2012 Constitutional Court Review 2011 4

CONSTITUTIONAL
COURT
REVIEW
(2011) 4
2012
Constitutional Court Review (2011) 4
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:

BusinessPrint
Pretoria
To order, contact:
PULP
Faculty of Law
University of Pretoria
South Africa
0002
Tel: +27 12 420 4948
Fax: +27 12 362 5125
pulp@up.ac.za
www.pulp.up.ac.za
Cover:
Yolanda Booyzen, Centre for Human Rights
ISSN: 2073-6215
© 2012
The financial contribution of the Konrad Adenauer Stiftung is gratefully

acknowledged.
CONSTITUTIONAL COURT REVIEW
The Constitutional Court Review, an initiative of Constitutional law of

South Africa (CLoSA), the Department of Public Law at the University
of Pretoria and the South African Institute for Advanced
Constitutional, Public, Human Rights and International Law (SAIFAC)
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 at least one leading
essay exploring broad themes arising from a given year’s
jurisprudence (each ± 20 000 words), with a response (± 5 000 words);
a number of shorter subject-specific articles (each ± 10 000 words).
Lead essays are solicited by the editors, as are some of the subject-
specific articles, but for the remainder unsolicited contributions are
invited. Such contributions must be sent to the editors at
stuart.woolman@wits.ac.za 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.
The financial contribution of the World Bank and the Konrad Adenauer
Foundation toward publication of the Constitutional Court Review is
gratefully acknowledged.
iii
SUBSCRIPTION
Subscriptions through agents or bookshops:

R380-00 (including VAT but excluding postage and packaging)
Subscriptions directly through PULP:

Subscriptions by registered students directly through PULP:

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: pulp@up.ac.za
Website: http://www.pulp.up.ac.za
iv
TABLE OF CONTENTS
The South African Constitutional Court and socio-economic rights 1

as ‘insurance swaps’
Rosalind Dixon & Tom Ginsburg
Rationality, reasonableness, proportionality: Testing the 31

use of standards of scrutiny in the constitutional review of
legislation
Christian Courtis
The democratic turn and (the limits of) constitutional patriotism 51

after the Truth and Reconciliation Commission Albutt v CSVR
Wessel le Roux
Customary (communal) land tenure in South Africa: Did Tongoane 73

overlook or avoid the core issue
Douglas Mailula
Rationality, the rule of law, and the sovereign return 113

Stewart Motha
Three-level games: Thoughts on Glenister, Scaw and 137

international law
Mkhululi D Stubbs
Confusing grace with amnesia: Reviewing acts of the head of state 167
Francois Venter
v
THE SOUTH AFRICAN CONSTITUTIONAL COURT
AND SOCIO-ECONOMIC RIGHTS
AS ‘INSURANCE SWAPS’
Rosalind Dixon* & Tom Ginsburg**
1 Introduction
Socio-economic rights are a central terrain of struggle in new

democracies.1 Often deemed essential for the legitimacy of the
constitution at the time of adoption, they are subject to downstream
pressures at the implementation stage as governments confront
limited budgets and the need for macroeconomic credibility. The
result is a gap between promise and reality. It is not surprising that,
in an age of judicialisation, socio-economic rights have become a
central topic of constitutional adjudication in many new democracies,
as courts struggle to balance normative commitments with
democratic prerogatives.2
The South African Constitutional Court, in the 2010 Term, heard

a number of important cases involving the socio-economic rights
provisions in sections 26-29 of the Constitution. In Nokotyana v
Ekurhuleni Metropolitan Municipality,3 the Court considered a claim
by the applicants to have access to upgraded toilets and lighting, as
* Assistant Professor, University of Chicago Law School, Professor, UNSW Faculty of

Law.
** Leo Spitz Professor of International Law and Professor of Political Science,
University of Chicago Law School, Research Professor, American Bar Foundation.
Our thanks to Sujit Choudry, Beth Goldblatt, Eric Posner and Theunis Roux for
extremely helpful comments on prior versions of the paper, and to Alex Bergersen
and Kristen McKeon for excellent research assistance.
1
The division between socio-economic and other rights is, of course, an artificial
one: see eg T Daintith ‘The constitutional protection of economic rights’ (2004) 2
International Journal of Constitutional Law 56 - 90. We use the labels, however,
simply as a short-hand for denoting a distinct set of rights.
2 RU Yepes ‘The enforcement of social rights by the Colombian Constitutional
Court’ in R Gargarella, P Domingo & T Roux (eds) Courts and social
transformation in new democracies (2006) 127; C Rodriguez-Garavito ‘Colombia:
The new left: Origins, trajectory and prospects’ in P Barret et al (eds) The new
Latin American left: Utopia reborn (2008) 129 - 157.
3 Nokotyana v Ekurhuleni Metropolitan Municipality 2010 4 BCLR 312 (CC).
1
2 The SA Constitutional Court and socio-economic rights as ‘insurance swaps’
part of the right of access to housing under section 26 of the

Constitution. In Juma Musjid Trust v MEC,4 the Court considered a
challenge to an order evicting a public school from privately-owned
land, based on the right to ‘basic education’ under section 29 of the
Constitution, and its potential horizontal application, under section
8(2), to a privately-owned trust owning land on which a public school
was located.5 In Tongoane v National Minister for Agriculture and
Land Affairs,6 the Court heard a challenge to various aspects of the
Communal Land Rights Act (CLARA) based on their inconsistency with
the right to legally secure land tenure under section 25(6) of the
Constitution.7 And in Bengwenyama Minerals (Pty) Ltd v Genorah
Resources (Pty) Ltd,8 the Court considered a challenge to the grant of
prospecting rights to the respondents, based on both a failure to
comply with relevant statutory provisions and the right of ‘equitable
access’ to natural resources in section 25(4) of the Constitution.9
The 2010 Term was also book-ended by numerous cases involving

sections 26-29. In 2009, in Mazibuko v City of Johannesburg,10 the
Court issued one of its most significant decisions to date involving
socio-economic rights, dismissing a challenge under section 27(1)(b)
of the Constitution (the right of access to sufficient water) to the
free-water allowance and pre-paid meter policy of the City of
Johannesburg (and Johannesburg Water (Pty) Ltd). In Mpumalanga
Department of Education v Hoërskool Ermelo,11 the Court dismissed
a challenge under section 29(2) of the Constitution to the decision of
the Minister to revoke the power of a public school board to
determine its own language policy. And in Joe Slovo v Thubelisha
Homes (Joe Slovo I),12 Joseph Leon v City of Johannesburg,13 Abhalali
v Premier of KZN, 14 and Machele v Mailulu,15 the Court addressed a
variety of questions relating to the scope of section 26(1).
4
Juma Musjid Primary School v Essay N.O. 2011 8 BCLR 761 (CC).
5 The reasons for decision in Musjid were handed down in 2011, but the initial
decision in the matter was given in 2010. See n 4 above, para 6.
6
Tongoane v National Minister for Agriculture & Land Affairs 2010 8 BCLR 741
(CC).
7
Constitution of the Republic of South Africa 1996 sec 25(6). The Court did not
ultimately find it necessary to address this claim, on its merits, because of its
finding that the legislation was invalid in its entirety, based on a failure to comply
with the proper procedures for enactment in terms of sec 76 of the Constitution:
see Tongoane (n 6 above) paras 109 - 116.
8 Bengwenyama Minerals (Pty) Ltd v Genorah Resources (Pty) Ltd 2011 3 BCLR 229
(CC).
9
For the relevant constitutional arguments, see Bengwenyama (n 8 above) paras 3
& 28.
10
Mazibuko v City of Johannesburg 2010 3 BCLR 239 (CC).
11
Mpumalanga Department of Education v Hoërskool Ermelo 2010 3 BCLR 177 (CC).
12 Residents of Joe Slovo Community, Western Cape v Thubelisha Homes 2009 2009
9 BCLR 847 (CC) (Joe Slovo I).
13 Joseph v City of Johannesburg 2010 3 BCLR 212 (CC).
14 Abahlali Basemjondolo Movement SA v Premier of the Province of Kwazulu-Natal
2010 2 BCLR 99 (CC).
15 Machele v Mailula 2009 8 BCLR 767 (CC).
(2011) 4 Constitutional Court Review 3
In 2011, in Gundwana v Steko Development CC, the Court again

confronted arguments based on section 26(1), this time in the context
of a dispute over the executability of mortgaged property under
various procedures for the ordering of default judgment in the High
Court.16 And in Joe Slovo II, it revisited questions raised in Joe Slovo
I about the right of the applicants to housing in terms of section
26(1).17
In several of these cases, the Court also confronted a potential

direct conflict between various socio-economic rights, such as the
right of access to land, housing and education, and the right to
property under section 25(1) of the Constitution. The best example of
this, in the 2010 Term, was Musjid, where the Court found that the
Trust was both under a duty not to impair relevant children’s ‘access
to basic education’ under section 29 and entitled to maintain and
enforce its right to private property under section 25.18 A similar
pattern also arose, however, in numerous cases decided in 2009 and
2011 involving the right of access to housing under section 26(1) of the
Constitution.
In Abhalali, the legislation in question was challenged by the

petitioners as in direct conflict with both section 26(1) of the
Constitution and the provisions under the Prevention of Illegal
Eviction from and Unlawful Occupation of Land Act (PIE) giving effect
to this right.19 At the same time, by requiring municipalities to
expend the political and economic resources necessary to institute
eviction proceedings against unlawful occupiers in various
circumstances, the legislation in question was also designed to
protect the right to property in section 25(1). Similarly, in Gundwana,
the petitions relied on section 26(1) to challenge various High Court
Rules permitting a High Court registrar to declare mortgaged property
specially executable, as part of granting default judgment, when such
rules were clearly designed to protect the right to property, by
allowing for more expeditious forms of legal enforcement of this
right.20
16 Gundwana v Steko Development CC and Others 2011 8 BCLR 792 (CC).

17
Residents of Joe Slovo Community, Western Cape v Thebelisha Homes 2011 7
BCLR 723 (CC) (Joe Slovo II).
18 A similar potential for this kind of direct conflict also arose in Bengwenyama,
given that, as part of its reasoning, the Court clearly affirmed both the ‘preferent
right’ of the second applicants — as a community that had had previously been
deprived of formal title to their land by racially discriminatory laws, but then had
those rights reinstated — to be granted prospecting rights over their own land;
and also the existence of both statutory and constitutional support for this
position, in light of constitutional provisions such as sec 25(4) that guarantee a
right of equitable access to mineral resources. The respondents, however, did not
explicitly rely on sec 25 in arguing that their prospecting rights should not be set
aside.
19
Act 19 of 1998.
20 n 16 above, para 37.
The 2009-2011 Terms, therefore, provide a natural opportunity to

revisit the political relationship between the origins of socio-
economic rights guarantees such as sections 26-29 and the right to
property, in section 25 of the Constitution. While the political origins
of rights such as section 25 have been theorised in prior work on the
‘insurance-based’ function of judicial review, the political origins of
other socio-economic rights have received relatively little
attention.21 A key aim of this essay, therefore, is to begin to fill this
gap in the literature — by expanding existing insurance-based theories
of judicial review so as to account for the political origins of various
socio-economic rights, other than property, and in particular: rights
of access to housing, land, mineral resources and collective
organisation and bargaining.
For left-wing parties to constitutional negotiations, the inclusion

of a constitutional right to property carries a clear risk: that courts
and others will interpret such a right to impede legislative attempts
to redistribute resources, or realise basic socio-economic rights, such
as the rights of access to housing, land or collective bargaining. One
solution to this problem will be for left-wing parties to argue for the
exclusion of a right to property from a constitution. This was the
strategy successfully adopted, for example, by the National
Democratic Party in Canada, in the negotiations leading up to the
adoption of the Charter of Rights and Freedoms 1982.22 Such a
strategy, however, will also often be impractical, given the demand
for political insurance on the part of conservative parties to
constitutional negotiations. Attempts by left-wing parties to ‘carve
out’ certain limits to constitutional property rights guarantees may
also fail for similar reasons, relating to bargaining costs.23
A more realistic alternative for such parties, therefore, will in

many cases be to argue for the inclusion of certain socio-economic
rights guarantees as a form of ‘insurance swap’, which insulates forms
of progressive legislation from future constitutional invalidation in
return for concessions on the constitutional protection of property
rights. This idea of an insurance swap has a close resemblance, we
suggest, to other forms of financial swap, such as interest-rate swaps,
exchange-rate swaps and credit default swaps: it can allow parties to
21
See T Ginsburg Judicial review in new democracies (2003); compare also R Hirschl
Towards juristocracy: The origins and consequences of the new constitutionalism
(2004). On socio-economic rights, see A Ben-Bassat & M Dahan ‘Social rights in the
constitution and in practice’ (2008) 36 Journal of Comparative Economics 103 -
19.
22
See A Alvaro ‘Why property rights were excluded from the Canadian Charter of
Rights and Freedoms’ (1991) 24 Canadian Journal of Political Science / Revue
canadienne de science politique 309 - 29.
23
For this idea of constitutional ‘carve-outs’, see R Dixon ‘Constitutional
definitions’ (Working Paper, 2011-12).
bargain in a way that is more efficient than in the case of a one-way

exchange.
This account of socio-economic rights is also helpful to

understanding the South African context in 1995-1996. The African
National Congress (ANC) had a number of reasons to support the
inclusion of socio-economic rights (or directive principles) in any
democratic constitution; but among these was a concern to prevent
an overly expansive reading of first generation rights, such as the right
to property. In 1995, opposition parties were also more willing to
make concessions to the ANC on socio-economic rights than on the
core of the right to property itself, which was viewed as a deal-
breaker. On one reading, therefore, provisions such as sections 26-29
had an important capacity to lower the decision costs, for all parties,
of reaching agreement on the final text of the Constitution.
For the Constitutional Court, an insurance swap-based theory of

this kind has potentially important implications for the interpretation
of sections 25-29. From a historical or ‘originalist’ perspective, it
suggests that a key task facing the Court will be the need to maintain
a balance between the right to property and other socio-economic
rights, which may potentially conflict with a right to property. To do
this, the Court will also need to do two things: one, invalidate any
statutory or common law presumption in favor of one or other sets of
right; and second, adopt reasoning that is as narrow and context-
sensitive as possible in all cases involving such rights. This will also
mean the Court avoiding broad statements in favor of either highly
expansive or deferential, or ‘strong’ or ‘weak’, approach to the
definition and enforcement of such rights.24
Our suggested approach also accords surprisingly well, we

suggest, with the actual approach of the Constitutional Court in both
recent and earlier cases involving socio-economic rights. This may be
pure coincidence, but nonetheless, points to an important source of
potential additional support for the Court, in the face of criticism of
certain aspects of its approach, such as its rejection of the idea of a
‘minimum core’ to various socio-economic rights.25
The essay proceeds in four parts, following this introduction. Part

1 outlines the basic contours of an insurance-based theory of judicial
24
On the distinction between strong versus weak-form review in this context, see M
Tushnet Weak courts, strong rights: judicial review and social welfare rights in
comparative constitutional law (2008); R Dixon ‘Creating dialogue about
socioeconomic rights: Strong-form versus weak-form judicial review revisited’
(2007) 5 International Journal of Constitutional Law 391.
25 See D Bilchitz ‘Giving socioeconomic rights teeth: the minimum core and its
importance’ (2002) 119 South African Law Journal 484, and discussion in Dixon (n
24 above).
review, and explains how such a theory can be expanded to account

for the origins of various socio-economic rights, in addition to the
right to property, as forms of ‘insurance swap’ for left-wing parties to
constitutional negotiations. Part 2 applies the theory to the South
African context in 1993-1996. Part 3 considers the implications for
such a theory for the interpretation and enforcement of various socio-
economic rights in South Africa and elsewhere. Part 4 offers a brief
conclusion, focusing on the potential further extension — and limits —
of insurance-based theories in the context of socio-economic rights
provisions.
2 Socio-economic rights as insurance swaps
Our theory draws from earlier work on the ‘insurance’ function of

constitutional review.26 The idea assumes, in a rational choice vein,
that constitutional designers will choose institutions based on their
prospective position in the post-constitutional order. A designer who
believes she is likely to be in the majority in a post-constitutional
election will favour majoritarian institutions. One who believes she
will be in the minority, on the other hand, will prefer to have
institutions that can check the majoritarian legislature. These might
include rights protections, judicial review, supermajoritarian
requirements, and guardian institutions such as human rights
commissions.
The insurance model has been applied with some success in

several contexts.27 On its face it predicts that constitutions written
by dominant parties, such as the African National Congress (ANC), will
have fewer rights protections, unless (as proved to be the case in
South Africa) the minority demands them as the price of agreement.
But as has been argued elsewhere, the veto power held by minority
parties in South Africa meant that there had to be some negotiated
compromise between the two sides. This is a standard account of the
reason for the inclusion of a constitutional right to property in South
Africa: it was a line in the sand for the National Party (NP).28
Socio-economic rights present some challenge for basic insurance

accounts because they are, generally speaking, majoritarian and
redistributive rather than minoritarian in character. We thus need a
different account of the inclusion of judicially enforceable socio-
26
Ginsburg (n 21 above).
27 J Finkel Judicial reform as political insurance (2006). But see SI Oseguera
‘Judicial reform in Mexico: Political insurance or the search for legitimacy?’
(2009) 62 Political Research Quarterly 753 - 66.
28 See M Chaskalson ‘Stumbling towards section 28: Negotiations over the protection
of property rights in the Interim Constitution’ (1995) 11 South Africa Journal of
Human Rights 222 - 240.
economic rights in a constitution. It is perfectly understandable why

a dominant majority party with a left-wing ideology would seek to use
the language of rights to signal their policy goals;29 it is equally
understandable why a right-wing party faced with sure electoral loss
would find judicial power attractive, along with property rights
protections. The puzzle is why we would observe the combination.
The property rights section of the 1996 South African constitution
contains competing imperatives, and is balanced by a broad set of
socio-economic rights in other sections. What might explain this
pattern?
In the ordinary context of constitutional negotiations, judicial

review was modelled as an insurance policy to minimise future losses
in the event of political defeat. All parties that see themselves as
potential losers in the electoral realm will want to ensure that they
have access to a court to challenge future policies that violate their
rights or liberties. As has been recognised, the availability of this
insurance lowers the stakes associated with political defeat, and
hence may make some constitutional bargains possible that would not
otherwise be so.30
For left-wing parties, however, there is also the danger that

agreeing to certain forms of basic rights-based insurance —
particularly in the form of rights to property or contract — can prove
extremely costly, greatly undermining the future ability to adopt
progressive legislation. If read sufficiently broadly, rights to property
or contract have the capacity to cast doubt on the validity of almost
any form of redistributive or progressive legislation. All that is
required for this to occur, in the context of a right to property, for
example, is that a court adopt a broad view of what is (in US language)
a ‘regulatory taking’, or in the language of the 1993 South African
Constitution, a ‘deprivation’ as opposed to ‘expropriation’ of
property. Given such an approach, almost any law imposing a tax, or
regulating the use of property, may be invalidated by a court for
inconsistency with the right to property. Where a constitution is
negotiated between left- and right-leaning parties, rather than
unilaterally adopted by a dictator or right-leaning party, rights to
property can therefore emerge as a significant source of
disagreement in constitutional negotiations.
Three basic responses emerge in the face of this kind of conflict.

One option will be non-inclusion of a right to property, perhaps in
29 On the signaling function of socio-economic rights provisions more generally, see

DA Farber ‘Rights as signals’ (2002) 31 The Journal of Legal Studies 83.
30 Ginsburg (n 21 above).
return for concessions on other issues.31 A second option will be for

the parties to include the guarantees sought by both parties, but to
negotiate language limiting the scope of the relevant political
insurance by appropriately clear ‘carve-outs’, so as to avoid any
obvious conflict between such rights and potential future progressive/
redistributive legislation favoured by a left-wing party to
constitutional negotiations.32 And a third option will be for the parties
to adopt both a right to property and a set of offsetting socio-
economic rights (or directive principles) that can help shield, or
immunise, progressive legislation or policies against the potential for
future constitutional attack.33
A good example of how socio-economic rights guarantees can play

this role — as constitutional ‘shields’ — is the decision of the SACC in
Minister of Public Works v Kyalami Ridge Environmental
Association,34 a case involving the right to housing as a mere shield,
and the better known case of Government of South Africa v
Grootboom,35 in which the applicants asserted a positive obligation
on the state to provide them with access to housing. The applicants
in Kyalami Ridge were property-owners who sought to restrain the
construction of temporary housing by the government near their
homes (for approximately 300 hundred people rendered homeless by
a flood in Alexandra), on the basis that the decision to commence
construction was unlawful. The government, however, defended the
legality of its actions by relying on a combination of its inherent right
as a property owner and its obligations under section 26(1) of the
Constitution.
Certain rights, such as the right of access to housing, or land,

mineral resources or collective bargaining, will have a particularly
strong capacity to play this role, because of their quite direct,
physical connection to the enjoyment of rights to real property.
However, almost any socio-economic right will have the capacity to
play such a role, given a sufficiently expansive approach by a court to
the right to property. (Take, for example, the role a right of access to
health-care, such as section 27(1), could play in helping support the
31
This, for example, is what happened in Canada in the context of conflict over the
inclusion of a right to property in the Canadian Charter of Rights and Freedoms.
The federal sponsors of the Charter (Prime Minister Pierre Trudeau and the
Liberal Party) agreed to drop their demand that such a clause be included, in
return for support from provinces such as Sascatchewan for the Charter package
as a whole: see eg Alvaro (n 22 above).
32
See eg Constitution of Zambia 1996 art 16. For this idea of constitutional ‘carve-
outs’, see Dixon (n 23 above).
33 The more freestanding such rights guarantees are, however, the more likely it is
that they will in fact be interpreted as having this kind of immunising effect.
34 Minister of Public Works v Kyalami Ridge Environmental Association 2001 7 BCLR
652 (CC).
35
Government of the Republic of South Africa v Grootboom 2000 11 BCLR 1169
(CC).
validity of a change in zoning law allowing the building of a nursing

home, or group-care facility, in a wealthy neighourhood.)36
For left-wing parties to constitutional negotiations, therefore, the

constitutionalisation of a broad range of socio-economic guarantees
can be seen as providing a form of insurance against the risk that a
particular concession (in relation to the right to property) will lead to
a very large cost in terms of the scope for adopting progressive
legislation. Because the demand for such insurance is linked to the
grant of reciprocal insurance to right-wing parties to constitutional
negotiations, such insurance can also be seen as contributing to a
form of constitutional ‘insurance swap’ arrangement.
In an economic context, a swap is a contract for an exchange of

future cashflows. Parties engage in them to reduce risk in markets
such as those for currency, commodities, and interest rates. For
example, a risk-averse party might want to exchange the obligation
to pay a floating rate of interest for the obligation to pay a fixed rate.
Another example is a credit default swap, in which one party transfers
the risk of a credit default to another party that is (presumably) in a
better position to accept the risk.37
Both parties under such an arrangement hold mutual cross-

collateralised promises that hedge their risk to a certain degree, but
leave to it downstream or ‘market’ agents (here, a court) to
determine the precise value of the hedge for both sides. Both parties
also pay some form of ‘premium’ for the insurance they obtain; a
premium is a small but certain cost today, in order to avoid a larger
and more uncertain loss tomorrow.38
Why might parties decide to adopt a ‘swap arrangement’ of this

kind? The reasons are clearly multiple, but the key explanation, we
suggest, lies in the presence of both high bargaining costs for parties
in attempting to negotiate a complete and fully articulated balancing
of the competing constitutional demands, and strong preferences for
one or other party over a particular constitutional domain.
36
For the conflict between property owners and such facilities, compare eg City of
Cleburne v Cleburne Living Center, Inc. 473 US 432 (1985).
37 RM Stulz ‘Credit default swaps and the credit crisis’ (2010) 24 Journal of
Economic Perspectives 73 - 92. There is, unsurprisingly, a large literature on the
costs and benefits of credit default swaps in the wake of the 2008 financial crisis,
in which swaps on subprime mortgage-backed securities played a major role.
Some argue that the presence of swaps improves the speed of transmission of
market information and makes markets more efficient. Others have argued that
the ability to offload risk reduces incentives for monitoring. Stulz 76.
38
The basic idea of insurance is that involves paying some definite cost (i.e.
premium) in order to avoid a potential uncertain liability in the future.
If bargaining costs are low, the parties will simply negotiate all
the details of a particular constitutional arrangement. Bargaining
costs, however, may be high for a number of reasons. One factor will
be asymmetric information among parties to a constitutional
negotiation process, which can lead to a failure to reveal the basis for
a bargain. Another problem is that constitutional bargaining
sometimes has the character of a bilateral monopoly, in which two
groups are thrown together by historical circumstance into a nation
and they have no possibility of divorce. This can lead to efforts to
‘hold-out’ for a better agreement, making it difficult to conclude a
bargain. Another source of trouble, is the existence of constitutional
‘passions’,39 which may lead parties to reject pareto-improving
trades or agreements. If bargaining costs are high, parties will also
often respond by ‘deciding not to decide’ all relevant constitutional
details,40 and instead adopting broad constitutional standards, or
vague constitutional language, requiring key issues be decided by
future legislators or courts.41
Parties, however, may also have a strong interest in providing —

or at least being seen to provide — some general guidelines for
downstream decision-makers in certain areas. They may have strong
historical reasons for wishing downstream decision-makers to be
constrained in particular ways. Alternatively, parties may have
publicly committed themselves to certain constitutional positions in a
way that means that they will demand a form of reputational
premium from the other side, before being willing to drop their
demand for particular constitutional language.
Preferences of this kind will mean that the parties to

constitutional negotiations are willing to pay a real political price in
order to obtain protection for their agenda (or policies) from a court,
in the event of a bad political outcome. The form this price takes will
generally be quite similar for both right- and left-wing parties,
namely, acceptance of downstream judicial power to impose certain
limits on their own freedom of action, when in government, in return
for certainty that their agenda (or policies) will receive at least
partial protection from the court in the event of a bad political
outcome. The only difference between the two contexts is that, in the
case of left-wing parties, the left may have greater freedom to choose
between more and less expensive (and thus comprehensive) forms of
insurance, via the choice between justiciable socio-economic rights
39
J Elster ‘Forces and mechanisms in the constitution-making process’ 45 Duke Law
Journal 364 (1995).
40 See T Ginsburg & R Dixon ‘Deciding not to decide: Deferral in constitutional
design’ I-Con: the International Journal of Constitutional Law (2011).
41 As above.
guarantees and mere ‘directive principles’ of state policy.42 That is,

the left can calibrate the strength of demanded rights guarantees.
One bit of evidence for the plausibility of such constitutional

insurance swap arrangements, we suggest, is the co-occurrence of
socio-economic and property rights in national constitutions. To
evaluate this relationship, we used a summary index of socio-
economic rights generated by Ben-Bassat and Dahan.43 They examine
different levels of constitutional protection for five sets of rights
(social security, health, education, housing, and workers’ rights) for
67 nations. Using data from the Comparative Constitutions Project,
we generated a variable to capture strong protection of private
property, which was coded 1 if a country’s constitution explicitly
provides for ‘full’ ‘adequate’ or ‘just’ compensation for takings. The
socio-economics rights index and the property rights variables are
positively correlated, and strong property rights are a significant
predictor of a higher socio-economic rights index in numerous
multivariate specifications, controlling for wealth, democracy, the
year the constitution was adopted, and whether the country has
ratified the International Covenant on Social and Economic Rights.44
Strong protection for property thus seems to go along with bundles of
socio-economic rights.
In the South African context, we argue, sections 25-27 of the

Constitution also ultimately reflected exactly this kind of swap
arrangement when it came to the protection of the right to property
and other socio-economic rights.
42 This assumes, of course, that both sets of guarantees give courts some textual
basis for imposing affirmative limits on government action, as is evidenced, for
example, by the approach of the Supreme Court to India to the directive
principles contained in the Indian Constitution. See eg Olga Tellis & Ors v Bombay
Municipal Council [1985] 2 Supp SCR 51. It also assumes, however, that courts are
more likely to impose affirmative limits on governments where the constitution
explicitly authorises this, than where it does not. See eg G Hogan ‘Directive
principles, socio-economic rights and the Constitution’ Irish Jurist, xxxvi (2001).
Right-wing parties, of course, also have the option of choosing more or less
comprehensive forms of insurance, via the selection of ‘strong’ versus ‘weak’
models of judicial review. Doubts have been raised, however, about the stability
of such a choice: see Tushnet (n 23 above); M Tushnet ‘The rise of weak-form
review’ in T Ginsburg & R Dixon (eds) (2011) Comparative Constitutional Law 321
- 333.
43 A Ben-Bassat & M Dahan ‘Socio-economic rights in the constitution and in
practice’ (2008) 36 Journal of Comparative Economics 103.
44
In some specifications, the significance was only at the 85% confidence level.
Data is available from authors.
3 The South African constitution and insurance

swaps
The 1996 Constitution was, of course, the product of a unique two-

stage process of constitutional drafting.45 The first stage involved the
negotiation, between the National Party (NP), African National
Congress (ANC) and other key players, of an interim constitution to
govern during a two-year period of transitional government; and the
second, the election of a Constituent Assembly to draft a final
democratic constitution. The two stages were also linked in a crucial
and novel way, by a form of constitutional ‘certification’
requirement, according to which a newly created constitutional court
was required to certify that the final Constitution was consistent with
32 fundamental constitutional principles set out in a schedule to the
interim Constitution.
The negotiation of the interim Constitution itself also took place

over many stages. The first formal multi-party talks began in 1991 at
the first Congress for a Democratic South Africa (CODESA I); and then
resumed, in 1992, at CODESA II. These negotiations, however, broke
down completely in June 1992, following a massacre of 49 black South
Africans at Boipatong. It took many more rounds of bilateral
negotiations between the NP and ANC, and a third round of formal
multi-party talks called the Multi-party Negotiating Process (MPNP) at
the World Trade Centre in Johannesburg in 1993, to finally come to
agreement on the constitutional transition process. A key decision
was to adopt an Interim Constitution that would include a bill of
rights, and to allow the judiciary to certify the final Constitution for
conformity with certain constitutional principles.
The constitutional protection of property was from the outset of

these constitutional negotiations one of the areas of sharp
disagreement between the parties.46 The ANC, for example,
expressed concerns from the outset of negotiations at CODESA I about
the effect of including any right to property in a South African
constitution. The concern was that such a right could make it
impossible — or at least prohibitively expensive — for a future
democratic government to restore land wrongfully taken under
apartheid, or even to redistribute land and resources with a view to
45 For an excellent summary of this process, from which we borrow in our summary
below, see H Ebrahim The soul of a nation: Constitution-making in South Africa
(1998).
46 Chaskalson (n 28 above)
addressing homelessness.47 A number of leading ANC thinkers

therefore argued that that ‘the only way to achieve a true balance
between the rights of property-holders and [the] property-less [was]
to weaken existing property rights, as a matter of deliberate
policy’.48 In the early stages of work by the Technical Committee on
Fundamental Rights, at Kempton Park in 1992, the ANC argued against
the inclusion of any form of constitutional protection for the right to
property.49
It soon became clear, however, that this position was politically

untenable. It went directly against the expansive approach toward
the drafting of a bill of rights taken by the ANC-appointed members
of the Technical Committee, and, also, the importance of the right to
property to the National Party, as a form of political insurance (or at
least perceived insurance).50
The National Party (NP) was, as Mathew Chaskalson has noted,

intent from the outset of constitutional negotiations ‘on ensuring that
the property of existing white owners would be safe from the
depredations of a future democratic government’.51 In 1993, it
therefore argued for the inclusion of an extremely strong form of
constitutional protection for the right to property, which required
both that any taking (or expropriation) of property had to be for a
‘public purpose’ and at full market value,52 and that sought to
prevent the imposition of any form of tax imposing ‘unreasonable
inroads upon the enjoyment, use or value of such property’.53
The ANC thus quickly responded by adopting an alternative

strategy, which sought to limit the scope of such a property guarantee
in at least three key ways: first, by making the content of the right to
property a matter to be ‘determined by law’; second, by preventing
the exercise of such rights in a manner contrary to the ‘public
interest’; and third, by making the compensation for any taking of
property an amount that achieved an ‘equitable balance’ between
the public interest and the interests of the property owner.
47
See eg G Budlender ‘The right to equitable access to land’ (1992) 8 South African
Journal on Human Rights 295 304; AJ van der Walt ‘Development that may change
the institution of private ownership so as to meet the needs of a non-racial
society in South Africa’ (1990) 1 Stellenbosch Law Review 26. (suggesting that
‘the inclusion of the right to private property [in a democratic constitution could]
serve to reinforce [an] abstract and absolute concept of ownership’, in a way
which would then stand in the way of the emergence of a non-racial society).
48
Budlender (n 47 above) 304.
49 Chaskalson (n 28 above) 229.
50
On the latter, see Chaskalson (n 28 above) 226 - 228.
51 Chaskalson (n 28 above).
52 Republic of South Africa ‘Government’s proposal on a Charter of Fundamental
Rights’ (2 February 1993) 11. See also discussion in Chaskalson (n 28 above) 224.
53 Government’s Proposal (n 52 above).
The property clause adopted in section 28 of the 1993 Constitution

was also ultimately a true compromise between the divergent
positions of the Government and ANC (and its allies) on these issues.54
On the issue of the restitution of land, for example, section 28

was silent in a way that reflected an important victory for the
Government on the issue of property rights. At the same time, by
linking the level of compensation payable for any expropriation of
land to the ‘history of its acquisition’, the clause also gave an
important victory to the ANC. In fact, on the issue of compensation
for the taking (or expropriation) of property more generally, section
28(3) established a near perfect compromise between the standards
advocated by the Government and the ANC. By requiring, for
example, that any such compensation be ‘just and equitable’, taking
into account both the market value of particular property and a range
of other relevant factors, section 28(3) clearly rejected the
Government’s preference for full market value to be the sole
determinant of the relevant constitutional standard.55 At the same
time, by directing attention to factors such as ‘the use to which
property [was] being put, the value of the investments in it by those
affected and the interests of those affected’, the clause also adopted
an approach to questions of justice and equity that largely rejected
the approach of the ANC, which wanted to require courts to focus on
(and thus impose on individual property owners the cost of
supporting) the social benefit of a particular expropriation. The
consideration of investment value obviously weighed in favor of the
NP, whose constituents were much more likely to have sunk
investments in property.
On the issue of regulatory takings (i.e. the regulation rather than

taking) of property, section 28(2) struck a compromise between the
positions of the Government and the ANC. It rejected, for example,
the ANC approach of making property rights subject to legislative
definition and an overriding public interest test; but also allowed the
regulation of property rights, subject to certain procedural
requirements (i.e. an ‘in accordance with law’ test).56
In 1995, however, the ANC effectively had the opportunity to

renegotiate the terms of this compromise, without the same degree
of danger of holdout from the NP. As a legal matter, the ANC majority
had broad freedom under the final Constitution to redefine the right
54
Chaskalson (n 28 above) 226.
55 The provisions require that consideration be given to ‘all relevant factors,
including the use to which the property is being put, the history of its acquisition,
its market value, the value of the investments in it by those affected and the
interests of those affected.’ Constitution of the Republic of South Africa, sec
28(3).
56 Constitution, sec 28(2).
to property.57 National constitutions, as the Constitutional Court

noted in the First Certification Case, define the right to property in
wide variety of different ways: some constitutions give very broad
protection to such rights, while others provide no express protection
whatsoever.58 The right to property also finds limited protection in
international human rights law, given the absence of such a right in
both the International Covenant on Civil and Political Rights (ICCPR)
and International Covenant on Economic, Social and Cultural Rights
(ICESCR).59 There is no clear minimum content to the right to
property as a form of ‘universally accepted fundamental righ[t]’,
which the final constitution was bound to respect by virtue of
Schedule 3 to the Interim Constitution.
Politically, the ANC also had far greater bargaining power than in
1992-93, by virtue of its strength in the Constitutional Assembly (CA).
In South Africa’s first democratic elections, in 1994, to elect the CA,
the ANC won approximately 60% of the vote, compared to the NP and
DP’s combined total of roughly 25%. To adopt a constitution, by 2/3
majority, the ANC therefore needed the support of only a small
number of additional members of the Assembly, which it could
achieve without obtaining the support of any of the other major black
or white political parties.60
This left the ANC with a clear and important choice: either it
could attempt to redraft the language of the property clause, so as to
closer conform to its preferred position in 1993, or demand additional
concessions from the NP and DP in the form of cross-collateralised
constitutional guarantees or insurance — i.e. socio-economic rights.
To a large extent, the ANC executive also chose to pursue the second
of these options.
In the context of the property clause, the ANC ultimately sought

only quite limited changes to the existing text of the Interim
57 Compare A Sachs http://www.nelsonmandela.org/omalley/index.php/site/q/

03lv00017/04lv00344/05lv01183/06lv01235.htm (accessed 8 August 2011)
(suggesting that the two stage process of constitution-making in fact gave ‘gave
[the ANC] a chance to come back in and have lots of different things
acknowledged and accepted’ in this context, that they were not able to achieve
in 1993).
58 Certification of the Constitution of the Republic of South Africa 1996 10 BCLR
1253 (CC) paras 72 - 73.
59 As above.
60 See eg S Daley ‘A new charter wins adoption in South Africa’ New York Time 09
May 1996 http://www.nytimes.com/1996/05/09/world/a-new-charter-wins-
adoption-in-south-africa.html?src=p (accessed 8 August 2011).
Constitution.61 It argued, for example, in its submission to the CA that

‘measures aimed at bringing about land reform for the benefit of
people previously disadvantaged by unfair discrimination’ should be
expressly excluded from the scope of the clause. And it sought to
reassert the idea that the public interest should be considered in
determining compensation for the expropriation of property,
alongside those factors set out in the interim Constitution.62
Otherwise, however, it rejected arguments from other members of
the governing ‘tripartite alliance’ (i.e. the Congress of South African
Trade Unions (COSATU) and the South African Communist Party (or
SACP)) that the property clause should be wholly excluded, or
radically redrafted.63
This ultimately meant that the core compromise made in 1993

regarding property remained largely intact in 1996. The ‘fair and
equitable’ standard adopted in 1993, for example, was retained in
1996.64 Section 25(1) of the 1996 Constitution also clearly continues
to apply even in the context of efforts at land restitution and
redistribution, in a way strongly opposed by the ANC in 1993.65
61
It argued, for example, that ‘measures aimed at bringing about land reform for
the benefit of people previously disadvantaged by unfair discrimination’ should
be expressly excluded from the scope of the clause. It also sought to reassert the
idea that the public interest should be considered in determining compensation
for the expropriation of property, alongside those factors set out in the Interim
Constitution. See ANC ‘African National Congress (ANC) preliminary submission on
land rights’ sec 3.1.2 (arguing that compensation should ‘establish an equitable
balance between the public interest and the interests of those affected’).
62 See ANC Land Rights (n 61 above) sec 3.1.2(a) (arguing that compensation should
‘establish an equitable balance between the public interest and the interests of
those affected’).
63 The final draft did also add an extensive set of sub-articles which addressed
demands for land reform in greater detail than had the 1993 text
64 Constitution sec 26(3). ‘The amount of the compensation and the time and
manner of payment must be just and equitable, reflecting an equitable balance
between the public interest and the interests of those affected, having regard to
all relevant circumstances, including: (a) the current use of the property; (b) the
history of the acquisition and use of the property; (c) the market value of the
property; (d) the extent of direct state investment and subsidy in the acquisition
and beneficial capital improvement of the property; and (e) the purpose of the
expropriation.’
65 Sec 25 of the Constitution reads in part ‘(4) For the purposes of this section:
(a) the public interest includes the nation's commitment to land reform, and to
reforms to bring about equitable access to all South Africa's natural resources;
and
(b) property is not limited to land.
(5) The state must take reasonable legislative and other measures, within its
available resources, to foster conditions which enable citizens to gain access to
land on an equitable basis.
(6) A person or community whose tenure of land is legally insecure as a result of
past racially discriminatory laws or practices is entitled, to the extent provided
by an Act of Parliament, either to tenure which is legally secure or to comparable
redress.
(7) A person or community dispossessed of property after 19 June 1913 as a result
of past racially discriminatory laws or practices is entitled, to the extent provided
by an Act of Parliament, either to restitution of that property or to equitable
redress.
Compared to 1993, however, the ANC in 1995 also placed far

greater emphasis on the constitutonalisation of various socio-
economic rights, as a means of checking an overly expansive reading
of the right to property.66
In 1993, the ANC argued for the recognition of a range of socio-

economic rights under the interim Constitution, including the ‘right to
enjoy basic social, educational and welfare rights’ for all men, women
and children; a land rights clause, which recognised ‘access to land’
as the ‘birthright of all South Africans’; and a negative right to
shelter, in the form of a right not to be removed from one’s home,
except by court order, and after consideration by the court of the
existence of potential reasonable alternative accommodation.67
This reflected the increasing belief among key ANC thinkers that
constitutional protections for socio-economic rights were ‘indivisible’
from and ‘interdependent’ with the recognition of civil and political
rights,68 and, in addition, the arguments by leading constitutional
thinkers, such as Etienne Mureinik and Nicholas Haysom, that if the
constitution were seen ‘to institutionalise and guarantee only
political/civil rights and ignore the real survival needs of the people’,
it would ‘find no lasting resonance among’ the majority citizens, as
‘the true guardians’ of the constitution.69 Or that, as Etienne Mureinik
put it, if a bill of rights contained only first-generation rights, it would
be ‘perceived to be elevating luxuries over necessities’ and thus as
simply a ‘charter of luxuries’70 that would find limited support from
the majority of black citizens who were not only deprived of civil and
political rights under apartheid, but also subject to severe forms of
economic deprivation at the hands of the apartheid state.
65
(8) No provision of this section may impede the state from taking legislative and
other measures to achieve land, water and related reform, in order to redress the
results of past racial discrimination, provided that any departure from the
provisions of this section is in accordance with the provisions of section 36(1).
(9) Parliament must enact the legislation referred to in subsection (6).
66 In this context the ANC reasserted its previous argument that the constitution
should contain ‘a positive right to land’, a right not to be ‘unlawfully evicted
from accommodation occupied by him/her without the legal process having been
invoked and a court order obtained’; and an obligation on the state ‘within the
limits of its available resources, to provide adequate shelter for all’. See ANC
‘Preliminary ANC submission Theme Committee 4 — Further socio-economic
rights’ sec 1.A.
67
[AB to insert].
68
This, of course, is the international law understanding. For discussion of this, and
the divide in the ANC over acceptance of the idea, see eg N Haysom
‘Constitutionalism, majoritarian democracy and socioeconomic rights’ 8 (1992)
South African Journal on Human Rights 451.
69 Haysom (n 68 above) 454.
70
E Mureinik ‘Beyond a charter of luxuries: Economic rights in the Constitution’
(1992) 8 South African Journal on Human Rights 464 465.
The ANC’s emphasis, however, was still largely on the recognition

of such rights as abstract goals or directive principles,71 to be realised
by a future democratic legislature under a series of ‘by law’ clauses,72
rather than on a role for courts in enforcing such rights as a constraint
on future democratic legislatures.
In 1995, by contrast, the ANC moved clearly toward support for

the constitutionalisation of socio-economic rights qua positive,
justiciable guarantees, arguing that ‘[t]he new Bill of Rights [should
not] shy away from including within the scope of its protection,
fundamental rights, which while posing difficulties in enforcement
reflect important principles in the promotion of a society based on
justice and equality, a society which seeks to redress the imbalances
of the past’.73 It also gave specific support, in its submission to the
Constituent Assembly, to the constitutionalisation of judicially
enforceable rights to shelter, health care, food, water and social
assistance.
The NP and DP, during this same period, also shifted markedly in
their position toward the constitutionalisation of such rights. In 1993,
the DP in particular argued strongly against the constitutionalisation
of such rights in 1993, especially rights such as the right ‘not to be
evicted from one’s lawful home’, as a ‘fundamental invasion of the
right to private property’.74 The DP was also instrumental, at the Ad
Hoc Committee stage, in ensuring that such a right was omitted from
the text of the 1993 Constitution.75 In 1995, however, both the NP and
DP dropped their opposition to the constitutionalisation of all relevant
71
For contemporaneous arguments in favour of a directive principles-based
approach, see eg DM Davis ‘The case against the inclusion of socio-economic
demands in a bill of rights except as directive principles’ (1992) 8 South African
Journal on Human Rights 475.
72
Art 11(2) of the draft bill of rights, for example, provided that ‘legislation shall
ensure the creation of a progressively expanding floor of minimum rights in the
social, educational and welfare spheres for all in the country’. Art 11(8) likewise
provided that, in the context of the right to health, that ‘a comprehensive
national health service shall be established linking health workers, community
organisations, state institutions, private medical schemes and individual medical
practitioners so as to provide hygiene education, preventative medicine and
health care delivery to all’. On the function of such by law clauses generally, see
Dixon & Ginsburg (n 40 above).
73
See eg ANC Theme 4 (n 66 above) (‘it is our firm belief that rights to social
assistance, food and water be included within the Bill of Rights’).
74
See DP submission 1993 327 - 328 (arguing that it would also jeopardise the
erection of new housing stock and deter financial institutions from granting bonds
prospective homeowners in the lower income category’).
75
R Spitz & M Chaskalson, The Politics of transition: A hidden history of South
Africa's negotiated settlement (2000) 328.
socio-economic rights, including the right not to be arbitrarily evicted

from one’s home.76
The inclusion of various socio-economic rights guarantees in

sections 26-29 of the Constitution, we suggest, also ultimately seems
to have had exactly the kind of effect an insurance swap-based theory
would predict, namely: to have promoted agreement between the
parties in an area that had involved significant decision costs in the
negotiation of the interim Constitution.
Further, for some within the ANC, at least, the inclusion of such
rights in the Constitution involved exactly the kind of political cost,
or constraint, implicit in an insurance-based theory. In 1993, such a
constraint would have appeared less salient (or costly) to key players
within the ANC, because the party as a whole was committed to the
economic approach embodied in the 1994 Reconstruction and
Development Program (RDP), which included a commitment to
meeting citizens’ ‘basic needs’ in much the same way contemplated
by socio-economic rights guarantees such as sections 26-29 of the
Constitution.77 By 1995, however, leading figures within the more
76
NP ‘National Party submission item 14: Socio-economic rights’ (1995), DP
‘Democratic Party submission on: Socio-economic rights’ (1995). For the initial
opposition of these parties to such clauses, see eg S Liebenberg ‘Socio-economic
rights’ in M Chaskalson et al (eds) Constitutional law of South Africa (1999).
77 The only ‘basic needs’ recognised under by the RDP, but not by the Constitution,
are needs relating to transport and telecommunications. ‘Land reform’, for
example, is one of the ‘basic needs’ listed in Chapter 2 of the 1994 RDP White
Paper; and sec 25(5) of the Constitution recognises a right on the part of citizens
‘to gain access to land on an equitable basis’, while sec 25(8) makes clear that
land reform is contemplated. The RDP recognises ‘housing and services [and]
water and sanitation’ as basic needs, while sec 26 of the Constitution recognises a
right of access to adequate housing, which some argue includes not only a right of
access to housing, but also related services such as sanitation, energy and
electrification. (This argument was in fact made in the 2010 Term itself, in Leon
Joseph v. City of Johannesburg, See para 32 in the context of a decision by
Johannesburg power to terminate the supply of electricity to the petitioners,
without direct notice to them as tenants (rather than to their landlord)).
Similarly, The RDP recognises ‘the environment, nutrition, health care, social
security and social welfare’ as core ‘basic’ priorities for the government; and sec
27(1) of the Constitution recognises a right of access to ‘health care services,
including reproductive health care; sufficient food and water; and social security,
including… appropriate social assistance’, while sec 24 recognises a right ‘to an
environment that is not harmful to their health or well-being’.
In areas in which the two overlap, the Constitution is also clearly informed by the
guidelines set out in various parts of Chapter 2 of the RDP. For example, in
endorsing ‘housing [as] a human right’ that government was ultimately
responsible for ensuring universal access to, the RDP acknowledged that meeting
this obligation would involve some delay, and the government’s approach to
housing must ‘take account of funding and resource constraints.’ See Restructure
and Development Program 1996 secs 2.5.5 - 2.5.6. The language in secs 26(2) and
27(2) of the Constitution requiring the state to take ‘reasonable legislative and
other measures, within its available resources, to achieve the progressive
realisation’ of various rights also directly parallels this understanding.
centrist or ‘pragmatic’ faction of the ANC,78 such as Thabo Mbeki (as

then Deputy-President) and (soon to be Minister for Finance) Trevor
Manuel, had already begun work on the pre-cursor to ‘GEAR’, the
‘Growth, Employment and Redistribution’ (GEAR) policy, as an
overlay to the government’s earlier Reconstruction and Development
Program (RDP)).79 While the aim of GEAR was to stimulate both
economic growth and job-creation (and thus said to be fully
consistent with the aims of the RDP),80 it included a number of ‘pro-
market’ policies aimed at reducing government borrowing, and
increasing private investment, that potentially conflicted with the
vision of the state as playing a ‘lea[d] role in building an economy
which offers to all South Africans the opportunity to contribute
productively’ — endorsed under both the RDP and sections 26-29 of
the Constitution.81
For this centrist faction in particular, therefore, there likely was

a potential for conflict between a decision to entrench socio-
economic rights qua positive rights in the Constitution, and the
freedom to pursue its preferred approach to economic management.
The fact that the party as a whole nonetheless took this step provides
suggestive evidence of the importance, for others within the ANC, of
preserving scope to pursue certain kinds of progressive/redistributive
economic measures — without fear of challenge under the property
clause.82
78 For this characterisation of the relevant factional politics, see eg T Roux The
politics of principle: The first South African Constitutional Court, 1995 - 2005
(forthcoming, Cambridge UP 2013).
79 This was admittedly in part due to increasing pressure on the currency, and
appointment of Trevor Manuel to the ministry of finance in 1996. See T Lodge
Politics in South Africa: From Mandela to Mbeki (2002) 26. However, the latter
development in particular was likely quite foreseeable to key COSATU and SACP
figures at an earlier stage. See eg T Lodge ‘The ANC and the development of
party politics in modern South Africa’ (2004) 42 Journal of Modern African
Studies 189-219; DT McKinley ‘Democracy, power and patronage: Debate and
opposition within the African National Congress and the Tripartite Alliance since
1994’ (2001) 8 Democratization 183 - 206. For a discussion of subsequent shifts
rightward within the ANC see also Roux (n 78 above) ch 8; D Davis ‘Socio-
economic rights in South Africa: The record of the Constitutional Court after ten
years’ (2004) 5 ESR Review: Economic and Social Rights in South Africa 3.
80 See eg T Manuel ‘Economic policy and South Africa’s growth strategy’ Ministry of
Finance 19 March 2007 http://www.treasury.gov.za/comm_media/speeches/
2007/2007031901.pdf (accessed 8 August 2011).
81 Reconstruction and Development Plan 1994 para 2.3.1 http://www.nelson
mandela.org/omalley/index.php/site/q/03lv02039/04lv02103/05lv02120/06lv02
126.htm (accessed 8 August 2011).
82 On the nature of this as a priority for many within the ANC, see eg G Budlender
‘The right to equitable access to land’ (1992) 8 South African Journal on Human
Rights 295 304; Chaskalson (n 28 above) 229.
4 Insurance swaps and judicial review
Given its potential relevance in South Africa in 1995, an insurance

swap-based theory of judicial review has potential normative
relevance for ‘originalist’ or backward-looking forms of interpretation
by a court such as the Constitutional Court.
Originalism, obviously, is a contested interpretive stance, and it

does not follow automatically from an insurance-based theory that
courts should pay attention to the founding bargain between parties
when interpreting a constitution. Parties must certainly believe that
courts will show some fidelity to the text and/or history behind
relevant constitutional guarantees, in order for an insurance-based
theory to operate. But this does not mean that courts must always
vindicate this belief, ex post, in order to facilitate efficient
constitutional bargaining. On the contrary, from a purely pragmatic
perspective, courts will often be free to depart from the terms of the
original political bargain, without undermining constitutional
efficiency. The only exception to this will be where constitutional
bargaining remains ongoing, by virtue, for example, of transitional
arrangements such as under the 1993 interim constitution in South
Africa, or a credible threat by a key constitutional player to ‘exit’ an
entire constitutional system, by resort to (say) violence, or the large-
scale withdrawal of capital.83
Courts, however, often choose to look to the historical context

behind various rights, either out of a sense of legal obligation, or as
a matter of interpretive preference. Indeed, the SACC has itself quite
clearly endorsed the relevance of attention to historical
understandings and context in the interpretation of sections 25-29. In
the context of section 26, for example, the Court has consistently
suggested that attention must be given to both the relevant ‘social
and historical context’ for the adoption of a right to housing.84 In the
context of the right to health, and specifically section 27(2), the
Court suggested, in Soobramoney, that provisions such as section 27
must be considered in their full context, ‘which includes the history
and background to the adoption of the Constitution, other provisions
of the Constitution itself and, in particular, the provisions of [the bill
of rights] of which [they are] part’.85 And in the context of the right
to water, in Mazibuko, the Court suggested that the first (and most
important?) background fact to consider in interpreting section 27(2)
was again the legacy of apartheid and its effect on the living
83 Machele (n 15 above); Joe Slovo II (n 17 above).

