National Director of Public Prosecutions and Others v Freedom Under Law (67/2014) [2014] ZASCA 58; 2014 (4) SA 298 (SCA); 2014 (2) SACR 107 (SCA); [2014] 4 All SA 147 (SCA) (17 April 2014)

About SAFLII
Databases
Search
Terms of Use
RSS Feeds
South Africa: Supreme Court of Appeal
SAFLII
>>
Databases
>>
South Africa: Supreme Court of Appeal
>>
2014
>>
[2014] ZASCA 58
|

|

National Director of Public Prosecutions and Others v Freedom Under Law (67/2014) [2014] ZASCA 58; 2014 (4) SA 298 (SCA); 2014 (2) SACR 107 (SCA); [2014] 4 All SA 147 (SCA) (17 April 2014)

Links to summary

THE
SUPREME COURT OF APPEAL OF SOUTH AFRICA
JUDGMENT
REPORTABLE
Case
No: 67/2014
In
the matter between:
NATIONAL
DIRECTOR OF PUBLIC
PROSECUTIONS
...........................................................................................
FIRST
APPELLANT
THE
HEAD: SPECIALISED COMMERCIAL
CRIME
UNIT
.............................................................................................
SECOND
APPELLANT
THE
NATIONAL COMMISSIONER: SOUTH
AFRICAN
POLICE
SERVICE
..............................................................
THIRD RESPONDENT
RICHARD
NAGGIE
MDLULI
..............................................................
FOURTH APPELLANT
v
FREEDOM
UNDER
LAW
.....................................................................................
RESPONDENT
Neutral
citation:
National Director of
Public Prosecutions v Freedom Under Law
(67/14)
[2014] ZASCA 58
(17 April 2014).
Coram:
Mthiyane DP, Navsa, Brand, Ponnan
et
Maya JJA
Heard:
1 April 2014
Delivered:
17 April 2014
Summary:
Review application – decisions
to withdraw criminal charges by National Prosecuting Authority –
reviewable on principle
of legality not under the
Promotion of
Administrative Justice Act 3 of 2000
– decisions by
Commissioner of Police to terminate disciplinary proceedings and lift
suspension of member – reviewed
and set aside under
s 6
of
PAJA – not competent for the high court to issue mandatory
interdicts to compel prosecution and disciplinary charges.
ORDER
On
appeal from:
North Gauteng High Court,
Pretoria (Murphy J sitting as court of first instance):
1 The appeal
succeeds only to the extent that paragraphs (b), (e) and (f) of the
order of the court a quo are set aside
2 The orders in
paragraphs (
a
), (
c
), (
d
), (
g
) and (
h
)
of the order by the court a quo are confirmed but re-numbered in
accordance with the changes necessitated by the setting aside
of the
orders in paragraph 1.
3 It is recorded
that the following undertaking has been furnished on behalf of the
first respondent:
(a)
To decide which of the criminal charges of murder and related crimes
that were withdrawn on 2 February 2012, are to be reinstituted
and to
make his decision known to the respondent within 2 months of this
order.
(b)
To provide reasons to the respondent within the same period as to why
he decided not to reinstitute some – if any –
of those
charges.
4.
There shall be no order as to costs in respect of the appeal.
JUDGMENT
Brand
JA
(Mthiyane DP, Navsa, Ponnan
et
Maya JJA concurring):
[1]
This is an appeal against an order of the high court granted at the
behest of the respondent. In substance the order reviewed
and set
aside four decisions taken by or on behalf of the first three
appellants in favour of the fourth appellant and directed
the first
three respondents to reinstate criminal prosecutions and disciplinary
proceedings against him. The appeal is with the
leave of the court a
quo. More precise details of the order appealed against will appear
from the exposition of the background
that follows. I find it
convenient to start that exposition by presentation of the parties.
The
Parties
[2]
The first appellant is the National Director of Public Prosecutions
(NDPP). Advocate Nomgcobo Jiba was appointed on 28 December
2011 as
the acting NDPP by the President of the Republic after the suspension
from that office of the then incumbent, Mr Menzi
Simelane in
consequence of a judgment of this court. The second appellant is
Advocate Lawrence Mrwebi (Mrwebi) who was appointed
on 1 November
2011 as Special Director of Public Prosecutions as the Head of the
Specialised Commercial Crimes Unit (SCCU) of the
National Prosecuting
Authority.
[3]
The third appellant is the National Commissioner of the South African
Police Service (the Commissioner). During the time period
relevant to
these proceedings that position was occupied first by General Bheki
Cele, thereafter by Lieutenant General Nhlanhla
Mkhwanazi, in an
acting capacity and finally by General Mangwashi Victoria Phiyega.
The fourth appellant, who took centre stage
in these proceedings, is
Lieutenant General Richard Mdluli (Mdluli) who held the office of
National Divisional Commissioner: Crime
Intelligence in the South
African Police Service (SAPS), a position also described as Head of
Crime Intelligence, since 1 July
2009.
[4]
The respondent, Freedom Under Law, is a public interest organisation,
registered as a non-profit company with offices in South
Africa and
Switzerland. It is actively involved, inter alia, in the promotion of
democracy and the advancement of respect for the
rule of law in the
Southern African region. Both its board of directors and its advisory
board are composed of respected lawyers,
judges and other leading
figures in society at home and abroad.
Background
[5]
It is common cause that on 31 March 2011 Mdluli was arrested and
charged with 18 criminal charges, including murder, intimidation,

kidnapping, assault with intent to do grievous bodily harm and
defeating the ends of justice. The murder charge stemmed from the

