+(,121/,1(

Citation: 10 S. Afr. J. on Hum. Rts. 31 1994

Content downloaded/printed from HeinOnline

Wed Feb 15 03:55:04 2017

-- Your use of this HeinOnline PDF indicates your acceptance

of HeinOnline's Terms and Conditions of the license
agreement available at http://heinonline.org/HOL/License

-- The search text of this PDF is generated from

uncorrected OCR text.

-- To obtain permission to use this article beyond the scope

of your HeinOnline license, please use:

Copyright Information
A BRIDGE TO WHERE?
INTRODUCING THE INTERIM
BILL OF RIGHTS

ETIENNE MUREINIK*

INTRODUCTION

What is the point of our Bill of Rights? The Bill is Chapter 3 of the interim
Constitution,1 which declares itself to be aspiring to be 'a historic bridge
between the past of a deeply divided society characterized by strife,
conflict, untold suffering and injustice, and a future founded on the
recognition of human rights, democracy and peaceful co-existence and
development opportunities for all South Africans, irrespective of colour,
race, class, belief or sex'. 2 If this bridge is successfully to span the open
sewer of violent and contentious transition, those who are entrusted with
its upkeep will need to understand very clearly what it is a bridge from,
and what a bridge to.

BSc LLB (Witwatersrand) BCL (Oxon), Professor of Law, University of the Witwatersrand,
Johannesburg.
1 Constitution of the Republic of South Africa Act 200 of 1993.
The interim Constitution emerged from the Multi-party Negotiating Process (MPNP) at the World
Trade Centre in Kempton Park during 1993. Political approval of the Constitution was the
responsibility of a body called the Negotiating Council, on which all the political parties participating
in the MPNP were represented. What the Council approved was enacted by Parliament, with
relatively minor amendments. The initial drafting was the responsibility of two committees: the
Technical Committee on Constitutional Issues and the Technical Committee on Fundamental Rights
during the Transition. The latter was responsible for Chapter 3 - the Bill of Rights - and the former
for the rest of the Constitution. This division of labour is visible in the final product. Thus, for
instance,
(a) Chapter 3 declares itself binding only on all legislative and executive organs of state (s 7(1)); but
Chapter 1 declares the Constitution as a whole to be binding on all legislative, executive and judicial
organs of state (s 4(2)); and
(b) Chapter 3 gives '[alny superior court' competence to inquire into the validity of the declaration
of a state of emergency and any action taken under it (s 34(3)), questions which plainly engage the
interpretation, protection and enforcement of the Constitution; but by Chapter 7 the Appellate
Division, a superior court, is deprived of jurisdiction over any matter within the jurisdiction of the
Constitutional Court (s 101(5)), and the latter court is given jurisdiction over 'all matters relating to
the interpretation, protection and enforcement' of the Constitution (s 98(2)).
Because of dissatisfaction with some of the earlier drafts of the Technical Committee on
Fundamental Rights during the Transition, an Ad Hoc Committee, consisting of lawyers representing
certain parties on the Negotiating Council, was formed. The Ad Hoc Committee had a hand in
redrafting some of the clauses in the Bill.
2 These words appear in the postamble to the Constitution, entitled 'National Unity and Reconcilia-
tion'.
 SOUTH AFRICAN JOURNAL ON HUMAN RIGHTS

What the bridge is from is a culture of authority. Legally, the apartheid

order rested on the doctrine of Parliamentary sovereignty. Universally,
that doctrine teaches that what Parliament says is law, without the need
to offer justification to the courts. In South Africa, since Parliament was
elected only by a minority, the doctrine taught also that what Parliament
said was law, without a need to justify even to those governed by the law.
The effect of these teachings, at the apogee of apartheid, was to foster an
ethic of obedience. The leadership of the ruling party commanded
Parliament, Parliament commanded its bureaucracy, the bureaucrats
commanded the people.
These teachings were stronger than the legal order they supported.
Obedience spilled over from law into practice. Apartheid policies were
implemented by governmental authorities, 3 private organizations and
individuals not required by law to do so. The consequence was a social
monolith, with no effective counterweights to government power.
This culture of authority was an indispensable nursery for the feature
of apartheid that most people would consider its defining characteristic:
the reduction to law of racial discrimination - differentiation on the
ground of race that is not justified. 4 Without a culture of authority it is
difficult to imagine how its gardeners could have cultivated the forest of
apartheid statutes whose most distinctive feature was their want of
justification.
If the new Constitution is a bridge away from a culture of authority, it
is clear what it must be a bridge to. It must lead to a culture of justification
- a culture in which every exercise of power is expected to be justified;
in which the leadership given by government rests on the cogency of the
case offered in defence of its decisions, not the fear inspired by the force
at its command. The new order must be a community built on persuasion,
not coercion.
If the Constitution is to be a bridge in this direction, it is plain that the
Bill of Rights must be its chief strut. A Bill of Rights is a compendium of
values empowering citizens affected by laws or decisions 5 to demand
justification. If it is ineffective in requiring governors to account to people
governed by their decisions, the remainder of the Constitution is unlikely
to be very successful. The point of the Bill of Rights is consequently to
spearhead the effort to bring about a culture of justification. That idea
offers both a standard against which to evaluate Chapter 3 of the interim

3 Such as local authorities and statutory bodies.

4 Racial discrimination is not a descriptive term, connoting any differentiation on the ground of race,
or even any disadvantage inflicted on account of race. If it were, it would be used to encompass racial
preferences that remedy the effects of racial disadvantage even by people who approve of such
preferences; but supporters of such preferences are anxious to distinguish them from racial
discrimination. Racial discrimination is a term of censure, used to condemn racial differentiation that
is not justified.
5 After this, the word 'decisions' will be used to embrace both laws and other kinds of decisions that
may be brought under scrutiny by the Bill of Rights. A law is of course a kind of decision.
INTRODUCING THE INTERIM BILL OF RIGHTS

Constitution and a resource with which to resolve the interpretive

questions that it raises.

