerations: ‘Determining whether conduct is wrongful or unlawful for the purposes of aquilian liability, as has been stated on many occasions, essentially involves a question of legal policy which must be answered in the light of the norms and values of our society and, since 1994, in the light of the norms of our Constitution. Even though the question is a normative one, it is not to be answered on the basis of 'an intuitive reaction to a collection of arbitrary factors', but instead requires an identification of the relevant norms followed by an analysis and, if necessary, balancing of those norms to determine the outcome.’ (parqa 70) (emphasis mine and footnotes omitted.) (2008) 1 Constitutional Court Review 347

6 The prinicple of proportionalty: A possible basis upon which the balancing of public policy considerations should be performed?

The principle of proportionality is firmly established in South African constitutional law. The invocation of the principle of proportionality has been experienced mainly within the context of limitations analysis.43 The most fundamental idea underlying the principle of proportionality is that courts will candidly assess the competing rights, values and interests reflected in both the asserted constitutional norm and in the existing law that allegedly impairs that norm.44 In other words, the principle of proportionality requires more than a mechanical exercise of summing up on some imagined scale the respective virtues of the two sides to the disputes. Chaskalson P, in S v Makwanyane, offers up a lucid exposition of the nature of the principle of proportionality and what it entails:

The limitation of constitutional rights for a purpose that is reasonable and necessary in democratic society involves the weighing up of competing values, and ultimately an assessment based on proportionality. This is implicit in the provisions of section 33 (1). The fact that different rights have different implications for democracy, and in the case of our Constitution, for an ‘open and democratic society based on freedom and equality’, means there is no absolute standard which can be laid down for determining reasonableness and necessity. Principles can be established, but the application of those principles to particular circumstances can only be done on a case by case basis. This is inherent in the requirement of proportionality, which calls for the balancing of different interests.45

The value of the principle of proportionality would appear to be most apparent in those areas or instances where the courts must address rather novel issues, where no clear legal rule or precedent exists and where important matters of principle are at stake.46 In such instances the principle of proportionality will require a court deciding upon such

43 Section 36 of the Constitution provides the test for the limitation of constitutional rights. 44 See S Woolman & H Botha ‘Limitations’ in Woolman et al (eds) (n 27 above) ch 34; I Currie & J de Waal The Bill of Rights handbook (5th ed, 2005) 172-8; G Erasmus ‘Limitation and suspension’ in D van Wyk et al (eds) Rights and constitutionalism: The new South African legal order (1994) 629. 45 1995 3 SA 391 (CC) para 104. This dictum has been cited with approval in many subsequent cases. See, eg, National Coalition for Gay and Lesbian Equality v Minister of Justice and Others 1999 1 SA 6 (CC) para 33. 46 An obvious example of such an issue is where a law is being challenged for violating the rights protected in the Bill of Rights. In most instances there will be some right demand — fundamental and therefore worthy of constitutional protection — being pitted against some law that will often seek to attain some important societal purpose. 348 Wanted: A principled approach to the balancing of public policy considerations an issue to make explicit, at least, the following: the exact nature of the conflict; what the competing or conflicting values or interests are and the importance of these; and the substantive reasons for choosing one set of values or interests over the other. In other words, the principle of proportionality demands that where the courts are required to engage in the exercise of balancing they should do so in a manner that clearly identifies all the pertinent competing values or interests, before weighing these up and providing fully substantiated conclusions. This type of application of the principle of proportionality is one that we have become accustomed to seeing at work in our bill of rights jurisprudence.

The application of the principle of proportionality or at the very least an approach that is closely akin to it is nothing novel in delict. Courts have often had to consider whether or not to extend liability. In Carmichele, Judges Ackermann and Goldstone allude to the importance of [the principle of] proportionality:47

... in determining whether there was a legal duty on the police officer to act, Hefer JA in Minister of Law and Order v Kadir referred to weighing and striking a balance between the interests of the parties and the conflicting interests of the community. This is a proportionality exercise with liability depending upon the interplay of various factors. Proportionality is consistent with the Bill of Rights, but that exercise must now be carried out in accordance with the ‘spirit, purport and objects of the Bill of Rights’ and the relevant factors must be weighed in the context of [a] constitutional State founded on dignity, equality and freedom and in which government has positive duties to promote and uphold such values. (emphasis added)

The core of the idea underlying the principle of proportionality and how it can add an important dimension to public policy considerations in the determination of wrongfulness was powerfully captured by Nugent JA in Van Duivenboden:

When determining whether the law should recognise the existence of a legal duty in any particular circumstances what is called for is not an intuitive reaction to a collection of arbitrary factors, but rather a balancing against one another of identifiable norms.48 (emphasis added)

A similar approach to rethinking public policy considerations grounded in an application of the principle of proportionality was also followed in Steenkamp (SCA).49 The Supreme Court of Appeal’s judgment — like the minority’s opinion in the Constitutional Court — illustrates how the principle of proportionality can be applied in the

47 Carmichele (n 26 above) para 43. 48 Van Duivenboden (n 26 above) para 21. 49 Steenkamp (SCA) (n 32 above). (2008) 1 Constitutional Court Review 349 determination of wrongfulness. The SCA’s decision accomplishes this task by meticulously identifying and engaging with the public policy considerations it deemed relevant to the determination of wrongfulness.50 The SCA focused on the public and societal interests that underlay both the appellant’s claim for a remedy and the respondent’s denial of liability. Although the final decision with respect to wrongfulness in the Supreme Court of Appeal accords with that of the majority of the Constitutional Court, it is certainly more palatable. It gives due consideration to the competing rights and interests of both the applicant and the respondent in an objective and judicious manner. In short, the conclusion reached as to the question of wrongfulness in the Supreme Court of Appeal decision is more transparently reasoned and does not leave one with an uneasy sense that the exercise of balancing masks the subjective whims of judges.51