killing of Mr Tefo Ramogibe (the deceased) on 17 February 1999. From
about 1996 until 1998 the deceased and Mdluli were both involved
in a
relationship with Ms Tshidi Buthelezi. The deceased and Buthelezi
were secretly married during 1998. Mdluli was upset about
this and
addressed the issue on numerous occasions with Ms Buthelezi and the
deceased and members of their respective families.
At the time Mdluli
held the rank of senior superintendent and the position of commander
of the detective branch at the Vosloorus
police station. Charges of
attempted murder, intimidation, kidnapping, et cetera, rested on
allegations by relatives and friends
of the deceased and Ms Buthelezi
that Mdluli and others associated with him – including
policemen under his command –
brought pressure to bear upon
them through violence, assaults, threats, kidnappings and in one
instance rape, with the view to
compelling their co-operation in
securing the termination of the relationship between the deceased and
Ms Buthelezi. According
to one of the complainants who is the mother
of the deceased, Mdluli had on occasion taken her to the Vosloorus
police station
where she found the deceased injured and bleeding. In
her presence Mdluli then warned the deceased to stay away from Ms
Buthelezi.
The deceased was killed a few days thereafter.
[6]
On 23 December 1998 the deceased was the victim of an attempted
murder. He reported the incident to the Vosloorus police station.
On
17 February 1999 the deceased and the investigating officer, Warrant
Officer Dhlomo, drove to the scene in Mdluli’s official
vehicle
for the stated purposes of the deceased participating in a
pre-arranged pointing out. According to Dhlomo they were attacked
by
two unknown assailants at the scene who shot at them and took away
his firearm and the vehicle in which they were travelling.
He ran to
a nearby tuck-shop to summon the police. Upon his return he found
that the deceased had been killed. At the time, the
matter never
proceeded to trial. Much of the original docket and certain exhibits
have since been lost or have disappeared.
[7]
Information about the discontinued investigation re-surfaced after
Mdluli was appointed the Head of Crime Intelligence in 2009.
Two
senior officers of the Directorate of Priority Crime Investigation
(the Hawks), Colonel Roelofse and Lieutenant-Colonel Viljoen,
were
appointed to assist in the renewed investigations and Mdluli came to
be arrested on these charges – to which I shall
refer as the
murder and related charges – on 31 March 2011. In the light of
the seriousness of these charges, the then Commissioner
of Police,
General Bheki Cele, suspended Mdluli from office on 8 May 2011 and
instituted disciplinary proceedings against him.
[8]
After Mdluli’s arrest on the murder and related charges, some
members of Crime Intelligence came forward with information

concerning alleged crimes committed by some of its members, including
Mdluli. Lieutenant Colonel Viljoen, who was involved in the

investigation of the murder and related charges, was instructed to
investigate these allegations in conjunction with Advocate C
Smith of
the Specialised Commercial Crime Unit (SCCU). Following upon these
investigations, Smith successfully applied for a warrant
for Mdluli’s
arrest on charges of fraud and corruption which was executed on 20
September 2011.
[9]
What emerges from the papers filed of record is that the charges of
fraud and corruption originate from the alleged unlawful
utilisation
of funds held in the Secret Service account – created in terms
of the Secret Services Act 56 1978 – for
the private benefit of
Mdluli and his wife, Ms Theresa Lyons. Broadly stated it is alleged
that one of Mdluli’s subordinates,
Colonel Barnard, purchased
two motor vehicles ostensibly for use by the Secret Service but
structured the transaction in such a
manner that a discount of
R90 000 that should have been credited to the Secret Service
account, was utilised for Mdluli’s
personal benefit. The
further allegation was that those two motor vehicles were then
registered in the name of Mdluli’s wife
and appropriated and
used by the two of them.
[10]
On 3 November 2011 Mdluli wrote a letter to President Zuma, the
Minister of Safety and Security and the Commissioner stating
that the
charges against him were the result of a conspiracy among senior
police officers – including the then Commissioner,
General
Bheki Cele, and the head of the Hawks, General Anwar Dramat. The
letter also stated, rather inappropriately, that ‘[i]n
the
event that I come back to work, I will assist the President to
succeed next year’ which was an obvious reference to the

forthcoming presidential elections of the ruling African National
Congress in Mangaung towards the end of 2012. The allegations
of a
conspiracy led to the appointment by the Minister of a task team
which later reported that there was no evidence of a conspiracy
and
that the police officers who had accused Mdluli of criminal conduct
had acted in good faith.
[11]
On 17 November 2011 Mdluli’s legal representatives made
representations to Mrwebi in his capacity as Special DPP and
head of
the SCCU, seeking the withdrawal of the fraud and corruption charges.
These representations again contended that the charges
against Mdluli
resulted from a conspiracy against him involving the most senior
members of the South African Police Service. The
representations also
indicated that a similar approach had been made to Advocate K M A
Chauke, the DPP South Gauteng,
for withdrawal of the murder and
related charges. Mrwebi, in response to the representations made to
him, requested a report from
Smith and his immediate superior,
Advocate Glynnis Breytenbach, who both responded with a motivation
that the charges should not
be withdrawn. Despite this motivation,
Mrwebi decided to withdraw these charges and notified Mdluli’s
representatives of
his decision to do so on or about 5 December 2011.
The circumstances under which Mrwebi’s decision was arrived at
is central
to one of the disputes in this case. I shall revert to
this in due course.
[12]
On 1 February 2012 Chauke decided to withdraw the murder and related
charges as well. He explained that after he received the