THE TEXT

Structure of the Inquiry: Limitation

The design of the Bill of Rights makes it plain that the rights which it
proclaims are standards of justification - standards against which to
measure the justification of the decisions challenged under them. The
rights in the Bill are not framed as absolutes, with which anything that
collides is automatically annulled. Each of the rights is capable of
restriction under s 33, the provision - known as the limitation clause -
which determines conditions under which limitation of the right is
justified. So a challenge under the Bill cannot be resolved merely by
finding the decision challenged to be in conflict with the idea represented
by the right announced in the Bill. A challenge under the Bill opens up an
inquiry into the justification of the decision challenged.
In at least two ways, however, it may be thought that the Bill's method
of opening decisions to an inquiry into their justification is somewhat too
elaborate.
First, s 33 creates no fewer than three standards of scrutiny under which
to test limitations upon rights in the Bill. The least protected category of
rights may be limited by law of general application to the extent that the
limitation is reasonable, justifiable in an open and democratic society
based on freedom and equality, and does not negate the essential content
of the right in question. 6 A limitation on a right in the most protected
category has to satisfy those criteria and also be 'necessary'. 7 Between
these two categories lies an intermediate and hybrid category. So far as
the rights listed in it relate 'to free and fair political activity', they enjoy
the highest level of protection; so far as they do not, they attract only the
8
lowest.
The wisdom of dividing the rights in the Bill into these three categories
will no doubt prove to be one of its most controversial features. Three
categories is both too few and too many. It is too few because the rich
spectrum of considerations which ought to be brought to bear in deciding
what constitutes a legitimate limitation to each of the well over a hundred
distinct rights that can be identified in the Bill (in the thousands of
contexts to which they will be applicable) cannot be compressed into a
triad. And it is too many because some of the rights which appear higher

6 Section 33 (1).
7 Section 33(1)(aa).
8 Section 33(l)(bb).
 SOUTH AFRICAN JOURNAL ON HUMAN RIGHTS

in this trilaminar system cannot be distinguished, in the abstract, from

others which appear lower. The work of identifying the justified
limitations on the Bill's rights calls for a supple and nuanced jurispru-
dence, not a rule of thumb, pinkie and index finger. The three categories
are mechanical and arbitrary.

Freedom of Speech
Possibly the most arbitrary categorization is the one which assigns
freedom of speech and expression, including freedom of the press and the
other media 9 to the intermediate category of protection. It goes without
saying that freedom of speech is foundational to democracy. In the
United States it is often said that it is no accident that free speech is
protected by the First Amendment to the Constitution: it is primary in
much more than a numerical sense. And if free speech is foundational to
an established and vigorous democracy, it is doubly so to a culture of
justification which is struggling for viability. Open debate is indispensable
to that aspiration. Without freedom of speech, there can be no ethic of
persuasion, only one of coercion.
One feature of the Bill of Rights which makes this categorization
particularly arbitrary is the decision of the drafters to put the freedoms of
conscience, thought, belief and opinion10 in the most protected category.
Freedom of speech and expression merge seamlessly with those free-
doms, and it is difficult to see why a line should be drawn between them,
or how it can be. What is the point of protecting freedom of thought if one
is unfree to speak one's thoughts? And is opinion really being protected
if it cannot be expressed? What is more, the freedoms of conscience,
thought, belief and opinion include 'academic freedom in institutions of
higher learning', and the freedoms of speech and expression include the
'freedom of artistic creativity and scientific research'. Freedom to
undertake scientific research is usually exercised as part of academic
freedom in an institution of higher learning, so this ordering of these
freedoms seems to blur even further the line between what is protected
less and what is protected more.
The danger is that a court may read the hybrid protection of free speech
-greater protection of speech related to free and fair political activity,
lesser protection of any other kind of speech - as disclosing an intention
liberally to permit state interference with press scrutiny of evils such as
labour exploitation, defective consumer products, threats to the environ-
ment, professional malpractice, perhaps even routine maladministration.

9 Section 15(1), which reads: 'Every person shall have the right to freedom of speech and expression,
which shall include freedom of the press and other media, and the freedom of artistic creativity and
scientific research.'
10 Section 14(1), which reads: 'Every person shall have the right to freedom of conscience, religion,
thought, belief and opinion, which shall include academic freedom in institutions of higher learning.'
INTRODUCING THE INTERIM BILL OF RIGHTS

The consequences of such a reading for openness and accountability

would be disastrous.
What is even more puzzling about the relegation of speech to the
intermediate category of protection is that, in another respect, the
drafters of the Bill evinced acute awareness of the importance of
cultivating an ethic of persuasion. Section 15(2) requires all media
financed by or under the control of the state to be regulated in a manner
'which ensures impartiality and the expression of a diversity of opinion'. 1
This is not a provision routinely included in constitutional clauses which
protect freedom of speech, and it reflects a more than ordinary level of
concern to enrich and unshackle public debate; a level of concern which
it is not easy to square with giving freedom of speech only second order
protection.