7Conclusion

Again. My discomfort with the Constitutional Court’s majority decision in Steenkamp is not with its conclusion that the conduct of the tender board was not wrongful. The matter really could have gone either way. However, my main concern and the reason for writing this note is to take issue with the unsatisfactory and somewhat opaque manner in which the majority reached its conclusion. The most profound shortcoming of this decision is that it leaves the Constitutional Court vulnerable to the critique that the ultimate conclusion as to wrongfulness reflects a mere subjective preference rather than a decision grounded in principle. That some may consider this charge to be unfair criticism is undoubted. However, as long as courts continue to engage in an amorphous balancing exercise, critiques such as mine will be difficult to rebut. A failure to develop a principled framework will always leave those who are sitting on the wrong side of a decision in which public policy considerations were considered with reason to be sceptical about what actually animated

50 See Steenkamp (SCA) (n 32 above) paras 29-44 51 See Woolman & Botha (n 44 above) 34-93 to 34-104 9: The authors provide a sustained critique of the concept of balancing and its relationship with proportionality; in particular, the inappropriateness of the balancing metaphor in constitutional adjudication. 350 Wanted: A principled approach to the balancing of public policy considerations a court’s decision.52

Steenkamp has given us an opportunity to consider how the exercise of balancing policy considerations might be profitably supplanted by a more principled framework. The principle of proportionality is a useful departure point in the creation of such a framework. The appeal of this principle, to my mind, is that it is already established in our law, and more importantly, that it accords with the unarticulated practices that have already taken hold in our courts under the new constitutional dispensation.

52 In a recent article Stuart Woolman has raised similar concerns about the Constitutional Court’s thin reasoning. Woolman objects to this minimalist mode of adjudicative reasoning by the Constitutional Court for several reasons, not least of which is that he considers it to be tantamount to abdication by the Constitutional Court of its responsibility to develop clearly articulated rules and to provide clearly substantiated reasons for its decisions. These clearly articulated rules and substantiated reasons are, according to Woolman, essential for purposes of developing a coherent constitutional jurisprudence that informs political discourse, as well as serving to guide the lower courts and the other coordinate branches of government. S Woolman ‘The amazing, vanishing Bill of Rights’ (2007) 124 South African Law Journal 762 784-787. See also F Michelman ‘On the uses of “interpretive charity”’: Some notes on application, avoidance, equality and objective unconstitutionality from the 2007 term of the Constitutional Court of South Africa’ (2008) 1 Constitutional Court Review 1 (Contends that while the judgments in question are thinly reasoned, they can, if patiently (and charitably) assessed, and if given sufficient expiation by commentators, be reconstructed so as to avoid some of the alleged problems with both their form and substance.) SEVERING THE UMBILICAL CORD: A SUBTLE JURISPRUDENTIAL SHIFT REGARDING CHILDREN AND THEIR PRIMARY CAREGIVERS

Ann Skelton

1Introduction

In S v M (Centre for Child Law as Amicus Curiae)1 the Constitutional Court overturned a High Court judgment in which a primary caregiver of children had been sentenced to correctional supervision, including a brief period of imprisonment. The appellant had been convicted on multiple counts of fraud in the Regional Court. The sentence handed down by the High Court on appeal required her to undergo imprisonment prior to possible release on correctional supervision. Her further appeal to the Constitutional Court focused on the duties of a court when sentencing a primary caregiver, given the constitutional injunction that a child’s best interests are of paramount importance in all matters concerning the child. The Constitutional Court’s judgment subtly adjusts the balancing exercise at sentencing, as governed by the so-called ‘triad of Zinn’ principles. The judgment also provides an endorsement of the usefulness of correctional supervision, one of the benefits of which is the scope that it allows for the application of restorative justice.

The more profound aspect of the judgment in S v M, however, lies in its development of child law, and in particular in what the Constitutional Court adds to its previous pronouncements on section 28 of the Constitution2 — that the section is a statement of realisable and enforceable rights. In S v M, the Court pays more detailed attention than previously to the ‘best interests of the child’ principle. The Court also expands for the first time on what the Constitution means when it says that the best interests of the child are of paramount importance in all matters concerning the child. This case note examines the focus in the judgment on the child’s rights as

1 2008 3 SA 232; 2007 12 BCLR 1312 (CC); 2007 2 SACR 539 (CC) (S v M). 2 Constitution of the Republic of South Africa, 1996 (the Constitution).

351 352 Children and their primary caregivers distinct from the rights of their caregivers. It is argued that this distinction presents a new development in the Court’s child rights jurisprudence, and that some of the Constitutional Court’s earlier pronouncements on children in relation to their caregivers should be re-examined in the light of S v M.

What are the duties of a sentencing court in the light of section 28(2) of the Constitution when the person being sentenced is the primary caregiver of minor children? This question was the central pivot of the arguments and judgment in S v M. The case wound its way up to the Constitutional Court by means of a long journey that began in the Wynberg Regional Court. The applicant in the case was M,3 the mother of three minor children. She was the sole caregiver of the children, and was also the main provider of financial support for their care. She had raised a bond on a modest home in which the family lived on the income she derived from two small businesses. She was convicted on various counts of fraud and theft and was sentenced initially by the Regional Court to four years imprisonment.4 On appeal, the High Court confirmed her conviction on 38 counts of fraud and three of theft amounting to a total value of just over R19 000. The High Court set aside her sentence and replaced it with a sentence of four years in terms of section 276(1)(i) of the Criminal Procedure Act 51 of 1977, a sentence which incorporates correctional supervision, subject to an initial portion of the sentence being served in prison. The High Court refused the applicant leave to appeal to the Supreme Court of Appeal, and that Court itself turned down a petition for leave to appeal.5

The effect of the High Court sentence was that the applicant would have had to serve no less than a total of eight months imprisonment before the Commissioner of Correctional Services could have considered releasing her on correctional supervision. When the applicant noted an appeal in the Regional Court against her sentence, that court denied her bail pending the determination of her appeal. Consequently, the applicant began serving her sentence on 29 May 2003. The High Court granted the applicant appeal bail on 25 July 2003, after she had served just under two months of her sentence. She remained on bail at the time when the matter was heard in the Constitutional Court. The fact that she had served two months in 2003 meant that she was in effect facing a period of no less than six months of further imprisonment in terms of the sentence imposed by the High Court.