representations by Mdluli’s legal representatives, he realised
that there was no direct evidence implicating Mdluli in the
murder
charge. He therefore decided that an inquest should be held before he
proceeded with that charge and that the murder charge
should
therefore be provisionally withdrawn pending the outcome of the
inquest. To prevent fragmented trials, so he said, he decided
that
the 17 charges related to the murder should also be provisionally
withdrawn, pending finalisation of the inquest.
[13]
I pause to record that at Chauke’s request the inquest was held
in terms of the Inquests Act 58 of 1959 by the magistrate
of Boksburg
who handed down his reasons and findings on 2 November 2012. His
ultimate conclusions make somewhat peculiar reading,
namely that:
‘
The
theory of Mdluli being the one who had orchestrated the death of [the
deceased] is consistent with the facts.’
And
that:
‘
The
death [of the deceased] was brought about by an act
prima
facie
amounting to an offence on the
part of
unknown persons
.
There is
no evidence
on a balance of probabilities implicating Richard Mdluli [and his
co-accused persons] in the death of the deceased.’
[14]
I say peculiar, because s 16(2) of the Inquests Act required the
magistrate to determine whether the death of the deceased
was brought
about by any act or omission amounting to an offence on the part of
any person. The evidence before him clearly established
a prima facie
case against Mdluli. That appears to be borne out by the first
conclusion. The second conclusion, which appears to
contradict the
first seems to be both unhelpful and superfluous. It was not for the
magistrate to determine Mdluli’s guilt
on a murder charge,
either beyond reasonable doubt or on a balance of probabilities. But
if Chauke had any uncertainty about the
import of the magistrate’s
findings he could have asked for clarification or even requested that
the inquest be re-opened
in terms of s 17(2) of the Inquests
Act. Furthermore, it is clear that the magistrate’s findings
were wholly irrelevant
to the 17 related charges. Nonetheless it is
common cause that no further steps have since been taken by the
prosecuting authorities
to reinstitute any of the 18 charges.
[15]
I return to the chronological sequence of events. On 29 February 2012
the Acting National Commissioner of Police at the time,
General
Mkhwanazi, withdrew the disciplinary proceedings against Mdluli and
on 31 March 2012 he was reinstated and resumed his
office as Head of
Crime Intelligence. In fact, shortly thereafter, his duties were
extended to include responsibility for the unit
which provides
protection for members of the national executive.
[16]
On 15 May 2012 FUL launched the application, the subject of the
present appeal. The notice of motion contemplated proceedings
in two
parts. Part A sought an interim interdict, essentially compelling the
Commissioner to suspend Mdluli from office pending
the outcome of the
review application in part B. In part B FUL sought an order reviewing
and setting aside four decisions, namely:
(a) The decision
made by Mrwebi on or about 5 December 2011 to withdraw the charges of
fraud and corruption.
(b) The decision by
Chauke on or about 2 February 2012 to withdraw the murder and related
charges.
(c) The decision by
the Commissioner of Police on or about 29 February 2012 to terminate
the disciplinary proceedings; and
(c) The decision by
the Commissioner on or about 31 March 2012 to reinstate Mdluli to his
office.
[17]
Apart from the orders setting aside the four impugned decisions, FUL
also sought mandatory interdicts:
(a) directing the
prosecution authorities to reinstate the criminal charges against
Mdluli and to ensure that the prosecution of
these charges are
enrolled and pursued without delay; and
(b) directing the
Commissioner of Police to take all steps necessary for the
prosecution and finalisation of the disciplinary charges.
On
6 June 2012 the interim interdict sought in part A was granted by
Makgoba J. The application for leave to appeal against that
order was
unsuccessful and the interim interdict is thus extant. The review
application came before Murphy J who granted an order
(a) setting
aside the four impugned decisions as well as (b) the mandatory
interdict sought together with (c), an order for costs
in favour of
FUL against the respondents. His judgment has since been reported sub
nom
Freedom Under Law v National
Director of Public Prosecutions & others
2014 (1) SA 254
(GNP).
FUL’s
locus standi
[18]
I now turn to the appellant’s contentions on appeal and I deal
first with those arising from challenges by the NDPP and
Mrwebi.
These relied mainly on formal and procedural objections rather than
the merits of the case. Included amongst these formal
objections was
a challenge to FUL’s legal standing. However, this challenge
was not pursued in argument. Suffice it therefore
to say that in my
view the objection to FUL’s standing was unsustainable from the
start. FUL’s mission to promote accountability
and democracy
and to advance respect for the rule of law and the principle of
legality in this country has been recognised by this
court (see eg
Freedom Under Law v Acting Chairperson
Judicial Service Commission & others
2011
(3) SA 549
(SCA) paras 19-21). In addition, I agree with the finding
by the court a quo that the matter is one of public interest and
national
importance (para 1 of its judgment). What I do find somewhat
perturbing is the court’s high praise for Dr Mamphela Ramphele

and Justice Johan Kriegler who deposed to FUL’s founding and
replying affidavits respectively (see para 4). It needs to be

emphasised that all litigants, irrespective of their status, should
be treated equally by our courts. Judges must therefore be
wary of
creating the impression – which would undoubtedly be unfounded
in this case – that they have more respect for
some litigants
or their representatives than for others.
Reviewability
of decisions to withdraw a prosecution
[19]
The next challenge by the NDPP, which was embraced by Mrwebi and
Mdluli, related to the reviewability of a prosecutorial decision
to
discontinue a prosecution. The issue arising from this is a narrow
one. This is so because it is not contended by the NDPP that

decisions of this kind are not reviewable at all. On the contrary,
the NDPP conceded that these decisions are subject to what has
become
known as a principle of legality or a rule of law review by the
court. The allied issue is whether these decisions are reviewable

under the provisions of the Promotion of Administrative Justice Act 3
of 2000 (PAJA). Although the answer to that question is by
no means
decisive of the matter. I nonetheless believe the time has come for
this court to put the issue to rest. This belief is
motivated by two
considerations. First, because the court a quo had pronounced on the
question and held that PAJA is of application
(paras 131-132 of the
judgment). Secondly, and more fundamentally, by the considerations
that appear from the following statement
by Ngcobo J in
Minister
of Health & another NO v New Clicks South Africa (Pty) Ltd &
others (Treatment Action Campaign & another
as Amici Curiae)
2006 (2) SA 311
(CC) paras 436-438:
‘
Our
Constitution contemplates a single system of law which is shaped by
the Constitution. To rely directly on s 33(1) of the

