---

CONSTITUTIONAL COURT OF SOUTH AFRICA

                                                                                                                        Case CCT 106/15

In the matter between:

MERAFONG CITY LOCAL MUNICIPALITY                                                                     Applicant

and

ANGLOGOLD ASHANTI LIMITED                                                                                   Respondent

Neutral citation:     Merafong City Local Municipality v AngloGold Ashanti Limited [2016] ZACC 35

Coram:                    Moseneke DCJ, Bosielo AJ, Cameron J, Froneman J, Jafta J, Khampepe J, Madlanga J, Mhlantla J, Nkabinde J and Zondo J

Judgments:              Cameron J (majority): [1] to [84]

                                  Jafta J (dissenting): [85] to [178]

Heard on:                 18 February 2016

Decided on:              24 October 2016

Summary:                collateral or reactive challenge by organ of state — inter governmental dispute — national and local functional areas of competence — exclusive municipal power to levy surcharge on water supply binding nature of administrative action — permissibility of collateral or reactive challenge by organ of state — justification of delay — distinction between declaration of constitutional invalidity and just and equitable remedy

ORDER

On appeal from the Supreme Court of Appeal (hearing an appeal from the High Court of South Africa, Gauteng Division, Pretoria):

1.                 Leave to appeal is granted.

2.                 The appeal is upheld.

3.                 The orders of the High Court of South Africa, Gauteng Division, Pretoria and the Supreme Court of Appeal are set aside.

4.                 In their stead, the matter is remitted to the High Court to determine, after the lodging of further affidavits as the applicant, Merafong City Local Municipality, and the Minister of Water Affairs and Forestry may consider appropriate, on the lawfulness of the Minister’s decision of 18 July 2005, and, if necessary, what remedy is to be granted.

5.                 The Minister is to lodge the record of the decision by 4 November 2016.

6.                 The further affidavits, if any, by the applicant are to be lodged by 18 November 2016 and by the Minister by 25 November 2016.

7.                 The respondent, AngloGold Ashanti Limited, may lodge its affidavits, if any, by 6 December 2016.

8.                 Costs are reserved for consideration by the High Court.

JUDGMENT

CAMERON J (Moseneke DCJ, Froneman J, Khampepe J, Madlanga J, Mhlantla J and Nkabinde J concurring):

Introduction

[1] At issue is a decision the Minister of Water Affairs and Forestry (Minister) took on 18 July 2005 in which she overturned a decision by the applicant, Merafong City Local Municipality (Merafong), to levy a surcharge on water for industrial use by the respondent, AngloGold Ashanti Limited (AngloGold).  In addition, on the portion of water used for domestic purposes, the Minister ruled that Merafong, the other mining houses affected and Rand Water, which is a statutorily established organ of state,[1] “should negotiate a reasonable tariff”.  Merafong seeks to overturn a decision of the Supreme Court of Appeal, dismissing an appeal to it from a decision of the Gauteng Division of the High Court, Pretoria,[2] which gave effect to the Minister’s decision.

[2] Merafong contends that the Minister’s decision was invalid because it intruded on an exclusive constitutional competence which section 156(1) of the Constitution confers on Merafong.[3]  This empowers Merafong to deal with “[w]ater and sanitation services limited to potable water supply systems and domestic waste-water and sewage disposal systems”.[4]  The Minister purported to act in pursuance of section 8(4) of the Act, which permits a party who has applied to a local authority for certain services to appeal to the Minister “against any decision, including any condition imposed”.  The Act empowers the Minister on appeal to “confirm, vary or overturn any decision” of a water services authority.[5]  Merafong was the water services authority in question.  It says there is a clash between its executive constitutional competences and the power the Minister purported to exercise.

Background and litigation history

[3] AngloGold owns mines in Merafong.  Its mining operations require water.  It also supplies water for domestic use to its employees staying on the mines.  From 1958 until 2004, that water came to AngloGold directly from Rand Water, with whom it had a number of written water supply agreements.  The water reached AngloGold through a system of reservoirs, pipelines and other apparatus maintained by Rand Water.

[4] In December 1997, when the Act came into effect, all this changed.  The Act gave statutory recognition to local government’s new constitutional authority to administer water and sanitation services.  It designated municipalities as water services authorities responsible for progressively ensuring access to water services by consumers in their areas of jurisdiction.  Municipalities flexed these new statutory muscles only from 1 July 2003.

[5] On 11 February 2004, Merafong sent a written notice to all the mining houses operating in its area, including AngloGold.  The notice said Merafong had been accorded the powers and functions of a water services authority.  It asked the mining houses to apply for approval for the supply of water for industrial use in terms of section 7 of the Act.  After the notice, at meetings with the mining houses, Merafong explained the implications of the Act and its role as a water services authority.  On 31 May 2004, Merafong sent another written notice informing the mining houses that new tariffs would come into operation from 1 July 2004.

[6] In response, on 8 April 2004, AngloGold sought Merafong’s approval “in terms of section 7 of the [Act] to continue obtaining water from Rand Water for its mining operations and associated domestic applications at the tariff set by, and under the conditions imposed by Rand Water”.[6]

[7] Merafong announced the new tariffs in May 2004.  They were a lot higher than Rand Water’s tariffs.  But Merafong couldn’t provide water services itself.  So it appointed Rand Water as its agent to do so and to collect payment.  After deducting its share, Rand Water would pay the balance – the surcharge Merafong had added – into the coffers of Merafong.  Merafong said it imposed the surcharge because it had to “find new sources of income to ensure [its] financial sustainability”.  Its executive Mayor justified the surcharge as necessary to fulfil “the promise of better service delivery to the total community,” and noted the impact of the surcharge on “the economy of the region”.

[8] Merafong told the mining houses they had a right to appeal to the Minister against the new charges.  AngloGold appealed to the Minister.[7]  It complained that the tariff Merafong proposed was much higher than the Rand Water tariff (an extra half a million rand per month) without “adding any value to, or assuming any responsibility for any aspect of the water supply”.  It also complained that Merafong failed to recognise AngloGold’s role as a water service provider or make any attempt, other than to ask for information on the mine’s consumption, to understand its economic situation.

[9] The Minister, purporting to exercise her powers under the Act, upheld AngloGold’s appeal.  She found the tariff increase unreasonable because Merafong would add no value to the services Rand Water provided to AngloGold.  On 18 July 2005, she made a ruling.[8]  She overturned the Municipality’s decision to levy the surcharge on water for industrial purposes.  She ruled that a surcharge could be levied only on the portion of water AngloGold was using for domestic purposes.  The Minister then directed Merafong, AngloGold and Rand Water to negotiate a reasonable tariff on this portion.

[10] In September 2005, Merafong obtained legal advice that the Minister’s decision was “void in law”.  As a result, in September 2007 it threatened to discontinue AngloGold’s supply of water unless it paid the disputed surcharge, plus arrears.  The cut-off would have been catastrophic.  So AngloGold paid, under protest, and has continued doing so.

[11] During 2006 and 2007 Merafong and the mining houses tried to work out an arrangement but their negotiations ran aground.  A draft interim agreement was drawn up and discussed.  It provided that the mines would be charged Merafong’s tariff (with surcharge) for domestic water use and Rand Water’s industrial tariff (without surcharge) for the mine hostels and operational water use.  But the agreement was never bolted down.  The Municipality remained adamant that the Minister’s decision was unlawful.  It refused to buckle down under it.

[12] That is where things stand.  AngloGold claims it has been unlawfully overcharged nearly R31 million for water,[9] and forced to pay up under an enterprise throttling threat of cut-off.  The Municipality continues to levy a surcharge on AngloGold’s industrial water use.

[13] Against this background, AngloGold, in April 2011, launched proceedings in the High Court to compel Merafong to comply with the Minister’s ruling.  In response, Merafong conditionally counter-applied for a declarator that it has exclusive executive authority to set, adopt and implement tariffs on the provision of water services in its jurisdiction, including surcharges.  It also sought a declarator that section 8 of the Act did not confer authority on the Minister “to interfere with a tariff set and implemented” by it for water services provided in its area of jurisdiction.  In the alternative, it sought to strike down the provision allowing appeals to the Minister, section 8(9), as unconstitutional and invalid.

[14] The High Court granted AngloGold’s application and dismissed Merafong’s counter application.  It found that AngloGold had validly applied to Merafong under the statute,[10] and that the Minister was vested with appellate power,[11] which she exercised.  But the High Court found that the Minister’s decision, even if impugnable, was in any event binding on the Municipality until set aside.  For this it relied on Oudekraal.[12]  The High Court accordingly ordered that Merafong comply with the Minister’s ruling.

[15] On appeal, with leave of the High Court, the Supreme Court of Appeal endorsed this outcome.[13]  It held that Merafong was obliged to approach the court to set aside the Minister’s ruling, and that it breached the principle of legality by simply disregarding it.[14]  Its failure to challenge the Minister’s ruling in judicial review proceedings, rather than the constitutional attack it launched against the empowering statutory provisions, posed an insuperable difficulty for it.[15]  Citing Kirland,[16] the Court held that the notion that an organ of state could remain inert or passive in the face of a ruling adverse to its powers, only to use the alleged invalidity of that ruling as a shield when it was sought to be enforced, was unacceptable.  Further that a collateral challenge to the validity of an administrative action is a remedy available only to an individual threatened by a public authority with coercive action: “[t]he notion that an organ of state can use this shield against another organ of state is simply untenable”.[17]

In this Court

[16] In this Court, as in the Supreme Court of Appeal, Merafong did not persist with its argument that section 8(9) of the Act is invalid, and the powers the Act conferred on the Minister are unconstitutional, though counsel said the argument “was alive on the papers”.[18]  Merafong also conceded that the supply of water for industrial purposes could competently be subjected to ministerial scrutiny.

[17] Merafong’s principal contention is that the Supreme Court of Appeal misapplied Oudekraal and Kirland.  This is because of an alleged fundamental distinction between decisions that fall within the scope of powers with which a public official is clothed, but are merely wrongly taken, and those that are “on their face, beyond the powers of the decision-maker”.  In the latter case, so the contention goes, the person or entity subject to the decision is entitled to ignore it until, as a matter of process, that decision is sought to be enforced against it.  Then, the person or entity is entitled to raise the nullity of the decision as a defence.  That, Merafong says, is what it did here.  In oral argument, Merafong also contended that the particular circumstances surrounding its defence permitted it to be raised in these proceedings.

[18] Rand Water initially filed a notice of opposition, which it later withdrew, indicating that it would abide in the litigation’s outcome.  Though the Minister was cited in the High Court, neither of the parties sought any direct relief against her.  She filed no papers and did not appear in either the High Court or the Supreme Court of Appeal.

[19] After oral argument, this Court on 1 April 2016 issued directions inviting the Minister to indicate whether it was proper for the Court to determine whether: (i) the Municipality has the exclusive competence to implement tariffs and levy surcharges; (ii) section 8(9) of the Act confers authority upon the Minister to interfere with tariffs; alternatively, (iii) if she has authority under section 8(9) to interfere, whether section 8(9) is unconstitutional and invalid.  The Minister was also directed to file written submissions setting out the grounds of her opposition if she opposed the determination of the constitutional issues.  The parties were given an opportunity to respond.[19]

[20] The Minister in her submissions contended that, as a matter of procedure, the Court has the power to decide Merafong’s conditional counter-application, and that, although an organ of state may not avail itself of a collateral challenge, this was no bar to the Court deciding the direct constitutional challenge.  She submitted that the Court should nevertheless not decide the counter-application because, on the facts of her ruling, it was unnecessary for the issue to be reached.

