South African corporate social investment follows scorecard points and legitimacy more than outcomes

Introduction

Ask a South African company what it spends on corporate social investment (CSI) and it can usually give you a figure. Ask what that money changed for anyone, and the answer tends to get vaguer (Pillay and Moodley, 2025). Preston and Peeroo (2023) estimated that a 2 per cent levy on after-tax profits of certain categories of companies would have raised an extra R3.111 billion for CSI, which they put at 1.2 per cent of the Department of Social Development’s 2022/2023 budget. The number is useful because it shows how big the spending question is. It does not say whether the spending does any good, and that gap sits at the centre of this paper. Lefike, Turpin and Matthee (2023) make a related point from the project end, noting earlier findings that CSI projects are often short-lived.

The argument here is that CSI in South Africa is shaped mainly by two pressures, the Broad-Based Black Economic Empowerment (B-BBEE) scorecard and the need to hold on to a social licence, and that both reward visible spend more than demonstrable community outcomes. Dreyer, Viviers and Mans-Kemp (2021) show how quickly the socio-economic development element of the scorecard moved among listed companies, while Matebesi and Twala (2023) show how mining firms can use community programmes to protect their own position. Measurement of what CSI achieves is thin by comparison (Pillay and Moodley, 2025). The paper starts with the terminology and the legal setting, moves to the scorecard, turns to outcomes and measurement, and then looks at legitimacy and the mining cases before concluding.

CSI, CSR and philanthropy in the South African setting

The three terms get used loosely, and in South Africa that looseness has practical consequences. Preston (2024) set out to build definitions of corporate social responsibility (CSR), CSI and corporate philanthropy specifically for the South African context. She started from section 18A of the Income Tax Act 58 of 1962, which allows a deduction for donations to defined public benefit organisations, and built the definitions through a systematic literature review, coding the themes in Atlas.ti and using the frequency of codes to shape each definition. The method is transparent, which is its main strength, though a definition built from how often researchers use a word records habit more than it records what the best practice ought to be. The point of the exercise was that researchers in this country no longer have to borrow general definitions from elsewhere. That matters more than it sounds, because a study of a company’s CSI is only comparable with another if they mean the same thing by it (Preston, 2024; Ngoepe-Ntsoane, 2023).

The ordinary usage in South Africa treats CSI as the part of CSR that takes the form of a defined social spend, usually on education, health, skills or community projects, as distinct from the wider conduct of the business (Preston, 2024). The Mauritian comparison makes the distinction clearer. Preston and Peeroo (2023) found that what is called mandatory CSR in Mauritius, a contribution to a CSR fund required under its Income Tax Act since 2009, is in fact mandatory CSI. A company is told to hand over a share of profit. It is not told how to treat its workers, suppliers or the environment. So when people argue about whether South Africa should legislate CSR, they are very often arguing about CSI without saying so (Preston and Peeroo, 2023; Ngoepe-Ntsoane, 2023).

South Africa has not legislated either. Ngoepe-Ntsoane (2023) used document analysis to argue that South African law does not oblige companies to fulfil their social responsibilities and that a national policy framework, possibly a standalone statute, is needed. Her argument sits inside a call for a national consensus on a policy framework tied to social cohesion and nation-building. Preston and Peeroo (2023) reached a similar place from the tax side and recommended that a levy along Mauritian lines be considered, on the ground that it would bring every profitable company into CSI. Neither paper is neutral, since both start from the view that voluntary arrangements have not been enough. That is a reasonable view, but it carries an assumption that more money leads to better results, and the later sections suggest this assumption has not been tested (Pillay and Moodley, 2025).

How the B-BBEE scorecard shapes what gets funded

The nearest thing to a legal push towards CSI in South Africa is not a CSI statute at all. It is the B-BBEE framework. Lefike, Turpin and Matthee (2023) describe CSI in South Africa as deployed as part of B-BBEE to assist and empower disadvantaged individuals and communities. Dreyer, Viviers and Mans-Kemp (2021) note that B-BBEE, for all the criticism it attracts, remains the most prominent means of uplifting black people in the country. Companies that want a good scorecard rating, whether for tenders, for customers who count their suppliers’ ratings, or for plain reputation, have a reason to spend on socio-economic development whatever their view of the philosophy behind it. In that sense a good deal of South African CSI is compliance behaviour, and the scorecard is the thing that gives the compliance a shape.

The best quantitative evidence on this comes from Dreyer, Viviers and Mans-Kemp (2021), who analysed 1767 observations for 379 companies listed on the Johannesburg Stock Exchange between 2004 and 2015. Overall B-BBEE scores rose, and the socio-economic development element showed the largest change over the period, while employment equity showed the smallest. Those two results sit awkwardly together. Employment equity is slow to shift because it depends on who gets hired and promoted. Socio-economic development can be raised by writing cheques and setting up programmes, which is a much quicker move. The authors suggest that the Department of Trade and Industry should reflect on how attainable the employment equity and ownership targets are. They do not claim that this explains the pattern, so the reading is mine, but it fits the idea that companies climb the scorecard where climbing is cheapest and quickest (Lefike, Turpin and Matthee, 2023).