Constitution and on common law when PAJA, which was enacted to give
effect to s 33, is applicable, is, in my view, inappropriate.
It
will encourage the development of two parallel systems of law, one
under PAJA and another under s 33 and the common law
. . .
Legislation enacted by Parliament to give effect to a constitutional
right ought not to be ignored. And where a litigant
founds a cause of
action on such legislation, it is equally impermissible for a court
to bypass the legislation and to decide the
matter on the basis of
the constitutional provision that is being given effect to by the
legislation in question . . . It follows
that the SCA . . . erred in
failing to consider whether PAJA was applicable. The question whether
PAJA governs these proceedings
cannot be avoided in these
proceedings.’
[20]
The domain of judicial review under PAJA is confined to
‘administrative action’ as defined in s 1 of the
Act. The definition starts out from the premise that ‘administrative
action’ is ‘any decision taken, or any failure
to take a
decision, by . . . a natural or juristic person . . . when exercising
a public power or performing a public function
in terms of an
empowering provision, which adversely affects the rights of any
person and which has direct, external legal effect
. . .’.
Mrwebi and Chauke derived their power to withdraw the criminal
charges against Mdluli from the provisions of the
National
Prosecuting Authority Act 32 of 1998 (the NPA Act). On the face of
it, their decisions sought to be impugned in this case
clearly
constituted ‘administrative action’. But s 1(
ff
)
of the definition excludes ‘a decision to institute or continue
a prosecution’. The question in the present context
is thus –
does the exception extend to its converse as well, namely a decision
not to prosecute or to discontinue a prosecution?
[21]
Cora Hoexter
Administrative Law in South Africa
(2 ed, 2012)
at 241-242 is of the firm view that the intention behind the
exception ‘was to confine review under the PAJA
to decisions
not
to prosecute. There is less need to review decisions to
prosecute or to continue a prosecution as types of administrative
action,
since such decisions will ordinarily result in a trial in a
court of law’. Thus far our courts have, however, been less
decisive.
In
Kaunda & others v President of the Republic of
South Africa & others
2005 (4) SA 235
(CC) para 84 Chaskalson
CJ acknowledged that:
‘
In
terms of the [PAJA] a decision to institute a prosecution is not
subject to review. The Act does not, however, deal specifically
with
a decision not to prosecute. I am prepared to assume in favour of the
applicants that different considerations apply to such
decisions [as
opposed to the decision to institute a prosecution] and that there
may possibly be circumstances in which a decision
not to prosecute
could be reviewed by a Court. But even if this assumption is made in
favour of the applicants, they have failed
to establish that this is
a case in which such a power should be exercised.’
[22]
The implication is therefore that decisions not to prosecute are not
necessarily excluded from the application of PAJA. Conversely,
in
Democratic Alliance & others v Acting National Director of
Public Prosecutions & others
2012 (3) SA 486
(SCA) para 27
Navsa JA stated:
‘
While
there appears to be some justification for the contention that the
decision to discontinue a prosecution is of the same genus
as a
decision to institute or continue a prosecution, which is excluded
from the definition of “administrative action”
in terms
of s 1(
ff
)
of PAJA, it is not necessary to finally decide that question. Before
us it was conceded on behalf of the first and third respondents
that
a decision to discontinue a prosecution was subject to a rule of law
review. That concession in my view was rightly made.’
[23]
The court a quo (in paras 131-132 of its judgment) found itself in
disagreement with what it described as the
obiter
dictum
of Navsa JA that a decision to
discontinue prosecution is of the same genus as a decision to
prosecute. ‘For the reasons
stated by Professor Hoexter’
so it held, ‘a decision of non-prosecution is of a different
genus to one to institute
a prosecution. It is final in effect in a
way that a decision to prosecute is not’.
[24]
However, unlike the court a quo I am not persuaded by the reasoning
advanced by Professor Hoexter for the view that she proffers.
To say
that the validity of a decision to prosecute will be tested at the
criminal trial which is to follow, is, in my view, fallacious.
What
is considered at the criminal trial is a determination on all of the
evidence presented in the case of the guilt or lack thereof
of the
accused person, not whether the preceding decision to prosecute was
valid or otherwise. The fact that an accused is acquitted

self-evidently does not suggest that the decision to prosecute was
unjustified. The reason advanced by the court a quo itself,
namely,
that a decision not to prosecute is final while a decision to
prosecute is not, is in my view equally inaccurate. Speaking

generally, both these decisions can be revisited through subsequent
decisions by the same decision-maker, by in the one case
re-instituting
the prosecution, and by withdrawing the prosecution in
the other.
[25]
What is called for, as I see it, is to focus on the policy
considerations that underlie the exclusion of a decision to institute

or continue to prosecute from the ambit of PAJA and to reflect on
whether or not the same considerations of policy will apply to
a
decision not to prosecute or to discontinue a prosecution. In
National Director of Public Prosecutions
v Zuma
[2009] ZASCA 1
;
2009 (2) SA 277
(SCA) para 35 fn
31 Harms DP cited a line of English cases that emphasised the same
policy considerations that underlie the exclusion
of decisions to
prosecute from the PAJA definition of administrative action. These
included
Sharma v Brown-Antoine and
others
[2006] UKPC 57
;
[2007] 1 WLR 780
(PC) para 14
and
Marshall v The Director of Public
Prosecutions (Jamaica)
[2007] UKPC 4
para 17. The first principle established by these cases, as I see it,
is that in England, decisions to prosecute are not immune
from
judicial review but that the courts’ power to do so is
sparingly exercised. The policy considerations for courts limiting

their own power to interfere in this way, appear to be twofold.
First, that of safeguarding the independence of the prosecuting

authority by limiting the extent to which review of its decisions can
be sought. Secondly, the great width of the discretion to
be
exercised by the prosecuting authority and the polycentric character
that generally accompanies its decision-making, including

considerations of public interest and policy.
[26]
As I see it, the underlying considerations of policy can be no
different with regard to decisions not to prosecute or to discontinue