3 The Constitutional Court issued an order on the day of hearing that the citation of the case name should include only the initial of the applicant’s surname, in order to protect the identity of her children. 4 S v M (n 1 above) para 3. 5 S v M (n 1 above) para 4. (2008) 1 Constitutional Court Review 353

The Constitutional Court decided to enrol her application for leave to appeal. The Chief Justice issued directions in which he posed three questions,6 and required that the parties should confine the scope of their submissions to answering these. In addition to the central question of defining the role of a sentencing court when sentencing a primary caregiver of minor children, the Chief Justice also directed that once the duties had been identified, there should be consideration of whether the duties were observed in this case. The final question posed was that if the Court were to find that the duties were not observed, what order (if any) should the Court make?

The Court had asked the Cape Bar to appoint a curator ad litem for the children, and the curator compiled a comprehensive report for the Court, in addition to making oral submissions. The Centre for Child Law, which had been granted leave to enter the case as amicus curiae, also presented ‘wide ranging written and oral submissions on the constitutional, statutory and social context in which the matter fell to be decided’.7

The applicant, the curator and the amicus all contended that the sentencing court had a duty to give specific and independent consideration to the effect that sentencing a caregiver to imprisonment would have on minor children.8 It was conceded that there would be cases where the best interests of the child would be outweighed by the imperative to sentence a caregiver to imprisonment where considerations of proportionality required that. On the facts of M’s case, however, it was argued that due consideration of the best interests of the children led to the conclusion that the appropriate sentence would be one that did not require M to spend any further time in prison. It was proposed that conditions could be added to the sentence of correctional supervision to redress the impact of the crime on the victims and to recognise the interests of society.

6 S v M (n 1 above) para 5. 7 S v M (n 1 above) para 6. 8 S v M was not the first case to find that a sentencing court had such a duty. In S v Kika 1998 2 SACR 428 (W) the High Court set aside a sentence of a fine which resulted in imprisonment of a primary caregiver, finding that the magistrate had not considered the impact that the sentence might have on the children. The judge caused the circumstances to be investigated, and sent the matter back to the court a quo for the setting of a fresh sentence, pointing out a court must, when sentencing a primary caregiver, be satisfied that the welfare needs of the children will be met. The direct link with section 28(1)(b) and section 28(2) of the Constitution was first articulated in S v Howells 1999 1 SACR 675 (C); 1999 2 All SA 233 (C) affirmed on appeal by the Supreme Court of Appeal in Howells v S 2000 JOL 6577 (SCA). In that case the court considered the effect that imprisonment of a primary caregiver convicted of fraud would have on the children, but nevertheless decided to set a sentence of imprisonment, whilst making ancillary orders to ensure the children would be adequately cared for. 354 Children and their primary caregivers

The response of the National Director of Public Prosecutions, supported in the main by the Department of Social Development and the Department of Justice and Constitutional Development, was that the current operation of the general principles of sentencing provided sufficient room for the consideration of the best interests of the primary caregiver’s children. Their submissions were that the courts a quo in this matter had given adequate consideration to the situation of the children, and that the sentence should not be interfered with.

The majority judgment of the Court was written by Sachs J. Characterised by ringing language and memorable passages, the judgment is the Court’s clearest and most detailed explanation to date of the content and scope of children’s rights as set out in section 28 of the Constitution. The judgment also makes some interesting observations about recent developments in the principles of sentencing, such as the growing recognition of restorative justice,9 and re-affirms the importance and usefulness of correctional supervision as a form of sentence.10 For the sake of completeness, this case note will provide an overview of the judgment, including a brief consideration of the Court’s findings and observations on sentencing. The major part of the note will expand on the Court’s pronouncements on children’s rights, and in particular the Court’s interpretation of the paramount importance of the best interests of the child as stated in section 28(2) of the Constitution.

2 Overview of the judgment

The majority judgment proceeds by delivering responses to the key questions that had been posed by the Court. Expounding on the duties of the sentencing court in the light of section 28(2) of the Constitution when the person being sentenced is the primary caregiver of minor children, the Court begins by considering the current approach to sentencing. Sachs J finds that the Zinn11 triad still retains the status of the ‘sentencing north star’.12 However, this departure point must now be considered in the light of the Constitution, which has transformed the traditional aims of sentencing.13 In the view of the

9 S v M (n 1 above) para 55. 10 S v M (n 1 above) para 58-59. 11 In S v Zinn 1969 2 SA 537 (A) at 540G-H the appellate division described what has to be considered at sentencing as ‘the triad consisting of the crime, the offender and the interests of society’. 12 S v M (n 1 above) para 10 fn 4. 13 S v M (n 1 above) para 10. The court makes reference to the case of Director of Public Prosecutions, KwaZulu-Natal v P 2006 3 SA 515 (SCA); 2006 1 All SA 446 (SCA); 2006 1 SACR 243 (SCA); and Brandt v S 2005 2 All SA 1 (SCA); S v B 2006 1 SACR 311 (SCA). These cases deal with the way in which the Constitution has changed the approach to sentencing of offenders who were below the age of 18 years at the time of the commission of the offence. (2008) 1 Constitutional Court Review 355

Court, the question to be asked was whether the rights of children as set out in section 28 of the Constitution ‘add an extra element’14 to the responsibilities of a sentencing court over and above those imposed by the Zinn triad, and if so, how are those responsibilities to be fulfilled? In order to examine this question properly, the Court turned its attention to the significance of section 28(2) of the Constitution, namely the ‘best interests of the child’ provision.15 The detail of this aspect of the judgment will be discussed below. For the moment it is sufficient to note that the Court reiterates that section 28(2) is a self-standing right, as well as being a guideline for the balancing of other rights. The question is not whether section 28 creates enforceable rights — which clearly it does — but rather ‘what reasonable limits can be placed on their application’?16 In exploring this, the court tackles the issue of what is meant by ‘paramount’. The judgment strives to establish how the paramountcy principle operates in practice, particularly when the child’s rights are being weighed against other important competing rights. The weighing exercise in M’s case was that between the rights of the children not to be separated from parental or family care and to have their best interests considered, and the rights of society to be protected by the punishment of criminal misconduct.