There are limits to what the study can show. It covers the 2004 and 2007 Codes of Good Practice, and the codes have been revised since, which may have changed what companies do. It also measures scores and not results. A rising socio-economic development score says that a company spent a defined share of its profit on qualifying activities, and tells us nothing about whether a school got better or a clinic stayed open (Dreyer, Viviers and Mans-Kemp, 2021). Lefike, Turpin and Matthee (2023) point to this gap from the other end, noting that earlier research found CSI projects to be often short-lived and unsustainable. A scorecard that counts spend would not be expected to pick up that kind of failure.

What CSI achieves and how anyone knows

The outcomes literature is small and mixed, and the studies disagree about what to measure (Pillay and Moodley, 2025; Lefike, Turpin and Matthee, 2023). Pillay and Moodley (2025) ran a quantitative longitudinal study over ten years and reported a negative statistical relationship between CSI and the Human Development Index in South Africa, where a 1 per cent rise in CSI went with a marginal fall in human development. They suggested that CSI spending is not reaching citizens, possibly because of systemic corruption. I would handle that finding with care. An aggregate national correlation cannot separate CSI from everything else moving the index, and the direction of cause could easily run the other way, with companies spending more where conditions are worsening. What the result does show is how little the available national data can say about CSI, which is a different and less dramatic point than saying CSI makes things worse. The study also leaves open which companies, sectors or kinds of project the spending covers, so it cannot say where the money went wrong.

Project-level work gives a more grounded picture. Lefike, Turpin and Matthee (2023) studied four CSI learning centres focused on information and communication technology in poor urban communities in Soweto. They built a systems framework drawing on soft systems methodology, Ubuntu philosophy and the idea of autopoiesis, meaning the ability of a system to produce and sustain itself. The centres lasted to the extent that community members supported each other and the centres could organise and collaborate on their own terms. Six guiding principles came out of it for future CSI projects of this type. The study is small and tied to one city and one kind of project, so it is best read as showing what a serious evaluation could look like and not as a result that applies everywhere. Still, it offers something the national figures cannot, namely an account of why a particular CSI project lasted (Lefike, Turpin and Matthee, 2023; Serfontein-Jordaan and Dlungwane, 2022).

The mining cases add a stakeholder angle. Serfontein-Jordaan and Dlungwane (2022) interviewed community development specialists at three South African mining companies and found that the companies aim for economic, social and environmental sustainability in their host communities, and that involving stakeholders in planning and decisions is seen as critical to making projects last. They also recommend that management balance competing stakeholder expectations. The weakness is who was asked. The respondents were company specialists, so the findings describe how the companies see their engagement, and the communities’ view of that engagement comes through only second-hand. Put next to the Soweto study (Lefike, Turpin and Matthee, 2023), the pattern is that sustainability depends on how much the community owns the project, and almost nobody measures that.

Legitimacy, risk and the mining evidence

If outcomes are hard to see, the legitimacy function of CSI is much easier to observe. Matebesi and Twala (2023) examined two conflict-ridden cases in South Africa using eighteen in-depth interviews and company and media reports. They concluded that mining companies mainly use CSR and the social licence to operate to assert and maintain corporate control while talking about local ownership and sustainable mining. The strategies gave temporary relief and gradually wore away the legal and political force of both ideas. That is a pointed conclusion, and it comes from two cases, so it should not be stretched across the whole industry. Still, it fits the way many mine-community disputes in the country are described, with the company offering projects at the point of protest, and it sits oddly beside the more hopeful picture of stakeholder engagement drawn by Serfontein-Jordaan and Dlungwane (2022).

Wider African evidence points in the same direction without being as harsh. Bezzola, Günther, Brugger and Lefoll (2022) built a dataset of CSR payments by 292 mining companies in 17 African countries over eight years. Companies were more likely to engage in CSR when they had been exposed to conflict in earlier years, and once the CSR was carried out it reduced tensions. But CSR did not bring conflict down to the level found in areas that had never received any, and the authors describe the effects as moderate. South Africa is one country among the 17, so the findings are not a direct statement about it. They do support the view that CSR behaves partly as operational risk management, which is a poor basis for deciding what a community actually needs. Matebesi and Twala (2023) and Bezzola et al. (2022) disagree on tone more than on substance.

The same logic shows up in how companies talk about CSI. Bezuidenhout and Kloppers (2023) interviewed ten senior communication practitioners in ten South African organisations. All of them accepted sustainability as the purpose of CSR, yet business sustainability came first and societal sustainability was supported rather than led. CSR communication was mainly used to present the company as a responsible corporate citizen and to protect its reputation. Their sample is small, and I found nothing in the ten sources that looks at CSI in the financial sector specifically, so claims about banks and insurers would have to rest on extension from mining and ICT cases (Matebesi and Twala, 2023; Lefike, Turpin and Matthee, 2023) and not on direct evidence.