a prosecution. This view is supported by English authorities dealing
with non-prosecution. So, for instance it was said in
R v Director
of Public Prosecutions, Ex Parte Manning
[2001] QB 330
para 23:
‘
[T]he
power of review is one to be sparingly exercised. The reasons for
this are clear. The primary decision to prosecute or not
to prosecute
is entrusted by Parliament to the [prosecutor] as head of an
independent, professional prosecuting service, answerable
to the
[National Director of Public Prosecutions] in his role as guardian of
the public interest, and to no-one else.’
And
by Kennedy LJ in
R v Director of Public Prosecutions, Ex Parte C
[1995] 1 Cr App R 136 at 139G-140A:
‘
It
has been common ground before us in the light of the authorities that
this Court does have power to review a decision of the
Director of
Public Prosecutions not to prosecute, but the authorities also show
that the power is one to be sparingly exercised.’
At
141B-C Kennedy LJ then continued to say:
‘
From
all of those decisions it seems to me that in the context of the
present case this court can be persuaded to act if and only
if it is
demonstrated to us that the Director of Public Prosecutions . . .
arrived at the decision not to prosecute . . .’
Whereupon,
he proceeded to set out the grounds recognised by the English courts
for interference in decisions not to prosecute.
Suffice it to say
these grounds are substantially similar to the ones recognised by our
courts as justification for a rule of law
review. The dictum from
Kaunda
does not indicate that a PAJA review might be
available, but on the assumption made, the suggestion appears to be
that in appropriate
circumstances a rule of law review might be
apposite.
[27]
My conclusion from all this is that:
(a) It has been
recognised by this court that the policy considerations underlying
our exclusion of a decision to prosecute from
a PAJA review is
substantially the same as those which influenced the English courts
to limit the grounds upon which they would
review decisions of this
kind.
(b) The English
courts were persuaded by the very same policy considerations to
impose identical limitations on the review of decisions
not to
prosecute or not to proceed with prosecution.
(c) In the present
context I can find no reason of policy, principle or logic to
distinguish between decisions of these two kinds.
(d) Against this
background I agree with the
obiter dictum
by Navsa JA in
DA
& others v Acting NDPP
that decisions to prosecute and not to
prosecute are of the same genus and that, although on a purely
textual interpretation the
exclusion in s 1(
ff
) of PAJA
is limited to the former, it must be understood to incorporate the
latter as well.
(e)
Although decisions not to prosecute are – in the same way as
decisions to prosecute – subject to judicial review,
it does
not extend to a review on the wider basis of PAJA, but is limited to
grounds of legality and rationality.
[28] The legality
principle has by now become well-established in our law as an
alternative pathway to judicial review where PAJA
finds no
application. Its underlying constitutional foundation appears, for
example, from the following dictum by Ngcobo J in
Affordable
Medicines Trust & others v Minister of Health & others 2
006
(3) SA 247
(CC) para 49:
‘
The
exercise of public power must therefore comply with the Constitution,
which is the supreme law, and the doctrine of legality,
which is part
of that law. The doctrine of legality, which is an incident of the
rule of law, is one of the constitutional controls
through which the
exercise of public power is regulated by the Constitution.’
[29]
As demonstrated by the numerous cases since decided on the basis of
the legality principle, the principle acts as a safety
net to give
the court some degree of control over action that does not qualify as
administrative under PAJA, but nonetheless involves
the exercise of
public power. Currently it provides a more limited basis of review
than PAJA. Why I say currently is because it
is accepted that
‘[l]egality is an evolving concept in our jurisprudence, whose
full creative potential will be developed
in a context-driven and
incremental manner’ (see
Minister
of Health NO v New Clicks SA (Pty) Ltd & others
2006
(2) SA 311
(CC) para 614; Cora Hoexter op cit at 124 and the cases
there cited). But for present purposes it can be accepted with
confidence
that it includes review on grounds of irrationality and on
the basis that the decision-maker did not act in accordance with the

empowering statute (see
Democratic
Alliance & others v Acting National Director of Public
Prosecutions & others
2012 (3) SA
486
(SCA) paras 28-30).
Impugned
decisions to withdraw criminal charges only provisional and not final
[30] This brings me
to the further technical challenge by the NDPP, namely that the
impugned decisions by Mrwebi and Chauke were
not final, but only
provisional. The contentions underlying this challenge will be better
understood against the statutory substructure
of these decisions
which is to be found in s 179 of the Constitution, read with the
relevant provisions of the NPA Act. Under
the rubric ‘prosecuting
authority’ s 179 of the Constitution provides in relevant
part:
‘
(1)
There is a single national prosecuting authority in the Republic,
structured in terms of an Act of Parliament, and consisting
of
(
a
)
National Director of Public Prosecutions, who is the head of the
prosecuting authority, and is appointed by the President, as
head of
the national executive; and
(
b
)
Directors
of Public Prosecutions and prosecutors as determined
by an Act of Parliament.
(2)
The prosecuting authority has the power to institute criminal
proceedings on behalf of the state, and to carry out any necessary

functions incidental to instituting criminal proceedings.
(3)
. . .
(4)
National legislation must ensure that the prosecuting authority
exercises its functions without fear, favour or prejudice.
(5)
The National Director of Public Prosecutions
(
a
)
(
b
)
(
c
)
(
d
)
may review a decision to prosecute or not to prosecute, after
consulting the relevant Director of Public Prosecutions and after

taking representations within a period specified by the National
Director of Public Prosecutions, from the following:
(i)
The accused person.
(ii)
The complainant.
(iii)
Any other person or party whom the National Director considers to be
relevant.’
[31]
The national legislation contemplated in s 179 of the
Constitution was promulgated in the form of the NPA Act. The power
to
institute and conduct criminal proceedings is given legislative
expression in s 20 which provides:
‘
(1)
The power as contemplated in section 179(2) and all other relevant
sections of the
Constitution
to –
(
a
) institute
and conduct criminal proceedings on behalf of the State;
(
b
) carry out
any necessary functions incidental to instituting and conducting such
criminal proceedings; and
(
c
)
discontinue criminal proceedings,
vests in the
prosecuting authority
and shall, for all purposes, be
exercised on behalf of the
Republic
.
(2) . . .
(3) Subject to the
provisions of the
Constitution
and
this Act
, any
Director
[defined in s 1 as a DPP] shall, subject to the
control and directions of the
National Director
, exercise the
powers referred to in subsection (1) in respect of –
(
a
) the area
of jurisdiction for which he or she has been appointed; and
(
b
)
. . ..’
[32]
Mrwebi and Chauke, who were both DPPs, were therefore authorised by
s 20(3), read with s 20(1)(
c
), to withdraw the
criminal charges against Mdluli. But because Mrwebi was appointed as
a special DPP his powers were limited by
the provisions of s 24(3)
which provides:
‘
A
Special Director shall exercise the powers . . . assigned to him or
her by the President, subject to the directions of the National