No Absolute Rights: Detention without Trial

The second respect in which the provision in the Bill for opening decisions
to an inquiry into their justification may be thought too elaborate is that
all the rights in the Bill invite such an inquiry, and none are absolute, not
even the rights to freedom from torture12 and servitude. 1 3 It is to be hoped
that that does not generate a jurisprudence of justifiable limitations on
freedom from torture and slavery. Much the same concern arises in
relation to non-emergency detention without trial.
Detention without trial is arguably the polar opposite of a culture of
justification. Detention without trial means that what many cherish as
their most important liberty, freedom of the person, may be removed
without a most basic kind of justification - proof of criminal conduct
established in a court of law. Despite that, most constitutional democrats
accept that a democratic order, or a democratizing one, may, in the face
of a genuine emergency, be forced by way of self-preservation to resort
to detention without trial. In the face of such a necessity, constitutional
democrats generally retreat from insisting that incarceration be justified
in a criminal court to a more modest kind of justification: they require
only that the genuineness of the emergency and the necessity for the
action taken under it be demonstrable.
Detention without trial under a state of emergency is provided for in
s 34. In a carefully drafted provision, the drafters have brought
emergency detention under the discipline of compulsory justification in
several ways. For one thing, the declaration itself of a state of emergency

11 This is an adaptation of Art 6(2) of the Democratic Party's Draft Bill of Rights (May 1993), which
requires the state, in respect of 'the exercise of its control, if any, over any public media... [to]
ensure diversity of expression and opinion'.
12 Section 11(2).
13 Section 12. The Afrikaans text renders 'servitude' as 'slawerny', the most obvious translation of which
is slavery.
 SOUTH AFRICAN JOURNAL ON HUMAN RIGHTS

expires after 21 days unless it is renewed by a two-thirds majority in the

National Assembly for three months at a time. 14 The effect of that would
ordinarily be to require the government to justify the declaration of a
state of emergency to at least one party other than the ruling party. 15 The
necessity, moreover, of both the declaration and anything done under it
is justiciable before any superior court. 16 And the necessity of a detention
is specifically made justiciable. 17 The combined effect of these protec-
tions is to put the government under strong obligations to justify its
actions to bodies independent of it - to justify the general necessity of the
state of emergency in court and to a supermajority in Parliament, and to
justify the particular necessity of any detention in court. These
obligations are strengthened by a constitutional duty on the state to
supply written reasons to justify any detention.' 8
It would be incongruous to bring emergency detention under so
rigorous a r6gime of justification, but continue to permit political
detention without trial outside a state of emergency. It would make no
sense to require the need for political detention without trial to be shown
by demonstrating the genuineness of the emergency to bodies indepen-
dent of the government, but then permit such detention without requiring
that need to be demonstrated. The corollary is that the denial of personal
freedom outside a demonstrably genuine emergency ought generally to
be capable of justification only by adducing the evidence necessary to
procure a criminal conviction. To that general corollary there may have
to be exceptions, to provide, say, for detention preparatory to extradi-
tion, or to secure the attendance of the accused at a criminal trial, or
prevent the spread of infectious diseases, or confine those whose mental
illness makes them dangerous. These kinds of cases are commonly
catered for by specific exceptions, 19 and a constitution which valued
freedom of the person highly would enact mechanisms for ensuring that
they could be invoked only when their application was justified.
The Bill of Rights does disclose a high degree of concern for freedom
of the person. Section 11 (1), which governs the position outside of a state
of emergency, is not content merely to entrench that freedom; it goes on

14 Section 34(2).
15 This constraint would obviously lose most of its value if the ruling party, on its own, commanded a
two-thirds majority in the National Assembly. But that kind of preponderance would in any event
bring the ruling party within striking distance of the two-thirds majority of a joint sitting of the
National Assembly and the Senate necessary to amend or rewrite most of the Constitution itself
(ss 62 and 73). No constitutional device can help an electorate which is determined not to protect itself
against tyranny.
16 Section 34(1), (3) and (4).
17 Section 34(6)(c).
18 Section 34(6)(g).
19 See, for example, Art 5(1) of the European Convention on Human Rights.
INTRODUCING THE INTERIM BILL OF RIGHTS

to outlaw detention without trial by name. 20 But the value of this strong
stand against non-emergency detention is put at some risk by the failure
to make it absolute, subject to whatever exceptions might be thought
indispensable. Like all the rights in the Bill, those protected by s 11(1) are
limitable. What is more, in two other sections, the Bill seems to envisage
non-emergency detention. Section 25 lists the rights of detainees
separately from those of arrested and accused persons, and s 30(2)
requires a child who is in detention to be treated in a manner that takes
account of his or her age.
Against these features, however, must be set the consideration that the
right in s 11(1) to freedom from non-emergency detention is included in
the category of rights most protected under the limitation clause. That
means that it cannot be limited except by law of general application which
(a) is reasonable and (b) is necessary and (c) is justifiable in an open and
democratic society based on freedom and equality and (d) does not
negate the essential content of the right in question. The right in question
is the right not to be detained without trial. It is difficult to imagine a
political detention law which could satisfy all these requirements. For one
thing, it would struggle to avoid negating the essential content of the right
not to be detained without trial. For another, it is difficult to imagine how
such a law could ever be necessary, since the Bill makes separate
provision for necessity - it caters fully in s 34 for a state of emergency.
From all of which it seems that, on a proper interpretation of the Bill,
s 11(1) outlaws non-emergency political detention altogether. If to that
the objection is raised that s 25 and s 30(2) appear to envisage
non-emergency detention, the answer must be that these provisions are
to be read as catering for the exceptional categories of non-emergency
detention mentioned above, not for non-emergency political detention.
On a proper construction of the Bill of Rights, therefore, it seems that
it outlaws non-emergency political detention altogether. It is nevertheless
a great pity that that has not been done explicitly, because there can be
no guarantee that the Bill of Rights will be given its proper construction.
The limitability of s 11 (1), coupled with the fact that the Bill does envisage
non-emergency detention, invites an executive-minded court to uphold a
non-emergency political detention law. If that happens, it will undermine
the Bill's contribution to a culture of justification, because detention
without trial not justified by a genuine emergency, demonstrated in the
kinds of ways provided for in s 34, is anathema to a real effort to bring
government action under the discipline of accountability.
In two respects, then, the methods employed by the Bill to foster an
inquiry into the justification of the decisions within its remit are too
elaborate, and the consequence is a tendency to undermine the

20 Section 11 (1) reads thus: 'Every person shall have the right to freedom and security of the person,
which shall include the right not to be detained without trial.'
 SOUTH AFRICAN JOURNAL ON HUMAN RIGHTS

contribution that the Bill makes to a culture of justification. The first is

the triocular scrutiny introduced by the limitation clause, which injects an
element of arbitrariness. The second is the failure to remove certain rights
from the arena of justification altogether by making them illimitable.