84
See eg Grootboom (n 35 above) para 25; Joe Slovo I (n 12 above) 191.
85 Soobramoney v Minister of Health (Kwazulu-Natal) 1997 12 BCLR 1696 para 16.
conditions of black South Africans in areas such as Phiri in Soweto.86
An insurance swap-based theory, therefore, also has potential

relevance in providing guidance to the Constitutional Court as to the
direction of relevant historical understandings.
4.1 A pro-balancing approach
In cases of direct conflict between property and other socio-economic

rights, an insurance swap-based theory suggests that, even from an
originalist perspective, South African courts have in fact been charged
with the task of creating, rather than giving effect to some pre-
existing, balance between rights.87 This is a consequence of the
existence of high bargaining costs between the ANC and NP and DP
over the issue of the constitutional protection of property, and the
consequent decision by both sides to ‘decide not to decide’ the
controversial details in this area. This assignment of balancing power,
as Justice Sachs has noted, also implies an important degree of
judicial freedom, and responsibility, in interpreting provisions such as
sections 25-29.88
At the same time, an insurance swap-based perspective suggests

that, from the drafters’ perspective, it is also centrally important that
courts should in fact attempt to maintain such a balance: this, after
all, was the key reason that both parties preferred an insurance swap
to an approach that involved a greater loss for one party, in return for
reciprocal concessions from the other party in another area. To
achieve this balance, we suggest, it will also be essential for a court
to adopt at least two basic strategies: first, to invalidate any statutory
or common law presumption in favour of one set of rights at the
expense of the other; and second, to ensure that, in its own reasoning
about these rights, it adopts as careful, context-sensitive approach as
possible.
Over time, almost any legal presumption will tend to give priority
to one set of rights over another. This will be true even for quite weak
presumptions, which are merely procedural in nature: with enough
litigation, the law of large numbers will inevitably mean that, without
some equally strong counter-pressure, there is a tilt in the law toward
results that reflect such a presumption. Thus, if a court allows
86
Mazibuko (n 10 above) para 10.
87 Port Elizabeth Municipality v. Various Occupiers 2004 12 BCLR 1268 (CC) paras 22
- 23 (noting that ‘the way in which the courts are to manage the process [of
balancing competing rights] has … been left as wide open as constitutional
language could achieve, by design and not by accident, by deliberate purpose and
not by omission’).
88 Port Elizabeth Municipality (n 87 above).
presumptions to operate, it will have little chance of maintaining a

true balance between competing rights, across the constitutional
system as a whole.
Similarly, broad forms of reasoning by a court about the scope or

priority of particular rights will have the potential to undermine the
balance between the relevant rights, across the constitutional system
as a whole. Courts cannot, of course, wholly avoid deciding in favor
of a particular right or rights claimant in a system of concrete review.
They can, however, seek to limit the system-wide consequences of
such decisions, by either avoiding particular constitutional questions,
or resolving particular constitutional controversies in as narrow,
context-specific a way as possible.89
This kind of commitment to context-sensitive balancing in cases

of direct rights conflict will also have important implications in cases
in which just one side of the relevant rights equation is involved.
Otherwise, courts may find that where a case of actual conflict does
arise, they are pre-committed to preferring a quite abstract, rule-like
resolution of the particular conflict. It will therefore be important for
courts, in all cases, to avoid broad statements in favor of either a
wholly non-deferential or deferential — or strong or weak — approach
to the enforcement of socio-economic rights.
In the South African context in particular, this is consistent with

the rejection of both the idea of a ‘minimum core’ or ‘minimum
content’ to various socio-economic rights and the suggestion that
courts should consistently defer to legislative or executive judgments
about reasonableness when in it comes to the realisation of these
rights. By definition, the idea of there being a ‘minimum core’ to
various rights, under sections 26(2) and 27(2) of the Constitution,
involves a relatively broad statement by the Court, about the
(presumptively) non-derogable scope of such rights in a particular
case. It also, however, inevitably involves the Court making even
broader, more abstract statements about the kinds of values that will
inform the interpretation of such rights, in future cases, because of
the relationship between how one defines such a core and the priority
given to other constitutional values, such as rights to life, dignity and
equality.90
On the other hand, suggestions by the Court that the government

has unlimited freedom to make judgments about reasonableness
89
As Frank Michelman has noted, such an approach allows for the resolution of a
particular case ‘without predetermining so many others that one “side”
experiences large-scale victory or defeat’, see F Michelman ‘Foreword: Traces of
Self-Government’ (1986) 100 Harvard Law Review 4 34.
90 Dixon (n 23 above).
under sections 26(2) and 27(2), will be equally detrimental to a

context-sensitive, pro-balancing type approach — by undermining any
claim a government has to be acting pursuant to predefined legal
obligation, rather than making a pure policy judgment, in cases where
its actions are challenged on constitutional property rights grounds.
4.2 The Court’s record thus far
The approach of the Constitutional Court, we suggest, has in fact

shown a quite striking degree of consistency to date with all these
various aspects of ‘pro-balancing’ approach.91
In cases involving statutory presumptions in favour of a right to

property, the Court has consistently voted to invalidate such
presumptions, in favor of a more case-by-case, contextual approach
to balancing. This is equally true for recent cases, such as Abahlali
Basemjondolo Movement of South Africa v Premier of KwaZulu-
Natal92 and Gundwana v Steko Development,93 as for earlier such as
Jaftha v Schoeman.94
In Abhalali, the provincial legislation under challenge (i.e. the

KwaZulu-Natal Elimination and Prevention of Re-emergence of Slums
Act 6 of 2007) purported to require, in certain circumstances, both
private property owners and municipalities to invoke the procedures
for eviction provided for by section 6 of the PIE, the security of tenure
law.95 It thus increased the likelihood that eviction proceedings would
be commenced without any engagement with unlawful occupiers, or
consideration of other alternatives. At the same time, the Act could
also be seen as aimed at protecting a right to private property in at
least two ways: first, by requiring municipalities to expend the
resources — and political capital — necessary to protect property
owners against unlawful occupation in areas deemed to be a ‘slum’;96
and second, by creating procedures to protect the value of property
in neighboring areas. It thus sought to create a form of procedural
presumption in favour of the right to property over existing rights of
access to housing.
91 We do not mean to suggest that, in doing so, the Court has necessarily been
influenced by an insurance-swap understanding, but merely, that it has acted
consistently with such an understanding.
92 Abahlali (n 14 above).
93
Gundwana (n 16 above).
94
Jaftha v Schoeman 2005 2 SA 140 (CC).
95 KwaZulu-Natal Elimination and Prevention of Re-emergence of Slums Act 6 of
2007, secs 9 - 13.
96 It should be noted that, in the view of the majority, the legislation may well have
applied more broadly, to areas that were simply informal settlements, but even if
confined in this way, as Yacoob J. dissenting favored, the majority found that it
was still invalids: see Abahlali (n 14 above) 107.
The majority of the Court, however, held that under section 26(2)
of the Constitution, such a procedural presumption was
unconstitutional. The ‘compulsory nature’ of the relevant provisions,
the Court held, ‘disturb[ed] [the] carefully established legal
framework’ for evictions established by both section 26(2) and the PIE
and national Housing Act.97 Under this framework, the Court further
held, property rights could legitimately be protected by eviction
procedures, but only providing that housing rights were
simultaneously protected by the giving of proper notice to unlawful
occupiers, and by insisting that decision-makers consider all other
possible alternatives.98
Even more recently, in Gundwana, the Court adopted the same

kind of approach to High Court Rules permitting a registrar to declare
mortgaged property specially executable, as part of granting default
judgment.99 By allowing for cheaper, and more expeditious,
execution proceedings in cases of default by a borrower, these rules
clearly sought to enhance the property rights of existing
mortgagees.100 They also did so in a way that gave clear procedural
priority to these rights over the rights of South Africans in economic
distress to retain access to existing housing. High Court procedures,
for example, clearly allow a borrower to apply to the court to have a
default judgment set aside. Borrowers, however, as Froneman J
noted, can often be unaware of such procedures, or ‘too poor to make
proper use of them’, so that an initial — more or less routine — order
for execution by a registrar effectively becomes final in many
cases.101
The Court’s response was, once again, to insist on the

inconsistency between this kind of procedural presumption in favor of
a right to property and the insurance (swap) created by sections 26(1)
and (3). The presence of sections 26(1) and (3) in the Constitution, the
Court held, implies that any order for execution in respect of
immovable property must be a ‘proportionate means’ of protecting
the right to property of mortgagees.102 For this to be the case, the
Court further held, any application for such an order must also be
subject to a careful ‘evaluation’ by a judge, rather than automatically
granted.103 Without such an evaluation, the Court suggested, it would
be impossible to determine whether there were other reasonable
means of satisfying a judgment debt, and thus whether the burden of
97
Abahlali (n 14 above) para 122.
98
As above.
99 For the relevant effect of the rules in this respect, see Gundwana (n 16 above)
para 35 - 36.
100 Gundwana (n 16 above) para 37.
102
Gundwana (n 16 above) para 54.
such an order on those ‘who are poor and at risk of losing their homes’
was in fact justifiable in the particular circumstances.104
This also directly parallelled the approach of the Court in Jaftha,

where the legislation under challenge was part of a broader scheme
dealing with the satisfaction of judgment debts that not only allowed
but actually required a clerk of the High Court to issue a warrant for
the execution of immovable property where they were satisfied that
there was insufficient moveable property to satisfy a judgment debt
(including a judgment entered by default).105 In striking down the
relevant law as an unjustifiable infringement of the right to housing
under sectoin 26(1) of the 1996 Constitution, Justice Mokgoro placed
strong emphasis on the degree to which such a requirement created a
constitutionally impermissible presumption in favor of the rights of
creditors (or property), as opposed to homeowners. Instead, the
justifiability of ordering the execution of a home depended on a range
of contextual factors, such as the economic position of the parties,
the hardship or prejudice to the homeowner of ordering execution,
and the degree to which the parties were acting in an informed and
good faith way in the relevant context.106
In other cases involving a potential direct conflict between

property and other socio-economic rights, the Court has consistently
and quite explicitly insisted on the need for a context-sensitive, ‘all
things considered’ approach to balancing competing rights. This
comports with the insurance-swap rationale.
In the first case in this category, Port Elizabeth Municipality,107

for example, in overturning an order for eviction under the PIE, the
Court noted both the degree to which the statute was designed to give
effect to constitutional imperatives; and the degree to which those
imperatives themselves called for ‘concrete and case-specific
solutions’ to the problem of balancing on the part of the Court.108
Thus, under provisions such as sections. 25 and 26(1) and (3), Justice
Sachs suggested, the judicial function was distinctly ‘not to establish
a hierarchical arrangement between the different interests involved,
privileging in an abstract and mechanical way the rights of ownership
over the right not to be dispossessed of a home, or vice versa’ but
‘rather … to balance out and reconcile the opposed claims in as just

105
See Magistrates’ Courts Act 32 of 1944 sec 67, and discussion of its effect at
Jaftha (n 94 above) paras 14 - 16.
106 Jaftha (n 94 above) para 43 (noting that the section is ‘sufficiently broad to allow
sales in execution to proceed in circumstances where it would not be justifiable
for them to be permitted’), paras 40 - 42 (setting out the range of potentially
relevant considerations).
107
Port Elizabeth Municipality (n 87 above).
108 Port Elizabeth Municipality (n 87 above) para 22.
a manner as possible taking account of all the interests involved and

the specific factors relevant in each particular case’.
In the 2010 Term, in Juma Musjid Trust v MEC,109 the Court took
a similar approach to balancing competing rights to private property
and education under sections 25 and 29 of the Constitution. The issue
facing the Court in Musjid was whether a private trust was entitled to
evict from its property a public school that it had helped establish,
but which the state had taken over, only to default on the rental
payments. The issue, the Court held, involved a direct conflict
between the property rights of the trust and the rights of students (or
‘learners’) to have ‘access to basic education’ under section 29, given
that the Trust itself was under a clear horizontal duty to respect such
rights. The test for the resolving this conflict, the Court suggested,
was inherently case-specific, involving an all-things-considered
judgment about reasonableness — or whether the ‘Trustees acted
reasonably in seeking an order for eviction’.110 In applying this test,
the Court also looked to a range of context-specific factors, such as
the degree to which the trust had sought to resolve the matter by
other means, given notice to the MED of its intention to proceed with
the order, and also been willing to delay the effect of any such
eviction.111
The Court has also taken a similarly narrow, context-sensitive

approach even in cases involving no such immediate conflict of rights,
and in doing so, rejected calls for it to endorse both a more absolutist
‘minimum core’ or ‘minimum content’-based and more deferential
approach. In Government of South Africa v Grootboom,112 for
example, in deciding whether the government’s failure to provide
basic shelter to the applicants (who were rendered homeless by a
prior eviction order) was in breach of section 26(2), the Court was
urged to endorse the idea of a ‘minimum core’ to the right to shelter.
The Court, however, rejected this approach, instead insisting on the
need for a more case-by-case approach to determining the contours
of reasonableness under section 26(2), designed to take account of
the actual resources available to the state, the diverse needs of
different groups in relation to access to housing, and the full range of
information available to the Court about the state of housing needs
and development.113 At the same time, it also rejected the suggestion
that it should give complete deference to the political branches,
under sections 26(2)-27(2), by holding that no housing policy could be
reasonable unless it constituted ‘a coherent public housing
109
Musjid (n 4 above).
110 Musjid (n 4 above) para 62.
111 Musjid (n 4 above) paras 62 - 65.
112
Grootboom (n 35 above).
113 Grootboom (n 35 above) paras 32 - 33.
programme directed towards the progressive realisation of the right

of access to adequate housing’, was ‘capable of facilitating the
realisation of the right’, was reasonable in design and
implementation, and provided some form of relief state must provide
for relief for ‘those in desperate need’.114
Similarly, in Minister for Health v Treatment Action Campaign (No

2),115 in the context of the right of access to health-care under
section 27(2), the Court rejected the idea of a minimum core in favour
of an incremental, context-sensitive approach to reasonable-ness,
suggesting that the Court itself was ill-equipped ‘to make the wide-
ranging factual and political enquiries necessary for determining …
minimum-core standards’.116 But in doing so it also affirmed that, to
be reasonable, a policy must not ignore ‘those whose needs are the
most urgent and whose ability to enjoy all rights therefore is most in
peril’, and take account of the special vulnerability of particular
groups, such as the poor in rural areas.117
More recently, in Mazibuko v City of Johannesburg,118 the Court

affirmed both aspects of this approach, this time in the context of the
right of access to water. It thus both rejected arguments that section
27(1) required the Court to quantify the amount of water sufficient
for dignified life,119 and insisted that the idea of reasonableness,
under section 27(2), had four basic minimum components: the
requirement that the government take some affirmative steps to
realise relevant rights; that these steps be reasonable; not
unreasonably fail to make provision for those most desperately in
need; or involve unreasonable limitations or exclusions; and be
subject to ‘continua[l] ... review ... to ensure that the achievement
of the right is progressively realised’.120 It also explicitly rejected the
idea that notions of ‘democratic deference’ required complete
deference by the Court to the government, or governing faction(s) of
the ANC, about notions of reasonableness, rather than a more
participatory democratic process in which the Court itself played a
much more central role in promoting democratic accountability.121
114 Grootboom (n 35 above) paras 40 - 41, 66.

115
Minister of Health and Others v Treatment Action Campaign and Others (No 2)
2002 10 BCLR 1033.
116 n 116 above, paras 32 - 39.
117
Minister of Health (n 116 above) paras 68 - 70.
118
Mazibuko (n 10 above).
119 Mazibuko (n 10 above) para 59 (suggesting that what sec 27(2) requires ‘will vary
over time and context’). For the similarities between this argument and the
minimum core argument in Grootboom and TAC, see Mazibuko (n 10 above) paras
51 - 52.
120
Mazikbuko (n 10 above) para 67.
121 Mazibuko (n 10 above) para 81. This also accords with what Etienne Mureink
argued in 1994 was a key benefit of including socio-economic rights in the 1994
Constitution, namely, their capacity to promote a culture of justification or
greater accountability, within the government. See Mureinik (n 70 above).
5 Conclusion
Constitutions often involve inter-temporal insurance guarantees,

whereby parties (and factions within parties) seek to reduce the risk
of downstream policy drift. We have argued that the complex bargain
among property and socio-economic rights in South Africa’s 1996
Constitution can be understood using the metaphor of insurance
swaps, in which two opposing factions agree to accept each others’
downstream guarantees, even if the two are in some tension with
each other. This effectively delegates the precise tradeoffs to
downstream constitutional courts. So long as the courts can be trusted
to act as good faith interpreters, they will facilitate constitutional
bargains that might otherwise be unachievable. In South Africa, the
apparent tensions between progressive redistribution and security of
property were highly salient, but the Constitution was nevertheless
adopted, in part because of the insurance swap mechanism in our
view.
This view also has potential normative implications for

constitutional adjudication. While originalism is a contested
interpretive stance, the Court in South Africa has indicated a
willingness to consider historical understandings and context in the
interpretation of sections 25-29; and an insurance-based theory
provides quite clear guidance to the Court as to the direction in which
these kind of historical understandings point. It suggests, for
example, that where an insurance swap arrangement has applied,
courts should consider context and tradeoffs, and avoid deciding
claims in ways that systematically abnegates one or the other part of
the swap arrangement. It thus counsels an approach to constitutional
adjudication that is plainly evident in various Constitutional Court
decisions in the 2009, 2010 and 2011 Terms, namely: an attention to
the context and tradeoffs, rather than more absolutist notions of the
scope of socio-economic rights, or their ‘minimum core’. At the same
time, it also lends support to the Court’s claim to actively enforce
such rights, in appropriate cases, rather than simply defer to the
government in respect of all matters of social and economic policy.
RATIONALITY, REASONABLENESS,
PROPORTIONALITY: TESTING THE USE OF
STANDARDS OF SCRUTINY IN THE
CONSTITUTIONAL REVIEW OF LEGISLATION
Christian Courtis*
At least three cases decided by the South African Constitutional Court

in the 2010 term deal with constitutional challenges to legislation. In
this brief overview, I discuss the Court’s choice regarding the
standard of scrutiny, and the way in which the selected standard is
applied in each concrete case critically. Issues to be explored include:
what is the justification for choosing a particular test? What are the
practical consequences of applying different standards of scrutiny?
Does it really make a difference? To which degree does the Court
actually employ the sets of questions or prongs that it uses to
characterise each test? How intense is the scrutiny made under the
chosen test?
1 Background: rationality, reasonableness and

proportionality as standards in the previous
case law of the Constitutional Court
The development and adoption of stock formulas for these different

standards and their respective tests are, of course, not new in the
jurisprudence of the Constitutional Court of South Africa. The first
precedent where the Court makes use of the rationality standard
dates from the interim Constitution period. The Court has gradually
defined its approach to their use in a series of cases which I summarise
here. The list has no pretension of being exhaustive.1
In Makwanyane (1995),2 decided under the Interim Constitution,

the Court considered a constitutional challenge against the death
* University of Buenos Aires.

1
For a comprehensive analysis, see A Price ‘The content and justification of
rationality review’ in S Woolman & D Bilchitz (eds) Is this seat taken?
Conversations at the Bar, the Bench and the Academy about the South African
Constitution (2012) 36.
2 S v Makwanyane 1995 3 SA 391 (CC) para 156.
31
32 Rationality, reasonablenss, proportionality
penalty, without any explicit mention in the interim Constitution

about it being forbidden. Justice Ackerman, in a concurring opinion,
introduced the rationality standard as follows: 3
In reaction to our past, the concept and values of the constitutional

state, of the “regstaat”, and the constitutional right to equality before
the law are deeply foundational to the creation of the “new order”
referred to in the preamble. The detailed enumeration and description
in section 33(1) of the criteria which must be met before the legislature
can limit a right entrenched in Chapter 3 of the Constitution emphasises
the importance, in our new constitutional state, of reason and
justification when rights are sought to be curtailed. We have moved
from a past characterised by much which was arbitrary and unequal in
the operation of the law to a present and a future in a constitutional
state where state action must be such that it is capable of being
analysed and justified rationally. The idea of the constitutional state
presupposes a system whose operation can be rationally tested against
or in terms of the law. Arbitrariness, by its very nature, is dissonant with
these core concepts of our new constitutional order. Neither arbitrary
action nor laws or rules which are inherently arbitrary or must lead to
arbitrary application can, in any real sense, be tested against the
precepts or principles of the Constitution. Arbitrariness must also
inevitably, by its very nature, lead to the unequal treatment of persons.
Arbitrary action, or decision making, is incapable of providing a rational
explanation as to why similarly placed persons are treated in a
substantially different way. Without such a rational justifying
mechanism, unequal treatment must follow.
Segments of this quote have been referred to by the Court in later

cases. However, the scope of the rationality standard proposed by
Justice Ackerman in this case was considerably more substantive than
the standard retained by the Court in its later jurisprudence. In fact,
it is unlikely that a very thin conception of rationality, as the mere
connection of a legitimate State purpose with the means chosen,
could have been of any use in a case where the challenge to the death
penalty was fundamentally based on variations of the substantive due
process notion.
In New National Party of South Africa (1999),4 the Court dealt

with a challenge to electoral rules that imposed the requirement of
specific identification documents to exercise the right to vote,
striking out some other valid IDs from the list. The majority of the
Court, following the opinion of Justice Yacoob, invoked a narrow
notion of ‘rationality’, as mere connection between a legitimate
State purpose and the means chosen, and found the legislative
scheme to be rational and thus constitutional. In a dissenting opinion,
3 As above.
4
New National Party of South Africa v Government of the RSA 1999 3 SA 191 (CC),
particularly paras 19 & 24.
Justice O’Regan criticised the approach of the majority and called for
a more contextual analysis that took into consideration the potential
discouraging effect of a restrictive regulation on the exercise of the
right to vote, particularly in that nascent stage of the South African
democracy. Justice O’Regan proposed a more substantive standard of
review, a ‘reasonableness’ standard, which in her view would have
enabled the Court to discuss the appropriateness of the regulation in
promoting the right to vote.
In United Democratic Movement (2000)5 the Constitutional Court

reviewed the constitutionality of a statute allowing members of
Parliament to change their political party under circumstances
specified by the law. The Court again made use of a thin rationality
test, finding that the statute passed constitutional muster as there
was a connection between a legitimate State purpose and the means
chosen to achieve that purpose.
In Pharmaceutical Manufacturers Association (2000)6 the

Constitutional Court found that State action — in the case, a
presidential decision to bring an Act of Parliament into force — did not
pass the thin version of the rationality test. In that instance, the
Parliament had adopted a new legislative scheme modifying the
Medicines and Related Substances Act. The new scheme made
obsolete the existing schedules that were referred to in the previous
scheme, and required, in order to be operative, the adoption of new
regulations and schedules. However, the President, acting upon the
powers granted by the legislative act, decided to bring it into force
even before the necessary regulations and schedules were adopted.
During the Constitutional Court proceedings, the Government itself
acknowledged the technical mistake. The Court refined the definition
of the rationality standard, adding that the determination of whether
the decision was rationally related to the purpose for which the power
was given called for an objective — rather than a subjective —
enquiry: that is, no matter the good faith or mistaken belief of the
public authority who took the decision. The Court considered that the
presidential decision could not be found ‘objectively rational on any
basis whatsoever’, that it was irrational and thus struck it down.
In Bel Porto School Governing Body (2002),7 the Court had to face
a constitutional challenge against a legislative scheme that
confronted the difficult task of achieving racial desegregation in
schools in a context of limitation of resources. The scheme privileged
5 United Democratic Movement v President of the RSA (1) 2000 11 BCLR 1179 (CC),
particularly paras 55 - 76.
6 Pharmaceutical Manufacturers Association of SA; In Re: Ex Parte Application of
President of the RSA 2000 2 SA 674 (CC), particularly paras 85 & 90.
7
Bel Porto School Governing Body v Premier of the Province, Western Cape 2002 9
BCLR 891 (CC), particularly para 45.
the situation of permanent staff over that of personnel who were

contracted by former only-white schools on a discretionary basis and
were thus not considered public employees, despite long working
periods. The majority of the Court was satisfied about the existence
of a link between legitimate State purposes and the means chosen,
and hence considered the legislative scheme constitutional. A joint
partial dissent by Justices Sachs and Mokgoro, and separate partial
dissents by Justice Ngcobo and Justice Madala did not agree on the
solution regarding the status of the contracted personnel. However,
their respective arguments were not based on the rationality of the
scheme, but rather on other substantive and procedural standards and
principles — fairness, administrative justice or lack of consultation.
In Affordable Medicines Trust (2005)8 the Court considered the

rationality and proportionality of a licensing scheme introduced by
the Government requiring a permit to dispense medication in
specifically licensed premises. The Court upheld the licensing
scheme, concluding that it was rationally related to the achievement
of the legitimate government purpose to increase access to medicines
that are safe for consumption by the public. In relation to the
challenges based on alleged breaches of the rights to choose and
practise a profession, dignity, freedom of movement and property,
the Court found that the licensing scheme did not limit those rights,
and thus did not even enter into the analysis of the justifiability of the
limitations.
Furthermore, the Court developed criteria to assess the

reasonableness of the measures taken by the State to achieve the
realisation of the rights to housing and of the right to health,
respectively,9 in two noted leading cases, Grootboom (2000)10 and
Treatment Action Campaign (2002).11 Similarly, the Court has also
approached the reasonableness of administrative action.12 Standards
that involve rationality and proportionality-related assessment have
also been developed by the Court in the context of arbitrary
deprivation of property.13
8
Affordable Medicines Trust v Minister of Health of RSA 2006 3 SA 247 (CC),
particularly paras 74 - 79.
9 For an enlightening discussion of this issue, see D Brand ‘Socio-economic rights
and courts in South Africa: Justiciability on a sliding scale’ in F Coomans (ed) The
justiciability of economic, social and cultural rights (2006) 207; S Liebenberg,
‘Socio-economic rights: Revisiting the reasonableness review/minimum core
debate’ in S Woolman & M Bishop (eds) Constitutional conversations (2011) 303.
10
Government of the Republic of South Africa v Grootboom 2001 1 SA 46 (CC).
11 Minister of Health v Treatment Action Campaign (No 1) 2002 5 SA 703 (CC).
12
See Bato Star Fishing (Pty) Ltd v Minister of Environmental Affairs and Tourism
2004 4 SA 490 (CC).
13 For a comprehensive discussion of this issue, see T Roux ‘The ‘arbitrary
deprivation’ vortex: Constitutional property law after FNB’ in S Woolman & M
Bishop (eds) (n 9 above) 265.
In summary, the approach adopted by the Constitutional Court has

been the following:14
- The exercise of public power — including lawmaking15 — should be

rational and not arbitrary. ‘Rationality’ is understood as a rational
relationship between the measures adopted and a legitimate
governmental purpose. ‘Rationality’ calls for an objective enquiry.
- When the challenge to general norms is based on the alleged
infringement of any of the rights included in the Bill of Rights, the
standard of review to be applied is in principle that of ‘proportionality’,
according to section 36(1) of the South African Constitution. The
‘proportionality’ test requires the Court to assess if limitations to rights
are reasonable and justifiable in an open and democratic society based
on human dignity, equality and freedom, taking into account all relevant
factors, including: the nature of the right, the importance of the
purpose of the limitation, the nature and extent of the limitation, the
relation between the limitation and its purpose; and the existence of
less restrictive means to achieve the legitimate purpose.
- ‘Reasonableness’ is a standard specific to some constitutional rights —
such as the right to health, or the right to housing — but it is also a
criterion to be taken into account when assessing both the
proportionality of the limitation of a constitutional right (as per section
36(1) of the Constitution), or the appropriateness of the action adopted
by the political branches of the Government to comply with
constitutional duties or achieve legitimate state purposes.16
In my view, two general conclusions can be drawn from the analysis

of the previous Constitutional Court case law:
- The apparent adoption of an agreed set of criteria regarding the choice

of the standard of review and the test entailed by each standard can be
14
This is, of course, an over-simplified summary. For a comprehensive overview, see
the discussion between M Bishop ‘Rationality is dead! Long live rationality! Saving
rational basis review’ and A Price ‘The content and justification of rationality
review’ both in S Woolman & D Bilchitz (n 1 above) 1 and 36, respectively. The
extent to which there is an overlap between ‘reasonableness’ and
‘proportionality’, and between ‘rationality’ and ‘proportionality’ deserves
further exploration, but the issue exceeds the purpose of this section, as it is not
particularly relevant for the cases that I will examine below.
15
Among other areas, rationality analysis has been employed by the Constitutional
Court to assess the acceptability of differentiations made by the legislative
branch under equality challenges (section 9(1) of the SA Constitution), when such
differentiations are not based on prohibited discrimination grounds mentioned in
Section 9(2). Bishop’s article focuses on this use. See Bishop (n 14 above). I did
not include such cases in the previous overview, as none of the judgments
discussed later deal with challenges based primarily on section 9(1) of the
Constitution. Interestingly, in one of the cases (Law Society of South Africa) the
need to remove arbitrary differentiations was mentioned by the Government as a
purpose to justify the challenged provisions of the statute.
16 For example, whether the State has complied with its obligations to facilitate
participation in the legislative process; whether the State has complied with sec
33 or whether the State has complied with its positive obligations under sec 7(2)
to ‘protect, promote and fulfil’ constitutional rights.
misleading. The analysis of case law shows that the adoption of a

seemingly accepted approach by the Court was not undisputed. A closer
look at the case law shows that in different cases, the scope of the
different formulas was not the same, and even when the tests referred
to are the same, the intensity of the scrutiny varies considerably.
Moreover, while the Court has in some cases tried to restrict the use of
the ‘reasonableness’ standard to only those cases involving
constitutional rights which explicitly include this standard in their
wording, ‘reasonableness’ keeps coming back as a criterion in different
ways.
- The use of the ‘rationality’ test has not been a particularly fruitful as a
standard to review general norms. The Court concluded only in a small
number of cases17 that the challenged norms or acts did not pass the
‘rationality’ test, and in fact in three of those cases the challenge was
not directed against a law, but rather against the Executive’s behaviour.
2 Poverty Alleviation: reasonableness of

measures to ensure participation, rationality
of a Constitutional amendment and
subsequent legislation
In Poverty Alleviation,18 the applicants challenged the constitutional

validity of a constitutional amendment and legislation adopted
thereafter with the aim of changing the boundaries between two
provinces and thus transferring municipalities from one province to
the other. According to the claim, the authorities involved (National
Assembly, National Council of Provinces, provincial legislature of
KwaZulu-Natal) failed to facilitate public involvement in the decision-
making process, and the constitutional amendment and legislation
were allegedly irrational. Interestingly, none of the claims were based
on the alleged violation of rights included in the Bill of Rights, but
rather on rules and duties governing the adoption of legislation or the
amendment of the Constitution (sections 74, 74(8), 59(1)(a), 72(1)(a)
and 118(1)(a) of the Constitution). The Court considered the first
challenge under a reasonableness scrutiny, and the second under a
rationality scrutiny. In both cases, the Court upheld the constitutional
validity of the amendments and legislation.
Regarding the claim on failure to comply with the duty to

adequately facilitate public involvement in the different instances of
the decision-making process, provided for in similar wording by
17 As of 2011, only five, according to Price’s account, while it has dismissed far
more. See Price (n 14 above) 64.
18
Poverty Alleviation Network v President of the Republic of South Africa 2010 6
BCLR 520 (CC).
sections 59(1), 72(1) and 118(1)(a), the Constitutional Court, relying

on precedent,19 decided to consider the issue under a reasonableness
enquiry — even if reasonableness is not mentioned as a standard in
any of those provisions. The Court stated that legislative bodies have
‘considerable discretion to determine how to fulfill’20 this duty, and
took into account in order to assess reasonableness the need to strike
a balance between ‘the need to respect parliamentary autonomy’ and
the ‘right of the public to participate in the legislative process’.21
Factors to be weighed to assess the reasonableness of the measures
adopted by the legislative bodies to facilitate public participation are
the nature, importance and urgency of the legislation, and the time
and expense that public involvement may require.
On these bases, the Court examined the legislative history of the

process of adoption of the Constitutional amendment and the ensuing
legislation, and concluded that the involved authorities had indeed
facilitated participation. The Court went on to examine three related
issues: whether facilitation of participation requires only the
consultation of the discrete group directly affected by the legislation,
with the exclusion of others; whether facilitation of participation
required the legislative bodies to receive oral submissions; and
whether the legislative bodies involved had considered the
representations made by the plaintiffs. Given the deference granted
by the Court to the legislative bodies in the determination of the
methods for facilitating public involvement, it is not surprising that
the answers to these three issues were negative. The Court concluded
that there is no requirement to consult solely with the discrete group
directly affected by the legislation — the legislature should only
afford them a reasonable opportunity to participate meaningfully.
They also rejected the claim that oral submissions were required,
leaving it to the legislature to decide whether oral submissions could
provide more clarity to the already received written representations.
Finally, the Court maintained that public involvement does not mean
that the inputs offered should necessarily have an impact on the
outcome legislation, but only that the opportunity to be heard is
granted.
Regarding the alleged irrationality of the constitutional

amendment and ensuing legislation, the plaintiffs made a number of
different claims: that the outcome legislation ignored the inputs
received from the public, that it was based on factual errors, that the
decision was pre-determined and that the lawmakers were instructed
19
Doctors for Life International v Speaker of the National Assembly 2006 6 SA 416
(CC) para 99.
20 Poverty Alleviation (n 18 above) para 35, quoting Doctors for Life (n 19 above)
para 99.
21 Poverty Alleviation (n 18 above) para 35.
to vote in a particular manner. The Court resorted to their standard

formula on rationality — rational connection of the measure with a
legitimate governmental purpose — and dismissed the claims made by
the applicants without much hesitation. They had no doubt that the
constitutional amendment and ensuing legislation were connected to
a legitimate governmental purpose, and summarily dismissed the
claims. The Court stated that the applicants’ argument muddled
procedure with substance, and pointed out that rationality concerns
itself with outcome, rather than with compliance with procedural
requirements. They also rejected the claim that the legislative bodies
erroneously considered the nature of the municipality at stake,
granting deference to the views adopted by the legislative bodies and
repeating that a court cannot interfere with legislation simply
because it disagrees with its purpose or believes that it should be
achieved in a different way. A court could only interfere ‘if it can be
shown that the objective is arbitrary, capricious or manifests naked
preferences’.22 In turn, the court considered that it cannot
investigate the motives of the legislative bodies, but should rather
stick to the examination of the rationality of the legislation itself.
A number of points can be made on the Court’s decision in Poverty

Alleviation. Regarding the choice of the standard of review, it is
interesting to note that while the Constitutional Court usually rejects
‘reasonableness’ as a general standard to evaluate challenges against
legislation not strictly based on the infringement of constitutional
rights, it spontaneously resorts to ‘reasonableness’ as a standard to
evaluate compliance with legislative procedural duties — such as the
duty to facilitate public involvement in the lawmaking process — even
if there is no mention of such a standard in the respective
constitutional provisions that were the basis of the claim.23
However, the scope of the ‘reasonableness’ test used in this case

proves to be a thin standard to assess the legislative conduct, as the
criteria offered to evaluate what is reasonable assumes from the
inception a broad deference to the legislature. It is fair to say that the
Court seems to devote more attention to considering — in a favourable
light — the evidence offered by the legislative bodies to show that it
complied with facilitating public involvement in the lawmaking
process, than to dig in any form into the arguments justifying the
legitimacy of the governmental purpose and responding to the other
claims related to ‘rationality’. But the consequences of choosing
22 Poverty Alleviation (n 18 above) para 71.

23
The argument can be made that the Court had done so in previous cases (eg
Doctors for Life and Matatiele) and that, thus, in Poverty Alleviation it is just
following its precedents. But this doesn’t affect the point made here — it might
only transfer the same question to the justification of the decision made in those
cases.
either ‘reasonableness’ or ‘rationality’ do not seem to make an

enormous difference in the case. Both the intensity of the use of the
‘reasonableness’ test for assessing the adequacy of the facilitation of
public involvement in lawmaking, and of the use of the ‘rationality’
test for assessing the outcome of the legislation appear to be mild.
The Court is in both cases far from probing the legislative bodies: it
appears to assume that the legislative record is prima facie valid,
unless it sees strong evidence to the contrary — which was not the
case in Poverty Alleviation.
3 Law Society of South Africa: the cold and the

hot in the same dish
In Law Society of South Africa,24 the applicants challenged the

constitutional validity of a legislative amendment of the Road
Accident Fund Act, a law regulating a public fund to provide
compensation to victims of car accidents. Specifically, allegations of
constitutional violations were made about the elimination in the new
statutory scheme of a residual common law claim allowing accident
victims to recover losses not compensable under the act from a
wrongdoer which existed in the previous statutory scheme, and to the
establishment of inadequate tariff levels to cover medical expenses
incurred as a consequence of car accidents. The legal claims were
two-fold. On the one hand, the applicants alleged that the new
statutory scheme was irrational, but requested the Court to move
away from a mere rationality test and to endorse a more substantive
standard — the assessment of whether the new legislation unfairly
deprives people of their constitutional protection. On the other
hand, the applicants alleged that the new legislation unjustifiably
limited several constitutional rights — the right to security of the
person, the right to property, the right to health and the right to an
adequate remedy.
In respect to the claim that the new statutory scheme was

irrational, the Court rejected the request to engage in a more
substantive test, and maintained its previous trend of adopting a
narrow rationality test — that is, rational connection between means
chosen and a legitimate governmental purpose, without considering
whether legislation is fair or reasonable or appropriate, or whether
there are other or even better means that could have been used.
According to the Court, the adequate space for a more substantive
test — such as the proportionality test — was the consideration of the
justifiability of alleged breaches to a fundamental right provided for
in the Bill of Rights under section 36 of the Constitution.
24 Law Society of South Africa v Minister for Transport 2011 1 SA 400 (CC).
However, even adopting a narrow ‘rationality’ test, when

confronted with the analysis of the rationality of the elimination of
the residual civil law claim, the Court had to struggle to find a
justification for the legislative amendments. According to the
government, the purposes of the new statutory scheme were to tackle
the increasing deficit of the fund, to remove arbitrary differentiations
(passengers on the one hand, and drivers and pedestrians on the
other) in the compensation of accident victims, and to integrate the
compensation of road accident victims into a comprehensive social
security scheme offering life, disability and health insurance cover for
all accidents and diseases. The government argued that this
legislative scheme is a gradual step towards the replacement of a
common law system of compensation with a set of limited no-fault
benefits of a broader social security net.
The applicants pointed out that the existence of a residual

common law claim does not influence the financial viability of the
scheme, as common law claims are not directed against the fund but
against the wrongdoer. Thus, its abolition would not further the end
of meeting the needs of every victim by making the scheme fully
funded. So, according to the applicants, there is no proper relation
between the means chosen and the alleged governmental purpose.
While the Court accepted that the elimination of the residual

claim does not worsen or improve the financial standing of the fund,
it went on to argue that the scheme must be seen as a whole and not
only in the light of the common law claim. According to the Court, as
the statutory scheme puts caps on the compensation to be paid by the
Fund, with the continued existence of a residual claim the liability of
wrongdoers would increase in proportion to the level of caps imposed,
thus making liable motorists bear the risk of substantially increased
residual claims from accident victims. The Court deems that the risk
to which the new cap exposes all drivers is disproportionate in
relation to the ‘relatively small inattentiveness or oversight that
could give rise to the risk’.25 The Court makes an additional
argument: that the retention of a common law claim does not sit well
with a social security compensation system aiming to provide
equitable compensation to all people regardless of their financial
ability. To prove this, the Court asserts: a) that the common law claim
would be actually recovered only from wrongdoers capable of paying
the compensation or of affording insurance, and b) that the right to
sue would be available only to those who can afford to pay legal fees
or who are granted legal aid.
25 Law Society of South Africa (n 24 above) para 50.

What is remarkable in this justification is that these reasons were

provided by the Court itself, and not by the Government. The Court
hardly weighed impartially the arguments offered by the Government
and the applicants as to whether there was a rational connection
between the elimination of the residual common law claim and the
purposes sought by the legislative scheme. Instead, it in fact
furnished a justification that indeed focuses on the protection of the
potential wrongdoers against the exposure to increased liabilities —
an argument the Government had not articulated.
The Court also gave some weight to the Government’s contention

that the new statute needs to be seen as a transitional scheme,
towards a no-fault compensation regime. The Court considered the
abolition of the common law claim as a ‘necessary and rational part’26
of an interim scheme aimed at achieving financial viability and more
effective and equitable access to social security services. But little
explanation was offered to support this conclusion: no reason is given
as to why it would be incompatible to have a capped no-fault
compensation system paid by the Fund, and to maintain a common
law claim against the wrongdoer, either fault or non-fault, requiring
of the prospective plaintiff a higher procedural burden and subject to
the risk of not recovering, but offering the possibility of a higher
compensation — which would in turn be more appropriate to satisfy
the health or disability-related needs ensuing from the accident.
In conclusion: not only did the Court choose the very narrow
rationality test to consider the constitutional challenge of the
legislative scheme, but the intensity of the application of the test was
indeed notoriously weak — to the point that the main arguments used
to assert that there was a rational connection between the impugned
provisions and the legitimate governmental purpose were in fact
provided by the same Court.
After dismissing the challenge on the basis of irrationality, the

Court moved on to consider the alleged breaches of several
fundamental rights. While the Court had previously stated that the
standard of scrutiny would be stricter when the challenge at stake
was the claim that a fundamental right was breached, it is not clear
that the Court completely lived up to this statement. Three of the
claims were dismissed, without digging much into the reasons offered
by the government to justify the impugned provisions, and apparently
without exhausting the steps of the analysis that — at least in theory
— should be prompted by the proportionality test — or by any other
scrutiny deemed to be more substantive.

According to the first challenge, the elimination of the residual

common law right to claim breached the duty to adopt positive
measures to ensure the right to security of the persons (section
12(1)(c) of the South African constitution, which enshrines a right ‘to
be free from all forms of violence from either public or private
sources’) and/or the right bodily and psychological integrity (section
12(2) of the South African constitution). The Court sticks to the
analysis of the claim based on 12(1)(c), engaging in the evaluation of
the challenge under a proportionality test. As to the first prong of the
test — whether the fundamental right at stake is compromised by the
impugned piece of legislation — the Court accepts that the protection
granted to road accident victims falls within the thrust of the State’s
positive obligations to respect, protect and promote the right to be
free from violence either from public or private sources and, hence,
that the State incurs section 12 obligations in relation to victims of
road accidents.
The next issue considered by the Court was whether the abolition
of the common law claim unjustifiably limited section 12(1)(c). In my
view, the question was wrongly presented by the Court. The language
of the justifiability of the limitation of rights, as stated in section 36
of the South African constitution, is mainly addressed to evaluate
restrictions upon or regulations of the enjoyment or exercise of
freedoms.27 Here, the issue at stake was instead the appropriateness
of the positive measures adopted by the State in order to protect the
right to be free from violence — an analysis which resonates more
closely with the issue of the reasonableness of the positive measures
adopted to realise socio-economic rights than with the framework of
the justifiability of the limitation of freedoms or negative rights.28 To
put it slightly differently: what the Court is assessing here is whether
the positive measures adopted by the State to protect the right to be
free from violence are adequate, and not if that right was
unjustifiably restricted. The analysis of a ‘restriction’ presupposes
that the content of the positive measures constitutionally due by the
State is already known — but this is exactly what is at stake here.
Moreover, the claim made by the applicants seemed to require an

inter-temporal evaluation from the Court: whether the new
legislation had unduly reduced the level of protection previously
granted by the existing legislative scheme — which allowed for a
residual common law right to sue. This idea can be presented both in
a ‘negative’ or ‘positive’ wording. In a ‘negative’ formulation — which
27
For a comprehensive analysis of Section 36, see S Woolman & H Botha
‘Limitations’ in S Woolman et al (eds) Constitutional law of South Africa (2nd
Edition, OS, 2006) Chapter 34. See also S Woolman & H Botha ‘Limitations: Shared
constitutional interpretation, an appropriate normative framework and hard
choices’ in S Woolman & M Bishop (eds) (n 9 above) 149.
might superficially resemble the wording of the limitation of rights —

the question would be whether the new legislative scheme unduly
limited or restricted the existing level of protection of the right. This
implies that restrictions of the already conceded level of legislative
protection are unjustifiable: that is, that once the legislator
recognised that a common law right to sue is part of the protection
for victims of road accidents, removing that right is forbidden. In
other contexts, this idea has been captured by the notions of
‘prohibition of retrogression’, ‘irreversibility’ or ‘standstill or ratchet
effect’.29 In a ‘positive’ formulation — which might rather resemble
the context of evaluation of the reasonableness of positive measures
— the question might rather be posed as whether reducing the level
of protection already conceded does meet the positive obligation to
protect the right. The Court seems to avoid any need to embark itself
in inter-temporal comparisons, and thus considers both the old and
the new schemes not as a chronological sequence, but rather as two
alternative options in a tabula rasa scenario. This approach fails to
capture the main point made by the plaintiffs, which was not that
article 12(1)(c) of the South African constitution necessarily requires
a common law action to protect the right, but that victims of road
accidents will, in the future, see reduced the options for
compensation that they had already enjoyed before the legislative
change.
By dealing with the abolition of the common law right to sue as a

mere limitation of rights, the Court finds that of course it diminishes
the victims’ capacity to protect and to enforce the right to the
28 By this I do not mean that rights other than socio-economic rights only impose
negative obligations on the State. They also impose positive obligations, even if
they do not usually carry the same qualifications (i.e. reasonableness, subjection
to available resources) as it is the case with socio-economic rights. The
Constitutional Court has discussed the scope of these positive obligations of rights
other than socio-economic rights in a few cases. See, for example, New National
Party of South Africa (n 4 above) paras 13 - 17, and O’Regan J’s dissenting opinion
at paras 118 - 119 (obligation to enact legislation and take positive measures to
ensure the enjoyment of the right to vote); President of the Republic of South
Africa and Another v Modderklip Boerdery (Pty) Ltd 2005 5 SA 3 (CC) paras 39 - 51
(positive obligations stemming from the right to access to courts). In any event,
my contention here applies to the standards to evaluate positive obligations
stemming from any constitutional right, regardless of whether it is or not
classified as socio-economic. Even if the term used is the same (ie
‘reasonableness’), its meaning in the context of limitations of rights as stated in
section 36 cannot be the same as when used as a standard to assess compliance
with positive obligations. In the former case, the issue at stake whether a
restriction to a freedom can be justified. In the latter, the issue at stake is
whether the measures adopted to ensure the enjoyment of the right are adequate
or appropriate, given certain circumstances.
29 See, for example, UN Committee on Economic, Social and Cultural Rights,
General Comment No 3 ‘The nature of States parties obligations (Art. 2, par.1)’
12/14/1990 para 9: ‘any deliberately retrogressive measures in that regard would
require the most careful consideration and would need to be fully justified by
reference to the totality of the rights provided for in the Covenant and in the
context of the full use of the maximum available resources’.
security of the person, but finds little trouble in considering that the
same justification given to uphold the rationality of the new
legislative scheme also justifies the limitation. The following
paragraph is graphic enough:30
As I understand it, the real complaint of the applicants is that they have
lost the common law right of recourse and have to contend with a new
ceiling on their claim for general damages and to loss of income or
support. They add that the common law right of recourse against
wrongdoers comes at no cost to the Fund. It is the wrongdoer and not
the fund that would be liable for the residual common law damage. The
Minister and the Fund have advanced adequate justification for this
limitation. They rehearse no fewer that eleven substantive grounds of
justification, most of which are cogent. The over-arching grounds are
the urgent need to make the Fund financially viable and sustainable, and
to make its compensation regime more inclusive, transparent,
predictable and equitable.
Interestingly, the arguments used by the Court to justify the

limitation of the fundamental right under the proportionality test are
exactly the same used before to assess the rationality of the scheme.
While some overlap might be explained by the inclusion of the means-
ends analysis as one of the criteria to assess proportionality, the
question left by the Court’s approach is whether it makes any
difference to invoke a rationality or a proportionality standards —
claiming indeed that the latter is a more substantive standard — when
the exact same reasons said to meet one test are repeated to consider
the second test passed. There is no trace in the proportionality
analysis engaged by the Court of the use of other criteria that the
proportionality test should have prompted according to section 36 of
the Constitution and to the stock formula usually used by the Courts:
the nature of the right; the importance of the purpose of the
limitation; the nature and extent of the limitation; the relation
between the limitation and its purpose; and the existence of less
restrictive means to achieve the purpose. Moreover, according to
section 36, these are only illustrative examples of ‘all relevant
factors’ that should be taken into account. The individual
consideration of each factor is poor, to say the least: the only factor
considered is the relation between the limitation and its purpose, and
this is done in the same narrow fashion as in the rationality test — that
is, mere connection of the means with the purpose, rather than
fulfilling the promise of a more substantive analysis. Little is said
about the weight of the other factors — nature of the right,
importance of the purpose of the limitation, nature and extent of the
limitation, existence of less restrictive means to achieve the purpose
— and there is no effort to identify the existence of other relevant

factors that should be taken into account. So instead of making use of

the flexible formula of section 36 to perform a context-based
evaluation of the reasons offered by both parties, the Court employs
the proportionality test in a rather formalistic manner, which makes
it difficult to distinguish any consequence of its use in comparison
with the use of a narrow rationality test.
Two additional constitutional challenges — regarding the rights to

property (section 25(1) of the Constitution) and the right to a remedy
(section 38 of the Constitution) were also summarily dismissed by the
Court. Regarding the right to property the claim is dismissed as the
Court finds no ‘arbitrary deprivation of property’31 — so not even the
first prong of the proportionality test bears fruit here.32 In relation to
the right to a remedy, the Court points out that is has already found
that the limitation placed by the new statutory scheme on this right
is reasonable and justifiable in all circumstances.
The Court’s approach to the violation of the right to health was

considerably different: the standard of scrutiny was notoriously
stricter, and the deference shown to the reasons presented by the
Government was very narrow. The plaintiffs’ contention was that the
tariff level determined by the legislative scheme for claims to be paid
by the Fund for hospital and other medical treatment was so low that
road accidents victims will not be able to obtain treatment from
private health care institutions. According to the applicants, the
regulation was irrational, deprived the innocent victim of an effective
remedy, and, in relation to the right to health recognised in section
27 of the Constitution, was retrogressive — as opposed to progressive
— and unreasonable.
Upon consideration of the expert evidence, the Court finds that

the tariff fixed by the regulation is ‘wholly inadequate and unsuited
for paying compensation for medical treatment of road accident
victims in the private health sector’,33 thus meaning that victims that
cannot afford private medical treatment by other means will have to
submit to treatment in public health establishments. It is also
established that public health institutions are not able to provide
adequate services crucial to the rehabilitation of accident victims
with permanent disabilities, including quadriplegic and paraplegic
victims. The Court also considers that the savings that such a tariff
would generate amount to 6% of the total compensation bill at the
most. According to the Court, depriving quadriplegic and paraplegic