Director: Provided that if such powers . . . include any of the
powers . . . referred to in section 20(1), they shall be exercised
.
. . in consultation with the Director of the area of jurisdiction
concerned.’
[33]
According to the NDPP’s argument, the withdrawal of the
criminal charges in this case must also be understood against
the
background of s 6 of the Criminal Procedure Act 51 of 1977 (the
CP Act). This section draws a distinction between the
withdrawal of
criminal charges, before an accused person has pleaded – in
s 6(
a
)
– and the stopping of a prosecution after the accused person
has pleaded, as contemplated in s 6(
b
)
.
The latter section provides that where the prosecution is stopped the
court is obliged to acquit the accused person, while a withdrawal
in
terms of s 6(
a
)
does not have that consequence. A charge withdrawn under s 6(
a
)
can therefore be reinstituted at any time.
[34]
The withdrawal of charges by Mrwebi and Chauke, so the NDPP’s
argument went, was covered by s 6(
a
)
and not by s 6(
b
).
In consequence, so the argument proceeded, these decisions were only
provisional and therefore not subject to review. Although
I am in
agreement with the premise of the argument, that both decisions to
withdraw were taken in terms of s 6(
a
),
my difficulty with its further progression is twofold. First, I can
see no reason why, at common law, a decision would in principle
be
immune from judicial review just because it can be labelled
‘provisional’ however illegal, irrational and prejudicial

it may be. My second difficulty is more fundamental. I do not believe
a decision to withdraw a criminal charge in terms of s 6(
a
)
can be described as ‘provisional’ just because it can be
reinstituted. It would be the same as saying that because
a charge
can be withdrawn, the institution of criminal proceedings is only
provisional. As I see it, the withdrawal of a charge
in terms of
s 6(
a
)
is final. The prosecution can only be recommenced by a different,
original decision to reinstitute the proceedings. Unless and
until it
is revived in this way, the charge remains withdrawn.
[35]
The NDPP’s second argument as to why the impugned decisions
were not final rests on the provisions of s 179(5)(
d
)
of the Constitution. Since in terms of this section the decisions
were still subject to review by the NDPP, so the argument went,
they
were only provisional. I have already expressed my reservations about
the proposition that because a decision is provisional
it is not
subject to challenge, based on legality or rationality. What the
NDPP’s argument based on s 175(5)(
d
)
mutated to was the contention that, because the impugned decisions
were subject to an internal review, FUL should have been non-suited

for failure to exhaust the internal remedies available to it. That,
of course, is a completely different case.
Exhaustion
of internal remedy
[36]
The NDPP’s final argument as to why review proceedings were not
competent, was that FUL had failed to exhaust an internal
remedy
available to it. What this contention relied upon was the provision
in s 179(5)
(d),
which enables the NDPP to review a decision not to prosecute at the
behest of any person or party who the NDPP considers to be
relevant.
Since I have found a review under PAJA unavailable, s 7(2) of
the Act, which compels exhaustion of internal remedy
as a
pre-condition to review, save in exceptional circumstances, does not
apply. At common law the duty to exhaust internal remedies
is far
less stringent. As Hoexter (op cit 539) explains, the common law
position is that a court will condone a failure to pursue
an
available internal remedy, for instance where that remedy is regarded
as illusory or inadequate.
[37]
In this case we know that Advocate Breytenbach made a request early
on to the NDPP, which was supported by a 200-page memorandum,
that
the latter should intervene in Mrwebi’s decision to withdraw
the fraud and corruption charges. In addition, the dispute
had been
ongoing for many months before it eventually came to court and,
during that period, it was widely covered by the media.
But despite
this wide publicity, the high profile nature of the case and the
public outcry that followed, the NDPP never availed
herself of the
opportunity to intervene. Against this background FUL could hardly be
blamed for regarding an approach to the NDPP
as meaningless and
illusory in a matter of some urgency.
Challenge
to decision to withdraw the fraud and corruption charges
[38]
FUL’s first challenge of this decision rests on the contention
that Mrwebi had failed to comply with the provisions of
s 24(3)
of the NPA Act in that he did not take the decision to withdraw the
charges ‘in consultation’ with the
DPP ‘of the area
of jurisdiction concerned’ as required by the section. As to
the legal principles involved, it has
by now become well established
that when a statutory provision requires a decision-maker to act ‘in
consultation with’
another functionary, it means that there
must be concurrence between the two. This is to be distinguished from
the requirement
of ‘after consultation with’ which
demands no more than that the decision must be taken after
consultation with and
giving serious consideration to the views of
the other functionary, which may be at variance with those of the
decision-maker.
[39]
An understanding of the factual basis for the challenge calls for
elaboration of the facts given thus far. The DPP of the area
of
jurisdiction concerned, as envisaged by s 24(3), was Advocate
Mzinyathi, the DPP of North Gauteng. Mrwebi’s version
in his
answering affidavit is that he briefly discussed the matter with
Mzinyathi on 5 December 2011, after which he prepared an
internal
memorandum addressed to Mzinyathi, setting out the reasons why, in
his view, the fraud and corruption charges should be
withdrawn.
Although Mzinyathi did not agree with him at that stage, there was a
subsequent meeting between the two of them, together
with Advocate
Breytenbach, on 9 December 2011. At that meeting, so Mrwebi said, the
other two were initially opposed to the withdrawal
of the charges,
but that all three of them eventually agreed that there were serious
defects in the State’s case and that
the charges should be
provisionally withdrawn. However, the problems with this version are
manifold. Amongst others, it is in direct
conflict with the contents
of Mrwebi’s internal memorandum of 5 December 2011 from which
it is patently clear that by that
stage he had already taken the
final decision to withdraw the charges. The last two sentences of the
memorandum bear that out.
They read:
‘
The
prosecutor is accordingly instructed to withdraw the charges against
both Lt-General Mdluli and Colonel Barnard immediately.’
And:
‘
The
lawyers of Lt-General Mdluli will be advised accordingly.’
[40]
An even more serious problem with the version presented in Mrwebi’s
answering affidavit, is that it was in direct conflict
with the
evidence that he and Mzinyathi gave under cross-examination at a
disciplinary hearing of Breytenbach. The transcript of
the hearing
was annexed to the supplementary founding affidavit on behalf of FUL.
The conflict is set out in extensive detail in
the judgment of the
court a quo (paras 47-48). I find a repetition of that recordal
unnecessary. What appears in sum is that Mrwebi
conceded in
cross-examination that he took a final decision to withdraw the
charges before he wrote the memorandum of 5 December
2011; that at
that stage he did not know what Mzinyathi’s views were; and
that he only realised on 8 December 2011 that Mzinyathi
did not share
his views, at which stage he had already informed Mdluli’s
attorneys that the charges would be withdrawn. According
to
Mzinyathi’s evidence at the same hearing, Mrwebi took the
position at their meeting of 9 December 2011 that the charges
had
been finally withdrawn and that he was
functus
officio,
because he had already
informed Mdluli’s attorneys of his decision.
[41]
In these circumstances I agree with the court a quo’s
conclusion (para 55) that Mrwebi’s averment in his answering