AdministrativeJustice
In important respects, however, the Bill gives a lead which, properly
followed, would put South Africa at the frontiers of the search for a
culture of justification. Potentially very prominent here is s 24, the
administrative justice clause.
This section reads thus:
'Every person shall have the right to-

(a) lawful administrative action where any of his or her rights or interests is affected or
threatened;
(b) procedurally fair administrative action where any of his or her rights or legitimate
expectations is affected or threatened;
(c) be furnished with reasons in writing for administrative action which affects any of his
or her rights or interests unless the reasons for such action have been made public; and

(d) administrative action which is justifiable in relation to the reasons given for it where
any of his or her rights is affected or threatened.'

The object of s 24(a) is to annul ouster clauses: since it creates a

constitutional right to lawful administrative action, it vitiates legislative
provisions which attempt to insulate administrative action from judicial
review, because administrative action which in the absence of such a
provision is reviewable - in the sense that it is liable to be set aside on
review - is unlawful. 21 The force of s 24(a), it seems, is consequently to
overrule decisions such as Staatspresidentv United Democratic Front,22
which decided that the ouster clause 23 in the Public Safety Act 195324 had

21 Pace the General Council of the Bar, which expressed the opinion, during the drafting process, that
'to entrench a "right" to "lawful" administrative decisions, is without content. The "right" to
"lawful" administrative decisions, appears to be simply the right to decisions that do not stand to be
impugned by way of review. The formulation therefore gives no content to the grounds of review that
may vitiate an administrative decision' (General Council of the Bar of South Africa Submissions to
Technical Committee on FundamentalRights during the Transition(Seventh ProgressReport: 29 July
1993) (10 August 1993) 18).
This opinion seems to overlook the fact that what was being drafted here was not a statute, but a
constitution. If it had been proposed to enact a right to lawful administrative decisions in a statute,
it might fairly have been said that the proposed cnactmcnt was doing no more than restating the effect
of the existing law, and consequently that the enactment was 'without content'. But the effect of
restating the existing law in the Constitution is to protect it against legislative interference (ss 4 and
7). That is why the effect is to block ouster clauses; a consequence very far from being without
content.
22 1988 (4) SA 830 (A).
23 Section 5B.
24 Act 3 of 1953.
INTRODUCING THE INTERIM BILL OF RIGHTS

the effect of defeating an attack upon an emergency regulation for

uncertainty. 25 If uncertainty is a ground of review at all - and no one says
it is not - its presence makes a regulation 26 contaminated by it unlawful,
contrary to s 24(a), for if the regulation were lawful, no court would be
entitled, in the absence of the ouster, to strike it down. 27 The right in
s 24(a) consequently brings within the reach of judicial review important
classes of administrative action which might otherwise lie beyond judicial
scrutiny; it therefore makes an important contribution to fostering the
28
ethic of justification.
The same is true of s 24(b), the right to procedurally fair administrative
action. At a minimum, this paragraph requires a court applying it to
consider whether any decision within its ambit can be taken to have been
decided fairly if the person affected has not first been heard. And the
person affected will not ordinarily be taken to have been heard if the case
which he or she has to meet has not been disclosed, and an opportunity
given to reply to it. The pressure which this paragraph consequently puts
on officials to disclose the case upon which they propose to act, and
debate it with those whom it affects, clearly also fosters the enterprise of
encouraging decisions to be justified.
Nowhere, however, is the contribution of s 24 to that enterprise more
conspicuous than in its paragraphs (c) and (d). Paragraph (c) puts a
general duty on officials to give reasons for their decisions, something that
South African administrative lawyers have sought for decades as an
essential prerequisite to close judicial review of administrative action.
And paragraph (d) expressly empowers a court to inquire into the
justification of administrative action.
The effect of this paragraph is to overrule two closely linked doctrines
which have stultified review for unreasonableness in our administrative
law. The first doctrine is that unreasonableness is not, in itself, a
reviewable defect; unreasonableness is material only so far as from it may
be inferred the existence of some other deficiency in the decision. 29 That
other deficiency has been described variously, but under inspection
generally comes down to abuse of discretion. 30 The second doctrine,

25 See also Natal Indian Congress v State President1989 (3) SA 588 (D), which extended the effect of
Staatspresidentv United DemocraticFront from uncertainty to unreasonableness.
26 1 take the expression 'administrative action' in s 24 to be wide enough to cover a regulation, or, for
that matter, any other kind of legislation made by an executive authority.
27 Cf the judgments of Rabie ACJ at 853 and Hefer JA at 872.
28 It remains open to the state, of course, to try to defend a particular ouster clause as a justifiable
limitation on s 24(a) under s 33. This is another unfortunate consequence of the failure to make some
rights in the Bill illimitable.
29 See, for example, Union Government v Union Steel Corporation(South Africa) Ltd 1928 AD 220 at
236-7.
30 See, for example, Union Government v Union Steel Corporation(South Africa) Ltd supra note 29 at
237; Administrator, Transvaaland The Firs Investments (Pty) Ltd v JohannesburgCity Council 1971
(1) SA 56 (A) at 80.
 SOUTH AFRICAN JOURNAL ON HUMAN RIGHTS

usually perceived to be a corollary of the first, 31 is that a merely ordinary

degree of unreasonableness does not suffice to vitiate the decision; there
must be some egregious degree of unreasonableness. 3 2 Prevailing
authority teaches that what is required is a decision which is 'grossly
unreasonable to so striking a degree' as to warrant the appropriate
inference; 33 entailed in which is that the decision must be worse than
strikingly grossly unreasonable.
Section 24 (d) deals this doctrine of deep deference a blow which all
friends of official accountability will wish to be mortal. It empowers a
court to review administrative decisions within its reach for justifiability
-which is to say, for reasonableness - on its own. 34 The courts will have