Jurisdiction and leave to appeal

[21] The dispute about the Minister’s power to intervene on a matter of municipal competence raises constitutional issues that fall within this Court’s jurisdiction.  It also raises arguable points of law of general public importance that the Court ought to consider.  Merafong’s arguments have substance, and leave to appeal should be granted.

Issues

[22] The principal question is whether the Supreme Court of Appeal and High Court were right to enforce the Minister’s ruling.  Behind that question is the broader issue of when a public authority may collaterally or reactively challenge an administrative act, like the Minister’s ruling, that is sought to be enforced against it, outside proceedings brought to review it.

Merits

            Collateral or reactive challenges

[23] As noted above, the Supreme Court of Appeal held against Merafong on the basis that, so long as an administrative decision has not been set aside, an organ of state may not raise its invalidity as a defence to proceedings against it to enforce that decision.[20]  Relying on the invalidity of an administrative act as a defence against its enforcement, while it has not been set aside, has been dubbed a collateral challenge[21] – “collateral” because it is raised in proceedings that are not in themselves designed to impeach the validity of the act in question.[22]  While the object of the proceedings is directed elsewhere, invalidity is raised as a defence to them.

[24] The Supreme Court of Appeal, citing its own jurisprudence[23] and Kirland,[24] held that Merafong could not invoke a challenge to the validity of the Minister’s ruling in reaction to AngloGold’s application at all.  This was because that remedy is available only to a person whom a public authority threatens with coercive action.  In those cases, the Court held, citing Oudekraal,[25] the rationale was that “the legal force of the coercive action will most often depend upon the legal validity of the administrative action in question”.[26]  That was why the disputed validity of the decision could be questioned in those proceedings; but not beyond that.

[25] This knockout blow to Merafong derived from a category-approach to who can raise a collateral challenge.  Only an individual whom a public authority threatens with coercive action can; and no one outside the category.  Never a public authority.  This approach squeezes collateral challenge into a rigid format – one that neither doctrine nor practical reason appears to warrant.

Collateral challenges pre-Constitution

[26] The first reported challenge to administrative action designed to ward off criminal proceedings was not strictly collateral, but reactive, for it was not raised in the criminal proceedings themselves.[27]  In Johnstone, the Appellate Division of the Supreme Court (the predecessor of the Supreme Court of Appeal) held that a corporation against which a charge of transgressing a wage determination had been laid could obtain a declaratory order, outside the criminal court, declaring that the determination was not applicable to it.  Schreiner JA warned that where a prosecution had already been instituted, a court asked to exercise its discretion by in effect declaring that the person is innocent “would do well to exercise great caution before granting such an order”.[28]  But where there was no dispute about the facts, an application for a declaration could be entertained, in a proper case, even though, if the applicant’s argument was wrong, he may already have contravened the law.[29]

[27] In Johnstone there was an impending prosecution.  But reactive challenges to administrative decisions can be raised in any coercive setting, not only criminal.  In Panasonic,[30] a trade union sought to attack collaterally the validity of industrial council proceedings on whose validity the employer’s lock-out depended.  The complaint, the Court held, had to be raised by way of review, and not collaterally in other proceedings.[31]  Conradie J located the reason in the nature of the employer/employee contest, in which the Court acts as a referee.  The referee may intervene if a blow is struck below the belt.  But the referee “would be astounded while the bout is in progress to receive a complaint that something had gone wrong with the weigh-in”.[32]  In other words, in contests of force between organised labour and employers, the fight had to proceed regardless of possible defects in the preconditions giving rise to it.

[28] But the Panasonic Court found that it was “impossible to lay down any fixed rule” on when a reactive challenge should be countenanced:

“Each case will depend on its own circumstances, in particular the nature of the alleged irregularity, the reason that it had not been raised earlier, the stage which the economic contest had reached and whatever other factors may be relevant.”[33]

[29] By contrast, in Photocircuit,[34] also with no prosecution in issue, a reactive attack on the validity of the establishment of an industrial council was allowed.  The council sought an order to force employers to render returns to enable it to exact contributions from them.  The employers responded by challenging the validity of the establishment of the council and of the extension to them of agreements requiring the contributions.  Scott J pointed out that in collateral challenges judicial scrutiny of an administrative act or subordinate legislation arises not because a discretionary remedy is sought, namely review or a declaratory order, but to determine “the entitlement of the party seeking enforcement or the guilt or innocence of an accused person”.[35]  He concluded that the defendant or accused in collateral challenge proceedings cannot be precluded from raising the invalidity defence merely on the grounds of delay.  But, as in Panasonic, the Court emphasised that whether a collateral attack will be permitted depends on the circumstances.[36]

[30] These cases predate the Constitution.  But they show that, in South African law, the permissibility of a reactive attack on administrative action has always been approached with a measure of flexibility.[37]  And its availability is not limited to those at risk of criminal conviction.[38]  A subject at risk of criminal conviction or other coercive action by the state may indeed raise a reactive or defensive challenge to the lawfulness of the administrative act on which the prosecution or coercion is based.  But reactive challenges in our law have never been limited to these circumstances.  In Panasonic, the challengers were employees facing a lockout.  They were denied a collateral challenge not because of the absence of state coercion, but because the challenge was inappropriate to the proceedings they brought. Given the power play taking place, it would have been unjust to allow the challenge to be brought.

The Constitution and collateral challenges

[31] The pre-Constitution approach was rooted in features of the common law remedy of review, which was discretionary, and which took account of delay, process and other considerations in determining whether relief should be granted.[39]  By contrast a party relying on a collateral challenge was insulated from the vicissitudes of the discretionary award of the review remedy, or its unavailability on grounds of delay.[40]  The defence should be capable of being raised simply because a forced levy was disputed.  Indeed, the very point of refusing to permit a collateral attack, on occasion, was not only that the source of the impugned act would be brought before a review court, but that unreasonable delays could be punished.[41]  The consequence was, and is, that delay may insulate irregular administrative actions from review.[42]  By contrast, to coerce a citizen into payment or prison on the basis of an unlawful administrative action, however ancient, would itself offend the rule of law.  Hence the continued permissibility of reactive or collateral challenges.

[32] So the remedies of review and collateral challenge differ distinctively in object, application and scope.  The Supreme Court of Appeal explained in Oudekraal that where the validity of an administrative act is challenged collaterally a court has no discretion to allow or disallow the raising of the defence: “the right to challenge the validity of an administrative act collaterally arises because the validity of the administrative act constitutes the essential prerequisite for the legal force of the action that follows”.[43]  It follows that the subject may not be precluded from challenging its validity.[44]  On the other hand, a court asked to set aside an invalid administrative act in proceedings for judicial review has a discretion whether to grant or withhold the remedy:

“It is that discretion that accords to judicial review its essential and pivotal role in administrative law, for it constitutes the indispensable moderating tool for avoiding or minimising injustice when legality and certainty collide.  Each remedy thus has its separate application to its appropriate circumstances and they ought not to be seen as interchangeable manifestations of a single remedy that arises whenever an administrative act is invalid.”[45]

[33] It is here, where “legality and certainty collide”, that AngloGold’s dispute with Merafong asks for answers that have not previously been provided.  Those must be grounded in the Constitution.  The Constitution is the supreme law, and conduct inconsistent with it is invalid.[46]  The Constitution provides that, when deciding a constitutional matter within its power, a court “must declare that any law or conduct that is inconsistent with [it] is invalid to the extent of its inconsistency”.[47]  To this injunction the Constitution adds a discretionary “may”: a court deciding a constitutional matter “may make any order that is just and equitable”.[48]

[34] In Bengwenyama,[49] this Court explored the Oudekraal paradox, that an unlawful act can produce legally effective consequences.  The apparent anomaly, Froneman J noted, “is not one that admits easy and consistently logical solutions”:

“But then the law often is a pragmatic blend of logic and experience.  The apparent rigour of declaring conduct in conflict with the Constitution and PAJA unlawful is ameliorated in both the Constitution and PAJA by providing for a just and equitable remedy in its wake.  I do not think that it is wise to attempt to lay down inflexible rules in determining a just and equitable remedy following upon a declaration of unlawful administrative action.  The rule of law must never be relinquished, but the circumstances of each case must be examined in order to determine whether factual certainty requires some amelioration of legality and, if so, to what extent.  The approach taken will depend on the kind of challenge presented – direct or collateral; the interests involved and the extent or materiality of the breach of the constitutional right to just administrative action in each particular case.”[50]

[35] In AllPay[51] the same judge, speaking on behalf of the Court, took this further.  He noted that there was a “clear distinction” between “the constitutional invalidity of administrative action”, on the one hand, and, on the other, “the just and equitable remedy that may follow from it”.[52]  It was for this reason that the Court declared invalid a tender whose award was riddled with suggestive irregularities, while nevertheless suspending the declaration of invalidity pending determination of a just and equitable remedy.[53]  Upsetting the award might have had disastrous consequences for millions of vulnerable grant recipients.  Hence it was just and equitable to keep the unlawful award temporarily in place by the exercise of the broad remedial powers the Constitution has vested in this Court.

[36] Hence the central conundrum of Oudekraal, that “an unlawful act can produce legally effective consequences”,[54] is constitutionally sustainable, and indeed necessary.  This is because, unless challenged by the right challenger in the right proceedings,[55] an unlawful act is not void or non-existent, but exists as a fact and may provide the basis for lawful acts pursuant to it.[56]  This leads to a logical corollary, which this Court recognised in Giant Concerts,[57] that an own-interest litigant may be denied standing “even though the result could be that an unlawful decision stands”.[58]

[37] These consequences follow from the wording of section 172(1) itself, which requires a court to declare any law or conduct inconsistent with the Constitution invalid to the extent of its inconsistency, but requires the court to do so only “when deciding” a constitutional matter within its jurisdiction.  The provision does not dictate to courts when or how they must decide.  It contemplates that a court may decline to decide a matter because the right complainant is not before it, or because the challenge is not warranted in the particular proceedings before it.

[38] All this illuminates whether Merafong should be permitted, in these enforcement proceedings, to raise the alleged invalidity of the Minister’s ruling.

