(Graphic: John McCann/M&G)
Shortly before the May 2024 general elections a potentially significant step in South Africa’s fight against corruption was taken. On 19 April, Democratic Alliance (DA) MP, Glynnis Breytenbach, announced her intention to introduce a Private Member’s Bill: the Constitution 21st Amendment Bill. As gazetted for public comment, the Bill proposes the establishment of a permanent anti-corruption commission (ACC) in chapter 9 of the Constitution.
South Africa certainly needs a standing ACC. However, for reasons we explain below, it should not adopt the model suggested in the Bill. While Breytenbach’s “initiation” of the Bill is commendable, the Bill itself is flawed and thus not in a “suitable state for…future action”.
On 19 May, the period for public comment on the Bill ended. Strictly speaking, the Bill could therefore be tabled as is in the current session of Parliament. However, such a move would require the tacit approval of the Government of National Unity. To our knowledge, this approval has not yet been given. Ironically, the DA would likely have been able to introduce the Bill more easily before the “dramatic shake-up of SA’s political landscape” following the 2024 elections.
The fact that the Bill has not yet been tabled does, however, provide a timely opportunity for further critical reflection on its strengths, weaknesses and prospects for refinement. We seize this opportunity and, in so doing, propose an alternative model for a constitutionally-entrenched, standing ACC. In contrast to the model set out in the Bill, our proposal is constitutionally compliant and more likely to be effective in combatting corruption.
In brief, a major issue with the Bill is that it unlawfully divests the National Prosecuting Authority (NPA) of its constitutional power, among others, to “institute criminal proceedings on behalf of the state”. In particular, the Bill usurps the NPA’s power to prosecute serious corruption and high-level organised crime, instead vesting this power in its proposed ACC. This move is inconsistent with the Constitution, which gives the NPA a distinctive and “pivotal role in the administration of criminal justice.” The Bill’s suggested ACC model also differs starkly to that recommended by the Zondo Commission and is at odds with international best practice. Further, the Bill is likely to be counter-productive in the long-term, practical mission to tackle corruption.
Relatedly, there are three serious, interconnected problems with the Bill. The first problem is that it interferes with the Constitution’s carefully constructed scheme for the division, and exercise, of state authority. The NPA is established alongside the Judiciary and the Judicial Service Commission in chapter 8, “the courts and administration of justice”. The NPA is a “hybrid” constitutional institution that falls under South Africa’s “Integrity and Accountability” (“I&A”) Fourth Branch of State– it is not part of the Executive.
The Constitution vests the NPA – and it alone – with the “power” to “institute criminal proceedings” and “carry out any necessary functions incidental thereto”. Like all organs of state, it must act in the public interest, and is thus constitutionally-mandated to do so “without fear, favour or prejudice”, insulated from outside interference with its independence.
On the matter of the NPA’s independence, it is important to note that as early as 2011, in the famous case of Glenister II, the majority of the Constitutional Court agreed with Adv Paul Hoffman SC (acting as counsel for Mr Glenister) that the disbandment of the DSO (or “Scorpions”) and its replacement with the DPCI (the “Hawks”) was constitutionally wanting. In short, this was because unlike the Scorpions, which were housed in the NPA, the Hawks were located in the SAPS and so did not enjoy the necessary structural and operational independence. To this end, the majority emphasised that, “[t]o understand our native conception of institutional independence, we must look to the courts, to Chapter 9 institutions, to the NDPP, and in this context also to the now-defunct DSO [within the NPA]”.
While this Glenister-II gloss on the traditional, tripartite conception of the separation of powers went unnoticed in academic literature at the time, it should no longer be overlooked. The bottom line is that as early as 2011, our apex court was signalling its appreciation of South Africa’s (rising) Fourth Branch of State, and that – contrary to mainstream misconceptions – this Branch is not limited to the chapter 9 institutions, but includes various other independent bodies, such as the NPA.