32
Despite its easy rejection in this case, deprivation of property is another fruitful
context to show the Court’s varying approach of rationality. Theunis Roux has
shown that the ‘arbitrariness’ test has in some cases been used much more
substantively than in others. See Roux (n 13 above) 265-281.
victims from adequate access to medical care to achieve such

negligible financial saving is unreasonable.34
Besides the finding, what is interesting here is the standard used

by the Court to assess the constitutionality of the scheme. The Court
asserts that the tariff fixed by the regulation
is incapable of achieving the purpose which the Minister was supposed to

achieved, namely a tariff which would enable innocent victims of road
accidents to obtain the treatment they require (…) It must follow that
the means selected are not rationally related to the objectives sought to
be achieved. That objective is to provide reasonable healthcare o
seriously injured victims of motor accidents.35
Additionally, the Court finds that the tariff is under-inclusive in

relation to the healthcare needs of quadriplegic and paraplegic road
accident victims, and hence ‘unreasonable and thus in breach of
section 27(1)(a) read together with section 27(2) of the
Constitution’.36
It seems that, inadvertently, the Court is using here a very

demanding standard to assess the progressive realisation of the right
to health — a much more substantive standard than the one used in
leading socio-economic rights cases such as Grootboom, Treatment
Action Campaign or Mazibuko.37 Instead of relying mostly on
procedural requirements, as it did in those cases,38 the standard used
here to assess the reasonableness of the measures adopted to
progressively realise the right to health is squarely the adequacy of
the measures to allow victims to access the treatment they require.
Moreover, it seems to convey the idea that the right to health includes
an opportunity to access treatment in private medical health care
facilities — an idea that is hardly reflected in section 27 of the
34
This reasoning can perhaps be compared to Khosa, where the Constitutional Court
stroke down legislation denying social assistance to non-citizens as discriminatory
and unreasonable. The Court rejected similar ‘savings’ arguments. See Khosa &
Others v Minister of Social Development & Others 2004 6 SA 505 (CC) paras 60 -
62. But, as it was mentioned before, the Court considered, among others,
‘savings’ arguments as a proper justification for the elimination of the residual
civil law claim.
35
Law Society of South Africa (n 24 above) para 99.
36 Law Society of South Africa (n 24 above) para 100. Section 27 of the South African
Constitution, in the relevant parts, reads as follows:
‘27.1) Everyone has the right to have access to health care services,
including reproductive health care; (…)
27.2) The state must take reasonable legislative and other measures,
within its available resources, to achieve the progressive realisation of
each of these rights’.
37
Mazibuko v City of Johannesburg 2010 4 SA 1 (CC).
38 For a sharp overview and critique of this approach see D Brand ‘The
proceduralisation of South African socio-economic rights jurisprudence, or: “What
are socio-economic rights for?” ’ in H Botha, A van der Walt & J van der Walt (eds)
Rights and democracy in a transformative constitution (2004) 33.
Constitution. One could wonder how the Court would react if

prompted to employ the same standard to assess the quality of the
general public health services offered by the Government to fulfil the
right to health.
A second remark refers to the use of the means-end analysis

made. Presenting the analysis as a means-end consideration seems to
eschew, rather than to reveal, the same nature of the reasoning of the
Court. Rather than a formal comparison of means and end, the Court
is making here a substantive assessment: it is testing the adequacy or
appropriateness of the level of tariffs fixed by the Government
against the substantively defined content of the required treatment.
Using the same name for this analysis and, say, the rationality analysis
made before, or the one made in Poverty Alleviation, seems
completely misleading.39
In summary: the same case shows two opposed approaches to

choose and apply the standard of scrutiny for deciding a constitutional
challenge of a legislative statute. The elimination of the residual
common law claim was dealt with through a very deferential
approach: the Court provided itself reasons to justify the connection
between the means chosen and the intended purpose, and hardly
bothered to probe further criteria to consider the alleged breached of
the rights to security of the person, property and adequate remedies.
In that account, the difference between the Court’s use of a
‘rationality’ and a ‘proportionality’ scrutiny seems to be negligible.
On the other hand, things change a lot when it comes to the level of
tariff fixed by the regulations to pay for health and other medical
treatment. Here — again, regardless of the name of given by the Court
to the type of scrutiny — the standard of review used is remarkably
more substantive and less deferential to the Government. My
suspicion is that the justice who wrote the decision first reached a
conclusion about the merits of each challenge, and then
accommodated the standard of scrutiny to his previously adopted
conclusion, rather than employing methodically the stock formulas
39 In an early review, Danie Brand pointed out the different levels of scrutiny
employed under the same rubric (reasonableness) in the Court’s socio-ecomomic
rights cases. See D Brand (n 9 above) 207-236. The point was particularly renewed
after the Court’s decision in Mazibuko, where reasonableness is used as mere
formal due process review. See, in this respect, R E Kapindu, ‘Reclaiming the
frontier of constitutional deference: Mazibuko v City of Johannesburg – a
jurisprudential setback’, in S Woolman & D Bilchitz (eds) (n 1 above) 319; L
Williams, ‘The role of courts in the quantitative-implementation of social and
economic rights: a comparative study’ (2010) 3 Constitutional Court Review 141.
and prongs for each type of scrutiny to analyse the various

constitutional challenges.40
4 Malachi: rationality postponed in favour of

proportionality
In Malachi,41 the Constitutional Court had the opportunity to consider

the constitutional validity of the legislative provisions of the
Magistrate’s Court Act which granted courts the power to order, at
the request of the creditor, the arrest of a debtor when it appears
that he/she intends to leave the country — the so-called tanquam
suspectus de fuga.
The applicant challenged the legislative provisions as contrary to

the right to security and freedom of the person in terms of section
12(1) of the South African constitution, particularly paragraph (a),
which refers to the right ‘not to be deprived of freedom arbitrarily or
without just cause’. The Court employed a proportionality scrutiny to
consider the constitutional challenge, following the prongs of the
stock formula: (a) does the arrest based on the impugned provision
limit the applicant’s right to freedom of the person arbitrarily or
without a cause?; (b) if the right is limited, is the limitation
justifiable?
Regarding the first prong of the test, the Court pointed out that
by definition, an arrest or detention limits the freedom of a person —
so that the relevant question is whether the limitation is without a
just cause. According to the reasoning of the Court in the case, for a
cause to be just, it requires a strict relation of necessity of the
measure with a legitimate purpose sought. The object of the arrest is
to ensure that the potential debtor remains within the jurisdiction of
the court until the court has given judgment in the matter. The Court
suggested that the arrest does not guarantee the satisfaction of the
judgment debt — the debtor can leave the country as soon as the
judgment is given, and the arrest does not render the judgment any
more executable to the creditor than would have been the case had
the debtor left the country. Moreover, the order for arrest is made in
a moment when the civil liability has not yet been established. As
prison for debt is forbidden, a fortiori there is no legal justification
for depriving of the personal freedom of a person that has not even
been proven to be civilly liable, and might never be. The Court found
40
Lucy Williams makes a similar point about the Constitutional Court’s use of
‘separation of powers’ arguments to avoid engaging with the evidence offered in
the case and thus weakening the potential substantive bite of the reasonableness
standards in Mazibuko. See Williams (n 37 above).
41 Malachi v Cape Dance Academy International (Pty) Ltd 2011 3 BCLR 276 (CC).
then that there is no just cause for the arrest in terms of the
impugned provisions.
Sticking perhaps too literarily to the stock formula, the Court

went on redundantly to analyse whether the limitation is justified —
when the previous analysis raised exactly the same question: whether
there was a just cause for the arrest, which was dealt with by the
Court as whether there is a substantive — rather than formal — legal
justification for the arrest. In any case, after engaging in a
determination of the existence of a just cause in terms of section
12(1), the Court moved on to analyse the justifiability of the
limitation in terms of section 36 of the Constitution. Here the same
reasons are presented in a slightly different manner, according to the
language of the factors described in section 36. Thus, while the Court
recognised that there is a connection between the arrest and the
objective to be achieved — facilitating debt collection — the
impugned provisions are not strictly necessary, have not considered
less restrictive options, and are over-inclusive, because even they
might facilitate the payment of debtors that are in a position to pay
but refuse to do so, they also strike debtors that have no means to pay
their debt. The threshold of R40 to trigger the order of arrest was also
seen as proof of the disproportion between means and purpose.
Finally, the Court weighed the importance of the right to personal
freedom. Taking all these factors into consideration, the Court held
that the limitation is not reasonable and justifiable in an open and
democratic society based on human dignity, equality and freedom,
and declared the provision unconstitutional.
Here, the Court seemed to deny the existence of a lexicographical

order of questions to address constitutional challenges to legislation
— that is, first rationality, then proportionality — as seems to be
suggested in other cases.42 Even if in previous cases the Court had said
that rationality is the minimum muster that all acts of the public
authorities should pass, here it simply avoids the question of the
rationality of the legislative scheme and tests the impugned
provisions with a proportionality analysis. Given the arguments used
to declare the statute unconstitutional under the proportionality test
— which include the lack of necessary connection between means and
purpose — a stringent application of a rationality test would probably
have lead to the same result. The fact that the solution of the case
looks relatively straightforward and surely elicits sympathy for the
applicant might also have moved the judge to use a standard of
scrutiny that allows for a more expansive deployment of arguments
against the constitutional validity of the statutory provisions — rather
than sticking to the minimum necessary.
42 See, for example, Affordable Medicines Trust (n 8 above) para 92.

The fondness to expand on arguments to strike down the statutory

provisions at stake seems to offset the silent avoidance to comment
on the seemingly dubious nature of the contract that linked the
parties. Ms Malachi, of Moldovan origin, was brought to South Africa
hired by Cape Dance Academy International and House of Rasputin
Properties to perform the duties of an ‘exotic dancer’. Upon arrival in
South Africa, the employers order her to surrender her passport to
them and refused to give it back unless she reimbursed them with the
money they allegedly spent on her pursuant to the terms of the
contract. After several months of work, instead of making a living, she
appears to be indebted to her own employers for R100 000. The
employers use the tanquam suspectus de fuga injunction to arrest her
in order to avoid her from leaving the country and force her to pay her
debt. Any bona fide reader of these facts can at least see hints of
human trafficking or forced labour. The Court might have felt
uncomfortable to frame the issue in those terms, and to distil any
legal consequences from such a framing — for example, to develop the
legal duty of Courts to eradicate and punish human trafficking or the
ensure compliance with the constitutional and international
prohibition of servitude and forced labour (section 13 of the South
African Constitution).
5 Conclusion
As I tried to show in the discussion, the examination of the case law

of the Court shows that rather than strict formulas that determine the
outcome of the constitutional challenges examined, the standard of
scrutiny chosen and the intensity of the scrutiny are just flexible
departing points to address the issues at stake, but rarely determine
the outcome of the case. The three cases commented on in this article
seem to confirm this idea. The distinctions between rationality,
proportionality and reasonableness as standards of scrutiny are
artificially stressed in some circumstances, while it is rather the
fluidness and the inter-linkages that are underscored in other. The
Court rarely sticks in the actual analysis of each case to the
apparently rigid stock formulas and prongs that it uses to define the
required standards of scrutiny. These points may rather call for an
enquiry into the reasons that might move the Court to grant either a
more deferential or a more stringent review of the action or inaction
of the political branches of government.
THE DEMOCRATIC TURN AND (THE LIMITS OF)
CONSTITUTIONAL PATRIOTISM AFTER THE TRUTH
AND RECONCILIATION COMMISSION:
ALBUTT V CSVR
Wessel le Roux*
1 Introduction
On 23 February 2010, the South African Constitutional Court

unanimously dismissed an appeal against an interim interdict which
prevented President Zuma from pardoning apartheid criminals, until
such time as the victims of the crimes in question had been given an
opportunity to make representations to the President.1
Given its subject matter, the Albutt judgment forms part of an

important quartet of recent cases in which the Constitutional Court
was asked to revisit the nature, scope and effect of the amnesty
process administered in the late 1990s by the Truth and Reconciliation
Commission (TRC) under the Promotion of National Unity and
Reconciliation Act 34 of 1995.2 An obvious way of reading the Albutt
judgment would therefore be to explore its place within the recent
constitutional retrospective of the TRC process. Fruitful and
important as such a reading might be, it is not one which I wish to
pursue here.
I wish to suggest, instead, that the Albutt judgment should be

read in the context of another series of recent Constitutional Court
* Professor in the Department of Public Law and Jurisprudence, University of the

Western Cape. Thanks to Jacques de Ville for raising important questions about
the theoretical perspective adopted in the discussion below. My attempt to
respond to these questions will unfortunately have to stand over until another
time.
1 Albutt v Centre for the Study of Violence and Reconciliation 2010 3 SA 293 (CC).
The interim interdict was granted by Seriti J in the North Gauteng High Court on
29 April 2009 (see Centre for the Study of Violence and Reconciliation v President
of the RSA [2009] ZAGPPHC 35 (29 April 2009)).
2
The quartet must be read together with Azanian Peoples Organization (AZAPO) v
President of the RSA 1996 4 SA 672 (CC) and, in addition to the Albutt judgment,
includes Du Toit v Minister for Safety and Security 2009 12 BCLR 1171 (CC),
Minister for Justice and Constitutional Development v Chonco 2010 4 SA 82 (CC)
and The Citizen 1978 (Pty) Ltd v McBride 2011 4 SA 191 (CC).
51
52 Albutt v CSVR
cases,3 or as part of a democratic turn on the South African

Constitutional Court.4 Precisely what this democratic turn entails and
in which direction it leads, have become vexing questions in South
African constitutional scholarship.5 As I argue in more detail below,
the Albutt judgment sheds important light on these questions. It
differs from the other cases of the democratic turn, in as far as it
takes the TRC process and the meaning of national reconciliation as
its point of entry into the limits and possibilities of post-apartheid
democracy.
The TRC process has famously been criticised by Mahmood

Mamdani and others for its failure to properly focus on the social, as
opposed to the political, dimension of reconciliation.6 While this
critique holds true as far as it goes, it does assume more or less
3
The series does not constitute a closed list and would include at least Minister of
Health v New Clicks South Africa (Pty) Ltd 2006 2 SA 311 (CC) para 621 - 630 (per
Sachs J); Doctors for Life International v Speaker of the National Assembly 2006
6 SA 416 (CC); Matatiele Municipality v President of the Republic of South Africa
(2) 2007 1 BCLR 47 (CC); SABC v National Director of Public Prosecutions 2007 1
SA 523 (CC) para 134 - 153 (per Sachs J).
4
The democratic turn is traced in more detail by T Roux ‘Principle and pragmatism
on the Constitutional Court of South Africa’ (2009) 7 International Journal of
Constitutional Law 106 130; and T Roux ‘The principle of democracy in South
African constitutional law’ in S Woolman & M Bishop Constitutional conversations
(2008) 79 94.
5 In the context of socio-economic rights, Danie Brand argues that it represents an
important counter-veiling tendency against the depoliticisation of social conflict
inherent in socio-economic rights jurisprudence; that is, provided that the turn
towards participatory democracy is not limited to participation in constitutional
institutions. For this reason he takes issue with Theunis Roux’s Dworkinian reading
of the democratic turn as nothing but a constructive interpretation of the
constitutional principle of democracy (D Brand ‘Writing the law democratically: a
reply to Theunis Roux’ in Stu Woolman & Michael Bishop (eds) Constitutional
conversations (2008) 97). Beyond this debate, Stu Woolman and Henk Botha
understand the democratic turn as a shift in the Court's jurisprudence from
private law dignitas to dignity as self-government (S Woolman & H Botha
‘Limitations’ in S Woolman et al (eds) Constitutional law of South Africa (2nd
Edition, OS, 2008) 34 - 116. In the context of value-based rights interpretation
and limitation analysis, this means that the former primacy of dignity as a value
has now been (or should be) replaced with a focus on democracy as the animating
value of the Constitution. The implications of this shift from dignity to democracy
as far as the theory of constitutional interpretation is concerned, remain largely
under-explored. The revival of legal process thinking in the USA represents a
democratic turn of its own. See further JH Ely Democracy and distrust; C Sunstein
The partial constitution; and S Breyer Active liberty (2005). It is perhaps worth
recalling that the latter work occupies a central place in the Doctors for Life
judgment.
6 M Mamdani ‘Amnesty or impunity? A preliminary critique of the Report of the
Truth and Reconciliation Commission of South Africa’ in S Benhabib et al
Identities, affiliations, and allegiances (2007) 325. Mamdani claims that the TRC
had the freedom to define its own mandate, and thus had to explore the link
between ‘the political reconciliation at Kempton Park’ and a larger social
reconciliation (359). The failure to do so narrowed the TRC perspective ‘to a
political reconciliation between state agents and political activists, individual
members of a fractured political elite’ as opposed to a ‘societal reconciliation
between perpetrators and victims’ (326). What the ‘political reconciliation at
Kempton Park’ entailed is not fully explored by Mamdani. It is the question to
uncritically that the TRC process worked with and resulted in an

adequate understanding of political reconciliation as a constitutional
value.7 The Albutt judgment provides a welcome opportunity to
problematise that assumption and to revisit the supposed success of
the TRC process on this score as well; or at least to reopen the
question exactly what model of political reconciliation the TRC
process introduced as constitutional standard for post-apartheid
South Africa.8
The discussion below of the judgment proceeds in three sections.

Each section is organised around a different and directly competing
re-interpretation of the TRC process and the model of political
reconciliation implied in that process. The first section explores the
re-interpretation undertaken by former President Mbeki as he
announced a Special Dispensation for apartheid criminals in 2007. The
second section explores the constitutional challenge against the
Mbeki interpretation and the alternative interpretation of the TRC
process adopted in Albutt (per Froneman J). The final section
explores the re-interpretation of the TRC as a counter-monument
suggested by Aletta Norval, in an attempt to assert the (im)possibility
of political reconciliation against both Mbeki and Froneman J's
understanding of democratic politics.
6
which Albutt seeks to provide an answer. Mamdani's critique is echoed by Karin
van Marle ‘Lives of action, thinking and revolt: a feminist call for politics and
becoming in post-apartheid South Africa’ in W le Roux & K van Marle (eds) Post-
apartheid fragments: law, politics, critique (2007) 34 38 - 40.
7
Mamdani's distinction between political reconciliation and social reconciliation
also informs the distinction between the ‘first transition’ (political) and the
‘second transition’ (social) in recent policy statements of the ANC (see M
Mataboge ‘ANC wants new constitution’ City Press 4 March 2012 1). For a defense
of national reconciliation understood primarily in a political sense, see D
Moellendorf ‘Reconciliation as a political value’ (2007) 38 Journal of Social
Philosophy 205. Moellendorf defends the idea (206) that ‘a political community in
which former strangers view and treat each other as equal citizens is partially
constitutive of reconciliation as a normative goal for political purposes. When the
erstwhile stranger is taken as a fellow citizen, treatment that might have been
thought permissible in the past will be proscribed. Now the person is a co-
participant in the political process, not one to be driven out, contained, or
suppressed. This approach takes reconciliation to be a normative political ideal
that is less than the whole of social justice, but that offers a basis for reasonable
hope that further justice is within the reach of those pursuing it by constitutional
means’.
8
This suggestion should not be misunderstood as a claim that it is fully possible or
even meaningful to distinguish between the political and the other elements of
reconciliation (as some theories of constitutional patriotism seem to imply).
Nancy Fraser, for example, argues that cultural recognition, economic
redistribution and political representation form three inseparable elements of a
comprehensive theory of social justice (Nancy Fraser Scales of justice:
Reimagining political space in a globalizing world (2009) 12 - 30.)
54 Albutt v CSVR
2 Thabo Mbeki and the TRC: towards ‘a new and

common patriotism’
The origins of the Albutt case date back to 21 November 2007, when
former President Thabo Mbeki announced before a joint sitting of
Parliament that he intended to deal with the ‘unfinished business’ of
the Truth and Reconciliation Commission (TRC) in a ‘flexible, decisive
and speedy manner’, so as to finally bring South Africa’s experiment
with transitional constitutionalism to a close.9 Mbeki explained that
some of the business of the TRC remained unfinished, because the
amnesty process was hemmed in by an arbitrary cut-off date which
did not take into account the ongoing violence which accompanied
the transition to democracy.10 This meant that 13 years after
apartheid, there remained thousands of sentenced criminals in prison
for crimes that originated in political violence immediately after the
end of apartheid.
To deal with these cases in the spirit of transitional

constitutionalism, it was necessary to extend the earlier cut-off date
with some five years to 16 June 1999. However, this would still
provide only part of the answer to the problem. Because the TRC
amnesty process could not be reopened, former President Mbeki
announced that he was willing, on application, to pardon all apartheid
criminals who would otherwise have qualified for amnesty under the
TRC process. Mbeki committed himself to deal with each application
on an individual basis, guided by the values of transitional
constitutionalism (‘nation-building and national reconciliation’) and
upholding the ‘principles, criteria and spirit that inspired and
underpinned’ the TRC amnesty process.
In order to confirm and to illustrate his own understanding of this

commitment to the integrity of the TRC process, Mbeki announced
that he would not exercise his Presidential prerogative at his own
discretion:11
I requested the convening of this Joint Sitting to inform the Hon

Members of Parliament that, considering what the nation sought to
9
T Mbeki ‘Address to the joint sitting of Parliament to report on the processing of
some Presidential pardons’ www.anc.org.za/show.php?id=4240 (accessed 20
January 2012).
10
The interim Constitution initially determined 6 December 1993 as cut-off date for
those political crimes that could be considered for amnesty. Because this date did
not accommodate the political violence that surrounding the 1994 general
elections, it was later extended to 10 May 1994 (the date of the inauguration of
former President Mandela). The new date, in turn, did not take into account that
political violence persisted well after the arrival of democracy (especially in
areas such as KwaZulu-Natal).
11 Mbeki (n 9 above).
achieve through the TRC process, I have decided to institute a special

process to assist me as I discharge my constitutional obligation to
consider the requests for pardon ... Further to entrench the practice the
nation has sought to cultivate, of acting in unity as it addresses the
crimes of the past, I would like the political parties represented in our
Parliament to assist me properly to discharge my constitutional
responsibility.
In an attempt to stay true to the spirit of the TRC process, and in the
best tradition of parliamentary or representative democracy, Mbeki
announced the establishment of a multi-party Parliamentary
Reference Group (PRG), with the task of making considered
recommendations to the President about each pardon application
that had been received. To ensure that the PRG was not divided by
the strategic pursuit of party political interests, Mbeki invited every
party represented in Parliament to appoint one single representative
to serve on the PRG. This principle of party political equality
effectively neutralised the power of the majority party in Parliament
on the PRG. As the leader of that party, Mbeki wished to symbolically
and practically illustrate that the true spirit of national reconciliation
and nation building found expression in the conscious turn away from
strategic interest group politics and majoritarian democracy, to an
alternative, deliberative or dialogical model of democracy.12
The establishment and composition of the PRG powerfully

symbolised the deeper deliberative nature of post-apartheid
constitutional democracy. It also embodied an alternative self-image
of the post-apartheid nation: as collectively engaged in a deliberative
political project in which political power has been traded for political
or constitutional principle. The reconciled nation appears here as a
‘community of principle’ in Dworkin's sense of the term; the PRG as
the deliberative telos of nation-building and national reconciliation.13
12
For our purposes the following definition of deliberative democracy by Frank
Michelman will suffice (‘Conceptions of democracy in American constitutional
argument: The case of pornography regulation’ (1989) 56 Tennessee Law Review
291 – 319 293):
Deliberative politics connotes an argumentative interchange among
persons who recognise each other as equal in authority and entitlement
to respect. [...] [I]t refers to a certain attitude towards social
cooperation, namely, that of openness to persuasion by reasons referring
to the claims of others as well as one’s own. The deliberative medium is a
good faith exchange of views in which all participants remain open to the
possibility of persuasion by others and in which a vote, if any vote is
taken, represents a pooling of judgements.
13 It is easy to see where this interpretation of Mbeki's own democratic turn could
lead. Dworkin suggests that once we accept the image of the nation as a
community of principle, then the Court can legitimately assume the task of
holding us true to that better image of ourselves, and to ensure that Parliament
always operates as a forum of principle (as opposed to power). The Constitutional
Court could then be understood as the permanent institutionalisation of the PRG,
as a fragile deliberative space within the context of otherwise strategic
parliamentary politics. As I suggest below, the value of Albutt is precisely that it
56 Albutt v CSVR
For Mbeki the PRG was thus from the start far more than merely the
functional heart of the Special Dispensation process. It was also the
symbolic expression of the true deliberative nature of post-apartheid
parliamentary democracy and national reconciliation.14
That we are dealing here with a counter-factual and demanding

conception of democracy is immediately clear from Mbeki's own
address. Having announced the Special Dispensation for apartheid
criminals, Mbeki immediately turned his attention to the opposition
parties in Parliament, in the hope of convincing and motivating them
to accept his invitation to participate in the PRG process. He did so
with a direct appeal to the Preamble of the Constitution and the TRC
process. Mbeki reminded Parliament that the victims of apartheid did
not ask for retribution, but for public remembrance (this was the
powerful message of the TRC process). Public remembrance in this
sense required that we put our own party political interests aside, and
recall the principles for which the victims of apartheid had suffered
and died in everything we do. Mbeki can even be read to suggest that
it was precisely in order to facilitate this public memory that those
principles were entrenched in the Constitution. In this way the
Constitution serves today as their memorial and our loyalty to the
constitution's founding values and principles as our commemoration of
the past.15 It was in this spirit that Mbeki called on all political parties
to stand together in unity behind the Special Dispensation, ‘moved by
a new and common patriotism’.16
13 enables us to resist this attempt to put the Court forward as a mirror of the
nation's ideal political self or its absent self-government.
14
See Democratic Alliance v Masondo 2003 2 SA 413 (CC) for a similar understanding
of the deliberative and dialogical character of representative politics, precisely in
order to secure national reconciliation and unity at the level of our deeply
divided cities (per O'Regan and Sachs JJ).
15
Johan Snyman presents a reading of memorial constitutionalism which serves to
undermine this legitimating rhetoric powerfully. The duty to remember the
principles for which the victims of apartheid suffered died can never be
discharged in the form of loyalty to the present Constitution or the present
interpretation of its founding values and principles (a Dworkinian engagement
with the Constitution). It requires a constant refusal of this kind of constitutional
closure and constitutional patriotism. As he says: ‘The politics of memory is never
completed, because the norm for what we have to do today can never be stated
in terms clear enough’ (Johan Snyman ‘Thoughts on dealing with the legacies of
radically unjust political behaviour’ in W le Roux & K van Marle (eds) Law,
memory and the legacy of apartheid (2007) 3 - 10.) For a fuller refection on this
radical incompleteness of constitutionalism as a politics of memory, see 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 (2007) 65 - 90. I return to this theme further
below in section 4 of this note.
16
Mbeki's fear that some political parties might refuse to participate in the PRG
must be understood in light of the reference earlier in his address to the political
violence between supporters of the ANC and IFP in KwaZulu-Natal, and the anger
caused by his failure, over a period of five years, to deal with a large number of
pardon applications by IFP members.The dispute between the IFP and the
President eventually ended in the Constitutional Court. See Minister for Justice
and Constitutional Development v Chonco 2010 (4) SA 82 (CC).
Mbeki did not further clarify what he understood by this ‘new

patriotism’ but the context in which the term was used, as I explained
it above, made clear enough what he had in mind.17 As Elsa van
Huysteen suggests, the ‘new patriotism’ celebrated by Mbeki (and
Nelson Mandela before him), must be understood as a local version of
the constitutional patriotism (Verfassungspatriotismus) that Jürgen
Habermas started popularising in the late 1980s as the only basis for
the political integration of post-communist Germany and Europe as a
whole.18 It was the spirit in which Mbeki wanted Parliament to
conduct its business in the shadow of the TRC process.
3 Albutt and the TRC: towards the ‘pervasive

demands for participatory living’
One of the apartheid criminals who could not apply for amnesty under
the TRC process was Ryan Albutt. When the Special Dispensation was
announced, Mr Albutt was serving an eight years prison sentence for
his part in a fatal attack by a white vigilante group on striking black
municipal workers.19 The attack was organised by the right wing
separatist Afrikaner Weerstandsbeweging (AWB) and took place in
Kuruman at the end of August 1995, well after the cut-off date for
amnesty under the TRC process (10 May 1994). Having being invited
by former President Mbeki to do so under the Special Dispensation
discussed above, Mr Albutt applied for a Presidential pardon during
the window period between 15 January 2008 to 31 May 2008.
17
This was of course not the first time that Mbeki had spoken about a ‘new
patriotism’ that united the post-apartheid nation. In fact, the phrase had
dominated the Presidential rhetoric and thinking of Nelson Mandela in die mid
1990s, when Mbeki was serving as deputy President and began adopting it into his
own vocabulary and constitutional thinking. At the unveiling of a mural in Cape
Town on 8 May 1996, to celebrate the adoption of the Constitution, former
President Mandela celebrated the ‘power of the New Patriotism’ and, just as
Mbeki would do a decade later, suggested that this patriotism originated in and
was sustained by the unique nature of public memory in post-apartheid South
Africa: ‘The moving testimony of witnesses before the Truth and Reconciliation
Commission has reminded us of the injustices of the past and of the great debt we
owe to those who suffered for freedom and justice. It has moved us all to renew
our resolve that never again shall racial discrimination be allowed to blight the
lives of our people’ (full speech available at http://www.info.gov.za/speeches/
1996/960513_0x772.htm (accessed 28 October 2012). Mandela's words clearly
borrowed from the Preamble of the Constitution where the idea of the
Constitution as an apartheid memorial found formal expression. His words also
remind us that the drafting process of the Constitution and the TRC process
coincided for a short but decisive period in the autumn of 1996.
18 E van Huyssteen ‘“The glowing fire of our new patriotism”. The Constitutional
Court, civil society and constitutional patriotism in South Africa’ in T Cruz Maria
& SA Sitas (eds) Gathering voices. Perspectives on the social sciences in Southern
Africa (1996) 73. See J Habermas The divided west (2006) 67 - 82 and J Werner-
Müller Constitutional patriotism (2007) 15 - 45.
19 Detail about the offence is contained in S v Whitehead [2007] ZASCA 171.
58 Albutt v CSVR
Having received Mr Albutt's application (along with more than

2000 other applications), the PRG began processing the cases one by
one, looking primarily at the record of the criminal proceedings.20 It
did so behind closed doors, without releasing the names of the
applicants to the general public. A number of civil society
organisations approached the PRG, and eventually also the President,
requesting that more detail about the applications be released and
that the victims of the crimes in question be given an opportunity to
make representations to the PRG. Both the PRG and the President
refused to accede to these requests. The civil society organisations
thereupon approached each of the political parties represented on
the PRG individually, in an attempt to secure greater openness and
transparency in the PRG process, but without any more success.
When it was announced in March 2009 that the PRG had finally
provided the President with a list of the apartheid criminals which it
recommended for pardon, a coalition of civil society organisations,
including the Centre for the Study of Violence and Reconciliation
(CSVR), applied on an urgent basis in the North Gauteng High Court for
an interim interdict preventing the President from granting any
pardon, unless and until the victims have been given access to the
applications and an opportunity to make representations. Mr Albutt,
whose name appeared on the list of approved applicants, intervened
in the litigation and opposed the application, arguing with the
President that the victims had no constitutional right under the PRG
process to a hearing.
Seriti J granted the interim interdict in favour of the NGO

coalition on 29 April 2009. He ruled that the decision whether to
pardon an offender constituted administrative action under the
Promotion of Administrative Justice Act 3 of 2000;21 that the right of
victims to make representations before a prisoner is released on
parole should equally apply when a prisoner is pardoned;22 and that
the President should be held to his public commitment to adhere to
the basic spirit and principles of the TRC process, which included
victim participation and transparency (the direct opposite of the PRG
process up to that point).23
Mr Albutt thereupon applied to the Constitutional Court for leave

to appeal against the interim interdict issued by Seriti J (an
application supported by the President). It was of course not the first
20
The factual background appears from the judgment of Seriti J in Centre for the
Study of Violence and Reconciliation v President of the RSA [2009] ZAGPPHC 35
(29 April 2009).
21 Para 7.3. This finding was set aside by the Constitutional Court para 76.
22 Para 7.4.2. This finding was set aside by the Constitutional Court para 76.
23
Para 7.4.3. This finding was confirmed by the Constitutional Court para 69. See
further below.
time that the Constitutional Court had been asked to review a

decision by the President to pardon (or not to pardon) a convicted
criminal. Previous case law of the Court had firmly established that
the exercise of this prerogative power is subject to judicial review,
and that such a review could focus on the legality of the decision (was
the power sourced in law?); the rationality of the decision (was the
exercise of the power related to its objective?); and the
reasonableness of the decision (does the exercise of the power
unreasonably limit any fundamental right?). While the issue of legality
dominated the Court's previous judgment on the issue in Chonco,24
and the issue of reasonableness the judgment in Hugo,25 the
judgment in Albutt turned solely on the question of rationality, or the
relationship between the means chosen (consideration behind closed
doors without any victim input) and the end to be achieved (national
reconciliation in the political sense of the term).26
While the Constitutional Court granted Mr Albutt leave to appeal,

the appeal itself was unanimously dismissed. Ngcobo CJ concluded
that the decision by the President and the PRG to exclude the victims
from participating in the process was irrational, given the objectives
of ‘national reconciliation and national unity’ that the President had
himself identified as objectives when he announced the Special
Dispensation. Even though rationality review is very much an internal
form of critique, measuring the threshold effectiveness of a policy or
dispensation against the objectives it set for itself, the process
requires a clear understanding of what the objectives in question
entail.
In as far as the South African understanding of ‘national

reconciliation and national unity’ was largely determined by the TRC
process, the arguments before the Court in Albutt turned for the most
part on a reinterpretation of the TRC process. What did the TRC
process really say about the nature of the nation as a reconciled unity?
What was the test for the successful political reconciliation of the
24
Minister for Justice and Constitutional Development v Chonco 2010 4 SA 82 (CC).
The question was whether the Minister had any legal power to grant or receive or
consider applications for Presidential pardons.
25
President of the Republic of South Africa v Hugo 1997 4 SA 1 (CC). The question
was whether reprieving only female prisoners with young children violated the
right to equality of fathers with young children.
26
The legality of the power of the President to pardon apartheid criminals and to
institute a Special Dispensation to do so, was universally accepted to be derived
from section 84(2)(j) of the Constitution (see Albutt para 52). It was also
universally accepted, given what was aid in Hugo, that the pardoning of apartheid
criminals could not be said to unreasonably limit any fundamental rights (save
possibly the right to fair administrative action, an issue which the Constitutional
Court deliberatively decided not to consider (see Albutt para 79 - 82). This left
only the rationality of the Special Dispensation as basis for constitutional review.
60 Albutt v CSVR
South African nation?27 Was former President Mbeki correct that the
PRG, as a non-majoritarian model of multi-party deliberative
democracy, best represented the political dimension of post-
apartheid reconciliation?
As an answer to this question and an insightful reinterpretation of

the TRC process as a model of political reconciliation, the judgment
of Ngcobo CJ is highly disappointing. Studded with rhetorical
platitudes, the judgment never explains why the TRC process,
including the central role of victim participation, provides a
condensed model (or as we will see later perhaps even an anti-model)
of the post-apartheid political process, and thus can serve as binding
model for other political institutions like parliament, or more
specifically the PRG. In order to do so, the court would have had to
directly engage the nature of national reconciliation as a political
concept, or, failing that, to carefully deal with the differences and
similarities between the TRC and the PRG.
There is no attempt in the judgment to undertake the first task.

An attempt to undertake the second eventually runs out of steam
when Ngcobo CJ announces that the effort to establish the
differences and similarities between the two processes is
misguided.28 The simple fact is that, even if there were indeed
significant differences between the two processes, former President
Mbeki was aware of those differences and had ‘[d]espite these
differences’ decided to apply the principles and values of the TRC
process to the PRG process. Irrespective of the merits of this decision,
‘the subsequent disregard of those principles and values without
explanation was irrational’.29 A great deal of the judgment is, on this
basis, simply devoted to a descriptive overview of the principles and
values that underpinned the TRC process (as if to remind Mbeki about
the implications of the decision he had taken, without giving Mbeki's
own interpretation of that commitment any time of day).
Ngcobo CJ does not find that the exclusion of victims is irrational,

let's say for argument's sake, because participation in all decision-
making processes is a necessary precondition of post apartheid
democracy, so that exclusion of this condition would be defeating the
object of the exercise. He finds instead that the decision is irrational,
27
The case turned on the political dimension of national reconciliation only. The
idea of constitutional patriotism as a theory of reconciliation assumes that this
constitutional or political dimension (citizenship) can be meaningfully
distinguished from other dimensions of reconciliation. This does not mean that
democratic citizenship, or what Nancy Fraser calls ‘participatory parity’, cannot
embody an exacting post-nationalist and post-national theory of social justice
(see further N Fraser ‘Re-framing justice in a globalizing world’ in T Lovell (ed)
(Mis)recognition, social inequality and social justice (2007) 17 – 35).
28
Albutt paras 66 - 68.
29 Albutt para 69.
simply because former President Mbeki had changed his mind without
any apparent reason. Ngcobo CJ turns what might at worst have been
a misunderstanding into a patent irrationality.
To be fair, Ngcobo CJ states over and over again that victim

participation is a necessary condition for national reconciliation. The
problem is that he never explains the reasoning behind what soon
turns into a mostly rhetorical statement. The closest Ngcobo CJ comes
to such an explanation is to list a number of instrumental
considerations which necessitated victim participation in the TRC
process (most notably the lack of available knowledge about the gross
human rights violations of the past,30 and to provide an opportunity
(in the absence of criminal trials) for those who were violated to
receive public recognition that they had been wronged).31
The weakness of this instrumental explanation of victim

participation in the TRC process, as basis for the claim that victims
must also participate in the PRG process, is exposed in the counter-
argument put forward by Albutt and the President.32 They argued that
victim participation was not instrumental to the PRG process because
all the applicants for Presidential pardons were already convicted
criminals. The truth about their crimes had been conclusively
established during criminal trials, and the victims of these crimes had
participated in those trials as witnesses, thus receiving public
recognition that they had been wronged. These crimes thus no longer
stood in the way of national reconciliation as numerous unearthed
apartheid crimes did at the end of apartheid, before the TRC process.
In fact, in contrast to the amnesty process of the TRC, justice had not
been compromised in these cases for the sake of the truth. This was
more than the TRC process had ever sought or managed to achieve.
As said above, Ngcobo CJ's final response to this argument is that

Mbeki knew about this crucial difference, but nevertheless decided to
apply the requirement of victim participation to the PRG process; and
that, as far as Ngcobo CJ was concerned, was the end of the matter.33
This is an extremely weak basis for the judgment, given, as I have
tried to point out above, that it is at least open to interpretation
whether Mbeki had indeed deviated from his original understanding of
what the principles and values of the TRC process had required of
him, given the unique circumstances of the Special Dispensation
process. In the absence of a deeper understanding of what
reconciliation means in a political sense, or a theory of post-apartheid
democracy, Ngcobo CJ seems unable to respond to the argument that
30 Albutt para 65.

31 Albutt para 60.
32
Albutt para 62.
33 Albutt para 67.
62 Albutt v CSVR
victim participation is at best a contingent feature of some

reconciliation processes, but not of all.
Perhaps it was because he realised that Ngcobo CJ's judgment did

not sufficiently engage the substance of national reconciliation as a
political concept, that is, that it lacked a working definition of
democracy, that Froneman J felt compelled to write a short
concurring judgment in which he directly confronted the deliberative
model of parliamentary democracy which Mbeki sought to
institutionalise with the creation of the PRG. Froneman J's starting
point is that it is necessary to look beyond the contingencies of our
recent history (which includes the contingent participatory features
of the amnesty process and the contingent way in which the pardon
process was introduced) to the deeper principle of democracy that is
at work in our broader constitutional history, stretching back as far as
pre-colonial times.34 This shift in perspective allows Froneman J to
present the centrality of victim participation in the TRC process as the
culmination of an old and pervasive understanding of democracy.
Borrowing a phrase from Amartya Sen, Froneman J presents this
broader constitutional history as a response to the ‘pervasive
demands for participatory living’.35 This demand constantly
resurfaces in that history, not as a binding model or foundational
moment, but as an inspirational leitmotiv.36 This demand is also
respected in our Constitution. As the democratic turn on the
Constitutional Court has made clear, ‘the democracy our Constitution
demands is not merely a representative one, but is also, importantly,
a participatory democracy’.37 That we live in a participatory
democracy with roots that reach back to pre-colonial times also holds
true when Presidential pardons are to be granted to apartheid
criminals:38
Council for the applicant argued that the requirement of victim

participation was met through the process set in place by the President
which involved all the political parties represented in Parliament. Put
differently, the argument was that representative democracy was
sufficient in the circumstances. It is not.
As this passage makes clear, Froneman J understood the case before

him as a direct confrontation between representative democracy and
participatory democracy, and the further question which of the two
conceptions best understands the meaning of the TRC process for
post-apartheid democratic politics. One could also say that the case
turns on different understandings of what constitutional patriotism
34
Albutt para 89.
35 Albutt para 92. See Amartya Sen The idea of justice (2009) 322.
36 Albutt para 91.
37
Albutt para 90.
38 Albutt para 90.
demands of post-apartheid citizens after the TRC. In as far as the

memory and constitutional recognition of past injustices (the TRC
process and the Preamble of the Constitution) precludes a certain
celebratory engagement with history and the revival of an
uncomplicated national pride, constitutional patriotism arises as the
only achievable, but even more so, only permissible form of political
identification for South Africans. Mbeki accepted as much when he
called on political parties, in the name of the TRC process, to convert
their party loyalties into a new constitutional patriotism. However, he
was deeply mistaken to assume that the privileged site for the
formation of this new patriotism could be Parliamentary politics, even
in the exacting form of deliberative democracy required by the PRG
process.
In contrast to Mbeki, Froneman J follows most champions of

constitutional patriotism, including Habermas himself, to insist that
the emergence of a proper constitutional patriotism in divided
societies depends on active citizen participation in a vibrant public
sphere. If we were to take the memory work of the TRC process
seriously, if the TRC became the eyes with which we look at the past,
forced backwards into the future, then we would have to do
everything we can to deepen and strengthen active citizenship and
participatory democracy in society. A narrow focus on the perfection
of the internal dynamics of representative democracy would be
misplaced, especially for somebody who understood and wanted to
take seriously Mbeki's call for a new constitutional patriotism.
Froneman J does not further elaborate on the different forms that

active citizen participation in a fragile post TRC public sphere could
take. We know from Albutt that active citizen participation would at
the very least include the right of apartheid victims to participate in
the PRG before a final decision about the granting of a pardon is
taken. Such participation would have constative and not merely
instrumental value. But how should we understand citizen
participation in the broader sense after Albutt? The question is
complex and cannot be fully pursued here. The task is not made easier
by the fact that Froneman J's judgment seems to lead us in two
different directions and two versions of participatory democracy. The
first opens a place for civil society participation or politics within the
constitutional matrix, in the name of a life in politics not limited by
or restricted to party political activities, like electioneering and
voting. The second shifts the focus to the tensions between politics
and die constitutional matrix, in the name of a life in politics not
limited by or restricted to constitutional activities, like making and
litigating rights claims or deepening the rights discourse generally,
but driven rather by the attempt to think and bear witness to the
64 Albutt v CSVR
political and activating the Differend (as Lyotard might have phrased
it).39
On the one hand, Froneman J explicitly relates his use of the term
participation to the democratic turn of the Constitutional Court in
Docters for Life and Matatiele.40 Both these cases read a requirement
of public participation into the legislative process, and thus
understood participatory democracy as the right to participate
constructively within the specific institutional spaces and
opportunities created for this purpose by the Constitution. Thus the
PRG was constitutionally defective because it did not create a similar
institutional opportunity for victim participation in the decision-
making process. However, the Constitution does not only create or
demand opportunities for first order institutional participation in this
form, it also recognise the importance and entrenches the possibility
of second order citizen participation in politics through constitutional
review and litigation. The Albutt case provides a good example of
this. The litigation provided an opportunity for and facilitated active
civil society participation in society, in an effort to ensure an
opportunity for victims of apartheid crimes to participate in the
decision-making process of the PRG. Like Doctors for Life and
Matatiele before it, the Albutt judgment vindicated the value and
legitimacy of non-representative and extra-parliamentary political
action and citizen participation in society; a life in politics not limited
to party political activities and the right to vote. The case became a
clear example of civil society flexing its muscles as is clear from Hugo
van der Merwe's criticism of parliamentary politics directly after the
judgment of the High Court:41
In Parliament [opposition parties] welcomed the announcement of the

new prosecutions policy, which was subsequently judged
unconstitutional. After they were co-opted into the reference group
steering the presidential pardon process, they somehow became
39 To deviate slightly from the discussion, Lyotard distinguished between two styles
or modes of politics. The one aimed at litigating constitutional claims; the other
at bearing testimony to and activating the Differend (note, not the differences,
as Lyotard's postmodernist injunction is often misunderstood). As the Differend
marks the inability to phrase something in a manner that would make litigation
within a common constitutional idiom possible, Lyotard was never attracted to
constitutional doctrine, but explored the possibilities that art and different styles
of modernist writing presented for bearing testimony to the Differend (ultimately
the inability of rephrasing normative claims (obligations) as constative claims,
made in the name of ‘We the people’). Postmodernism as a resistance to the
hegemony of scientific discourse, and the demand for the phrasing of justiciable
knowledge claims as model for phrasing within the political and ethical language
games. See further Jean-François Lyotard The Differend: Phrases in dispute
(1988). I return to this issue in the final section below.
40 Albutt para 90.
41 H van der Merwe ‘No thanks to the opposition for brake on special pardons’
Sunday Independent 3 May 2009 www.armsdeal-vpo.co.za/articles14/
pardons.html (3 November 2011).
resistant to pleas to help bring survivor's voices into the process.

Political parties were at best silent spectators, at worst collaborators, in
this process that has also been declared to be inconsistent with our
constitution. In both these cases, it has been civil society organisations
and survivors themselves who have had to speak out, and eventually turn
to the courts to curb the abuse of power. … While much attention has
been focused on the elections, we should keep in mind that a vibrant
civil society remains a pillar of democracy and guarantor of human rights
protection.
In the mind of the applicants at least, participatory democracy after

Albutt is synonymous with the re-emergence of a vibrant civil society
outside the institutional framework of representative democracy,
capable of undertaking public interest litigation.42 This positive view
of the balance of political power that emerges from the Albutt
judgment, has, predictably, not convinced everybody.
As the Zuma presidency gradually entrenches itself, Albutt's court

centred version of participatory democracy has increasingly come
under pressure from a growing constitutional populism. According to
the champions of this populism, the court centred model of
participatory democracy evidenced in Albutt is fundamentally
subversive of the democratic political process. Shortly after the
Albutt judgment was delivered, Minister of Justice, Jeff Radebe,
warned a group of young lawyers about the rise of the new civil
society (‘the unbridled rush to create my own nongovernmental
organisation’) and described a litigation hunger civil society as the
real threat to the constitution (as opposed to the hypothetical threat
that the ANC under President Zuma supposedly posed to the
Constitution).43 A few months later, the ANC's General Secretary,
Gwede Mantashe, again spoke out against the tendency of civil society
groups and minority parties to abuse the Constitutional Court to
sustain opposition politics against the electoral mandate of
Parliament.44
These comments signal the return of the counter-majoritarian

difficulty to the centre stage of post-apartheid constitutional
debates. There is of course by now a ready answer available to the
charge that civil society driven constitutional review is a counter-
democratic force in society. That answer is that the Court and its
review powers are, correctly understood, proto-democratic or hyper-
42
The possibilities and limitations of public interest litigation has become a key
theme in the aftermath of Albutt. See, for example, the special volume of South
African Journal on Human Rights (2011) dedicated to the issue.
43
‘NGO forces real threat to constitution — Radebe’ www.businessday.co.za/
articles/Content.aspx?id=127404 (accessed 20 January 2012).
44 ‘Full Interview: ANC's Mantashe lambasts judges’ www.sowetanlive.co.za/news/
2011/08/18/full-interview-ancs-mantashe-lambasts-judges (accessed 20 January
2012).
66 Albutt v CSVR
democratic in nature. This answer rests on the claim that

parliamentary democracy is notoriously suboptimal as a matter of
democratic process (Ely),45 political principle (Dworkin),46 and
practical rationality (Michelman).47 For these reasons, the review of
Parliament by an independent Constitutional Court is actually
democracy reinforcing or democracy perfecting.
While civil society organisations will no doubt take heart from

these attempts to vindicate constitutional litigation as a viable form
of democratic politics, the new parliamentarians are not the only
ones who have expressed concern about a model of participatory
democracy and constitutional patriotism that revolves primarily
around the litigation of constitutional rights claims. Having carefully
studied the dynamics of socio-economic rights litigation before the
Constitutional Court, Danie Brand concluded recently that
constitutional rights adjudication is at best a limited and at worst a
limiting instrument for political transformation.48 Brand relies on the
work of Emilios Christodoulides,49 to suggest that South Africans
check their uncritical ‘constitutional optimism’ (read constitutional
patriotism) and start exploring alternative possibilities of agonistic
politics in society.50 Karin van Marle extends and radicalises Brand's
reading of socio-economic rights to a critique of modern law and
constitutional democracy in general. Also relying on the work of
Christodoulides, Van Marle suggests that we develop a memorial
45
JH Ely Democracy and distrust: A theory of judicial review (1980) calling on
Constitutional Courts to clear the cannels of political change (105 - 134) and to
facilitate the representation of minorities (135 - 180), the two constitutional
pathologies of a democratic process of representation. Ely wished to restrict the
work of the Court to these two democracy reinforcing tasks in an attempt to
escape the intrusion of values into the adjudication process.
46 R Dworkin ‘The forum of principle’ (1981) 56 New York University Law Rev 469; R
Dworkin Freedom's law: The moral reading of the American Constitution (1997) 1
- 38. According to Dworkin, it is perverse for a majority who has violated the
conditions or moral foundation of democracy (treating everybody with equal
concern and respect) to appeal to democracy to inslulate itself from
constitutional review.
47 F Michelman ‘Law's republic’ (1988) 97 Yale Law Review 1493 1531 - 1532 arguing
that the civil rights movement shows that the primary arenas of jurisgenerative
politics lie outside the major, formal channels of electoral and legislative politics
in civil society and street life. This non-state centred notion of republican
citizenship, not tied to centralised majoritarian power, provides a rejoinder to
the majoritarian and populist objection to judicial review: ‘The Court helps
protect the republican state- that is the citizens politically engaged - from lapsing
into a politics of self-denial […] to assume their own moral completion as they
now are and thus to deny to themselves the plurality on which their capacity for
transformative self-renewal depends’ (1532). In ‘The Supreme Court 1985 term -
Foreword: Traces of self-government’ (1986) 100 Harvard Law Review 4,
Michelman argues that the task of the Constitutional Court is to virtually
represent to the people their otherwise absent democratic self-government.
48
Brand (n 4 above) 101 - 106.
49 E Christodoulides ‘Constitutional irresolution: law and the framing of civil
society’ (2003) 9 European Law Journal 401.
50
D Brand (2009) Courts, socio-economic rights and transformative politics
unpublished LLD thesis, Stellenbosch University, 2009, 41.
constitutionalism in contrast to the monumental constitutionalism

that she sees implied in our ‘constitutional optimism’,51 or
‘constitutional patriotism’.52 She seeks to explore the possibility of
non oppressive forms of community that are not grounded in or
mediated by law. To this end she insists that we keep the possibility
of political action alive, outside the framework of the constitution,
and thus not as the making and vindication of rights claims. Van Marle
does not simply mean that constitutional democracy (or community)
is an oxymoron, and thus that democracy or community can constitute
a self-present alternative to the mediation of law. It is a question of
writing or opening the possibilities of community and political action
in spite of the law, not in place of the law.53
I suggested above that Froneman J's judgment can also be read to

further the more agonistic and less institutional understanding of
participatory democracy suggested by Brand and Van Marle.
Froneman J's reliance on the work of Amartya Sen points us in the
direction of a second interpretation of the demand for participatory
living.
In the Introduction to his book The idea of justice Sen

distinguished between two traditions in the history of justice.54 The
first is a transcendental tradition which aims to define perfect justice
and to focus on the stable institutional arrangements that would
ensure justice in society. The second is a comparative tradition which
relies on comparisons between already existing societies in order to
identify and remove manifest injustices. Sen defends the second
tradition in his book. In this tradition the idea of justice is sustained
by public dialogue about the lives that people are actually capable of
realising at different times and in different societies. In part IV of the
book, the part cited by Froneman J, Sen points out that the difference
between institutional and dialogical approaches to justice is mirrored
by a difference between institutional (or representative) and
dialogical (or participatory) models of democracy. Sen uses this
distinction to undermine the claim that democracy is a uniquely
Western concept without application in non-Western contexts. This
might be true as far as constitutional democracy as a presumed
institutional perfection of democracy is concerned, but it is not true
of the participatory ethos that has informed the idea of justice since
ancient times.55
51 Van Marle (n 6 above) 53.

