IN THE HIGH COURT OF SOUTH AFRICA

(WESTERN CAPE DIVISION, CAPE TOWN)

CASE NO: 19976/2019

In the matter between:

 

SECHABA PROTECTION SERVICES CC (PTY) LTD   HIGH GOALS INVESTMENTS cc t/a CHUMA SECURITY SERVICES   SUPREME SECURITY SERVICES CC   vusa-isizwe security (pty) ltd	First Applicant   Second Applicant   Third Applicant   Fourth Applicant
and   PASSENGER RAIL AGENCY OF SA LTD   BONGISIZWE MPONDO   #UNITEDBEHIND	First Respondent   Second Respondent   Amicus Curiae

Coram: Bishop, AJ

 

Dates of Hearing: 19 October 2023

 

Date of Judgment: 3 November 2023

 

 

 

 

JUDGMENT

 

 

 

BISHOP, AJ

 

[1] Like the railways that crisscross the country, this case has a long history that has engaged at least seven judges of this Division. In what is hopefully the penultimate stop on its long journey, I was initially asked to decide whether the First Respondent (PRASA) had complied with an earlier supervisory order of this Court. That Order required PRASA to continue employing the Applicants to provide security services until it fulfilled certain conditions. PRASA approached this Court this year claiming the conditions were met, supervision should end, and it should be permitted to terminate the Applicants’ contracts. PRASA ultimately accepted that those conditions had not been met. Accordingly, the parties agreed that some form of ongoing supervision was required to see the case to its final destination. This judgment plots that route.

 

The Background: 2005 - 2019

[2] While this litigation only departed the station in 2019, the need for its voyage was sparked by an earlier case decided by this Court in 2003. In 2001, a group of concerned Cape Town rail commuters launched an application to compel the state entities responsible for urban commuter railways to ensure that commuters were safe when travelling by train. The litigation followed increasing incidents of violent crime on Cape Town’s trains. This Court – in a judgment of Davis and Van Heerden JJ –declared that Metrorail and the South African Rail Commuter Corporation Ltd (SARCC) had a duty to ensure commuter safety. The Court found that the Respondents had failed to fulfil that duty and directed them to take the necessary steps to address the failure.1 On appeal, the SCA produced three judgments, but effectively upheld the appeal and set aside the High Court’s remedy.2

[3] The case made its way north to Braamfontein. In a 2004 judgment of O’Regan J,3 the Constitutional Court unanimously concluded that Metrorail and SARCC “have an obligation to ensure that reasonable measures are taken to provide for the security of rail commuters whilst they are making use of rail transport services provided and ensured by, respectively, the first and second respondents.”4 It granted no further relief. That duty arose from the South African Transport Services Act 9 of 1989, read with the constitutional rights to dignity, to life, and to be free from public and private violence, together with the constitutional value of accountability. O’Regan J reasoned:

Institutions which are organs of state, performing public functions and providing a public service of this kind, should be held accountable for the provision of that service. It is for this reason that the Constitution affirms accountability as a value governing public administration. Metrorail has the obligation to provide rail commuter services in a way that is consistent with the constitutional rights of commuters. In the absence of a public law obligation of the kind contended for by the applicants, there is no way of ensuring that Metrorail complies with this duty.5

[4] The Court emphasized that “[i]t does not matter who provides the measures as long as they are in place.”6 But the “responsibility for ensuring that measures are in place” rested with Metrorail and SARCC.7

[5] Fast forward to 2019. SARCC had been effectively replaced by PRASA. PRASA relied on several companies in the Western Cape to assist it to fulfil its constitutional duty to take reasonable measures to ensure the safety of its commuters. Those companies included the Applicants who were initially appointed to provide security services in 2011 for one year. That year came and went, but the Applicants were kept in place through regular month-to-month extensions of their contracts. For nearly seven years.

[6] But in 2018 PRASA decided it was time for a change. It began reviewing its security operations. In April 2019, it issued a new tender for security services (the 2019 Tender). All the Applicants submitted bids. PRASA planned to award the 2019 Tender in July. But there were delays.

[7] While the 2019 Tender remained unawarded, PRASA continued to use the Applicants’ services. On 1 October 2019, it had extended the contracts for another month until 31 October 2019. As their contracts had been regularly extended month-to-month since 2012, the Applicants quite reasonably expected that they would continue to be renewed until the 2019 Tender was finalised.

[8] Instead, on 1 November 2019, PRASA called the Applicants to instruct them to cease providing services from midnight that day. The Applicants were, understandably, bemused. They sought clarification or reconsideration. But PRASA was unmoved – the contracts were over, and the Applicants were ordered not to come back to work.

[9] The Applicants refused. They continued to send their employees to work because the 2019 Tender had not yet been awarded and no other plan existed to replace them with new service providers. While PRASA had accumulated significant inhouse security resources (I return to the details later) it could not fulfil its constitutional duty without the assistance of external security companies.

[10] The First to Third Applicants (the Sechaba Applicants) approached this Court to address the impasse. On 8 November 2019, they launched an urgent application for an order that would, in effect, allow them to continue to render services until the 2019 Tender was completed, or PRASA had reported to the Court on affidavit to explain how it would “secure the railway system … in the absence of utilising the services of the applicants.”

[11] This application was naturally brought in the Sechaba Applicants’ own interest. It would allow them to continue earning money for the provision of their services. But it was also expressly brought in the public interest. The Sechaba Applicants relied on the Constitutional Court’s judgment in Rail Commuters Action Group and explained that “it is in the public interest that those utilising the trains do so in a safe environment”. Without their services – they argued – PRASA could not meet its obligation to protect its commuters.

[12] PRASA opposed, largely on contractual grounds. It argued that it was entitled to cancel the Sechaba Applicants’ contracts, and that the Court had no business preventing it from doing so.

[13] A civil society organization, #UniteBehind, intervened in the application. #UniteBehind is a coalition of various organisations. One of its objectives is to fix the ailing commuter rail services. It supported the Sechaba Applicants’ basic proposition that, without their services, PRASA could not ensure the safety of rail commuters. But it contended for different relief. It argued that the Court should grant a structural interdict that would allow it retain supervision to ensure the Sechaba Applicants were not removed until adequate safety measures were in place. It also argued that PRASA should be required to continue to use their services not only until the 2019 Tender was complete, but until “an adequate contingency safety plan approved by the Railway Safety Regulator” was submitted to the Court, and the Court was “satisfied with the adequacy of the plan”.

[14] The matter was heard by Hlophe JP on 19 November 2019. He gave an order on the day that it seems was an amalgam of the Sechaba Applicants’ and #UniteBehind’s proposed orders. It is central to the present dispute. The relevant parts read:

3. That, pending the completion and implementation of [the 2019 Tender] and/or an adequate contingency plan (as referred to in paragraph 4 below), [PRASA] shall continue to utilise the services of the applicants, on the same terms and conditions, as contracted for previously.

4. Within 30 days the respondent shall report to this Honourable Court, on affidavit –

4.1. the status of the completion and implementation (sic) of [the 2019 Tender]

4.2. to present an adequate contingency safety plan approved by the Railway Safety Regulator, which is to include the following:

4.2.1. A description of the preventative and response measures to be used to manage theft, vandalism, assault, sexual assault and other criminal acts or other sources of harm;

4.2.2. A description of the measures to ensure the protection of commuters, PRASA employees and assets;

4.2.3. The allocation of security roles (including the South African Police and Metro Services) and responsibilities to appropriate personal (sic).

4.3. That should the respondent fail to adhere to the time constraints provided, that an affidavit be filed with the Court and the parties to this matter, explaining the reason for the non-compliance.

4.4. That should this court be satisfied with the adequacy of the plan referred to above, such plan be implemented forthwith, but after the applicants have been provided a reasonable notice period to be determined by this court.

5. That, to the extent necessary, after receiving the affidavit(s) referred to in paragraph (4) above, that the parties are granted leave to approach this Honourable Court on the same papers, duly supplemented, for further appropriate relief, as may be necessary in the circumstances.