Having established the scope and significance of section 28, the inquiry shifted to what the proper approach of a sentencing court should be where the convicted person is the primary caregiver of minor children. Is it enough to consider the plight of the children as part of the Zinn triad — that part of the triangle dealing with the criminal and his or her circumstances? Counsel for the Director of Public Prosecutions had argued that the principles of sentencing provided sufficient scope to consider the situation of the children, and that the law needed no development. The Court took a different approach, favouring the view of the amicus that a child of a primary caregiver is not a ‘circumstance’ but an individual whose interests needed independent consideration.17 The weight that would be accorded to those interests would depend on the facts of each case, but the Court determined that there was a duty on any court sentencing a primary caregiver to give separate and specific consideration to the impact that the intended sentence would have on the children. The Court took cognisance of information that had been placed before it about the effects on children of the imprisonment of their primary caregivers. Also considered was the international legal framework, the UN Convention on the Rights of the

14 S v M (n 1 above) para 11. 15 Section 28(2) reads as follows: ‘A child’s best interests are of paramount importance in every matter concerning the child’. 16 S v M (n 1 above) para 14. 17 S v M (n 1 above) para 30. 356 Children and their primary caregivers

Child, and in particular the African Charter on the Rights and Welfare of the Child. Article 30(1) of that Charter contains a unique provision dealing expressly with ‘Children of Imprisoned Mothers’. Of particular relevance were clauses calling on states parties to ‘ensure that a non- custodial sentence will always be first considered when sentencing such mothers’; and to ‘establish and promote measures alternative to institutional confinement of such mothers’. Article 30(1) concludes by reminding the states parties that ‘the essential aim of the penitentiary system will be the reformation, the integration of the mother to the family and social rehabilitation’.

The Court held as follows: [F]ocused and informed attention needs to be given to the interests of children at appropriate moments in the sentencing process. The objective is to ensure that the sentencing court is in a position adequately to balance all the varied interests involved, including those of the children placed at risk. This should become a standard preoccupation of all sentencing courts.18

The result of the judgment in S v M, therefore, is that in each case the sentencing court must give specific attention to the impact the sentence will have on the child or children of a primary caregiver. This does not mean that a primary caregiver will never, henceforth, be given a custodial sentence. The judgment explains quite clearly that the choice of the sentencing option least damaging to the interests of the children is made ‘within the legitimate range of choices in the circumstances available to the court’.19 In other words, if a non- custodial option is clearly appropriate, the court must impose such a sentence, bearing in mind the interests of the children. If there is a range of appropriate sentences under consideration, the likely negative impact of imprisonment on the children of the primary caregiver will generally tip the balance in favour of a community- based sentence.

If the court cannot reasonably consider a non-custodial sentence in the circumstances, the court’s responsibilities do not end there. The judgment gives step-by-step guidelines indicating that

if, on the Zinn triad approach the appropriate sentence is clearly custodial and the convicted person is a primary caregiver, the court must apply its mind to whether it is necessary to take steps to ensure that the children will be adequately cared for while the caregiver is incarcerated.20

18 S v M (n 1 above) para 33. 19 As above. 20 S v M (n 1 above) para 36. (2008) 1 Constitutional Court Review 357

Having enumerated the duties, the majority then applied them to the facts of M’s case and decided to overturn the High Court sentence. It was replaced with a sentence that allowed her to serve no further time in prison. The remainder of the sentence was suspended, on various conditions, including correctional supervision for a period of three years, and repayment to the victims of the crime of the monies of which they had been defrauded. In addition to the need to protect the children’s best interests, the fact that M had not committed offences for a long time, and had made demonstrable efforts to be a contributing member of society were strong factors influencing the choice of sentence. The Court pondered whether it should, having decided to set aside the order of the High Court, replace the sentence or refer the matter back for the court a quo to reconsider. It decided to set a new sentence itself, due to the fact that a significant amount of up-to-date and good quality information had been placed before it.

The Court made several interesting observations regarding sentence. The judgment weighs up the advantages of correctional supervision versus a custodial sentence.21 Correctional supervision is described as ‘a multifaceted approach to sentencing comprising elements of rehabilitation, reparation and restorative justice.’22 It is also a sentence that, if applied to those who are likely to respond positively, can protect society without the destructive impact that imprisonment can have on innocent family members. The Court remarked on prison overcrowding, pointing out that rehabilitation is difficult to achieve under such circumstances, and that therefore correctional supervision is a sentence which is more likely to result in rehabilitation.23 The Court went on to say that ‘[a]nother advantage of correctional supervision is that it keeps open the option of restorative justice in a way that imprisonment cannot do’. The court

21 The judgment recalls the Court’s earlier judgment in S v Williams and Others 1995 3 SA (CC), per Langa J as he then was: ‘The development of this process [of correctional supervision] must not be seen as a weakness, as the justice system having “gone soft.”’ 22 S v M (n 1 above) para 59. The Court refers to the following judgments regarding correctional supervision: S v R 1993 1 SA 476 (A); S v Siebert 1998 1 SACR 554 (SCA); S v Ingram 1995 1 SACR 1 (SCA). 23 S v M (n 1 above) para 61. The Court cites S v Lebuku 2006 JOL 17622 (T) para 13 – 15 with approval, where Webster J refers to the 2003/2004 Annual Report of the Inspecting Judge of Prisons in which Judge Fagan recommended the use of non- custodial sentences in appropriate cases, in order to avoid further over-crowding. 358 Children and their primary caregivers explained that central to the notion of restorative justice24 is the recognition that the community, and not the criminal justice system, that is the site of crime control. The sentence handed down by the Court demonstrates a restorative justice lens, as in addition to being confined to her home when not working, and being required to do 10 hours of community service per week, M was required to pay back her victims, which she had offered to do. With regard to this, a particular point is made: It would have special significance if she is required to make the repayments on a face-to-face basis. This could be hard for her, but restorative justice ideally requires looking the victim in the eye and acknowledging wrongdoing.25

Writing for the minority,26 Madala J confirmed the principles decided upon by the majority. He fully agreed that the sentencing court does have a duty to consider the rights of the children when sentencing a primary caregiver. However, applying that approach, Madala J came to a different conclusion. The fact that M was a repeat offender, and that she had committed further offences whilst on bail were factors that, according to the minority, outweighed the children’s interests. The overall judgment is not weaker for the minority judgment. The principles are common to both, although the judges differed on the outcome after all the relevant factors had been considered.