31 But see my 'Administrative Law in South Africa' (1986) 103 SALJ 615 at 629-30.
32 See, for example, Union Government v Union Steel Corporation(South Africa) Ltd supra note 29 at
237; Administrator, Transvaaland The FirsInvestments (Pty) Ltd v JohannesburgCity Councilsupra
note 30 at 80. This is now subject to the exception (the limits of which are unclear) recognized in Hira
v Booysen 1992 (4) SA 69 (A) at 93.
33 National Transport Commission v Chetty's Motor Transport (Pty) Ltd 1972 (3) SA 726 (A) at 735.
34 Paragraph (d) uses the word 'justifiable' instead of'reasonable'. The wording derives from dissension
during the drafting. In one of their early reports, the drafters of the Bill opposed the entrenchment
of review of administrative decisions for unreasonableness on the ground that that would have
'far-reaching consequences for South African administrative law' (Technical Committee on
Fundamental Rights during the Transition Seventh ProgressReport (29 July 1993) 6). This was a very
curious objection - after all, much of the Bill, and the Constitution of which it is part, will have
cataclysmic consequences for South African law - and it sparked fierce debate.
At that stage there appeared to be two sources of resistance among political negotiators to the notion
of review for unreasonableness. The one was the usual - and legitimate - anxiety that that idea
might be abused by the courts to usurp the policymaking prerogatives of the executive. The other was
an anxiety arising from a somewhat irrational distrust of the word 'reasonable'.
In an effort to allay these concerns, I proposed a compromise administrative justice clause to certain
political negotiators. The compromise clause tried to meet the first anxiety by defining the scope of
the review jurisdiction in a way which attempted to guard against the kind of abuse feared. Paragraphs
(2) and (3) of the compromise clause (below) embody this effort. And it tried to meet the second
anxiety by replacing the word 'reasonable' with 'justifiable', which I take to mean the same thing, but
which is apparently freer of the connotations which dismay those who fear 'reasonableness'.
The compromise clause read as follows:
'Compromise Administrative Justice Clause:
18.(1) Every person shall have the right not to be affected adversely by a decision made in the exercise
of public power which is unlawful, procedurally unfair, or not justifiable.
(2) A decision shall not be considered justifiable unless:
(a) a plausible answer can be given to any reasonable objection to the decision;
(b) a plausible explanation can be given why the decision was chosen in preference to any arguably
superior alternative; and
(c) a rational connection can be shown between the premises of the decision (including any evidence
or argument on which it purports to be based) and the decision itself.
(3) In deciding whether a decision is justifiable, the [designated authority] shall not usurp the
prerogative of the decisionmaker to make such policy choices as the decisionmaker considers
desirable in the interests of good governance, and it shall respect and uphold every such choice.
(4) Every person shall have the right to be furnished with reasons in writing for any decision, made
in the exercise of public power, which affects him or her adversely.'
In the end, as can be seen from the administrative justice clause actually enacted (quoted in the text
above), the legitimate anxiety was ignored, and no wording was adopted to guard against it; but the
irrational anxiety was met, and the word 'justifiable', the sole survivor of the compromise clause,
somehow found its way into the final product.
(The expression 'designated authority' was used to fit with the text then current. At that stage it was
still an open question which courts would be entrusted with the enforcement of the Bill.)
INTRODUCING THE INTERIM BILL OF RIGHTS

to develop a theory of what makes a decision justifiable, and what makes

it unjustifiable. A good starting point would be to recognize that the
justifiability of an administrative decision is a matter not of second-
guessing the policy choices that it entails, which is the prerogative of the
decisionmaker, but rather of the soundness of the process of deciding
which went into its making. It is suggested that an administrative decision
cannot be taken to be justifiable unless (a) the decisionmaker has
considered all the serious objections to the decision taken and has
answers which plausibly meet them; (b) the decisionmaker has consi-
dered all the serious alternatives to the decision taken, and has discarded
them for plausible reasons; and (c) there is a rational connection between
premises and conclusion - between the information (evidence and
argument) before the decisionmaker and the decision taken. 35 A
decisionmaker who is conscious, while deciding, that a court may
eventually scrutinize his or her decisions under these criteria is under
serious pressure to deploy a thorough and disciplined decisionmaking
36
process; one likely to yield wise decisions.
So understood, s 24(d) would lift the shutters that now veil the inner
workings of so much administrative decisionmaking, and expose it to the
lights of the ethic of justification. In one respect, however, s 24(d) may be
thought to render administrative action excessively vulnerable: it
requires administrative action to be justifiable 'in relation to the reasons
given for it'. This qualification might be read as referring back to s 24(c),
which entitles one to demand reasons in writing for administrative action.
If s 24(d) requires administrative action to be justifiable by the reasons
supplied in response to a demand made under s 24(c), that might wreak
administrative havoc, for it would require officials to supply reasons
under s 24(c) which are carefully enough drafted to take the decision in
due course past judicial review under s 24(d).
Suppose, in other words, that an official, faced with some urgent
demand for written reasons under paragraph (c), responds carelessly,
omits one of the reasons which actually moved the decisionmaker to take
the decision, and consequently furnishes a set of written reasons which is
not full enough to demonstrate justifiability under paragraph (d). If on
judicial review under paragraph (d) the decisionmaker has to show the
decision to be justifiable on the basis of the careless statement of reasons,
the court would have to strike down the decision, even although it is one
which might in fact be justifiable, just because the written statement is
incomplete. On this reading of s 24, the decision would stand or fall by the
written statement of reasons, and that statement could not be augmented
with reasons not mentioned in it but which did actually move the