[15] Despite a request from PRASA, Hlophe JP did not provide reasons for his order. It therefore needs to be interpreted on its own terms, and in light of the evidence and argument that served before the Judge President.

[16] It is worth, at this stage, identifying five key elements of this order, as the dispute before me is (or more accurately, was) whether PRASA had met the conditions to permit it to terminate the Applicants’ services.

[17] First, the primary operative paragraph is paragraph 3. It requires PRASA to continue to use the Sechaba Applicants’ services until it completes and implements (a) the 2019 Tender; “and/or” (b) “an adequate contingency safety plan”.

[18] Second, what constitutes “an adequate contingency safety plan” is effectively defined by paragraph 4. It must be one that: (a) is “approved by the Railway Safety Regulator”; (b) includes the requirements in paragraphs 4.2.1 to 4.2.3; and (c) that this Court is satisfied is adequate.

[19] Third, to establish that: (a) it had completed and implemented the 2019 Tender; (b) the Railway Safety Regulator had approved the safety plan; and (c) (presumably) to satisfy this Court the plan was adequate, PRASA was required to file an affidavit within 30 days.

[20] Fourth, if it filed the affidavit late, PRASA had to explain why.

[21] Fifth, the parties were given leave to approach this Court for further relief.

[22] Given that PRASA had been working on reviewing its safety plan from 2018, and that the 2019 Tender had been issued in April 2019, the parties expected that PRASA would complete the tender and get its safety plan approved within a matter of months. None of the parties envisioned that, four years later, the Applicants would still be working for PRASA under the 2011 agreement.

 

The Background: 2020 to 2021

[23] There were regular and bitter skirmishes between the parties after the Hlophe Order was granted. It is not necessary to delve into all of them. But a general account is necessary to appreciate the current position, and to understand some of the relief that is now sought.

[24] The first important development is that the Fourth Applicant (Vusa Isizwe) intervened as an applicant in accordance with an order granted by Savage J on 21 January 2020. The result of her order was to extend the Hlophe JP Order to apply to it as well as the original Sechaba Applicants.

[25] PRASA partially complied with the Hlophe JP Order in the sense that it allowed the Applicants to continue to perform their services. It did not, however, pay them for those services. Nor did it file the affidavit required by paragraph 4 of the Hlophe JP Order.

[26] This led to two orders granted by Ndita J on 6 March 2020. The first directed PRASA to pay the Sechaba Applicants’ unpaid invoices. It also required PRASA to explain its failure to comply with Hlophe JP’s Order, to specify what further steps should be taken, and to show cause why it should not be found to have not complied with the Hlophe JP Order, or be found in contempt. The second order permitted Vusa Isizwe to intervene in the Sechaba Applicants’ enforcement proceedings and obliged PRASA to pay Vusa Isizwe its outstanding invoices.

[27] PRASA paid the Sechaba Applicants as required by Ndita J, but not Vusa Isizwe. It again failed to pay later invoices for all the Applicants. The Applicants therefore went back to court and obtained an order from Gamble J on 5 June 2020. His order – like Ndita J’s first order – not only required payment, but also set a return date for PRASA to establish why it should not be held in contempt of court, and why it should not be declared to have breached ss 165(5), 195(1)(f) and/or 237 of the Constitution.

[28] In the meantime, PRASA tried to launch a new tender process for security services. The Applicants applied to stay the tender pending the satisfaction of the Hlophe JP order. PRASA resisted the application on the basis that the Hlophe JP Order only applied in the Western Cape. The application came before Le Grange J. On 30 June 2020, he ordered that “the order granted by the Judge President on 19 November 2019 has universal application and is therefore not only applicable to the Western Cape”. He also temporarily stayed the new tender process, but ordered a return day to allow PRASA to explain why the stay should not be extended until the Hlophe JP Order was satisfied.

[29] The return dates of the Gamble J Order, and the Le Grange J Order were set down for hearing on 24 August 2020.

[30] A month before that hearing, on 20 July 2020, PRASA finally filed an affidavit providing the updates required by paragraph 4 of the Hlophe JP Order. The affidavit was deposed to by Jonas Rakau, PRASA’s Acting Head: Group Security. It provides a detailed account of the steps that PRASA had taken to ensure the security of its assets and passengers. It also sought to explain the delay in the filing of the affidavit. In short, the delay was caused by the Minister dissolving PRASA’s interim board and appointing an administrator in its place (the Second Respondent), and the impact of the Covid-19 lockdown.

[31] PRASA reported that, shortly after the Hlophe JP Order was granted, it cancelled the 2019 Tender. It had also prepared a new Annual Safety Improvement Plan and a Security Plan in terms of the National Railway Safety Regulator Act 16 of 2002, and anticipated that the Regulator would issue it with a safety permit pursuant to those plans. Finally, it had halted the emergency procurement that triggered the Le Grange J Order.

[32] The bottom line was that, while PRASA had done substantial planning work to improve its security, it had not found a way to replace the Applicants’ services. It had not completed and implemented the 2019 Tender, or any other tender. It had not had a safety plan approved by the Regulator, as the Hlophe JP Order required, which could satisfy this Court that it no longer required the Applicants’ services. The Rakau affidavit does not contend that it had.

[33] That brings us back to the return date for the Gamble J and Le Grange J Orders of 24 August 2020. That came before Hack AJ. He delivered judgment on 9 February 2021 and ultimately decided four issues. He ordered PRASA to pay the Applicants in terms of invoices issued to them. He finally interdicted the emergency procurement until the “final determination” of the application – a reference to the conclusion of the supervisory relief under the Hlophe JP Order. Because of its multiple failures to comply with court orders, he ordered PRASA to pay the Applicants’ various costs on an attorney and client scale. Lastly, he declined to order the Second Respondent – PRASA’s administrator – to pay costs personally.

[34] But Hack AJ did not find PRASA in contempt, did not declare that it had failed to comply with any court order, and did not declare that it had breached any constitutional obligation. His judgment did not explain why he did not make those orders, that were expressly before him as a result of the Gamble J Order.

[35] That is where matters stood in early 2021. Most of this is now water under the bridge. But it provides the deeper context to what followed and is part of the explanation for the order I make.

 

The Present Application

[36] Things then went quiet. In the 28 months from the delivery of Hack AJ’s judgment on 9 February 2021 until 19 June 2023, nothing happened. Then PRASA filed an application for the “discharge” of the Hlophe JP Order. Its argument was that it had demonstrated compliance, or imminent compliance, with the Hlophe JP Order and that it should therefore be allowed to terminate the Applicants’ services on 60 days’ notice.

[37] PRASA contended that it had satisfied the requirement to obtain the approval from the Regulator. It had, from July 2020, been regularly granted a safety permit by the Regulator. In addition, on 25 January 2023, it had written to the Regulator asking it to “approve” its safety plan as required by the Hlophe JP Order. The letter – addressed to the Regulator’s CEO – notes that the Regulator is not ordinarily required to approve a safety plan, and that the request was made solely to comply with the Hlophe JP Order. Mr Kgomari, an official of the Regulator, responded on 24 February 2023 stating: “The Regulator has received various submissions demonstrating compliance with the requirements of the ruling and has satisfied itself that PRASA has developed and implemented adequate security management plans for the Western Cape operations.” Because it had obtained the Regulator’s approval, PRASA argued, it had complied with paragraphs 3 and 4.2 of the Hlophe JP Order.

[38] With regard to actually replacing the Applicants’ services, PRASA admitted that it had not completed the 2019 Tender. But it claimed that “soon, after the delivery” of its application, it would “commence with an open and competitive procurement process” (the 2023 Tender). It estimated it would take 60-90 days (by mid-September) to complete that tender process. The 2023 Tender was in fact initiated as RFP HO/SEC/002/05/2023 on 14 July 2023 – a month after the application was launched, but prior to the Applicants filing their answering affidavits.