3 Development of Child Law

S v M has developed child law significantly, through its detailed enunciation of the scope and application of section 28. The concept of best interests is put under the spotlight, and the Court spells out the way in which ‘paramountcy’ referred to in section 28(2) should be

24 The court cites the following sources on restorative justice: D Pinnock ‘What kind of justice?’ University of Cape Town, Institute of Criminology Occasional Paper Series 4-95 (1995) available at http://web.uct.ac.zadepts/sjrp/publicat/ whatkind.htm, accessed on 16 August 2007; and M Batley in Maepa (ed) Beyond retribution: Prospects for restorative justice in South Africa Institute for Security Studies, Monograph 111 ‘Restorative justice in the South African context’ 21 http://www.iss.co.za/pubs/Monographs/No111/Chap2.htm (S v M (n 1 above) n 67). See also A M Skelton ‘The influence of the theory and practice of restorative justice in South Africa with special reference to child justice’ unpublished LLD thesis, University of Pretoria, 2005; A Skelton & M Batley ‘Charting progress, mapping the future: Restorative justice in South Africa’ Institute for Security Studies and Restorative Justice Centre (2006). 25 In Dikoko v Mokhatla 2006 6 SA 235 (CC), in the dissenting judgment of Sachs J, at para 114, the key elements of restorative justice are listed as encounter, reparation, reintegration and participation. See also the dissenting judgment of Mokgoro J, particularly at para 68. Since S v M was heard there have been three other judgments reported which expound upon restorative justice, see S v Maluleke 2008 1 SACR 49 (T); S v Shilubane 2008 1 SACR 295 (T); S v Saayman 2008 1 SACR 393 (E). 26 Nkabinde J and Navsa AJ concurred with the dissenting judgment. (2008) 1 Constitutional Court Review 359 understood and applied. The Court acknowledges that the concept of best interests is not new to South African courts, as it has long been used as the applicable standard in matters such as care, contact or maintenance. In the family law field it is everyday currency, its most common application being where adult parties are in dispute about the care of or contact with a child. In such cases, the decisions will always be guided fundamentally by what is in the child’s best interests.

However, in the new constitutional order, ‘the scope of best interests has been greatly enlarged’.27 Section 28(2) declares the best interests of the child to be of paramount importance, not just in family matters, but in all matters concerning the child. At first blush this formulation seems to have such an all-encompassing reach that questions should be asked about how it can be effectively used in practice. To illustrate this, Sachs J quoted from the judgment of Jooste v Botha28 in which Van Dijkhorst J was so concerned that section 28(2) would over-ride all other legitimate interests, that he felt constrained to interpret the provision only as a general guideline and not as a rule of law of horizontal application. Sachs J points out that Van Dijkhorst J was wrong about this. He explains that while section 28 undoubtedly serves as a general guideline to the courts, its ‘normative force’ does not stop there.29 In fact, the Constitutional Court has previously described section 28(2) as an independent, self- standing right,30 and as ‘an expansive guarantee’.31 Sachs J sums up the unequivocal position of the Court: section 28 establishes a set of children’s rights that courts are obliged to enforce.32

This leads to another question: If the ambit of children’s rights is so wide, in what circumstances can such rights be limited? It is a question that has been pondered by academics for a number of years.

27 S v M (n 1 above) para 12. 28 2000 2 SA 199 (T). 29 S v M (n 1 above) para 14. 30 In Minister of Welfare and Population Development v Fitzpatrick and Others 2000 3 SA 422 (CC); 2000 3 BCLR 713 (CC) per Goldstone J: ‘The plain meaning of the words clearly indicates that the reach of section 28(2) cannot be limited to the rights enumerated in section 28(1) and section 28(2) must be interpreted to extend beyond those provisions. It creates a right that is independent of those specified in section 28(1).’ 31 Sonderup v Tondelli 2001 1 SA 2001 (CC); 2001 2 BCLR 152 (CC) (Sonderup) para 29. 32 S v M (n 1 above) para 14. 360 Children and their primary caregivers

Clark,33 writing in 1998, took the view that children’s rights would trump opposing rights. However, her comments related to family law, where the best interests principle is considered the determining factor. Bekink and Bekink considered the matter more broadly, arguing that ‘paramount’ means that in weighing up competing interests, the scales must tip in favour of the child.34

In the case of De Reuck v Director of Public Prosecutions, Witwatersrand Local Division and Others35 the Constitutional Court provided some clarity on the issue. In the High Court decision in the same matter, the Court had found that children’s best interests can never be trumped by the rights of others.36 The Constitutional Court disagreed, reiterating its approach that constitutional rights are mutually interrelated and interdependent, forming a single constitutional value system. Citing its earlier judgment in Sonderup,37 the Constitutional Court held that section 28(2), like other rights enshrined in the Bill of Rights, is subject to limitations that are reasonable and justifiable in accordance with section 36.

4 What is the meaning of ‘paramount importance’ of children’s best interests?

The question that remained hovering in the air even after the Constitutional Court had ruled in De Reuck was: What is to be understood by the phrase ‘paramount importance’? Friedman and Pantazis38 posed the question thus: ‘Indeed, if a child’s best interests are not always supreme, what is the point of section 28(2)?’