35 See, further, my 'Reconsidering Review: Participation and Accountability' 1993 Acta Juridica35 at
40-43.
36 Which is not to say that the idea of justifiability entitles a court to strike down a decision merely on
the ground of its unwisdom.
 SOUTH AFRICAN JOURNAL ON HUMAN RIGHTS

decisionmaker to take the decision and which, together with the written
reasons, do in fact show the decision to be justifiable.
So the drafters, who initially resisted the idea of entrenching review for
unreasonableness altogether, 37 have, somewhat perversely, in the end
embodied it in a form which may be read as being among the most
demanding versions possible of that idea. Indeed, this version is so
demanding that it may sabotage the idea altogether. There can be few
civil services anywhere which have the capacity regularly to produce
statements of reasons that, without pre-litigation revision, are sufficiently
meticulously drafted to assure passage past judicial review. Our own civil
service, in the nature of the transition that South Africa is undergoing,
will be relatively inexperienced, and it will be dealing with a great tide of
expectations, and an immense volume of work. To burden it with the kind
of rigour which this reading requires might undermine government
altogether.
And that might put the very idea of review for unreasonableness in
jeopardy, because it would give the government a case for a constitutional
amendment to free itself of the burden imposed by this reading. The
result might well be the repeal altogether of s 24(d). And that would of
course gravely subvert the contribution that s 24 makes to the advent of
a culture of justification.
If the courts can read the Bill of Rights in a way which averts this
consequence, they should: if the point of the Bill is to foster a culture of
justification, it is the courts' duty to read the Bill in the way which best
preserves its capacity to do that. And there are ways of avoiding the
reading of s 24(d) which would put the Bill's contribution to such a culture
in jeopardy. The courts might note, for instance, that s 24(c) and s 24(d)
govern different classes of administrative action: the right to written
reasons in s 24(c) applies to administrative action which affects one's
rights or interests, but the right to justifiable decisions in s 24(d) applies
to administrative action which affects or threatens one's rights. This
discrepancy is a cogent argument for refusing to read the two paragraphs
together; for refusing to read 'justifiable in relation to the reasons given
for it' in paragraph (d) as referring to the written reasons made
compulsory by paragraph (c). That would free a court to read 'the reasons
given for it' in paragraph (d) as referring to the reasons supplied in judicial
review itself by way of justifying the decision. The effect would be to
make s 24 an instrument for improving governance, not frustrating it. A
reading of s 24 which achieves that is plainly to be preferred.
This point draws attention to the most conspicuous idiosyncrasy in s 24:
that each of the four rights that it confers applies to a different class of
administrative action: paragraph (a) confers a right to lawful administra-
tive action where one's rights or interests are affected or threatened;

37 See note 34 above.

INTRODUCING THE INTERIM BILL OF RIGHTS

paragraph (b) confers a right to procedurally fair administrative action

where one's rights or legitimate expectations (but not one's interests) are
affected or threatened; paragraph (c) confers a right to written reasons
where one's rights or interests are affected (but not threatened); and
paragraph (d) confers a right to justifiable administrative action where
one's rights (but not one's interests or legitimate expectations) are
affected or threatened. The courts will have their work cut out developing
a coherent theory to account for these variations.

Freedom of Information
The drafting is also problematic in s 23, the right of access to official
information. Access to official information is a matter of the utmost
importance to any effort to bring about a culture of justification. A
government which can close its files will be under much weaker pressure
to justify its decisions than one which has to open them.
If the Bill of Rights entrenched a general right to information, that
would of course not expose all official information to public scrutiny.
Even the most committed advocates of open government acknowledge
that there are always sensitive classes of information - such as
information necessary to protect personal privacy, or commercial
confidentiality, or national security, or law enforcement - which
government is entitled to conceal. But the limitation clause is available to
protect such information: government could procure legislation to
preserve the confidentiality of these classes of information, and then
justify that legislation under the limitation clause. Such legislation would
be of the same form as a typical Freedom of Information Act. The main
value of a general right to information would consequently be to force
government to define and defend the classes of information which it
asserts have to be secret. Have the drafters bequeathed us a general right
of information?
Section 23 gives every person a 'right of access to all information held
by the state or any of its organs at any level of government'; but only 'in
so far as such information is required for the exercise or protection of any
of his or her rights'. The drafters' decision to limit the right to information
38
required for the exercise or protection of one's rights is unfortunate. It
invites a conservative court to read the right as merely ancillary to
litigation: as conferring entitlement only to the information necessary to
vindicate one's rights at law - no more, in other words, than a

38 The limitation appears to be derived from Hugh Corder et al A Charter for Social Justice (1992) 54.
In that work the authors propose the following access-to-information clause:
'(1) Everyone has the right of access to information held by any authority performing governmental
functions.
(2) Everyone has the right of access to that information which is necessary for the implementation of
his or her rights.'
 SOUTH AFRICAN JOURNAL ON HUMAN RIGHTS

constitutional right to discovery. There is nothing, however, in the text

which precludes the courts from reading the right more broadly. One may
seek to exercise or protect one's rights by means other than litigation -
for instance through publicity or by resort to the political process - and
information may be required for that purpose. On that, broader, reading,
s 23 would come substantially closer to conferring a general right of access
to official information.
In choosing between a narrower reading and a broader one, the courts
should have regard to the point of the Bill of Rights - to foster a culture
of justification. That point will usually argue for a broader reading. So the
drafting is not entirely irretrievable, although it is to be regretted that the
opportunity to enact an unambiguous general right to information was
squandered.