52
K van Marle ‘Constitutional patriotism or constitutional nationalism? A response to
Wessel le Roux's paper’ (2009) 24 SA Public Law 400.
53 Van Marle (n 51 above) 406.
54
A Sen The idea of justice (2009) 1 - 30.
55 Sen (n 53 above) 326.
68 Albutt v CSVR
When Froneman J follows Sen by tracing the roots of democracy

to pre-colonial times, he is clearly not interested in the institutional
dimension of democracy, in constitutional democracy as such, but in
democracy as a participatory mode of living or being in the world. In
as far as democracy in the sense of the demand for participatory living
attaches to the idea of justice and precede the constitution, it also
exceeds it, and can thus mark the limits of the constitutional attempt
to embody or contain that demand and the idea of justice. There is
thus also a second meaning of participatory democracy latent in
Froneman J's judgment, one that resonates with the critique by Van
Marle and others against the the monumentalisation of institutional
democracy that is implicit in strong forms of constitutional review.56
Froneman J does not pursue this second interpretation of

participatory democracy any further. However, we do have an
indication of how he might have proceeded had he decided to do so.
In an essay on the work of Frank Michelman, which was written some
time before he became a Constitutional Court judge, Froneman J
spoke powerfully along these lines about the ‘[i]mpossibility of
constitutional democracy’.57 In the same volume of essays, Henk
Botha developed a critique of Michelman's defence of court centred
republicanism and spoke about ‘[r]ights, limitations and the
(im)possibility of self-government’.58 These reflections on the
impossibility of any stable or monumental constitutionalisation of
democracy also brings to mind an earlier essay on the TRC, in which
56 My own attempts to explore the limits of a court centred or hyper-democratic

solution to the role of the Constitutional Court has largely been inspired by Jean
François-Lyotard's critique of institutionalised (or constitutionalised) politics, in
the republican mold of Habermas and Michelman. Lyotard's basic point is that the
discursive symmetries implied by democratic self-government, as a solution to
the modern problem of justifying obligations, fails to account for the discursive
asymmetry implied in the very idea of obligation and responsibility. It goes
without saying that for Lyotard this essentially ethical asymmetry, and the modes
of being and forms of community it opens onto, cannot be explored though law
and the constitutional perfection of democracy as a discursive language game.
See W le Roux ‘Six (individually-named) notes on the counter-aesthetics of
refusal’ in K van Marle (ed) Refusal, transition and post-apartheid law (2009) 57 -
78. See also W le Roux ‘Bridges, clearings, and labyrinths: The architectural
framing of post-apartheid Constitutionalism’ in Le Roux & Van Marle (eds) (n 6
above) 59 - 100 for an attempt along similar lines to explore the architectural
design of the Constitutional Court building as a labyrinth, as opposed to a clearing
prepared under a lekgotla tree for the deliberative gathering of the community
(as the icon of the Court suggests). My previous reading of the Constitutional
Court building and its urban environment as an apartheid memorial follows a
similar argument (Le Roux (n 14 above) 65 - 90). Contrast my earlier attempt to
present the Court as a place on the street, in which the idea of street democracy
was still very strongly influence by Michelman's hyper-democratic ideal of the
Constitutional Court as a democratic institution (W le Roux ‘From acropolis to
metropolis: The Constitutional Court building and South African street
democracy’ (2001) 16 SA Public Law 139).
57 J Froneman ‘The impossibility of constitutional democracy’ in H Botha et al (eds)
Rights and democracy in a transformative constitution (2004) 93.
58
H Botha ‘Rights, limitations and the (im)possibility of self-government’ in Botha
et al (eds) (n 56 above) 13.
Aletta Norval suggested that the effect of victim participation in the

TRC process is to mark the ‘(im)possibility of reconciliation’ and so to
open the possibility of a truly post-apartheid and non-identitarian
form of democracy.59
4 Aletta Norval and the TRC: towards ‘the

(im)possibility of reconciliation’
In her essay, Norval discusses the memory work of the TRC and the
centrality of victim participation in the process. Like Ngcobo CJ and
Froneman J, she claims that the most significant aspect of the
memory work is the way in which it has offered an occasion for
survivors to gain recognition of their plight in full public view.60 It is
the focus on the everydayness of injustice and the reoccupation of
memory sites by ordinary citizens that reveals the real significance of
the TRC hearings. However, Norval also suggests that the
participation of ordinary citizens made the memory work of the TRC
different from the standard nationalist uses of memory and
monuments. From this insight she derives an understanding of
reconciliation as a political concept that contrasts fruitfully with both
the deliberative and participatory models celebrated by Mbeki and
Froneman J respectively.
Norval relies on the work of John Gillis to locate the TRC in a

broader history of commemoration in society; a history which reveals
pre-national, national and post-national forms of commemoration.61
For our purposes it is sufficient to reflect on the differences between
the last two forms of commemoration. National memories tend to
focus on the construction of unity and continuity and locate public
memory in archival sites or public monuments. To capture the
difference between national memories and post-national memories,
Norval likens the difference to that between the Catholic practice of
locating the sacred in certain ritualistic times and places, and the
anti-ritualistic demand of Protestantism that the sacred by brought
into everyday life itself. Post-national memories represent a similar
iconoclasm in the sphere of public memory and ‘attempts to
desacralise the nation-state, to democratise memory, and to retrace
59
A Norval ‘Memory, idenity and the (im)possibility of reconciliation: The work of
the Truth and Reconciliation Commission in South Africa’ (1998) 5 Constellations
250 - 265.
60
Norval (n 58 above) 258.
61 Norval (n 58 above) 255.
70 Albutt v CSVR
a multiplicity of pasts’.62 Modern iconoclasm and the

deinstitutionalisation of memory therefore results in a shift away
from archival and monumental sites, to new civic spaces where
individuals and groups can come together to debate and negotiate the
past and, through this process, define their future. Post-national
commemoration is characterised by a move away from traditional
monuments and memory-places, precisely because monuments tend
to become the opposite of what they set out to be. Instead of
facilitating participation and contestation, they reduce us to passive
spectators and consumers of national myths.
One of the most striking examples of this shift is presented by the

counter-monument movement in Germany during the late 1980s.
Norval refers in this regard to the memorial against fascism designed
by Jochen and Esther Gerz in Hamburg. The memorial denies or
subverts its own status as monument or memory place and its ability
to stand in for or represent the otherwise absent public memory of
citizens. It does so by gradually sinking into the ground. An inscription
at the top of the disappearing column reads as follows:63
One day it will have disappeared completely, and the site of the
Hamburg monument against fascism will be empty. In the end, it is only
we ourselves who can rise up against injustice.
Given this typology of memorial cultures and forms of

commemoration, Norval's central claim is that the TRC must be seen
as a post-national form of memory making and thus as an opening onto
a post-national political identity. Norval claims that the TRC shares
many of the features of the counter-monument in Hamburg, and
should be understood as an institutional counter-monument against
apartheid. Three features provide the key to its post-national status.
The first is the nature of the transition and the complex layering (as
opposed to the radical break) that it implies. The second is the central
role that was given in the process to participation of ordinary citizens.
The third is the fact that the existence of the TRC is limited by
statute. This prevents it from becoming a permanent fixture of the
institutional landscape. Like the Hamburg counter-monument, it will
disappear and leave citizens with the responsibility to continue the
memorial process it would have started.
62
Norval (n 58 above) 256. In this sense, the Constitutional Court building stands
fully in the sign of this post-national iconoclasm. The design deliberately
undermines all inherited iconographies of justice and refuses its status as an icon
or monument to the modern post-apartheid state. I previously described this
feature of the design as a shift from the iconic acropolis on the hill (the Union
Building in Pretoria, for example) to the inner city metropolis (the Constitutional
Court building in Braamfontein). In this sense the design calls provocatively for an
iconoclastic jurisprudence that seeks to desacralise and democratise legal
culture. See Le Roux (n 55 above) 139.
As the key to an alternative interpretation of participatory

democracy in Froneman's Albutt judgment, the second feature
identified by Norval, victim participation, is of most significance.
Norval distinguished in this regard between the participation in the
TRC process by political parties, on the one hand, and ordinary
citizens as perpetrators and victims, on the other.64 As she reads the
submissions by the political parties, they all served to provide the
historical context or grand narratives which could explain, and so
justify, the actions that were perpetrated in their names.65 According
to Norval, this kind of historiography belied the complexities and
ambiguities that marked the transition process. The transition was
simply not a linear process with a clear break between the past and
the present that allowed for this kind of re-monumentalisation of the
present. In the submissions by the perpetrators and victims, by
contrast, the focus fell on the everydayness of injustice and the
reoccupation of memory sites by ordinary citizens. If apartheid was
characterised by an identitarian mythology of the past, in which
competing histories were at best kept apartheid and at worst violently
repressed, a truly post-apartheid society would begin by
deconstructing this identitarian logic in memory making and politics
alike.
Norval suggest that the memory work of the TRC, precisely

because of its participatory character, is characterised by a double
signification. It is a remembrance as such, which already reminds us
about the incompleteness of the present given the injustices of the
past, and a remembrance of the logic of closure, or of a different way
of remembering.66 It is in this sense that Norval links the TRC to the
aesthetics of the counter-monument. Counter-monuments also have
a dual signification. They are memorials and thus call to
remembrance, but in the same process also marks the impossibility of
achieving finality or monumentality in doing so (of becoming a
monument). As a result of victim participation, the work of the TRC
did not result in a new master narrative of the past (as the political
parties desired), nor in a simple compilation of plural pasts, each with
their own set of heroes and victims, in the spirit of toleration and
national reconciliation. The TRC pointed to a fundamental
impossibility of completion as such:67
In recognising that identity relies upon traces of the not-now, not-here,

it opens identity out onto a beyond which is to be post-apartheid and
post-national ... The TRC potentially celebrates and commemorates not
completion and national myths of origin in their full splendor, but the

66
Norval (n 58 above) 259.
72 Albutt v CSVR
impossibility of identity, of the purity of origins, and also of

reconciliation.
Victim participation in the TRC process thus inaugurated a kind of

memory work in which the past and identity is continually reworked
as a palimpsest of traces.68 The palimpsest is a kind of double
signification. It is a writing that immediately draws attention to its
own future erasure. Could what Norval says about victim participation
and the memory work of the TRC also be applied to citizen
participation in the PRG and other democratic institutions?
At the juncture in our constitutional history marked by Albutt (a

resurfacing of the centrality of Parliament and Party as
representative institutions, as reflected in the displacement of the
TRC process to the PRG process), it is perhaps not enough merely to
remind us off the centrality of participatory democracy and to call for
a renewal or intensification of our constitutional patriotism (as
alternative to the more populist patriotisms of party and patriarch
that have reasserted themselves), but to also start exploring counter-
constitutional moments or better, to understand Froneman J call for
participatory democracy a mode of being politically that is counter-
democratic or counter-constitutional in the sense that Norval would
understand the term. Participation in a participatory democracy, and
the ‘participatory living’ celebrated by Froneman J in Albutt, would
then also assume a dual signification. It would not simply mean living
under different, democratically vindicated laws, but more
importantly, living differently under the law.
68
The palimpsest is a key memorial metaphor in memory work of the District Six
Museum. The first exhibition of the museum, ‘Streets: Retracing District Six’
opened on 10 December 1994 and remains the central memorial device of the
museum. It consists of two related elements: a series of old street signs from the
area, and a large laminated map which covers the central floor space in the old
church. Both elements represent an urban neighbourhood that no longer exists.
However, as the plaque at the entrance of the museum says, in remembering the
exhibitions ‘do not want to recreate District Six but to work with its memory’.
The map on the museum floor is an interactive space where ex-residents are
invited to write their names, return their houses to the map, and add comments,
poems, descriptions and stories. Empty space is thus reanimated and community
invoked there is no linear narrative or prescribed order or completed exhibition.
The museum is permanently in process and exclusively on the written comments
of ex-residents and the fragmentary remains of their possessions. In this process,
even the map of District Six, a so-called objective representation of the urban
environment, becomes a palimpsest. In this sense the installation participates in
the counter-monumental tradition of memorial sites to which Norval refers.
CUSTOMARY (COMMUNAL) LAND TENURE
IN SOUTH AFRICA: DID TONGOANE
OVERLOOK OR AVOID THE CORE ISSUE?
Douglas Mailula*
I conceive that land belongs to a vast family of which many are dead,
few are living and countless members are unborn**
1 Introduction
In a developmental African state such as South Africa, judges face

land issues which are sensitive, complicated, challenging, and
controversial.1 When adjudicating upon these issues, the judiciary
should courageously face the challenges head-on and deal with the
issues decisively as and when they arise, in order to, among others,
play their transformative role, and also ensure legal certainty. This
task is particularly important in a constitutional state, with a
transformative mandate, like South Africa. However, in doing so, the
judiciary must be sensitive to the unique and complicated character
of land holding, the controversies around it, and the sensitivities of
the land holders.
The sensitivity of land issues arises from the fact that African
communities share special, intimate, and intricate relationships with
* B proc, LLB, LLM (UNISA), Senior Lecturer, Department of Public, Constitutional

and International Law, Unisa. I would like to thank Danie Brand, Stu Woolman, &
Theunis Roux for inviting me to make a contribution to this publication.
** A Nigerian chief's submission to the West African Land Commission in 1912.
1
E Johnson ‘Communal land and tenure security: analysis of the South African
Communal Land Rights Act 11 of 2004’ unpublished LLM dissertation, Stellenbosch
University, 2009 15; B Cousins ‘Contextualising the controversy: dilemmas of
communal tenure reform on post-apartheid South Africa’ in A Claassens & B
Cousins (eds) Land, power and custom: controversies generated by South Africa’s
Communal Land Rights Act (2008) 3; B Cousins ‘Legislating negotiability: tenure
reform in post apartheid South Africa’ in K Juul & C Lund (eds) Negotiating
property in Africa (2002) 67.
73
74 Customary (communal) land tenure in South Africa
their land. To Africans, land is more than a mere asset of economic

value. It is not merely ‘a material and productive resource that
enables survival, livelihoods, and agricultural production. It is also an
important symbolic resource that heavily influences status, rites of
passage, and identity.2
Vermu explains that ‘land is deeply laden with cultural and

spiritual meanings that are context and culturally specific’.3
Therefore, in Africa, land embodies significant cultural and spiritual
values, or what Dannenmaier refers to as a ‘unique or distinctive
connection to the land with deep social, cultural, and spiritual
meaning.4 The African Union also indicates that, ‘land [in Africa] is
regarded not simply as an economic or environmental asset, but also
as a social, cultural, and ontological resource ... embodied in the
spirituality of society’.5 This view is shared by Anaya. He indicates
that, ‘indigenous [African] peoples’ rights over land and natural
resources flow not only from possession, but also from their
articulated ideas of communal stewardship over land and a deeply felt
spiritual and emotional nexus with the earth and its fruits’.6
Stavenhagen attests to this unique and significant cultural and
spiritual connection between Africans and their land. He testifies
that
the relationship between indigenous [African] peoples and the land is an

essential tie which provides and maintains the cultural identity of those
peoples. One must understand that the land is not a mere instrument of
agricultural production, but part of a geographic and social, symbolic
and religious space, with which the history and current dynamics of
those peoples are linked.7
Nkosi takes this spiritual and cultural nexus to land even further. He
explains that:
[i]n many African families the umbilical cord of a new born baby is
buried. In other communities when a boy is circumcised, the foreskin
2
R Vermu ‘“Without land you are nobody”: critical dimensions of women’s access
to land and relations in tenure in East Africa’ IDRC Scoping study for East Africa
on women’s access and rights to land and gender relations in tenure 2007 http://
www.landcoalition.org/program/wa_programmes_afr_4.html (accessed 15
November 2011).
3
n 2 above, 55.
4
Dannenmaier refers to this as the ‘distinctive connection doctrine’.
E Dannenmaier ‘Beyond indigenous property rights: exploring the emergence of a
distinctive connection doctrine’ (2008) 86(53) Washington University Law Review
101.
5 African Union, 2009 ‘Framework and guidelines on land policy in Africa’ (Addis
Ababa: African Union) adopted by the Thirteenth Ordinary Sessions of the African
Union. Sirte, Libya 1 - 3 July 2009, Assembly/AU/DEC.1(XIII).
6 SJ Anaya’s Indigenous peoples in international law (2004) 141.
7
R Stavenhagen Social classes in agrarian societies (1975) 83, as quoted in
Dannenmaier (n 4 above) 90.
and blood is also buried. The sacredness of land in Africa is further

linked to the fact that our ancestors are buried in it. Without land, we
[Africans] would not have a home for a dead body. That is why we kneel
barefooted next to the grave when we want to communicate anything to
our ancestors, showing a lot of respect for the land on which they lie.
When death strikes in a family, no one is allowed to till the land. We
mourn until that person is buried. After a funeral, in some cultures, we
do not touch the soil with a hoe, do not plough or till the land until a
ritual of cleansing the family is performed. Some communities like the
AmaZulu, do not till the land for a year when a member of a royal family
has passed away. The Zulu tribe believes that the elders and young men
must go to hunt so that a sacrifice can be made to the ancestors before
the land where a leader is to be buried is touched.8
This is also echoed by Lilongula. She highlights that before the

Korekore people of Zimbabwe touch their land, they go to the spirits,
which are said to be linked to certain animals or trees.9 Similarly, an
indigenous ‘declaration’ to the 2002 World Water Forum in Kyoto
proclaimed that indigenous African people reaffirm their relationship
to ‘Mother Earth’, on which they are placed in a sacred manner.10
Their relationship with their traditional lands and territories is
proclaimed to be the fundamental physical, cultural, and spiritual
basis for their existence.11
It is, indeed, this unique and special cultural and spiritual

connection between Africans and their lands that poses serious
challenges and complexities when it comes to the regulation of land
issues, including land holding and security of tenure. In Africa, these
issues are regulated under a complex system of customary law. Under
this system, Africans do not hold land simply as individuals.12 The
concept of individual land ownership is, therefore, generally foreign
to Africans. Shipton and Coheen highlight that ‘personal land claims
always depend on broader social entities, or combinations of them:
whether on extended homestead families, lineages, villages,
chiefdoms, ethnic sections, or other groups or networks’.13 In African
societies, land is generally held communally or collectively.14 It
8 Z Nkosi ‘Spirituality, land and land reform in South Africa’ http://www.wcc-

coe.org/wcc/what/jpc/echoes-16-05.html (accessed 01 December 2011) 1.
9
R Lilongula, Statement in Voices of the earth: cultural and spiritual values of
biodiversity (2000) http://www.unep.org/OurPlanet/imgversn/105/voices.html,
quoted in Dannenmaeir (n 5 above) 88.
10
Third World Water Forum, Kyoto, Japan, March 2003, Indigenous Peoples Kyoto
Water Declaration 1 http://www.indigenouswater.org/user/IPKyotoWaterDeclara
tionFINAL.Pdf, quoted by Dannenmaeir (n 4 above) 88.
11
n 10 above, 88.
12 P Shipton & M Goheen ‘Understanding African land-holding: power, wealth, and
meaning’ (1992) 62(3) Journal of the International African Institute 307 308.
13 n 12 above, 307.
14 DD Ndima ‘Judicial review and the transformation of the South African
jurisprudence with specific reference to African customary law’ (2007) 21(1)
Speculum Juris 84.
belongs to some group, be it a household, chiefdom, community, or

even a combination of these groups, whether the living, the living-
dead15 (ancestors), or the unborn. Awuah-Nyamekye and Sarfo-
Mensah explain that ‘[land] ownership is often tied to the living, the
dead and unborn. In the various tribal societies that constitute the
traditional areas, the living with the chief as trustee is said to be
holding the land in care for their ancestors and the unborn’.16 It is this
complex system17 that poses challenges to modern land law reform.
These challenges are varied. However, relevant for purpose of this

article, is the challenge identified by Cousins and Claassens, namely,
‘how to recognise and secure land rights that are clearly distinct from
“Western legal” forms of private property but are not simply
“customary”, given the impacts of both colonial policies and of past
and current processes of rapid social change’.18 According to du
Plessis,19 the problem is how to recognise and secure tenure rights.20
Should it be secured in the private ownership paradigm, or should
15
According to Gehman ‘the phrase “the living-dead” expresses the living
relationship between the living and their dead ancestors: for the ancestral spirits
are the “living dead” and not the “dead ancestors”. The living dead for five
generations enjoys a state of “personal immortality”, while they are being
personally remembered by first name by their living descendants “their process
of dying is not yet complete”. They are partly spiritual and partly human with one
foot in the spirit world and one foot in the world of the living. When the last
person dies who could remember them, however, the living dead have completed
the dying process and move into the past (zamani) as impersonal spirits ...’
RJ Gehman African traditional religion in biblical perspective (2005) 217.
16
S Awuah-Nyamekye & P Sarfo-Mensah ‘Ensuring equitable distribution of land in
Ghana: spirituality or policy? A case study from the forest-savanna agro-ecological
zone of Ghana’ (2011) 10(21) The International Indigenous Policy Journal 1; See
also Okoth-Ogendo 2005 www.plaas.org.za (accessed 20 November 2011).
17 H Mostert ‘Tenure security reform and electronic registration: exploring insights
from English law’ (2011) 14(3) PER/PELJ 85; M Adams & S Tenure ‘Legal dualism
and land policy in Eastern and Southern Africa’ UNDP-International Land Coalition
Land rights for African Development: From Knowledge to Action Nairobi, October
31 - November 3 conference proceedings www.undp/drylands (accessed
30 January 2012). See also DW Nabudere ‘Cheikh Anta Diop: The social sciences,
humanities, physical and natural sciences and transdisciplinarity’ (2007) 2(1)
International Journal of African Renaissance Studies — Multi-, Inter and
Transdisciplinarity 6 18; Nisbert Taringa ‘How environmental is African traditional
religion? http://enviro.lclark.edu/resources/EastAfrica/Hadzabe/How.pdf43
(accessed 02 May 2012); Workineh Kelbessa Traditional Oromo attitudes towards
the environment 25 - 26 http://www.ossrea.net/publications/images/stories/
ossrea/ssrr-19-p-3.pdf (accessed 02 05 2012).
18
B Cousins & A Claassens ‘More than simply “socially embedded”: recognising the
distinctiveness of African land rights’ in Claasens & Cousins (n 1 above) 3.
According to them, this is always the ‘central question’ on land issues in Africa.
19
WJ du Plessis ‘African indigenous land rights in a private ownership paradigm’
(2011) (14)7 PER/PELJ 46.
20 Land tenure refers to ‘the terms and conditions on which land is held, used and
transacted’. In simple terms, it means how the right to land is obtained and
distributed in a particular society or legal system. Land tenure reform, on the
other hand, refers to changes or improvement of these terms and conditions,
such as the amendment of the terms of contracts between land owners and
tenants, or the conversion of more informal tenancy into formal property rights.
(M Adams et al ‘Land tenure reform and rural livelihoods in Southern Africa’
(1999) 39 Natural Resources Perspectives 1; D Mzumara ‘Land tenure systems and
indigenous forms of land tenure be fully recognised and thus

protected?21 It is submitted that this is exactly the main challenge
that the Constitutional Court was confronted with in Tongoane,22 and
as du Plessis correctly indicates, despite the Tongoane judgment, the
problem as to secure indigenous land rights remains unresolved.23
In this note I do not purport to provide an answer to this difficult

question. Rather it is I seek to highlight that Tongoane deliberately
avoided this core issue which was raised by the applicants, namely,
that their use and occupation of the lands under dispute, as regulated
by customary law,24 is being threatened by the introduction of
CLARA.25 I will, therefore, critique the Court’s avoidance of this core
or substantive issue, and then also provide a very brief comment on
the procedural issues raised. As a point of departure, I provide a very
brief contextual historical background to the complex land issues in
South Africa, including security of communal land tenure.26 This is
followed by a discussion of the Tongoane case, after which a critical
comment is made. In this critical comment, I start with the procedural
issues raised in Tongoane, namely, the issue of proper tagging; and
the issue of Parliament’s failure to facilitate public involvement in its
law making process. This is followed by a critique of the Court’s
decision not to deal with the substantive issue, namely whether the
introduction of CLARA threatens security of communal land tenure as
provided for under customary law. As indicated earlier, it is not the
purpose of this note to provide an answer to this difficult question or
to attempt to resolve this complex matter. Instead the article is
essentially focussed around two issues (a) the Court’s focus on
tagging, a procedural matter, as opposed to the substantial issue of
tenure reform, and (b) the Courts’ non-development of the customary
law.
20
sustainable development in Southern Africa’ ECA/SA/EGM Land (2003) 2), as was
the case with CLARA, as the applicants argued in Tongoane. A fundamental goal of
land tenure is to enhance, secure and protect the people’s land rights against
arbitrary evictions, expropriation and landlessness in general (Awuah-Nyamekye
& Sarfo-Mensah (n 16 above) 6). This also serves to ensure a sustainability usage
of land and a complete peace of mind to rights holders who make considerable
investment in the land in question.
21
Du Plessis (n 19 above) 46.
22 Tongoane v National Minister for Agriculture and Land Affairs 2010 6 SA 214 (CC)
(Tongoane) para 33.
23
n 19 above, 46.
24 Tongoane (n 22 above) paras 31 - 33.
25
The Communal Land Rights Act 11 of 2004.
26
Here, I do not intent to go into a comprehensive historical narration but merely to
give a concise contextual background to put the issues in their proper context.
This is covered comprehensively in Tongoane (n 22 above) para 9 - 27. As will be
demonstrated later, it is quite bizarre that the Court spend such a substantial
time investing in a historical narration of the consecutive historical land tenure
system, which later becomes useless as the Court does not ultimately address the
issue.
2 Contextual background
The genesis of the challenges relating to communal security of land

tenure in South Africa can be traced to its initial colonisation by
Britain, and to its own subsequent apartheid regime, both of which
resulted in communal land dispossession.27 Land dispossession was, in
fact, the actual linchpin of the apartheid policy a few decades ago.28
Lahiff captures the essence of land dispossession in South Africa when
he says that:29
[t]he extent of land dispossession of the indigenous population in South

Africa, by Dutch and British settlers, was greater than any other country
in Africa, and persisted for an exceptionally long time. European
settlement began around the Cape of Good Hope in the 1650s and
progressed northwards and eastwards over a period of three hundred
years. By the twentieth century, most of the country, including most of
the best agricultural land, was reserved for the minority white settler
population, with the African majority confined to just 13% of the
territory, the ‘native reserves’, later known as African Homelands or
Bantustans ...30
For centuries prior to colonisation and apartheid in South Africa, the

majority of land rights in African societies, including land tenure, was
27
B Cousins ‘More than socially embedded: the distinctive character of ‘communal
tenure’ regimes in South Africa and its implications for land policy’ (2007) 7(3)
Journal of Agrarian Change 281 283; Du Plessis (n 19 above) 51; Dannenmaeir (n 4
above) 72; Shipton & Goheen (n 12 above) 317.
28 Since 1913 the notorious apartheid government enacted various pieces of
legislation that generally had the cumulative effect of dispossessing the black
majority of their land and putting their security of tenure over the remnants of
such lands, or the reserves, in a precarious position. The principal legislative
instruments of land dispossession included, among others, the Native Land Act 27
of 1913, and the Native Trust and Land Act 18 of 1936, both of which restricted
the African population to 13% of the total land area of South Africa; the Group
Areas Act 41 of 1950, which allocated certain areas to specific race groups; the
Natives Laws Amendment Act 54 of 1937, which served to prohibit Africans from
buying land in urban areas; the Bantu Authorities Act 68 of 1951, which allowed
the establishment of tribal, regional, and territorial authorities; the Prevention of
Illegal Squatting Act 52 of 1951, which allowed the government to establish
resettlement camps for surplus people evicted from white farms; the Blacks
Resettlement Act 2 of 1954, to give the state the authority to remove Africans
from any area in the magisterial district of Johannesburg and adjacent areas; the
Promotion of Bantu Self-Government Act 46 of 1959, to establish the Bantustans
and make the reserves the political homeland of black South Africans.
29 The history of land dispossession in South Africa is well documented (and,
therefore, beyond the scope of this note). See for instance, MA Yanou
Dispossession and access to land in South Africa: an African perspective (2009);
MC Lee Unfinished business in the Southern African Development Community:
the land crises in South Africa (2003).
30 E Lahiff ‘Land reform and poverty in South Africa’ Research Report 36 of the
Institute for Poverty, Land and Agrarian studies (PLAAS) at University of the
Western Cape 18 June 2007, 1 - 2, available at http://www.plaas.org.za/pubs/rr/
PLAAS_RR36_Lahiff.pdf/ (accessed 02 May 2012).
managed and controlled under a system of customary law.31 As

Cronkleton and others state, the
customary tenure systems by definition have evolved over long periods

of time in response to local specific conditions. And in the process of
recognition, such customary systems have been ignored, subordinated
or, at times, effectively accommodated. The scholarly debate on
whether to accept one legal system over the others, or what their
respective weights should be, continues. There is a call for a paradigm
shift from legal pluralism, which recognizes parallel systems to legal
integration which would mesh them. Integration would require
understanding of the major constituents of each other. 32
According to Cousins, pre-colonial land tenure was both ‘communal’

and ‘individual’, and could be seen as ‘a system of complementary
interests held simultaneously’.33 This meant that different interests
in the same property could vest in different holders.34 These
complementary interests were, (and still are) also dynamic and ‘ever
changing’.35
This system was (and still is) generally referred to either as

‘customary tenure,’36 ‘customary land tenure’,37 or ‘communal land
tenure’,38 as it is an important component of customary law or
indigenous law.39 Customary tenure refers to a ‘set of rules and norms
that govern community allocation, use, access, and transfer of land
and other natural resources’.40 As Freudenberger indicates,
the term “customary tenure” invokes the idea of “traditional” rights to

land and other natural resources — the tenure usually associated with
31
TM Mends & J de Meijere ‘A study of the institution of the customary land tenure
system in the supply of property rights for urban development — an example of
Accra, Ghana’ 3 http://www.fig.net/pub/accra/papers/ts03/ts03_01_mends
_demeijere.pdf (accessed 1 October 2010).
32 P Cronkleton et al ‘The devolution of management rights and co-management of
community forests’ in AM Larson et al (eds) Forests for people: community rights
and forest tenure reform (2010) 70; quoted, with approval, by Awuah-Nyamekye
& Sarfo-Mensah (n 16 above) 12.
33
B Cousins ‘Characterising “communal” tenure: nested systems and flexible
boundaries’ in Claassens & Cousins (n 1 above) 111; Du Plessis (n 19 above) 49;
TW Bennett Customary law in South Africa (2007) 381.
34
35
36 Adams et al (n 17 above).
37
J Pottier ‘Customary land tenure in sub-Saharan Africa today: meanings and
contexts’ in C Huggins & J Clover (eds) From the ground up: land rights, conflict
and peace in sub-Saharan Africa (2005) 55 - 76; Cousins & Claassens (n 18 above)
4.
38
Cousins & Claassens (n 18 above) 4; C Boone ‘Property and constitutional order:
land tenure reform and the future of the African state’ (2007) African Affairs
576; Shipton & Coheen (n 12 above) 311.
39 In this article the phrases ‘customary law’ and ‘indigenous law’ are used
interchangeably.
40
MS Freudenberger ‘The future of customary tenure’ United States Agency
International Development (USAID) Issue Paper (2011).
indigenous communities and administered in accordance with their

customs, as opposed to statutory tenure usually introduced during the
colonial period.41
As du Plessis puts it:42
African indigenous law in property was more concerned with relationship

status and people’s obligation towards one another in respect of the
property rather than the rights to people’s ownership of the property.
She explains that the relationships between people were more

important than an individual's ability to assert his or her interest in
property against the world.43 ‘Entitlements to property were more in
the form of obligations resulting from family relationships rather than
a means to exclude people from the use of certain property’.44
In African customary law, although a measure of individual control

over the broad interests that were embedded in land is recognised,
the paramount title to land is ‘perceived as vested above society and
whatever rights any one person had to the land were subordinate to
the entire community’s rights’.45 This is what Allot refers to as the
institution of ‘paramount control of land by “tribes”, “village
communities”, and other territorial groupings, “family land” or “clan
land”, where the individual’s enjoyment of land may be fettered by
the superior rights of the social group to which he belongs’.46
This system was changed by colonial rule, which often tried to

retain a form of ‘communal’ land tenure that could suit its
interests.47 As a result, African communities’ security of land tenure
has, since colonial times,48 remained precarious and insecure. This
41
n 40 above.
42 n 19 above, 49.
43
As above.
44
As above.
45 P Kameri-Mbote ‘The land has its owners! Gender issues in land tenure under
customary law in Kenya’ International Environmental Law Research Centre
(IELRC) Working Paper No. 9 of 2005, http://www.ielrc.org/content/w0509.pdf
(accessed on 02 November 2011).
46
AN Allot ‘Towards a definition of “absolute ownership”’ (1961) 5(2) Journal of
African Law 99.
47 Claassens & Cousins (n 33 above) 111.
48
The history of land dispossession under African colonisation in general, and
colonisation and apartheid in South Africa in particular, is well documented. See
for instance, in the broader African context, Pottier (n 37 above); Mends & de
Meijere (n 31 above); and in the South African context, L Ntsebeza & R Hall (eds)
The land question in South Africa: the challenges of transformation and
redistribution (2007); WD Thwala ‘Land and agrarian reform in South Africa’, in
P Rosset et al Promised land: competing visions of agrarian reform (2006) 57;
Awuah-Nyamekye & Sarfo-Mensah (n 16 above); Du Plessis (n 19 above).
was exacerbated by gradual land dispossessions,49 or what Adams and

others refer to as ‘enforced land alienation at the hands of
Europeans’.50 As Cousins puts it,
[t]his history [of land dispossession] has involved major modification and
adaptation of indigenous land regimes, but seldom their complete
destruction and replacement. Conquest and settlement in the colonial
period, followed by twentieth-century policies of segregation and
apartheid, saw white settlers and their heirs take possession of most of
the land surface of South Africa. State policies attempted to reconfigure
the livelihood and land tenure systems of the indigenous populations in
ways that served the interests of the dominant classes. African
“reserves” were created as a way to contain resistance and to facilitate
the supply of cheap labour for the emerging capitalist economy. They
also functioned to lower the cost of colonial administration through a
system of indirect rule, within which traditional leaders undertook local
administration on behalf of the state — often in a highly authoritarian
manner, termed “decentralized despotism” by Mamdani (1996).51
This threat to communal security of land tenure still remains to date.

This fact was acknowledged in Tongoane, as follows:52
[W]hat emerges ... from [the Bantu Areas Land Regulation] is that (a)
the tenure in land which was subject to the provisions of the Black Land
Act and the development Trust and Land Act and which was held by
African people was precarious and legally insecure; (b) indigenous law
governed succession to land in these areas ...
At paragraph 26, Ngcobo CJ continued as follows:
The Bantu Homelands Citizenship Act, 1970 and the Bantu Homelands
Constitution Act, 1971 further entrenched land dispossession as a key
policy of the apartheid edifice. African people would, as a consequence,
have no claim to any land in “white” South Africa ... They had
precarious title to the land they occupied to remind them of the
impermanence of their residence in “white” South Africa.
Ngcobo CJ concluded that African people were relentlessly

dispossessed of their land and given legally insecure tenure over the
land they occupied.53
49
B Cousins ‘More than socially embedded: the distinctive character of ‘communal
tenure’ regimes in South Africa and its implications for land policy’ 7(3) Journal
of agrarian change 281 283. Tongoane (n 22 above) paras 9 - 33; R Hall ‘Land and
agrarian reform in South Africa’ a Status Report of the Programme for Land and
Agrarian Studies, University of the Western Cape (2004) 48.
50
Adams et al ‘Land tenure reform and rural livelihoods in Southern Africa’ 39
(1999) Natural resources perspective 3.
51 Cousins (n 27 above) 283.
52
Tongoane (n 22 above) para 21.
53 Tongoane (n 22 above) para 27.
Our Constitution has as one of its objects the reversal of this

history. It requires the restoration of land to people and communities
that were dispossessed of land by colonial and apartheid laws after 19
June 1913.54 It also requires that people and communities whose
tenure of land is legally insecure as a result of racially discriminatory
colonial and apartheid laws be provided with legally secure tenure or
comparable redress.55 CLARA was enacted to ‘provide for legal
security of tenure’.56 Ironically,57 it is for this very reason that the
constitutional validity of CLARA was challenged in Tongoane.
In this case, as Ngcobo CJ succinctly observed, there were

basically two broad categories of grounds/objections: (a) procedural
and (b) substantive.58 The procedural issue was twofold: (1) whether
CLARA was correctly tagged, and (2) whether Parliament complied
with its constitutional obligation to facilitate public participation in
its law-making process. The substantive issue, on the other hand,
involved whether the provisions of CLARA, instead of providing legally
secure tenure, actually undermines it. As he correctly identified:59
What lies at the heart of the confirmation proceedings is the question

whether CLARA undermines the security of tenure of the applicant
communities. The applicants submit that it does, and that for this
reason CLARA is inconsistent with section 25(6) read with section 25(9)
of the Constitution which requires Parliament to enact legislation to
provide for legally secure tenure or comparable redress ...
However, the Court opted to deal with the procedural issue only and
decided not to deal with the substantive issue. In this regard the
Court, per Ngcobo CJ, explained as follows:60
Once it is concluded that CLARA is unconstitutional in its entirety

because it was not enacted in accordance with the provisions of section
76, it seems to me that that is the end of the matter. Although the
anxiety of the applicants to finalise the matter in the light of the energy
and time they invested in it is understandable, there is nothing left for
this Court, as a court of final appeal, to consider.
As will be demonstrated later, it is submitted that the Court was

misguided in this regard.
54
Section 25(7) of the 1996 Constitution (the Constitution).
55 Section 25(6) of the Constitution.
56
Preamble to CLARA.
57
H Mostert ‘South Africa’s Communal Land Rights Act: a plea for restraint in
reform’ (2010) 54(2) Journal of African Law 301.
58
See also W du Plessis & JM Pienaar ‘The more things change the more they stay
the same: the story of communal land tenure in South Africa’ (2010) 16(1)
Fundamina 86.
59
3 The Tongoane judgment
The applicant communities submitted that, far from securing their

land tenure, CLARA actually undermines it and makes it more
insecure. Ngcobo CJ summarises this argument as follows:61
The communities are concerned that their indigenous law-based system

of land administration will be replaced by the new system that CLARA
envisages. They are concerned that this will have an impact on the
evolving indigenous law which has always regulated the use and
occupation of land they occupy. They are further concerned that their
land will now be subject to the control of traditional councils which, as
is apparent from the record, they consider to be incapable of
administering their land for the benefit of the community. All the
communities claim that the provisions of CLARA will undermine the
security of tenure they presently enjoy in their land, and those who own
the land fear that they will be divested of their ownership of the land.
While some of these claims are disputed by the government
respondents, what is not disputed is that the land occupied by the
communities is administered in accordance with indigenous law, and
that traditional leaders, in particular the tribal authorities, play a role in
the administration of communal land. There is some issue as to the
extent to which the role of traditional leaders and tribal authorities
accords with indigenous law.
On this basis, it was argued on the applicants’ behalf, that CLARA is

inconsistent with section 25(6) of the Constitution, and therefore
invalid.
3.1 The issues
The applicants sought confirmation of an order of the North Gauteng

High Court,62 which declared certain provisions of CLARA
unconstitutional for undermining the security of tenure of certain
communities63 in respect of their lands, in contravention of sections
25(6), read with section 25(9), of the Constitution. Read together,
these sections require Parliament to enact legislation to provide for
61
62 Tongoane v The National Minister of Agriculture and Land Affairs 2010 8 BCLR
838 (GNP).
63
These were four communities that occupied land to which the provisions of CLARA
would apply, namely the Kalkfontein community, which owns and occupies two
farms known as Kalkfontein B and C in the Mpumalanga province; the Makuleke
community which owns and occupies a piece of land known as the Pafuri Triangle
in the Limpopo province; the Makgobistad community, which allegedly
established rights in respect of land in the area known as Mayayane in the North
West province; and the Dixie community, which occupies and independently
control the farm known as Dixie 240 KU, in the Pilgrims Rest District in the
Limpopo province. In all cases, the land falls under a tribal authority’s
jurisdiction and the use and occupation thereof is regulated by indigenous law.
legally secure land tenure or comparable redress.64 CLARA was the

legislature’s response to this constitutional requirement.65 Ironically,
CLARA is challenged as contravening the very same constitutional
requirement it seeks to accomplish, namely ‘to provide for legal
security of tenure’.66
In addition, the applicants sought leave to appeal against the

same judgment, which dismissed their application to have CLARA
declared unconstitutional in its entirety, for Parliament’s failure to
enact it in accordance with the procedure prescribed by section 76,
rather than section 75, of the Constitution. Relying on the Liquor
Bill67 case, the applicants argued that, as CLARA affects the
provinces, it should have been tagged or classified as a section 76 Bill,
because its ‘provisions in substantial measure fall within a functional
area listed in Schedule 4’.68 They submitted that the provisions of
CLARA, in a substantial measure, deal with ‘indigenous and customary
law’, and ‘traditional leadership’, which are functional areas that are
listed in Schedule 4,69 that is, functional areas on which the national
and provincial legislatures have concurrent legislative competence.
Parliament, on the other hand, argued that the test for tagging a Bill
was the substance of the legislation which was referred to as the ‘pith
and substance’ test.70 As Ngcobo CJ indicated, the phrase, ‘pith and
substance’, is borrowed from other jurisdictions and refers to the
‘substance’, the ‘purpose and effect’, or the ‘subject-matter’ of
legislation and was developed by the Constitutional Court to
determine whether the National Assembly or provincial legislature has
the competence to legislate in a particular field.71 Based on this test,
Parliament argued that the ‘pith and substance’ of CLARA was land
tenure, and any provision of CLARA dealing with indigenous law or
traditional leadership is incidental to land tenure and, therefore,
irrelevant for tagging purposes.72
Lastly, the applicants lodged an application for direct access to

the Constitutional Court seeking an order declaring CLARA
64
Section 25(6) of the Constitution provides that ‘a person or community whose
tenure of land is legally insecure as a result of past racially discriminatory laws or
practices is entitled, to the extent provided by an Act of Parliament, either to
tenure which is legally secure or to comparable redress’; sec 25(9) provides that
‘Parliament must enact the legislation referred to in subsection (6)’.
65
Cousins & Claassens (n 18 above) 8; Cousins (n 27 above) 284.
66
The Preamble to CLARA, Tongoane (n 22 above) para 28.
67 Ex Parte President of the Republic of South Africa: In re constitutionality of the
Liquor Bill 2000 1 SA 732 (CC); 2000 1 BCLR 1 (CC).
68
Tongoane (n 22 above) para 49. Schedule 4 of the Constitution contains a list of
matters that fall within the concurrent legislative competence of the national
and provincial legislatures.
69 n 61 above.
71
Tongoane (n 22 above).
72 Tongoane (n 22 above) para 50 read with para 95.
unconstitutional on the ground that Parliament failed to comply with

its constitutional mandate to facilitate public involvement in its
legislative process in contravention of sections 59(1)(a) and 72(1)(a)
of the Constitution.73
3.2 The decision
On the issue of tagging, after reconsidering the Liquor Bill case, the
Court held that a distinction should be drawn between determining
whether the National Assembly or NCOP has the competence to
legislate in a particular field, that is, the ‘characterisation of a Bill’;
and determining how a Bill ought to be properly tagged and ultimately
enacted.74 It was held that these are two different processes for
which two different tests must be applied.75 According Ngcobo CJ,
[t]here is an important difference between the “pith and substance”

test and the “substantial measures” test. Under the former, provisions of
the legislation that fall outside of its substance are treated as
incidental. By contrast, the tagging test is distinct from the legislative
competence. It focuses on all the provisions of the Bill in order to
determine the extent to which they substantially affect functional areas
listed in Schedule 4 and not on whether any of its provisions are
incidental to its substance.76
Based on this important distinction, the Court upheld the applicants’

contention, thus reaffirming the test for the determination of the
procedure to be followed in enacting a Bill as developed in the Liquor
Bill case, namely that any Bill of which the provisions in a substantial
measure fall within a functional area listed in Schedule 4, must be
dealt with under section 76.77 It was held that while the main subject-
matter of a Bill, which is a key factor in determining legislative
competence, may not affect provinces, some of its provisions may,
nevertheless, have a substantial impact on the interests of
provinces.78 The Court held that the test for the tagging of Bills must
be informed by the need to ensure that the provinces exercise their
appropriate role, fully and effectively, in the process of considering
national legislation that substantially affects them.79 After analysing
73 Tongoane (n 22 above) paras 3, 43, 113, and 114. Both sections 59(1)(a) and
72(1)(a) of the Constitution provide for public participation in Parliament’s law
making process. Section 59(1)(a) obliges the National Assembly to facilitate
public involvement in its legislative and other processes including those of its
committees. Section 72(1)(a) makes exactly the same provision in respect of the
National Council of Provinces.
75
78
the provisions of CLARA, Ngcobo CJ held that the inescapable

conclusion is that various provisions of CLARA affect, in substantial
measure, indigenous law and traditional leadership, both being areas
of concurrent national and provincial competence.80 He found that
CLARA replaces the living indigenous law regime which regulates the
occupation, use, and administration of communal land.81 He further
found that it replaces both the institutions that regulated these
matters and their corresponding rules. He accordingly concluded that
Parliament followed an incorrect procedure in enacting CLARA.82
In considering the appropriate remedy, Ngcobo CJ held that where

the Constitution prescribes a legislative procedure, that procedure
must ordinarily be followed. Enacting legislation that affects the
provinces in accordance with the procedure prescribed in section 76
is a material part of the law-making process relating to legislation
that substantially affects the provinces. He held that failure to
comply with the requirements of section 76 renders the resulting
legislation invalid. He accordingly held CLARA to be unconstitutional
and invalid for want of compliance with the procedures set out in
section 76 of the Constitution.
4 Comment
The confirmation of the invalidity of CLARA in Tongoane did not come

as a surprise to many in South Africa, as CLARA clearly fails to meet,
among others, an important constitutional mandate to provide for
security of tenure over land. The controversial nature of CLARA was
well acknowledged and appreciated by some in the academic
fraternity83 and other sectors.84 For instance, commenting on the
provisions of CLARA, Cousins wrote that ‘the controversies over land
tenure reform in post-apartheid South Africa resonate strongly with
those raging elsewhere in Africa’.85 To me at least, what did come as
a surprise, as indicated earlier, is the Court’s decision to entertain
only one procedural issue and not to entertain the substantive issue
raised by the applicants at all. I will revert to this issue later.