The authors went on to observe that section 28(2) is highly unusual, because it is the only section in the Constitution that applies to a group of people in relation to all aspects of their lives. In their view the section appears to be aimed at creating a right for children

33 B Clark ‘Competing custody rights: New concepts of family and the best interests of the child’ (1998) 35 Comparative and International Law Journal of South Africa 288 290. See further J Heaton ‘Some general remarks on the concept “best interests of the child”’ (1990) 53 Tydskrif vir Hedendaagse Romeins Hollandse Reg 96; B Clark ‘A golden thread? Some aspects of the application of the standards of the best interests of the child in South African family law’ (2000) Stellenbosch Law Review 3; F Mahlobogwane ‘South African courts and the “best interests of the child” in custody disputes’ (2005) 38 Comparative and International Law Journal of South Africa 246. 34 B Bekink & M Bekink ‘Defining the standard of the best interest of the child: Modern South African perspectives’ (2004) 37 De Jure 21 26. 35 2004 1 SA 406 (CC); 2003 12 BCLR 1333 (CC) para 55. 36 De Reuck v Director of Public Prosecutions, Witwatersrand Local Division 2003 3 SA 389 (W) (De Reuck) para 45. 37 2001 1 SA 1171 (CC); 2001 2 BCLR 152 para 27-30. 38 A Friedman & A Pantazis ‘Children’s rights’ in S Woolman et al (eds) Constitutional law of South Africa (2nd Edition, OS, 2006) 47-34 - 47-35. (2008) 1 Constitutional Court Review 361 as children — addressing the vulnerability of children, and ensuring that their rights do not, as in the pre-constitutional era, frequently have to give way to the rights of others. Friedman and Pantazis concluded that section 28(2) implies that in every matter where a child’s rights are (substantially) involved, those interests must be taken into account. In addition, they said that a child’s interests have a leg up vis-à-vis other rights and values, though this does not amount to these rights necessarily acting as trumps.39 The judgment in S v M seems to have proved them right.

Sachs J states that what lies at the heart of section 28 is the right of the child to be a child and enjoy special care. In a poetic passage of the judgment, he explains what he means by this:

Every child has his or her own dignity. If a child is to be constitutionally imagined as an individual with a distinctive personality, and not merely as a miniature adult waiting to reach full size, he or she cannot be treated as a mere extension of his or her parents, umbilically destined to sink or swim with them. The unusually comprehensive and emancipatory character of section 28 presupposes that in our new dispensation the sins and traumas of fathers and mothers should not be visited on their children. Individually and collectively all children have the right to express themselves as independent social beings, to have their own laughter as well as sorrow, to play, imagine and explore in their own way, to themselves get to understand their bodies, minds and emotions, and above all to learn as they grow how they should conduct themselves and make choices in the wide social and moral world of adulthood. And foundational to the enjoyment of the right to childhood is the promotion of the right as far as possible to live in a secure and nurturing environment free from violence, fear, want and avoidable trauma.40

This forms the crux of the Court’s decision that sentencing courts do have a duty to consider the rights of the children when sentencing the primary caregiver. The state’s action — imprisonment of an adult — causes the removal of the caregiver from the family home. This directly impacts on the child’s right to family and parental care. The Court went on to say that although no constitutional injunction can completely protect children from life’s hard knocks, the law can create conditions to protect them from abuse or neglect. In S v M, the Court defined the legal duty as ensuring that the best interests of the child are considered by the sentencing court wherever there is a likelihood that the primary caregiver will be removed from the family through imprisonment. The likely effect, where the range of available sentencing options permits it, is to give children’s rights ‘a leg up’ by keeping the caregiver in the family. Where a custodial option is considered by the sentencing court to be the only legitimate course,

39 Friedman & Pantazis (n 38 above) 47-35. 40 S v M (n 1 above) para 18. 362 Children and their primary caregivers then the court must nevertheless be satisfied that the children will be properly cared for and may make orders to minimise the negative effects on the children.

If children’s rights do not always win out over other rights, what is the meaning of ‘paramount importance?’ The judgment embarks on an exploration of this thorny question. Sachs J comments that the very expansiveness of the paramountcy principle appears to promise everything but deliver little in particular.41 The best interests concept is indeterminate, resulting in differing views about its meaning by professionals and judges. These problems accepted, the Court has recognised that it is precisely the contextual nature and inherent flexibility of section 28 that constitutes the source of is strength. The determination will depend on the circumstances of each case, and this is not a weakness, but a strength. A truly child-centred approach requires an in-depth consideration of the needs and rights of the particular child in the ‘precise real-life situation’ he or she is in. To apply a pre-determined formula for the sake of certainty, irrespective of the circumstances, would in fact be contrary to the best interests of the child.42

The exercise of weighing up the best interests of the child has been clearly articulated by the Court. S v M has built on the Court’s previous jurisprudence,43 creating a nuanced model that can be applied in a range of different legal and factual contexts. A more difficult question, concedes Sachs J, is to ‘establish an operational thrust for the paramountcy principle’. S v M goes further than any previous judgment, though it still defines the principle more by stating what it is not, than by saying what it is. It is not an ‘overbearing and unrealistic trump’,44 it cannot be interpreted ‘to mean that the direct or indirect impact of a measure or action on children must in all cases oust or override all other considerations.’ Sachs J concludes that ‘the fact that the best interests of the child

41 S v M (n 1 above) para 23. See also H Reece ‘The paramountcy principle: Consensus or construct?’ (1996) 49 Current Legal Problems 267; D Matlala ‘The law reports: Parent and child — paramountcy of a child’s best interests’ (2006) June De Rebus 41. 42 The Court has re-stated this position in the subsequent judgment of AD and Another v DW and Others (Centre for Child Law as Amicus Curiae; Department for Social Development as Intervening Party) 2008 3 SA 183 (CC), per Sachs J para 55: ‘Child law is an area that abhors maximalist legal propositions that preclude or diminish the possibilities of looking at and evaluating the specific circumstances of the case ... This means that each child must be looked at as an individual, not as an abstraction. It also means that unduly rigid adherence to technical matters, such as who bears the burden for proof, should play a relatively diminished role; the courts are essentially guarding the best interests of a child, not simply settling a dispute between litigants.’ 43 Minister of Welfare and Population Development v Fitzpatrick and Others 2000 3 SA 422 (CC); De Reuck (n 35 above); Sonderup (n 30 above). 44 n 1 above, para 25. (2008) 1 Constitutional Court Review 363 are paramount does not mean that they are absolute.’45 To acknowledge all of these realities is important because if the best interests principle is spread ‘too thin’ it risks becoming devoid of meaning, instead of promoting the rights of children as it was intended to do.46