The import of Oudekraal and Kirland

[39] In Oudekraal, a provincial executive officer, the Administrator, had granted Oudekraal township development rights.  That was in 1957, more than 40 years before the dispute.  Now Oudekraal asked the City Council to approve its engineering services plan for the township.  The City Council refused.  It said the Administrator’s approval, decades back, was bad for lack of compliance with the provincial law under which it was granted.  Was the City Council entitled to refuse approval because the underlying grant of development rights was bad?  The High Court said Yes.  The Supreme Court of Appeal said No.  The City Council could not simply treat the Administrator’s act as though it did not exist.  Until it was properly set aside by a court of law, the approval engendered legal consequences.  The Court however refused Oudekraal the declaratory relief it sought.  This was because the approval was vulnerable to being set aside in proceedings properly brought for judicial review, and that had to be done first.[59]

[40] In Kirland, an official had refused an authorisation Kirland sought, but, before that decision was communicated, an acting stand-in, in dubious circumstances, granted the permission.  The question was whether the Department of Health, Eastern Cape (Department) knowing the grant was dubious, could treat it as non-existent.  This Court, applying Oudekraal, said No.  Doing so affected Kirland’s rights.  Even though the second decision (which was communicated first) might be defective, government should generally not be exempt from the forms and processes of review.  It could not take a shortcut across Kirland’s legal and constitutional protections.[60]  Kirland may have acted to its detriment in reliance on the second decision.  It would be unfair to Kirland to allow government to ignore the decision.  In addition, government in bringing proceedings to set aside the suspect decision would have to explain its dilly-dallying, which suggested acquiescence.[61]  It therefore had to apply formally for a court to set aside the defective decision, so that a court could properly consider its effects on those subject to it, as well as the Department’s delay in making the challenge.[62]

[41] The import of Oudekraal and Kirland was that government cannot simply ignore an apparently binding ruling or decision on the basis that it is invalid.  The validity of the decision has to be tested in appropriate proceedings.  And the sole power to pronounce that the decision is defective, and therefore invalid, lies with the courts.  Government itself has no authority to invalidate or ignore the decision.  It remains legally effective until properly set aside.[63]

[42] The underlying principles are that the courts’ role in determining legality is pre-eminent and exclusive; government officials, or anyone else for that matter, may not usurp that role by themselves pronouncing on whether decisions are unlawful, and then ignoring them; and, unless set aside, a decision erroneously taken may well continue to have lawful consequences.  Mogoeng CJ explained this forcefully, referring to Kirland, in Economic Freedom Fighters.[64]  He pointed out that our constitutional order hinges on the rule of law:

“No decision grounded [in] the Constitution or law may be disregarded without recourse to a court of law.  To do otherwise would ‘amount to a licence to self-help’.  Whether the Public Protector’s decisions amount to administrative action or not, the disregard for remedial action by those adversely affected by it, amounts to taking the law into their own hands and is illegal.  No binding and constitutionally or statutorily sourced decision may be disregarded willy-nilly.  It has legal consequences and must be complied with or acted upon.  To achieve the opposite outcome lawfully, an order of court would have to be obtained.”[65]

[43] But it is important to note what Kirland did not do.  It did not fossilise possibly unlawful – and constitutionally invalid – administrative action as indefinitely effective.  It expressly recognised that the Oudekraal principle puts a provisional brake on determining invalidity.  The brake is imposed for rule of law reasons and for good administration.  It does not bring the process to an irreversible halt.  What it requires is that the allegedly unlawful action be challenged by the right actor in the right proceedings.  Until that happens, for rule of law reasons, the decision stands.

[44] Oudekraal and Kirland did not impose an absolute obligation on private citizens to take the initiative to strike down invalid administrative decisions affecting them.  Both decisions recognised that there may be occasions where an administrative decision or ruling should be treated as invalid even though no action has been taken to strike it down.[66]  Neither decision expressly circumscribed the circumstances in which an administrative decision could be attacked reactively as invalid.[67]  As important, they did not imply or entail that, unless they bring court proceedings to challenge an administrative decision, public authorities are obliged to accept it as valid.[68]  And neither imposed an absolute duty of proactivity on public authorities.  It all depends on the circumstances.

Merafong’s reactive challenge

[45] Against this background, the question is whether, when AngloGold sought an order enforcing the Minister’s decision, Merafong was entitled to react by raising the invalidity of her ruling as a defence.  AngloGold lodged its application on 19 April 2011.  Merafong lodged its counter-application some months after, on 3 August 2011.  At that time, the Minister’s ruling, which was given on 18 July 2005, had stood for more than six years.

[46] This lapse of time deserves context.  In September 2005, some two months after the Minister’s ruling, Merafong obtained a detailed opinion from its attorneys.  This concluded that the Minister had no power to prescribe a specific tariff or to interfere in the tariff-setting functions of a municipality.  Hence any act on her part to interfere with those functions was “void in law”.  Furthermore, both Merafong and the Minister had “misdirected themselves by seeking to subject the tariff-setting powers of Merafong to an appeal process”.  The attorneys recommended that Merafong should engage the Minister further and “point out that both Merafong and the Minister have misconstrued their positions in law” and that the Minister “must withdraw the decision to set aside the decision of Merafong to levy tariffs on the mines”.

[47] Soon after receiving this opinion, Merafong sent it to the Minister on 31 October 2005.  And it drew her attention to the opinion again, on 30 March 2006 and on 24 October 2007.  Meanwhile, it tried repeatedly to secure meetings with the Minister.  As the second part of the Minister’s ruling required, Merafong negotiated with the mining houses in its area, including AngloGold.  From the time of the ruling, until October 2007, it held numerous meetings with the mining houses and Rand Water.  Draft agreements on tariffs were negotiated.  But they were never signed.  The parties remained at odds.  In its letter to the Minister of 24 October 2007, Merafong complained that the mines “have not responded favourably to our proposals”.  In its answering affidavit, it described the mines’ attitude during these negotiations as “increasingly obstructive and uncooperative”.

[48] As a last resort, on 30 March 2006, Merafong sought to declare a formal dispute with the Minister.  The Minister’s response was that the parties should try to resolve the matter through dialogue before declaring a dispute.  She proposed a meeting.  But this never materialised.  Merafong says this was “through no lack of trying” on its part.

[49] But after that, activity ceased.  Merafong says that after failing to reach agreement with the mining houses, and with the Minister “also not complying with the principles of cooperative government”, it “imposed and implemented the tariffs”.  It did so by holding its hand on the tap, leaving it to AngloGold to institute proceedings to enforce the Minister’s ruling.  In its affidavit opposing AngloGold, Merafong complained about AngloGold’s “inordinate” delay in bringing the proceedings.  They were brought, Merafong claimed, “out of time”.  The application, it said, should be dismissed on this ground alone.  As will more fully emerge, coming from Merafong, which itself delayed for nearly five years after the impasse arose, and responded only when AngloGold went to court, that was rich.

[50] Merafong argued it should be permitted to raise a reactive challenge to AngloGold’s attempt to enforce the Minister’s ruling because there is a fundamental distinction between decisions that fall within the scope of powers with which a public official is clothed, but are merely wrongly taken, and those that are palpably and obviously beyond the powers of the decision maker.  In the latter case, where a decision “lacks the facial imprimatur of lawfulness”, a person subject to the decision is entitled to ignore it until, as a matter of process, that decision is sought to be enforced against it.  At that point the nullity of the decision may be raised as a defence.  Counsel contended that decisions of this nature “on their face fall beyond the ostensible scope of the powers conferred upon a public officer [and] have no validity and should be treated as such even though they have yet to be set aside on review”.[69]

[51] But this approach is not sound.  Neither the Constitution, nor the Promotion of Administrative Justice Act[70] (PAJA), offer warrant for it.  And logic offers no support either.  Merafong accepts that the Minister’s ruling constituted administrative action.  PAJA provides that administrative action may be reviewed if the administrator who took it “was not authorised to do so by the empowering provision”.[71]  In addition, administrative action may be set aside if it “contravenes a law or is not authorised by the empowering provision”.[72]  The grounds on which administrative action may be set aside are consonant with the statute’s definitions.  “Decision” means any decision “under an empowering provision”.[73]  “Empowering provision”, in its turn, means a law or other source of authority “in terms of which an administrative action was purportedly taken”.[74]

[52] From this it can be seen that even decisions “purportedly taken” under a statute, but which in fact lack authorisation, are subject to review under PAJA.  The plain premise of PAJA is that remedies are available to all who are affected by unlawful administrative action, whether the unlawfulness resides in a process-defect or in the absence of authority.

[53] In addition, Merafong’s argument suggests that decisions that are invalid on their face are limited to those that are taken entirely without statutory or other authority.  That is wrong.  A decision taken within statutory authority may be equally plainly vitiated – for instance by palpable fraud, or error of law, or mistake of fact.

[54] If we were to sustain Merafong’s argument that it was entitled to ignore the Minister’s decision until it was sought to be enforced, this must extend to all cases of patent invalidity.  This would suggest that an official may ignore a decision, taken under statutory power (intra vires), that is tainted by patently improper influence or corruption.  But that is precisely what happened in Kirland – and the self-help argument was not countenanced.  What is more, not only would what is or is not “patently unlawful” be decided outside the courts, but there would be no rules on who gets to decide and how.  If failure to review a disputed decision is defensible on the basis that the decision was considered patently unlawful, the rule of law immediately suffers.  So the argument is not tenable.

[55] Nevertheless, our case law offers little support for a rigid doctrinal limitation upon the viability of a reactive challenge.  While reactive challenges, in the first instance, and perhaps in origin, protect private citizens from state power, good practical sense and the call of justice indicate that they can usefully be employed in a much wider range of circumstances.  There is no practical, or conceptual, justification for strait-jacketing them to private citizens.  It is readily conceivable, for instance, that an organ of state may through legal proceedings seek unjustly to subject another organ of state to a form of coercion.  Where appropriate, that other should be able to raise a defensive or reactive challenge.  Categorical exclusions should be eschewed.  A reactive challenge should be available where justice requires it to be.  That will depend, in each case, on the facts.

[56] In oral argument, Merafong suggested that the permissibility of its challenge should be judged by whether the right person brought the challenge in the appropriate tribunal, timeously, citing the decision-maker, with all the relevant evidence available, in light of rule of law considerations.  It seems to me that the approach in Bengwenyama,[75] albeit in the context of review remedies, offers practical guidance here.  The permissibility of a reactive challenge by an organ of state must depend on a variety of factors, invoked with a “pragmatic blend of logic and experience”.  And – as in Bengwenyama – it would be imprudent to pronounce any inflexible rule.

[57] It is correct that, in contrast to the government department in Kirland, Merafong here did formally counter-apply for a declarator that the Minister had no authority to issue her ruling.  It asked for the Minister’s ruling to be set aside.  Its notice of motion sought a declaration that it was vested with exclusive executive authority to set and adopt tariffs relating to the provision of water services within its area of jurisdiction and to impose and recover service fees and surcharges.  It also sought a declaration that section 8 of the Act did not confer authority on the Minister “to interfere with a tariff set and implemented” by Merafong.  And the Minister was served with the papers.  Though not appearing in either the High Court or the Supreme Court of Appeal, she signified in this Court that she had no objection to the reactive challenge being determined, though she said the Court should not reach it.

[58] The Supreme Court of Appeal in effect imposed a duty of proactivity on Merafong, though it did so without the benefit of the Minister’s views before it.  It held that Merafong could not simply ignore the Minister’s ruling.  Once it concluded the Minister’s decision was wrong, it was duty bound to initiate proceedings to set it aside – and until it did, the decision remained binding on it.  So far, so good.  But the Court went further.  It held that Merafong’s inaction in the face of a ruling it considered invalid later disqualified it entirely from resisting AngloGold’s application to enforce that ruling.  To counter-apply in the enforcement proceedings AngloGold had initiated was not competent.  The ruling had to stand because Merafong, though disputing it, had never itself gone to court to have it set aside.

Constitutional citizenship and co-operative governance

[59] Was the Supreme Court of Appeal correct to disbar Merafong from raising a reactive defence because it failed to take the initiative?  The answer is No – but the path to that answer must first be cleared.  First, as a matter of practice, and good constitutional citizenship, it is undoubtedly so that Merafong should have gone to court to set aside the Minister’s ruling.  As a state organ, Merafong had the resources, and the responsibility, to obtain judicial clarity in its dispute with AngloGold about the ruling.  Instead of doing so, it threatened to cut off AngloGold’s water.  That was not nice.  Worse, it was not good constitutional citizenship.

[60] As a good constitutional citizen, Merafong should either have accepted the Minister’s ruling as valid, or gone to court to challenge it head-on.  AngloGold did what Merafong advised it to do – it appealed to the Minister.  On legal advice, Merafong later recanted its view that AngloGold was entitled to appeal.  But that didn’t give it warrant to bully one of its ratepayers.  In enforcing its view of the Minister’s disputed ruling, Merafong was resorting to a form of self-help.