affidavit, to the effect that he consulted and reached agreement with
Mzinyathi before he took the impugned decision, is untenable
and
incredible to the extent that it falls to be rejected out of hand.
The only inference is thus that Mrwebi’s decision
was not in
accordance with the dictates of the empowering statute on which it
was based. For that reason alone the decision cannot
stand.
[42]
The court a quo gave various other reasons why Mrwebi’s
impugned decision cannot stand. These are comprehensively set
out in
the judgment of the court a quo under the heading ‘the
withdrawal of the fraud and corruption charges’ (para
141 et
seq). However, in the light of my finding that the decision falls to
be set aside on the basis that it was in conflict with
the empowering
statute, I find it unnecessary to revisit these reasons. Suffice it
to say that, in the main, I find the court’s
reasoning
convincing and nothing that has been said in arguments before us
casts doubt on their correctness.
The
decision to withdraw the murder and related charges
[43]
This brings me to the decision by Chauke to withdraw the murder and
related charges. It will be remembered that on Chauke’s

version, he withdrew the murder charge pending the outcome of the
inquest that he had requested and that he withdrew the 17 other

related charges to avoid a fragmented trial. The contention by FUL
was in essence that this decision was irrational. However, as
I see
it, the contention has not been substantiated in argument. On the
face of it the decision that the findings at an inquest
could perhaps
enable him to take a more informed view of the prospects of the
State’s case with regard to the murder charge,
was not
irrational. It is true that the outcome of the inquest could have no
impact on the 17 related charges. But Chauke never
thought that it
would. As I understand his reasoning, he always intended to reinstate
at least some of the charges after the inquest,
with or without the
murder charge. What he tried to avoid, so he said, was a
fragmentation of trials. That line of reasoning I
do not find
irrational either, particularly since the evidence supporting the
related charges would also impact on the murder charge.
It is true
that he could have asked for a postponement of the 17 related charges
pending the inquest, but we know that a postponement
is not for the
asking. It could be successfully opposed by Mdluli, in which event
the fragmentation, which Chauke sought to avoid
for understandable
reasons, may have become a reality.
[44]
FUL’s real argument, which found favour with the court a quo
(para 183) is that Chauke’s failure to proceed with
the murder
and related charges after the findings of the inquest became
available, was irrational. But that decision – or
really his
failure to apply his mind afresh to the matter after the conclusion
of the inquest – was not the subject of the
review application.
It will be remembered that the review application started in May 2012
while the results of the inquest only
became available in November of
that year. Stated somewhat more concisely: I do not believe the
earlier decision to withdraw the
charges – which is the
impugned decision – can be set aside on the basis that a
subsequent decision, taken in different
circumstances, not to
reinstate all or some of those charges, was not justified. To that
extent the appeal must therefore succeed.
[45]
However, having said that, senior counsel for the NDPP conceded,
rightly and fairly in my view, that there is no answer to
the
proposition that at least some of the murder and related charges are
bound to be reinstated. In the light of this concession
he undertook
on behalf of his client – which undertaking was subsequently
elaborated upon in writing:
(a) That the NDPP
will take a decision as to which of the 18 charges are to be
reinstated and will inform FUL of that decision within
a period of 2
months from this order.
(b) If the NDPP
decides not to institute all 18 charges, he will provide FUL with his
reasons for that decision during the same
period.
I
can see no reason why this undertaking should not be incorporated in
this court’s order and I propose to do so.
Jurisdiction
of the high court to review the decision to terminate disciplinary
proceedings
[46]
This brings me to the decisions by the Commissioner of Police, to
terminate the disciplinary proceedings against Mdluli and
then to
reinstate him to his position on 27 March 2012. Not unlike the NDPP,
the Commissioner’s response to FUL’s challenge
to these
decisions focused mainly on technical objections, rather than to
defend the decisions on their merits. The first technical
objection
was that the high court lacked jurisdiction to review the impugned
decisions by virtue of
s 157
of the
Labour Relations Act 66 of
1995
. The court a quo found this argument fundamentally misconceived
(para 227) and I agree with this finding. The argument rests on
the
premise that this is a labour dispute, which it is not. It is not a
dispute solely between employer and employee. The mere
fact that the
remedy sought may impact on the relationship between Mdluli and his
employer does not make it a labour dispute. It
remains an application
for administrative law review in the public interest, which is
patently subject to the jurisdiction of the
high court.
Mootness
[47]
The Commissioner’s next technical objection was that the
impugned decision had become moot. The factual basis advanced
for the
contention was that, shortly after the application had been launched,
disciplinary charges were again initiated against
Mdluli –
which charges are currently pending – and that he was again
suspended from office, which suspension is still
in force. It is
common cause, however, that the new disciplinary charges do not
pertain to the murder and 17 related charges. Nor
do they correspond
with the fraud and corruption charges that were withdrawn by Mrwebi.
In this light I can find no merit in the
mootness argument. The fact
that disciplinary proceedings had been instituted on charges A and B
obviously does not render moot
the challenge of a decision to
terminate disciplinary proceedings on charges Y and Z.
Review
of a decision to terminate disciplinary proceeding
[48]
The Commissioner’s powers to institute disciplinary charges and
to suspend members of the police derive from regulations
published
under the South African Police Services Act 68 of 1995. These powers
can be traced back to s 207(2) of the Constitution
which
requires the Commissioner to manage and exercise control over the
SAPS. These powers are clearly public powers. That is why
they were
promulgated by law and not merely encapsulated in a contract between
the parties. The Commissioner took the decision
to institute
disciplinary proceedings against Mdluli and to suspend him pursuant
to these powers. When he decided to reverse those
decisions, he did
so in the exercise of the same public powers. It follows that the
latter decisions constituted administrative
action, reviewable under
the provisions of PAJA.
[49]
As the factual basis for the challenge of these decisions, FUL relied