Later in the judgment the operation of the principle is explained in a more positive manner. The judgment states that sentencing officers should pay appropriate attention to the children of a primary caregiver and take reasonable steps to minimise damage. It continues thus:

The paramountcy principle, read with the right to family care, requires that the interests of children who stand to be affected receive due consideration. It does not necessitate overriding all other considerations. Rather, it calls for appropriate weight to be given in each case to a consideration to which the law attaches the highest value, namely the interests of children who may be concerned.47

5 The child’s rights separate from adult’s rights: a subtle shift

A further important feature of the majority judgment in S v M is the fact that it carefully focuses on the best interests of the child, and does not get entangled in a debate about the issues relating to the rights of the primary caregiver. The discourse centres on children’s rights to family and parental care, and their right to have their best interests given appropriate weight. Counsel for the National Director of Public Prosecutions attempted to argue that if the sentencing court considers children’s rights, then primary caregivers will gain an unfair advantage in the sentencing process. The majority of the Court rejected this out of hand, saying that ‘the issue is not whether parents should be allowed to use their children as a pretext for escaping the otherwise just consequences of their own misconduct.’48 That, said the Court, would be a mischaracterisation of the interests at stake. The purpose of emphasising the duty of the sentencing court to consider the best interests of the child is not to allow errant parents to escape punishment, but rather to protect innocent children from avoidable harm.

45 S v M (n 1 above) para 26. 46 S v M (n 1 above) para 25. 47 S v M (n 1 above) para 42. 48 S v M (n 1 above) para 35. See, however, Madala J’s dissenting judgment at para 117, where he states that the Court must guard against creating a perception that encourages offenders to use the interests of children as ‘a tool in the judicial process’. 364 Children and their primary caregivers

This focus on the child’s rights separate from the rights of their caregivers heralds a subtle but significant shift in the Constitutional Court’s child rights jurisprudence. Read together with the strong statements earlier in the judgment about children being possessed of their own dignity, and being separate beings who are not merely destined to sink or swim with their parents, the Court’s ‘new’ enunciation of children’s rights calls for a re-examination of some of its earlier pronouncements.

The High Court decision in Grootboom v Oostenberg Municipality & Others49 concerned an application brought by a group of 390 adults and 510 children (all of whom were living with their parents) who had not been provided with adequate housing and who, at the time of the initial application, were living on a sports field, following their eviction from land destined for low-cost housing. The applicants succeeded in their argument that because children have an unqualified right to shelter in terms of section 28(1)(c), coupled with the fact that they have a right to be cared for by their families or parents, children had the right to be provided with shelter (together with their parents) until such time as their parents could provide them with shelter. Davis J was careful to point out that the children were the bearers of the right. The fact that they should not be separated from their parents in order to access their rights did not amount to a derivative right for the parents (as had been argued by the applicants), but rather that ‘an order that enforces a right to shelter should take account of the need of the child to be accompanied by his or parent’.50

The approach of the majority in S v M is reminiscent of this way of thinking. The similarities are that in both cases the judgments make it clear that the rights accrue to the child, but because of the imperative to keep children in their families wherever possible, the adult caregiver might also benefit from the required order.

However, the Constitutional Court did not uphold the judgment of Davis J. In Government of the Republic of South Africa & Others v Grootboom & Others,51 the Court found that section 28(1)(c) did not create any primary obligations on the State to provide shelter on demand to children and their parents, but rather that the primary obligation to care for children lay with their parents, and that the state only has the obligation in the alternative to provide shelter

49 2000 3 BCLR 277 (C) (Grootboom 1). 50 Grootboom 1 (n 48 above) 289D. 51 2001 1 SA 46 (CC); 2000 11 BCLR 1169 (CC) (Grootboom 2). (2008) 1 Constitutional Court Review 365 when, for example, children are removed from their families.52 The Court found that responsibilities of the state to children in the care of their families extended to providing the legal and administrative infrastructure to ensure that children are protected in terms of section 28(1). Section 28(1)(c), although it contains no internal qualifier suggesting that it is subject to progressive realisation, does not, according to Grootboom, provide children with a direct and immediately enforceable right to housing.53 The Court was concerned that ‘[c]hildren could become stepping stones to housing for their parents instead of being valued for who they are’. This statement jars with the fact that children in South Africa receive other poverty or care-related benefits, such as child support grants,54 and there is no serious suggestion that this might lead to them being valued for the money that their presence brings into the household rather than for who they are.55

In S v M the majority of the Court does not get distracted by any ideas that primary caregivers might use their children as ‘stepping stones’ to avoid imprisonment, although this kind of argument was advanced by counsel for the National Director of Public Prosecutions. The minority judgment in S v M does, however, get mired in this way of thinking. Although Madala J agrees with the majority on the principle that the children of primary caregivers should have their best interests considered and protected, he is concerned about children being ‘used’ by adults seeking to avoid imprisonment. He cautions that ‘[t]his Court should be wary of setting a precedent that creates a perception that courts will give primary caregivers a sentence that is disproportionate to what they deserve and which encourages them to use the interests of children as a tool in the judicial process.’56 These remarks, in contrast to the explanation in the majority judgment that the children are the bearers of the rights

52 Grootboom 2 (n 51 above) para 77. The judgment did not state clearly that children living with their parents will not have a claim for the enforcement of their rights in terms of section 28(1)(c), as it uses the term ‘for example’ when referring to children removed from their primary caregivers. 53 The Court came to the puzzling conclusion that there was no difference between the right to ‘adequate housing’ provided for in section 26 and ‘shelter’ in section 28. Friedman & Pantazis (‘Children’s Rights’ in Woolman et al (n 37 above) 46-8) have criticised this interpretation by the court, firstly because it ignores the plain meaning in the two different phrases, and also because the Constitutional drafters’ choice of different terminology suggests different conceptual extensions. 54 In the case of Khoza and Others v Minister of Social Development and Others; Mahlaule and Another v Minister of Social Development and Others 2004 6 BCLR 569 (CC) the Court recognised that denial of support to South African born children of foreign permanent residents in South Africa was not only discrimination on the grounds of the parents’ nationality, but also trenches on the rights that children have under section 28(1)(c). 55 Friedman & Pantazis (n 38 above) 47-6 n 2: ‘In any case, the state is committed to providing all needy children with child care grants.’ 56 S v M (n 1 above) para 117. 366 Children and their primary caregivers and to suggest otherwise is a mischaracterisation, strike a chord that is an echo of children being used as ‘stepping stones’.