[61] This was out of kilter with Merafong’s duty as an organ of state and a constitutional citizen.  This Court has affirmed as a fundamental principle that the state “should be exemplary in its compliance with the fundamental constitutional principle that proscribes self-help”.[76]  What is more, in Khumalo,[77] this Court held that state functionaries are enjoined to uphold and protect the rule of law by, inter alia, seeking the redress of their departments’ unlawful decisions.[78]  Generally, it is the duty of a state functionary to rectify unlawfulness.  The courts have a duty “to insist that the state, in all its dealings, operates within the confines of the law and, in so doing, remains accountable to those on whose behalf it exercises power”.[79]  Public functionaries “must, where faced with an irregularity in the public administration, in the context of employment or otherwise, seek to redress it”.[80]  Not to do so may spawn confusion and conflict, to the detriment of the administration and the public.[81]  A vivid instance is where the President himself has sought judicial correction for a process misstep in promulgating legislation.[82]

[62] Section 41(3) of the Constitution requires an organ of state involved in an inter governmental dispute to “make every reasonable effort” to settle it before it approaches a court.  The provision’s cognate statute, section 40(1) of the Intergovernmental Relations Framework Act[83] (Framework Act) similarly requires organs of state to avoid legal proceedings in favour of co-operative amelioration.  This Merafong did.  But, when amelioration failed, the Constitution and the statute required more of it.

[63] In abjuring legal proceedings on the part of organs of state until there is no reasonable alternative, both section 41 and the Framework Act imply a corollary.  This is that, when all reasonable measures and alternative remedies have been exhausted, an organ of state to which a contested ruling applies should ordinarily go to court to have the legal rights and wrongs of the ruling determined.  In the circumstances, without holding that Merafong was under a stand-alone duty to clarify the Minister’s decision, once Merafong disputed the decision, and decided it did not wish to comply with it, Merafong owed a duty to AngloGold, which relied on the decision.  That duty was to seek clarification from the courts.  What it could not do was sit on its hands or defy the ruling by enforcing its own unilateral view.

[64] So although the principle of intergovernmental cooperation negates recourse to courts of law until every other avenue has been exhausted, when this point has been reached, the Constitution may require responsible governmental citizens to take recourse to law.

[65] So Merafong should not have been content to hold its hand over the tap, ready to turn it off if AngloGold got stroppy.  But is this enough to disqualify Merafong’s reactive defence now that it raises it in response to AngloGold’s attempt to enforce the Minister’s ruling?  I would say No.  For considerations springing largely, but not solely, out of convenience, I would permit Merafong’s challenge to be decided.

[66] First, even if belatedly, Merafong’s “conditional counter-application” in substance challenges the Minister’s ruling.  Unlike Kirland,[84] where the government department made no effort to counter-apply, Merafong formally sought an order embodying its approach to the dispute.  That order did not seek the setting aside of the Minister’s ruling, but a declaration only that she had no statutory or constitutional power to take it.  To that extent Merafong has joined issue with AngloGold.  And to that extent its merits in seeking to have that challenge determined in these proceedings deserve to be taken into consideration.

[67] Second, there is the obtrusive fact that dismissing Merafong’s appeal, without more, would further protract a dispute that has already run very long.  It would force Merafong to reinstitute proceedings, with all the expense, further delay and multiplication of formalities, pleadings and lawyerly exchanges this would entail.

[68] Since Merafong’s status as an organ of state does not categorically exclude it from a reactive challenge, I would not close the court’s door in its face in these proceedings.  For the reasons that follow, I would however remit its challenge to be decided afresh by the High Court.

[69] First, we must note that Merafong’s reactive challenge has distinctive attributes.  These render it different from those a subject raises when the state threatens imprisonment or coerces payment.  In these cases, which we may call “classical” collateral challenges, delay plays no role.  The subject is entitled, as of right, to scrutinise the lawfulness of coercive action because the rule of law requires that official power not be exercised against the liberty or property of a subject unless it is lawfully sourced.[85]

[70] The virtue of “classical” reactive challenges lies precisely in the fact that they provide a defence to parties who face the enforcement of the law but who never previously confronted it.  And it is for this reason that they may sometimes be disallowed.  Where a statute provides for an appeal or other remedy, and the disputed decision was specifically directed to the challenging party, our courts have forbidden a collateral challenge.[86]

[71] The point of these cases is that the ruling or decision was not directed to the world at large.  It was specific.  It was known to the subject.  They stand in contrast to instances where the law is of general application, and is possibly unknown to the person against whom it is sought to be enforced.  There, delay cannot be a disqualifying consideration.

[72] Here, Merafong was well aware of the Minister’s decision, which was specifically addressed to it.  It does not dispute that it knew that a legal challenge was immediately available to it.  This means that Merafong’s reactive challenge is of the category that necessitates scrutiny in regard to delay.

[73] A further and related reason to remit the matter to the High Court is this.  Merafong accepts that the Minister’s ruling was administrative action.  Whether under PAJA, or legality review,[87] it was obliged to institute proceedings to review the decision without unreasonable delay.[88]  The rule against delay in instituting review exists for good reason: to curb the potential prejudice that would ensue if the lawfulness of the decision remains uncertain.  Protracted delays could give rise to calamitous effects.  Not just for those who rely upon the decision but also for the efficient functioning of the decision-making body itself.  Had Merafong instituted a review application, as it ought, the Court hearing it would have had to consider whether the delay precluded its challenge.[89]

[74] In Khumalo, this Court held that state functionaries must act without delay in setting right their own legal missteps.[90]  The Court also held that an assessment of a plea of undue delay involves examining: (i) whether the delay is unreasonable or undue (a factual enquiry upon which a value judgment is made in the light of all the relevant circumstances); and if so (ii) whether the court’s discretion should be exercised to overlook the delay and nevertheless entertain the application.[91]

[75] The approach in Khumalo is apposite here.  It is not necessary to decide whether Merafong would or could have obtained an extension, for the present proceedings, which were initiated by AngloGold, are not under PAJA.  But the factors this Court endorsed in Khumalo require consideration.  And there is nothing pertinent before this Court.  Merafong’s answering affidavit in the High Court nowhere addresses its failure to bring proceedings, and hence the delay in its counter-challenge.  That, of course, was because Merafong was content to keep its hand on the tap.  It awaited AngloGold’s move, which eventually came in April 2011.

[76] But, on remittal to the High Court, with appropriate explanation in further affidavits, there may be convincing reasons for Merafong’s approach.  These may arise from the fact that it was advised, as its answering affidavit in the High Court averred, that it was “entitled” to ignore the ministerial ruling and compel AngloGold to pay the surcharge.  Ordinarily this would not be an excuse.  But, Merafong may urge, the law was not clear.  The reasons may conceivably relate, also, to the quality of Merafong’s good faith belief in the lawfulness of its powers and approach.  Further information will enable the hurdles of lateness and potential and actual prejudice to be raised, debated and properly considered.

[77] What is more, because delay, though relevant, need not be conclusive in a reactive challenge, Merafong may also invoke the fact that the Minister in these proceedings has expressed the view that its challenge can be considered (though decided against it).

[78] A third reason to remit for the High Court to decide the validity of the Minister’s decision is that Merafong’s counter-challenge was bare.  It relied solely on the statutory and constitutional setting to decry the Minister’s invocation of appellate powers over its surcharge.  The record of the Minister’s decision was not before it.  The High Court nevertheless decided the statutory and constitutional issues against Merafong.  But it did so without considering whether Merafong’s delay should disqualify its challenge.  That was because it held that Merafong was categorically excluded from raising a collateral challenge.  And the Supreme Court of Appeal dismissed Merafong’s appeal solely on the basis of Merafong’s categorical exclusion.

[79] Once Merafong’s categorical exclusion has been set aside, the proper process is for the High Court to have before it the Minister’s record of decision, together with Merafong’s explanation for the delay, on appropriate terms for AngloGold to respond, and for the question of delay to be decided first.  Thereafter the validity of the Minister’s decision, based on anything arising pertinently from the record, in its statutory and constitutional setting, may be considered.

[80] A fourth and final reason to remit the matter to the High Court is for it to consider argument and, if necessary, evidence on the question of remedy, should Merafong’s challenge be time-barred or otherwise fail.  This Court has had no submissions on this important issue, and the High Court, though rejecting Merafong’s challenge, did not consider it.

[81] This approach, instead of shutting the door in Merafong’s face, conditionally permits its challenge to proceed.  In doing so, this Court recognises that a range of reactive or collateral challenges exists.  In some, delay is axiomatically irrelevant.  In others, it counts.  It is not necessary, here, to consider the factors specific to each kind: those will be developed in the case law.  For now, it is enough to free Merafong’s challenge from the categorical disqualification, and to allow it to go ahead on appropriate terms.

[82] I should add that it is also inapposite for this Court to determine Merafong’s constitutional challenge.  Merafong avowedly did not persist in this before the SCA.  Before us, it did not mention the issue in its written argument, nor did it allude to it in oral argument.  When counsel for Merafong was asked about it, he averred simply “it’s alive on the papers”.[92]  This Court invited submissions from the Minister, who had not appeared in the High Court and SCA.  The Court itself here inquired about the constitutional point.  The Minister urged that the point not be decided.  But Merafong now seized the opportunity to assert that it could be decided.  That is belated opportunism the Court should not countenance.  Since Merafong had in effect let the point lie, so far as not even to make written or oral submissions on it, it is not in the interests of justice to allow it to now try to resuscitate it.  In any event, counsel for Merafong submitted in oral argument that the constitutional point was “conditional on this Court finding that the Minister’s decision was lawful - that she had jurisdiction in terms of the Act to make the decision”.  Since, for the reasons I have set out we relay that very question to the High Court, it follows that, even on Merafong’s approach, the constitutional point should be decided only later.

Summary

[83] Unlike the Supreme Court of Appeal, I would not disqualify Merafong’s reactive defence because it is an organ of state.  Merafong must be permitted to raise a challenge to the Minister’s decision, but on appropriate terms that call for it to explain properly its delay in challenging that decision.  In those circumstances, the most equitable order as to the costs is to allow the reviewing court to determine them. Because AngloGold was seeking to enforce an order of an organ of state it should not, according to the well-known principle in Biowatch,[93] be made to pay the costs of its effort to do so, thus far.

Order

[84] I propose the following order:

1.                  Leave to appeal is granted.

2.                  The appeal is upheld.

3.                  The orders of the High Court of South Africa, Gauteng Division, Pretoria and the Supreme Court of Appeal are set aside.

4.                  In their stead, the matter is remitted to the High Court to determine, after the lodging of further affidavits as the applicant, Merafong City Local Municipality, and the Minister of Water Affairs and Forestry may consider appropriate, on the lawfulness of the Minister’s decision of 18 July 2005, and, if necessary, what remedy is to be granted.

5.                  The Minister is to lodge the record of the decision by 4 November 2016.

6.                  The further affidavits, if any, by the applicant are to be lodged by 18 November 2016 and by the Minister by 25 November 2016.

7.                  The respondent, AngloGold Ashanti Limited, may lodge its affidavits, if any, by 6 December 2016.

8.                  Costs are reserved for consideration by the High Court.

 

JAFTA J (Bosielo AJ and Zondo J concurring):

Introduction

[85] I have read the judgment of my colleague Cameron J (first judgment) and agree that leave must be granted and that the orders of the High Court and the Supreme Court of Appeal be set aside.  However, I do not support the further orders proposed in the first judgment.