in its founding affidavit on a statement by the then
Acting
Commissioner, Lieutenant-General Mkhwanazi, in Parliament that he was
instructed by authorities ‘beyond’ him
to withdraw
disciplinary charges and reinstate Mdluli in his office. FUL added
that in doing so Mkhwanazi had failed to make an
independent decision
which rendered his actions reviewable. Though Mkhwanazi filed an
answering affidavit in the interim interdict
proceedings in part A of
the notice of motion, he did not deal with these allegations. In the
answering affidavit filed in part
B, the present Commissioner,
General Phiyega, said the following in response to these allegations
by FUL.
‘
General
Mkhwanazi was quoted out of context. As I understood and this is what
he later clarified was that his response was in relation
to the issue
of the withdrawal of charges, which falls within the domain of the
NPA, which invariably in his view affected the
purpose of the
continued suspension and disciplinary charges then. General Mkhwanazi
never received any instructions from above.
His confirmatory
affidavit will be obtained in this regard. Should time permit, I will
ensure that the copy of the Hansard being
the minutes or the
transcription of the parliamentary portfolio committee meetings is
obtained and filed as a copy which will clarify
the issue.’
[50]
But despite these undertakings, no confirmatory affidavit was filed
by Mkhwanazi nor was a copy of Hansard provided. In argument
before
the court a quo, the Commissioner’s representatives again
undertook to file an affidavit by Mkhwanazi, but this undertaking
was
later withdrawn (para 213 of the judgment a quo). In the premises the
court a quo held (para 214) that the Commissioner’s
explanation
was untenable and stood to be rejected. I do not believe this finding
can be faulted. Moreover, after all is said and
done, neither
Mkhwanazi nor Phiyega gave any reasons for the impugned decision. The
inevitable conclusion is thus that the decisions
were either dictated
to Mkhwanazi or were taken for no reason at all. In either event they
fall to be set aside under s 6
of PAJA. This means that the
appeal against the court a quo’s order to that effect cannot be
sustained.
Appropriate
remedy
[51]
What remains are issues concerning the appropriate remedy. As we
know, the court a quo did not limit itself to the setting
aside of
the impugned decisions. In addition, it (a) ordered the NDPP to
reinstate all the charges against Mdluli and to ensure
that the
prosecution of these charges are enrolled and pursued without delay;
and (b) directed the Commissioner of Police to reinstate
the
disciplinary proceedings and to take all steps necessary for the
prosecution and finalisation of these proceedings (para 241(
e
)
and (
f
)).
Both the NDPP and the Commissioner contended that these mandatory
interdicts were inappropriate transgressions of the separation
of
powers doctrine. I agree with these contentions. That doctrine
precludes the courts from impermissibly assuming the functions
that
fall within the domain of the executive. In terms of the Constitution
the NDPP is the authority mandated to prosecute crime,
while the
Commissioner of Police is the authority mandated to manage and
control the SAPS. As I see it, the court will only be
allowed to
interfere with this constitutional scheme on rare occasions and for
compelling reasons. Suffice it to say that in my
view this is not one
of those rare occasions and I can find no compelling reason why the
executive authorities should not be given
the opportunity to perform
their constitutional mandates in a proper way. The setting aside of
the withdrawal of the criminal charges
and the disciplinary
proceedings have the effect that the charges and the proceedings are
automatically reinstated and it is for
the executive authorities to
deal with them. The court below went too far.
Costs
[52]
As to the court a quo’s costs order against the appellants in
favour of FUL, I can see no reason to interfere. Although
I propose
to set aside some of the orders granted by the court a quo, it does
not detract from FUL’s substantial success
in that court. On
appeal the position is different. Here it is the appellants who
achieved substantial success. Ordinarily this
would render FUL liable
for the appellants’ costs on appeal. But it has by now become
an established principle that in constitutional
litigation
unsuccessful litigants against the Government are generally not
mulcted in costs, lest they are dissuaded from enforcing
their
constitutional rights. (See eg
Biowatch
Trust v Registrar, Genetic Resources & others
2009
(6) SA 232
(CC).) Although the rule is not immutable, I find no
reason to deviate from the general approach in this case. Hence I
shall make
no order as to the costs of appeal.
[53]
The order I propose should therefore reflect the intent:
(a) To confirm the
setting aside of Mrwebi’s decision to withdraw the fraud and
corruption charges in para (
a
) as well as the setting aside of
the Commissioner’s decision to terminate the disciplinary
proceedings against Mdluli in
para (
c
) as well as the setting
aside of Mdluli’s reinstatement by the Commissioner on 28 March
2012 in para (
d
) of the order of the court a quo.
(b) To reverse the
setting aside of Chauke’s decision to withdraw the murder and
related charges in para (
b
) of that order.
(c) To set aside the
mandatory interdicts in paras (
e
) and (
f
) of the order;
(d) To confirm the
costs order in paras (
g
) and (
h
) of the order; and
(e)
To give effect to the undertaking on behalf of the NDPP with regard
to the reinstitution of the murder and related charges.
[54]
In the premises it is ordered that:
1 The appeal
succeeds only to the extent that paragraphs (b), (e) and (f) of the
order of the court a quo are set aside
2 The orders in
paragraphs (
a
), (
c
), (
d
), (
g
) and (
h
)
of the order by the court a quo are confirmed but re-numbered in
accordance with the changes necessitated by the setting aside
of the
orders in paragraph 1.
3 It is recorded
that the following undertaking has been furnished on behalf of the
first respondent:
(a)
To decide which of the criminal charges of murder and related crimes
that were withdrawn on 2 February 2012, are to be reinstituted
and to
make his decision known to the respondent within 2 months of this
order.
(b)
To provide reasons to the respondent within the same period as to why
he decided not to reinstitute some – if any –
of those
charges.
4.
There shall be no order as to costs in respect of the appeal.
F
D J BRAND
JUDGE
OF APPEAL
.APPEARANCES:
For
the 1
st
and 2
nd
Appellants: L Hodes SC (with him N Manaka, E Fasser)
Instructed by:
State Attorney,
Pretoria
c/o
State Attorney, Bloemfontein
For
the 3
rd
Appellants: W R Mokhali SC (with him M Zulu)
Instructed by:
State Attorney,
Pretoria
c/o
State Attorney, Bloemfontein
For
the 4
th
Appellant: I Motloung
Instructed by:
Maluleke Seriti
Makume Matlala Inc, Germiston
c/o
Peyper Attorneys, Bloemfontein
For
the Respondent : S Yacoob (with him I Goodman, N van der Walt)
Instructed
by:
Cliffe
Dekker Hofmeyr, Johannesburg
Matsepes,
Bloemfontein