Since Glenister II, the Constitutional Court has continued to reiterate the importance of the NPA’s “constitutional guarantee of independence”, such that “any legislation or executive action inconsistent therewith would be subject to constitutional control by the courts”. Section 32 of the NPA Act and the Code of Conduct underscore this “guarantee” by requiring the NPA to “be free from political, public, and judicial interference”.
As it stands, the Bill contemplates an unlawful and irrational “legislative interference” with the NPA’s constitutional mandate and independence. Why? Because it divests the NPA of key prosecutorial powers, and assigns them wholesale to its proposed ACC. The Bill thus also flouts the rule of law, as well as the separation of powers, which requires the NPA and other I&A bodies to play their distinctive roles in protecting our constitutional democracy.
Section 191A of the Bill is particularly objectionable. If inserted into the Constitution, this provision would give the proposed ACC the power, to the exclusion of the NPA, to investigate and prosecute certain corrupt conduct, namely that which “deprives a particular social group or substantial part of the population of the Republic of a fundamental right”, or “causes financial loss or damage that is significant to the Republic”.
Neither of these jurisdictional requirements is defined in the Bill, let alone self-evident. By linking the prosecution of serious corruption and high-level organised crime to such complex definitional enquiries, the Bill would cause much confusion – particularly insofar as the distinction between the ACC’s jurisdiction and that of the NPA will be unclear.
This raises the second problem with the Bill: its substance is such that it would likely lead to forum-shopping between the ACC and the NPA. This irrational outcome would undermine, rather than enhance, the objective of systematically and expeditiously tackling corruption. All public power must be exercised lawfully and rationally, which, at a minimum, means that legitimate objectives ( or “ends”) must be pursued using lawful measures (or “means”) logically linked to the attainment of the “ends”.
By excluding the NPA from prosecuting serious corruption and organised-crime-related offences, the measures proposed in the Bill are not rationally connected to its stated purpose of fighting corruption. If anything, a properly capacitated and well-functioning NPA is essential in this fight. After all, “[i]f you subvert the criminal justice system, you subvert the rule of law and constitutional democracy itself”, which is precisely why “the NPA Act proscribes improper interference with the performance of prosecutorial duties.
The Bill’s explanatory memorandum is telling. It reveals one of the Bill’s motivations as being a deep distrust in the NPA based on its past performance: “[i]t is common cause that the NPA was one of the institutions that suffered the most at the hands of the State Capture project.” However, it does not follow that now excluding the NPA entirely from the prosecution of serious corruption offences and high-level organised crime is a rational way of addressing these social scourges. On the contrary, such an approach is irrational and, if implemented, would probably impede the long-term project of preventing and redressing corruption
To the extent that the Bill aims to address any ongoing dysfunctionality of the NPA, this purpose would less invasively, and more rationally, be achieved through a simple amendment to the allegedly problematic “ministerial control provision” in section 179(6) of the Constitution. Such a minimalistic change would quash any lingering doubt about the NPA’s de facto independence without undermining the Constitution’s basic institutional scheme and founding values.
Further, while the NPA may have been dysfunctional during the state-capture years, the errors of yesterday should not stymie the good-faith efforts of today. Several such efforts are well underway. They are targeted, coordinated and seek to enhance the NPA’s role and independence as an important I&A body. These efforts should not be impeded prematurely and unjustifiably – an unfortunate outcome likely to ensue if the Bill were to be passed in its current form.
The third problem with the Bill is its lack of alignment with the solid groundwork laid in the Zondo Report, as well as international best practice on standing ACCs. Zondo recommended the creation of a “Permanent Anti-State Capture and Corruption Commission” to “investigate and publicly expose” corruption “in the way that … [the Zondo] Commission did”. The Zondo Commission was an “I&A Commission” of the highest order that fulfilled its “checking” function with the right mix of rigour and restraint.
