South Africa’s Global Minimum Tax Act defends the tax base but barely reaches non-resident digital companies

Introduction

South Africa has had a global minimum tax on its statute book since 1 January 2024, at least formally. The Global Minimum Tax Act 46 of 2024 and the Global Minimum Tax Administration Act 47 of 2024 were enacted in December 2024 and January 2025, and the first of them applies to fiscal years beginning on or after that date (Republic of South Africa, 2024; Republic of South Africa, 2025). It is tempting to read all this as the country’s answer to the digital economy, since the OECD sold its two-pillar package as a response to the tax problems of digitalisation (Sengwane, 2023). That reading is wrong, or at the very least badly incomplete.

This paper argues that the Act is a sensible piece of tax-base protection and a weak instrument against non-resident digital companies. The minimum tax asks how lightly a group’s profits are taxed in a jurisdiction. It does not ask whether a foreign platform earns money from South African users without any presence here, which is what the digital tax debate was about in the first place (Sengwane, 2023; Legwaila, 2024). The work of reaching those companies still sits with the market-side rules of Pillar One and with domestic choices that South Africa has so far put off (Banga and Beyleveld, 2024). The sections below explain what Pillar Two actually measures, set out the South African legislation as it stands in October 2026, weigh the developing-country argument about domestic top-up taxes, and then look at what the market-side rules might yield and what else the country could do.

What the minimum tax measures and what it leaves alone

Pillar Two applies to groups with consolidated revenue of at least EUR 750 million and aims at a 15 per cent effective rate in every jurisdiction where the group operates (Sengwane, 2023). Where the rate falls short, a top-up tax is charged, either by the parent’s home state or by the state where the low-taxed entity sits. The test is about rate. It says nothing about nexus. A foreign streaming service that sells to millions of South Africans from a server elsewhere, and pays 20 per cent on its profits in that other country, clears the test comfortably, and South Africa collects nothing extra from it (Legwaila, 2024). This whole thing is easy to miss because the headline word is ‘minimum’, which sounds like it must catch the big digital names.

Legwaila (2024) goes further and argues that the rules were designed by and for a small group of wealthy countries. The income inclusion rule gives first call on undertaxed profit to the parent’s home state, so any benefit for developing countries is secondary and hard to pin down. The estimated global gain is about USD 150 billion a year. Developing countries mostly have statutory rates somewhere between 20 and 40 per cent, which means a 15 per cent floor is rarely a problem for an ordinary company, and what the minimum tax really bites on is tax incentives (Legwaila, 2024). Titus (2022) reaches a similar worry from the African side, arguing that Pillar Two carries negative implications for corporate tax policy in African countries and may even affect progress on the Sustainable Development Goals.

The design is also less of a flat 15 per cent than the label suggests. Schjelderup and Stähler (2023) show that the substance-based income exclusion lets a low-taxed subsidiary deduct payroll and tangible asset costs twice from the top-up base, so the full minimum is not reached whenever that exclusion is positive, and for a large enough cost share of labour or capital it behaves like a production subsidy. Perry (2023) says something similar about Sub-Saharan Africa, describing the outcome as a far cry from the 15 per cent minimum usually portrayed, because of the exceptions and the complexity. For an asset-light digital business the exclusion is probably small, though neither paper tests that directly.

The problem that put digital companies on the agenda in the first place is nexus. Sengwane (2023) lays it out clearly. South African residence and source rules mostly depend on physical presence, a website is not a permanent establishment, and a company can sell cloud services into the country without a server, an office or an employee here. Pillar One was meant to fix that, by handing market jurisdictions a share of the residual profit of very large groups (Sengwane, 2023; Baraké and Le Pouhaër, 2024). Treating the two pillars as one package is where a lot of the confusion starts.

What South Africa has legislated

The Global Minimum Tax Act contains an income inclusion rule in section 4, which makes a South African constituent entity liable for top-up tax on low-taxed foreign operations, and a domestic minimum top-up tax in sections 6 to 19, which brings the South African profits of foreign-headed groups up to 15 per cent (Republic of South Africa, 2024). The charging provisions of the undertaxed profits rule are switched off, in sections 5 and 9 of the same Act (Republic of South Africa, 2024). SARS opened registration and notification on eFiling on 16 March 2026, with 30 April 2026 as the extended deadline for notifications and 30 June 2026 for the first GloBE information return (South African Revenue Service, 2026). So the machinery is actually running now.

Treasury does not expect much from it. The 2026 Budget Review cut the estimated revenue from the global minimum tax for 2026/27 from R8 billion to R2 billion, using the OECD’s updated rules and newer data on companies’ operations (National Treasury, 2026). Set against corporate income tax collections of about R346.6 billion in 2025/26, R2 billion is well under one per cent (National Treasury, 2026). That sits in the same small range as the figure Banga and Beyleveld (2024) give for Amount A, which I come back to below. On the numbers, neither pillar looks likely to change South Africa’s fiscal position.