81
Tongoane (n 22 above) paras 74 - 97.
82
Tongoane (n 22 above) paras 98 - 110.
83 See for instance, H Mostert ‘(n 57 above) 398 & 298; Cousins & Claassens (n 18
above) 1; AJ van der Walt Constitutional property law (2005) 334.
84 For instance, non-governmental organisations such as the Legal Resources Centre
and some communities whose security of land tenure was threatened as a result
of the introduction of this piece of legislation.
4.1 Tagging
The Tongoane Court should be commended for reaffirming the test for
tagging as the ‘substantial measure’ test. The Court made a very
important distinction for the test for tagging or classification,86 which
is a procedural matter, and the test for characterisation87 or
competency, which is a jurisdictional matter. In other words,
determining how to tag or classify legislation, for the purpose of
enactment, is a step in the process of law making and, therefore, a
procedural matter. Characterising a Bill, on the other hand, is a
jurisdictional issue, which involves determining whether a particular
legislative authority has legislative competence on a particular
matter. The test for the two can, therefore, not be the same. The
Court indicated the important difference between the two, namely
that in determining classification or tagging, what matters is whether
a particular Bill, in a substantial measure, falls within a functional
area listed in Schedule 4 of the Constitution; while in determining
competence, characterisation or jurisdiction, the test is ‘the subject
matter’ or the ‘pith and substance’88 of the particular Bill.
The whole idea of tagging a Bill correctly is to allow provincial

interests to be adequately considered and promoted in the national
legislative process. If a Bill, in a substantial measure, affects the
interest of the provinces, such a Bill must be tagged as a section 76
Bill, and the procedure provided for in section 76 should be followed
in passing that Bill. If it does not, it must be tagged as a section 75 Bill
and the less onerous procedure provided for in section 75 should be
followed. On the other hand, the idea behind legislative competence
is to ensure co-operation between different spheres of government
and to avoid or minimise potential legislative conflicts or disputes
between two different spheres of government, namely, the national
and the provincial spheres.89 However, as Murray and Simeon
indicate, the fact that many potential jurisdictional disputes are
‘avoided’ by the granting of concurrent powers to provinces and the
new national sphere does not avoid the need to classify or categorise
new national laws.90 In other words, both the classification and the
competency issues are important aspects that need to be considered
in any given case.
It is for this reason that I submit that although different, the

classification, categorisation or tagging of Bills and their

87
C Murray & R Simeon ‘“Tagging” Bills in Parliament: section 75 or section 76?’
(2006) 123 South African Law Journal 242.
88 Murray & Simeon (n 87 above) 245.
89
Murray & Simeon (n 87 above) 232.
90 Murray & Simeon (n 87 above) 233.
characterisation are inextricably entwined and mutually and

reciprocally dependent. In order to determine whether a particular
legislative authority has competence to legislate on a particular
matter, one needs to look at the Schedules, including Schedule 4.
Although the test for legislative competence is the determination of
the subject-matter or the ‘pith and substance’ of the particular Bill,
this cannot realistically be done without also looking at whether that
subject matter, in a substantial measure, affects the interests of the
provinces, so that the section 76 procedure should be adopted. In
other words, it is difficult to imagine a situation in which the subject
matter of a particular Bill is clearly within the concurrent legislative
competence of both legislatures, but the provisions of the particular
Bill do not in a ‘substantial measure’ fall within a functional area
listed in Schedule 4 of the Constitution. By the same token, it cannot
be practically possible to determine the subject matter of a particular
Bill without having to look at how it substantially affects the interests
of the province.
However, the judgment should be commended for making this

important distinction between the test for tagging and the test for
characterisation, and authoritatively setting the test. It should be
noted, however, that this test for tagging was actually pre-empted by
Murray and Simeon, prior to Tongoane. They suggested that
[t]he same argument, (that the substantial measure test, rather than
the “pith and substance” test should be used in tagging CLARA) applies
to the Communal Land Rights Act. As we note above, it was tagged to
follow the s 75 process because the parliamentary law advisers
concluded (probably correctly) that its pith and substance is “the
provision of legal security of tenure by transferring communal land to
communities, or by awarding comparable redress”. The law advisers
commented that substantive provisions in the bill that referred to
customary law did not “render the Bill a s 76 Bill” but, at most, were
“matters incidental to the ‘pith and substance’ of the Bill”. But the Act
has a direct impact on matters relating to traditional leadership and
customary law. Moreover, a Bill cannot escape being tagged s 76 simply
because its ends fall outside schedule 4.91
This was echoed by Claassens and Cousins, also commenting on

CLARA, as follows:92
… [T]he “pith and substance test” does not provide appropriate criteria
for tagging Bills. The question instead must be whether or not provisions
of the law in a substantial measure fall within a functional area listed in
schedule 4. It is clear that they do in this case [of CLARA].
91
Murray & Simeon (n 87 above) 254.
92 Claassens & Cousins (n 1 above) 81.
What is not clear from this test is what ‘substantial measure’

entails. It is submitted that the phrase ‘substantial measure’ is vague
and open to varied interpretations. Does it refer to the impact of the
individual provisions of the Bill looked at individually or does it refer
to the cumulative impact of the different provisions of the Bill? It is
submitted that ‘substantial measure’ refers to the extent or degree
to which a particular Bill deals with a matter of concurrent legislative
competence as listed in Schedule 4. To put it differently, it refers to
the extent of, or the degree of impact of the particular Bill on
provincial interests as listed in Schedule 4. The impact or effect of the
Bill on the interests of the provinces must be of substantial extent,
degree or measure for it to be tagged as a section 76 Bill. In my view,
what is important should be the cumulative or combined effect of the
Bill rather than how individual provisions individually affect a
particular concurrent matter. Although the individual provisions will
be considered, it is their cumulative or combined impact that is
ultimately determinative of whether a particular Bill in a substantial
measure falls within a concurrent legislative competency of the two
spheres of government.
In Tongoane, the Court spent a considerable amount of time and

energy on the issues of tagging for the first time. In fact, it is no
exaggeration to state that the case could mistakenly be perceived as
being solely about tagging. The other issues were pushed to the
periphery. However, it still remains uncertain as to what exactly this
loaded test of ‘substantial measure’ entails. Although it is conceded
that this issue will probably be answered authoritatively in a
subsequent case, it is submitted that the Court has missed a good
opportunity to provide clarity and certainty on this issue once and for
all.
4.2 Constitutional obligation to facilitate public

participation in legislative processes
Parliament’s obligation to facilitate public participation in its

legislative processes arises from sections 59(1)(a) and 72(1)(a) of the
Constitution.93 Read together, these sections require both houses of
Parliament to facilitate public involvement in the legislative and
other processes including processes in their respective committees.94
93
See also sec 118(1)(a) of the Constitution.
94 Section 59(1)(a) of the Constitution obliges the National Assembly to ‘facilitate
public involvement in the legislative and other processes of the Assembly and its
committees’. Section 72(1)(a) and 118(1)(a) echoes the same obligation to
facilitate public involvement in the legislative and other processes of the
Assembly and its committees in respect of the National Council of Provinces and
provincial legislatures, respectively.
This obligation is informed by our government system of

participatory democracy.95 Participatory democracy requires decision
makers, including the legislatures, to involve the people in all
decisions affecting them.96 This system is one of the values upon
which the new constitutional dispensation is based. The 1996
Constitution of South Africa is founded on numerous values, designed
to shape a particular type of society South Africa envisions.97 Public
participation in the law making process98 is a cornerstone of modern
democracy.99 It ensures that people are listened to during the law
making process and thus treated with dignity and respect. As
Czapanskiy and Manjoo state,
... in the interest of promoting human rights [including human dignity]

and democracy, the legislative duty to facilitate public participation is
an important one. Hence, Doctors for Life may provide valuable lessons
with respect to citizen participation in the law-making process, thereby
further promoting human rights values of, amongst others, dignity and
respect ...100
The idea of democracy is about respect for the people, it’s about
people themselves having a say in matters that affect their lives, and
it’s about self-determination or people determining what is best for
them. This is in accordance with both the classic and modern
advocates for participatory democracy, who hold that the more that
citizens are engaged in self-governance, the more they gain in self-
respect, autonomy and empathy for others.101 It should, therefore,
always be conceived and understood from a bottom-up perspective.
95 KS Czapanskiy & R Manjoo ‘The right of public participation in the law-making

process and the role of the legislature in the promotion of this right’ (2008) 19 (1)
Duke journal of comparative & international law 1.
96
A du Plessis ‘Public participation, good environmental governance and fulfillment
of environmental rights’ PER/PELJ 11(2) (2008) 174.
97 Section 1 of the Constitution of the Republic of South Africa. These values include
freedom, equality and human dignity.
98
See generally Czapanskiy & Manjoo (n 95 above); R Scott ‘An analysis of public
participation in the South African legislative sector’ unpublished MA dissertation,
Stellenbosch University, 2009; C Evans & S Evans ‘Evaluating the human rights
performance of legislatures’, (2006) 6 Human Rights Law Review 545 548; I
Buccus ‘Towards developing a public participation strategy for South Africa’s
provincial legislatures’ Special Focus: Centre for Public Participation (2006); L
Nyati ‘Public participation: what has the Constitutional Court given the public?’
(2008) 12(2) Law, Democracy & Development 102; T Karl & QC Hudson-Phillips ‘A
case for greater public participation in the legislative process’ http://
slr.oxfordjournals.org/content/8/2/76.full.pdf (accessed 04 May 2012); T Roux
‘Democracy’ in S Woolman et al (eds) Constitutional Law of South Africa (2006
2nd edition part II) 10 - 18.
99
See generally for the importance of and the reasons for public participation D
Brynard Public participation in local government administration: bridging the
gap (2009); SX Hanekom Public policy (1987); JE Anderson Public policymaking
(1990); S Langton Citizen participation in America (1978).
100 Czapanskiy & Manjoo (n 95 above) 3 - 4.
101
C Pateman Participation and democratic theory (1970) 22 - 44 cited by Czapankiy
& Manshoo (n 95 above) 15.
In other words, it should be about what people at the grassroots level

need and not what a particular authority wants for the people.
The issue of determining whether the legislature had complied

with the constitutional requirement of facilitating public
participation in the law making process was left open in Tongoane.
However, the judiciary has on numerous occasions dealt with this
issue in the past. For instance in Matatiele, Ngcobo J (as he then was)
held that
[o]ur constitution contemplates a democracy that is representative, and

that also contains elements of participatory democracy. As the Preamble
openly declares, what is contemplated is “a democratic and open
society in which government is based on the will of the people”.
Consistent with the constitutional order, section 118(1)(a) calls upon the
provincial legislatures to facilitate involvement in [their] legislative and
other processes’ including those of their committees. As was held in
Doctors for Life International v Speaker of the National Assembly and
Others (CCT 12/05), our Constitution calls for open and transparent
government and requires legislative organs to facilitate public
participation in the making of laws by all legislative organs of the
State.102
In addition to Matatiele and Tongoane, in 2010 alone the Court spent

quite some time on issues of participatory democracy, including the
right to be heard. In at least four of the 24 cases decided in that year,
the Court either directly or indirectly, dealt with the issue of
participatory democracy. These cases included Poverty Alleviation
Network,103 which dealt with the issue of facilitating public
participation in Parliament’s law making process; Tongoane, in which
a similar issue was raised but not dealt with; Albutt,104 which dealt
with the issue of victim participation in the consideration of
presidential pardons (the right of victims of crime to be heard before
the offenders are pardoned); and Bengwenyama,105 which dealt with
a local community’s right to be consulted on mining and prospecting
activity in respect of their land.
Although this matter was left open in Tongoane, I am of the view

that it was exhaustively and adequately dealt with in previous cases
including the 2010 cases highlighted above. I, therefore, express no
opinion on this issue.
102
Matatiele Municipality v President of the Republic of South Africa 2007 1 BCLR 47
(CC) (Matatiele) para 40.
103
Poverty Alleviation Network v President of the Republic of South Africa and
other Case 2010 6 BCLR 520 (CC).
105
Bengwenyama Minerals (Pty) Ltd v Genorah Resources (Pty) Ltd 2011 4 SA 113
(CC).
4.3 Avoiding the core/real issue: whether CLARA, rather

than providing security of tenure communal land,
actually threatens it.
As soon as the constitutional judge’s decision is issued, the debate

ceases.106
At first blush, the Tongoane judgment might seem very convincing and
solid. At the very beginning of this judgment, Ngcobo CJ
acknowledged and appreciated the significance of CLARA as
legislation intended to meet a very sensitive outstanding
constitutional imperative of providing legally secure tenure or
comparative redress to African communities whose tenure of land is
illegally insecure as a result the racist policies of apartheid, that were
imposed under the colour of the law.107
Ngcobo CJ emphasised the importance of this legislation by

outlining, from the outset, the important constitutional mandate
given to Parliament to address the issue of insecurity of communal
land rights. He did this by quoting in the very second paragraph,
section 25(6) of the Constitution, which requires Parliament to
provide persons and communities whose security of tenure is
insecure, with secure land tenure or comparable redress.108
Ngcobo CJ, ‘in order to put the issues presented into context, ...
consider[ed] it desirable to sketch briefly the legislative scheme
which brought about our colonial and apartheid geography and which
facilitated land dispossession of African people, the resultant
insecure land tenure for the majority of our country, and the history
of land occupation by the four applicant communities’.109 This
historical background is comprehensive, accurately narrated, and
well written.
However, one finds it difficult to understand the relevance of this

long historical narration. Despite spending so much time, energy, and
effort on identifying the substantive issue and providing a proper
contextual background, Ngcobo CJ regrettably failed to deal with it.
He failed to determine the real or the substantive issue that was
raised by the applicants, namely that the provisions of CLARA actually
undermine security of communal land tenure instead of protecting it,
106
L Favoreu ‘Actualité et légitimité du contrôle jurisdictionnel des lois en Europe
ocidentale’ (1984) Revue de troit public et del la science politique 1195 trans
E Rosenfeld (ed) Constitutionalism, identity, difference, and legitimacy;
theoretical perspectives (1994) 272.
108
as required by the Constitution. The historical narration given by the

CJ is, therefore, at best obiter and at worse irrelevant. This history
has very little (if anything) to do with ‘tagging’ which is the only issue
the Court ultimately considers. It is submitted that the Court took an
easy way out of a difficult issue.
After clearly setting out the issues for determination,110 Ngcobo

CJ indicated that ‘it is convenient to consider first (own emphasis)
whether Parliament followed the correct procedure in enacting
CLARA, or the classification or tagging question’.111 This creates the
impression that the judge will proceed to deal with the second and
then the third or last issue. This amounts to a reasonable expectation
that all the issues raised, or at least more than one of the issues, will
ultimately be considered. However, as one reads further, it is
discovered that in the Ngcobo ‘s opinion, ‘once it is concluded that
CLARA is unconstitutional in its entirety because it was not enacted in
accordance with the provisions of section 76, it seems ... that that is
the end of the matter’.112 As he continued, ‘although the anxiety of
the applicants to finalise the matter in the light of the energy and the
time invested in it is understandable [but not appreciated],113 there
is nothing left for this Court, as a Court of final appeal, to
consider’.114
He does not rely on any authority for this statement. Despite a

diligent search, I could not find any established legal principle to this
effect.115 It is submitted that this statement is derived from what
could at best be a rule of logic, and at worst a rule of convenience.
However, a rule of logic to this effect could only arise if the remaining
issues are irrelevant or unnecessary. In my view, the Chief justice
merely found it convenient not to deal with this difficult and,
perhaps, also political issue.
4.3.1 No reason why the Court failed to deal with the core issue
Although there are various reasons why a court would avoid dealing
with a particular issue raised by an applicant in any given case, I could
not find any reason for this in Tongoane. Normally such reasons could
include instances where the applicant does not have standing; where
the case is not ready for adjudication or when the matter is not ripe;
or cases where the matter does no longer require adjudication or the
110
112
113 Author’s insertion.
115
This does not appear in the rules of the Constitutional Court, the Uniform Rules,
any related rules or statute, or case law.
matter is moot.116 There was no dispute about the parties’ standing

and this matter was beyond doubt in Tongoane. As Vrancken and
Killander observe, the issue of standing and ripeness often overlap
because an applicant often lacks sufficient interest if the matter is
not ripe.117
The principle of ripeness requires that if, in any given case,
a constitutional issue can be dealt with more conveniently at a later

stage [rather than at the stage in which it is raised] and the applicant
will get no tangible advantage from an earlier ruling, ... the applicant
[must] wait until the court can ground its decision on concrete relief’.118
According to Hoexter, the idea behind ripeness is that a complainant
should not go to court before the offending action or decision ripe for
adjudication.119
The issue of threat to security of tenure in Tongoane, which the Court

avoided, was ripe. The issue of ripeness was not raised as an objection
by the respondents and consequently not considered by the Court in
Tongoane. For this reason, ripeness as a reason for not dealing with
threat to security of tenure cannot be sustained.
With regard to whether a matter is moot, the Constitutional Court

has stated that a matter is ‘moot and therefore not justiciable if it no
longer represents an existing or live controversy which should exist if
the court is to avoid giving advisory opinions on abstract propositions
in law’.120 For instance, in JT Publishing the Court refused to declare
that certain provisions of the Publications Act 42 of 1974 were
unconstitutional and invalid, because the Act, which was in force
when the action was instituted, was in the process of being repealed
when the Court heard the case.121 In contrast to JT Publishing, CLARA
was not in the process of being repealed when the matter was heard
in Tongoane and the matter was not moot. The court merely relied on
a statement of the Minister,122 to the effect that CLARA would be
repealed in toto, despite caution from the applicant’s counsel.123
Ncgobo CJ could ‘not see anything wrong with the Minister informing
[the] Court that CLARA, as it stands, is not consistent with
116 P Vrancken & M Killander ‘Human Rights Litigation’ in A Govindjee (ed) et al

Introduction to human right law (2009) 251 258 - 259.
117
n 116 above, 258 – 259.
118 Currie & de Waal Bill of rights handbook (2005) 92, referring to SAPT v Director:
Directorate for Organised Crime and Public Safety 2000 2 BCLR 200 (C) 2061; also
quoted by Vrancken & Killander (n 116 above) 258.
119 C Hoexter et al (eds) The new constitutional and administrative law (2003) 302.
120
National Coalition of Gay and Lesbian Equality v Minister of Home Affairs 2000 2
SA 1 (CC) (National Coalition of Gay and Lesbian Equality) para 21 fn 18.
121 JT Publishing v Minister of Safety and Security 1997 3 SA 514 (CC) para 16.
122
Minister for Agriculture and Land Affairs, who was one of the respondents.
government policy’.124 This was a surprising and disappointing

misdirection on the part of the Court. The Minister’s mere political
statement has no legal effect. The Court is also not accountable to the
Minister or to ‘government policy’. It is subject only to the
Constitution and the law which it must apply without fear, favour or
prejudice.125 This Constitution is the supreme law of the land,126
binding all organs of state including the executive and the legislature.
Government policy, like all law or conduct, is subject to the
supremacy of the Constitution. Furthermore, although the Minister
made such a political statement, he is not a law-maker. The Court
could not even rely on the statement if it was made by Parliament,
because Parliament itself is a deliberative organ which might or might
not decide to repeal CLARA in toto. Parliament follows democratic
processes and there is no guarantee that CLARA will be repealed in
toto.
In any event, the Court may, indeed, decide to deal with a matter
even if the matter has become moot in the course of the proceedings
with regard to the parties concerned, if it is in the interest of justice
to do so. For instance, in Pillay,127 despite that fact that the issue was
found to be moot at the time it reached the Constitutional Court,
because the party concerned had completed high school, the
Constitutional Court, nevertheless, granted leave to appeal because
it was in the interest of justice to hear the case as it raised
vital questions about the extent of protection afforded to cultural and

religious rights in the school setting and possibly beyond. The issues are
both important and complex, as is evidenced by the varying approaches
of the courts below as well as courts in foreign jurisdictions. Extensive
argument has been presented, not only from the parties but from three
amici curiae. There is accordingly no doubt that the order, if the matter
is heard, will have a significant practical effect on the School and all
other schools in the country, although it will have no direct impact on
Sunali [the pupil concerned]. It is therefore in the interests of justice to
grant leave to appeal.128
In Langeberg, Yacoob J held that
[e]ven though a matter may be moot as between the parties in the sense
defined by Ackermann J [in National Coalition for Gay and Lesbian
Equality n 120], that does not necessarily constitute an absolute bar to

125
Section 165 of the Constitution.
126
Section 2 of the Constitution.
127 MEC for Education: KwaZulu-Natal v Pillay 2008 2 BCLR 99 (CC).
128
Pillay (n 127 above) para 35. In this case the issue was whether prohibiting a high
school pupil from wearing a nose ring in accordance with her religious and
cultural beliefs violated the Bill of Rights, but before the matter reached the
Constitutional Court, the pupil had finished high school and the decision would
therefore not have any direct impact on her.
its justiciability. This Court has discretion whether or not to consider it.
Langa DP, in President, Ordinary Court-Martial and Others v Freedom of
Expression Institute and Others, throws some light on how such
discretion ought to be exercised. The conclusion in that judgment is that
section 172(2) of the Constitution does not oblige this Court to hear
proceedings concerning confirmation of orders of unconstitutionality of
legislative measures which have since been repealed but has a discretion
to do so and should consider whether any order it may make will have
any practical effect either on the parties or on others.129
As Yacoob indicated, in high courts, sitting as courts of appeal, and in

the Supreme Court of Appeal (SCA), the situation is governed by
section 21A of the Supreme Court Act 59 of 1959.130 A number of SCA
judgments131 have dealt with the exercise of discretion in terms of
this section. However, the Natal Rugby Union v Gould case132 is
illustrative of the application of this section. It concerned the
correctness of the procedure that had been followed in electing the
president of the Union. By the time the appeal was heard, the Union
had held a re-election in accordance with the procedure contended
for by the respondent. Howie JA held that
[h]ad there been no appeal the judgment of the Court below would in all
probability have continued to influence the procedure adopted in
respect of office bearer elections at future union meetings. There was,
of course, nothing irregular or unfair in the procedures adopted at the
re-election meeting, viewed purely in isolation, without regard to the
constitution. But the union does have this constitution. It is the chosen
instrument by which the union’s affairs are to be regulated and the
union, its office bearers and council members are entitled to have it
interpreted in order to guide them in the future. In the circumstances I
consider that determination of the appeal will, quite apart from the
issue of costs in the Court below, have a practical effect or result within
the meaning of s 21A of the Supreme Court Act.133
This is to be contrasted with the position in this Court where there is

no equivalent statutory provision. As Yacoob J indicated in
Langeberg:134
129 Independent Electoral Commission v Langeberg Municipality 2001 3 SA 925 (CC);

2001 9 BCLR 883 (CC) para 9 (Langeberg). In this case the respondent resisted an
appeal on the ground of mootness and contended that the elections have come
and gone and that the IEC has, by creating an additional voting district after the
judgment of the High Court had been given, removed the possibility of any live
dispute between the parties.
130 Langeberg (n 129 above) para 10.
131
McDonalds Corporation v Joburgers Drive-Inn Restaurant 1997 1 SA 1 (A) 14C;
Premier, Provinsie Mpumalanga v Groblersdalse Stadsraad 1998 2 SA 1136 (SCA)
1143A - B; Western Cape Education Department v George 1998 3 SA 77 (SCA);
84G; Simon NO v Air Operations of Europe AB 1999 1 SA 217 (SCA) 226I - 227A;
Natal Rugby Union v Gould 1999 1 SA 432 (SCA) 444I - 45B.
132 n 131 above.
133
Natal Rugby Union (n 131 above) para 45B.
[T]his Court has a discretion to decide issues on appeal even if they no

longer present existing or live controversies. That discretion must be
exercised according to what the interests of justice require. A
prerequisite for the exercise of the discretion is that any order which
this Court may make will have some practical effect either on the
parties or on others. Other factors that may be relevant will include the
nature and extent of the practical effect that any possible order might
have, the importance of the issue, its complexity, and the fullness or
otherwise of the argument advanced. This does not mean, however, that
once this Court has determined one moot issue arising in an appeal it is
obliged to determine all other moot issues.
It is submitted that the substantive issue raised by the applicants in

Tongoane was not moot at the time of hearing in Tongoane. Even if it
was found to be moot, which is not the case, the Court could have
exercised its discretion and dealt with this issue. The Court was,
therefore, misguided in relying on a mere statement of a cabinet
minister to whom it is not accountable.
4.3.2 The need for judicial activism in the context of

transformative constitutionalism
Diescho correctly indicates that
through judicial activism, judges influence the direction of the law. This
happens when their interpretation of the law goes beyond the mere
words of the texts at hand and beyond the matters mentioned: when
they interpret the law purposefully, and say the unsaid through their
interpretation.135
According to du Plessis ‘judicial activism is, therefore premised on the

belief that judges have a creative role to play in the interpretation
and application of enacted law’.136 As Dugard proposes, judges in a
post-apartheid and constitutional democratic South African society
should play an activist role or function.137 Justice O‘Regan also stated
in Fourie, that ‘the power and duty to protect constitutional rights is
conferred upon the courts and courts should not shrink from that
135
See JD Diescho ‘The paradigm of an independent judiciary: Its history,
implications and limitations in Africa’ http://www.kas.de/upload/auslands
homepages/namibia/Independence_Judiciary/diescho.pdf, (accessed 19 January
2012) 38; H Botha ‘Rights, limitations and the (im)possibility of self government’
in H Botha et al (eds) Rights and democracy in a transformative constitution
(2003) 14; RE Levy & RL Glicksman ‘Judicial activism and restraint in the Supreme
Court's environmental law decisions’ (1989) 42 Vanderbilt law review 343.
136 L du Plessis ‘Some of Frank Michelman’s prospects for constitutional
interpretation in South Africa — in retrospect’ in H Botha et al (n 135 above) 89;
C Hoexter Administrative law in South Africa (2007)).
137 J (Jackie) Dugard ‘Judging the judges: towards an appropriate role for the
judiciary in South Africa’s transformation’ (2007) 20 Leiden Journal of
International Law 965.
duty’.138 Moseneke also indicates that the task to fulfil the

transformative design of the Constitution is a collective one, and falls
to be obeyed by all state functionaries including the judiciary.139 He
continues that the judiciary should acknowledge and recognise that
the Constitution was developed within a particular historical and
social context and that the Constitution is set to redress this legacy.
It was a context of social exclusion, unequal power relations and
material dispossession.140
However, Dugard concluded that both during apartheid and post-

apartheid, South African judges ‘have not assumed an activist role in
society’,141 as they should, and she correctly blames this (in the post-
apartheid constitutional democracy) on what she calls
‘jurisprudential conservatism’.142 She indicates that the South
African Constitution is not only ‘moral’, but also inherently
transformative, and that it thus obliges the judiciary to play an active
role in advancing socio-economic equality.143 Quoting Moseneke,144
she notes that the Constitution ‘enjoins the judiciary to uphold and
advance its transformative design’.145 In his seminal article on
transformative constitutionalism, Klare also highlights what he terms
the ‘inherent conservatism’ of South African legal culture.146 As he
correctly indicates, this approach to legal interpretation is ‘highly
structured, technicist, literal and rule bound’ as opposed to the
‘policy-oriented and consequentialist’ approach that he favours.147
Klare observes that formalist legal reasoning operates to mystify the
choices that judges make in their interpretative work.148 According to
him, this discourages appropriate constitutional innovation and leads
to less generous or innovative interpretation, which may result in
even progressive judges comfortable with an activist role taking for
granted limitations on their interpretative scope that owe more to
138
Minister of Home Affairs v Fourie 2005 1 SA 524 (CC) para 171.
139 D Moseneke ‘Transformative adjudication in Post-apartheid South Africa — taking
stock after a decade’ (2007) 21(1) Speculum Juris 6.
140
Moseneke (n 139 above) 6.
141 Dugard (n 137 above) 967.
142
Dugard (n 137 above) 967.
143
Dugard (n 137 above) 970.
144 D Moseneke ‘The fourth Bram Fischer Memorial Lecture’ (2002) 18 South African
Journal on Human Rights 314.
145
Dugard (n 137 above) 971. See also M Pieterse ‘What do we mean when we talk
about transformative constitutionalism?’ (2005) 20 SA Public Law 155 157.
146
K Klare ‘Legal culture and transformative constitutionalism’ (1998) 14 South
African Journal on Human Rights 146. See also H Botha ‘Freedom and constraint
in constitutional adjudication’ (2004) 20 South African Journal on Human Rights
249; A van der Walt ‘Tradition on trial: a critical analysis of the civil-law tradition
in South African property law’ (1995) 11 South African Journal on Human Rights
169; S Mothupi ‘Transformation of the judicial system: the debate continues’
(2006) 47(2) Codicillus 1.
147 Klare (n 146 above) 168.
148 Klare (n 146 above) 170 – 173. See also D Brand ‘The proceduralisation of South
African socio-economic rights jurisprudence, or “what are socio-economic rights
for?”’ in Botha et al (n 135 above) 51.
tradition than to responsibility and obligations of their role under the

Constitution.149
It is submitted that the Tongoane judgment is a classic example

of this ‘jurisprudential conservatism’. It is further submitted
Tongoane regrettably resorted to the classical English law doctrine of
legal positivism.150 As Ndima observes, unfortunately this positivistic
approach of Western law, ‘... continues to bedevil judicial efforts to
transform African jurisprudence today’.151 Tongoane was a typical
case of mechanical,152 and technical or formal judicialism. It
exemplifies what Hoexter refers to as ‘extreme judicial conservatism
or restraint’.153 Botha summarises Michelman’s illustration of a
commitment to a style of adjudication that is
based on practical reason rather than the mechanical application of

legal rules; that is sensitive to context and the concrete circumstances
in which litigants find themselves; that resist the bureaucratic impulse
to repress social difference in the name of order, uniformity and control;
and that acknowledges the responsibility of judges for their decisions,
rather than deferring to external authority.154
As Christiansen indicates:155
The [Constitutional] Court‘s expansive power to advance substantive

justice comes from institutional characteristics as much as from the
generous enumeration of political and social rights ... the Court has very
broad jurisdiction over constitutional matters and has far-reaching,
discretionary remedial powers ... These procedural characteristics form
a critical aspect of the power and authority of the judiciary and the
Court. In some ways these qualities were necessary for a transition like
South Africa‘s, but they also reflect the conscious vesting of authority in
the Constitutional Court.
As Moseneke indicates, a judicial exercise must not be merely

positivistic, but should rather be value-drenched and sensitive to the
broader social context and peculiarities presented by each case at
hand.156 Dugard correctly points out a ‘judge is not a mere automaton
149
Klare (n 146 above) 170 - 173.
150 See for instance, J (John) Dugard ‘Should judges resign? — a reply to Professor
Wacks’ (1984) 101 South African Law Journal 286; B Schlink ‘German
constitutional culture in transition’ in Rosenfeld (n 106 above) 198.
151 Ndima (n 14 above) 83.
152
D Rousseau ‘The constitutional judge: the master or the slave of the constitution’
in Rosenfeld (n 106 above) 262.
153 Hoexter (n 119 above) 135.
154
n 146 above, 17 - 18.
155 E Christiansen ‘Transformative constitutionalism in South Africa: creative uses of
constitutional court authority to advance substantive justice’ (2010) 13 Journal of
Gender, Race & Justice 581.
156 n 143 above.
who declares law ... he has a wide range of options open to him in fact
finding ...’157
4.3.4 Lessons from socio-economic rights jurisprudence?
Although the constitutional property clause,158 which includes access

to land, is from a purely legalistic point of view not a socio-economic
right, it is linked either directly or indirectly to some socio-economic
rights. This link is attributed to at least two factors. Firstly, similarly
to access to other socio-economic needs such as adequate housing,159
access to basic health care services, sufficient food and water, and
social security,160 access to land is a basic socio-economic need.
Actually access to most socio-economic needs such as housing, food,
and water, depends on access to land for their realisation. As Roodt
puts it, ‘in theory these rights allow citizens to demand from the state
access to basic needs, such as adequate land, housing, education,
health care, nutrition, and social security’.161 Relying on a quote from
Dumbutshena,162 Dugard suggests that South Africa needs an activist
judiciary that purposively pursues transformative adjudication with
the goal of achieving socio-economic equality.163 Secondly, both the
right of access to land and the socio-economic rights are meant to
redress South Africa’s past of racial discrimination and to advance the
constitutional values of freedom, equality and human dignity.164
Thirdly, this link is clear from language used in the Constitution for
protecting and promoting typical socio-economic rights. Section 25(5)
of the Constitution provides that ‘the state must take reasonable
legislative and other measures within its available resources, to foster
conditions which enable citizens to gain access to land on an equitable
basis’. This language is similarly used in sections 26(1),165 26(2),166
27(1),167 and 27(2)168 of the Constitution. At the risk of overstretching
it, one could also argue that the textual juxtaposition of these
constitutional provisions indicates to some extent the indirect link
157
n 137 above, 286.
158 Section 25 of the 1996 Constitution.
159
See sec 26(1) of the Constitution.
160
See sec 27 of the Constitution.
161 MJ Roodt ‘Impediments to the delivery of socioeconomic rights in South Africa’
Paper delivered at SASA (South African Sociological Association) Conference,
Stellenbosch 2008, 1.
162 E Dumbutshena ‘Judicial activism in the quest for justice and equality’ in
B Ajibola & D van Zyl (eds) The judiciary in Africa (1998) 188.
163
n 150 above, 979.
164 See the Preamble to and sec 1 of the Constitution.
165
This section provides that everyone has a right to have access to adequate
housing.
166 This section provides that the state must take reasonable legislative and other
measures within its available resources, to achieve the progressive realisation of
each of these rights.
167 This section provides that everyone has the right to have access to health care
services, including reproductive health care; sufficient food and water; and social
security ...
between these rights. It is submitted that security of communal land

tenure involves the provision and security of a critical socio-economic
resource, namely land. Rights relating to the provision of socio-
economic provision are classed in section 25-29 of the Constitution.
Therefore, when adjudicating upon land issues such as the threat

to security of tenure, courts could to some limited extent also be
influenced by the socio-economic rights jurisprudence.
Socio-economic rights, judicial activism and separation of powers
The jurisprudence of the Constitutional Court regarding the use of

remedial powers in respect of socio-economic rights demonstrates
judicial activism.169 According to Hoexter, this means that judges
cannot, and should not, try to avoid involvement in the adjudication
of social and economic policy.170 However, Dugard notes that the
Constitutional Court has failed, in its adjudication of socio-economic
rights, to promote meaningfully the realisation of these rights for poor
South Africans, and as such, it has not met the challenge to uphold
and advance the Constitution’s transformative design.171 She argues
that the post-apartheid judiciary shies away from an activist role in
a socio-economically divided society, claiming separation of powers in
a legitimate democracy as a shield against playing a more activist role
in the transformation of South Africa.172 Liebenberg argues that a
rigid model of separation of powers in socio-economic rights
adjudication is particularly ill-suited to the South African
transformative Constitution, in which co-operation between the three
branches of government is essential for empowering government to
take the necessary developmental and redistributive measures
envisaged by the Constitution, particularly in relation to socio-
economic rights.173 Brand criticises the Constitutional Court for
‘proceduralising its adjudication of socio-economic rights’,174 which,
he points out, has certain important negative practical consequences,
including its potential to discourage future socio-economic rights
litigation, providing limited tools for the Court to deal with possible,
really difficult, future cases, and its failure to set substantive
168 This section provides that the state must take reasonable legislative and other
measures within its available resources, to achieve the progressive realisation of
each of these rights.
169 Dugard (n 137 above) 965; Christiansen (n 155 above) 589; M Pieterse ‘Coming to
terms with the judicial enforcement of socio-economic rights’ (2004) 20 South
African Journal on Human Rights 383; Hoexter (n 121 above) 142.
170 Hoexter (n 119 above) 139.
171
n 137 above, 971.
173 S Liebenberg ‘Towards a transformative adjudication of socio-economic rights’
(2007) 21(1) Speculum juris 52.
174 Brand n (148 above) 33.
standards to guide future social- and economic policy-making.175 He

further argues that such a proceduralisation approach is theoretically
disappointing from a transformative constitutional point of view
because it fails to give any form of expression to the substantive
political philosophy underlying the Court’s socio-economic rights
judgments.176 In addition to Tongoane, the Court, has thus,
deliberately avoided to provide a satisfactory description of the
substance or content of socio-economic rights in three important
socio-economic judgments.177
A counter argument could be made that courts should exercise

restraint or deference, in the light of separation of powers. Although
this is often only viewed from the perspective of administrative law,
with particular reference to the courts exercising deference when it
reviews administrative conduct for unreasonableness, or an enquiry
about expertise and separation of powers,178 Davis would prefer this
to extend ‘towards a proper consideration of the role of a court in a
constitutional democracy.’179 However, restraints and deference
applies in the context of competency.180 For instance, in the Phambili
Fisheries case, Schultz JA emphasised that judicial deference ‘does
not imply judicial timidity or unreadiness to perform the judicial
function’.181 This was endorsed by O’Regan J in Bato Star 182 by
making it clear that ‘respect’, which is the term she preferred to
‘deference’, does not mean simply rubberstamping an unreasonable
decision in recognition of the complexity of the decision or the
identity of the decision-maker.
The court has played a limited but pivotal activist role as far as
socio-economic rights are concerned.183 As Liebenberg indicates, one
of the major obstacles to the judicial enforcement of socio-economic
175
Brand (n 148 above) 37.
176
As above.
177 Soobramoney v Minister of Health (Kwazulu-Natal) 1998 1 SA 765 (CC)
(Soobromoney); Government of the Republic of South Africa & Others v
Grootboom & Others 2000 1 SA 46 (CC) (Grootboom); Minister of Health v
Treatment Action Campaign (No. 2) 2002 5 SA 721 (CC) (TAC).
178
Y Burns & M Beukes Administrative law under the 1996 Constitution (2006) 31; C
Hoexter ‘The current state South African administrative law’ in C Hoexter & L van
der Vyver (eds) Realising administrative justice (2002); C Hoexter ‘The future of
judicial review in South African administrative law’ (2000) 17 South African Law
Journal 484 510 - 512.
179 D Davis ‘To defer and when? administrative law and constitutional democracy’
(2006) Acta juridica 40 - 41.
180
Hoexter (n 119 above) 143 - 144.
181 Minister of Environmental Affairs and Tourism v Phambili Fisheries (Pty) Ltd 2003
6 SA 407 (SCA) para 50.
182 Bato Star Fishing (Pty) Ltd v Minister of Environmental Affairs 2004 4 SA 490 (CC)
para 48.
183
For example, Soobromoney (n 177 above); Grootboom (n 177 above); and TAC (n
177 above).
rights is formalistic conceptions of the separation of powers

doctrine.184
4.3.6 Failure to comply with the Constitutional obligation to

develop customary law
In South Africa, customary law is given constitutional,185 statutory,186

and judicial recognition.187 For instance, section 39(2) of the
Constitution provides that, a court, tribunal or forum must, in an
adjudicative context, develop customary law in accordance with the
spirit, purport and objects of the Bill of Rights; the Traditional
Leadership and Governance Framework Act188 which recognises the
role and institution of traditional leadership in the administration of
justice; and in Alexkor the Constitutional Court held that ‘the courts
are obliged by section 211(3) of the Constitution to apply customary
law when it is applicable, subject to the Constitution and any
legislation that deals with customary law’.189
What is customary law?
As indicated earlier, communal land tenure is regulated under

customary law. There are many definitions of customary law,190 but
184
Liebenberg (n 173 above) 53.
185 See for instance, sections 39(2) & 211 of the Constitution.
186
For instance, the Recognition of Customary Marriages Act, as amended by the
Judicial Matters Second Amendment Act 42 of 2001, which recognizes customary
marriages; the Traditional Leadership and Governance Framework Act 41 of 2003
which recognises the role and institution of traditional leadership in the
administration of justice; and the long outstanding Traditional Courts Bill [B 15-
2008] published in GG 30902 of 27 March 2008, which provides for the structure
and functioning of traditional courts.
187 Bhe v the Magistrate, Khayelitsha; Shibi v Sithole Case CCT69/03; South African
Human Rights Commission v President of the Republic of South Africa 2005 1 SA
580 (CC) (Bhe) para 43; Alexkor Ltd v Richtersveld Community 2003 12 BCLR 1301
(CC) (Alexkor) para 51; Pharmaceutical Manufacturers Association of SA: In re Ex
Parte President of the Republic of South Africa 2000 2 SA 674 (CC) para 44;
Mabuza v Mbatha 2003 4 SA 218 (C) (Mbatha) para 32; Mthembu v Letsela 2000 3
SA 867 (SCA) (Letsela); D Mailula ‘Abdication of judicial responsibility, cultural
self-determination and the development of customary law: Lessons from
Shilubana’ (2008) 23(2) SA Public Law 215.
188 Act 41 of 2003.
189
M Herbst & W du Plessis ‘Customary law v common law marriages: a hybrid
approach in South Africa’ (2008) 12(1) Electronic Journal of Comparative Law
http://www.ejcl.org (accessed 15 November 2011) 4.
190
See, for instance, Kameri-Mbote (n 45 above); M Chanock ‘Neither customary nor
legal: African customary law in an era of family law reform’ (1989) 3
International Journal of Law and Policy 72 - 88 74; and A Clarke ‘Securing
communal land rights to achieve sustainable development in Sub-Saharan Africa:
critical analysis and policy implications’ (2009) 5(2) Law, Environment and
Development Journal 130 134, wherein it is defined as ‘legal rules and processes
that have become an intrinsic part of accepted legal conduct and arise from
social practices rather than positive law’.
the following summary by Herbst and Du Plessis suffices for the purpose of
this note:191
African customary law in the modern sense of the word (i.e., with
Western influence):
denotes all those legal systems originating from African societies as part
of the culture of particular tribes or groups that have been maintained,
supplemented, amended and or superseded in part by:
(a) changing community views and the demands of the changing world;
(b) contact with societies with other legal systems;
(c) contact with and the influence of other legal systems; and
(d) the direct and indirect influence of foreign (non-indigenous)
government structures.
It is generally accepted that customary law is characterised by

unwritten192 rules, practices or principles; it is dynamic and not
static,193 it relates to and binds a specific indigenous community or
area;194 and it generally regulates private relations.195
Although one accepts the first two characteristics, in the light of

the new constitutional dispensation, it is difficult to accept the last
two, namely that indigenous or customary law only relates to or binds
specific indigenous communities within which it applies, and that it
relates only to private relations or applies only within the private
sphere.
Firstly, from a conflict of law perspective,196 there is no

legislative or judicial certainty as to how to resolve a conflict between
the application of customary law and common law in a particular
situation. Bennett set out the following rules or guidelines as to how
to deal with this conflict:197 where there is an agreement between
the parties, the courts must enforce such agreement; where there is
a dispute on the choice of law, the court must be guided by the nature
of a prior transaction; where the transaction is known to both systems
of law, the court must look at the circumstances of the particular case
to determine general cultural orientation by looking at the subject
191
Herbst & du Plessis (n 189 above) 3.
192 L Cotula (ed) Changes in ‘customary’ land tenure systems in Africa (2007) 10.
193
M Herbst & W du Plessis (n 189 above) 3; Pottier (n 37 above) 55.
194
Section 1 of the Traditional Leadership and Governance Framework Act 41 of
2003.
195
C Rautenbach ‘South African common and customary law of intestate succession:
a question of harmonisation, integration or abolition’ (2008) 12(1) Electronic
Journal of Comparative Law http://www.ejcl.org (accessed 15 November 2011)
4.
196 TW Bennett ‘Conflict of laws — the application of customary law and the common
law in Zimbabwe’ (1981) 30(1) The International and Comparative Law Quarterly
59.
197 Bennett (n 33 above) 53 - 57.
matter and environment in which the transaction was concluded;

where the transaction is known to both legal systems, the parties use
of a form peculiar to one system may be indicative of an intention to
abide by that system; the parties ways of living and their overall
cultural orientation also has a strong influence on the choice of law;
and the parties exemption from customary law.198
I find it difficult to agree with these rules. Customary law and

common law are constitutionally placed on the same level as equal
systems of law in South Africa by section 39(2) of the Constitution.
Why is it necessary for the parties to agree to the application of
customary law if they are not given the same choice in respect of the
application of the common law? If the application of the common law
to everybody is compulsory, why should the application of customary
law be dependent on some rules such as the parties consent, the
nature of their prior transaction, their general cultural orientation or
way of living? Why should they be exempt from the application of
customary law, if they cannot be exempt from the application of the
common law?
The Constitution does not limit the application of customary law

to specific indigenous communities or limit its application to their
culture. It only recognises indigenous law as a system of law applying
parallel to the common law.199 What is required is that its application
must be consistent with the purport, spirit and object of the Bill of
Rights.200 Section 211(3) of the Constitution expressly states that
‘courts must apply customary law when that law is applicable’. What
determines whether customary law applies in a particular matter is,
in my view, not the subject of the matter or the person, but the
nature of the transaction, irrespective of who the subjects are, where
the subjects are resident, what their general cultural orientation is,
whether they chose it to apply, or whether they choose to be exempt
from its application. Whether indigenous law applies should,
therefore, be determined with reference to the transaction itself. In
other words, the question should be whether indigenous law applies
in a particular transaction rather than whether it applies to a
particular person.
Secondly, although one concedes that ‘the traditional public/

private law divide in Western legal systems201 is unknown in
customary law’,202 we cannot take it for granted that all relationships
to which indigenous law applies, are of a private nature and within a
198
Bennett (n 33 above) 57.
199 Section 211 of the Constitution.
200 Section 39(2) of the Constitution.
201
Actually this division is also artificial in common law. Hoexter (n 119 above) 136.
202 Rautenbach (n 195 above) 4 n 19.
private sphere.203 Under customary law, the relationship between a

traditional authority such as a chief, traditional council, or headman
and his or her subject is not a private matter if that authority is
exercising his official powers. That is certainly a public matter
regulated under customary public law. For instance, when a chief
allocates residential, agricultural or grazing land to his or her subject,
he or she is exercising a public power or performing a public function
and thus the relationship is of a public nature, it is unequal and it is
regulated by public law rather than private law. This is, of course, one
of the confusions we often observed not only from the judiciary, but
also from some academics.204 For instance, by declaring
unconstitutional the principle of male primogeniture in the context of
succession to chieftaincy in Shilubana,205 the Constitutional Court
failed to recognize and acknowledge the difference between the
application of this principle in private law relationships of estate
succession or inheritance of private property, as was the case in the
Bhe,206 and the application of this principle in the public law context
of succession to official title of chief, as was the case in Shilubana.207
It is in this context of the public law nature of customary law that the
Tongoane judgment was involved, namely customary or communal
land law with specific reference to security of land tenure.
The court acknowledged that the security of tenure of communal

land is regulated by customary law, as follows:208
[W]hat is not disputed is that the land occupied by the communities is

administered in accordance with indigenous law, and that traditional
leaders, in particular the tribal authorities, play a role in the
administration of communal land.
Despite the Court acknowledging this fact, it failed to deal with the
core issue and as a result contravened its constitutional mandate to
203 Contrary to this, Rautenbach believes that customary law generally deals with
private relationship and therefore operates in the private sphere only. She
mentions that it pertains to limited areas of law such as family law, law of
property, law of delict, traditional leadership and courts, and finally, intestate
succession. How she arrives at the conclusion that traditional leadership is a
private law matter leaves much to be desired. See Rautenbach (n 195 above) 4 -
5.
204
Rautenbach (n 195 above) 4 - 5.
205 Shilubana v Nwamitwa 2008 (9) BCLR 914 (CC) 926 (Shilubana).
206 Bhe (n 187 above).
207
Shilubana (n 205 above).
208 Tongoane (n 22) para 33.
develop customary law209 as required by section 39(2) of the

Constitution.
Failure to comply with the constitutional obligation to develop

customary law
On several occasions, the judiciary has attempted to comply with the

constitutional mandate to develop customary law.210 However, the
Constitutional Court has, at least once,211 arguably failed to live up
to this constitutional imperative. As I have argued elsewhere, the
Constitutional Court has abdicated its responsibility.212 It is submitted
that this abdication is becoming a worrying trend as it resurfaced in
Tongoane.213 As indicated earlier, the Constitutional Court has, yet
again, conveniently overlooked the real issue, namely the
development of communal security of land tenure under customary
land law. This development should involve the ‘integration’ or
‘harmonisation’ of customary and statutory law.214 As to how to
develop customary law, the Court could, and should have learned and
borrowed from its own jurisprudence on the development of common
law.
As indicated earlier, the aim of this note is not to suggest specific

recommendations as to the manner in which the development of
customary law should take place. However, section 39(2) of the
Constitution provides that, a court, tribunal or forum must, in an
adjudicative context, develop customary law or the common law, in
accordance with the spirit, purport and objects of the Bill of Rights.
Although successful in the development of the common law, the Court
is not doing well in the sphere of customary law. This is despite the
fact that the common law and customary law have been placed at the
same level in section 39(2), namely, at the level where they are both
recognised as part of our legal system and both need to be developed
209 This argument relates only to security of communal land tenure as provided for in
CLARA and not the other areas beyond customary law, which are nevertheless
covered by CLARA. The land control areas that are not covered exclusively by
customary law but also by other legislative measures, but which nevertheless fall
within the ambit of CLARA include areas covered by the Upgrading of Land Rights
Act 112 of 1991 and Proclamation R188 of 1969 which is yet to be repealed. See in
this regard, Du Plessis & Pienaar (n 59 above) 81.
210
Alexkor (n 187 above) para 51; Pharmaceutical Manufacturers Association of SA:
In re Ex Parte President of the Republic of South Africa 2000 2 SA 674 (CC) para
44; Mbatha (n 187 above) para 32; Letsela (n 187 above).
211
Shilubana (n 205 above).
212
See Mailula (n 187 above). See also Hoexter (n 161 above) 143; H Botha ‘Rights,
limitations and the (im)possibility of self government’ in H Botha et al (eds)
Rights and democracy in a transformative constitution (2003) 18; See also
S Woolman ‘The amazing, vanishing Bill of Rights’ (2007) 124 South African Law
Journal 762.
213
See Mailula (n 187 above).
in accordance with the values and ideals of the constitution. The

obligation to develop the two systems of law similarly or equally is
peremptory rather than discretionary. In fact, this is an inherent
power of the courts.215 As indicated in Alexkor
while in the past indigenous law was seen through the common law lens,
it must now be seen as an integral part of our law. Like all law it depends
for its ultimate force and validity on the Constitution. Its validity must
now be determined by reference not to common law, but to the
Constitution.216
If the courts have developed the common law, it should not be

difficult to develop customary law.
The development of the common law was dealt with

comprehensively by Ackerman and Goldstone JJ in Carmichele,217 in
which the duty of the courts which is derived from sections 7,218
8(1),219 39(2), and 173220 of the Constitution was emphasised. The
peremptory nature of this obligation was acknowledged in Masiya, as
follows:221
It needs to be stressed that the obligation of the Courts to develop the

common law, in the context of section 39(2) objectives, is not purely
discretionary. On the contrary, it is implicit in s 39(2) read with s 173
that where the common law as it stands is deficient in promoting the s
39(2) objectives, the courts are under a general obligation to develop it
appropriately.
The judges were, however, reminded in Carmichele, that when

developing the common law in accordance with section 39(2) and 173,
they ‘must be mindful of the fact that the major engine for law
reform should be the legislature and not the judiciary’.222 With
approval, the Court quoted the following remarks by Iacobucci J in
R v Salituro:223
215 Section 173 of the Constitution.

216
Alexkor (n 187 above) para 51.
217
Carmichele v Minister of Safety and Security (Centre for Applied Legal Studies
Intervening) 2001 4 SA 938 (CC). In this case, one of the issues was whether the
Court should develop the common law of delict in order to afford the applicant a
right to claim damages if the police or the prosecutor were negligent.
218 This section deals with the general significance of the Bill of Rights within the
framework of the new constitutional dispensation.
219
This section provides that the Bill of Rights applies to all law, and binds the
legislature, the executive, the judiciary, and all organs of state.
220
This section makes provision for the inherent powers of the Court to develop the
common law.
221 Masiya v Director of Public Prosecutions 2007 5 SA 30 (CC) para 39. See also
Carmichele (n 217 above) para 33.
222 Carmichele (n 217 above) para 36.
223 (1992) 8 CRR (2d) 173; [1991] 3 SCR 654, cited in Carmichele (n 217 above) para
36 from a similar citation by Kentridge AJ in Du Plessis v De Klerk 1996 3 SA 850
(CC) para 61.
Judges can and should adapt the common law to reflect the changing
social, moral and economic fabric of the country. Judges should not be
quick to perpetuate rules whose social foundation has long since
disappeared. Nonetheless there are significant constraints on the power
of the Judiciary to change the law ... In a constitutional democracy such
as ours it is the Legislature and not the courts which have the major
responsibility for law reform ... the Judiciary should confine itself to
those incremental changes which are necessary to keep the common law
in step with the dynamic and evolving fabric of our society.
However, in Carmichele the Court held that ‘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 ... whether or not the parties in any particular case request the
Court to develop the common law under s 39(2)’.224 This
jurisprudential approach was also followed in Masiya,225 where the
Court, in the interest of justice and in order to promote the values,
ideals and principles underlying the Constitution, developed the
common law definition of rape to include non-consensual intentional
anal penetration of females but not males.226 As Mienie says, although
the Court acknowledged and appreciated that the non-consensual
penetration of a male victim is ‘no less degrading, humiliating and
traumatic’227 than that of a female victim, it nevertheless held that
law reform is the major responsibility of the legislature.228 The Court
held that ‘the power given to the court to adopt a method of common
law development which is closer to codification than incremental fact
driven development codification’.229 However, as Dersso argues the
‘partial development of the rights [dignity, equality, freedom and
bodily integrity] may ... be considered as amounting to a failure on
the part of the Court to discharge its mandate and obligations.’230
224
Carmichele (n 217 above) para 36.
225
Masiya (n 221 above).
226 The Masiya judgment has received criticism from several academics. For
instance, Mienie argues that the definition of rape should have been extended to
include the non-consensual intentional anal penetration of any other person (not
only females), particularly in the light of the fact that the victim in this case was
a young child. According to her, ‘the refusal to engage the judiciary in legal
reform is both disappointing and puzzling, given the Court’s own celebration of
the centrality of individual rights in rape law reform’. CA Mienie ‘Child rape and
the development of the common law: Masiya v Director of Public Prosecutions’
(CC 54/06; unreported) (2007) 22 SA Public Law 576. For criticism of the Masiya
judgment, see also SA Dersso ‘The role of courts in the development of the
common law under sec 39(2): Masiya v Director of Public Prosecutions Pretoria
(The State)’ http://www.saifac.org.za/docs/res_papers/RPS%20No.%2035.pdf
(accessed 09 May 2012); N Ntlama ‘Masiya, gender equality and the role of the
common law’ (2009) 3(2) Malawi Law Journal 303.
227
Masiya (n 221 above) para 30.
228 Mienie (n 227 above) 579.
229
Masiya (n 221 above) para 31.
230 SA Dersso ‘The role of courts in the development of the common law under
s 39(2): Masiya v Director of Public Prosecutions Pretoria (The State) and Another
http://www.saifac.org.za/docs/res_papers/RPS%20No.%2035.pdf (last visited 20
August 2012) 13.
As the Court has managed to develop the common law, there is no

reason why it should be unable or unwillingly to develop customary
law, particularly because it has an obligation to do so in terms of
section 39(2). As Van der Walt indicates, ‘during the apartheid era
customary law was largely ignored and underrated, especially in the
context of property law [including communal land tenure].231
Rautenbach also indicates that ‘customary law was initially ignored by
the colonials, then later tolerated and eventually recognised, albeit
with certain reservations and conditions’.232 In the light of its current
explicit constitutional recognition, customary law must be developed
in accordance with the spirit, purport and objects of the Bill of Rights.
As Hlophe J states:233
If one accepts that African customary law is recognised in terms of the

Constitution and relevant legislation to give effect to the Constitution,
... there is no reason, in my view, why the courts should be slow at
developing African customary law. Unfortunately one still finds dicta
referring to the notorious repugnancy clauses as though one were still
dealing with a pre-1994 situation. Such dicta, in my view, are
unfortunate. The proper approach is to accept that the constitution is
the supreme law of the Republic ... In line with this approach, my view
is that it is not necessary at all to say African customary law should not
be opposed to the principles of public policy or natural justice. To say
that it is fundamentally flawed as it reduces African law (which is
practiced by the vast majority in this country) to foreign law — in Africa!
5 Conclusion
As indicated earlier, the main challenge facing issues of land in Africa

is always, how to ‘recognise and secure land rights that are clearly
distinct from “Western legal” forms of private property but are not
simply “customary”, given the impacts of both colonial policies and of
past and current processes of rapid social change’.234 As indicated
earlier, in Tongoane, this challenge was identified, acknowledged,
confirmed and summarised by Ngcobo CJ as follows:
[t]he communities are concerned that their indigenous-law-based

system of land administration will be replaced by the new system that
CLARA envisages. They are concerned that this will have an impact on
the evolving indigenous law which has always regulated the use and
occupation of land they occupy. They are further concerned that their
land will now be subject to the control of traditional councils which, as
is apparent from the record, they consider to be incapable of
231 AJ Van der Walt Constitutional property law (2005) 402.