In so doing, the Zondo Commission revealed the value of certain commissions of inquiry as truth-seeking mechanisms that can effectively foster peace, nation-building and restorative justice. Additionally, it showed how such commissions can lay foundations for systemic and institutional reforms, as well as retributive justice. In the latter respect, while commissions of inquiry – whether permanent or ad hoc – can and should expose wrong-doing, they should not impose civil or criminal liability for this is a judicial function within the separation of powers. The Bill contravenes this maxim, and relatedly, flouts the rule of law and presumption of innocence.
The adoption of a revised version of the Zondo Commission’s model for a standing ACC is preferable. Zondo’s suggestions pertaining to the role, functions, organisational structure, staffing, and independence of the ACC model are constitutionally sound. In addition, we suggest that the standing ACC furnish its reports not only to the President, but also to Parliament, and that it do so on an annual basis. (See, Lauren Kohn ‘Integrity & Accountability Commissions of Inquiry: A South African Perspective’ (2024) Utrecht Law Review [forthcoming]). Such an approach preserves the NPA’s prosecutorial authority while simultaneously integrating the new ACC into the existing network of I&A institutions, some old and some new. Examples of new I&A institutions include: the statutorily-entrenched, permanent Investigating Directorate Against Corruption; and, pursuant to the new Public Procurement Act, 2024, the independent Procurement Office (housed in the Treasury); and the Public Procurement Tribunal of record, which should operate as a kind of “specialised, fast-track court”. Importantly, like the NPA itself, all these new bodies are required to act independently, and hence, “without fear, favour or prejudice”.
Our refined proposal for a standing ACC accords also with contemporary international best practice. For example, under the Australian National Anti-Corruption Commission (NACC) Act, the NACC may not usurp the functions of, among others, “commonwealth integrity agencies” and its mandate excludes criminal prosecution. Rather, the NACC’s mandate covers the likes of preliminary investigations, reporting, public education, recommendations of reforms, and it may “refer” certain “corruption issues to Commonwealth agencies and State or Territory government entities”. Such a coordinated, multistakeholder model better suits the South African context than that proposed in the Bill. The Australian Act’s definition of corruption (in section 8) is also instructive. In essence, it entails a “breach of public trust” – this was both the cause and result of state capture in South Africa – and it aptly captures the definitional remit of the kind of corruption a standing South African ACC should focus on.
The GNU is thus encouraged to go back to the legislative drawing board and constitutionally entrench a permanent ACC as one of the country’s I&A fourth-branch institutions along the lines we have suggested. Unlike the Bill’s model, our proposal will ensure the ACC’s constitutionality, practical utility and legitimacy. It will empower the ACC with investigatory, educational, reporting and remedial functions in a lawful and rational way – that is to say, in a way that is constitutionally sound and logically connected to the purpose of combating “serious and systemic corruption”. This purpose would be premised on a clear definition of serious and/or systemic corruption, namely conduct linked to a “breach of public trust in state officialdom”. Further, on exposing such instances of corruption, the ACC must refer them to the NPA for prosecution, and/or “non-trial redress”, in accordance with the existing constitutional and regulatory frameworks.
Thus established, the ACC would be able to execute its mandate effectively by alleviating, among others, the investigatory complexities of serious corruption cases, while synergising with – rather than undermining – the (good) work of established institutions. This approach would ensure that the ACC operates in a constitutional manner and for constitutional ends.
Lauren Kohn is a Legal Scholar and Young Research Fellow at UCT; an admitted Advocate of the High Court of SA; and co-founder of the access-to-justice platform, www.SALegalAdvice.co.za. Before joining the Academy in 2013, Kohn practiced for several years as an Attorney at Webber Wentzel (in alliance with Linklaters), primarily in public-law fields.
Theunis Roux is Professor of Law and Head of the School of Global and Public Law at UNSW, Sydney and a leading global scholar in comparative constitutional law. Roux was the founding director of the South African Institute for Advanced Constitutional, Public, Human Rights and International Law (SAIFAC).