[86] With regard to jurisdiction and leave the first judgment asserts that “the dispute about the Minister’s power to intervene on a matter of municipal competence raises constitutional issues that fall within this Court’s jurisdiction”.  It adds that Merafong’s arguments in this regard have substance.  But in addressing the merits of the case the first judgment does not reach the very dispute that motivated it to grant leave nor does it consider Merafong’s arguments which it holds, have substance.

[87] However, I agree with the first judgment that the enforcement of the Minister’s decision of 18 July 2005 in terms of which she overturned Merafong’s decision to levy a surcharge on the supply of water within its area of jurisdiction is the core issue in this matter.  Merafong impugns the Minister’s decision on two grounds.  First, it argues that, properly construed, the provision in terms of which the Minister claims to have taken the decision does not vest her with the necessary power.  Second and if it does, the provision is inconsistent with the Constitution and for that reason is invalid.  I agree that on the face of it, there is substance in these arguments.  Consequently, they must be addressed to test that substance.

[88] Additional to the reasons advanced by the first judgment for leave is the fact that this case presents this Court with an opportunity of defining the reach of the principles in both Oudekraal[94] and Kirland[95] in the context of the Constitution and the principle of constitutional supremacy.  This is because the two decisions derive their validity, as every precedent and other law, from the Constitution.  Here the judgments of the High Court and Supreme Court of Appeal illustrate a misapplication of the two decisions.  For example, both Courts held that even if it was illegal, the Minister’s decision was valid and binding for as long as it remained in existence.

[89] The proposition that both Courts distilled from Oudekraal and Kirland was that an invalid administrative act that exists in fact is binding and enforceable until set aside by a competent court.  This proposition collides head-on with the principle of legality which is an integral part of the rule of law.  And the first judgment endorses this proposition.[96]  I disagree for reasons I shall outline in a moment.

[90] A further issue raised is the question whether, as the High Court and Supreme Court of Appeal concluded, a collateral challenge is a defence not available to an organ of state.  This Court is yet to express its view on the issue.  The other Courts already have asserted that the right to raise a collateral challenge is limited to individuals because liberty or property of an individual may be taken away on authority of a lawful decision.  The first judgment prefers to leave the question open for now.

[91] I propose to begin with the question whether in law an organ of state is prohibited from raising a collateral challenge.  It is necessary for this Court to put this issue to rest, one way or the other.  Leaving it open means that courts in this country are bound by decisions of the Supreme Court of Appeal that say a collateral challenge is not available to the state.  Therefore, it is in the interests of justice that there be certainty on whether this Court affirms that conclusion or not.  Next in analysis will be the scope and content of the Oudekraal and Kirland principles before addressing the merits.

Collateral challenge

[92] A collateral challenge is raised to resist coercive compliance with a legal act or law.  It is nothing else but an argument advanced in proceedings before a court, to the effect that the legal decision or law sought to be enforced is invalid and therefore the party against whom enforcement is pursued must not be ordered to comply with an invalid act or law.

[93] As the Supreme Court of Appeal held in Helderberg Park Development,[97] a collateral challenge may be raised only when the impugned act is invoked to coerce a party raising it into compliance.  This is so because the target of the challenge is the compulsion to comply.  This means that there is no time limit within which the defence may be invoked.  The defendant may have been supine until an attempt to compel is made and only then she can contend that she should not be ordered to comply because the act is itself invalid.  In Oudekraal the Supreme Court of Appeal stated that a person may mount a collateral challenge “because the legal force of the coercive action will most often depend upon the legal validity of the administrative act in question”.

[94] Later the Court emphasised the point in these words:

“[T]he right to challenge the validity of an administrative act collaterally arises because the validity of the administrative act constitutes the essential prerequisite for the legal force of the action that follows and ex hypothesis the subject may not then be precluded from challenging its validity.”[98]

[95] This statement underscores the fact that, for an administrative act to be enforceable, it must be valid.  And for it to be valid it must, among other requirements, be lawful.  If it is illegal, an administrative act cannot be enforced because it would be inconsistent not only with the rule of law but also with section 33 of the Constitution which guarantees “an administrative action that is lawful, reasonable and procedurally fair”.[99]

[96] Therefore a successful collateral challenge promotes the objects of the rights guaranteed by section 33 of the Bill of Rights and the rule of law by prohibiting enforcement of invalid and illegal decisions.  Ordinarily our courts do not enforce illegal decisions because their duty is to uphold the Constitution and the law.[100]  Therefore, no court may close its eyes to a collateral challenge raised by a party with interest if there is substance in the defence.  It is against this backdrop that the proposition that a collateral challenge is not available to the state must be evaluated.

Collateral challenge by the state

[97] Relying on Kirland and its decision in Cable City,[101] the Supreme Court of Appeal here endorsed the conclusion that it is not competent for the state to raise a collateral challenge.  The Court stated:

“It is clear from these dicta that Merafong was obliged to approach the court to set the Minister’s ruling aside and that it breached the principle of legality by simply disregarding it.  And the collateral challenge it sought to mount against the ruling does not avail it because it is an organ of state.  It is established in our law that a collateral challenge to the validity of an administrative action is a remedy available to a person threatened by a public authority with coercive action precisely because the legal force of the coercive action will most often depend on the legal validity of the administrative action in question.  The notion that an organ of state can use this shield against another organ of state is simply untenable.  These findings dispense with the need to deal with the substantive issues raised in the matter.  The appeal must fail.”[102]

[98] But the referenced paragraphs in Kirland and Cable City do not support this proposition.  Paragraph 35 of Kirland on which reliance was placed is part of a minority judgment and deals with the question whether the order sought by Kirland could be granted despite a finding that the decision it sought to confirm had been held to be invalid.  No reference was made to a collateral challenge in that paragraph.  Whereas in Cable City the Court merely stated the principle that “a party has a right to raise a collateral challenge to the validity of an administrative act if he was threatened by a public authority with coercive action”.  In that case it was a private company that raised a collateral challenge and therefore the statement in paragraph 15 must be read in that context.

[99] In Kirland the majority too did not address the question of a collateral challenge, let alone that the defence could not be raised by an organ of state.  While Cable City considered and upheld the collateral challenge raised there, the Court did not determine whether the state could advance that defence.

[100] The fact that in both Oudekraal and Cable City, the Supreme Court of Appeal stated that a citizen threatened with coercive action by a public authority is entitled to raise a collateral challenge does not, without more, mean that the defence is not available to an organ of state threatened by another organ of state with coercive action based on an illegal administrative decision.  Nor does it mean that, when coercive action is exacted by a private party against an organ of state, the latter may not raise a collateral challenge.

[101] I can think of no reason in logic or principle that militates against the state raising a collateral challenge where it faces a claim that it should comply with an illegal decision.  Take for example, a decision that is inconsistent with the Constitution, it can hardly be said that an organ of state is precluded from resisting its enforcement by showing that it is invalid by reason of being in violation of the Constitution.  And the prohibition being that the party raising it is precluded from doing so on the sole basis that it was an organ of state and nothing more.

[102] The proposition becomes more untenable if the fact that the same organ of state may advance exactly the same ground in a counter-application in the same proceedings.  On the authority of Kirland, if a public official harbours an opinion that an administrative act is invalid, she may not simply disregard the act.  She must institute a review application to have it set aside.

[103] In Kirland Cameron J said:

“PAJA requires that the government respondents should have applied to set aside the approval, by way of formal counter-application.  They must do the same even if PAJA does not apply.  To demand this of government is not to stymie it by forming upon it a senseless formality.  It is to insist on due process, from which there is no reason to exempt government.  On the contrary, there is a higher duty on the state to respect the law, and to fulfil procedural requirements and to tread respectfully when dealing with rights.”[103]

[104] To insist that, if an organ of state challenges the validity of an administrative act, it is compelled to do so only by a counter-application would amount to placing form above substance.  That would be a narrow technical approach that serves nothing but that a particular form should be followed.  This is because the same objection could be raised regardless of whether it is framed as a counter-application or a collateral challenge.  This sort of formalism is not required by PAJA or rules of procedure for our courts.

[105] The Constitution recognises the inherent power our courts have to regulate and protect their process if the interests of justice so demand.[104]  This power enables courts to depart from their rules in the interests of justice.  In PFE International this Court remarked:

“Since the rules are made for courts to facilitate the adjudication of cases, the superior courts enjoy the power to regulate their processes, taking into account the interests of justice.  It is this power that makes every superior court the master of its own process.  It enables a superior court to lay down a process to be followed in particular cases, even if that process deviates from what its rules prescribe.  Consistent with that power, this Court may in the interests of justice depart from its own rules.”[105]

[106] I conclude that an organ of state may raise a collateral challenge.  This brings me to the scope of the principles in Oudekraal and Kirland.

Principles

[107] Because of the misapplication of the principle laid down in Oudekraal it has become necessary for this Court to determine the scope and content of that principle.  Its misapplication has muddled up our law, turning on its head basic principles like: an illegal administrative act has no legal force and as such cannot be enforced.  This is a principle that flows from the rule of law principle of legality which is to the effect that an illegal administrative act, although it may exist in fact, does not exist in law and consequently it may not be enforced because it is not binding.  This is so because an administrative act derives its legal force from its validity.  Simply put an invalid act is unenforceable.

[108] These principles have been part of our law from time immemorial and no authority is necessary to be cited.  But since the adoption of the Constitution, these principles have been reinforced by making the rule of law a foundational value and part of the Constitution.  Significantly what this means for present purposes is that the rule of law is entrenched as part of our Constitution and in turn that means that any administrative act inconsistent with the rule of law is invalid and therefore has no legal force and consequently cannot be enforced.  This is because the Constitution is our supreme law and any conduct that is inconsistent with it is invalid.

[109] This is what Affordable Medicines Trust emphatically tells us.  In that case Ngcobo J said:

“Our constitutional democracy is founded on, among other values, the ‘[s]upremacy of the constitution and the rule of law.’  The very next provision of the Constitution declares that the ‘Constitution is the supreme law of the Republic; law or conduct inconsistent with it is invalid’.  And to give effect to the supremacy of the Constitution, courts ‘must declare that any law or conduct that is inconsistent with the Constitution is invalid to the extent of its inconsistency’.  This commitment to the supremacy of the Constitution and the rule of law means that the exercise of all public power is now subject to constitutional control.

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.  It entails that both the legislature and the executive ‘are constrained by the principle that they may exercise no power and perform no function beyond that conferred upon them by law.’  In this sense the Constitution entrenches the principle of legality and provides the foundation for the control of public power.”[106]

[110] Consistent with these principles this Court held that administrative acts that were ultra vires (beyond the powers) were in breach of the doctrine of legality and as a result were not binding as they lacked the necessary force.  This Court outlined the invalidity of such acts in these terms:

“In exercising the power to make regulations, the Minister had to comply with the Constitution, which is the supreme law, and the empowering provisions of the Medicines Act.  If, in making regulations the Minister exceeds the powers conferred by the empowering provisions of the Medicines Act, the Minister acts ultra vires (beyond the powers) and in breach of the doctrine of legality.  The finding that the Minister acted ultra vires is in effect a finding that the Minister acted in a manner that is inconsistent with the Constitution and his or her conduct is invalid.  What would have been ultra vires under common law by reason of a functionary exceeding his or her powers, is now invalid under the Constitution as an infringement of the principle of legality.  The question, therefore, is whether the Minister acted ultra vires in making regulations that link a licence to compound and dispense medicines to specific premises.  The answer to this question must be sought in the empowering provisions.”[107]

[111] This legal position is further buttressed by section 33 of the Constitution which guarantees “administrative action that is lawful, reasonable and procedurally fair”.  This means that an unlawful administrative act is not an act contemplated in the Constitution.  An act of that kind would be inconsistent with the Constitution and for that reason invalid.