232 Rautenbach (n 195 above) 1.
233
Mbatha (n 187 above) para 30.
234 Cousins & Claasens (n 18 above) 3.
administering their land for the benefit of the community. All the
communities claim that the provisions of CLARA will undermine the
security of tenure they presently enjoy in their land, and those who own
the land fear that they will be divested of their ownership of the land.
While some of these claims are disputed by the government
respondents, what is not disputed is that the land occupied by the
communities is administered in accordance with indigenous law, and
that traditional leaders, in particular the tribal authorities, play a role in
the administration of communal land. There is some issue as to the
extent to which the role of traditional leaders and tribal authorities
accords with indigenous law.235
However, the Court conveniently failed to consider the core issue

raised by the applicants, namely the threat to security of customary
land tenure pursuant to the introduction of CLARA. This is despite the
Court acknowledging that, ‘what lies at the heart of the confirmation
proceedings is the question whether CLARA undermines the security
of tenure of the applicant communities’.236 Surely this was the
substantive matter, or core issue that the Constitutional Court was
confronted with. The applicants challenged CLARA mainly because
the new statutory system of land tenure it introduced sought, among
others things, to replace the customary land tenure or land tenure
under customary law, and this threatens and puts in a precarious
position, their security of tenure over the lands they use or occupy. It
is upon this basis that CLARA was argued to be unconstitutional.
However, this issue was not dealt with by the Court, probably because
it was ‘not disputed ... that the land occupied by the communities is
administered in accordance with indigenous law’.237 Moseneke has
indicated that judges are required to, and should at all times stay
faithful to the facts before them.238 However, in Tongoane the judges
deliberately ignored some of the facts before them and focused only
on others.
It is submitted that the Court should have addressed this issue

because it has a constitutional mandate not only to redress societal
inequalities caused by the apartheid regime, but most importantly to
develop the indigenous law system as a parallel system to the common
law and statutory law, and actually integrate it into, and harmonise
it with the new constitutional dispensation. In other words, the Court
failed to ‘recognise and secure land rights that are clearly distinct
from “Western legal” forms of private property but are not simply
“customary”, given the impacts of both colonial policies and of past
and current processes of rapid social change’.239 The Court should
235
238
Moseneke (n 139 above) 8.
239 Cousins & Claassens (n 18 above) 3.
play what Hoexter refers to as ‘the transformation enhancing role of

the judiciary’.240 Furthermore, it is submitted that the sensitivity of
land issues, and in particular the security of tenure of communal land,
warrants a different approach from other issues in order to ensure
legal certainty.
However, the Court found it convenient to avoid the core issue

and to deal with the incidental or secondary procedural issues only as
discussed earlier. It is submitted that it was incorrect for the Court to
avoid dealing with the substantive issue of security of tenure. As
highlighted in Port Elizabeth Municipality v Various Occupiers, ‘the
procedural and substantive aspects of justice and equity cannot
always be separated. The managerial role of the courts may need to
find expression in innovative ways’.241 It is submitted that the Court
has failed to play this critical ‘managerial role’ in Tongoane.
This was either a misdirection, or a deliberately and concerted

effort to avoid facing the core substantive challenge head-on
probably because of the sensitivity of the issue, its complexity, and
its political nature.
240
Hoexter (n 119 above) 138.
241 2005 1 SA 217 (CC) para 39.
RATIONALITY, THE RULE OF LAW,
AND THE SOVEREIGN RETURN
Stewart Motha*
Spectres haunt the rule of law in South Africa. Law is deeply

implicated in constituting and sustaining the structures and practices
of the colonial and apartheid regimes. Despite this legacy, law is at
the heart of transformation and renewal. The constitutional
dispensations of 1994 and 1996 expressed a great faith in law —
especially in ‘the law of the law’.1 The early transformative
metaphors of the ‘book’ and the ‘bridge’ in the Epilogue to the
interim constitution evidence the aspiration that the introduction of
a fundamental law and constitutional supremacy would help a
fractured and wounded polity to turn the page, ‘open a new chapter’,
and journey to a new social and political order.2 The abandonment of
parliamentary sovereignty in favour of constitutional supremacy
placed the rule of law and its custodians — judges — at the apex of
the juridico-political order. It can of course be argued that this
dispensation was arrived at through a representative and deliberative
process where the peoples’ delegates constituted a new order, which,
through democratic means, posited the constitution as supreme. This
reflects a feature of many modern liberal constitutional orders that
claim to logically sustain the tension between constitutional
supremacy and the sovereignty of the ‘people’, or indeed to sustain a
* Reader in Law, Birkbeck College, University of London; s.motha@bbk.ac.uk. I am

grateful to the editors of the Constitutional Court Review for arranging the
productive symposium at which this paper was first presented in Johannesburg in
December, 2011. My special thanks to Danie Brand, Karin van Marle, and Helen
Carr for comments and numerous discussions. I also benefitted from the
comments of the two anonymous referees.
1
See J Derrida ‘The laws of reflection: Nelson Mandela, in admiration’ in J Derrida
& M Tlili (eds) For Nelson Mandela (1987).
2
AJ van der Walt ‘Dancing with codes – protecting, developing and deconstructing
property rights in a constitutional state’ (2001) 118 South African Law Journal
258 259; M Antaki ‘The bridge and the book’ in K van Marle & S Motha (eds)
(forthcoming) Genres of critique: Aesthetics and liminality in post-Apartheid
jurisprudence.
113
114 Rationality, the rule of law, and the sovereign return
productive tension about where sovereignty is located.3 This very

liberal story is replete with tensions and contradictions — ones that
are amplified and potentially unsustainable and undesirable in the
South African setting. In this essay I consider aspects of the rule of law
and constitutional supremacy as a feature of neoliberal governance. I
contrast this with approaches to plurality that can inform the rule of
law in ways that are more consistent with the aspiration of renewing
the social and juridical order.
Neoliberal governance exhibits a strong preference for the rule of

experts, the accountability and predictability of rules, and the
subjection of state power to judicial control. Private power and the
exigencies of capital face relatively few constraints.4 Democratic and
parliamentary forms of accountability and state institutions are
viewed as overly bureaucratic centres of potential corruption. The
‘rule of law’ is thus posited as an unrivalled good when compared to
the ‘will’ of sovereign authorities whether that is located in an office,
assembly, or population. I am not suggesting that the rule of law
should not be a cornerstone of a constitutionally delimited state. The
problem is rather one of considering the significance of the sovereign
‘will’ and its relationship to the conditions of plural existence.
Plurality and law (especially the ‘law of the law’ or the grounds of
law) are co-originary. When law is turned into instrumental rule-
governed existence, or mediated by the reason of judges alone, the
conditions of law’s origination are neglected. This essay explores how
law might remain open and attentive to plurality especially in
decisions where political coexistence is what is at stake.
Reconciliation and Amnesty were central to the post-apartheid

constitutional dispensation. Despite the work of the Truth and
Reconciliation Commission, many South Africans convicted of crimes
on the margins of the political crisis continued to languish in prison.
In 2007 President Mbeki proposed that he deal with this ‘unfinished
business’ through the exercise of the Presidential pardon power. What
followed was a bureaucratisation of the sovereign will to pardon on
the one hand (the President’s process), and the subjection of this
process to judicial review on the other. The South African
Constitutional Court’s decision in Albutt v Centre for the Study of
Violence and Reconciliation is an exemplary instance of the tension
between sovereign and judicial power.5 In this essay I examine the
wider implications for the rule of law that follow from subjecting the
pardon power to judicial review on the grounds of rationality. I place
3 P Kahn Political theology: Four new chapters in the concept of sovereignty (2011)
8-17.
4
See D Harvey A brief history of neoliberalism (2005) discussed below.
the tension between sovereign and judicial power alongside the

question of the grounds of law, or the ‘law of the law’ as plurality.
1 Being reasonable
1.1 Rationalising the pardon power
In Albutt the applicants challenged the power of the President to

grant a pardon pursuant to section 84(2)(j) of the Constitution to
people who claim to be convicted of offences with a political motive.
The powers and functions of the President are set out in section 84 —
and includes the responsibility for ‘pardoning or reprieving offenders
and remitting any fines, penalties or forfeitures’ (section 84(2)(j)).
The question regarding the exercise of the pardon power framed by
the parties and the Constitutional Court was whether the President,
before granting a pardon to a group of convicted prisoners, is required
to afford the victims of the offences a hearing. The dispute was
essentially between ‘the state’ in the form of the President and the
Minister for Justice and Constitutional Development, and a coalition
of non-governmental organisations (NGOs) whom, amongst other
things, represented victims of crimes committed by those potentially
benefitting from a pardon.
The background to the case is as follows. In a speech to a joint-

sitting of Parliament on 21st November, 2007 President Mbeki
announced the special dispensation for people convicted of politically
motivated offences. The dispensation was aimed at dealing with the
‘unfinished business’ of the Truth and Reconciliation Commission (the
TRC), including the question of amnesty for those who had not
participated in the TRC process. The process adopted by the President
included setting up a multi-party ‘Pardon Reference Group’ (the
PRG). The PRG had a limited lifespan, and was required to consider
applications by persons ‘convicted and sentenced solely on account of
allegedly having committed politically motivated offences before
June 16, 1999’.6 The Terms of Reference of the Special Dispensation
also set out further details of how an applicant would qualify,
including that they needed to have been sentenced to a period in
prison or a fine for an offence, act or omission associated with a
political objective committed in the course of offences of the past.
While the PRG was charged with advising the President – he made it
clear that he would form an independent opinion based on the
material placed before him. President Mbeki had also stated that in
making the decision with respect to the pardon he would be guided by
the principles and values that underpin the Constitution. These
6 Albutt (n 5 above) para 5.

included ‘the principles and objectives of nation-building and

national reconciliation’.7 The Presidential pardon was to be used to
deal with the ‘unfinished business’ of the TRC.8 NGOs representing
victims made attempts to facilitate their participation in the special
dispensation process. The President and the PRG refused to provide
victims or their representatives the opportunity to participate and
thus litigation followed.
Although multiple grounds were asserted for contending that

victims had a right to participate in the process, the Constitutional
Court crystallised the question as being whether the ‘decision to
exclude the victims from participating in the special dispensation
process is irrational’ (emphasis added).9 The Court applied the
principle that the exercise of the pardon power, in addition to being
subject to the Constitution as the supreme law, and to the doctrine
of legality, also ‘must be rationally related to the purpose sought to
be achieved by the exercise of it’.10 Thus, Chief Justice Ngcobo who
wrote the main judgment surmised that the President’s approach to
exclude victims from the process of the special dispensation must be
rationally related to the achievement of the objective of granting the
pardons. The means used had to be rationally related to the ends
sought. This introduces an instrumental logic between means-ends in
cognising the nature of the pardon power. Ngcobo CJ thus proceeded
to assess the rationality of the dispensation process as a question of
the suitability of the means used to achieve the stated ends.
The special dispensation had sought to attain national unity and

reconciliation through the application of ‘principles and values that
underpin the constitution’. These objectives (ends) and their
formulation were principally drawn from President Mbeki’s speech of
21st November 2007 to the Joint-Sitting of Parliament. Ngcobo CJ
reasoned that as the special dispensation was compared by the
President to the Amnesty process of the TRC, and the involvement of
victims had been central to the work of the TRC, such participation
should be regarded as central to ‘rebuilding a nation torn apart by an
evil system and promoting reconciliation’.11 If the latter were the
ends sought, to deny victim participation would constitute irrational
means. The Chief Justice summed up the court’s approach thus:12
7
Address by the President of South Africa, Thabo Mbeki to the joint sitting of
Parliament, Cape Town, 21st November, 2007 http://www.anc.org.za/show.php?
id=4240 (accessed 17 November 2011).
8
As above.
11
Albutt (n 5 above) para 56.
Once it is accepted, as it must be, that the twin objectives of the special
dispensation process are nation-building and national reconciliation and
that the participation of victims is crucial to the achievement of these
objectives, it can hardly be suggested that the exclusion of the victims
from the special dispensation process is rationally related to the
achievement of the objectives of the special dispensation process.
He went on to say that victim participation in the process for granting

pardons was ‘the only rational means to contribute towards national
reconciliation and national unity’.13 It was also held that the context
of the special dispensation process, involving the establishment of
whether the relevant offence was committed with a political motive,
required the victim to be granted a hearing.14 ‘Accountability,
responsiveness, and openness’ enshrined in section 1(d) of the
Constitution required that political parties alone could not be the
ones permitted to make submissions to the PRG or the President. In a
separate judgment that concurred with that of the Chief Justice,
Froneman J extended the basis of this conclusion to ‘African’ notions
of ‘participatory democracy’. But this latter tradition was only
permitted the status of a ‘further legitimisation’ of the main
judgment, and was not treated as a ‘direct authority’.15 If African
tradition was being invoked, it needed, it seems, to be given a
subordinate status.
The ends of nation-building and reconciliation were not in

question, but a process that excluded the participation of victims (the
means used) was deemed irrational. The pardon process was thus
subjected to the court’s standard of rationality, which, as with many
courts around the world, is driven by the judicial approval of the
means used to achieve legitimate or just ends.
The judicial assessment of the means-ends nexus is often

presented as an objective test. The reasoning in such cases is often
circular. Recall the preeminent English authority on rationality — the
Wednesbury case.16 In that case an unreasonable or irrational
decision or judgment was characterised by Lord Greene MR as a
decision ‘so unreasonable that no reasonable body could make such a
decision’. A putative rational being is invoked who can stand in as the
objective standard to be compared with the impugned decision-
maker. A simpler way of putting it would be to say ‘it is such a bad
decision that no good decision-maker could have made it’, and the
court will quash bad decisions. This circular reasoning is a powerful

14
16 Lord Greene in Associated Provincial Picture Houses v Wednesbury Corporation
[1947] 1 KB 223; see also Lord Diplock in Council of Civil Service Unions v Minister
for the Civil Service [1985] AC 374.
rhetorical gesture. It enacts an ‘as if’ — ‘as if’ courts can provide a
measure for the full range of rational decisions.17 The nexus between
‘means’ and ‘end’ is that putative measure. Rationality in these cases
is a notion formed through performative rhetorical gestures that
place the substantive facts and circumstances of a case alongside the
processes, strategies, and techniques of governance (means). The
means used to reach particular objectives are then judged to be
legitimate or not. It is worth exploring the relationship of means to
ends further before moving to alternative characterisations of how
the rule of law might interact with an exceptional juridical device
such as the pardon power.
1.2 The means/ends logic and rationality
On what basis is a process (means used) regarded as legitimate? What

is the relative relationship of means to ends? Can the ends be justified
by the means, or vice-versa? As Walter Benjamin put it in his seminal
essay ‘A critique of violence’ in 1921: ‘the most elementary
relationship in any legal system is that of ends to means, and further,
that violence can first be sought only in the realm of means, not of
ends’.18 It might at first glance seem rather straightforward to assess
whether violence (in the widest sense of the German term Gewalt) is
being used for just or unjust ends. For instance, it might be assumed
that a person convicted of a crime can legitimately be incarcerated
for protecting the community or a range of other ends such as
rehabilitation. The violence of confinement and deprivation of liberty
would be justified on the basis that a judicial process established guilt
and the proportionality of the sentence. Benjamin suggests that such
a conclusion is not so self-evident. We would still be troubled by
whether violence could even be used for just ends. Nor is it possible
to resolve the problem in the realm of means.
There are two standard resolutions offered for this means-ends

conundrum — from natural and positive law. Benjamin goes on to
dismiss both schools: 19
[They] meet in their common basic dogma: just ends can be attained by
justified means, justified means used for just ends. Natural law
attempts, by the justness of the ends, to ‘justify’ the means, positive
law to ‘guarantee’ the justness of the ends through the justification of
the means. This antinomy would prove insoluble if the common dogmatic
17 The centrality of the notion of ‘as if’ introduced to modern philosophy and
juridical systems by Immanuel Kant is the subject of a longer study by me which is
forthcoming.
18 W Benjamin ‘Critique of violence’ in W Benjamin One-way street and other
writings (1978) 132-154.
19 Benjamin (n 18 above) 133.
assumption were false, if justified means on the one hand and just ends
on the other were in irreconcilable conflict. No insight into this problem
could be gained, however, until the circular argument had been broken,
and mutually independent criteria both of just ends and of justified
means were established.
Benjamin interrupts and rejects the assumption that the relationship

of means to ends can be taken for granted. There may well be just
means and just ends. The problem lies in leaping from the
justification of means to the justifications of ends and vice versa. We
are thus called on to emphatically reject the natural law notion that
means used can be legitimated by just ends.20
The Constitutional Court in Albutt applies the positivist dogma

that the legitimacy of ends can be ‘guaranteed’ by just means. Such
just means are equated with a process of consulting victims before
the pardon power is exercised. The legitimacy of the exercise of the
pardon power would then be guaranteed by the process of
consultation. The Court, however, fails to countenance the fact that
means and ends might exist in different orders of legitimacy and
judgment. Moreover, they do not entertain the fact that means and
ends might actually exist in irreconcilable conflict. This takes us to
the heart of the problem. The means-end logic risks evading the
question of responsibility and judgment faced by a public official or
judge.
The pardon power vested by the Constitution in the President, and

the circumscribed process he adopted for exercising it in this
particular instance (through a bureaucratic process of consulting the
PRG), was a means of attending to the ‘unfinished business’ of
transition from Apartheid and pursuing the ends of nation-building
and reconciliation. Neither the Court nor the President entertain the
possibility that these means and ends may be irreconcilable, and
justly so. The pardon power is invoked to contend with a crisis, to
suture a tear in the polity. The victims can justly claim that they
should be involved in the process that addresses the individual
wounds, and especially the persons who have inflicted the scars.
However, ‘finishing the business of reconciliation’ through the
exercise of the pardon power hardly seems to grapple with the infinite
20 Walter Benjamin (n 18 above) went onto to discuss violence as a means in terms

of ‘law making’ and ‘law preserving violence’, comparing revolutionary general
strikes and the police power to determine when to intervene for ‘security
reasons’ (141-42). It is beyond the scope of our inquiry to pursue Benjamin’s
messianic account of the emergence of a non-state form of ‘divine violence’
which would break the reign of violence as means (153-54). However, he does
point to the significance of the relationship between means and ends in a legal
system, and interrupts the all too regular deployment of rational means as a basis
for judging the legitimacy of ends.
demands of responsibility that need to be attended to in South

Africa.21
We thus need to examine whether a law-sanctioned power such as

the pardon can be subjected to forms of judgment or responsibility.
What is the nature of responsibility and judgment in the context of a
power to pardon? The Court’s conclusion in Albutt was that nation-
building and reconciliation, reference to the TRC’s Amnesty process
in the President’s 2007 speech, the deliberative nature of the South
African democratic settlement, and the need to ascertain whether
convictions were in fact for political crimes all point to consultation
of victims as the only legitimate means. The validity of the judgment
on whether to pardon depends on the process adopted when making
the decision. It is a procedural approach to the question of judgment
and responsibility. But is a procedural approach sufficient in the face
of a power as irregular as the pardon? What does it mean to regularise
the pardon power? More widely, what are the implications for
transformative constitutionalism if the question of judgment and
responsibility is reduced to procedure? Before addressing these
questions I need to place the pardon power in a wider juridical and
political frame than the one entertained by the Constitutional Court’s
means-end approach to a rational decision.
1.3 Undoing the pardon power
The state through the office of the President has the power to pardon.
More widely the Albutt case poses the question of what the
relationship is between the state and the law. The Constitution grants
the power to the President. The source of the law is the Constitution.
Nonetheless, there is a question of what governs the incidence of the
exercise of the pardon power. Is the Presidential pardon power
subject to other norms — such as the version of reasonableness and
rationality administered by the Constitutional Court? Or does the
Presidential pardon power express the presence of a ‘sacral’ or
‘infinite’ element in the juridical order? The more delimited approach
to the pardon power was expressed in the 1997 ruling of the
Constitutional Court in The President of the Republic of South Africa
v Hugo.22 The Court said:23
(Pardoning a sentenced person) is not a private act of grace in the sense

that the pardoning power in a monarchy may be. It is recognition in the
Interim Constitution that a power should be granted to the President to
21 I return to this problem of a crisis of judgment and responsibility and an

alternative critical approach below.
22
President of the Republic of South Africa and Another v Hugo 1997 4 SA 1 (CC).
23 Hugo (n 22 above) para 44.
determine when, in his view, the public welfare will be better served by
granting a remission of sentence or some other form of pardon.
This is an instrumental rationalisation for the existence of the pardon

power, and a utilitarian account of when it should be exercised.
Beyond this narrow ambit the extent to which the pardon power
should be regulated by the ordinary law of the land is the subject of
extended scholarly discussion.
According to Paul Kahn — who renews the significance of Carl

Schmitt’s thinking for constitutional and legal disputes in
contemporary settings — we need to take seriously the fact that the
existence of the state admits of an existential and sacral element of
sovereignty in any juridical order.24 Famously for Schmitt, ‘Sovereign
is he who decides the exception’.25 Moreover, the exception is not
lawless chaos but a borderline situation that is part of the juridical
order. The medieval monarch manifested the miraculous and sacral
aspect of sovereignty through the notion of the ‘King’s two bodies’.26
The King was at once a natural and finite person, and infinite in the
reach of his office — ‘the King is dead, long live the King’. The
monarch could also defy the laws of nature by curing individuals
through the laying of hands, and step outside the law of the land to
address the social body through equity.27 The pardon power is one
sense in which the sacral power of the mediaeval monarch persists.
As Kahn puts it:28
This is a remnant of the sovereign power to decide on the exception to

the law. It always verges on lawlessness as we try to find a ground for
mercy that does not appear to be mere partiality. The ground can only
be care, which is always personal and unbound by rules. We may feel
that we need a pardon power; yet if we cannot speak of care, love, or
the sacred, we are at a loss to offer a justification that is consistent with
our other beliefs about the rule of law. Our ordinary inclination, then, is
to displace pardon by a system of “earned probation”, administered by a
bureaucratic board. We seek to normalise the exception.
Kahn identifies the tension between the rule of law, and the need for
an exceptional power that is not regulated by the normal law. As
Grant Gilmore put it: ‘In Hell, there will be nothing but law’.29 The
pardon is an event, a decision, which quintessentially should be
outside law: ‘it is a gift that comes as if from nowhere. Indeed if we
24
Kahn (n 3 above) 37ff.
25
C Schmitt Political theology: Four chapters on sovereignty ([1922] 1986).
26 See E Kantorowicz The King’s two bodies: A study in mediaeval political theology
(1957); and E Santner The royal remains: The People’s two bodies and the
endgames of sovereignty (2011).
27 Kahn (n 3 above) 37.
28
Kahn (n 3 above) 38.
29 G Gilmore The ages of American law (1977) 111, quoted in Kahn (n 3 above) 38.
can give an account of its exercise in a particular case — that is, if we

can offer a causal explanation of how the pardon came to be granted
— we are more likely to judge it corrupt’.30
Importantly, Kahn also points to the reluctance in a liberal legal

order to invoke notions of ‘care, love, or the sacred’ when justifying
or legitimating the exercise of exceptional powers. The tendency is to
seek a bureaucratic resolution — or as we see in Albutt, a means-end
rationalisation, as well as instrumental reasons such as public welfare
as in Hugo. Reliance is placed on a form of technical expertise
(probation) or mode of reasoning (means-end) where judgment and
responsibility are replaced with a logic of substitution. In the face of
what remains of a fearsome power (the sovereign exception)
utilitarian reasons (of social cohesion or welfare) and procedural
forms are asserted. As I have been emphasising, this evades the
problem of judgment and responsibility.
In this context it is worth recalling the hopes and aspirations that

were laid out in relation to the rule of law and adjudication in the
‘new South Africa’. With constitutional supremacy came a call for
judges to take an active part in the transformative aspirations of the
Constitution. The rule of law needed to be given teeth and content —
but with a clear sense of what the South African post-apartheid
setting demanded. It is worth quoting at length, then, Karl Klare’s
articulation of the ‘postliberal’ approach to constitutional
adjudication and the rule of law: 31
Do traditional accounts of legal constraint and the rule-of-law ideal

make sense in the new South Africa? Does the rule-of-law ideal imply a
depoliticised conception of law inconsistent with the aspiration to
develop adjudicative methods that will contribute to egalitarian social
change? Or, must we develop a revised, perhaps somewhat more
politicised, understanding of the rule of law and adjudication that can
be consistent with and support transformative hopes? Can we conceive
practices of constitutional interpretation that acknowledge and fulfil the
duty of interpretive fidelity and yet that are engaged with and
committed to “establish[ing] a society based on democratic values,
social justice and fundamental human rights,” a society that will
“[i]mprove the quality of life of all citizens and free the potential of
each person[,]” in the words of the Preamble? Can we describe a method
of adjudication that is politically and morally engaged but that is not
illicit “judicial legislation”? Is there a postliberal account of the rule of
law [emphasis added] suitable to the political challenges South Africa
has set for itself?
30 As above.
31
K Klare ‘Legal culture and transformative constitutionalism’ (1998) South African
Journal of Human Rights 146 150.
Klare was articulating the need to break from traditional models

of adjudication, and the sense that the rule of law during apartheid
was a system-preserving institution with habits of adjudication and
legal practice from which a deliberate and conscious break was
required. It was not sufficient for South Africa to develop a
sophisticated constitutional framework with transformative
aspirations. None of this would be of much consequence if judges saw
their role in conventional liberal terms as those who merely applied
and interpreted the law ‘on the books’ (as if that were ever
possible!). What, then, does it mean for judges to be politically and
morally engaged in a transformative constitutional enterprise?
I cannot say whether Klare would approve of the Court’s approach

to the Presidential pardon in Albutt. That is partly due to the limits
of the notion of ‘postliberal’ which connotes both the preservation of
liberal rights and values, as well as expecting a departure from them.
It may be that the Court’s stance of requiring that victims be accorded
a hearing conforms to the constitutional aspiration — visible also in
the constraints on emergency powers (section 37) — of holding the
executive to account in a setting where executive power has been the
instrument of multiple excesses. What emerges, then, is the sense
that according victims a hearing cannot be the end of the matter
when it comes to determining the role and content of the rule of law
in a post-colonial or transformative setting. We are simply left with
the conclusion that the Court has been effective in reviewing the
exercise of the pardon power because the President has been held to
the Court’s standard of rationality. I would like to pose the problem
rather differently.
A general tension has emerged between what can be described as

neoliberal forms of governance, and the necessary recourse to the
exceptional and the infinite in any juridical order. In A brief history
of neoliberalism David Harvey points to the juridicalisation of politics
as a key feature of neoliberalism.32 He suggests that neoliberal
governance takes place through experts and elites who are often
suspicious of and discourage more democratic and parliamentary
modes of decision-making. With the horizons of social agonism and
the possibility of consensus limited or discredited, political conflict is
shifted to the courts. The rule of law’s protection of individual rights
becomes the key mediator and remedy to wider political and social
disagreements. This rights-centred governance is also accompanied
by a potentially contradictory appeal to populism in times of crisis.
Immigration, crime, migrant labour, refugees, and the post 9/11
security agenda are all dealt with as if there is a ‘society’ and
cohesive nation-state being harmed by an external threat. Neoliberal
32 Harvey (n 4 above) 66-67.

institutions take the individual as primary and elevate the rule of law
as the key protector and guardian of individual interests. Experts such
as judges are increasingly deployed to check political decisions and to
assess whether they conform to increasingly globalised standards of
governance. This rule-governed regime is consistently inattentive to
the grounds of political community. In the next part of this essay I
consider the significance of being attentive to the problem of
plurality in the context of responsible decision making and the
exercise of judgment in hard cases.
2 Plural becoming
2.1 The sovereign return
What was hidden in this constitutional conflict over the pardon was a
fear of the power of the President. It is a fear of an archaic sovereign
remainder to be found in all modern democratic states, even though
the framers of the South African Constitution have done much to
delimit it. Plurality may hold the key to addressing this fear, as well
as to developing a different conception of judgment and responsibility
through the rule of law. We can only begin to explore that here.
Addressing the archaic sovereign remainder which presented itself in
Albutt is one place to start.
In the opening pages of Rogues: Two essays on reason, Jacques

Derrida refers to ‘The wolf and the lamb’, a poem by La Fontaine.33
This poem expresses the question ‘who has the ability, right or power
to decide on the law’, and with what force. Derrida restates it as
follows:34
But just who has the right to give or take some right, to give him- or
herself some right or the law, to attribute or to make the law in a
sovereign fashion? Or to suspend the law in a sovereign way?
In the South African setting this is presented as a tension between the

assertion of the President to govern with a popular mandate, and the
role of the courts in constitutional adjudication. Carl Schmitt has one
of the most influential responses to this question in the modern
tradition. He gives an account of the laicised theology of sovereign
power, drawn from the ‘outer most sphere’ of limit-situations as he
called it. The pardon is one such limit situation — one that the Court
in South Africa attempted to discipline through the means-end
account of rationality. It also drew on the fact that one of the
fundamental principles in the Constitution is participatory democracy
33
J Derrida Rogues: Two essays on reason (2005).
34 Derrida (n 33 above) xi.
– and that principle avows consultation and the opportunity to present

a view. The real fear, however, is the President’s ability to say ‘I can’,
and ‘I will’. This is the ipseity of sovereignty that grounds and
threatens the constitutional order. It is what is at stake when
President Jacob Zuma asserts that: ‘[t]he powers conferred on the
Courts cannot be regarded as superior to the powers resulting from a
mandate given by the people in a popular vote’.35
Ipseity is akin to auto in Greek. It is from this ipse that one

extracts the possibility of giving oneself law, or asserting self-
determination. Democracy can only be imagined with the assertion of
this ipse/auto — the autonomous, self-same subject. This ipseity is
taken up in liberal ontology, and this is what is called into question by
Derrida. This possibility of an ‘I can’ by myself, that is, this ipseity, is
named here in order to call it into question. This involves calling into
question the ‘assembling’ of the ‘resembling ensemble’, the
simulacra of resemblance, the simulation that is the act of making
similar.36 This is at the heart of what is potent about the presidential
assertion — ‘I can’, ‘I will’, ‘in the name of the people, their
democratic mandate’ and so on. To say ‘I can’ also has a social
manifestation and is a key condition of many liberal and modern
concepts (albeit with a pedigree that can be traced back to
antiquity): possession, property, power, husband, father, son,
proprietor, seignior, sovereign, host, or master. Think, also, of the
possessive individual from Hobbes and Locke. There can be no
liberalism, and no liberal democracy, not to mention all those social
contract theories, without this notion of ipseity.37
The exercise of presidential power and the assertion of its

independence from other constitutional branches is the eruption of
what threatens democracy from within. In modern democracy the
individual autonomous being becomes one with a people or nation,
authorises subjection to a sovereign, or holds sovereignty as one-of-
the-many citizens. Democracy is a force, a force in the form of a
sovereign authority (as reason and decisiveness), and re-presentation
of the power and ipseity of a people. The authorisation of the
exercise of power in modern democracy must constantly return to its
source, its authorisation. While the axiomatic of democracy as circle,
sphere, ipseity, autos of autonomy, symmetry, homogeneity,
semblance and similarity, and God which is the analogy in the
American Declaration, are all ways of expressing the autonomy of the
35 Address to the National Assembly bidding farewell to Chief Justice Ngcobo and
welcoming Chief Justice Mogoeng, 1 November 2011, cited in G Budlender
‘People’s power and the courts. Bram Fischer Memorial Lecture, 2011’ (2011) 27
South African Journal on Human Rights 582 582.
36
Derrida (n 33 above) 11.
37 Derrida (n 33 above) 11-12.
political, Derrida identifies the double bind within this tradition of

democracy. Each of these elements are incompatible with, and clash
with, the ‘truth of the democratic’, namely the other, heterogeneity,
dissymmetry, multiplicity, the anonymous ‘anyone’, the ‘each one’.38
Although participatory democracy is an important principle in the

South African Constitution it hardly begins to tackle what might be
rotten in the core of democracy. Democracy is an ipso-centric order
where the autonomy of the political undoes itself when it is given
expression through a sovereign form. The double bind of ipseity is the
clash of the ‘I can’ with the autonomy of ‘everyone’. This is what
Derrida termed the auto-immunity of democracy — the possibility that
like a body’s immune system which can turn on itself, the principle of
autonomy and a representative mandate can destroy the very
conditions of its own existence when the sovereign intervenes in the
name of the community to suspend democratic participation. This is
effectively what was attempted by President Mbeki in relation to the
pardon power.
How should a post-liberal legal system respond to this sovereign

return? In constituting a legal order, sovereignty is in movement
towards a frame of reference, a normalised political condition. The
illimitable thus moves towards and by way of a limit. It could also be
said that the illimitable exists in and through a limit. As Derrida puts
it, once the indivisible is divided, and the illimitable has been limited,
sovereignty as the ‘undivided’ and ‘unshared’ becomes an impossible
possibility. This is in contrast to the singular plenitude of sovereignty
which is often asserted:39
[I]s it not the very essence of the principle of sovereignty everywhere

and in every case, precisely its exceptional indivisibility, its illimitation,
its integral integrity? Sovereignty is undivided, unshared, or it is not. The
division of the indivisible, the sharing of what cannot be shared: that is
the possibility of the impossible.
What is crucial here is the insight that whatever divides — and the
sovereign limit divides — also ‘shares itself’ in this partition.40 That is
to say, the assertion of something like the plenitude of the pardon
power exposes the frontier and fault line of sovereignty. That is
precisely why cases such as Albutt are limit cases in more than one
sense. They expose the return of the archaic sovereign form as well
as test the efficacy of limiting the illimitable, dividing the indivisible.
The question, now, is whether that ‘impossible possibility’ as Derrida
called the task of undoing sovereign plenitude can be driven deeper
38 Derrida (n 33 above) 14.

39
J Derrida Without alibi (2003) xx.
40 As above.
into the re-imagination of the democratic polity than the

instrumental logic of means-end rationality which the court deployed
and we examined above.
2.2 Rule of law: Rationality or plurality?
Tackling the persistent sovereign return can be undertaken through

the court examining the rationality of the decision. We have observed
the limits to this approach, and will further explore it below. If
plurality is to be at the heart of a democratic and participatory polity
as an alternative approach, what will anchor it in the juridical order?
In this section I explore the deeper limits to the rationalist approach
as well as point towards how plurality might be installed at the heart
of the rule of law guaranteed by the constitutional order. Through the
notion of plurality I wish to deepen the constitutional grounds for
delimiting sovereign power. The epistemic resources for this move
might be drawn from tradition, custom, ubuntu, as well as a critical
approach to judgment and responsibility.
Albutt fails to reflect the ontological plurality (of law and political
existence) when giving content to the rule of law in the South African
setting. The exercise of the pardon power was certainly linked to the
need to build a new polity and achieve reconciliation. This approach,
reflected in the amnesty process of the TRC, was the relatively unique
compromise that was struck in the negotiations that lay the ground for
a post-apartheid order. But the TRC process need not have been the
only way to attend to the ‘unfinished business’ of reconciliation.
Indeed, a new and different process may have been more apt for
dealing with a large number of incarcerated persons languishing in
prisons while the state and NGOs debated the ‘rationality’ of the
process. The ‘law of the law’ may well have suggested, and indeed
required, that a swift and effective process was the most apt way to
deal with the people who had applied for a pardon for political
crimes. Albutt reflects a Court exercising a notion of the rule of law
too caught up with a Kantian notion of rationality, rather than one
reflecting the ontological plurality reflected in ubuntu or other ways
of expressing the same values. However, ubuntu or an ontology of
plurality as an approach to what I have termed the ‘sovereign return’
does not necessarily lead to a different outcome in the case. The
approach of plurality may well deem consultation and the opportunity
to be heard as central to what the rule of law might require. Though
the outcome might be similar or the same, how one get’s there is
significant. In what follows I delve deeper into this argument.
Linda Meyer’s Justice as mercy contrasts the Kantian approach

with more pluralist approaches to pardon, mercy, and forgiveness.41
She begins by contrasting the retributive objection to reconciliatory
theories of justice which involve mercy, pardon, and grace. The
standard view from those who promote retribution as a basis for
punishment is that anything else is demeaning of the accused, and is
inegalitarian. Justice Clarence Thomas of the US Supreme Court
epitomises this latter view, and he has put it like this: ‘A system that
does not hold individuals accountable for their harmful acts treats
them as less than full citizens. In such a world, people are reduced to
the status of children or, even worse, treated as though they are
animals without a soul’.42 A utilitarian response to this retributivist
stance is to claim that just desert can be manipulated to produce the
best outcome for the greatest number by treating crime harshly or
leniently – whichever produces the best outcome for the greatest
number. However, there is then Justice Thomas’s Kantian objection
that people are being treated as objects and not as full moral agents
capable of making choices with consequences. As Meyer
demonstrates, responding to these arguments requires a move away
from the Kantian grounds of responsibility, ethics, and community.43
Meyer argues that reason cannot be a ground of responsibility,

ethics, and community as ‘reason itself requires a prior stance of
being with others’. This political condition of ‘being-in-common’ is
akin to the ‘law of the law’ to which I have alluded to above.44 With
this move towards the notion of ‘being with’ as ontologically prior to
rationalist orders of law and justice (or means-end logics), Meyer
argues that mercy is not the exception to law (as Paul Kahn has put
it) or a practical compromise (as utilitarians would have it), but in
fact the ground of justice: ‘the basis on which justice itself is
possible’. A simplified Kantian position, she argues, has become a
legal ‘catechism’ or ‘Kanticism’. If we understand ‘reason’ as a
system of logically consistent rules, then a community or law that
operates according to reason, or in a reasonable manner, might be
thought of as a community that is held together by reason.45 This
Kantian position is a conception of ‘justice as reason’, and of
community as held together by reason:46
The ideal of justice as a system of rules is deeply embedded in our legal

system. Reason is the touchstone for law. Irrational laws are
unconstitutional; irrational people are not criminally responsible.
Differences in treatment must be either explained as reasonable or
41 L Meyer Justice as mercy (2010).

42
Meyer (n 41 above) 2.
43 Meyer (n 41 above) 3.
44 See J-L Nancy Being singular plural (2000).
45
46 As above.
eliminated. ‘Interest’, ‘feeling’, or ‘opinion’ is not universal and

therefore not a reason; selfish prudence is not a reason.
Meyer draws on the later Kant of the Critique of judgment, as well as

Heidegger and Levinas, to give an account of the relationship between
justice and grace. The problem with Kanticism is that it makes us
believe that reason arises before our connection to each other. But
reason is not the ‘glue’ that links one person to another. Meyer argues
that mercy does not follow rules of reason and cannot be
universalised. Once the switch to Heidegger and Levinas is made — it
is ethical community that is the given, and reason that is a derivative.
Our being-with is not derived from reason, it is ontologically prior.
Community is not derived from being reasonable, it is already given.47
‘Justice as reason’, and reason as the ‘glue’ that holds community

together is the view that Meyer challenges. As she explains, the
Kanticism of law is based on Kant’s Groundwork on the metaphysics
of morals which set out propositions that have become deeply
embedded in legal systems. Kant’s motivation for making reason the
foundation of ethics and community was driven by an urge to free
human will from causal necessity.48 Principles of logic and consistency
were thought to free reason from the ‘arbitrary power of nature’s
relations of cause and effect’.49 So for Kant: ‘Everything in nature
works according to laws. Reasonable [vernunftiges] beings alone have
the faculty of acting according to the conception of laws, that is
according to principles, ie have a will’.50 Once the will is freed of the
impulses of nature, its actions can be made to conform to a universal
law. From this is derived the first categorical imperative of always
acting in accordance with reason. Once reason is universal and an end
in itself, and like cases are treated alike — we can move to the second
categorical imperative that reason in oneself or others is an end in
itself and not a means to an end.51 Reason then becomes the basis of
freedom, responsibility, and community.52 An ethical state and
harmonious community are formed on the basis of reason and its
‘kingdom of ends’.53
Kant’s account of judgment was driven by the conviction that

reason is an assumption – and that the gap between reason and nature
needed to be bridged. The application of rational principles to the
world is what Kant called ‘judgment’.54 This judgment cannot be
48
49
As above.
50 I Kant Groundwork on the metaphysics of morals (1785) 40 cited in Meyer (n 41
above) 11.
52 As above.
53
brought under a rule — for if it were a rule it would in turn demand

guidance from judgment.55 Judgment is thus a practiced talent and
cannot be taught.56 Imagination is an ‘arresting faculty’ that makes
us ‘passive’ so we can have the space to take in the world when
making a judgment:57
Our desire for order, pattern, form, and simplicity, for a world that
makes sense as though it were designed by an intellect like our own, is
the aesthetic core of judgment. Kant notes that when we discover such
unities, we experience spontaneous pleasure and satisfaction, common
to all human beings.
The sublime, the sense of awe, ‘the feeling of disjunction between

ourselves and the world, the feeling of not being at home in the
world, the feeling that the world is beyond our power to
comprehend’58 gives us an intimation of infinity. We can intuit this
idea even though we cannot experience or understand it. Justice is an
encounter with this sublime, but it ‘evades our grasp’.59
Meyer claims that we need an account of the ‘we’ in order to care

for the merely finite ‘human’.60 This account of the ‘we’ in South
Africa is a problem of origin and inauguration as I pointed out above.
The Constitutional order, as well as the approach to reconciliation,
mercy, and forgiveness adopted at the time of transition remain
contested. There are renewed calls for a constitution that is apt for
present needs rather than averting the crisis of transition in the early
1990s. The transitional regime had a specifically African or Bantu
inflection in that ubuntu was placed at the heart of the ethos of
reconciliation, as well as being central to key early decisions of the
Constitutional Court.61 Being-with each other, and continuing to live
as if South Africa belongs to ‘all who live in it’ is central to being and
becoming in post-apartheid South Africa. To the extent that the Court
in Albutt reduces the processes of amnesty and the work of the TRC
to a participatory model and a rational means of accomplishing the
ends of reconciliation, they distort the philosophy that informed the
transition. They forget the grounds of community. The new
constitutional dispensation was informed by a conception of
community not as communitarian essence, but a dynamic form of
being-with which should continue to inform the decisions of the
Court.
55
As above.
56
58
61
See S Motha ‘Archiving colonial sovereignty: From ubuntu to a jurisprudence of
sacrifice’ (2009) 24 SA Public Law 297-327.
Drucilla Cornell and Karin van Marle have commented on how

ubuntu might be regarded as an interactive ethic that stands behind
the law in the ‘new’ South Africa.62 On their account, ubuntu is not
only an account of being or existence. It is also an ‘ontic orientation
in which who and how we can be as human beings is always being
shaped in our interaction with each other’.63 Ubuntu is distinguished
from communalism or communitarianism — terms that suggest the
privileging of community over the individual. For Cornell and van
Marle what is at stake in ubuntu’s ontic orientation is the ‘process of
becoming a person’, and how one is given a chance to become a
person.64 Community is not some static entity ‘outside’ the
individual: ‘The community is only as it is continuously brought into
being by those who “make it up”’.65 Cornell and van Marle explain
how this ontic orientation of ubuntu can be deployed so that freedom
can be understood as indivisible. With the Constitutional Court’s
decision in Makwanyane in mind, they explain how a society that
allows the death penalty institutionalises a form of a vengeance as the
field in which we must all operate. A conception of freedom drawn
from ubuntu ‘is not freedom from; it is freedom to be together in a
way that enhances everyone’s capability to transform themselves in
their society’.66 Given ubuntu is an ‘ontic orientation within an
interactive ethic, it is indeed a sliding signifier whose meaning in
terms of a definition of good and bad is always being re-evaluated in
the context of actual interactions, as these enhance the individual’s
and community’s powers’.67 While some might call this imprecise,
unpredictable, or a dangerous basis on which to curtail state violence
(such as the death penalty), Cornell and van Marle argue that the
‘bloatedness of ubuntu’ is actually its strength.68 One person’s
freedom may still be destroyed by the community. This will endure as
long as there are competing freedoms, and especially in the realm of
punishment. But ubuntu is an African principle of transcendence
which provides a mode of attending to the moral fabric of an
aspirational community.69
In a recent book Cornell and Muvangua have asserted that ubuntu

should be regarded as the constitutional Grundnorm, or the law of the
law in South Africa, and one that is distinct from the heavy Kantian
62
D Cornell & K van Marle ‘Exploring ubuntu: Tentative reflections’ (2005) 5 African
Human Rights Law Journal 195 205.
63
As above.
64
Cornell & Van Marle (n 62 above) 206.
65 As above.
66
Cornell & Van Marle (n 62 above) 207.
67 As above.
68 As above.
69
D Cornell ‘Ubuntu, pluralism, and the responsibility of legal academics to the new
South Africa’ (2009) 20:1 Law and Critique 43 48.
inflection that this notion usually receives.70 Regretting the fact that
the concept of ubuntu is not explicitly present in the Constitution of
1996, Cornell and Muvangua nonetheless draw on John Murungi,
Justice Mokgoro, Justice Sachs and others to argue that ubuntu is the
law of the law.71 It is, they claim, the unifying motif of the Bill of
Rights, and evident in a range of court decision from the Magistrates
courts to the Constitutional Court.72 Ubuntu is not just a ‘traditional’
aside, but a clearly operational, structured, and institutional notion
that gives shape to what is meant by justice and reasonableness.73
What are the implications of this institutionalisation? What will
become of ubuntu when it moves from the realm of grounds to that of
reason? Cornell and Muvangua do not address these latter questions.
How will ubuntu avoid the fate of rationality and its inevitable means-
ends logics discussed in the first part of this essay? A deeper answer
to these questions has been provided by Mogobe Ramose. Cornell and
Muvangua’s failure to discuss his work in their textbook on ubuntu and
the law is a regrettable omission.
As Ramose has comprehensively argued, and I have discussed

elsewhere, the whole-ness that the philosophy of ubuntu is supposed
to inspire is not the absolute of community-as-law or
communitarianism.74 Rather whole-ness through ubuntu is the
recognition that be-ing is not fragmented as the subject/noun ‘be!’
as it is in (some) western ontologies.75 Ubuntu philosophy undoes the
abstract human subject of western legal thought.76 It eschews the re-
presentation of the subject as the abstract representation of the
‘subject-verb-object’ structure of language/law. It does so by de-
centring the nounal subject from the fragmentation subject/object.
African law:
is law without a centre since the legal subject here is an active but
transient participant in the be-ing, that is, the musical flow of law ...
Ubuntu law is not only the ontology of the do-ing subject. It is
contemporaneously the epistemology of the dicern-ing subject
continuously harmonising the music of the universe. In this sense,
ubuntu philosophy of law is a dynamology. Law here is thus dynamic
because it is in the first place rheomodic.77
70 D Cornell & N Muvangua (eds) Ubuntu and the law: African ideals and
postapartheid jurisprudence (2011) 20.
71
Cornell & Muvangua (n 70 above) 18.
72 As above.
73
Cornell & Muvangua (n 70 above) 25.
74 See Motha (n 61 above).
75 M Ramose African philosophy through ubuntu (2002) 46-47.
76
Ramose (n 75 above) 92.
77 As above.
The subject is then not obliged to live ‘within the law’ as with the
western legal subject, but to ‘live the law’.78 The object of law
inspired by ubuntu is to maintain equilibrium.79
Being-becoming grounded on ubuntu may be a far more durable

source of ensuring the participatory nature of a democracy. If
embedded in the rule of law, it might demand that a person in the
community affected by a decision be treated inclusively before a
decision is reached. Moreover, the persistence of the sovereign return
as a feature of modern liberal democracy would be marginalised
through ubuntu’s renunciation of the ipseity discussed above. This is
the potential of plurality over rationality as a core value animating
the rule of law.
2.3 The rule of law as critique
I would finally like to turn to the problem of responsibility and

judgment that is also provoked by the means-end rationality posited
in the Albutt case. If the rule of law is to have any teeth in the face
of the sovereign return, then it must do so with a firm sense of the
primacy of judgment as the response to a crisis. To be responsible, as
I suggest here, is to exercise judgment as critique.
I will not be the first to suggest that there is something inherently

restorative or re-constitutive in the work of critique as a response to
a political crisis or demand. Wendy Brown builds an argument for
critical theory as a hope rather than a luxury in dark times.80 Through
a helpful etymological memoriam of the term ‘critique’ which derives
from the Greek term krisis, Brown disassociates critique from
negativity, and scholasticism.81 She suggests that there could be no
such thing as ‘mere’, ‘indulgent’, or ‘untimely’ critique. In ancient
Athens krisis was ‘a jurisprudential term identified with the art of
making distinctions’.82 These distinctions were essential for arriving
at a judgment and rectifying disorder in democracy. Krisis was a scene
where the object, agent, and process for judging were intermingled
as the defendant might at once be a citizen and member of the
Senate: ‘Procedurally, juridical krisis thus consisted of recognising an
objective crisis and convening subjective critics who then passed a
critical judgment and provided a formula for restorative action’.83
This is a far cry from the idea of a critic as disinterested, or radically
opposed to the system or social organisation. There are of course
78 Ramose (n 75 above) 93.

79
Ramose (n 75 above) 93-94.
80 W Brown Edgework: Critical essays on knowledge and politics (2005).
81 Brown (n 80 above) 5.
82
As above.
83 As above.
many political and philosophical reasons to be suspicious of critique

as the restoration of a fractured or adulterated whole. But we still
cannot get away, as Brown insists, from the task of distinguishing
‘true from false, genuine from spurious, beautiful from ugly, right
from wrong’.84 We weigh pros and cons, we judge evidence — that is
to say, we deliberate. So judgment, distinction, and deliberation are
essential aspects of critique.
Can critique (in the form of judgment) ever be untimely —

especially in the context of a judgment that concerns the pardoning
of persons who have been incarcerated for many years? Does critique
renew the time that calls it forth? Does it go past the event, the crisis
that was the instance of critique? Brown eschews the notion of
‘untimely critique’, and affirms a critical enterprise directed at
‘setting the times right again’, for instance by repairing a ‘tear in
justice’.85 In this mode there is a clear restorative element to critique
— a repairing that would sit comfortably with the exigencies of ubuntu
discussed above. The very possibility of going forward demands that
the contours of the crisis be articulated, that it be inscribed as a
problem of justice. Critique is then the response, the gesture,
however imperfect and unjust, that decides and distinguishes
between a variety of infinite responses. With the force of a (critical)
distinction, the finding and inscribing of a difference, time is adjusted
and a new condition of plurality is made possible. We know from
Derrida that these adjustments, these legal or other decisions are
always already unjust, deconstructable.86 But the finitude of the
event of critique exposes the crisis, the tear in justice. The time of
judgment as critique is, then, at once finite and infinite. The
impossibility of a just response (a tear in justice that cannot be
repaired), renews critique through the permanence of the crisis, the
‘again and again’, which marks the enormity of the demand for
justice.
Also at stake in critique is the status of the critic, and her

relationship to the order that has fallen into crisis. This also entails a
move away from the narrowly adjudicatory sense of critique. Critique
entails a spatial dimension here — it is a matter of ‘critical distance’
from the object of critique. Socrates is the early model that Brown
invokes — a critic who distanced himself and the task of discerning
individual virtue and political justice from the domain of political and
judicial institutions. In contemporary traces of the old usage of
critique there is still:
84 As above.
86
J Derrida ‘The force of law: The mystical foundations of authority’ in G Nidjar
(ed) Acts of religion (2001) 230-298.
sustained linking of the objective and subjective dimensions of critique,

the ways in which a worldly event or phenomenon, whether a collapsed
empire or a diseased body, connects a specific condition with an
immediate need to comprehend by sifting, sorting, or separating its
elements, to judge, and to respond to it.87
While this quality of repair and restoration, sorting and separating,

may seem to pull the radical edge out of critique, what Brown takes
away from it is the sense of urgency, and the fact that a response to
the crisis is not optional. What she seeks to establish is that critique
is not merely negative or academic, nor is it dependent on some
regime of transcendent Truth.
But is critique as Brown renders it too readily posed in a juridical

grammar, and too redolent of seeking a (desperate) political
restoration? Can a body of law, the body before the law, or the body
politic be so readily restored through the devices and practices of
critique as judgment? And if we are to eschew the juridico-political
grammar of critique, why do we do this? And who is this ‘we’? Who is
the agent of critique beyond critique as adjudication, and what does
she seek? I want to suggest that in taking up these questions we must
direct attention to the instance of critique, to the event that calls
forth the critical response. It is the singularity of this event that drives
the critical encounter, informs its ethics, and demands a politics
which cannot be legislated in advance.
3 Conclusion
The Presidential pardon, the problem of mercy, the infinite

calculation that might have been open to the President in exercising
his power draws attention to a polity in crisis. The sovereign return
was no arbitrary exercise of a gratuitous will, but the response to a
tear in justice. In this essay I have argued that the means-end
approach to rationality as the court’s response to the people excluded
from the President’s deliberations hardly addresses the crisis of
democracy at the heart of this problem. A participatory democracy
will always have the problem of who counts, how many will be
counted, who will be heard. I have characterised this as the problem
of ipseity or the eternal return of the sovereign. In responding to this
crisis I have suggested that an alternative approach to plurality can
ground a critical judgment or what it means to respond to this crisis.
There will be no end to the rule of law grappling with this

sovereign return. However, in South Africa the epistemic and political
resources for attending to the impossible task of regulating the
sovereign return take multiple forms. In Albutt we witnessed the

pervasive and globally dominant discourse of rationality and
reasonableness. A postcolonial and post-apartheid judiciary would
seek to draw on epistemic and political discourses such as ubuntu to
address problems that go to the heart of the political and juridical
order. I examined how ubuntu might be regarded as the ‘law of the
law’ in South Africa — but it would only retain this status with any
meaning and significance if judges develop principles through the
multiple discourses of ubuntu rather than deploy it as an occasional
rhetorical flourish or nod to tradition. Re-treating the conditions and
possibilities of plural existence is at the heart of inaugurating a
postcolonial juridical and social order. Albutt was a case that raised
the problem of plurality and reconciliation in a direct and immediate
way. The relationship between the sovereign return and the problem
of suturing a tear in justice was clear. Rather than mustering all the
intellectual resources at its disposal the Constitutional Court opted
for the most mundane inquiry of examining the relation of means to
ends. Such banalities will prove to be no defence when the sovereign
returns again, as it has threatened to do, to assert the primacy of
populism over the rule of law.
THREE-LEVEL GAMES: THOUGHTS ON
GLENISTER, SCAW AND
INTERNATIONAL LAW*
Mkhululi D Stubbs**
[The Constitution’s] approach to adjudication requires an acceptance of

the politics of law. There is no longer place for assertions that the law
can be kept isolated from politics. While they are not the same, they are
inherently and necessarily linked.1
1 Introduction
This case comment considers certain specific effects of courts’
decisions on national politics and in international political arenas. As
we shall see, the Constitutional Court in Glenister v President of the
Republic of South Africa2 upset the accepted separation of powers
applecart in the domain of international political dynamics. Put
pithily, the decision brought the courts more pervasively into a
playing field ordinarily the domain of politicians and the political
branches.
However, in the commentary below the focus is not upon whether

this incursion is good or bad in normative terms. The objective here
is descriptive in nature. I analyse certain consequences of Glenister
in the context of existing dynamics between political negotiations and
municipal law-making in South Africa in order to better understand
what is at stake, strategically, when certain issues pertaining to
international agreements are in play. I shall contend that the
majority’s approach in Glenister is a new development in regard to
the way in which courts in South Africa approach the interpretive
* Though the errors remain my own, I am indebted to Shoneez Rugan, Daniel

Stewart, Mieke Krynauw, Raisa Cachalia, Pholani Ngwenya and two anonymous
referees for their constructive criticism and, particularly, to Stu Woolman,
without whom this paper would not have been possible.
** Associate at Bowman Gilfillan Incorporated.
1 Langa (2006) Stell LR 351 353.
2
Glenister v President of the Republic of South Africa and Others 2011 3 SA 347
CC.
137
138 Thoughts on Glenister, Scaw and international law
import and effect of international law. Whilst deferring analysis of

certain questionable features of the judgment for another day, this
comment emphasises the extent to which the courts in South Africa
have emerged as strategic actors with respect to international
political decision-making.
Two challenges exist for even this limited descriptive endeavour.