[39] In the gap between termination of the Applicants’ services, and completion of the 2023 Tender, PRASA indicated that it would make use of security service providers that formed part of the panel established by the Airport Company of South Africa (the ACSA Panel). As PRASA put it – companies appointed from the ACSA Panel would “fill the void left by” the Applicants. PRASA would also rely on a further 3 100 security personnel it had employed, 698 of which had been deployed to the Western Cape. These employees worked in conjunction with the 887 security personnel provided by the Applicants, and 1 344 Metropolitan Protection Services officials. For all the insourced security personnel, they were not sufficient to meet PRASA’s needs. That is precisely why it still sought to tender for external assistance.

[40] In short, PRASA argued that it had complied with the requirement for Regulatory approval and would shortly replace the Applicants through the 2023 Tender and had mechanisms to ensure safety in the interim. Therefore, there was no need to compel it to continue to use the Applicants’ services, or for ongoing judicial supervision.

[41] The Applicants, unsurprisingly, opposed. They argued that, properly interpreted, the Hlophe JP Order required PRASA to demonstrate both that it had implemented the 2019 Tender and obtained the Regulator’s approval for its safety plan. They differed slightly on precisely what the Order meant. I address those differences below.

[42] However, they acknowledged that it was no longer possible for PRASA to implement the 2019 Tender, and that PRASA could instead seek to implement a new tender to comply with the Hlophe JP Order. But PRASA, they argued, had approached the Court far too early. It had come to court before it even issued the RFP, and would likely be heard before the 2023 Tender would be completed and implemented. Vusa Isizwe and the Royal Sechaba Applicants therefore launched counter-applications to amend the Hlophe JP Order to refer to the 2023 Tender, but still maintaining court supervision until PRASA had in fact completed and implemented that tender.

[43] The Applicants also disputed that PRASA had obtained the Regulator’s approval for its safety plan. Vusa Isizwe questioned whether the letter from Mr Kgomari could indeed be accepted as an approval by the Regulator itself. Approval from the Regulator, it argued, had to come from its Board or a properly delegated official, not from an employee without clear delegation. It also argued that the Hlophe JP Order, read with the Le Grange J Order, required the Regulator’s approval of a national safety plan, not merely one for the Western Cape.

[44] The Applicants also took issue with PRASA’s attempt to rely on the ACSA Panel. Vusa Isizwe argued that PRASA was not permitted by regulation 16A.6 of National Treasury’s Regulations8 to the Public Finance Management Act.9 The Sechaba Applicants questioned PRASA’s power to use the ACSA Panel, and pointed out that the Panel would terminate on 30 September 2023, whereas PRASA’s application to discharge the Hlophe JP Order would only be heard in October. How then, they asked, could the ACSA Panel serve as a stopgap pending the 2023 Tender?

[45] The Sechaba Applicants also brought an application to review and set aside the award of a security tender to Mzansi Securifire Group (Pty) Ltd. The tender, which was to assist PRASA to recover the Central Line in Cape Town, was awarded on 26 October 2022 and was to run until September 2023 (although I was informed in October that it was still in operation). The Sechaba Applicants argued that this was in violation of the Hlophe JP Order, and therefore unlawful. They also, unusually, included a specific prayer to compel PRASA to provide it with a Rule 53 record for its decision. PRASA countered that the appointment of Mzansi Securifire was an ad hoc tender, for a specific service, and therefore did not violate the Hlophe JP Order. Before me, the only issue was whether PRASA should be ordered to provide the Rule 53 record.

[46] Finally, the Sechaba Applicants sought an order declaring that PRASA had “breached its obligations in terms of sections 165; 195 and 237 of the Constitution”. They argued that PRASA had failed to comply with the Hlophe JP Order, by failing to pay them, by failing to report on time, by failing to take any further steps to allow it to terminate the Applicants’ services, and then by seeking the discharge of the Order prematurely.

[47] Those, then, were the basic issues in dispute when the matter came before me.

[48] For reasons that will become apparent, it is necessary, to set out the order in which papers were exchanged, and what occurred at, and after, the hearing of the application.

[49] After Vusa Isizwe had filed its answering affidavit and counter-application, but before the Sechaba Applicants had done so, the parties arranged a case management meeting before Le Grange ADJP (as he now is). He issued an order setting the matter down for hearing on 19 October 2023, and regulating the filing of further papers. That order provided for Vusa Isizwe to file a reply to PRASA’s answer to its counter-application. As the Sechaba Applicants had not yet launched a counter‑application, it did not provide a similar opportunity for them.

[50] When they filed their answer, the Sechaba Applicants also filed a counter-application (seeking amendments to the Hlophe JP Order, a review of the appointment of Mzansi Securifire, and a declaration of unconstitutional conduct). PRASA filed separate replying affidavits in its application, which were also answering affidavits in the counter applications. Both Vusa Isizwe and the Sechaba Applicants filed replying affidavits.

[51] Two days before the hearing of the application, PRASA served an application to strike out most of Vusa Isizwe’s replying affidavit. It complained that Vusa Isizwe had included matter that went beyond the permissible scope of a reply. On the day of the hearing, PRASA launched an application to strike out the Sechaba Applicants’ whole replying affidavit. It argued that they were not entitled to file a reply at all as it was not provided for in Le Grange ADJP’s order. Both applications were only provided to me in chambers just before the hearing.

[52] On the morning of the hearing, PRASA sought to introduce a supplementary affidavit setting out the status of the 2023 Tender. This was not a surprise as it had previously indicated its intention to do so. The bottom-line of the affidavit is that PRASA’s Corporate Bid Adjudication Committee required further information to make a decision. It anticipated it would receive that information on 20 October 2023, and finally adjudicate the tender on 21 October 2023. However, it would still be necessary to obtain further, unspecified approvals, which PRASA anticipated would occur by 31 October 2023.

[53] However, the affidavit did not indicate by when PRASA anticipated concluding contracts with the successful bidders, nor by when those bidders would be in a position to “put boots on the ground”. I put this difficulty to PRASA’s counsel, Mr Jacobs SC, at the hearing. As the point of the Hlophe JP Order was to ensure there was no gap in the provision of services, it was vital to know not only when the 2023 Tender would likely be awarded, but also when it would be implemented. He accepted that the Court needed this information, and took an instruction from his client on how to address the Court’s difficulty.

[54] The outcome of that instruction was that PRASA would agree to an order that would require it to report to the Court on when the 2023 Tender would in fact be implemented, and that it would not terminate the Applicants’ services until that date was known. I also asked Mr Jacobs whether his client would have any objection to obtaining additional confirmation from the Regulator that it had in fact approved PRASA’s safety plans. He undertook to obtain that instruction.

[55] In light of the position adopted by PRASA, I indicated that I intended to retain the file, and invited the parties to submit either an agreed draft order, or separate draft orders, for how to manage the further supervision of the application. Evidently the parties could not agree. PRASA and the Applicants submitted two separate draft orders a few days after the hearing. I deal with the content of those draft orders when addressing the just and equitable remedy.

 

The Issues

[56] Against that extensive backdrop, I was required to consider the following primary issues:

[56.1.] Does the Hlophe JP Order require PRASA to complete and implement the tender, or obtain the Regulator’s approval of a safety plan, or both?

[56.2.] Did PRASA adequately establish that the Regulator had approved its safety plan?

[56.3.] At the time the application was heard, had PRASA completed and implemented a tender process?

[56.4.] What order should the Court make to regulate and realise the resolution of the matter?

[57] In addition, the following secondary issues arose:

[57.1.] Should PRASA’s striking out applications be granted?

[57.2.] Could PRASA make use of ACSA’s Panel?

[57.3.] Should an order be made to compel PRASA to file a Rule 53 Record in the Sechaba Applicants’ review of the appointment Mzansi Securifire?

[57.4.] Should the Court make an order that PRASA had failed to comply with its constitutional obligations?