Much disappointment has been expressed with regard to the Constitutional Court’s decision in Grootboom, viewed from a children’s rights perspective.57 However, it has also been acknowledged by several writers58 that the effect of the decision in Grootboom — vis-à-vis children’s rights — underwent some adjustment in the case of Minister of Health & Others v Treatment Action Campaign & Others.59 In that case, which dealt with access to treatment to avoid mother-to-child transmission of HIV/AIDS, the Court held as follows:

The state is obliged to ensure that children are accorded the protection contemplated by s 28 that arises when the implementation of the right to parental or family care is lacking. Here we are concerned with children born in public hospitals and clinics to mothers who are for the most part indigent and unable to gain access to private medical treatment which is beyond their means.60

So the approach adopted in Grootboom to the effect that children living with their parents will have to look to their parents, rather than the state, for the fulfilment of their section 28(1)(c) rights was ameliorated.61 Although the two cases are clearly distinguishable by the fact that shelter is arguably easier for parents to provide than a specialised form of health care, it was significant that the TAC Court

57 J Sloth-Nielsen ‘The child’s right to social services, the right to social security, and primary prevention of child abuse: Some conclusions in the aftermath of Grootboom’ (2001) 17 South African Journal on Human Rights 210; E Bonthuys ‘The South African bill of rights and the development of family law’ (2002) 119 South African Law Journal 748; M Pieterse ‘Reconstructing the private/public dichotomy? The enforcement of children’s constitutional social rights and care entitlements’ (2003) Tydskrif vir Suid-Afrikaanse Reg 1; S Liebenberg ‘Taking stock: The jurisprudence of children’s socio-economic rights and its implications for government policy’ (2004) 5:4 ESR Review 2; and K McLean ‘Housing’ in S Woolman et al (eds) Constitutional law of South Africa (2nd Edition, OS, 2006) 55.51 – 55.54. 58 J Sloth-Nielsen ‘Children’ in D Davis & H Cheadle (eds) The South African constitution: The bill of rights (2nd ed [date]) 421; Friedman & Pantazis (n 38 above) 47-13; E Bonthuys ‘Children’ in I Currie & J de Waal (eds) The bill of rights handbook (5th edition [date]) 613. 59 2002 1 BCLR 1033 (CC) (TAC). 60 TAC (n 59 above) para 79. 61 S Liebenberg (‘The judicial enforcement of social security rights in South Africa’ in E Riedel Social Security as a Human Right (2007) 69) points out that the Court did not conclude that children enjoyed an unqualified, direct claim to the provision of basic health care services, but rather based their finding on the fact that the government’s policy was unreasonable because it excluded a particularly vulnerable group. (2008) 1 Constitutional Court Review 367 gave clarity that section 28(1)(c) rights do not only set an obligation for the state where the child is separated from the parent.62

6Conclusion

S v M appears to herald a new era in which the fear of children being used as instruments by their parents to advance their parents’ rights have been supplanted by a clear recognition of children as rights bearers. Comparisons have been adduced in this article to show that this approach may indicate a subtle shift by the Court to see children as separate from their parents and therefore not ‘umbillically destined to sink or swim with them’.

There are, of course, significant differences between cases such as Grootboom and TAC on the one hand, and S v M on the other. The first-mentioned cases deal with socio-economic rights of children. This means that the orders given in those matters potentially had significant cost implications. S v M does not create major cost implications, it requires only an investment of time and care on the part of state officials already employed. In such cases it is clearly easier for the Court to make the kind of pronouncements that it has made in S v M. Grootboom and TAC can themselves be separated out from one another. Health care is not something that parents can ever fully provide unless assisted either by private health care providers, or if the parents are poor, by the state. Although the judgments do not fully explain the difference, commentators have mentioned that nutrition and shelter are more likely to be within the scope of what parents can be expected to deliver, whilst health care and social services are more likely to require input by the state.63

Children of primary caregivers who are due to be sentenced pose an interesting counter-point. They are still living with their parents, but are at risk of being separated from them. S v M found that the state has a duty to create a positive legal environment, and thus sentencing courts now have clear duties that have been articulated by the Court. These arise from the rights of the children, and the majority judgment had no difficulty in dispelling any concerns about

62 The jurisprudence related to the socio-economic rights of children who are separated from their parents has gradually developed, with the courts finding that such children do have an immediate and directly enforceable claim against the state for the rights included in section 28(1)(h): see Centre for Child Law and Another v Minister of Home Affairs and Others 2005 6 SA 50, which dealt with the rights of unaccompanied foreign children to receive social services, and Centre for Child Law and Others v MEC for Education, Gauteng and Others 2007 1 SA 223 (T), which dealt with children who had been removed from state care and placed in a school of industries where they were living in extremely poor conditions, without adequate social services. 63 Friedman & Pantazis (n 38 above) 47-12; J Sloth-Nielsen (n 56 above) 225. 368 Children and their primary caregivers the primary caregivers reaping unfair advantages from the fact that they had children. The only distinction that can be made is that S v M deals with the right to family or parental care, and does not deal directly with socio-economic rights of the children.

The effect of S v M is ultimately to force the state to create a legal environment to facilitate children’s rights being recognised and their best interests being met as far as possible through the orders of sentencing courts. To that extent, the judgment accords with the role of the state as outlined in Grootboom, which was to create the regulatory environment for the achievement of children’s rights. However, in the way that it ‘constitutionally imagine[s]’64 children as separate rights bearers with separately identifiable and enforceable rights, S v M takes children’s rights a jurisprudential step forward.

64 S v M (n 1 above) para 18.