By Joshua Eveleigh and Astra Christodoulou

Introduction

In 2016, an important amendment to the Competition Act 89 of 1998 (the “Act”) was brought into operation. The amendment introduced criminal liability for cartel conduct in the form of section 73A of the Act, which states:

“(1) A person commits an offence if, while being a director of a firm or while engaged or purporting to be engaged by a firm in a position having management authority within the firm, such person –

(a) caused the firm to engage in a prohibited practice in terms of section 4(1)(b); or

(b) knowingly acquiesced in the firm engaging in a prohibited practice in terms of section 4(1)(b).”

Section 74 states further:

“Any person convicted of an offence in terms of this Act, is liable –

(a) in the case of a contravention of section 73(1), or section 73A, to a fine not exceeding R500 000-00 or to imprisonment for a period not exceeding 10 years, or to both a fine and such imprisonment; or

(b) in any other case, to a fine not exceeding R10 000-00 or to imprisonment for a period not exceeding six months, or to both a fine and imprisonment.”

In other words, a manager or director who steers a firm into price fixing, market division or collusive tendering can, under section 74 of the Act, face a fine up to R500,000 or imprisonment for up to 10 years, or both. Section 73A(5) of the Act would provide that a consent order or a finding by the Competition Tribunal (the “Tribunal”) or Competition Appeal Court (the “CAC”) that a firm is guilty of price fixing, dividing markets or colluding on tenders could be used as prima facie proof in criminal proceedings against any of that firm’s directors or managers. It must be noted, however, that while section 73A(1) to (4) came into operation on 1 May 2016, subsections (5) and (6) (the former dealing with the prima facie evidentiary effect, and the latter prohibiting a firm from paying or indemnifying a convicted individual) have not yet been brought into operation.

Despite section 73A having been in force for over a decade, it has not yet been put into practice.

One of South Africa’s largest and ongoing corruption cases, the investigation into the Tembisa Hospital Scandal, offers an opportunity to consider why this is the case. The case involves collusive manipulation of South Africa’s procurement system to the tune of billions of rands; however, the criminal cartel offence and operation of the Act have played no part thus far in the State’s investigation. Setting aside the criminal offence, section 59 of the Act gives the Commission a route against the firms involved that carries administrative penalties measured against their turnover, and that could be pursued independently of the criminal process. Tembisa’s sharpest lesson may lie not in the dormant criminal route but in the unused civil one.

The criminal cartel offence

Section 73A criminalises individual involvement in hardcore cartel conduct prohibited in section 4(1)(b) of the Act, namely price fixing, the division of markets and collusive tendering. It does not address criminal liability for the firm, which remains subject to the civil regime, but for the natural person who directed it.  Liability extends beyond registered directors to anyone “engaged or purporting to be engaged by a firm in a position having management authority”; therefore, the whole management chain may be exposed.

The offence may be committed in one of two ways: by causing the firm to engage in the prohibited practice, or by knowingly acquiescing in it. Section 73A(2) indicates that acquiescence requires actual knowledge of the conduct; therefore, mere negligence or a failure to detect a cartel operating below management level will not found liability.

The consequences of conviction are found in section 74, which provides for a fine of up to R500,000, imprisonment for up to 10 years, or both. Any fine must be paid personally by the convicted individual, and section 73A(6) would, once in operation, prevent a firm from paying it or indemnifying the person against it. A conviction also carries collateral consequences under company law, including potential disqualification from serving as a director.

Crucially, an individual cannot be charged in a vacuum. Section 73A(3) provides that a person may only be prosecuted once the firm has admitted, in a consent order, that it engaged in a prohibited practice under section 4(1)(b), or once the Tribunal or the CAC has made a finding to that effect. The finding against the firm is therefore a precondition to the criminal charge against the individual. Section 73A(5) would then allow that finding to serve as prima facie proof in the criminal proceedings against the individual. That feature, though not yet in operation, has attracted sustained criticism on the basis that it may amount to a reverse onus and could raise constitutional concerns regarding the rights protected in section 35 of the Constitution of the Republic of South Africa, 1996, including the right to be presumed innocent and the right to a fair trial

Enforcing the offence

The difficulty lies not in the definition of the offence but in the conditions attached to prosecuting it. South Africa operates two separate enforcement processes for cartel conduct. The civil process is directed at the firm, is investigated and prosecuted by the South African Competition Commission (the “Commission”) and is adjudicated by the Tribunal. The criminal process is directed at the individual and may be prosecuted only by the National Prosecuting Authority (the “NPA”).

Due to the precondition in section 73A(3), the civil process must ordinarily conclude, whether by consent order or by a Tribunal or CAC finding, before the criminal process against an individual can begin. Cartel matters are frequently litigated for years on technical and jurisdictional grounds before any such finding is made, therefore meaning the ability to prosecute criminal liability tends to arise long after the underlying conduct has occurred.