[58] I intend to address the issues in this order:

[58.1.] The striking out applications;

[58.2.] The and/or issue;

[58.3.] Whether the Regulator has approved PRASA’s safety plan;

[58.4.] Whether PRASA has completed and implemented a tender;

[58.5.] The just and equitable remedy;

[58.6.] The ACSA Panel and the Mzansi Fire issues;

[58.7.] Whether PRASA has failed to fulfil its constitutional obligations; and

[58.8.] Finally, costs.

 

The Striking Out Applications

[59] On the day of the hearing, PRASA’s counsel arrived with two striking out applications. As I mentioned earlier, the first was directed at certain sections of Vusa-Isizwe’s replying affidavit which, PRASA claimed, strayed beyond the permissible bounds of reply. The second was directed at the whole of the Sechaba Applicants’ replying affidavit on the basis that, when regulating the further conduct of the matter, Le Grange J did not make provision for them to file a reply.

[60] After hearing counsel, I dismissed both applications with costs, including the costs of two counsel. These are my reasons.

[61] Applications to strike out matter from affidavits are regulated by rule 6(15). It allows a party, at any time, to apply to strike out matter in an affidavit on a variety of grounds, including that it constitutes new matter in reply.10 But the “court may not grant the application unless it is satisfied that the applicant will be prejudiced if the application is not granted.” An applicant seeking to strike out matter in an affidavit must show both that the averment is liable to be struck out, and that it will suffer prejudice if the averment is not struck.

[62] PRASA’s two applications must fail, primarily, because they do not allege or identify what prejudice PRASA suffers. They were brought solely on notice, without a supporting affidavit. I was referred to authority that, because rule 6(15) requires a showing of prejudice, “a founding affidavit will ordinarily be required” in an application to strike out.11 This seems sound practice to me. While there may be cases where prejudice will be apparent without evidence, in most cases evidence will be necessary to establish prejudice. It also allows the other party an opportunity to address and attempt to rebut the claim of prejudice.

[63] Even if it is permissible in some cases to seek to strike out under rule 6(15) without an affidavit establishing prejudice, this is not such a case. I asked Mr Jacobs what prejudice his client would suffer if the matter was not struck out. He struggled to identify any. Prejudice was not immediately apparent from the notice. The matter that PRASA sought to strike out in Vusa Isizwe’s affidavit was primarily legal argument that could be, and was in fact, dealt with in heads of argument and counsel’s oral submissions. While it may not have strictly belonged in a replying affidavit, it was not factual material that prejudiced PRASA’s ability to make its case.

[64] The problem with the attempt to strike out the Sechaba Applicants’ replying affidavit runs deeper. PRASA sought to strike out the entire affidavit on the basis that there was no provision for it in Le Grange ADJP’s order. That is so, but there are two difficulties. One, PRASA was still required to establish prejudice. It did not do so. Two, the application was not, in truth, a striking out application. It was a claim that the Sechaba Applicants’ reply constituted an irregular step. I have my doubts whether it was an irregular step – they had brought their own application, which was answered, and they were entitled to reply without express permission from Le Grange ADJP. But even if it was, given the nature of its objection, PRASA should have followed the process under Rule 30 or 30A to object to its filing.

[65] There was a further unsatisfactory element of both applications – they were only provided to the court on the morning of the hearing. The Vusa Isizwe application was served two days before the hearing, and the Sechaba Applicants application only the morning of the hearing. While it is so, as Mr Jacobs argued, that striking out applications must be set down for hearing at the same time as the main application, it does not follow that they can or should be provided to the Court on the same day. The applications to strike out should have been launched a reasonable time after the replying affidavits were received. Vusa Isizwe’s was filed on 26 September 2023, and the Sechaba Applicants’ on 9 October 2023. At that stage PRASA knew the application would be heard on 19 October 2023. Yet it waited until 17 and 19 October 2023. As there was no affidavit, there was no explanation for the delay, which was obviously prejudicial to the Applicants. It also placed the Court in an invidious position. I would have been tempted to dismiss the applications for this reason alone. But as the late applications were, in any event, without merit, I need not decide that issue.

 

“and/or”

[66] One of the key debates between the parties on the papers was whether the Hlophe JP order required PRASA to complete and implement the tender and obtain the Regulator’s approval for its contingency safety plan, or whether it could do just one of the two, before it was permitted to terminate the Applicants’ contracts. PRASA argued it need only complete one of the two requirements to satisfy the Hlophe J order. The Applicants contended that PRASA was required to do both.

[67] The debate fizzled out somewhat at the hearing, because PRASA’s counsel agreed that the Court should retain supervision until it had completed the new tender process, and in its draft order undertook to obtain proof from the Regulator that it had approved the safety plan.

[68] Nonetheless, it seems to me that the Applicants were correct in their reading of the Hlophe JP Order. Orders, like other legal documents, must be interpreted purposively.12 Despite request, Hlophe JP chose not to give reasons for his order, so the purpose can only be ascertained from considering the papers that served before him. Those reveal that the primary purpose of that application was to ensure that PRASA’s passengers and infrastructure were protected.

[69] Would the order achieve that purpose if it required only the implementation of a tender or the approval of a safety plan? As Vusa Isizwe pointed out, that depends on the safety plan. If the safety plan provided for PRASA to insource 100% of its safety needs, then it may not be necessary to complete the tender. If outsourcing remained part of PRASA’s plan, then the Order’s purpose could only be achieved if the tender to replace the Applicants was complete.

[70] These hypotheticals reflect reality, both then and now. Mr Nacerodien explained that it was never contemplated when this matter was before Hlophe JP that PRASA would not continue to outsource at least some of its security needs – that is why the tender is mentioned in the order. Today, PRASA still plans to outsource work through the 2023 Tender. So both when the order was granted, and now when its discharge is sought, PRASA accepts it cannot ensure safety without completing a tender to outsource some of its security needs.

[71] Accordingly, even if the parties had not agreed that both an implemented tender, and an approved safety plan are required, I would have reached that conclusion. In my view, it is not an error in the Hlophe JP Order that requires correction. Rather, the Order, needs to be properly interpreted in light of its context and purpose. It could not mean that PRASA was entitled to cease using the Applicants’ services when it did not have a plan in place to replace them completely either with insourced security, or new external service providers.

 

The Regulator’s Approval

[72] Did PRASA establish that the Regulator had approved its safety plan? PRASA rightly pointed out that this was not ordinarily part of the Regulator’s statutory role. The Regulator is governed by the Safety Act which provides for the Regulator to issue safety permits and monitor safety compliance. But the requirement for the Regulator’s approval of a contingency safety plan comes from the Hlophe JP Order, not the Safety Act. That is so. But it remains a requirement.

[73] PRASA initially contended that it had met the requirement because it had written to the Regulator requesting approval in terms of the Hlophe JP Order, and the Regulator had responded indicating its approval. Mr Mahenye of PRASA wrote to the CEO of the Regulator on 25 January 2023, asking it to approve PRASA’s Security Plan. The letter made clear that the request was made to comply with the Hlophe JP Order. It also indicated that the Regulator’s officials had informed PRASA the request had to be made to the CEO.

[74] But the CEO did not respond. Instead, Mr Kgomari, the Regulator’s Acting Head: Safety Permit Administration, responded on 24 February 2023. I quoted his response earlier. He indicated that the Regulator was “satisfied with the adequacy and effectiveness” of PRASA’s safety plans. On its face, this appears to constitute approval by the Regulator as required by the Hlophe JP order.

[75] The Applicants did not contend that the Regulator was wrong to be so satisfied. They did not attack the substance of PRASA’s safety plans.