First, given how recently the judgment was handed down, an
empirical study of the extent of its real-world effect would be an
inappropriate metric. Secondly, a purely legal analysis of the
strategic effect of the majority decision would, understandably, be
similarly inapposite. Given that the real impact of Glenister is still in
its nascent stages, a meaningful discussion of the subject would
require a theoretical tool that could account for dynamic strategic
realignments across all three governmental functions. For this
purpose, a more useful model is two-level game theory.3
Two-level game theory is a method of political modelling designed

for the analysis of conflict resolution and other interactions between
domestic domains and international planes. As normally deployed,
however, the courts as political actors are left largely out of two-level
game theory. Glenister suggests that this lacuna in the theory should
be reconsidered. By comparing the reasoning applied in Glenister to
a ‘control sample’ decision — International Trade Administration
Commission v SCAW South Africa (Pty) Ltd4 — I hope to illustrate the
broad scope of courts’ discretionary influence in the domestic-
international political process. Once we have delineated the ambit of
this influence, its effect will be discussed in the context of two-level
game theory, which, in turn, will enable us to speculate upon
Glenister’s potential consequences with respect to decision-making in
the other branches of government. I will then argue that the emerging
role of the courts is sufficiently pervasive to justify the claim that
they sit at a third and separate level through which they interact with
players at the other two levels. In the final section of this paper, I will
set out at least one hypothetical legal consequence of the majority’s
decision by transposing the reasoning applied in Glenister to the
factual matrix presented in Government of South Africa v
Grootboom. The results of this second comparison are both striking
and illuminating.5
The comment is split into three parts. Part II begins by setting out
the South African legal framework insofar as it concerns the
application of international law on the domestic plane. Thereafter, it
3 RD Putnam ‘Diplomacy and domestic politics: The logic of two-level games’

(1988) 42 International Organization Foundation 427.
4
2010 5 BCLR 457 (CC).
5 2001 1 SA 46 (CC).
proceeds to examine the Court’s holdings in both Glenister and Scaw.

Part III explains the ‘mechanics’ of two-level game theory and
describes some of its applications. In the light of conclusions drawn in
Parts II and III, Part IV then examines the cogency of a three-level
claim, as well as the potential import of such a claim for South African
constitutional politics.
2 The Constitution and relevant case law
2.1 The South African legal system for international

incorporation
The process of concluding an international agreement is fraught with

many concerns and complexities, not least of which is a satisfactory
understanding of the likelihood that the ‘other side’ will do as it says
it will do. There is no point in expending the political capital required
to propel a given country (‘Country A’) into negotiations, only to
negotiate with another country (‘Country B’), which lacks the
inclination, or the power, to give effect to Country A’s legitimate
demands.
In an international negotiation, where the issue regarding the

negotiator sitting across the table concerns her inclination to
implement an agreement, the question is one of trust, which may be
informed by the previous conduct of Country B’s negotiator, or of the
country that she represents (amongst other informative factors). But
where the issue is whether one’s negotiating counterpart has the
power to give effect to an agreement (if one is reached), then the
question becomes one of credibility, which may be informed by the
legal mechanics of incorporation applicable to international
agreements in Country B. As the section below will show, even the
promises or assent of the most trustworthy negotiator, who herself
has the best intentions, will lack credibility if domestic
implementation of the agreement in that negotiator’s country may
subsequently be stymied by legal rules or political processes
(depending in part upon whether Country B’s legal system tends
towards a ‘monist’ or a ‘dualist’ system of incorporation). When the
potential political cost of failed negotiations is weighed against the
potential benefits of a successfully implemented international
agreement, clearly all countries must have an interest in
understanding the municipal legal systems of their negotiating
counterparts.
Understanding the municipal pitfalls to implementation in South

Africa is no simple task. As John Dugard explains, the relationship
between international and municipal (or ‘domestic’) law ‘troubles
both theorists and courts’.6 Traditionally, there are two approaches

to the incorporation of international agreements, the monist
approach and the dualist approach.7 In reality, however, no country
ever follows an approach that exemplifies a pure model of either.
On the dualist approach, a clear dividing line must be drawn

between international law and domestic law. International covenants
entered into by sovereign states are binding on those states in the
international arena, but that they are not effective as municipal law
without additional domestic legislative action. As an example,
imagine that New Zealand (or some other nation, selected at random)
enters into an international covenant,8 and one of the terms of the
covenant is that New Zealand will incorporate certain obligations of
that covenant into its domestic law. If New Zealand does not
incorporate the terms of the covenant as promised, then the
institutions in the country that ought to be required to act in
accordance with the covenant will, in fact, not be bound by them. If
New Zealand’s domestic institutions fail to comply with the terms of
the international covenant (due to a failure to incorporate the
covenant, or for other reasons), New Zealand would be in breach, and
the legal consequences of the breach would be brought to bear upon
New Zealand on the international plane. The result: New Zealand
would be excluded from the benefits of the covenant entirely. Take a
simple illustrative example. If, in an international negotiation, New
Zealand agreed with South Africa to the mutual lowering of
agricultural trade barriers, but New Zealand ultimately failed to
lower its trade barriers as a result of its inability to implement the
negotiated agreement successfully, then the terms of the agreement
might allow South Africa to withdraw future concessions and also to
recover value for any concessions already provided.
Under the monist model, international law and domestic law are
far less distinct. Take the aforementioned example. If New Zealand
were a monist state, agreement to the same international covenant
would automatically transform the terms of that agreement into
binding obligations on the domestic plane. In other words, the
6
J Dugard International law: a South African perspective (2000) 43. For an
examination of the debate see, inter alia, JG Starke ‘Monism and dualism in the
theory of international law’ (1936) 17 British Yearbook of International Law 66;
and C Roodt ‘National Law and Treaties’ (1987-8) 13 South African Yearbook of
International Law 72.
8
There are, of course, many different kinds of international agreements. Some
agreements may constitute an obligation in principle only. Other agreements may
take effect and command obedience immediately, as would a contract between
individual persons. Yet other agreements may come into effect at a later date, or
come into effect in parts over a period of time. For the sake of simplicity, this
paper uses the term ‘international agreement’ as having immediate effect,
subject to the qualifications imposed by the applicable international-municipal
system, whether monist or dualist.
relevant institutions of New Zealand would be bound to act in

accordance with the terms of the covenant from the moment that
there was agreement upon the international plane, and without any
act of municipal incorporation. The monist approach would allow New
Zealand to avoid contravention of the trade agreement and,
accordingly, no South African entitlement would arise.
The Constitution of the Republic of South Africa, 1996 (‘the

Constitution’) tends toward dualism (albeit a harmonised form of
dualism, which requires a certain degree of consistency between
domestic law and international law).9 As a general matter,
international obligations are treated as agreements entered into by
sovereign states, and the obligations to which those agreements give
rise are owed to the parties thereto — that is, the sovereign states —
and no one else. Disputes regarding the interpretation of these
agreements are either resolved by diplomatic means or, otherwise, by
an appropriately mandated international tribunal. In turn, obligations
which are binding at the domestic level will be those enacted by
Parliament, as supplemented by the common law. When individuals
break the municipal laws, they are beholden to the state.
Section 231 of the Constitution specifically engages international

negotiation and the incorporation of international agreements.10 The
national executive is vested with the responsibility to negotiate and
enter into international agreements. International agreements that
are not of a ‘technical, administrative or executory nature’ only
become binding upon the Republic on the international plane once
9 One example of this ‘harmonised’ or ‘mixed’ legal regime in South Africa is the
fact that, pursuant to section 232 of the Constitution, ‘South African common law
treats international law as part of [domestic] law’ (see Dugard (n 6 above) 51).
Moreover, section 232 of the Constitution provides as follows: ‘When interpreting
any legislation, every court must prefer any reasonable interpretation of the
legislation that is consistent with international law over any alternative
interpretation that is inconsistent with international law.’
10
Section 231 of the Constitution provides as follows:
(1) The negotiating and signing of all international agreements is the
responsibility of the national executive.
(2) An international agreement binds the Republic only after it has been
approved by resolution in both the National Assembly and the National
Council of Provinces, unless it is an agreement referred to in sub-
section (3).
(3) An international agreement of a technical, administrative or
executive nature, or an agreement which does not require either
ratification or accession, entered into by the national executive, binds
the Republic without approval by the National Assembly and the National
Council of Provinces, but must be tabled in the Assembly and the Council
within a reasonable time.
(4) Any international agreement becomes law in the Republic when it is
enacted into law by national legislation, but a self-executing provision of
an agreement that has been approved by Parliament is law in the Republic
unless it is inconsistent with the Constitution or an Act of Parliament.
(5) The Republic is bound by international agreements which were
binding on the Republic when this Constitution took effect.
they have been approved by resolution in both houses of

Parliament.11 For an international agreement to have domestic effect
it must be enacted as legislation. It is for this reason that it is said that
an act of incorporation by Parliament is required for any agreement
to ‘transform’ from an agreement between states into a law that
regulates people.12 Failure to incorporate an international agreement
does not render that agreement irrelevant to a court’s decision.
However its influence, if unincorporated, is certainly somewhat less
direct.13
When interpreting the Bill of Rights, a court, tribunal or forum ... must
consider international law ... [and] [w]hen interpreting any legislation,
every court must prefer any reasonable interpretation of the legislation
that is consistent with international law over any alternative
interpretation that is inconsistent with international law.
Under the Constitution, the executive wears more than one hat. Aside
from being responsible for the negotiation and execution of
international agreements, it is also vested with the authority to
develop and to implement domestic policy and is responsible for the
preparation, initiation and implementation of national legislation.14
The Constitution thereby makes provision for executive-inflected
influence on both the domestic level and the international level. On
a dualist approach, then, the executive straddles the divide that
dualism itself creates between domestic and international affairs. But
the executive’s power is limited. In order to give effect to the
agreements that the executive negotiates with its international
11 This excludes international agreements which are held to be ‘self-executing’. See

section 231(4) of the Constitution (see n 10 above). This term that has not yet
been given clear content by the courts. See President of the Republic of South
Africa v Quagliani; President of the Republic of South African v Van Rooyen;
Goodwin v Director-General, Department of Justice and Constitutional
Development 2009 8 BCLR 785 (CC). The Court did not reach the issue of deciding
whether certain terms of an international covenant were ‘self-executing’ as
contemplated in section 231 of the Constitution, as the outcome was identical on
either of competing interpretations on offer.
12
For a discussion of the adoption and transformation theories see F Morgenstern
‘Judicial practice and the supremacy of international law’ (1950) 27 British
Yearbook of International Law 42.
13
See Sections 39(1)(b) and 232 of the Constitution.
14
Section 85(2) of the Constitution. Section 85 of the Constitution provides as
follows:
(1) The executive authority of the Republic is vested in the President.
(2) The President exercises the executive authority, together with the
other members of the Cabinet, by:
(a) implementing national legislation except where the Constitution or
an Act of Parliament provides otherwise;
(b) developing and implementing national policy;
(c) co-ordinating the functions of state departments and
administrations;
(d) preparing and initiating legislation; and
(e) performing any other executive function provided for in the
Constitution or in national legislation.
counterparts, it must have the ‘buy-in’ of the electorate, in one

manner (such as an act of incorporation by Parliament through the
promulgation of legislation) or another (through the absence of a
critical mass of collective protest by affected interest groups, or,
impliedly, through the operation of a self-executing provision.)15
If the executive in a dualist system negotiates an international

agreement, and the terms of that agreement anticipate domestic
application, but the legislature fails to enact that agreement into law,
then, in theory, the executive will lose political traction on both
planes. How so? An executive that over-promises and under-delivers
may lose credibility on the international plane in future negotiations
as well as political or popular support domestically. On the other
hand, if ratification is successful, the executive may gain on both
planes. If the executive is wise (or simply strategically savvy), it will
avoid accession to international covenants unless, in its judgment,
they are likely to pass muster back home.
On a monist approach, by contrast, the political calculations will

be significantly different. In a monist system, executive agreement on
the international plane is effectively a domestic law-making function.
The downside here is that the executive has no option but to take full
political responsibility for an unpopular agreement. It will not have
the ratification of the legislature to buttress its decision.16 The upside
is that all the credit will redound to the executive. If it enters into an
international agreement which is celebrated in its home country, then
it will not need to share the resulting political spoils with its coequal
branches of government. Accordingly, in a monist system, one might
say that the political stakes of international negotiation are often
higher.
Within the South African framework (and in other jurisdictions

with relevant similarities) the relationship between the executive and
legislature is, at first blush, and to the extent that both branches of
government act in accordance with the law, an issue of purely
political interest. To leave out the third branch of government would
be a mistake. Whilst our courts do not, themselves, enter into
international negotiations or promulgate laws, they are called upon to
interpret the international agreements that are entered into and the
15 Section 231(4) (see n 10 above).

16
Nor will the executive be in a position to ‘hide behind’ the legislature — there
may be times when it would be politically advantageous for the executive to
agree in principle to terms on the international plane, knowing full well that the
legislature will not pass the agreement. If the subject matter of the agreement is
of sufficient political interest and the terms of the agreement were popular
amongst a significant proportion of the electorate, the executive may gain
political ‘kudos’ for its attempt, and the legislature, or the political parties which
it comprises may shoulder the blame for failing to pass that law.
legislation that is subsequently enacted. When courts do so, their

decisions may affect decisions made by the other branches. To the
extent that they do have such an influence, they act, at least, as
shadow players on the domestic and international levels. It will
invariably be the case that the influence of the courts is not
particularly obvious when court’s decisions align with the state’s
expectations. Scaw is an example of an invisible court, a court whose
decision reinforces the judgment of the executive (in entering into
the relevant agreement) and the legislature (in enacting the relevant
laws). Glenister exemplifies the opposite proposition: the Court calls
out the executive for its failure to abide by international agreements
and the provisions of our own basic law. Whilst Scaw and Glenister
engage very different subject matter, an analysis of the reasons for
the success of the state’s argument for the enforcement of ratified
international obligations in the one case, measured against the
reasons for the failure of the state’s argument against the application
of an un-ratified international covenant in the other should
conveniently delineate the ambit of the courts’ influence. To this
end, the two cases are discussed in greater detail below.
2.2 The cases: Scaw and Glenister
Scaw was concerned with the operation and duration of international

trade remedies. (The case was, however, ultimately decided on the
principle of the separation of powers.) At issue were certain anti-
dumping duties provided for in terms of national legislation and
enacted pursuant to international agreement. The question before
the Court was whether it was appropriate for a company to interdict
an appropriately authorised executive officer from making a
recommendation about whether or not continued anti-dumping duties
were commercially justified.
Pursuant to South Africa’s accession to the General Agreement on

Tariffs and Trade (the ‘GATT’) and the subsequent World Trade
Organization Agreement (the ‘WTO agreement’),17 the government is
permitted, in defined circumstances, to take measures to protect its
domestic producers by preventing foreign commercial entities from
‘dumping’ goods onto the domestic market. This is achieved through
17
This agreement was approved by the South African Parliament through the
Geneva General Agreement on Tariffs and Trade Act 29 of 1948. South Africa
acceded to the subsequent World Trade Organisation Agreement on 1 January
1995. The international rules relating to dumping are contained in Article VI of
the GATT and the Anti-Dumping Agreement. See Scaw (n 4 above) para 25.
the imposition of anti-dumping duties.18 Anti-dumping duties are

therefore a legitimate means, in terms of the GATT and the WTO
agreement, for a state to protect against the harm that could be
caused to domestic suppliers by the introduction of goods from
foreign suppliers at a fraction of the price at the former are able to
produce the same goods. In essence, then, anti-dumping duties
provide a temporary ‘safe haven’ for domestic producers from foreign
competition.
The salient facts of the case are as follows. SCAW South Africa
(Pty) Ltd (‘SCAW’) approached the High Court for an interim interdict
restraining the Minister of Trade and Industry from accepting a
recommendation by the International Trade Commission (‘ITAC’)
concerning the continued operation of anti-dumping duties against
one of SCAW’s major commercial competitors. That application was
successful. The effect of the High Court interdict was to restrain the
Minister of Finance from terminating the existing anti-dumping duty,
as per ITAC’s recommendation, pending the final determination of an
application to be instituted by SCAW to review and to set the
recommendation aside. ITAC took the High Court decision on appeal
that ultimately found its way to the Constitutional Court.
What is important to note about the Scaw appeal for present

purposes is the Court’s affirmation of the domestic applicability of
South Africa’s international obligations was explicitly predicated on
the fact that the GATT and the WTO agreement had been
incorporated into domestic law:19
[South Africa’s international] obligations are honoured through domestic

legislation that governs the imposition of anti-dumping duties and other
trade remedies. In the main the legislation consists of the International
Trade Administration Act, 2002 ...; the Anti-Dumping Regulations ... The
[International Trade Administration] Act is the primary domestic
legislation for controlling anti-dumping duties and other harmful trade
practices associated with international trade.
The express object of the International Trade Administration Act 71

of 2002 (‘ITA Act’) is to establish ‘an efficient and effective system
for the administration of international trade.’20 Regulation 53 of the
18
See Scaw (n 4 above) para 1. The term ‘normal value’ is, itself, malleable — but
the question as to whether or not the relevant goods were indeed priced at
‘normal value’ is beyond the scope of this note. ‘Dumping’ occurs when goods are
introduced into a country ‘or its common customs area at an export price less
than the normal value of those goods.’
19
See Scaw (n 4 above) paras 2 & 31.
20 See Scaw (n 4 above) para 31.
Anti-Dumping Regulations provides for the subsistence of anti-

dumping duties for a maximum of five years.21 In addition, Regulation
20 provides that ‘[a]ll investigations and reviews shall be finalised
within 18 months after initiation.’
Taking the above into account, the Court held that the ‘setting,
changing or removal of an anti-dumping duty ... is a patently
executive function’.22 It followed, on the Court’s reasoning, that
allowing the frustration of the expiration of an anti-dumping period
by sustaining that period throughout the inevitable delays of the
litigation process would be tantamount to allowing for an ‘indefinitely
elastic term of duties’.23 When viewed in this light, allowing the
courts to frustrate what is properly an executive policy function
would be a breach of the doctrine of the separation of powers, as the
delays through the courts would have the effect of tampering with a
‘power [that properly] resides in the kraal of the national executive
authority’.24 The interdict was thus unanimously set aside.
The decision in Glenister was not unanimous and the Court split
by a five to four margin. Mr Glenister argued that the planned
disbandment of the Directorate of Special Operations (popularly
known as the ‘Scorpions’), which was located within the National
Prosecuting Authority (‘NPA’), and its replacement with the
Directorate for Priority Crime Investigation (popularly known as the
‘Hawks’), located within the South African Police Services, was
unconstitutional. The Court was required to decide, first, whether the
state, through, inter alia,25 the effect of the United Nations
Convention against Corruption,26 had a positive duty to create an
‘independent anti-corruption unit’ and, secondly, whether the
21
Regulation 53 of the Anti-Dumping Regulations to the International Trade
Administration Act provides as follows:
53.1 Anti-dumping duties shall remain in place for a period not exceeding
5 years from the imposition or the last review thereof.
53.2 If a sunset review has been initiated prior to the lapse of an anti-
dumping duty, such anti-dumping duty shall remain in force until the
sunset review has been finalised.
22
Scaw (n 4 above) para 102.
23 Scaw (n 4 above) para 80.
24
25
The majority also referred to the Prevention and Combating of Corrupt Activities
Act 12 of 2004, citing that Act as ‘the fullest recital of the insidious scourge of
corruption on society and the need to prevent and eliminate it’. However, the
Court dismissed the question of domestic incorporation as irrelevant as, in its
view, the enactment of the United Nations Convention into domestic legislation
could not, in any event, give rise to constitutional obligations (See Glenister (n 2
above) para 3). In this sense, this observation is irrelevant to the question of
incorporation and, for that reason, is not discussed in further detail here.
26
See http:///www.unodoc.org/documents/treaties/UNCAC/Poblications/Conven
tion/08-50026_E.pdf (accessed 14 February 2012). The UN Convention was
adopted on 31 October 2003 and entered into force on 14 December 2005. South
Africa signed the Convention on 9 December 2003 and ratified it on 22 November
2004.
creation of the Hawks as the extant anti-corruption unit, and its

substitution for the Scorpions, constituted the fulfilment of that
duty.27
The minority decision, written by Ngcobo CJ, would have

dismissed the applicant’s challenge (largely) on the following
grounds:28
The Constitution is not prescriptive ... as to specific mechanisms through

which corruption must be rooted out, and does not explicitly require the
establishment of an independent anti-corruption unit ... Lest I be
misunderstood, while I am prepared to hold that there is a constitutional
obligation for the state to take effective measures to fight corruption, I
am not prepare to narrowly construe the options available to the state in
discharging that obligation.
As this passage shows, the Chief Justice was indeed cognisant of the
fact that relevant international covenants had indeed been entered
into by the executive.29 He wrote, as a general proposition, that,
when viewed in the light of sections 233,30 39(1)(b)31 and
37(4)(b)(i),32 international agreements, and especially those
concerning human rights, are of significant interpretive value in the
resolution of disputes. However, Ngcobo CJ emphasised that this did
not mean that such agreements could ‘create rights and obligations in
the domestic legal space’.33 Whilst it was true that international
agreements were not meant to be ‘merely platitudinous or
ineffectual’,34 to allow such unincorporated agreements to create
domestic rights and obligations would be ‘tantamount to
incorporat[ing] the provisions of the unincorporated convention into
our municipal law by the back door.’35 Accordingly, in the absence of
27
28
29 Scaw (n 4 above) para 95.
30
Section 233 of the Constitution provides as follows:
When interpreting any legislation, every court must prefer any reasonable
interpretation of the legislation that is consistent with international law
over any alternative interpretation that is inconsistent with international
law.
31 Section 39(1)(b) of the Constitution provides as follows:
(1) When interpreting the Bill of Rights, a court, tribunal or forum:
...
(b)must consider international law.
32
Section 37(4)(b)(i) of the Constitution provides as follows:
(4) Any legislation enacted in consequence of a declaration of a state of
emergency may derogate from the Bill of Rights only to the extent that:
... (b)the legislation:
(i) is consistent with the Republic’s obligations under international law
applicable to states of emergency.
33
34 Minister of State for Immigration and Ethnic Affairs v Teoh (1995) HCA 20; 1995
[183] CLR 273 291.
35
n 34 above, para 98. Also see Minister of State for Immigration and Ethnic Affairs
v Teoh [1995] 183 CLR 273 286-287.
a directly applicable international standard, the minority would have

held that the impugned state action was constitutionally permissible.
The majority of the Court disagreed. It held that, in the absence

of an explicit domestic-law definition, the standard of ‘necessary
independence’ from the executive, which was the majority’s
yardstick for the constitutionality of the legislation, found its source
both in international law and in the Constitution.36 The majority
opined that the international standard was relevant because the
Constitution demanded as much: ‘necessary independence’ was
drawn ‘into the very heart’ of the Constitution,37 and, in this way,
became the ‘[domestic] measure of the state’s conduct in fulfilling its
obligations in relation to the Bill of Rights’.38 The Court thus held that
it was the Constitution (and not international law) that created a
positive obligation on the state to establish and to maintain an
independent body to combat corruption and organised crime.
Moreover, in the view of the majority, the Hawks lacked the
‘necessary independence’39 to discharge that obligation effectively
and, accordingly, their substitution for the Scorpions was held to be
unconstitutional.
The success of Mr Glenister’s appeal reflects an unusually

influential role for unincorporated international law in the domestic
legal space. The Court created a third way whereby international
obligations could have overriding domestic legal effect through, inter
alia, the interpretive injunction in section 39(1)(b) and its
concomitant effect on section 7(2) of the Constitution, the latter of
which concerns the positive obligations of the state. Ostensibly, the
state was held to a constitutional standard, but the standard was, in
reality, no different to than the international benchmark.
So we might say that, in Scaw, the executive was ‘freed up’, and,
in Glenister, it was ‘beat up’. How to explain this difference in
outcome? In Scaw, the relevant duty of the executive was the
regulation of economic policy, and the relevant range of permissible
state conduct was the discretion to impose or to eliminate anti-
dumping duties at its reasonable discretion, were aligned. So the
executive achieved a ‘win’. In Glenister, on the other hand, the state
lost because its conduct was held to fall short of an internationally-
informed constitutional standard. If we presume that the state would
have preferred to avoid the time and the cost inevitably incurred in
36
The Court relied upon its prior holding in S v Makwanyane 1995 3 SA 391 (CC)
paras 34 - 35. It had concluded in Makwanyane, and now reaffirmed, that it was
entitled to ‘consider both binding and non-binding instruments of international
law’.
38
As above.
39 As above.
bringing its anti-corruption measures within the prescribed standard

of independence. In other words, if we can presume that the state
would have an interest in ‘winning’ in future, then we might speculate
as to how it might best go about doing so. But to do so in an informed
manner would first require a better understanding of the overall
strategic framework within which the state operates. This framework
is the subject of Part III below.
3 Putnam’s two-level game theory
3.1 A metaphorical framework
As a matter of law, an examination of the effect of a judgment or a

decision in the arena of international negotiations is not relevant to
the adjudication of a dispute, even if the dispute has international
implications. A judge may not, for example, shrink away from finding
against a company incorporated in a sovereign state which is a
potential trade partner with the state of that judge on the basis of an
apprehension, however well-founded, that her adverse finding would
be harmful to the prospects of the conclusion of a bilateral
investment treaty with that state in future.
However, whilst courts should not be partial to government

litigants in interpreting the law, the executive and the legislature
remain the organs of state who, for the most part, make the law that
the courts interpret. If effective law-making is a legitimate
component of effective governance, then the non-adjudicative
branches of government, namely the executive and the legislature,
will obtain more efficient outcomes if they play the political game
‘within the white lines’. For when a court overrules government
action which has been taken in misapprehension of the law, the
executive and legislature are sent back to the drawing board, and all
of the time and the resources spent by both branches of government
go to waste.
Given the courts’ duty of qualified disinterest in regard to the

practical effect of their judgments in the political arena, a purely
legal analysis of this effect would be of limited utility. A more useful
model would set out a theoretical framework by which the practical
effect of these decisions, once made, may be measured.
Game theory is well-suited for this purpose. It involves ‘the study

of mathematical models of conflict and cooperation between
intelligent rational decision-makers’,40 or, in other words, the study
40 R Myerson Game theory: Analysis of conflict (1991) 1.

of hypothetical games. In any given game, the theory postulates a

defined number of players, a number of moves that may be made by
those players, payoffs that flow from certain outcomes which
themselves are often influenced by the independent choices made by
other players in the game.
There are zero-sum games and non-zero-sum games. Take the

example of a game of chess, where one player pits her wits against
another. Both players will be aim to win the game, and for one of the
players to win, the other player will inevitably have to lose. (Let’s
bracket, for argument’s sake, the possibility of a draw.) These
conditions obtain even if a third party (‘party C’) provides prize
money (say R100) is brought into the example. Neither advantage nor
opportunity for cooperation exists that would improve the outcomes
for both players, so the match is a zero-sum game. It is ‘winner takes
all’. (So even where draws are permitted in a championship (and each
player is awarded a half-point), both players still aim to win the
minimum number of games necessary to become the champion.)
Suppose now, instead, that the chess players were playing for prize
money of R100 to be given by party C and, in addition, party C acted
as a bookie, taking bets on the result. Suppose that party C ‘fixed’ the
game and bribed both chess-players to play to a stalemate, promising
to pay them each R150 for the tied result. Both chess-players will have
more to gain from playing to a draw (which would earn them R150
each, plus half of the prize money for a total of R200) than they could
gain from winning (the winner would take home R100) or, obviously,
from losing. In addition, party C will enjoy the benefit of taking bets
on odds that are favourable to what he knows to be a predetermined
result, thereby maximising his own profits. The game has now become
a non-zero-sum game because if the chess-players to cooperate with
party C, then they will achieve an outcome that is better for all of
them in the aggregate. In short, and because the game is non-zero-
sum, they can achieve more together than they would alone.
Two-level game theory,41 which is derived from game theory, is

a political model for international conflict resolution. If we use the
chess-player analogy above, then two-level game theory can be said
to concern games that involve interactions and potential synergies
between players across two game-boards, instead of one. The game-
boards in the metaphor represent the national and international
levels of political interplay. Given the contrasting dynamics between
international negotiations and domestic politics, the type of game
41
See Putnam (n 3 above). For the purposes of this hypothetical it is assumed that
negotiations involve only one issue, as opposed to being a multi-issue negotiation.
being played at each board will be different.42 As Putnam explains:43
At the national level, domestic groups pursue their interests by

pressuring the government to adopt favourable policies, and politicians
seek power by constructing coalitions among those groups. At the
international level, national governments seek to maximise their own
ability to satisfy domestic pressures, while minimising the adverse
consequences of foreign developments. Neither of the two games can be
ignored by central decision-makers, so long as their countries remain
interdependent, yet sovereign.
Consider the following hypothetical scenario as an example (‘the

original hypothetical’).44 When the Presidents of South Africa and
Brazil (collectively, ‘the Presidents’) meet in order to negotiate an
international agreement, they are both subject to the constraint that
any agreement that they reach will have to be ratified by their
respective constituents (for the sake of this example, we also assume
that Brazil, like South Africa, follows a dualist approach to
international law).45 If we assume further that the Presidents
themselves have no independent policy preferences other than to
achieve a ‘ratifiable’ agreement (as that term is defined by
Putnam),46 then the Presidents’ negotiations can be separated into
two stages: (a) the bargaining between the Presidents which might
lead to a tentative agreement at the international level (‘Level 1’);
42 Putman (n 3 above) 435. Putnam phrases his argument in terms of two levels for
the sake of ‘simplicity of exposition’ (449). He recognises that ‘many institutional
arrangements require several levels of ratification, thus multiplying the
complexity of ... analysis’.
43
Putnam (n 3 above) 434.
44 This scenario is an adaptation of an example expounded in greater detail by
Putnam himself, (n 3 above, 435). Only the names have been changed for the
purposes of this paper.
45 This hypothetical assumes, for the sake of simplicity, that the Presidents are the
sole negotiators of any international agreement that might result.
46
‘Ratification’, for the purposes of Putnam’s theory, may include a formal voting
procedure at Level 2, such as an act of incorporation envisioned in section 231 of
the Constitution. But it may also include any other decision-process at Level 2
that is required for the implementation of a Level 1 agreement, whether formally
or informally. So, trade unions may ‘ratify’ an agreement if they cooperate as
agreed at Level 1, and ‘ratification’ may be said to have failed where such
cooperation is withheld. The process of ratification need not even be democratic
on Putnam’s theory, and it is used generically here. See Putnam (n 3 above) 436.
There is one formal constraint on the ratification process: ‘ ... since the identical
agreement must be ratified by both sides, a preliminary [Level 1] agreement
cannot be amended at Level 2 without reopening the Level 1 negotiations. In
other words, final ratification must be simply ‘voted’ up or down; any
modification to the Level 1 agreement counts as a rejection.’ (437). Putnam also
discusses credibility-related strategic implications at Level 1 to what he calls
‘voluntary’ as against that produce a given product (say titanium, for example)
‘involuntary defection’ at Level 2 (438), especially where negotiators are
engaged in a repeated game (ie, they will negotiate again in future). These
implications are, however, beyond the scope of this note and are not discussed
here.
and (b) the process of engagement and ratification at the domestic

level (‘Level 2’).
3.2 ‘Win sets’
Where there is successful agreement at Level 1, that agreement can

be called a ‘win’. The same will apply for successful ratification at
Level 2. A ‘win-set’ can thus be said to occur where there is a win at
both Levels. If we imagine Level 1 and Level 2 as game-boards, we can
surmise that the Presidents will both appear at each other’s game-
boards (at Level 1), and also that they will interact directly with their
respective domestic constituents at Level 2. For this reason, when the
Presidents negotiate, their decisions will be steered by their
perception of the preferences of relevant domestic interest groups.
Suppose, in the original hypothetical, that the President of South
Africa wants to enter into a free trade agreement concerning titanium
and aircraft manufacture with the President of Brazil. At the Level 1
game-board, such an agreement might be a rational move for the
Presidents to make. Brazil will have an interest in the lowering of
trade barriers, as its industries, which could form a powerful domestic
interest group in themselves, will gain access to foreign markets and
Brazil’s economy might benefit through the consequent creation of
jobs. Furthermore, this agreement could result in a political gain for
the President of Brazil in an upcoming political campaign for re-
election to office. The President of South Africa would expect
reciprocal benefits of some sort — the lowering of barriers in
agriculture, or perhaps skills-transfer programmes in aircraft design
or manufacture. In order to reach agreement, however, the
Presidents will need to make trade-offs as well. For South Africa, one
trade-off might be the entry into the South African market of Brazilian
companies with better economies of scale than some of South Africa’s
incumbent producers of the same product. The market presence of
these Brazilian companies would constitute a commercial threat, as
well as a threat to the employees of those incumbent producers. If
these ‘threatened’ employees were unionised, and their respective
trade unions comprised a powerful domestic interest group,
collectively, then the President of South Africa would be faced with a
far more complex choice. Indeed:
[t]he political complexities for the player in [a] two-level game are
staggering. Any key player at the international table who is dissatisfied
with the outcome may upset the game board, and conversely, any leader
who fails to satisfy his fellow players at the domestic table risks being
evicted from his seat.47
47 Putnam (n 3 above) 434.

So, in the original hypothetical, it would follow that the more

powerful the trade unions (all other things being equal),48 the more
constrained the discretion of the President of South Africa will be on
the international plane. Where the trade unions are powerful, the
President will be averse to upsetting the trade unions for fear of the
political consequences (we might refer to this outcome as the ‘result
of the original hypothetical’). Conversely, if the agreement at Level
1 pertains to something else entirely, say, for example, the importing
or the exporting of wine, and if the winemakers and their employees
are not a powerful domestic interest group, then the number of
available win-sets for the President of South Africa will be larger than
it would have been in the original hypothetical. Larger win-sets at
Level 2 therefore make Level 1 agreement more likely. A large range
of Level-2 wins will augur very well for international cooperation.
In the same vein, it can also be said that the relative size of
ratifiable Level 2 win-sets (or, alternatively, the President of Brazil’s
perception thereof) will affect the distribution of the joint gains
available from the international bargain.49 As a corollary to the
flexibility allowed by the absence of resistance at Level 2, the greater
the number of win-sets for a given negotiator at game-board Level 1,
the more easily that negotiator may be ‘pushed around’.50 So while it
is true that a large number of potential win-sets will make
international cooperation more likely, it is equally true that if these
win-sets are unevenly distributed — that is, if one negotiator holds
more win-sets than the other — the negotiator with the greater
number of win-sets in hand will have the weaker bargaining power.
Putnam illustrates this in the following way:
In a simple zero-sum game, and on the facts of the original

hypothetical, let us imagine that X represents South African interests,
and Y represents Brazilian interests. Xm and Ym in Figure 1 will then
represent the best outcomes for each set of interests, and X1 and Y1
48
To prevent repetition, the phrase ‘all other things being equal’, or ceteris
paribus, will not be repeated for each proposition that follows.
49 This proposition assumes, for the purposes of this paper, that there is perfect
information as between negotiators.
will represent the countries’ respective bottom lines: the point

beyond which any agreement will be un-ratifiable at Level 2.
Accordingly, the distance between Xm and X1 and Ym and Y1 will
represent the spectrum of ratifiable win-sets available to each
negotiator:
Recall the result of the original hypothetical, which was a vociferous

and negative reaction from the trade unions of South Africa
(country X) to the news of negotiations with the President of Brazil
(country Y). As a result of this reaction, South Africa’s overall
inclination at Level 2 to enter into an agreement might feed through
to Level 1 and cause the contraction of the range of ratifiable win-sets
from X1 to, say, X2. Accordingly, agreement upon the subject-matter
represented by the space between the two points would no longer be
feasible. And this contraction would simultaneously strengthen X’s
bargaining position:
Now suppose that instead of attempting to pass through an

agreement pertaining to titanium and aircraft manufacture, as we did
in the original hypothetical, the President of South Africa changed
tack at Level 1 and negotiated terms that made trade-offs in an even
more politically and economically volatile zone at Level 2 in South
Africa, such as gold production. Because of the political strength in
that area of the mining sector, the range of ratifiable win-sets for X
might be even more severely contracted than it was in Figure 2. If the
scope of agreement is reduced to, say, X3, then the negotiations are
deadlocked, and no agreement is possible (‘the second
hypothetical’):
As depicted in Figure 4 above, the President of South Africa is

therefore constrained from assenting at Level 1, because it would be
impossible to ratify at Level 2. This outcome explains (or supports)
Putnam’s assertion that ‘the stronger a state is in terms of autonomy
from domestic pressures, the weaker its relative bargaining position
internationally.’51
‘Autonomy’, in the sense used by Putnam, is a descriptive term

which quantifies the executive’s independence from the parochial
interests of its domestic constituents. Often, if not in the majority of
cases, the executive will not be particularly independent. But this is
not, in principle, a bad thing. It only means that the political party
whom the given President represents is not so firmly entrenched that
his political mistakes will not have consequences at the ballot box. A
state’s ‘bargaining position’ and its ‘autonomy’ are thus interrelated.
And, as we have seen, the more autonomy or independence a state
enjoys from domestic pressures, the more easily it will be pushed
around at Level 1. Accordingly, albeit somewhat perverse, although
negotiators at Level 1 will have a strong interest in the popularity of
their fellow negotiators, the more autonomous the President of South
Africa is (in the sense that he is free of domestic pressure), the more
likely Brazil is to reach an agreement on more favourable terms

(because this lack of domestic pressure will weaken the President of

South Africa’s bargaining position).52
Furthermore, and as if the decision-making terrain were not

complex enough already, the entanglements between Level 1 and
2 may have bilateral feedback effects as well. Moves at the
international game-board might trigger domestic realignments and
vice versa. For the state, in some cases, this effect may be
manipulated to achieve domestic gains which would be impossible had
it been acting at Level 2 alone.53 So if a state is clever, then it will
exploit these synergies:54
Economic interdependence multiplies the opportunities for altering

domestic coalitions (and thus policy outcomes) by expanding the set of
feasible alternatives in this way — in effect, creating political
entanglements across national boundaries. Thus, we should expect
synergistic linkage (which is, by definition, explicable only in terms of
two-level analysis) to become more frequent as interdependence grows.
Glenister suggests that the increased frequency of synergistic linkage

is not all that we should expect. In Glenister, the Court assumed a
Level 2 role over and above its ordinary adjudicative function. It
acted as the ratifying link through which an international agreement
gained legal influence and effect in South Africa’s domestic policy
space. On a purist’s dualist approach, this outcome does not reflect
the natural order of things. This result strongly suggests that there is
a further layer of complexity unaccounted for by two-level analysis.
If we consider this anomaly against Putnam’s theory, it would seem
that the courts are emerging at a level of their own. This potential
development is discussed in further detail below.55

53
If, for example, the President of South Africa’s initial negotiations had been on
gold instead of titanium, and word of these first negotiations were leaked to the
public, we would expect that, as shown above in Figure 4, there would be no
possibility of international agreement. If the President of South Africa had then
consequently changed tack and negotiated on titanium, the public might be more
inclined to accept the agreement. If he then proceeded to sign an agreement on
titanium, he will have achieved the benefits flowing from increased titanium
exports at very little political cost.
54 Putnam (n 3 above) 447, 448.
55
One concerning recent example of the interdependence of economies is the
series of events that have come to be known as the European sovereign debt
crisis. The result of a complex combination of factors, including the globalisation
of finance, high-risk lending and borrowing practices and the subsequent
assumption of private debt burdens (such as bank bail-outs, for example), the
crisis has grown to become, according to the Organization for Economic
Development and Cooperation, a ‘key risk to the world economy’. See L Alderman
& S Castle ‘Dire warnings are building on European debt crisis’ by http://
www.nytimes.com/2011/11/29/business/global/moodys-warns-of-escalating-dan
gers-from-europes-debt-crisis.html?_r=1&pagewanted=all (accessed 27 February
2012).
4 Three-level games
4.1 The courts as third-level players
So what is the QED for a third level? What would be required to show
that the court is in fact participating, albeit as a shadow player, in the
domestic-international arena? (Let’s call this question the ‘First
Issue’.) And how might one reliably distinguish between a fully-
fledged third level and something closer to a subtle influence on the
dynamics at the first or second? (Let’s call this question the ‘Second
Issue’.) For obvious reasons, Putnam was not explicit about what a
third level would look like. But insofar as he delineates the essential
characteristics of Levels 1 and 2, he proffers an implicit idea of what
it is not. I propose that the courts possess at least three sufficiently
distinguishable characteristics that demonstrate the existence of the
First Issue and the Second Issue identified above.
The first two characteristics pertain to the First Issue. To begin,

then, for a court to be deemed to sit at its own game-board, it must
at least be capable of acting independently of the judgment, the
wishes or the interests of the players at the other two game-boards.
So the first essential characteristic is whether the courts can be called
‘rational agents’. Put slightly differently, this question requires one
to assess the extent to which a court is capable of acting
independently and in a goal-directed manner. Expressed in yet
another way, the issue is whether a court can, in law, come to a
conclusion (regarding the effectiveness of an international
agreement) based on its own idiosyncratic preferences (which
preferences might be, amongst other things, that the cases that are
brought to it are resolved, and also that they are resolved with a view
to doing justice in accordance with the Constitution and the rule of
law).
The second characteristic is that the imprimatur of a court must

be a necessary condition for the effectiveness of an international
agreement. The significance of this characteristic is that it shows that
the courts play a meaningful role not only in adjudication, but also in
the incorporative process, or the ‘game’, that is played between
Level 1 and Level 2 in the transformation from tentative agreement
into law. Because if a court is not required to take a view before a
ratified tentative agreement becomes unequivocally effective, then
the court is not important to the process of making that law
unequivocally effective. In short, the courts will not be players if they
cannot affect the result.
The third distinguishing characteristic, which pertains to the

Second Issue, is as follows. On Putnam’s account, Level 1 is the arena
for the international negotiation of a tentative agreement. Once

agreement is successfully reached, the reason that it is tentative is
because it will only be of real domestic force and effect once it is
ratified, whether formally (through the promulgation and enactment
of a law) or informally (via the absence of meaningful domestic
pressure from powerful interest groups).56 If an agreement at Level 1
is not subsequently ratified, then there are two possible
consequences — the agreement must either be abandoned at Level 1,
or negotiations must be re-opened at the same level. So Level 1
negotiates, and Level 2 confirms ‘up or down’ (by which it is meant
that there is no subsequent amendment that is made at Level 2 — the
function at that level is to ‘take [the agreement] or leave it’). For one
to assert that the courts sit at a third game-board, they will have to
do something different, or at least do something in a different way,
than what occurs at Levels 1 and 2.
It might be salutary to raise a flag at this point to acknowledge,

for the purposes of clarity, that the First and Second Issues are
separate and distinct, in that they can be resolved independently of
one another. To elucidate, if courts can be said to possess the first
two characteristics listed above, then the First Issue will be resolved,
while the Second Issue may be resolved, inter alia, by possession of
the third characteristic. But courts need not possess either of the first
two characteristics in order to resolve the Second Issue, and,
similarly, the presence of the third characteristic is irrelevant to the
resolution of the First Issue. The two Issues are accommodated
together in this paper deliberately: both are relevant for the purposes
of a three-level claim, and both arguments sit comfortably side by
side. To explain further, if courts are rational agents and are
legitimately involved in the process of bringing international
agreements into effect, but simultaneously constrained to the same
options as Level 2 players (that is, they have the power only to vote
‘up or down’ and are devoid of their ordinary power to determine the
meaning and legitimacy of any single given provision of an
international agreement), then the process will be indistinguishable
from the process of ratification as defined by Putnam. The concern
which is mitigated by the grouping of the two Issues is that the
independent existence of the courts as players in the game-metaphor
used here will have been established by the presence of the first two
characteristics, however the work done by the courts, in the absence
of the third characteristic, will be more appropriately ascribed to
Level 2 activity, as opposed to activity on a level of its own.

If the reasoning above has merit, the conditions which must be

satisfied by the courts to justify a three-level claim may be
summarised as follows. The Court:
(i) must be a rational agent;

(ii) must play a necessary role in the effective implementation of any
international agreement;57 and
(iii) must have an affect, in the overall game, which is sufficiently
distinguishable from that of the others.58
In regard to condition (i) above, compare the result in Glenister with

the conclusion in Scaw. In the latter case, the Court’s independence
was less visible, as it was the state that had an interest in the
enforcement of an international covenant which had subsequently
been transformed into a domestic obligation. As set out above, the
Court and the state were aligned in that case, and the state won. In
the former case, however, the state argued, implicitly, against
allowing an international covenant to have a bearing on the content
of the state’s municipal legal obligations, and it did so, inter alia, on
the basis that the agreement had not been incorporated at Level 2. In
Glenister, notwithstanding South Africa’s dualist approach, the Court
held against the state. This result constitutes a credible demon-
stration of the Court’s independent discretion.
Condition (ii) is satisfied because, whilst courts might not

negotiate or enact law, if a law resulting from an international
agreement is held to be constitutionally invalid, then the effect of the
international agreement disappears. In this sense, the court, ‘rejects’
the agreement, albeit after the fact. Now, on a narrow view, the
courts are not players, because a ratified international agreement can
be (and usually is) validly entered into without them. But on a broader
view, which is the preferred or more desirable view, an agreement (or
a given provision thereof) which is ratified is still tentative until it is
confirmed by the courts. Moreover, it would seem that courts’ assent,
on the facts of Glenister, might be of even greater strategic value to
the state than the more generalised ratification function at Level 2.
For on a dualist approach, whilst ratification from Parliament is,
indeed, a necessary condition for general domestic legal validity (in
the sense that the given agreement may broadly be said to have been
incorporated into South African law), the affirmation of the courts
alone, on the basis of Glenister, at least, may be sufficient for
effective application at Level 2 (via the confirmation or
57
Satisfaction of the conditions (i) and (ii) resolves the First Issue, as explained
above.
58 Satisfaction of condition (iii) resolves the Second Issue, as explained above, and,
together with affirmative resolution of the First Issue, the courts will be
legitimate third-Level players.
interpretation of a single provision, or group of provisions, as the case

may be). On the majority holding in Glenister, then, there may be
some instances in which the courts may act both independently, on
Level 3, and as a contemporaneous substitute for the ratification
function at Level 2. Thus, on a dualist approach, and in the context of
Glenister, the three-level game may be represented as follows in
Figure 5:
The final condition, condition (iii), is the distinguishability of the

courts’ function from the functions ascribed to actors at the other two
Levels. The courts satisfy this condition as well. They are different
because they play a distinctly interpretive role when they give effect
to the international agreements placed before them. What’s more,
they do not negotiate agreements, and they are not forced to ‘take
[an agreement] or leave it’, as players at Level 2 must do. They are,
rather, imbued with the power to give meaning to any single provision
or, alternatively, to interpret the effect of the entire agreement,
depending only upon the relevance of that interpretive exercise to
the effective resolution of the case that is before the court at that
time.59
The three conditions set out earlier are accordingly fulfilled, and
the First and Second Issues are thereby resolved. The satisfaction of
these conditions indicates the unique influence of the courts at an
independent level of game-play, and so it follows that courts can be
59
Moreover, though not relevant to the question concerning the nature of the
courts’ function, the courts are actuated by motives that are wholly distinct from
those that drive Levels 1 and 2, in that courts (unlike Parliament at Level 2) have
no ‘political indifference curve’, because they do not have a range of interests
against which to trade-off: an agreement is either legally compliant or it is not.
And, unlike the executive at Level 1, their decisions are not subject to the
Damoclean sword of ratification (or at least not in the same sense) — they are
‘accountable only to the Constitution’ and fulfill their duty ‘without fear, favour
or prejudice’ (see section 165 of the Constitution and De Lange v Smuts NO 1998 3
SA 785 (CC) para 178).
said to be legitimate third-level players. Accordingly what remains is

to examine some of the courts’ potential Level-3 effects through an
analysis of relevant and recent South African case law.
4.2 The three-level effect
Scaw was a win for the state because it held a win-set. It had tentative
agreement to, and ratification of, the GATT and the WTO agreement,
coupled with full domestic ratification in the form of the ITA Act. In
that case, the Court held that the state’s conduct fell within the
proper purview of its authority in terms of the Constitution. So the
state won at Level 3 because it had already won at Levels 1 and 2. If
the effect of Scaw were an example of the only manner in which an
international agreement could gain domestic effect, then we could
say the following about the state:
(a) if the state is inclined to conduct itself in a manner which is

beneficial to its interests; and
(b) if the state tends to benefit from entering into international
agreements (based on its experience in Scaw);
then
(c) the state will tend to be inclined to enter into international
agreements.
If the state is indeed so inclined, then there will be a corresponding

increase in the range of ratifiable win-sets, coupled with a decrease
in the state’s bargaining power on the international plane. Provided
that there is no negative feedback from the electorate at Level 2,
Scaw will be facilitative of international cooperation. We can
represent this increased inclination in the following manner:
The counterexample is, of course, Glenister, where the state lost.