The interface between the two authorities must also be considered. The Commission cannot prosecute individuals and cannot grant them immunity from criminal prosecution. Under section 73A(4), the Commission is limited to certifying that a person is “deserving of leniency”, in which case it may not itself seek or request that person’s prosecution, and it may make submissions to the NPA in support of leniency. The final decision, however, is that of the NPA, which is not bound by the Commission’s view. Effective enforcement therefore depends on close coordination between the Commission, the NPA and the South African Police Service.

This division of functions establishes a tension with the Commission’s Corporate Leniency Policy, which has historically been its most effective tool for uncovering cartels. Therefore, if a firm’s admission in a consent order can be used as the basis for criminal prosecution of the firm’s own directors and managers, the incentive to cooperate and settle diminishes. Further, as section 73A remains untested in practice, there is also the unresolved tension as to what extent the NPA may access the material a firm hands over in the course of seeking leniency.

Tembisa Hospital and the potential use of the Act

The Tembisa Hospital scandal concerns the alleged looting of more than R2 billion from the procurement budget of the Tembisa Provincial Tertiary Hospital in Gauteng. The scheme is said to have exploited the Request for Quotation system for purchases below the R500,000 threshold (set in National Treasury Practice Note No 8 of 2007/2008 “Supply Chain Management: Threshold Values for the Procurement of Goods, Works and Services by Means of Petty Cash, Verbal / Written Price Quotations or Competitive Bids”, since amended), deliberately splitting orders to circumvent the formal tender process. Investigators have identified several syndicates operating through large networks of companies. One alleged network is said to have controlled dozens of entities, the majority of which were irregularly appointed as suppliers; another is alleged to have funnelled contracts through a cluster of commonly linked companies.

The use of many nominally separate companies gives the matter a competition-law dimension. From the perspective of Tembisa Hospital, there would have appeared to be several separate entities competing for each quotation. In reality, those entities were, unbeknownst to the hospital, allegedly controlled by a single mind, which caused them to submit coordinated and inflated quotations, ensuring that the lowest of several artificially high bids would win. That is the classic mechanism of bid rigging. The appearance of competition is manufactured to conceal an allocation of business that has already been decided among the bidders. Where a single controlling mind allegedly causes several ostensibly independent firms to submit competing quotations, thereby creating the appearance of competition while allocating business among themselves, the conduct falls squarely within collusive tendering or market division of the kind prohibited by section 4(1)(b). On that characterisation, the individuals directing those firms could, in principle, fall within the reach of section 73A. The concealment of common control is foundational in the competition-law analysis because it is what allowed the entities to hold themselves out as independent competitors. The Commission would still need to establish an agreement or concerted practice between them, but the coordinated submission of rigged quotations is itself compelling evidence.

The difficulties of using the Competition Act at Tembisa

Several obstacles arise when considering whether Tembisa is a suitable vehicle for the first practical use of section 73A.

The first difficulty is that the core issue of the Tembisa scandal has been corruption, not cartelisation. Its core is the corrupt relationship between officials inside the hospital and the suppliers they favoured, a relationship between buyer and seller rather than between competitors. The cartel offence is aimed at competitors who secretly agree to rig a market against a buyer, not at insiders who capture a procurement system from within. Whatever residual element of collusive tendering might be found among the front companies is, for criminal purposes, overshadowed by the established offences of fraud, corruption and money laundering.

Furthermore, the structure of the offence makes the utilisation of the route slow. The criminal case against the individual depends on a prior admission by, or adverse finding against, the firm. The Commission would first have to investigate, refer and litigate to a conclusion its case against the firm. Cartel proceedings of that kind practically run for years. One of South Africa’s longest-running cartel investigations, into the alleged manipulation of the rand, lasted close to a decade before the Constitutional Court had ruled only on preliminary questions of pleading and jurisdiction, without yet deciding whether any collusion had occurred at all. A procurement-corruption matter of this severity and national importance cannot sensibly wait on that timetable.

A further consideration is the sanction available under section 73A, which may be viewed as modest against the scale of the conduct. Section 74 caps the penalty at a fine of R500,000 and imprisonment for up to 10 years. Against an alleged scheme running to billions of rand, a fine of that order may be seen as negligible, and the competition route offers nothing equivalent to the asset-recovery machinery that is at the core of the State’s response, namely the forfeiture and preservation of the proceeds of the fraud. By contrast, the ordinary offences of fraud, corruption and money laundering carry heavier penalties and unlock the confiscation of the stolen funds.

This modest cap, however, attaches only to the criminal sanction against the individual. The Act’s civil regime is another matter entirely. An administrative penalty imposed on the firms, by section 59 of the Act, can reach 10% (ten per cent) of the firm’s annual turnover, a figure that, on contracts totalling billions of rand, would be anything but negligible. While imprisonment for up to ten years remains a significant sanction, section 73A offers little by way of asset recovery and may therefore be less attractive to prosecutors than the ordinary offences. The availability of a substantial administrative penalty, by contrast, raises a distinct question about the regulator’s own role.