[76] Instead, Vusa Isizwe raised two objections. First, it pointed out that Mr Kgomari could not speak on behalf of the Regulator. Under the Safety Act, the Regulator is “governed and controlled by a board of directors”.13 When Hlophe JP required the approval of the Regulator, he required the approval of the Board. While the Board can delegate and assign powers,14 there was no evidence that it had delegated or assigned Mr Kgomari to approve PRASA’s safety plan. Vusa Isizwe raised this complaint squarely in its answering affidavit and invited PRASA to put up a resolution of the Board, under cover of an affidavit, to establish that the Regulator had in fact approved the Safety Plan. PRASA did not take up that invitation. Instead, it contended that the letter from Mr Kgomari was adequate proof that the Regulator had approved the safety plan.

[77] In my view, Mr Kgomari’s letter would ordinarily be sufficient. But having been challenged to confirm his authority, PRASA ought to have done so. PRASA itself stated in its letter that Mr Kgomari had advised that the letter should be addressed to the CEO. Vusa Isizwe’s request was not unreasonable. If indeed the Regulator had approved the Safety Plan – either because the Board had itself approved it, or had properly delegated the power – it would have been a simple task for PRASA to obtain confirmation. It had several months to do so. The failure to meet the challenge leaves a degree of uncertainty.

[78] If these were ordinary motion proceedings, I would probably still have accepted PRASA’s version. But these are not ordinary proceedings. The Court is exercising its supervisory jurisdiction in order to ensure the protection and fulfilment of constitutional rights for railway commuters. It needs to be satisfied that the requirements set by Hlophe JP have in fact been met. The unusual nature of this application require PRASA to meet a higher than ordinary standard of proof. In light of the challenge posed by Vusa Isizwe, PRASA has not met that higher burden.

[79] PRASA accepted, in its proposed draft order, that it would provide confirmation that the Regulator – that is, the Board of the Regulator – had approved its safety plan. My order reflects that.

[80] Vusa Isizwe’s second objection was that the Hlophe JP Order required approval of a national safety plan, not merely a safety plan for the Western Cape. That is so, Mr Solomon argued, because Le Grange J had held that the Hlophe JP Order had national application beyond the Western Cape’s borders. Mr Kgomari’s letter reflected approval only insofar as PRASA’s safety plan concerned the Western Cape.

[81] I do not accept that the Hlophe JP Order required the Regulator to approve a national safety plan. While there was some mention of the position in other provinces, the evidence before Hlophe JP was limited primarily to the Western Cape because the Sechaba Applicants operated only in the Western Cape. The evidence introduced by #UniteBehind also related almost exclusively to the security issues in the Western Cape.

[82] Le Grange J’s order that the Hlophe JP Order had national application must be understood in context. The context was given by a later judgment of a Full Bench of this Court that included Le Grange J (sitting with Samela and Francis JJ) in a different case15 in which security service providers operating in Gauteng sought to take advantage of the Hlophe JP Order. That Full Bench held that “the universality, or the nationwide applicability, of the order is confined to the very simple proposition that PRASA’s procurement processes apply nationally.” Those service providers – which never provided security guards in the Western Cape – were not entitled to the protection of Hlophe JP’s order merely because PRASA tendered on a national basis. That being the case, Mr Jacobs argued, there was no need for the Regulator to approve a national safety plan.

[83] I agree. What the Hlophe JP Order requires is the approval of a “contingency safety plan”. That plan was meant to be a basis on which to permit PRASA to cease using the Applicants’ services. There would be no reason to oblige PRASA to continue using service providers in the Western Cape when it had obtained the Regulator’s approval for a safety plan in this province, merely because it had not obtained similar approval in other provinces. The fulcrum of the Hlophe JP Order was compelling PRASA to employ the Applicants. The need for an approved plan must be linked to that goal.

[84] Accordingly, assuming that PRASA can demonstrate that the Regulator has in fact approved its safety plan for the Western Cape, it will have met this part of the Hlophe JP Order.

[85] I accept that the Hlophe JP Order anticipated that this Court would also satisfy itself of the adequacy of the safety plan. The Applicants appear to accept that, if indeed the Regulator approved PRASA’s safety plan, that it was not necessary for the Court to also substantively interrogate it. They certainly did not suggest there was any substantive reason to conclude it was not adequate. Given the current state of the matter, and the position of the parties, I do not intend to separately evaluate whether PRASA’s safety plan is “adequate”. If the Regulator approves it, that will be sufficient.

 

The Tender was Not Completed or Implemented

[86] PRASA’s application in this Court was premised on the notion that it had complied with the Hlophe JP Order because it had started the tender process, and anticipated that it would be completed before the application was determined. The Applicants objected, arguing that until the tender was in fact awarded, the application was premature.

[87] The Applicants are plainly correct. Hlophe JP required PRASA to “complete and implement” the 2019 Tender. While that specific tender was quickly abandoned, PRASA did not take the point this rendered that part of the Order inoperative. Rather, the parties accepted that the Order should be read to require completion and implementation of a comparable tender.

[88] The reason for that requirement was to avoid a gap between the termination of the Applicant’s services, and the appointment of new service providers. That is exactly the threat that faced Hlophe JP in 2019 – PRASA had terminated the Applicants’ services before appointing their replacements under the 2019 Tender. Allowing PRASA in 2023 to terminate the Applicants’ contracts while the 2023 Tender was still in process would create precisely the same risk.

[89] The stop-gap measure of using the ACSA Panel could not save PRASA. Leaving aside its legality – which I address next – it was no longer a practical solution because it was not available after 30 September 2023. So PRASA could not use it to fill a gap left by the Applicants if it terminated their services before the 2023 Tender was implemented. PRASA presented no other way to fill that gap.

[90] Even though, by the time of the hearing, the 2023 Tender seemed close to completion, many things could still go wrong. No bids had yet been approved. No contracts with bidders had been concluded. No dates for the appointed service providers to place boots on the ground had been set. There was still a risk the tender process could collapse, or that an unsuccessful tenderer could interdict its implementation. On PRASA’s own version, it was still seeking further information. At the time PRASA launched the application, and still at the time of the hearing, there was uncertainty about whether, and especially when, the 2023 Tender would be implemented. Until that date is known with some degree of certainty, this Court cannot responsibly allow PRASA to terminate the Applicants’ services.

[91] PRASA’s counsel ultimately accepted as much at the hearing. He took an instruction from his client that it would agree to an order that would pend permission to terminate the Applicants’ services until the actual implementation of the 2023 Tender. The concession was well made. Whether it should have been made earlier is an issue I return to when I address costs.

[92] The consequence is that it was ultimately common cause that PRASA had not satisfied this leg of the Hlophe JP Order. As it was required to satisfy both legs, it is not yet entitled to terminate the Applicants’ services. My order will address how this Court will determine when the condition has been fulfilled.

 

The Just and Equitable Order

[93] In my view, it is not enough to dismiss PRASA’s application. All the Applicants sought alternative relief to vary the Hlophe JP Order to align it with the past and the present. They recognized that the Order could not be used to allow them to supply their services to PRASA indefinitely. What is required is an order that will enable the Court to determine when PRASA is in a position to terminate the Applicants’ services, because it has put other measures in place.

[94] As I mentioned earlier, in light of the argument at the hearing of the matter, I invited the parties to propose draft orders. The parties could not agree on an order. They filed separate proposals. Each proposed structural interdicts that would empower the Court to retain supervision until such time as PRASA had met the requirements of the Hlophe JP Order, entitling it to cancel its agreements with the Applicants.

[95] PRASA proposed a simple order that would require it, by 30 November 2023, to report on the status of the 2023 Tender and confirm that Mr Kgomari’s letter reflected the view of the regulator. The Applicants would respond, and the Court would issue further directions if necessary. The Applicants jointly proposed a similar, slightly more detailed supervisory order. Their order expressly provides that, as I had indicated at the hearing, I would retain the file.

[96] I intend to grant an order that has elements drawn from both proposals. To explain why, I set out a few fundamentals.