Notwithstanding South Africa’s dualist approach, and the fact that
there was no ratified agreement in place, the Court held that the
state should have done more. Justifiable intentions were not
sufficient — the state’s conduct still had to meet the appropriate
constitutional standard, whose substantive content was informed by
international law. This finding was, for the state, a costly mistake. If
the effect of Glenister were taken as a legitimate signal of the courts’
stance on international agreements in future, as we did with Scaw
above, then the state might be less inclined to enter into
international agreements at Level 1. In other words:
(a) if the state is inclined to conduct itself in a manner that is beneficial

to its interests;
and
(b) if entering into international agreements tends to be an overall cost
the state;
then
(c) the state will not be inclined to enter into international agreements.
And the narrower the range of win-sets that can be ratified at Level
2, the more intransigent the negotiators at Level 1 will be. With
contracted negotiation-flexibility, they will be less likely to
compromise and come to terms:
In the Scaw analysis captured above in Figure 6, after showing that

there was a positive effect on the range of ratifiable win-sets at
Level 1, we presumed that there was a similar effect at Level 2 and
concluded, on that presumed basis, that there would be an overall
positive effect on international cooperation. If we were to relax this
presumption for Glenister, then the state’s task of optimising
synergistic linkages at Levels 2 and 3 becomes even more complex.
So, instead of presuming that Glenister’s effect at Level 2 is

negative, as it is at Level 1, suppose that Level 2 reacts positively and
welcomes the decision. There will be a contraction of political will at
Level 1 (resulting from the state’s failure to win), coupled with an
enlargement of the range of ratifiable agreements at Level 2. The
consequence is a conflict of interest at Level 1. The executive will
have to elect either to avoid further international negotiations, or to
give in to pressure from Level 2, with the former option having the
advantage of averting further adverse interpretations at Level 3, and
the latter having the advantage of a measure of political quid pro quo
from the electorate at Level 2 in the medium to long term. Under
these circumstances, decisions at Level 1 are more likely to be ad hoc
short-term manoeuvres than principled decisions for maximal benefit
in the long term. This conclusion is not meant as a cynical statement
about politics and politicians, but merely recognises the fact that
exigent circumstances produce exigent decisions.
4.3 The import of the decisions in Glenister and Scaw
International relations are a non-zero-sum game. Players at all levels

have more to gain from exploiting synergies in a symbiotic manner
than to operate in ignorance of them. But each player at each Level
will be situated differently and will, accordingly, have different
interests. As with almost any court decision, Glenister and Scaw
would have effects that were good for some, and not so good for
others. So, if one had to decide which decision were preferable, one
could only answer that ‘it depends’.
At least one thing that ‘it depends’ upon is the Level from which
the question is answered. Glenister may have been a difficult decision
for the state to swallow, but it may also have had a very positive
effect at Level 2. Scaw might even have had a negative effect at Level
2, despite the positive effect that it had at Level 1. But if we leave
the issue of perspective aside for a moment, and instead, assess the
two by virtue of their utility for legal certainty, the answer is
different. Legal certainty is a more important measure for present
purposes, as it is an end in which all Levels have an overlapping
interest. The executive, at Level 1, can be far more certain of what
it is and is not allowed to do. The ‘people’ at Level 2, will, in turn,
benefit from a more acute and corresponding awareness of the line
beyond which the state will have overstepped its bounds. And the
credibility of the courts at Level 3 will consequently be reinforced:
the other players can take some comfort in courts ‘doing what they’re
supposed to do’.
From the perspective of legal certainty, Scaw, with respect,

outshines Glenister. The former case follows the dualist approach to
the letter: it interprets a national enactment of an international

agreement in a manner which is consistent with the purport of both
the international and the consequent domestic obligations arising
from laws at each Level. For its part, the latter case also reaches an
outcome which is consistent with international law, but it does so
through a more complicated hermeneutic process. Notwithstanding
South Africa’s dualist tendency, the Court, in Glenister, held the
state to an obligation at Level 2 which had not been ratified.
Considering the fact that the decision in Glenister is, as a matter of
law, unusual, it is submitted that it would be reasonable to assume
that the result was an unexpected one for the state.
To illustrate the interpretive difficulty further, if one were to

transpose the approach followed in Glenister, and to apply the
reasoning of the majority to the facts of Grootboom, which was the
Court’s famous judgment concerning the state’s positive obligations
in regard to the right of access to adequate housing, then the outcome
of the former case might very well have been different. In the original
Grootboom (that is, Grootboom without Glenister’s reasoning), the
Court required the state to be take reasonable steps to ‘meet the
needs of those living in extreme conditions of poverty, homelessness
or intolerable housing’.60 This standard was less onerous than the
‘minimum core’ obligations found in international agreements
ratified by the state.61 In Grootboom, the court held that the
‘minimum core’ international obligations were distinguishable from
the constitutional desiderata and that adherence to the former
standard was thus unnecessary. But if the Court had opted for a
Glenister-styled approach, or, in other words, if ‘minimum core’ were
a standard that the Constitution took up for itself and imposed upon
the state, then, presumably, the state would have been required to
provide Mrs. Grootboom and the others with immediate housing. The
problem here is not that such a decision would have been a bad thing.
The difficulty is that Glenister, despite its similar international-

61
Article 11(1) of the United Nations’ International Covenant on Economic, Social
and Cultural Rights (which entered into force on 3 January 1976) provides as
follows:
The States Parties to the present Covenant recognise the right of
everyone to an adequate standard of living for himself and his family,
including adequate food, clothing and housing, and o the continuous
improvement of living conditions. The States Parties will take appropriate
steps to ensure the realisation of this right, recognising to this effect the
essential importance of international co-operation based on free consent.
This Article must be read with Article 2.1 (see Grootboom (n 5 above) para 27),
which provides as follows:
Each State Party to the present Covenant undertakes to take steps,
individually and through international assistance and co-operation,
especially economic and technical, to the maximum of its available
resources, with a view to achieving progressively the full realisation of
the rights recognised in the present Covenant by all appropriate means,
including particularly the adoption of legislative measures.
agreement subject matter, does not explain how to justify the

difference in outcome. If the reasons for the outcome are not clearly
understood by the state, then playing at Level 1 becomes more of an
unknown quantity, and thus a less justifiable risk.
As set out above, however, this paper does not presume to take a
normative position on whether the reasoning in any of Scaw, Glenister
or Grootboom should be preferred for one reason or another. The
upshot is, instead, that the effect of the incorporation of
international law in Scaw and Glenister is noticeably different and
that the conclusions in Glenister and Grootboom appear, with
respect, to be mutually exclusive of one another, which the majority
decision in Glenister does not comprehensively explain. The
discussion set out above suggests that at least two important
consequences attach to this analysis. The first consequence is that, as
a legal matter, the principles concerning incorporation are somewhat
less clear. The second is that, as a matter of strategy, the uncertainty
created by Glenister (when viewed in the light of previous
jurisprudence and particularly of South Africa’s dualist approach) is
likely to have the unfortunate effect of rendering consequent inter-
level realignments more tentative and temporary.
So it would seem that Glenister, relative to the decision in Scaw,

poses a challenge for future international interaction at two levels of
government. The challenge for the state, in the short term, will be to
formulate a strategic response at the international level that takes
account of the current uncertainty and delivers the most efficient and
effective result possible. The challenge for the courts, in the longer-
term, will be to develop a clear and principled approach to
international incorporation in this still-nascent constitutional area of
overlapping state functions.
CONFUSING GRACE WITH AMNESIA:
REVIEWING ACTS OF THE HEAD OF STATE
Francois Venter*
1 Introduction
In terms of section 83 of the South African Constitution the President

is both head of state and head of the national executive. Section 85
vests the executive authority of the Republic in the President, who
exercises it ‘together with other members of the Cabinet’. In terms
of section 86 the President is elected by the majority in the National
Assembly, meaning in practice that the leader of the majority party
becomes President. Section 91(2) provides that the President appoints
the Deputy President and Ministers and assigns their powers and
functions, and may dismiss them’. In law there are therefore no
limitations on the leader of the majority party regarding the
selection, dispatch and control of members of Cabinet. Limitations on
the President that may exist in this regard therefore depend almost
wholly upon the political culture of the governing party. As long as a
president remains in command of the majority party, he or she may
require Cabinet and Parliament to do his or her bidding, albeit —
thankfully — within the bounds of the Constitution.1
Section 84(2)(j) of the Constitution provides that the ‘President is

responsible for ... pardoning or reprieving offenders and remitting any
fines, penalties or forfeitures.’ This ‘responsibility’ is entrusted to the
* Professor of Law and Dean, Faculty of Law, Potchefstroom Campus, North West
University.
1 This is a feature ‘inherited’ from pre-constitutional South Africa and may be
described as the Constitution's Achilles heel: it harks back to an era in our
constitutional history in the 1980's when the State President was in effect
endowed with executive sovereignty. However, the current dispensation is
different from that era in that our Constitution is supreme and justiciable.
Nevertheless, the arrangement holds the danger of the destruction of
constitutionalism should a president endowed with supremacy over not only the
Executive, but also over a legislative majority capable of amending the
Constitution, be so inclined.
167
168 Reviewing acts of the head of state
President under the heading of ‘Powers and Functions of the

President.’ This is far from unique in the world and the provision
emerges, even in the South African context, from a long history. When
it comes to the exercise of the pardon, it is possible for a President as
guarantor of ministerial incumbency to ensure that any advice given
would accord with his or her personal preference. If advice or
recommendations received do not please the President, they may
easily be circumvented without any constitutional or other legal
consequences.
The discussion below begins with a sketch of the chronology and

outcomes of the cases of Chonco and Albutt, followed by a brief
exposition of the distinction between amnesty and pardon. In section
4 the political activities surrounding the cases are outlined and in
section 5 some substantive constitutional and administrative law
issues related to the circumstances of the cases, including the
reviewability of discretionary presidential actions, are discussed
before conclusions are drawn.
2 Chonco and Albutt
The presidential pardon received the attention of the Constitutional

Court more than once in 2010. Behind the eventual judgments that
were handed down lies a long trail of political, executive,
administrative, parliamentary and judicial activity. An understanding
of the chronology of all of these activities is necessary for an
understanding of the jurisprudential issues involved. The cases of
Chonco2 and Albutt3 came to the Court along different routes, but
they were eventually woven into the same fabric through the
manoeuvrings of politicians and litigants.
The Truth and Reconciliation Commission (TRC) commenced its

work in December 1995, and was dissolved at the end of 2001. The
Commission's Amnesty Committee considered more than 7000
applications for amnesty, and then granted or refused the
applications. After the dissolution of the TRC 33 members of the ANC
and PAC who were unsuccessful in their applications for amnesty were
pardoned by President Mbeki. It is significant that the 33 were not
granted amnesty, but received the presidential pardon.
2
Chonco v Minister of Justice and Constitutional Development 2008 4 SA 478 (T)
('Chonco High Court'); Minister of Justice and Constitutional Development v
Chonco 2009 6 SA 1 (SCA) ('Chonco SCA'); Minister of Justice and Constitutional
Development v Chonco 2010 4 SA 82 (CC) ('Chonco 1'); Chonco v President of the
Republic of South Africa 2010 6 BCLR 511 (CC) ('Chonco 2').
3 Centre for the Study of Violence and Reconciliation v President of the Republic
of South Africa North Gauteng High Court in Case No 15320/09; Albutt v Centre
for the Study of Violence and Reconciliation 2010 3 SA 293 (CC).
Moved by this, Mqabukeni Chonco, convicted and serving a

sentence for various serious crimes for which he was sentenced to
death in 1989,4 applied in May 2002 for a presidential pardon on the
grounds that his crimes were committed ‘for political objectives’.
Chonco was a member of the IFP, a political party that expressly
refrained from participating in the activities or the TRC. Chonco was
joined in his application by 383 prisoners in a similar situation, with
the IFP assisting the applicants. The application was duly lodged with
the Minister of Justice and Constitutional Development, but nothing
came of it until questions regarding the matter were raised in
Parliament. More than three years later, in September 2005,
President Mbeki stated that he had not yet seen the applications and
that the Minister was to process and submit them to him. When still
nothing was done, the IFP — supported by the PAC — approached both
the Human Rights Commission and the High Court for assistance.
On 21 November 2007 President Mbeki informed a joint sitting of

Parliament that a special process was to be set in motion to deal with
a large number of applications for presidential pardon for crimes
allegedly committed with a political motive arising from the conflicts
of the past. This ‘special dispensation’ included the establishment of
a multi-party ‘Presidential Reference Group’ (PRG) which was to
consider all the pending applications and make (non-binding)
recommendations to the President. Although the Chonco-related
applications would be included in this process, the subsequent
litigation on behalf of that group did not centre upon the special
dispensation, but rather on the executive prevarication over their
original (2002) applications for pardon.
On 11 February 2008 the North Gauteng High Court ordered the

Minister to do, within three months, whatever is necessary to make it
possible for the President to exercise his responsibility to consider the
pardoning applications of Chonco and others.5 The Minister appealed
against this order. The Supreme Court of Appeal dismissed the
Minister's appeal in March 2009,6 pointing out that the Minister was
constitutionally bound to perform the executive functions concerned.
The Court construed a right for prisoners to apply for a pardon, and
placed an obligation on the Executive to respond to such an
application. The Minister then applied for leave to appeal to the
Constitutional Court.
In what eventually became the Albutt case, the North Gauteng

High Court was approached in April 2009 by seven non-governmental
organisations for an interdict against the President on an urgent basis
4 The death sentence was later commuted to life imprisonment.

5
Chonco High Court (n 2 above).
6 Chonco SCA (n 2 above).
to prevent him from granting any pardon in terms of the Mbeki special
dispensation of 2007. The NGO's wished the victims of the crimes for
which the applicants had been convicted to be given access to the
applications and to the proceedings and recommendations of the PRG
and to be given the opportunity to make representations to the
President in that regard. The application was based on the fact that
the PRG did not function transparently since it had refused to reveal
any information about the applications for pardon submitted to it or
about the recommendations it had made and had also refused to
entertain representations to it on whether or not a pardon should be
granted. Furthermore the President had also refused to disclose which
applications for pardon he was considering and would not allow
victims or other persons affected by the offences concerned to make
representations to him.7 This was the precursor to the Albutt
judgment delivered by the Constitutional Court on 23 February 2010
in which it was held that the victims of the crimes concerned were
indeed entitled to be given the opportunity to be heard before the
President decided to grant pardon under the special dispensation.8
In the mean time the Constitutional Court had been seized with
the Minister's application for leave to appeal against the judgment of
the SCA in favour of Chonco and his co-applicants. In its judgment of
30 September 2009, later referred to by the Court as ‘Chonco 1’9 the
Constitutional Court granted the Minister leave to appeal, upheld the
appeal and overturned the judgments of both the High Court and the
Supreme Court of Appeal. Essentially this finding was based on the
Court's view that Chonco had wrongly sued the Minister while the
action should have been brought against the President. The applicants
promptly applied to the Constitutional Court for direct access and
sought an order declaring that the President had unreasonably
delayed in considering and deciding their applications for presidential
pardon, and directing the President to decide their applications
within a month. Judgment was delivered in what might be referred to
as 'Chonco 2' on 16 March 2010.
3 Amnesty and the presidential pardon
The TRC process emanated from the ‘postamble’ of the transitional

Constitution of 1993,10 the fifth paragraph of which provided as
follows:
7
Paras 44-46 of Albutt (n 3 above). At the time Kgalema Motlanthe was the
President. His term of office was between September 2008 and May 2009.
8 Albutt (n 3 above).
9
Chonco 1 (n 2 above).
10 Constitution of the Republic of South Africa, Act 200 of 1993.
In order to advance such reconciliation and reconstruction, amnesty

shall be granted in respect of acts, omissions and offences associated
with political objectives and committed in the course of the conflicts of
the past.
A very clear exposition of the nature of amnesty is provided by

O'Brien.11 Amnesty, he says, ‘is an expression of the sovereign power
of the lawmaker to provide immunity from the usual operation of the
law’, which can take various forms and which has an age-old history
as a means of settling civil and military conflicts. Its purpose is
essentially to bring about a formal forgetting (the root of the word
‘amnesty’ is the same as that of ‘amnesia’) of past occurrences by
means of awarding immunity from liability in law for deeds done in
the course of war or civil strife.
In line with the nature of amnesty, section 20(7) of the Promotion

of National Unity and Reconciliation Act 34 of 1995 provided that a
person who was granted amnesty by the TRC’s Amnesty Committee
for particular acts, omissions or offences associated with a political
objective, would no longer bear any liability for such acts, omissions
or offences. In terms of subsection (8), sentences imposed upon
persons convicted for crimes for which amnesty was granted would
lapse.
The constitutional ‘responsibility’ of the President in terms of

Section 84(2)(j) to pardon or reprieve offenders and to remit any
fines, penalties or forfeitures has its historical roots in the royal
prerogative of the British sovereign since at least the 17th Century.12
In the 18th Century Blackstone described the prerogative of mercy as
one of the great advantages of monarchy, but added:13
In democracies, however, this point of pardon can never subsist, for

there nothing higher is acknowledged than the magistrate who
administers the laws; and it would be impolitic for the power of judging
and of pardoning to centre in one and the same person.
The notion of the pardon should be understood to be a mechanism for

correcting judicial errors when effecting such corrections is beyond
the jurisdiction of a court. Originally it was deemed to be a royal act
of grace, the Crown being the fountain of justice for the whole realm.
Where it was originally a matter of pure regal discretion performed in
the absence of the opportunity to appeal a judicial conviction and
sentence, since the early 20th Century the consideration and granting
11
R O'Brien 'Amnesty and international law' (2005) 74 Nordic Journal of
International Law 261, especially section 2 at 262-265.
12 Se eg DEC Yale (ed) Sir Matthew Hale's The prerogatives of the King (1976) ix-xi.
13
Blackstone's Commentaries on the laws of England in four books Vol 2 (1753)
Chapter XXXI 398.
of the pardon in England has been located in Cabinet and therefore

became a function of the Executive.14
Before South Africa became a Republic, English colonial

constitutional law caused a similar pattern to be followed, leading
Hahlo and Kahn to state in 1960 that the Executive (Cabinet) had
‘secured control over the prerogative’ in that it could be exercised
without reference to the sovereign.15
When the Union of South Africa was converted into a Republic,

section 7(4) of the Republic of South Africa Constitution Act 32 of 1961
conferred such powers and functions as were immediately prior to the
commencement of this Act possessed by the Queen by way of
prerogative upon the newly established head of state, the State
President, and section 7(3)(f) provided that he had the power ‘to
pardon or reprieve offenders, either unconditionally or subject to
such conditions as he may deem fit, and to remit any fines, penalties
or forfeitures’. Under section 16(1) of the 1961 Constitution Act, the
State President was bound to act on the advice of the executive.
However, the State President of the Republic of South Africa

Constitution Act 110 of 1983 was not merely head of state, but also
head of government. The effect of this arrangement was that it was
possible for the State President to exercise the prerogative of pardon
without the advice — although 'in consultation' — with the Cabinet.16
The State President was in an extremely powerful position in that
cabinet ministers were fully dependent upon him for their
appointment or dismissal resulting in the advice rendered by ministers
to be likely not to challenge his opinions. This poor example of
modern constitutionalism was unfortunately replicated in both the
1993 and 1996 Constitutions.17
Presumably Chonco and his co-applicants approached the Minister

of Justice and Constitutional Development for presidential pardons in
2002 in terms of the provisions of section 327 of the Criminal
Procedure Act 51 of 1977. In effect section 327 allows a person who
has been convicted of an offence regarding which new evidence
becomes available after all avenues of recourse to the courts have
been exhausted, to petition the Minister of Justice and Constitutional
Development, who may refer it back to court for consideration and
14
Cf eg ECS Wade & AW Bradley Constitutional and administrative law 10th ed
(1985) 360 and C Turpin & A Tomkins British government and the Constitution 6th
ed (2007) 464-465.
15
HR Hahlo & E Kahn South Africa — The development of its laws and Constitution
(1960) 128.
16 Cf eg D Basson & H Viljoen Suid-Afrikaanse staatsreg 2nd ed (1988) 60, the
relevant provisions of Act 110 of 1983 being secs 6(4) and 19(1)(b).
17 Cf note 1 above.
may make a recommendation to the President regarding the granting

of a ‘free pardon’. Ministerial and presidential actions in this regard
are in terms of section 327(7) not subject to appeal or review. Given
the pre-constitutional origin of Act 51 of 1977 it may be that this
provision will, if challenged, be found to be unconstitutional or in
need of being read down in some respects, for example regarding the
exclusion of judicial review of the Minister's actions, and also because
section 84(2)(j) of the Constitution requires the petition to be
directed at the President and not the Minister, the latter not having
to be involved at all in terms of the Constitution.
Obviously amnesty as it was conceived in the 1993 Constitution is

different from pardon as provided for in Section 84(2)(j) of the 1996
Constitution. It would however appear that neither the Office of the
Presidency nor the courts recognised this distinction properly while
dealing with the Chonco and Albutt matters.
When the Constitutional Court was called upon to certify the text
of the draft Constitution of 1996, Section 84(2)(j) was objected to on
the grounds that the presidential pardon would undermine the
supremacy of the Constitution and violate the separation of powers.
To these objections the Court responded,18 with reference to the
historical roots of the pardon, that the President did not derive this
'power' to pardon from antiquity, but from the Constitution itself.
Responding to the concerns about the separation of powers, the Court
pointed out that it had never been part of the functions of the
judiciary to ‘pardon and reprieve offenders after justice has run its
course,’ but that that was a function ‘ordinarily entrusted to the head
of state in many national constitutions’ where the supremacy of the
constitution and the separation of powers are not considered to be
threatened thereby. The Court reconfirmed this view in the Hugo
case.19 There the Court (referring to the provision in the 1993
Constitution's equivalent of the current Section 84(2)(j)) expressly
stated that:20
... the exercise by the President of his powers under s 82(1) may be
subject to review by Courts of appropriate jurisdiction in the same way
as the exercise by him of other constitutional powers would be subject
to review.
Section 2 of the 1996 Constitution declaring the Constitution to be

‘supreme’ and rendering ‘law or conduct inconsistent with it’ invalid,
as well as section 7(2) requiring the state to ‘respect, protect,
18
Ex Parte Chairperson of the Constitutional Assembly: In re Certification of the
Constitution of the Republic of South Africa, 1996 1996 4 SA 744 (CC) paras 116 &
117.
19
President of the Republic of South Africa v Hugo 1997 4 SA 1 (CC) paras 5, 7 & 8.
20 Para 13 of the Hugo judgment (n 19 above).
promote and fulfil’ the rights in the Bill of Rights and section 8(1)
binding ‘the legislature, executive, the judiciary and all organs of
state’ to the Bill of Rights, read together, should, in line with the
dictum from the Hugo judgment cited above, render decisions by the
President to grant or withhold the pardon subject to judicial review.
Add to this that section 167(4)(e) grants the Constitutional Court the
jurisdiction to ‘decide that Parliament or the President has failed to
fulfil a constitutional obligation’. To these considerations we return
in section 5 below.
4 The travails of political litigation
Given the fact that President Mbeki pardoned 33 unsuccessful

amnesty applicants before, it is difficult to fathom why, when he
announced his special dispensation for the more than a thousand
applicants for pardon who did not take part in the TRC amnesty
proceedings but alleged that they were convicted for offences
committed while moved by political considerations, chose in effect to
infuse amnesty into the pardon. Also given the inordinate delay in
dealing with the applications up to that point and its eventual
outcome (albeit under a new presidency) one can reasonably only
infer that the President himself was driven by political motives as
opposed to considerations of legality or constitutionality. The
administrative equivocation by three presidents and their ministers
over eight years and the argumentative stances of the Executive in
the concomitant litigation strengthen this inference.
In Mr Mbeki's announcement of the special dispensation on 21

November 200721 there was a deliberate conflation of amnesty and
the presidential pardon. On the one hand he stated that:
the President has an obligation to consider all requests made to him or

her to pardon or reprieve offenders and remit any fines, penalties or
forfeitures. At the same time, having thus applied his or her mind, the
President is under no obligation to accede to the requests made to him
or her, provided that she or he proceeds in a rational manner.
On the other hand he said that he would, in deciding on the

applications:
be guided by the principles and values which underpin the Constitution,

including the principles and objectives of nation-building and national
reconciliation; and, uphold and be guided by the principles, criteria and
spirit that inspired and underpinned the process of the Truth and
21
The text of the speech can inter alia be found on the ANC's website at http://
www.anc.org.za/show.php?id=4240 (accessed 21 March 2012).
Reconciliation Commission, especially as they relate to the amnesty

process.
The Department of Justice and Constitutional Development

subsequently published the PRG's Terms of Reference,22 according to
which its work had to be completed by no later than 30 November
2008. It is noteworthy that the establishment of the PRG and its terms
of reference were not based on any statutory or constitutional
foundation: it was a unique mechanism created administratively to
assist the President in his consideration of the specific set of
applications for pardon in a manner not binding him in his decision-
making. Despite the unusual nature of this mechanism, it can hardly
be argued that the establishment of the PRG lay beyond the lawful
authority of the President. What was extraordinary was the decision
to half-revive the TRC amnesty process presumably as unfinished
business, and to have this process cast in a new format whereby
applications for amnesty (as pardon) were ostensibly to be weighed
objectively, though not transparently.
The PRG had not been established long when it was approached in
February 2008 by the NGO's acting on behalf of the victims of the
crimes for which the applicants for amnesty/pardon were convicted,
requesting access to the process. The PRG rejected these requests in
August 2008 stating that it was not obliged to allow such
participation, and that the President was the ‘custodian’ of the
process, to whom the applicants were therefore referred.23
After Mbeki was ousted by the ANC in September 2008, his

temporary successor for almost eight months, Kgalema Motlanthe,
stated in an affidavit that he ‘intend[ed] to deal with applications for
pardon ... in line with the approach outlined by the then President’.24
However, as has been mentioned above, Motlanthe refused to allow
victim participation. Later, in the Albutt case before the
Constitutional Court the President averred that he never actually
refused victim participation, but the Court had no doubt on this
account and described the President's statement as ‘argumentative in
tone’.25
When President Zuma took office in May 2009 the situation was
that the PRG was functus officio. Also, there were the Chonco
judgments of the High Court and the Supreme Court of Appeal
requiring the Executive to expedite consideration of the pardon
applications as well as an interim interdict granted by the High Court
22
To be found on the internet at http://www.info.gov.za/events/2008/
tor_pres_pardon.pdf (accessed 26 March 2012).
24
25 Albutt (n 3 above) para 44, also paras 45 & 46.
prohibiting the President from pardoning any of the applicants before

allowing victims of the crimes concerned to be heard. The application
to the Constitutional Court by Ryan Albutt (a member of the Afrikaner
Weerstandsbeweging (AWB)) for leave to appeal against the interim
interdict of the High Court was (ironically) supported by the President
and the Minister. At that time Zuma therefore was, as was Motlanthe,
opposed to the notion that the victims should be allowed a hearing.
This was an indication that none of the three presidents involved were
really serious with the revival of the TRC amnesty process in which the
public evidence of victims had played a key role.
In September 2009 the Minister's appeal against the Chonco

judgments succeeded in the Constitutional Court (Chonco 1).
However, the Court did open the possibility for subsequent action by
the applicants against the President, which they promptly took in the
hope of getting a response from President Zuma regarding their
applications pending since 2002. The judgment in Chonco 1 ended
with the obiter remark that the six years of delay in dealing with the
applications was ‘unacceptable’ in view of section 237 of the
Constitution26 and because ‘[g]ood governance and social trust are
premised at least partly on reasonable and responsive decision
making’ and ‘this kind of delay is out of kilter with the vision of
democratic and accountable governance’.27
The manner in which the President responded to being sued by

Chonco and others crowned the irony of the course of events in which
political moves were alternated with the exhaustion of all means
offered by litigious appeal using public resources. The Constitutional
Court was stopped in its tracks during the proceedings in Chonco 2 by
an announcement of the President on 4 February 2010 under the
heading ‘President Zuma has completed processing of Inkatha
Freedom Party (IFP) pardon applications’.28 The statement made it
generally known that the President had filed an affidavit in the
Chonco matter in which the Constitutional Court was asked that the
relief sought should not be granted in view of the fact that all 384
applications for pardon had been processed by him, rendering the
matter 'academic.' It was announced that the President had rejected
230 of the applications. 146 of the remaining applications, in which
application was also made in terms of President Mbeki's ‘special
dispensation’, could not be decided upon because of the order
granted in the North Gauteng High Court on 6 April 2009 prohibiting
the President from making a final decision in applications under the
26
Sec 237 provides: 'All constitutional obligations must be performed diligently and
without delay.'
27 Chonco 1 (n 3 above) para 45.
28
Available at http://www.thepresidency.gov.za/pebble.asp?relid=1267 (accessed
27 March 2012).
special dispensation without victims having had an opportunity to

make representations regarding the applications.29 The remaining
eight applications, not affected by the court order, were said to be so
closely linked to the others, that they could, in fairness to all
concerned, also not be considered. The statement ended with an
expression of the Presidency's hope that the announcement would
‘bring the matter currently before the Constitutional Court to an
end,’ which it did.
5 Presidential discretion and administrative

justice
Despite the absence of satisfactory distinctions between amnesty and

pardon in the Chonco and Albutt judgments, various strands of
analysis of profound doctrinal interest were touched upon in these
cases by the various courts, such as the distinction between head-of-
state and executive actions performed by a president (including the
question whether head-of-state actions are judicially reviewable),
the nature of the pardon in relation to amnesty and whether
executive conduct might be administrative action. Conclusive
guidance on none of these issues was provided in the judgments, but
it is instructive to follow the reasoning employed by the various
judges.
5.1 Head-of-state and executive acts of the President
Chonco's case was made on the basis that the delay in the President's
dealing with the applications for pardon was caused by a
constitutionally reproachable failure on the part of the Minister. The
Constitutional Court even speculated that this was a ‘litigation
strategy ... in order to break the seeming logjam in the process of
ministerial consideration’.30 Chonco's argument was based on the fact
that the President's decision on granting a pardon had to be preceded
by the exercise of normal executive functions performed in terms of
section 85(2)(e) of the Constitution, which provides:
The President exercises the executive authority, together with the other
members of the Cabinet, by:
(e) performing any other executive function provided for in the
Constitution or in national legislation.
29 The statement was not very accurate regarding dates: Mbeki's announcement was
made in 2007 (the statement said it was in 2008) and the court order concerned
was issued in the High Court not on 6 April 2009 but on 29 April 2009 (n 3 above).
This argument convinced both the High Court and the Supreme Court
of Appeal31 in view of section 92(3)(a) of the Constitution, which
requires ministers to ‘act in accordance with the Constitution’ and
section 237, which requires all constitutional obligations to be
performed 'diligently and without delay.'
The Constitutional Court now had good reason to explore the

distinction between the constitutional functions of the President as
head of state and head of the national executive. In affirmation of its
judgments in Hugo32 and SARFU33 that the presidential powers listed
in section 84(2) of the Constitution34 ‘originate historically from the
royal prerogative and were exercised by the Head of State rather than
the head of the national executive,’ but that they ‘are now clearly
original constitutional powers,’35 the Court defined a function as ‘a
tasked duty to act in terms of the Constitution or legislation’.36 Again
with reference to its previous judgments, the Court confirmed that
fundamental to the ‘principle of legality’ was the ‘sourcing of public
power’ in the Constitution or national legislation.37 Neatly putting
together dicta from three previous judgments of the Court,38 the
‘doctrine of legality’ was authoritatively described to be understood
as follows:
• the exercise of all public power must comply with the Constitution;
• legality is 'an incident of the rule of law';
• legality is a constitutional control regulating the exercise of public
power;
• neither the Legislature nor the Executive may exercise a power or
perform a function not conferred upon them by law, and
31 Cf Chonco 1 (n 3 above) para 13.

32
Hugo (n 19 above).
33
President of the Republic of South Africa v South African Rugby Football Union
2000 1 SA 1 (CC).
34
Assenting to and signing Bills; referring a Bill back to the National Assembly for
reconsideration of the Bill's constitutionality; referring a Bill to the Constitutional
Court for a decision on the Bill's constitutionality; summoning the National
Assembly, the National Council of Provinces or Parliament to an extraordinary
sitting to conduct special business; making any appointments that the
Constitution or legislation requires the President to make, other than as head of
the national executive; appointing commissions of inquiry; calling a national
referendum in terms of an Act of Parliament; receiving and recognising foreign
diplomatic and consular representatives; appointing ambassadors,
plenipotentiaries, and diplomatic and consular representatives; pardoning or
reprieving offenders and remitting any fines, penalties or forfeitures, and
conferring honours.
35
Chonco 1 (n 3 above) para 30.
36
38
As above. The previous dicta were drawn seriatim from Pharmaceutical
Manufacturers Association of SA: In re Ex parte President of the Republic of
South Africa 2000 2 SA 674 (CC), Affordable Medicines Trust v Minister of Health
2006 3 SA 247 (CC) and AAA Investments (Pty) Ltd v Micro Finance Regulatory
Council 2007 1 SA 343 (CC).
• the validity of the exercise of a public power must be 'clearly sourced

in law.'
However, the Court held that a power being ‘sourced in law’ did not
necessarily require express legislative listing of such powers. In view
of section 84(1) of the Constitution which provides that ‘(t)he
President has the powers entrusted by the Constitution and
legislation, including those necessary to perform the functions of
Head of State and head of the national executive,’ the Court stated
that the President bore more than mere decision-making powers, but
was also endowed with ‘auxiliary powers,’ which ‘would include the
power to request advice as well as the power to initiate the processes
needed to generate that advice, such as receiving and examining
applications for pardon’.39
The exploration of the head-of-state and head-of-the-national-

executive powers of the President, involving the determination of the
difference between sections 84 and 85 of the Constitution
respectively, produced the Court's conclusion in this regard that the
President as head of state acts alone, while as head of the national
executive his actions are performed collectively by him and the
Cabinet.40
The purpose of the Court's analysis was firstly ‘to identify the
source of the power to carry out the preliminary process, prior to the
head-of-state decision and, secondly, to determine to whom that
power accrues’.41 The conclusion reached was that the ‘preliminary
process,’ that is, the instruction to the Minister to receive and process
the applications for pardon, fell within the ambit of the President's
auxiliary powers.42 In effect, therefore, the President was found not
to have been hampered in the exercise of his sole authority to
consider the requests for pardon, since the instruction to the Minister
to do the preparatory work could have been withdrawn by the
President, transferred somewhere else, or he might even have
bypassed the procedure completely or might have ignored any advice
on the matter.43 Because there was no collective executive action in
the sense of the section 85 functions, the Minister was accountable to
the President alone — who also has the power to discipline or dismiss
the Minister and therefore there was no individual or collective
accountability to Parliament due to the laxity of the Minister.44
39 Chonco 1 (n 3 above) paras 31 & 32.

40
43
Chonco 1 (n 3 above) paras 38 & 39.
44 Chonco 1 (n 3 above) paras 36, 38 and 40.
While this interpretation is strictly speaking correct, it is under

the circumstances inconceivable that the President(s) were not aware
of, and by implication condoned, the failure of the Minister(s) to
perform their function diligently.
5.2 Judicial reviewability of presidential actions
The findings of the Court in Chonco 1 also brought up the question

whether the presidential powers entrusted by the Constitution
exclusively to the President were reviewable. The Court explicitly
stated that they were, in the sense that the Constitutional Court
(alone) was empowered by the Constitution to determine whether the
President ‘has failed to fulfil a constitutional obligation’.45 However,
at this juncture the Court did not expand upon the criteria within
which such a review might be conducted since it was not the
President's conduct, but the Minister’s that was the target of Chonco's
application. But soon afterwards the Court was called upon in Albutt
to consider this question.
In the Albutt case before the North Gauteng High Court it was
argued for the President that he had unfettered discretion regarding
the granting of pardon, while the applicants' position was that the
granting of pardon was reviewable administrative action for purposes
of the Promotion of Administrative Justice Act 3 of 2000 (PAJA). On
the basis of the inclusio unius rule of interpretation the Court
accepted the argument that the constitutional pardon was
administrative action since it was not excluded from the definition in
the PAJA. The Court considered the application of the common law
rule of interpretation not to offend any of the values and principles of
the Constitution and — in a distressing reliance on pre-constitutional
hermeneutics — held ‘that the legislature did not intend to exclude
the President's power of Pardon from the definition of an
“administrative action”’.46
Referring to the fact that the Service charter for victims of

crime47 required that input was to be obtained from victims when a
prisoner was being considered for parole, the High Court deemed the
practical effect of parole and pardon to be the same and that it was
therefore unjustified to deal with applicants for pardon in a manner
45 Chonco 1 (n 3 above) paras 41 and 43, referring to sec 167(4)(e) of the 1996
Constitution.
46
Centre for the Study of Violence and Reconciliation (n 2 above) section 7.3 of the
judgment.
47
Available at http://www.npa.gov.za/files/Victims%20charter.pdf (accessed 24
March 2012). The Charter is a document published by the Department of Justice
and Constitutional Development inter alia ‘To provide for the consolidation of the
present legal framework in South Africa relating to the rights of and services
provided to victims of crime’ informing victims of crime of their rights.
different to other prisoners. The President was therefore required in

law to consider all relevant information, including input from victims
and other interested persons before granting pardon. This, the Court
said, would be consonant with the basic values and principles
governing public administration set out in section 195, and the right
to administrative action protected by section 33 of the Constitution.48
Finally the Court considered it relevant that the TRC process was
transparent and allowed participation by victims49 and granted the
interim interdict applied for against the President preventing him
granting any pardon in terms of the ‘special dispensation’ before
victims had been provided with the relevant information and had
been given the opportunity to make representations.
Albutt approached the Constitutional Court for leave to appeal

directly against the decision of the North Gauteng High Court to grant
the interim interdict against the President. In the event of the Court
finding, like the High Court did, that the exercise of the constitutional
pardon was indeed administrative action, an order was sought for
declaring section 1 of the PAJA invalid. Even though it was against an
interim interdict, leave to appeal directly to the Constitutional Court
was granted in view of the fact that, in the words of the Chief Justice
the application raised ‘questions of considerable constitutional
importance concerning the powers of the President to grant pardon
under section 84(2)(j)’.50 With reference to Chonco 1 the Court
reiterated that ‘[w]hile there is no right to a pardon, the applicants
for pardon are at least entitled to have their applications considered
without delay’.51 Furthermore the Court found it necessary to deal
with the High Court's conclusion that the President's constitutional
function regarding the pardon ‘goes beyond the special dispensation
process’.52
The Court conceded the respondents' (the coalition of NGO's)

position that the President's refusal to allow representations to be
made to him by victims was irrational and that the context of the
special dispensation required the victims being given a hearing, but
preferred to leave the question whether the pardon constituted
administrative action open.53 The reasoning in the judgment is
significant.
48
Centre for the Study of Violence and Reconciliation (n 2 above) section 7.4.2 of
the judgment.
49 Centre for the Study of Violence and Reconciliation (n 2 above) section 7.4.3 of
the judgment.
52
Taking the supremacy of the Constitution as point of departure,

the Court again professed the ‘doctrine of legality, which is part of
the rule of law’. From this it follows that ‘the exercise of the power
to grant pardon must be rationally related to the purpose sought to be
achieved by the exercise of it’.54 Since the President's power to
pardon is derived from the Constitution, the manner in which it is
exercised must pass constitutional muster. To do so, the decision not
to allow victims to make representations in the special dispensation
process ‘must be rationally related to the achievement of the
objectives of the process’.55 Whereas the courts may not interfere in
the Executive's choice of the methods by which it performs its
functions, they must, when called upon to do so, objectively
determine ‘whether the means selected are rationally related to the
objective sought to be achieved’. The jurisdiction to do so equally
applies to the presidential pardon. 56
This finding opens, consistent with previous judgments,57 the

question whether the non-applicability of the PAJA (were that to be
decided) would have made a substantial difference.
5.3 The nature of the presidential pardon
In dealing with facts related to the ‘special dispensation’ announced

in 2007, the Court sought guidance from the history and underlying
sentiments of the TRC process and the objectives of ‘nation-building
and national reconciliation’58 and concluded that victim participation
was, as for the TRC process ‘fundamental to the special dispensation
process’.59
The Court rejected the suggestion, ironically made by the

President and the Minister's counsel, that there was a fundamental
difference between the pardon process and the TRC amnesty
process.60 In the seeming absence of arguments before the Court on
the acceptability of the President's tailoring of the pardon
54
55
57
Eg the Pharmaceutical judgment (n 37 above) where the Court held (para 79) that
the President's function to determine the date of the coming into effect of and
Act of Parliament in terms of the Act itself, was not administrative action but
that it lay 'between the law making process and the administrative process' but
that the exercise of the function was nevertheless reviewable in terms of the
rationality of the President’s decision (paras 89 and 90).
58
Albutt (n 3 above) paras 53-61. In para 52 the Court also pointed out that the
President, in terms of section 83(c) has a duty to promote 'the unity of the nation
and that which will advance the Republic.'
59
60 Albutt (n 3 above) paras 62, 63,& 65-67.
temporarily to fit the dimensions of amnesty,61 one must suppose that

the Court's underlying assumption here was that it was acceptable. By
implication the Court endorsed the temporary construction of pardon
as amnesty and struck down the High Court's approach that there was
no fundamental difference between the normal form of pardon and
the specially created amnesty-pardon.62 Fortunately, in his judgment
the Chief Justice specifically pointed out that he wished to deal only
with pardons under the ‘special dispensation’:63
What distinguishes this category from others not before us is that the
crimes in respect of which pardons are sought are alleged to have been
committed with a political motive; the objective of these pardons is to
promote national unity and reconciliation; and the crimes concerned
were committed in a particular historical context. Different
considerations may very well apply to other categories of applications
for pardon. This judgment does not therefore decide the question
whether victims of other categories of applications for pardon are
entitled to be heard. That question is left open.
This allows for the assumption that the presidential pardon is not in
future again to be construed as a form of amnesty. In a footnote in
the supporting judgment of three justices, reference is made to
judgments of courts in Trinidad and Tobago, the United Kingdom, New
Zealand and the United States apparently for the purpose of
demonstrating the deviation of the TRC amnesty process and Mbeki's
special dispensation from the normal conception of executive
pardon.64 This amounts to recognition that the post-TRC approach to
political pardon as amnesty confused pardon with amnesty.
It is to be hoped that this confusion will end with these cases and
that the presidential pardon will in future be distinguished from
political amnesty. If this distinction does not prevail, the doors may
be opened to political abuse of a constitutionally regulated
presidential function.
5.4 Pardon as executive action
Despite finding it to be unnecessary to decide the question whether

the exercise of the power to grant pardon constitutes administrative
action involving the PAJA, the Court remarked that:65
61 Proffering the argument would probably not have been considered by any of the
litigants to serve their interests in the case.
62 The Court decided (para 76) that the High Court had erred in this regard.
64
Albutt (n 3 above) para 88 footnote 2.
65 Albutt (n 3 above) para 80. Also paras 76 and 82.
[i]f one has regard to our jurisprudence, there is a substantial measure

of doubt as to whether the exercise of the pardon power constitutes
administrative action. Yet if this question is decided in the negative, a
more difficult question arises, namely, whether PAJA, upon its proper
construction, includes within its ambit the exercise of the power to
grant pardon. And if the answer to this question is in the affirmative,
more complex questions arise. Those questions are whether: (a) PAJA
merely regulates the exercise of the power or whether in effect it
reclassifies executive action as administrative action; and (b) whether it
is constitutionally permissible for the legislature to do either of these.
Having already answered what it considered to be the essential

questions in the case before it, the Court declined to say any more
about the nature of the general pardon power, whether it was
administrative action or whether the PAJA's constitutionality was in
question on this score.
However,remarkably, in considering the question whether the

President was justified in refusing victims to be heard in the process
of considering the applications for pardon under the special
dispensation, the Court utilised exactly such considerations as would
have been relevant if section 6(2)(f)-(h) of the PAJA were to apply,
namely reasonableness and rationality. However, for these
considerations to apply to the review of a public function, the PAJA is
not required: the constitutional imperative of legality suffices.
Whereas the term ‘reasonable’ is used in many provisions of the

Constitution,66 the word ‘rational’ is not to be found in its text. The
phrase ‘justifiable in an open and democratic society based on
freedom and equality’ in section 26(2) of the 1993 Constitution,
famously replicated in the limitation clause, section 36(1) of the 1996
Constitution and enhanced with ‘human dignity’, gave the
Constitutional Court cause for reading a rationality test into the
Constitution.67 Reasonableness and rationality — the latter being the
opposite of arbitrariness — have thus become judicial standards by
means of which the constitutionality of measures and actions of
organs of state taken in terms of policy, often where discretion comes
into play, is determined. This determination is required to be done
objectively. The approach was further canonised later in 2010 in the
judgment dealing with the amendment of the Road Accident Fund Act
56 of 1996 where paragraph 51 of the Albutt judgment was cited as
66
Sections 24-27, 29, 32, 33, 35-37, 41, 44, 59, 72, 80, 100, 103, 118, 122, 139, 146,
150, 160, 231 and 233 of the 1996 Constitution.
67
An early exposition of this test is to be found in S v Lawrence 1997 4 SA 1176 (CC)
para 41: 'The requirement that the measures be justifiable in an open and
democratic society based on freedom and equality means that there must be a
rational connection between means and ends. Otherwise the measure is arbitrary
and arbitrariness is incompatible with such a society.'
confirmation that it was now well settled law.68 The constitutional

demand for rationality in the performance of public functions, also
those of the head of state, clearly does not apply only to
administrative action but to all acts or omissions performed by those
bearing authority, including the President.
Not allowing the victims of applicants for pardon under the special
dispensation process to present the President with their views, was
held to be irrational against the background of the TRC amnesty
process where victims were given ample opportunity to testify and
make their feelings known. In view of the values of accountability,
responsiveness and openness enshrined in section 1(d) of the
Constitution, deciding whether there was a political motive behind
the commission of the crime concerned required more from the
President than merely considering the applicant's statement. ‘As a
matter of rationality’ the victims had to be heard.69
The considerations of reasonableness and rationality were

however not applied by the Court only to the special dispensation
situation. In the Albutt judgment the Court reconfirmed what was
stated in Chonco 170 regarding all forms of the pardon, therefore also
in the case of ‘ordinary’ pardons, namely that ‘the exercise of the
power to grant pardon must be rationally related to the purpose
sought to be achieved by the exercise of it’.71 Thus the undue delay
by the President to attend to the Chonco applications had to be
considered to be unreasonable.72
Applied to the presidential power of pardon, the reasonable-and-

rational test should mean that a refusal by the President to grant a
pardon or reprieve in an unreasonable and irrational manner may be
subjected to judicial review. If it is indeed true that the prerogative
nature of this ‘responsibility’ — as it is characterised by section 84(2)
of the Constitution — has been expunged and that it has become fully
constitutionalised, it follows that an applicant for pardon should be
provided with reasons when the application is refused in order to
enable a court to determine the reasonableness and rationality of the
refusal regardless of whether it is styled an ‘administrative act’ or
whether section 1(d) of the Constitution, section 33 of the
Constitution or the PAJA is applied.
68
Law Society of South Africa v Minister for Transport 2011 1 SA 400 (CC) paras 34-
37.
69
Albutt (n 3 above) paras 70-72 and 74.
71 Albutt (n 3 above) para 49 and 51. Cf the wording of sec 6(2)(f)(ii)(aa) of the
PAJA.
72 Cf also the wording of sec 6(2)(g) of the PAJA.
The TRC and Mbeki's special dispensation processes were unique

in that intangible considerations such as the political motivation of
the perpetrators and the ideals of national unity and reconciliation
were to be at the core of the decisions to grant or refuse amnesty and
pardon. Were the granting of pardon to political supporters and
refusal to others based on a president's political prejudices, or
conversely on a view that political opponents should be appeased by
granting pardons to their supporters, a court would have been able to
determine that it was irrational and therefore unreasonable.
This raises the question whether the Presidency's statement of 4

February 2010, which effectively stopped the proceedings in Chonco
2 merely by announcing that the applications for pardon had been
processed, the majority had been rejected and others were delayed
for reasons of similarity, should be considered to be reasonable. If so,
the power to pardon retains the essence of its historical nature as
‘royal prerogative’, that is, an act of discretionary grace based on the
subjective sentiment of mercy or clemency in the compassionate
heart of a king. If not, it must be considered to be exposed to judicial
review.
Furthermore, the granting or refusal of a pardon might even —

also in view of the fact that section 82(2)(j) of the Constitution is
pointedly not excluded from the definition of ‘administrative action’
in section 1 of the PAJA — be considered to be administrative action
subject to the giving of reasons under section 5 of the PAJA.
An argument that refusal of pardon on application does not affect

a right since a convicted person does not have a right to be pardoned,
would not be persuasive: the granting or refusal of such an application
at the very least affects the convicted person's — albeit justifiably
limited, but not deprived — right to freedom of movement.
6 Have we lost our way between amnesty and

pardon?
While it is to be hoped that future consideration of applications for

presidential pardon will no longer be complicated by the notion of
amnesty it might be asked what it takes to nudge an intractable
President into performing a constitutional function affecting a right
should he fail to do so expeditiously: pleading correspondence,
political pressure, oppositional litigation, or a strategy combining all
of these? As the facts of the Chonco case show, that may be the case
at the present stage of the development of South African politics and
constitutional development, but the real question should be: is the
position in law that the President's constitutional function to grant or
withhold the pardon an unqualified discretion in the nature of the
ancient royal prerogative and therefore beyond the reach of judicial

review? A positive response to this question would be unconscionable
in the constitutional state that the Republic of South Africa aspires to
become in terms of its Constitution.
President Mbeki's ‘special dispensation’ announced in 2007 linked

applications for pardon to the considerations underlying the unique
process of amnesty contemplated in the 1993 postamble. This, it is
submitted, confused the issue by imbuing the consideration of
applications for pardon with the element of political motivation. With
an attitude that reasons need not be given by the President for
granting or not granting such applications, it follows that refusal of an
application based on motives unpalatable to the President renders the
decision subjective, arbitrary and therefore irrational. That political
motives did influence the pardon of the 33 ANC and PAC members
cannot in the absence of reasons given be proven, but it would under
the circumstances prima facie appear to be the case. Similarly the
granting of parole as soon as possible within the framework of the
applicable legislation to convicted persons such as Shabir Shaik,73
Alan Boesak74 and Tony Yengeni75 — albeit not by the President, but
certainly within the reach of powerful presidential influence — can
hardly be separated from political considerations. Can it be
constitutionally acceptable to allow a president to pardon those with
political motives which he approves of and not to pardon others with
different political motives?
Historically and comparatively the pardon is intended merely to

provide a corrective on the judicial process where judicial
intervention has become impossible and good reasons exist for
executive intervention. In addition to the complication of the matter
by requiring political motivation behind the commission of the crimes
concerned in the Chonco and Albutt cases, the constitutional nature
of the pardon as being either a head-of-state, an executive or an
administrative act, has not been settled by the Constitutional Court.
For the purposes of determining whether the granting or not of an
application for pardon is judicially reviewable, the categorisation
thereof should however not make a difference.
If the consideration and grant of pardon is rooted solely in the

Constitution — as the Constitutional Court has indeed held it to be —
73
Released on medical parole only two years and four months after commencing a
15 year sentence incurred on the basis of a fraudulent relationship with Jacob
Zuma.
74
Who was found guilty of fraud and sentenced to three years' imprisonment, but
was released after one year in prison and received a presidential pardon four
years after his release, whereby his criminal record was expunged.
75
Also convicted of fraud but released on parole five months after commencing a
four year sentence.
its prerogative origins of royal discretion and grace dependent upon

the mercy of the monarch cannot still be a component thereof in any
form: that such a result would be incompatible with democracy was
already pointed out by Blackstone in 1753!76 In fact, the supremacy
of the Constitution and the ubiquitous relevance of legality speak
loudly for the need to reduce the presidential pardon to the level of
an act which requires rational, reasonable and motivated action fully
open to judicial review.
76
Blackstone Commentaries as quoted in n 13 above: see italicised phrase in the
citation.

2012 Constitutional Court Review 2011 4

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2012 Constitutional Court Review 2011 4

Uploaded by

Copyright:

Available Formats

CONSTITUTIONAL

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

Printed and bound by:

The financial contribution of the Konrad Adenauer Stiftung is gratefully

The Constitutional Court Review, an initiative of Constitutional law of

Subscriptions through agents or bookshops:

Subscriptions directly through PULP:

Subscriptions by registered students directly through PULP:

Postage and packaging charges are:

Subscriptions should be directed to the publisher:

The South African Constitutional Court and socio-economic rights 1

Rationality, reasonableness, proportionality: Testing the 31

The democratic turn and (the limits of) constitutional patriotism 51

Customary (communal) land tenure in South Africa: Did Tongoane 73

Rationality, the rule of law, and the sovereign return 113

Three-level games: Thoughts on Glenister, Scaw and 137

Rosalind Dixon* & Tom Ginsburg**

Socio-economic rights are a central terrain of struggle in new

The South African Constitutional Court, in the 2010 Term, heard

* Assistant Professor, University of Chicago Law School, Professor, UNSW Faculty of

part of the right of access to housing under section 26 of the

The 2010 Term was also book-ended by numerous cases involving

In 2011, in Gundwana v Steko Development CC, the Court again

In several of these cases, the Court also confronted a potential

In Abhalali, the legislation in question was challenged by the

16 Gundwana v Steko Development CC and Others 2011 8 BCLR 792 (CC).

The 2009-2011 Terms, therefore, provide a natural opportunity to

For left-wing parties to constitutional negotiations, the inclusion

A more realistic alternative for such parties, therefore, will in

bargain in a way that is more efficient than in the case of a one-way

This account of socio-economic rights is also helpful to

For the Constitutional Court, an insurance swap-based theory of

Our suggested approach also accords surprisingly well, we

The essay proceeds in four parts, following this introduction. Part

review, and explains how such a theory can be expanded to account

2 Socio-economic rights as insurance swaps

Our theory draws from earlier work on the ‘insurance’ function of

The insurance model has been applied with some success in

Socio-economic rights present some challenge for basic insurance

economic rights in a constitution. It is perfectly understandable why

In the ordinary context of constitutional negotiations, judicial

For left-wing parties, however, there is also the danger that

Three basic responses emerge in the face of this kind of conflict.

29 On the signaling function of socio-economic rights provisions more generally, see

return for concessions on other issues.31 A second option will be for

A good example of how socio-economic rights guarantees can play

Certain rights, such as the right of access to housing, or land,

validity of a change in zoning law allowing the building of a nursing

For left-wing parties to constitutional negotiations, therefore, the

In an economic context, a swap is a contract for an exchange of

Both parties under such an arrangement hold mutual cross-

Why might parties decide to adopt a ‘swap arrangement’ of this

Parties, however, may also have a strong interest in providing —

Preferences of this kind will mean that the parties to

guarantees and mere ‘directive principles’ of state policy.42 That is,

One bit of evidence for the plausibility of such constitutional

In the South African context, we argue, sections 25-27 of the

3 The South African constitution and insurance

The 1996 Constitution was, of course, the product of a unique two-

The negotiation of the interim Constitution itself also took place

The constitutional protection of property was from the outset of

addressing homelessness.47 A number of leading ANC thinkers

It soon became clear, however, that this position was politically