[112] Notwithstanding these fundamental principles, the High Court here held:

“The Municipality has not in its papers sought to review or overturn the Minister’s decision and thus based on the Oudekraal principle the Minister’s decision stands until set aside by a court of law.  The decision is therefore binding and enforceable and the municipality should abide by it.”[108]

[113] For its part the Supreme Court of Appeal held that, even if the Minister’s decision was ultra vires and unlawful, Merafong was bound by it.  This conclusion is approved by the first judgment.  It says:

“The Supreme Court of Appeal in effect imposed a duty of proactivity on Merafong, though it did so without the benefit of the Minister’s views before it.  It held that Merafong could not simply ignore the Minister’s ruling.  Once it concluded the Minister’s decision was wrong, it was duty bound to initiate proceedings to set it aside – until it did, the decision remained binding on it.  So far, so good.  But the Court went further.”[109]

[114] With respect the conclusion reached by the High Court and Supreme Court of Appeal, endorsed by the first judgment, is incorrect.  It suggests that an unlawful administrative act that exists in fact, and not in law, has legal force and is binding for as long as it is not set aside.  This is in direct conflict with established authority like Affordable Medicines Trust and many others.  But most importantly, the conclusion pays no regard to the supremacy of the Constitution which expressly declares that conduct that is inconsistent with it is invalid.

[115] The source of this wrong statement of the law, as the High Court suggests, is said to be Oudekraal.  While the Supreme Court of Appeal placed reliance on the decision of this Court in Kirland.  Therefore it is necessary to consider these cases.

Oudekraal

[116] I must state at the outset that Oudekraal is not authority for the proposition that an invalid administrative act is binding as long as it is not set aside by a competent court.  No court has the power of converting an unconstitutional and invalid act with no legal force into a valid act with binding effect.  This is so, it must be stressed, because the Constitution is supreme and it declares that conduct inconsistent with it is invalid.  That which is proclaimed to be invalid by the Constitution cannot be overruled by any court.  Courts are established and derive their powers from the Constitution which is binding on all arms of government, including the Judiciary.

[117] Consistent with its supremacy, in section 172(1)(a) the Constitution obliges every court, in deciding a matter within its competence, to declare law or conduct inconsistent with it invalid.[110]  In Bengwenyama[111] this Court affirmed that the duty imposed by the section is mandatory.  This Court went further to dispel the common law notion that a court in review proceedings has a discretion to withhold the remedy and declared that once it is proved that an administrative act is inconsistent with the Constitution and PAJA; a court must declare it invalid.[112]  In this regard no court has discretion.  But the consequences of the declaration of invalidity may be ameliorated by a further granting of a just and equitable order.  It is at this stage that a court has a discretion to exercise.

[118] In Bengwenyama Froneman J remarked:

“The apparent rigour of declaring conduct in conflict with the Constitution and PAJA unlawful is ameliorated in both the Constitution and PAJA by providing for a just and equitable remedy in its wake.  I do not think that it is wise to attempt to lay down inflexible rules in determining a just and equitable remedy following upon a declaration of unlawful administrative action.  The rule of law must never be relinquished, but the circumstances of each case must be examined in order to determine whether factual certainty requires some amelioration of legality and, if so, to what extent.  The approach taken will depend on the kind of challenge presented – direct or collateral; the interests involved and the extent or materiality of the breach of the constitutional right to just administrative action in each particular case.”[113]

[119] Therefore, the judgments in Oudekraal and Kirland must be read in the context just outlined here.  Oudekraal lays down the principle that in the limited situation of consecutive administrative decisions and if the empowering provision requires, as a  pre-decision condition, that the first act be in existence for the second act to be made, the mere factual existence of the first act would be enough for the validity of the second act.  Depending on the terms of the empowering provision, the validity of the second act may not be challenged on the ground that the first act was substantively invalid even though it was not set aside.

[120] Stating the ratio in Oudekraal the Supreme Court of Appeal said:

“Thus the proper enquiry in each case – at least at first – is not whether the initial act was valid but rather whether its substantive validity was a necessary precondition for the validity of consequent acts.  If the validity of consequent acts is dependent on no more than the factual existence of the initial act then the consequent act will have legal effect for so long as the initial act is not set aside by a competent court.”[114]

[121] What happened in Oudekraal was that a landowner had obtained approval to establish a township on its property subject to a number of subsequent decisions.  One of them was the approval of a general plan of the proposed township by the Surveyor-General and the other was lodgement of the approval with the Registrar of Deeds.  These acts had to occur within a “specified period, failing which the approval to establish the township would lapse”.  However, the period could be extended and indeed it was purportedly extended but after the expiry of the prescribed period.  During the purported extension the general plan was approved and lodged with the Registrar of Deeds.  This was obviously unlawful because after the expiry of the prescribed period the relevant functionary had no power to extend.

[122] Years later the landowner sought approval of the engineering services plan, as a further step towards the establishment of the township.  The Municipality refused to approve this plan, citing the fact that the landowner’s right to establish the township lapsed when it failed to obtain approval of the general plan and its lodgement within the prescribed period.  In an application to the High Court to compel the Municipality to approve the engineering services plan, the Court declined to grant the relief sought on the ground that its effect would be to proclaim that an “illegal action had somehow evolved into a legal decision and that would undermine the principle of legality”.

[123] The question that arose was whether the approval to establish the township which was shown on the facts to have been unlawfully granted, could lead to subsequent valid acts.  It was in this context that the Supreme Court of Appeal said an invalid administrative act may have legal consequences.  For this proposition the Court cited as authority the following statement:

“[A]n invalid administrative act may, notwithstanding its non-existence [in law], serve as the basis for another perfectly valid decision.  Its factual existence, rather than its invalidity, is the cause of the subsequent act, but that act is valid since the legal existence of the first act is not a precondition for the second.”[115]

And the Court further quoted and emphasised this:

“The crucial issue to be determined is whether that second actor has legal power to act validly notwithstanding the invalidity of the first act.  And it is determined by an analysis of the law against the background of the familiar proposition that an unlawful act is void.”[116]

[124] Consecutive administrative acts referred to in Oudekraal occur where, for example, the empowering provision requires the existence of a recommendation as a pre-condition for the making of an administrative decision.  In this instance both the recommendation and the ensuing decision constitute consecutive administrative acts.  As this Court observed in Walele^[117] a provision that empowered a municipality to approve building plans required the municipality when exercising that power to consider a recommendation by a building control officer.  In this instance the approval of the plans must be preceded by the recommendation which constitutes a jurisdictional fact.

[125] If the empowering provision requires the mere existence of the recommendation as was the case in both Walele and Bakgatla, then a recommendation that is invalid in law but exists in fact is capable of giving rise to a valid approval.  This is what Oudekraal means by the proposition that an invalid administrative act may have legal consequences.  In this context the invalid act refers to the recommendation and not the approval.  But if the approval itself is illegal, it can never have legal consequences flowing solely from its factual existence.  That is not what Oudekraal pronounced.

[126] But even in the case of consecutive acts, if the empowering provision requires the recommendation to be valid, its factual existence alone will not result in a valid approval.  This is because the legal existence of the recommendation as opposed to its factual existence is a precondition for the approval.  This much is clear from the analysis of Prof. Forsyth’s work quoted in paragraph 36 above on which Oudekraal relies.  It all comes down to what the empowering provision permits.  What is clear though is that the Oudekraal principle is limited to the situation of consecutive acts and even then to the first act.

[127] But most importantly, Prof. Forsyth himself accepts the simple proposition that an invalid administrative act does not exist in law and that an unlawful act is void.  A plain consequence of this is that such administrative act is not binding because it has no legal force.  This reflects an accurate position in our law.  Therefore the conclusion reached by the High Court, based on Oudekraal arose from its mistaken understanding of Oudekraal.

Kirland

[128] The majority in this Court in Kirland, with respect, extended the reach of Oudekraal far beyond its scope when it declared:

“The essential basis of Oudekraal was that invalid administrative action may not simply be ignored, but may be valid and effectual, and may continue to have legal consequences, until set aside by proper process.”[118]

[129] Happily this statement was not part of the ratio.  The issue that arose in Kirland was whether an administrative act which was facially unlawful could be set aside in those proceedings without a formal application from the MEC for Health to have a fraudulent administrative approval set aside.  It was in this context that the majority said:

“In summary: having failed to counter-apply during these proceedings, the Department must bring a review application to challenge the approval granted to Kirland, which remains valid until set aside.  In those proceedings, the Department will no doubt explain its dilly-dallying by accounting for the long months before it acted.  As respondent, Kirland will in turn be entitled to defend the decision, whether on the ground of its validity, or on the ground that it should not be set aside, even if it is invalid.”[119]

[130] The true position in our law is as set out in Affordable Medicines Trust.  An illegal administrative act is inconsistent with the Constitution and the rule of law.  The inconsistency renders it invalid, regardless of the fact that it is not set aside, because in our constitutional order the Constitution is supreme.  In our law an unlawful act is void ab initio and thus it can have no legal force and effect.

[131] As a result, Kirland must be construed as laying down the principle that public officials must, for the sake of certainty in law and preventing innocent parties from acting on invalid acts, apply for the setting aside of acts which they consider to be invalid.  They should not ignore such administrative acts.  But this does not mean, if they fail to do so, the invalid act automatically morphs into a valid act with binding legal effect.  The inaction on the part of the official cannot render valid an act that is inconsistent with the Constitution.  There is no constitutional or other legal basis for such evolution which is at odds with the constitutional supremacy.

[132] The one difficulty with the statement that says an invalid administrative act may be valid and binding until set aside is how it is applied in practice.  Bearing in mind that what gives validity to such act is not the correct exercise of power by the decision-maker, but the failure to seek the setting aside of the invalid act, the question that arises is: at what point of the invalid act does the failure trigger validity and binding effect?  Surely it cannot be as soon as the invalid decision is made.  It is not clear to me whether the invalid act automatically changes into a valid act or some further decision must be taken.

[133] With regard to an illegal administrative act, the difficulty is at what point does the legality principle cease to operate.  As observed in Affordable Medicines Trust, an illegal act “is now invalid under the Constitution as an infringement of the principle of legality”.  How then can it be said that an illegal act becomes valid merely because public officials failed to have it set aside?  Does this imply that section 33 of the Constitution which requires that administrative action be lawful ceases to apply if there is a failure to set the action aside?  All these questions are not addressed in the first judgment.

[134] These questions were also not addressed by the majority in Kirland.  Moreover, no reasons were advanced in Kirland for departing from the unanimous decision of this Court in Affordable Medicines Trust.  The principle in Kirland is also at odds with the constitutional rule of objective invalidity.  Conduct that is inconsistent with the Constitution is invalid from the moment the decision is taken and remains invalid, regardless of whether it is set aside or not.