[97] This is a constitutional matter. The justification for the Hlophe JP’s initial order was not, primarily, the Applicants’ commercial interests, but commuters’ constitutional rights. The Order could only have been granted in terms of s 172(1)(b) of the Constitution which enables any court determining a constitutional matter to grant relief that is just and equitable. Hlophe JP presumably satisfied himself that cancelling the Applicants’ services without a plan to replace them risked violating those constitutional rights. But the power to grant just and equitable relief is not contingent on a finding, under s 172(1)(a), that PRASA’s conduct was unconstitutional and invalid.16

[98] The power to grant just and equitable relief is “so extensive that [courts] ought to be able to craft an appropriate or just remedy even for exceptional, complex or apparently irresoluble situations.”17 The primary limit on that wide power is that a s 172(1)(b) remedy must provide effective relief – it must be “practically effective judicial intervention”.18 Sometimes that will be simple – directing an organ of state to cease unconstitutional conduct. But often it will be difficult, because the nature of the constitutional violation does not permit a simple solution. Courts then have a role to ensure that the order they give will meaningfully resolve the constitutional harm. Supervisory interdicts are one of the tools available to achieve that end.

[99] Therefore, the goal of supervisory relief is not punitive, but pragmatic.19 A court retains supervision because it cannot adequately resolve the dispute between the parties, or adequately protect the public interest through only a once-off order. Supervision is a recognition that all “three branches of government are engaged in a shared enterprise of fulfilling practical constitutional promises to the country’s most vulnerable.”20 An order of supervision is a judicial commitment to work together with other branches to resolve a constitutional infringement, or realise a constitutional commitment.

[100] Supervision can be necessary for different reasons. The court may need further information before it is in a position to take a final decision. The court may wish to ensure that a party in fact does what the constitution requires, because it has previously failed to comply with court orders, or to perform its legal functions. Or the problem may simply be so enduring or complex that it can only be resolved through regularly revisiting how the ultimate constitutional goal can be achieved. Supervision recognizes that circumstances change, that what may seem a reasonable solution today, may turn out to be unworkable. It allows government and courts to adapt, rather than to be constricted by court orders that may become impractical.

[101] Supervision is often a far more effective means to ensure compliance in constitutional cases than a simply mandamus coupled with the threat of contempt. The Supreme Court of Appeal recognized this in Meadow Glen, holding that “[c]ontempt of court is a blunt instrument to deal with” complex problems, and that “courts should look to orders that secure on-going oversight of the implementation of the order.”21 It does not help to threaten to hold a government official in contempt when she lacks the skills, resources or capacity to implement a court order.

[102] One difficulty in supervisory orders is that the judge who grants them generally does not retain control of the file. Instead, the file is returned to the general pool to emerge only when the parties require that it be brought before another judge for a decision. While judges are entitled to retain the file, there is no rule or practice that requires them to do so. In this case, the supervision of the Hlophe JP Order travelled from Savage J to Ndita J to Gamble J to Le Grange J to Hack AJ to me.

[103] This may not be a problem in commercial litigation requiring supervision where the Court’s role is limited to resolving a dispute between the parties. But in constitutional litigation where supervision is granted not only to serve the litigants, but to fulfil the rights of non-litigants and the broader public interest, the position is different. In those cases – and this is one such case – abandoning control of a file after ordering supervision has a variety of undesirable consequences.

[104] First, each judge must read the file afresh and acquaint themselves with the history of the matter. In this case, the record ran to 1600 pages. That unnecessarily consumes scarce judicial resources.

[105] Second, even a diligent judge will not have the same understanding of the matter as a judge who was involved in the matter from the beginning. They do not have the benefit of argument from counsel at earlier stages. And recreating exactly how and why decisions were made along the way – particularly when there is an order without reasons – is challenging. That is so here, where Hlophe JP gave no reasons for the order, and the parties disagreed (until very late in the day) on what it meant. If a different judge must rule each time judicial supervision is necessary, there is also an increased risk of inconsistent approaches, which will be unlikely to lead to effective solutions.

[106] Third, when the file is returned to the general pool, the result is that it is no longer their responsibility. Instead, it is left to the parties to drive the matter. Until a party decides that judicial involvement is necessary, nothing will happen. There may be cases where that is appropriate. But where, as here, the order is made not only to protect the parties’ interests, but the public interest, relying solely on litigant’s self‑interest to ensure successful supervision is unlikely to be effective.

[107] Again, this case makes the point. Having obtained the Hlophe JP Order and ensured regular payment, the Applicants had no commercial interest in PRASA satisfying its requirements. As long as the Order remained in place, they continued to enjoy the fruits of their 2011 Tender. Why would they push to end that boon? That is not a criticism of the Applicants, it is a stark fact. One would have hoped that PRASA would have taken the necessary steps to satisfy the Hlophe JP Order, so that they could implement their new plans to ensure rail safety. But for reasons good or bad, it didn’t. The result was that virtually nothing happened for three years. An order that was meant to last a few months, has been in place for almost four years. That is plainly undesirable and ineffective.

[108] If Hlophe JP, or Hack AJ, or any of the other six judges who dealt with the matter, had retained the file, and taken steps to monitor compliance, that likely would not have happened. At the least, PRASA would have been forced to regularly explain its delay along the way, which almost certainly would have hastened the resolution. That would have been in the interests of commuters and PRASA. This is not a criticism of those judges – hindsight is always 20/20. At the time those orders were given, it likely seemed that resolution was imminent and active supervision unnecessary. But it is a demonstration of why a judge should generally keep a file that requires supervision, even when it may seem unnecessary, because things seldom work out as anticipated.

[109] Fourth, if a judge retains the file it is far easier and swifter to address issues that arise along the way. The judge can issue new orders or directions when necessary, either in chambers, or after a hearing arranged with the parties. If no judge retains the file, the parties will have to obtain a new allocation, and a new judge each time they require judicial intervention. That causes inevitable delay, not of weeks, but often of months.

[110] Fifth, where parties – particularly state parties – ignore or fail to comply with supervisory orders, it undermines the rule of law and the integrity of the judiciary. Because these types of orders are not given solely for the benefit of another party, it is often substantively imprudent for judges to delegate the responsibility for monitoring and enforcement entirely to the litigants. Having determined that judicial supervision is necessary, a judge must also ensure that it is effective. That will often only be possible if she retains the file.

[111] In short, an order of judicial supervision is not the end of a judge’s role, but the beginning. By ordering supervision, a judge assumes the responsibility to ensure that their order is carried out. If she gives up the file, she delegates that obligation to the parties.

[112] I do not mean to lay down any ironclad rule that judges must retain files when they grant supervisory orders in constitutional matters. Certain cases may not require active judicial monitoring. If there is no public interest component, then the parties may not need – or be entitled to – the benefits of active judicial supervision. And there may be cases where it is inappropriate or impossible for the judge that granted a structural interdict to supervise it. But it should be the default position. Absent some factor pointing against retaining the file, a judge should do so.

[113] For those reasons, I made it clear to the parties at the hearing that, if there was to be further supervision, I intended to retain the file until the matter was resolved. They expressed no objection.

[114] That leaves only the details of the supervision. PRASA suggested that it should have until 30 November 2023 to report to the Court on the status of the 2023 Tender, and whether it had obtained confirmation of approval from the Regulator. That seems like a reasonable period of time. It is at least possible that, by then, PRASA will know when the newly appointed service providers can begin. Once it has filed its report, the Applicants should have an opportunity to respond to it.

[115] What happens thereafter is harder to predict. If the parties agree PRASA has complied, then there should be an order permitting it to terminate the Applicants’ services, and discharging the Court’s supervision. If there is disagreement on compliance, it may be necessary to arrange another hearing, or it may be possible for me decide the issue on the papers. If PRASA concedes, or I conclude, it has not yet complied then it will be necessary to set further reporting dates. The order I make recognizes these possibilities, but does not seek to govern them in detail. A more detailed order may well be unnecessary, and it seems unwise to be too prescriptive in advance.