Special Immunities: Labourand Religion

Despite these deficiencies, the rights in Chapter 3, as the above account
reveals, do comprise a set of principles with considerable power to fuel a
vigorous inquiry into the justification of decisions within their purview;
too much power for some, as the efforts of various lobbies to procure
special exemptions from such inquiry attests. 39 The impact of two such
lobbies is conspicuous in the Bill. Section 33(5)(a) states that the
'provisions of a law in force at the commencement of this Constitution
promoting fair employment practices, orderly and equitable collective
bargaining and the regulation of industrial action shall remain of full force
and effect until repealed or amended by the legislature'. The intention is
apparently to insulate existing labour legislation from inquiry under the
Bill into its justification, although the apparently deliberate elimination
of wording expressly giving this exemption primacy over the rights in the
Bill,40 which was present until late in the drafting process, 41 might, on a
theory which attached importance to the drafting history, 42 be thought to
weaken the evidence of that intention.
The other lobby is the religious. Section 14(1) guarantees the right to
freedom of religion. That idea might be thought to entail freedom from
governmental practices which favour one religious perspective over
another. 43 If it did, it would open inquiry into the justification of

39 The most notorious effort was the attempt by traditional leaders represented in the drafting process
to insulate African customary law from question under the Bill's guarantee of gender equality.
40 'Notwithstanding the provisions of this Chapter...'.
41 Technical Committee on Fundamental Rights during the Transition Tenth Progress Report
(5 October 1993) clause 34 (5).
42 As to which, see note 54 below.
43 In the United States, this question is sometimes thought to be the exclusive province of the
Establishment Clause of the First Amendment to the Constitution ('Congress shall make no law
respecting an establishment of religion'), and consequently not one which can arise under the Free
INTRODUCING THE INTERIM BILL OF RIGHTS

governmental practices which might be thought to favour one religion

over another, or religion in general over the absence of religion, such as
compulsory school prayer, religious education in state schools, state
support for private religious education, and the observance of religious
rituals in state institutions.
No doubt anticipating these possibilities, the drafters provided, in
s 14(2), that 'religious observances may be conducted at state or
state-aided institutions under rules established by an appropriate
authority for that purpose, provided that such religious observances are
conducted on an equitable basis and attendance at them is free and
voluntary'. The intention of this exemption appears to be to insulate
religious practices in public institutions from inquiry of the kind just
outlined. As it is drafted, though, the exemption seems to protect very
little."4
For one thing, it works only where attendance at a religious observance
is free and voluntary; but it is coming widely to be recognized that
religious observances which carry the endorsement, tacit or otherwise, of
a state institution are inherently coercive. State endorsement of a
religious perspective - be it only the perspective that religion is to be
preferred to irreligion - turns those who adhere to that perspective into
insiders, and those who do not into outsiders. 45 That alone puts pressure
- governmental pressure, since it comes from the state - on
non-adherents which may be considered coercion. 46 Furthermore, the
exemption applies only where religious observances 'are conducted on an
equitable basis'. But state endorsement of a religious perspective might
be thought inequitable to any who do not share that perspective.
The result is that a religious observance can be taken to be safely within
the exemption in s 14(2) only where the state does not endorse a religious
perspective, whether explicitly or tacitly. It will seldom be easy to
structure a religious observance at a state or state-aided institution to
meet this condition.
What is more, the exemption in s 14(2) is not framed to prevail over
freedom of religion; nor, for that matter over the freedoms of conscience,
thought, belief and opinion. On the contrary, it is expressed not to

Exercise Clause ('Congress shall make no law ... prohibiting the free exercise [of religion]'). From
the particular textual shape of the First Amendment, however, it does not follow that the question
cannot arise under the concept of freedom of religion, properly understood.
44 Potentially more succesful is s 14(3), which provides:
'Nothing in this Chapter shall preclude legislation recognizing -
(a) a system of personal and family law adhered to by persons professing a particular religion; and
(b) the validity of marriages concluded under a system of religious law subject to specified
procedures.'
The main point of this subsection is apparently to insulate the religious practices within its purview
from challenge under the guarantee of gender equality in s 8.
45 Cf Wallace v Jaffree 472 US 38 (1984) at 69 (O'Connor J).
46 Cf Lee v Weisman (1992) 120 L Ed 2d 467 (S Ct).
 SOUTH AFRICAN JOURNAL ON HUMAN RIGHTS

derogate from the generality of those freedoms. 47 This weakens the force
of the exemption. Nor is there any indication that the exemption is to
prevail over the prohibition in s 8 - known as the equality clause - on
unfair discrimination, direct and indirect, on the ground of religion,
conscience or belief. 48 This weakens even further the capacity of the
exemption to insulate from scrutiny state practices that favour certain
religious perspectives over others.
There is therefore ample support in the Bill for a reading of s 14(2)
which gives it a narrow effect. And a court should be inclined to give that
support great weight. Special exemptions from the rights in the Bill are
special immunities from the standards of justification that it announces.
They tend, therefore, to undermine the Bill's primary object of
inaugurating a culture of justification. They should consequently be
construed as narrowly as possible. It follows that any textual support in
the Bill for a narrow reading of the exemptions should be given as much
weight as possible.

Resolving Ambiguity: Affirmative Action

The same kind of approach, as I have already suggested, should govern
any ambiguity in the Bill: any ambiguity should be resolved in the way
which optimizes the Bill's capacity to foster a culture of justification.
One such ambiguity may be thought to arise in the field of affirmative
action. Section 8 (the equality clause) guarantees both equality before the
law and equal protection of the law, and prohibits unfair discrimination
both generally and on the particular grounds of race, gender, sex, ethnic
or social origin, colour, sexual orientation, age, disability, religion,
conscience, belief, culture or language. It then continues:
'This section shall not preclude measures designed to achieve the adequate protection and
advancement of persons or groups or categories of persons disadvantaged by unfair
discrimination, in order to enable their full and equal enjoyment of all rights and
49
freedoms.'