[135] Section 2 of the Constitution declares its supremacy and provides that “law or conduct inconsistent with it is invalid”.  According to the principle of objective invalidity, pre-Constitution legislation which is inconsistent with the Constitution became invalid on the date the Constitution came into force.  But legislation that was passed after the Constitution came into operation became invalid from inception.  This is because section 2 tells us that such legislation is invalid.  This principle applies with equal force to administrative action which constitutes conduct contemplated in section 2.  In light of this section there can be no justification for treating conduct differently from law because both are mentioned in in one sentence.  Both are subject to the supremacy of the Constitution.  This principle is buttressed by section 172(1) which obliges every court, when deciding a constitutional matter within its power, to declare law or conduct inconsistent with the Constitution invalid to the extent of the inconsistency.

[136] But the declaration of invalidity of law or conduct does not invalidate the law or conduct in question.  The court order merely declares the law or conduct invalid.  In contrast it is the Constitution itself which invalidates laws or conduct inconsistent with it.  In Ferreira[120] this Court proclaimed:

“The Court’s order does not invalidate the law; it merely declares it to be invalid.  It is very seldom patent, and in most cases is disputed, that pre-constitutional laws are inconsistent with the provisions of the Constitution.  It is one of this Court’s functions to determine and pronounce on the invalidity of laws, including Acts of Parliament.  This does not detract from the reality that pre-existing laws either remained valid or became invalid upon the provisions of the Constitution coming into operation.  In this sense laws are objectively valid or invalid depending on whether they are or are not inconsistent with the Constitution.  The fact that a dispute concerning inconsistency may only be decided years afterwards, does not affect the objective nature of the invalidity.  The issue of whether a law is invalid or not does not in theory therefore depend on whether, at the moment when the issue is being considered, a particular person’s rights are threatened or infringed by the offending law or not.”[121]

[137] Similarly the invalidity of administrative action depends on whether it is inconsistent with the Constitution.  If the decision was taken after the Constitution had come into effect, the decision becomes invalid at the time it is taken.  The fact that the question of its inconsistency is determined years after it was taken “does not affect the objective nature of invalidity”.  Nor does the failure to institute a review on the part of a public official render the act valid.

[138] By declaring the duty on public officials to apply for the setting aside of invalid administrative decisions, Kirland sought to avoid uncertainty arising from self-help, whereby public officials ignore decisions they regard to be invalid.  But by adding that an invalid administrative act that is not set aside “may become valid and effectual”, Kirland achieves quite the opposite by introducing uncertainty in our law with regard to legal force of illegal administrative decisions.  The questions raised here illustrate this point.

[139] But over and above uncertainty the proposition that an invalid act becomes valid and binding is a licence to subverting the Constitution and the rule of law.  It means that the longer an invalid act is not set aside, the Constitution and its supremacy cease to apply to the act which becomes valid and binding.  And once such act has morphed into being valid, there is no reason in logic and principle that precludes the person in whose favour the valid act was made, from enforcing it not only against the state but against the whole world.  This is so because in our law we do not have administrative decisions that are enforceable against the state only.  Administrative decisions confer rights on those persons in whose favour they are made and such rights once created are enforceable against everybody.

[140] But that proposition is not supported by any specific clause in the Constitution.  As mentioned, the proposition is inconsistent with sections 2 and 33 of the Constitution.  Whereas properly construed the principle in Oudekraal acknowledges that even in the case of consecutive administrative acts, the first invalid act is not enforceable but may result into a valid second act if the empowering provision requires a mere factual existence of the first act when the decision on the second act is made.[122]

[141] In contrast Kirland introduces a new brand of administrative action into our law.  The origin of the new action is not the exercise of public power but a default position.  It derives its validity not from any empowering provision but from the failure to institute review proceedings.  This brand of administrative action exists outside the Constitution which first and foremost is the source of all public power.  The Constitution does not envisage administrative action that springs from inaction by public officials.

[142] This approach suggests that there are two systems of law which regulate validity of administrative actions.  The one system is sourced from the Constitution. In terms of this system an unlawful administrative action is invalid.  The other system flows from case law.  In terms of this system an unlawful administrative action becomes valid and binding, for as long as it is not set aside on review.  In the same vein the Supreme Court of Appeal in Container Logistics declared:

“Judicial review under the Constitution and under the common law are different concepts.  In the field of administrative law constitutional review is concerned with the Constitution and legality of the administrative action, the question in each case being whether it is or is not consistent with the Constitution and the only criterion being the Constitution itself.  Judicial review under the common law is essentially also concerned with the legality of administrative action, but the question in each case is whether the action under consideration is in accordance with the behests of the empowering statute and the requirements of natural justice.”[123]

[143] In Pharmaceutical Manufacturers this Court rejected the view that there are two systems of law that control the exercise of public power.  This Court held that the common law principles have been subsumed under the Constitution from which they gain their force.  Chaskalson P proclaimed:

“I cannot accept this contention which treats the common law as a body of law separate and distinct from the Constitution.  There are not two systems of law, each dealing with the same subject matter, each having similar requirements, each operating in its own field with its own highest court.  There is only one system of law.  It is shaped by the Constitution which is the supreme law, and all law, including the common law, derives its force from the Constitution and is subject to constitutional control.”[124]

[144] But notably of the important principles emerging from Pharmaceutical Manufacturers is that under both the common law and the Constitution, an unlawful administrative action is invalid.  The Court stated:

“What would have been ultra vires under the common law by reason of a functionary exceeding a statutory power is invalid under the Constitution according to the doctrine of legality.  In this respect, at least, constitutional law and common law are intertwined and there can be no difference between them.  The same is true of constitutional law and common law in respect of the validity of administrative decisions within the purview of section 24 of the interim Constitution.  What is ‘lawful administrative action’, ‘procedurally fair administrative action’ and administrative action ‘justifiable in relation to the reasons given for it’, cannot mean one thing under the Constitution, and another thing under the common law.”[125]

[145] The irony here is that Kirland invokes the rule of law requirement for certainty in support of the proposition that an invalid administrative act becomes valid if not set aside.[126]  But this is done in breach of legality which also forms part of the rule of law.  In plain terms the rule of law is invoked to justify its breach and the principle of constitutional supremacy.  The purpose of constitutional supremacy is to prevent abuse of public power.  This is achieved by declaring invalid any conduct inconsistent with the Constitution.  Yet the Kirland principle creates space for corrupt public officials to abuse public power for their selfish interests.

[146] Two examples are sufficient to illustrate the point.  Take a case of a corrupt head of department who, in breach of section 217 of the Constitution and the relevant statutory framework, illegally extends a contract for procurement of services for 10 years, at a cost of R50 million per month.  He ensures that for five years the extension is not set aside because the power to institute legal proceedings by the department vests in him.  According to the principle in Kirland, the unconstitutional and illegal extension of the corrupt functionary becomes valid and binding for as long as it is not set aside.  This means that the service provider in whose favour the decision was made may enforce it with impunity for the duration of the extension.  This is because the extension would, according to Kirland, have become valid and effectual.  These facts reflect what occurred in Tasima where, on authority of Kirland the Supreme Court of Appeal held:

“According to general principles laid down by this Court in Oudekraal (paragraph 26) administrative decisions must be treated as valid until set aside, even if actually invalid.”[127]

[147] In a similar vein here too the Supreme Court of Appeal held that even if the Minister’s decision was ultra vires and unlawful, it was valid for as long as it was not set aside in a review application.[128]  But I have already pointed out here that the ratio of Oudekraal is in paragraph 31 and not 26.

[148] Yet in Prodiba the Supreme Court of Appeal declared an extension made by the same head of department in identical circumstances to be void ab initio.  This was done even though the review was brought two years after the decision to extend was taken.  There the Supreme Court of Appeal said:

“By not embarking on a competitive bid process, particularly given the nature and scale of the services to be provided, including the cost implications, Mr Mahlalela erred fundamentally.  By concluding the agreement without the approval of his employer and political principal and/or of the Cabinet, he acted without authority.  By concluding the agreement and incurring a liability for which there had been no appropriation, he not only erred, but acted against mandatory statutory prescripts and against the constitutional principles of transparent and accountable governance.  For all these reasons the agreement is liable to be declared void ab initio.  Consequently the appeal must be upheld.”[129]

[149] The Court in Prodiba could not have reached the conclusion that the extension was void ab initio if during the course of the two years, between the time the decision was taken and the stage at which it was set aside, that extension had become valid.  On the contrary this conclusion shows that the Court regarded the extension to have been invalid from inception up to the time it was set aside.

[150] Consequently, an illegal or unlawful administrative act may not be enforced or complied with, regardless of how long it remains in existence.  The High Court and the Supreme Court of Appeal here erred in concluding that, an illegal act is binding until it is set aside.

[151] What is said here about Kirland is in line with what was stated in Economic Freedom Fighters.[130]  There Kirland was cited to make the point that “no decision grounded on the Constitution or law may be disregarded without recourse to a court of law”.  The emphasis was on the fact that decisions based on the Constitution may not be disregarded.  This point was made clearer in these terms:

“The rule of law requires that no power be exercised unless it is sanctioned by law and no decision or step sanctioned by law may be ignored based purely on a contrary view we hold.  It is not open to any of us to pick and choose which of the otherwise effectual consequences of the exercise of constitutional or statutory power will be disregarded and which given heed to.  Our foundational value of the rule of law demands of us, as a law-abiding people, to obey decisions made by those clothed with the legal authority to make them or else approach courts of law to set them aside, so we may validly escape their binding force.”[131]

[152] It is apparent from this statement that in Economic Freedom Fighters this Court reaffirmed the rule of law to the extent that “no power [may] be exercised unless it is sanctioned by law and no decision or step sanctioned by law may be ignored based purely on a contrary view we hold”.  The statement goes further to state that everybody must “obey decisions made by those clothed with legal authority to make them”.  This is an antithesis of the proposition that illegal administrative decisions are valid and binding if not set aside.  Therefore reliance placed on Economic Freedom Fighters is misplaced.  Moreover, that case was concerned with the question whether the remedial action taken by the Public Protector was binding until set aside on review.  Furthermore, there the President conceded that the remedial action was binding.  The power to take remedial action comes directly from the Constitution.[132]

Merits

[153] The facts have been comprehensively set out in the first judgment and there is no need to repeat them here, except to the extent of making this judgment understandable.  Soon after the coming into force of the Water Services Act[133] (Act) which gave effect to municipalities’ authority to supply potable water and sanitation services, Merafong gave notice to mining companies operating within its area to apply for approval for the supply of water.  In May 2004 Merafong imposed new tariffs which were higher than those of Rand Water which until then had been providing water services.

[154] The source of the power for Merafong to impose tariffs is section 229 of the Constitution.  This section sets out municipal fiscal powers and functions.  It provides:

“(1)      Subject to subsections (2), (3) and (4), a municipality may impose—

(a)        rates on property and surcharges on fees for services provided by or on behalf of the municipality; and

(b)        if authorised by national legislation, other taxes, levies and duties appropriate to local government or to the category of local government into which that municipality falls, but no municipality may impose income tax, value-added tax, general sales tax or customs duty.

(2)        The power of a municipality to impose rates on property, surcharges on fees for services provided by or on behalf of the municipality, or other taxes, levies or duties—

(a)        may not be exercised in a way that materially and unreasonably prejudices national economic policies, economic activities across municipal boundaries, or the national mobility of goods, services, capital or labour; and

(b)        may be regulated by national legislation.”

[155] It is apparent from this provision that the Constitution allocates to municipalities the power to impose property rates and surcharges on the fees for services, irrespective of whether those services were provided by the municipality itself or its agent.  Barring income tax, value-added tax, general sales tax and customs duty, a municipality may also impose other taxes and levies assigned to it by national legislation.  However the power to impose taxes and surcharges on fees for services may not be exercised—

“in a way that materially and unreasonably prejudices national economic policies, economic activities across municipal boundaries or the national mobility of goods, services, capital or labour.”