 

The ACSA Panel

[116] It is, fortunately, not necessary to delve into whether PRASA would have been entitled to use service providers on the ACSA Panel, as interesting as that issue might be. By the time I heard the application on 19 October 2023, the issue was moot. The ACSA Panel had been disbanded on 30 September 2023. PRASA no longer sought to use it as a stopgap to provide service pending the termination of the Applicants’ services and thereby justify discharging the Hlophe JP Order before the 2023 Tender is implemented.

 

Mzansi Securifire

[117] To recall, as part of their counter-application, the Sechaba Applicants launched a review of the appointment of Mzansi Securifire to provide security services. The only relief they sought before me was to compel PRASA to provide a record of its decision in terms of rule 53.

[118] There are two reasons I decline to grant that order.

[119] First, it was not sought through the proper procedure. It is not necessary to obtain an order to compel a party in a review to file a rule 53 record. Rule 53(1)(a) requires a notice of motion initiating a review call upon the decision-maker to dispatch the record of the decision. The respondent then has 15 days to file that record. If it does not do so, the applicant must employ the usual mechanism to compel compliance with the rules – an interlocutory application in terms of rule 30A.22 In my view, it is not open to an applicant to circumvent that procedure – and the procedural and substantive requirements of rule 30A – by seeking an order to compel production of the record that anticipates non-compliance.

[120] Second, the entire review was incorrectly brought under this case number. The core of the Hlophe JP Order was to prevent the termination of the Applicants’ services until PRASA had demonstrated it could replace them. That did not create a monopoly for the Applicants over security services for PRASA in the Western Cape. PRASA retained its ordinary power to procure supplementary security services to fill new security needs. That – on the version before me – is all it did. It is not appropriate to use a court’s supervisory jurisdiction as a vehicle for disputes between the parties that are not relevant to the purpose of supervision.

[121] Mr Nacerodien – appearing for the Sechaba Applicants – argued that the appointment of Mzansi Securifire may in fact have displaced his clients from providing services. They required the record to determine whether that was the case. This is akin to a fishing expedition. For the review to have been incorporated as part of the Court’s supervisory role, there would have had to be evidence that in fact it was interfering with the purpose of the Hlophe JP Order – ensuring commuter safety. There was no such evidence in the affidavits supporting the counter-application.

[122] The Sechaba Applicants’ application to compel the provision of the review record is therefore dismissed.

[123] That leaves the tricky question of what to do with the review of the Mzansi Securifire tender. It should not have been part of this case number, and will not form part of this Court’s further supervision. But I have not considered its merits and do not wish to preclude the Sechaba Applicants from proceeding with that review in separate proceedings if they are so advised. I intend to make an order to that effect.

 

PRASA’s Constitutional Obligations

[124] Has PRASA failed to fulfil its constitutional obligations? The Sechaba Applicants identify three. First, the obligation under s 165(5) which provides that a court order “binds all persons to whom and organs of state to which it applies.” Second, s 195(1)(f) which provides that the public administration “must be accountable”. And third, s 237 which requires that “[a]ll constitutional obligations must be performed diligently and without delay.”

[125] PRASA’s conduct in this matter leaves much to be desired. It began with terminating the Applicants’ contracts in 2019 without a replacement. It continued with its failure to pay the Applicants, its tardiness in filing the Rakau affidavit, and its attempt to circumvent the Hlophe JP Order through a secret tender. It then failed to take any steps to fulfil the requirements of the Hlophe JP Order between 2020 and 2023. The premature timing of this application caps off its unsatisfactory conduct.

[126] But does that conduct warrant a declaration that it has breached the Constitution? I think it is necessary to distinguish between two periods – before Hack AJ’s judgment, and after it.

[127] PRASA’s failure to pay and its failure to timeously file an affidavit were squarely before Hack AJ. He was considering the return dates of the Gamble J Order. It required PRASA to give reasons why it should not be declared that it had not complied with the Ndita J Order, “why it should not be declared that [it has] breached sections 165(5), 195(10(f) and or 237 of the Constitution”, and why it should not be found in contempt. The Ndita J Order required PRASA to “show cause why the first and second respondent should not be found to be in contempt of court” for its conduct up to that point.

[128] Hack AJ was required to consider all of this. He explains in his judgment that the parties agreed that “the main application would not proceed but I was to determined (sic) certain aspects of two orders and the money claims.” It is not clear what that means. I was informed by Mr Nacerodien that the issues of constitutional non-compliance and contempt were argued before Hack AJ, and that he could not explain why they were not dealt with.

[129] Hack AJ was severely critical of PRASA. He held its “contemptuous conduct … over so many years … are in my view, an egregious example of disdain for the court which [it] must realise undermines the judiciary and therefore the entire fabric of our society and our constitutional democracy”. These are strong words.

[130] But Hack AJ did not find that PRASA had not complied with Ndita J’s Order, did not declare it in breach of the Constitution, and did not hold it in contempt. Instead, the criticism of PRASA’s conduct was to justify his order that it pay the Applicants’ costs on an attorney and client scale. Unfortunately, Hack AJ did not expressly explain why he declined to declare that PRASA did not comply with court orders, did not breach the Constitution, or was not in contempt of court. I was informed by Mr Nacerodien that the issues were in fact argued before him.

[131] Despite the lack of explanation, I do not believe it is open to me to reconsider what occurred before Hack AJ. He was expressly required to decide whether to grant those orders, and he decided not to do so. He decided instead that PRASA’s conduct justified only the punitive costs awards. The necessary implication is that he determined that further orders of contempt or breach of constitutional obligations were not warranted. As a result, the Sechaba Applicants cannot re-argue that PRASA’s conduct prior to 24 August 2020 warrants a declaration of non-compliance with the Constitution.

[132] That leaves PRASA’s conduct – or rather lack thereof – since the hearing before Hack AJ. Does that warrant a declaration of constitutional breach? PRASA’s primary sin was not to arrange a new tender to replace the Applicants sooner than July 2023. Its secondary sin was the timing of this litigation – an issue more appropriately dealt with as an issue of costs. It seems necessary to consider whether PRASA’s inaction justifies a finding that it has breached each of the three constitutional provisions raised by the Sechaba Applicants.

[133] PRASA certainly initially failed to comply with the Hlophe JP Order. It did not pay the Applicants, and it was extremely late in filing the Rakau affidavit. But I am only concerned with its conduct from 24 August 2020. PRASA’s torpidity must be deprecated. But the Order did not oblige PRASA to conduct a new tender process. It merely prevented it from terminating the Applicants’ contracts unless and until it had found a way to replace them. It required it to file one affidavit, but did not oblige it to engage in further reporting. If PRASA was content to continue to use the Applicants’ services indefinitely, that did not breach the Hlophe J Order’s terms (although it certainly undermined its intent). I therefore do not believe that it breached s 165(5).

[134] Similarly, I do not see why the failure to conduct a new tender renders PRASA unaccountable, contrary to s 195(1)(f). The Sechaba Applicants never explained precisely why s 195(1)(f) was breached by PRASA’s failure to act. Presumably it is because it took no steps to replace the Applicants. But it was never expressly required to do so. It was always open to the Sechaba Applicants to approach this Court to compel PRASA to replace them. But having secured regular payment in 2020, they had no motivation to do so, and took no steps until PRASA launched this application. I am not convinced it then lies in their mouth to claim PRASA was unaccountable. Because no judge retained the file, the Applicants were the only institutional mechanism available to place pressure on PRASA.

[135] I am also not convinced that, by failing to initiate a tender, PRASA failed to diligently fulfil a constitutional obligation. PRASA had an obligation to protect commuters. But it fulfilled that duty by using the Applicants’ services. The Sechaba Applicants did not explain what other constitutional obligation PRASA failed to diligently fulfil.