This provision is intended to save affirmative action programmes -

programmes to undo the disadvantages caused by apartheid - from
challenge under the equality clause. The desirability of such a provision
was not very controversial during the drafting process. All three political
parties which published proposed bills of rights prior to the process
affirmed the need for such a provision. 50 But an important controversy

47 The exemption granted by s 14(2) is prefaced with the words, 'Without derogating from the generality
of subsection (1)...'. Subsection (1) guarantees the freedoms of conscience, religion, thought, belief
and opinion.
48 Section 8(2).
49 Section 8(3)(a).
50 See the African National Congress's Draft Bill of Rights (Preliminary Revised Version February
INTRODUCING THE INTERIM BILL OF RIGHTS

during the drafting process raised the question how closely a court
applying this provision could scrutinize something purporting to be an
affirmative action measure to make sure that it really was one.
The controversy was sparked by an earlier draft, in which the place of
51
the words 'designed to achieve' was taken by the words 'aimed at'.
Those words raised the concern that they could be understood to
authorize a court to ask whether a measure was aimed at undoing
disadvantage, but not whether it was actually likely to undo disadvantage.
The concern, in other words, was that the provision might be construed
to empower the court to review the intentions of the authors of the
measure, but not its effects. If, say, the government reserved 95 per cent
of senior civil service jobs for Africans, it might be impossible to contest
the government's assertion that its aim was to protect and advance
categories of persons disadvantaged by unfair discrimination, in order to
enable their full and equal enjoyment of all rights and freedoms; but it
might be a lively question whether that was the probable effect of the
reservation. The anxiety was that the courts would have been confined to
scrutinizing the ends sought by the authors of the measure, and precluded
from scrutinizing the means chosen to realize those ends.
It was to meet this concern that the words 'aimed at' were replaced with
the words 'designed to achieve'. On its face, however, the new wording
is far from an unambiguous resolution of the controversy. The words
'designed to achieve' can mean (a) 'intended to achieve', in which case
they mean much the same as 'aimed at'; or (b) they can mean 'constructed
so as to achieve'. In this latter sense, the wording refers to measures which
are not only intended to achieve something, but the design of which
makes it objectively probable that they will in fact achieve that
something; which are structured, in other words, to achieve that thing. In
this sense 'designed to' embodies much of the ambivalence of 'calculated
to', which, although it can mean 'intended to', the courts often interpret
as 'likely to';5 2 indeed, it incorporates both those meanings.
In which sense is 'designed to achieve' used here: (a) or (b)? For what
it is worth, the drafting history rules out (a); if that had been the sense
intended, there would have been no point in discarding the original
wording, 'aimed at'. 53 But a court may, and probably should, decide that

1993) Art 14; the National Party's Government's Proposals on a Charter of Fundamental Rights
(2 February 1993) s 6 (3); the Democratic Party's Draft Bill of Rights (May 1993) Art 2 (2). See also
the KwaZulu Legislative Assembly's proposed Constitution of the State of KwaZulu/Natal
(1 December 1992) clause 10 (b).
51 Technical Committee on Fundamental Rights during the Transition Seventh Progress Report (29 July
1993) clause 2 (3). In this earlier draft, there were also other differences in wording.
52 Or some minor variation on that expression: see, for example, Amoils v JohannesburgCity Council
1943 TPD 386 at 389; R v Lord 1926 CPD 262 at 264; Brink v Lichtenburg Liquor Licensing Board
1944 TPD 161 at 171; JohannesburgLiquor Licensing Board v Short 1946 AD 713 at 722-3; Rossouw
v Norton NO 1950 (2) SA 1 (C) at 5; S v Beyleveld 1964 (1) SA 269 (T) at 272-3.
53 What is more, the political negotiators who settled the wording declined to insert the word
 SOUTH AFRICAN JOURNAL ON HUMAN RIGHTS

the drafting history is irrelevant. 54 A better reason for preferring sense

(b) over sense (a) is simply that, since it brings not only ends but also
means under judicial scrutiny, it brings the whole measure, rather than an
arbitrary fragment of it (the mere intentions of its authors) into the
enterprise of justification. Sense (b) consequently makes the provision
operate as a fuller and more coherent contribution to the cultivation of an
ethic of justification. That is a powerful reason for preferring it over sense
(a). On this view, the point of the Bill of Rights - to foster a culture of
justification - is again a resource which helps resolve ambiguities in the
Bill's text.

CONCLUSION

The point of the Bill of Rights is to foster a culture of justification. This

is something which appears not only from our history, but from the text
of the Bill itself. Measured against that point, the Bill suffers from
significant weaknesses. Fortunately, the point of the Bill of Rights is not
just a criterion for evaluating the Bill, but also a powerful guide for
answering the questions of interpretation that it generates. If the courts
use that guide properly, they can overcome most of the deficiencies in the
Bill, and make of it a potent weapon for bringing about democracy.

'reasonable' before 'measures' or the word 'reasonably' before 'designed', evidently because they had
been assured by the technical committee which drafted the Bill that 'the rationality of the measures
is sufficiently catered for by the qualification which follows the words "in order to" ' (unpublished and
undated drafts generated by the Technical Committee on Fundamental Rights during the Transition
during September 1993). Whatever one thinks of the soundness of the technical committee's opinion,
it seems clear that its effect was that the wording finally adopted was accepted on the understanding
that the rationality of affirmative action measures was to be reviewable. That would entitle a court
to scrutinize not just the intentions of the authors of the measures, but also the effects of the measures.
Consequently, so far as a court considers the drafting history relevant, this feature of that history
appears to demonstrate that the text was adopted on the understanding that not just the ends of
affirmative action measures, but also their means, were to be amenable to judicial review.
54 For several reasons, among them that people governed by a text are entitled to order their lives on
the strength of what they can see; on the strength, in other words, of the published words and
published authoritative commentary, without having to burrow among obscure and inaccessible
drafting sources. Other reasons for discounting the drafting history appear from a rich literature. See,
for example, Ronald Dworkin Law's Empire (1986) chapter 10; Life's Dominion (1993) chapter 5.