What would better practice look like? The sources give pieces of an answer, though nobody has put them together for South Africa. Lefike, Turpin and Matthee (2023) drew six guiding principles from the Soweto centres, which suggests that evaluation can ask whether a project is able to sustain itself and not only whether it was funded. Serfontein-Jordaan and Dlungwane (2022) put the weight on stakeholder buy-in, collaboration and involvement in decisions. Bezuidenhout and Kloppers (2023) argue that bringing the principles of development communication into the theory of CSR communication would let it serve societal sustainability and not just the company’s reputation. None of these is a measurement system in itself. They do share a preference for asking what communities can do with a project once the company has stepped back, and a scorecard that counts spend is poorly placed to ask that.

Taken together, the scorecard and legitimacy evidence point to the same structure of incentives. A company is rewarded for spending, for being seen to spend, and for defusing conflict, and it faces almost no consequence for spending badly (Dreyer, Viviers and Mans-Kemp, 2021; Matebesi and Twala, 2023). That does not mean companies are cynical. Many of the people running CSI programmes presumably care a great deal. It means that the system gives them very little help in telling the good projects from the weak ones, and the measurement gap noted above leaves them without the tools to find out (Lefike, Turpin and Matthee, 2023).

Where does this leave the proposal for a CSI levy? The Mauritian model would put more money into the system, and that is its attraction (Preston and Peeroo, 2023). But if scorecard points and licence protection are already the main drivers, and if outcomes are barely measured (Pillay and Moodley, 2025), then a levy on its own would probably add to the volume of spending without changing what the spending does. A levy paired with outcome reporting, independent evaluation and community involvement in design would be a different proposition, though none of the sources tested such a package in South Africa.

Conclusion

South African CSI is best understood as a spending practice organised by the B-BBEE scorecard and by the need to keep a social licence. The scorecard evidence shows rapid movement in socio-economic development scores among listed companies (Dreyer, Viviers and Mans-Kemp, 2021), and the mining evidence shows community programmes serving company interests as well as community ones (Matebesi and Twala, 2023; Bezzola et al., 2022). Definitions have improved, since Preston (2024) now offers South African ones, but the law still leaves CSI to company choice (Ngoepe-Ntsoane, 2023).

The weakest part of the picture is outcomes. National data suggest no clear benefit and project studies show what depends on community ownership, yet few companies appear to measure either (Pillay and Moodley, 2025; Lefike, Turpin and Matthee, 2023). Anyone planning to raise CSI spending through a levy, or to rely on it as a development tool, should first ask for better measurement. Without it, the figure of R3.111 billion (Preston and Peeroo, 2023) is a number about money, and the question of what the money does stays open.

Reference list

Bezuidenhout, L. and Kloppers, E. (2023) Corporate social responsibility communication for both business and societal sustainability in ten South African organisations. Communicare: Journal for Communication Studies in Africa, 42(2), pp. 17-29. doi:https://doi.org/10.36615/jcsa.v42i2.1538

Bezzola, S., Günther, I., Brugger, F. and Lefoll, E. (2022) CSR and local conflicts in African mining communities. World Development, 158, article 105968. doi:https://doi.org/10.1016/j.worlddev.2022.105968

Dreyer, J.A., Viviers, S. and Mans-Kemp, N. (2021) Reflecting on compliance with Broad-Based Black Economic Empowerment codes of good practice: trends and suggestions. South African Journal of Business Management, 52(1), article a1963. doi:https://doi.org/10.4102/sajbm.v52i1.1963

Lefike, M., Turpin, M. and Matthee, M.C. (2023) A systems framework to analyze the impact of corporate social investment projects with an information technology focus. The Electronic Journal of Information Systems in Developing Countries, 89(5), article e12273. doi:https://doi.org/10.1002/isd2.12273

Matebesi, S. and Twala, C. (2023) Evolution of mining company responses to civil society mobilization in South Africa. Enterprise & Society, 25(2), pp. 358-375. doi:https://doi.org/10.1017/eso.2023.23

Ngoepe-Ntsoane, M.J. (2023) Rethinking and reimagining corporate social responsibility as legislation in South Africa using a qualitative document analytics approach. International Journal of Research in Business and Social Science, 12(1), pp. 1-13. doi:https://doi.org/10.20525/ijrbs.v12i1.2294

Pillay, S. and Moodley, F. (2025) The impact of corporate social investment on human development in South Africa. International Journal of Business Ecosystem & Strategy, 7(5), pp. 326-333. doi:https://doi.org/10.36096/ijbes.v7i5.944

Preston, M.J. (2024) Defining corporate social responsibility, corporate social investment and corporate philanthropy for South Africa. South African Mercantile Law Journal, 36(3), pp. 398-422. doi:https://doi.org/10.47348/samlj/v36/i3a3

Preston, M.J. and Peeroo, S. (2023) An analysis of the possibility to implement a CSI tax levy in South Africa: lessons from Mauritius. De Jure, 56(1), pp. 280-308. doi:https://doi.org/10.17159/2225-7160/2023/v56a19

Serfontein-Jordaan, M. and Dlungwane, S. (2022) Achieving sustainable corporate social responsibility outcomes: a multiple case study in the South African mining industry. Communitas, 27, pp. 1-20. doi:https://doi.org/10.38140/com.v27i.6071