[136] If PRASA breached any constitutional obligations, it would seem to be the obligations in ss 165(4) and 217:

[136.1.] Section 165(4) requires that organs of state “assist and protect the courts to ensure the independence, impartiality, dignity, accessibility and effectiveness of the courts.” This imposes a “heightened duty”23 on organs of state to not only comply with court orders (as dealt with in s 165(5)) but to take positive steps to enhance the effectiveness of the courts. Here, it arguably required PRASA to not only comply with the letter of the Hlophe JP Order, but to fulfil its purpose. But the Sechaba Applicants did not rely on this provision, and PRASA was not called to defend itself on this score.

[136.2.] Section 217 obliges organs of state to ensure a “fair, equitable, transparent, competitive and cost-effective” procurement process. By keeping the Applicants in place for 11 years past the end of their initial tender, PRASA seems to have dismally failed to ensure it was obtaining the service it needed at a reasonable price. But the Sechaba Applicants, unsurprisingly, did not contend that the continued use of their services rather than their competitors constituted constitutional non-compliance.

[137] Accordingly, while PRASA may have breached other constitutional obligations, I am not convinced it has breached the obligations on which the Sechaba Applicants rely.

[138] That should not be read to condone PRASA’s conduct. It could and should have moved much faster to replace the Applicants. It did not provide a coherent explanation for why it took so long. But to my mind, the failure is also a judicial one. It shows that the supervision failed because it relied too much on the parties’ self-interest. Once they secured regular payments, the Applicants had no incentive to fulfil the purpose of the Hlophe JP Order – their own replacement. Having accepted that supervision was appropriate, this Court too had a duty to ensure that its order was not misused to undermine its basic purpose.

 

Costs

[139] The rule on costs in constitutional matters is that where a private party is successful against the state, it is ordinarily entitled to its costs.24

[140] In this instance, the Applicants have largely been successful. Not only did they resist the discharge of the Hlophe JP Order, but the primary relief they sought in their counter-applications – amending that Order – is in substance very similar to the relief PRASA conceded is necessary, and that I have ultimately granted.

[141] Vusa Isizwe has been substantially successful on all the issues it raised. It is entitled to 100% of its costs. The Sechaba Applicants sought relief concerning Mzansi Securifire that I have dismissed. It is not entitled to its costs for that part of the application. But that issue was limited. I determine that it is entitled to 90% of its costs from PRASA.

[142] The next question is on what scale those costs should be granted. Vusa Isizwe never sought a punitive scale, and so it shall receive its costs on the ordinary scale. But the Sechaba Applicants sought costs on the attorney and client scale. I am not inclined to grant costs on that scale.

[143] Punitive costs awards exist “to counteract reprehensible behaviour on the part of a litigant.”25 Because an ordinary costs award does not cover all the costs of litigation, punitive costs are justified when “it would be unfair to expect a party to bear any of the costs occasioned by litigation.”26

[144] I accept that PRASA’s application was imprudent. It ought to have waited until it had greater certainty about the fate of the 2023 Tender before seeking to discharge the Hlophe JP Order.

[145] But, without reasons to explain it, the Hlophe JP Order was certainly capable of an interpretation that the Regulator’s approval alone was sufficient. And while I ultimately conclude that Mr Kgomari’s letter is insufficient, that is only because Vusa Isizwe’s challenge was not answered, not because it was manifestly unreasonable for PRASA to rely on that letter. It was, therefore, at least plausible for it to believe it was entitled to the discharge of the Hlophe JP Order based solely on that letter, without having completed and implemented a tender.

[146] On the tender, PRASA also faced a very real predicament – albeit one partially of its own making. If it waited until it had certainty about when the 2023 Tender would be implemented before seeking this Court’s permission to terminate the Applicants’ services, it risked waiting too long. It would only have been able to launch the application after the conclusion of contracts with the new service providers, but it could not guarantee to those service providers when it would be permitted to employ their services. Moreover, it would have to wait for an indeterminate time for the matter to be allocated to a judge, for a hearing, and an order. It sought instead to short circuit that process by approaching the Court before the 2023 Tender was finalized in the hope that, by the time the Court was seized with the issue, new companies would have been appointed. As it turns out, it was guilty of excessive optimism in its own abilities.

[147] Of course, PRASA could have sought a less blunt order. It could have sought an order along the lines I now grant, or even of the sort the Applicants proposed in its counter-application. That type of application may not have elicited opposition from the Applicants at all. It ultimately proposed just such an order after the hearing. Arguably, it should have reached the realization some further supervision was necessary sooner. But approaching a court for a variation of the supervision order, only to have to approach it again when the tender was complete, would cause additional costs and delay. Especially without a judge to take control of the file.

[148] I am satisfied that an ordinary costs award is sufficient. It was ultimately proper for PRASA to approach this Court to revise the Hlophe JP Order. The fact that it could have done so more effectively or efficiently does not, in my view, justify a punitive costs award.

[149] All parties employed two counsel, and I am satisfied that was justified.

 

Conclusion and Order

[150] This case is nearly at the end of the line. After four long years, it is almost possible for the parties to disembark, and move on, unencumbered by the Hlophe JP Order. But it is not quite there. I hope that this matter will be finalized by the end of this year. But whatever course it now takes, I will retain the file to ensure it reaches its final destination as soon as possible.

[151] Accordingly, I make the following order:

 

1. That the First Respondent’s application dated 17 October 2023 to strike out portions of the Fourth Applicant’s replying affidavit is dismissed with costs, including the costs of two counsel.

2. That the First Respondent’s application dated 19 October 2023 to strike out the First to Third Applicants’ replying affidavit is dismissed with costs, including the costs of two counsel.

3. That the First Respondent’s application dated 19 June 2023, for the discharge of the order granted by Hlophe JP on 19 November 2019 (the Hlophe JP Order) is dismissed.

4. That the Hlophe JP Order is replaced with the following order:

4.1. That the First Respondent shall file an affidavit on or before 30 November 2023 that:

4.1.1. Reports on the status of its procurement process in bid number HO/SEC/002/05/2023, and in particular by what date any security service provider appointed in terms of that tender will be able to commence providing security services in the Western Cape; and

4.1.2. Advises whether the letter sent to it by Mr Kgomari on 24 February 2023, reflects the view of the National Railway Safety Regulator, such to be established by a letter or affidavit from the Regulator’s Chairperson or Chief Executive Officer.

4.2. That the Applicants may, by 14 December 2023, file an affidavit in response to PRASA’s affidavit.

4.3. That the Court shall consider the affidavits filed by the parties and either:

4.3.1. Terminate its supervision and permit PRASA to terminate the Applicants’ services on 60 days’ notice; or

4.3.2. Issue further orders or directions, including requiring the parties to file further reports, or setting the matter down for hearing.

4.4. That any party may approach this Court, on supplemented papers and appropriate notice to the other parties, for the variation of this order.

4.5. That Acting Justice Bishop shall retain the file until the discharge of this order.

5. That the First to Third Applicants’ application to compel PRASA to file a rule 53 record in its review of the appointment of Mzansi Securifire Group (Pty) Ltd is dismissed. The review of the appointment of Mzansi Securifire shall not form part of this Court’s continuing supervision of the dispute between the parties. This order does not preclude the First to Third Applicants from reviewing that decision in separate proceedings.

6. That the First to Third Applicants’ application to declare that the First Respondent had failed to fulfil its constitutional obligations in terms of ss 165(5), 195(1)(f) and 237 of the Constitution, is dismissed.

7. That the First Respondent is directed to pay the Fourth Applicants costs, including the costs of two counsel

8. That the First Respondent is directed to pay 90% of the First to Third Applicants’ costs, including the costs of two counsel.

 

____________________

M J BISHOP

Acting Judge of the High Court

 

 

 

Counsel for First to Third Applicants: Adv A Nacerodien

Adv N Nyathi

Attorneys for Applicant Mark Hess Attorneys

Counsel for Fourth Applicant: Adv R Solomon SC

Adv A Brink

Attorneys for Applicant Dunster Attorneys

 

Counsel for First Respondent: Adv D Jacobs SC

Adv A Coetzee

Attorneys for Respondents Norton Rose Fulbringht SA Inc