[156] The section prescribes conditions that apply to the exercise of municipal fiscal powers and functions.  It also empowers Parliament to pass legislation to regulate the exercise of the power by municipalities themselves.  The Act is such legislation.

[157] Although the Constitution limits the municipal competence on water services to potable water supply systems,[134] the Act empowers municipalities to supply water even for industrial use by prohibiting procurement of water from anybody else except a municipality duly designated under the Act as a water services authority or its nominee.[135]

[158] The Act goes further to regulate the power of municipalities to approve the supply of water for both domestic and industrial purposes.  It proclaims that a water services authority may not unreasonably withhold the approval and may give the approval subject to reasonable conditions.[136]

[159] However section 8 of the Act is not restricted to regulating the exercise of power by municipalities.  It goes further to subject decisions of municipalities on approvals to the Minister of Water Affairs.  The section achieves this by granting a person aggrieved by the municipality’s decision a right of appeal to the Minister who may confirm, vary or overturn a decision of the municipality.  The Minister is also empowered to prescribe the procedure “for conducting an appeal” under section 8 which must be lodged within 21 days of the appellant becoming aware of the impugned decision.  In addition the section authorises a relevant province to intervene as a party in the appeal lodged in terms of the Act.[137]

[160] Dissatisfied with the tariffs imposed by Merafong, AngloGold appealed to the Minister.  In her ruling the Minister upheld the appeal.  The Minister’s ruling, which is dated 18 July 2005, reads:

“1         Your Notice of Appeal dated 4 June 2004 lodged in terms of section 8(4) of the Water Services Act, 1997 has reference.

2          Decision

2.1        I differ with the conditions imposed by the Merafong City Local Municipality when approving the application.  I regard a tariff increase of 62% with no value added as unreasonable.  Water Services Authorities are required to exercise their rights in a manner that is fair, equitable and reasonable, and support national fiscal and economic policy.  Where a Water Services Authority adds no value to the services provided to a person or institution from another source it would be unreasonable to impose a fee, charge, surcharge or levy on the services provided.

2.2        Since water for industrial use is not defined as a municipal service in terms of section 1(xxv) of the Water Services Act, 1997 no surcharge can therefore be levied on water for industrial use.  Surcharges may only be levied on the portion of water that the mines are using for domestic purposes.

2.3        The Merafong City Local Municipality is of the view that it appropriately consulted with the mines and also considered the mines economic assessment presented by the Chamber of Mines on behalf of the mines.  Based on the appeal submitted to me, it is debatable whether the mines support the view that appropriate consultation has taken place.  When considering the merits of the appeals, I am not convinced that the Municipality has provided a reasonable opportunity for the mines to present themselves.

2.4        In terms of the power vested on me by section 8(9) of the Water Services Act, 1997, I therefore overturn the decision of the Merafong City Local Municipality to levy a surcharge on water for industrial use and rule that the Municipality, the Mines and Rand Water should negotiate a reasonable tariff on the portion of water that the mines are using for domestic purposes.”

[161] The Minister motivates her decision by asserting that the tariff increase of 62% was unreasonable and that where a municipality adds no value to the service provided by another person “it would be unreasonable to impose a fee, charge, surcharge or levy on the service provided”.  The Minister proceeds to express her opinion on circumstances under which surcharges may be levied.  She said since water used for industrial purposes is not defined in the Act as a municipal service, no surcharge can be levied for its supply.  A surcharge may only be levied, she concluded, on the portion of water that AngloGold was using for domestic purposes.

[162] The Minister’s reasons reveal the misconception of the legal position under which she laboured when she considered the appeal.  She completely overlooked the fiscal power conferred on municipalities by section 229(1) of the Constitution which expressly declares that a municipality is entitled to levy a surcharge on services provided by it and also those that are provided by a third party on its behalf.  Moreover, section 7 of the Act authorises the Municipality where the supply of water for industrial use is not done by itself, to nominate a service provider.  Here it is common cause that Rand Water was so nominated.

[163] Therefore on the Minister’s understanding, even where water for industrial use is supplied by the Municipality, it may not levy a surcharge for the service because that service is not defined as a municipal service in the Act.  The Minister was wrong on both fronts.  Section 229 authorises municipalities to levy a surcharge on services provided by it or on its behalf, regardless of whether there is value added or not.

[164] Taking the view that the Minister acted unconstitutionally, Merafong disregarded the ruling and insisted on the payment of the surcharge.  AngloGold paid under protest.  When resolution of the dispute eluded the parties, AngloGold approached the High Court for relief.  It sought an order directing Merafong to comply with the Minister’s decision of 18 July 2005.

[165] For its part Merafong opposed the relief and counter-applied that the decision be set aside.  In support of its counter-application Merafong advanced two grounds.  First, it contended that the Minister acted without power.  Second, in the event that the Court held that she had the authority, Merafong challenged the constitutionality of section 8 of the Act, to the extent that it authorised the Minister to interfere in the exercise of a municipal fiscal power.

[166] The Minister was cited as a party in the proceedings but this notwithstanding she declined to participate.  The stance taken by the Minister was indeed unusual.  Legislation she was given the responsibility to administer was being challenged and yet she chose not to get involved.  Despite the fact that it was her decision that was under attack.  This was not in line with the duty members of the Executive have towards courts when the validity of legislation they administer is impugned.  That duty arises even when a Minister concedes invalidity.  They are obliged to furnish the court with any information which may help it adjudicate the claim of constitutional invalidity.  It is through Ministers that information on impugned legislation passed by Parliament, reaches the courts.  An applicant does not have to cite Parliament every time legislation is impugned.  After all Ministers enjoy the right to introduce legislation in Parliament.

[167] Following its interpretation of the relevant provision, the High Court held that here the Minister exercised power duly conferred on her when she reached the impugned decision.  The Court proceeded to consider the constitutional challenge.

[168] The High Court accepted that Merafong exercised a municipal fiscal power contained in section 229 of the Constitution in imposing the surcharge.  But the Court concluded that the tariff increase was unreasonable and as a consequence, the Minister’s intervention was justified.

[169] In this regard the Court reasoned thus:

“With regard to water for domestic use, the authority of a municipality to impose surcharges on fees for the supply of water for domestic use provided by or on behalf of a municipality emanates from the Constitution itself.  This, as already stated, is provided for in section 229(1)(a) read with section 229(2) of the Constitution.  The parties are agreed that in terms of the Constitution a municipality has the original legislative and executive powers to impose tariffs and surcharges on fees for services provided by or on behalf of the municipality in respect of the supply of domestic water to members of the community within its area of Jurisdiction.  See Fedsure Life Assurance v Greater Johannesburg TMC [1998] ZACC 17; 1999 (1) SA 374 (CC) at para [39].

In this instance, it is common cause that the Municipality imposed tariffs and surcharges in respect of water supplied for domestic use to AngloGold.  It is also common cause that AngloGold was not satisfied with the Municipality’s decision to impose the surcharges on these services and appealed to the Minister in respect thereof.  When considering the appeal the Minister made a finding that where a water services authority adds no value to the services provided to a person or institution from another source it would be unreasonable to impose a fee, charge, surcharge or levy on the services provided, based on the appeal submitted to her, the Minister was not convinced that the Municipality provided a reasonable opportunity for the mines (including AngloGold) to present themselves.  This in my view resulted in an unreasonable tariff being set by the Municipality.  And as a result the Minister ruled that the Municipality, the Mines and Rand Water should negotiate a reasonable tariff on the portion of water that the mines are using for domestic purposes.”[138]

[170] Then the Court concluded:

“As already stated the power to impose the surcharges on fees for the supply of water for domestic use is subject to national legislation.  The [Act] is the regulatory framework by which the Minister assumes responsibility to regulate the exercise of the executive municipal power for water for domestic use.”[139]

[171] The fundamental flaw in this conclusion lies in the fact that the Court overlooked the principle of separation of powers that applies among the spheres of government.  This Court has proclaimed in a long line of cases that municipalities enjoy exclusive powers in relation to competencies allocated to them by the Constitution.[140]  In those cases this Court held that national and provincial spheres of government may not arrogate to themselves the power to exercise municipal competencies, by simply passing legislation authorising the exercise of municipal powers.

[172] Contrary to that authority, the High Court here concluded that it was constitutionally compliant for the impugned provision to empower the Minister to intrude into the terrain of Merafong and overturn a decision taken by it in the exercise of its municipal fiscal power.  On the authority of the Constitution as interpreted by this Court in the cases referred to, this was incorrect.  Instead the constitutional challenge should have been upheld.

[173] The fact that this claim may not have been pursued in the Supreme Court of Appeal does not bar this Court from adjudicating it.  Indeed the Minister and Merafong in the written submissions filed in response to our directions, agree that it is competent for this Court to consider this claim.  However, the Minister disputes that in overturning Merafong’s tariff in terms of section 8(9), she exercised Merafong’s exclusive power.  There is no merit in this argument.  Municipal fiscal power falls within the exclusive domain of municipalities and may be exercised by them and by them alone, subject to conditions listed in section 229 of the Constitution.

[174] The first judgment holds that Merafong did not persist with the constitutional attack before us.[141]  This is incorrect.  In its written submissions in response to our directions, Merafong argued:

“[S]ince the Minister has no jurisdictional competence to consider an appeal against the imposition of a municipal tariff on potable water for domestic consumption, she is debarred from entering this terrain for any purpose . . .

Since, as a matter of jurisdiction, she cannot entertain the matter, she cannot ‘confirm, vary or overturn’ the municipal decision.  So much is rightly conceded on behalf of the Minister too.  By parity of reasoning, it must follow that she cannot give a decision. . . .  We submit, therefore, that the constitutional issue can properly be dealt with . . .  moreover that it should be dealt with.”

[175] AngloGold concedes that it is competent for this Court to determine a constitutional challenge such as the one mounted by Merafong but argues that in the present circumstances it is not necessary to reach it.  The reasons advanced for this proposition are: that the Minister has not dealt with the attack “as an impermissible intrusion on the powers of local government”; Merafong’s failure to pursue the challenge in the Supreme Court of Appeal denied this Court the benefit of having the views of that Court; the invalidity claim amounted to a collateral challenge which may not be advanced by Merafong as an organ of state and that the constitutional challenge was nothing else but a disguised review, instituted out of time.

[176] The argument lacks substance.  I fail to appreciate how an invalidity challenge directed at a statutory provision on the ground that it is inconsistent with the Constitution could be both a review envisaged in PAJA and a collateral challenge.  A collateral challenge is not subject to any time bar but a PAJA review is.  Moreover, a constitutional attack against the validity of legislation does not constitute administrative action contemplated in PAJA.  Accordingly PAJA does not apply to such claims.  With regard to the proposition that a collateral challenge is not available to an organ of state, I have already held that it may competently be raised by the state.

[177] As it cannot be gainsaid that the section impermissibly empowers the Minister to exercise a municipal power, it follows that the section is inconsistent with the Constitution.  Accordingly this Court is obliged to declare it invalid to the extent of the inconsistency.

[178] I would have granted leave, upheld the appeal and declared section 8 of the Water Services Act to be inconsistent with the Constitution and set it aside.

For the Applicant: M S M Brassey SC, E S J Van Graan SC and J A Motepe instructed by De Swardt Vogel Myambo Attorneys

For the Respondent: N J Graves SC and I B Currie instructed by Knowles Husain Lindsay Inc

---

---