Taken together, these difficulties explain why the criminal cartel offence, for all its apparent reach, may be seen as a slow and, on its criminal side, ill-suited instrument for a matter such as Tembisa, which largely explains the criminal route the State has taken.

The route the State has taken  

The State has thus far opted not to utilise the Competition Act. The response has been driven by the Special Investigating Unit (the “SIU”), which was authorised by Presidential Proclamation 136 of 2023, to conduct a forensic investigation. Under the Special Investigating Units and Special Tribunals Act 74 of 1996, the SIU refers evidence of possible criminal conduct to the NPA for prosecution and is separately empowered to bring civil proceedings in the High Court or the Special Tribunal to recover the State’s losses.

On the criminal side, the charges brought and contemplated are the offences of fraud, theft, corruption, money laundering, forgery and uttering, and contraventions of the Public Finance Management Act. In August 2026, one of the alleged syndicate masterminds was arrested on his return to South Africa and appeared before the Specialised Commercial Crimes Court facing dozens of counts, before being released on bail; the NPA has indicated that further prosecutions arising from the three identified syndicates are to follow. On the civil side, the Asset Forfeiture Unit has obtained substantial forfeiture orders against assets linked to the alleged networks, and the SIU continues to pursue recovery of the diverted funds.

The logic of this route is straightforward. The ordinary criminal law reaches the corrupt heart of the scheme, which competition law cannot; its offences are broader and well established; it requires no prior finding by the Tribunal as a precondition to charging individuals; and it allows the various enforcement agencies to coordinate a single, faster response.

The route the Commission has not taken

Even if the criminal cartel offence is seen to be ill-suited to Tembisa, the Act’s civil prohibition is not. Section 4(1)(b), enforced by the Commission and the Tribunal, catches the collusive tendering described above directly, and without the features that make section 73A difficult to utilise. It does not depend on the NPA, on the criminal burden of proof, or on the outstanding subsections of section 73A, and it is directed at the firms rather than at the individuals. On a finding of contravention, it carries an administrative penalty of up to 10% (ten per cent) of a firm’s annual turnover, which on contracts of this value would be substantial.

Prosecuting cartel conduct with the goal of administrative penalties is a financial route the Commission can pursue independently of both the NPA’s prosecutions and the Asset Forfeiture Unit’s proceedings under the Prevention of Organised Crime Act. An administrative penalty differs from recovering the specific proceeds of the fraud, since it is calculated on turnover rather than on the amount looted, and it is paid into the National Revenue Fund rather than restored to the hospital. It is not, therefore, asset recovery in the sense that forfeiture is, but it is a substantial financial consequence for the very firms that profited. It operates alongside rather than instead of the forfeiture the State is already pursuing.

On this front, Tembisa is close to an ideal case for the Commission. The NPA is already going forward with prosecutions arising from the same collusive bidding; the conduct has been investigated in exhaustive detail; and the sums involved are vast. It is therefore difficult to understand why the Commission, as the regulator charged with enforcing the very prohibition in question, has not itself moved to secure a finding and the administrative penalties that would follow. Delay is the obvious objection, since a referral to the Tribunal could take years; but that is a reason for the Commission to begin now having its prosecution running alongside the criminal one. The penalty does not turn on the criminal outcome, and the passage of time weakens neither the turnover figures on which it is based nor the public interest in imposing it. A parallel referral would not compete with the criminal prosecutions but complement them and would signal that rigged public tenders attract a competition-law response as well as a criminal one.

What Tembisa means going forward

Tembisa is a revealing test of the criminal cartel offence precisely because it was never invoked. It shows that section 73A is unlikely to become the vehicle for the headline procurement-corruption cases that draw national attention, for the reason that corruption is the core issue the NPA wish to prosecute rather than cartel cases, and the ordinary criminal law is better fitted in this regard.

However, the more pointed lesson of Tembisa concerns the Commission rather than the NPA. The same conduct that supports the criminal charges is also a civil contravention of section 4(1)(b) that the Commission can pursue in its own right. That route asks none of the questions that make section 73A so hard to use. The weight of the administrative penalty alone is justification for the Commission pursuing prosecution in its own right. Over and above the administrative penalties, the message it would send to firms participating in public procurement would be that uncompetitive conduct will not go unpunished by the competition authorities.

Where a genuine horizontal cartel sits behind a public tender, section 73A remains available, and a future consent order or Tribunal finding could still open the door to the prosecution of the directors and managers responsible. The real test of the criminal offence will arise when the Commission secures a clear finding under section 4(1)(b) and the NPA elects to prosecute an individual on the strength of it. Until then, the deterrent value of the offence will remain largely theoretical, depending as it does on the interface between the Commission and the NPA being made to work, on the dormant provisions being brought into operation, and, above all, on a prosecution finally being brought.

The more immediate lesson, however, concerns not the dormant criminal offence but the civil route beside it. The administrative penalty route requires only a civil referral against the firms; it can be pursued by the Commission alone, and it is available now. That is why Tembisa speaks less to an untested criminal provision than about the unused civil one.