A modest figure is what one would expect. South Africa’s corporate rate is 27 per cent for years ending on or after 31 March 2023, well above the floor (Sengwane, 2023), so top-up tax arises mainly where a foreign group has a South African entity whose tax has been pushed down by incentives or timing differences. Legwaila (2024) adds a mining point. Extraction produces most of South Africa’s profit and profit shifting, and handing undertaxed income to the parent’s home state is awkward when the parent is not where the economic activity happens. The domestic top-up tax answers that complaint fairly directly, because South Africa can collect the top-up itself, and the OECD (2026a) has described these domestic taxes as a primary mechanism for protecting local tax bases, particularly in developing countries.

The international rules also moved after the Act was passed. On 5 January 2026 the 147 members of the Inclusive Framework agreed a side-by-side package with simplification measures, a safe harbour for substance-based tax incentives, safe harbours for groups headquartered in jurisdictions with an eligible minimum tax regime, and a stocktake process (OECD, 2026a). The detailed package states that the side-by-side system leaves qualified domestic minimum top-up taxes unaffected (OECD, 2026b). For South Africa that is probably the best news in the whole saga. The home-grown top-up tax is the one piece the January deal left alone. South Africa also never switched on the undertaxed profits rule.

Whether a developing country should bother with a domestic top-up tax

The literature does not agree on whether a country like South Africa should have a domestic top-up tax at all. Perry (2023) concludes that Sub-Saharan countries should adopt one if a critical mass of advanced countries implements Pillar Two, while she treats the benefits of refundable credits and treaty renegotiation as more ambiguous. Das (2024) says close to the opposite. The tax is called unsuitable for developing countries because of its unfamiliar design, complexity, limited revenue gain, peer review and resource burdens, with incentive reform and an across-the-board minimum tax offered as the alternative.

Mollagee (2024) lands somewhere in between, arguing that the worry about domestic top-up taxes multiplying across Africa may be exaggerated, since the extra revenue does not justify the extra capacity investment, and that the main job for African states is to reassess their preferential regimes. Read together, Das (2024) and Mollagee (2024) seem to share a diagnosis: the top-up tax is a symptom of generous incentives, and a country that redesigns its incentives has less to top up. Perry’s (2023) position leans on a different premise, that someone else will collect the tax if the host state does not. Which premise is right depends on whether the parent’s home state has a working income inclusion rule, and South Africa cannot control that.

Titus (2022) and Oguttu (2022) widen the frame. Titus argues for an arrangement in which all countries stop what is labelled virtual tax competition, while developing countries keep engaging in real tax competition so long as their incentives meet defined criteria. Oguttu (2022) sets out to help developing countries decide whether the Model Rules suit their economies and whether they should adopt alternative provisions that serve the same aims. Both treat suitability as an open question and not a settled one, which is a useful corrective to the official line that adoption is simply the responsible thing to do.

My own view is that South Africa’s decision to legislate sits closer to Perry than to Das, though for a reason Perry does not stress. The Act arrived with a 27 per cent rate and an administration that already runs registration through eFiling, so the extra capacity cost is probably lower than in the low-income states Das and Mollagee have in mind. Sengwane (2023) still warns of semi-skilled administration and limited resources in South Africa, and the first information return only fell due on 30 June 2026 after an extension (South African Revenue Service, 2026). The January safe harbour for substance-based incentives (OECD, 2026a) also gives the country room to keep incentives that are tied to real spending, which speaks to the sovereignty worry Legwaila (2024) raises.

Where the digital companies actually sit

The market-side numbers are small, and they move with design choices. Baraké and Le Pouhaër (2024) model Amount A, which covers groups with at least EUR 20 billion in revenue and profitability above 10 per cent, and put the aggregate additional tax revenue between EUR 5.7 billion in 2020 and EUR 10.9 billion in 2022. The marketing and distribution safe harbour cut redistributed profits by an average factor of about 2.6 over 2016 to 2022, and in 2022 upper-middle-income jurisdictions, the group South Africa belongs to, gained about 0.5 per cent of their corporate tax revenue in the central estimates. The authors stress that these are first-round, static estimates, and that Amount A mostly moves taxing rights around. That fits Legwaila’s (2024) complaint that the benefits for developing countries are secondary and uncertain.

For South Africa in particular, Banga and Beyleveld (2024) put the gross gain from Amount A at about US$60 million, roughly 0.25 per cent of corporate income tax in the 2018 tax year. They also expect the country to lose revenue on net under the reallocation approaches they compare, because South Africa hosts multinationals whose foreign revenue would be reallocated. That is an uncomfortable finding. It means the country that waited for the OECD (Sengwane, 2023) may end up with a market-side rule that gives little and could take something back, and a minimum tax that never aimed at digital companies at all.

Other developing countries did the sums differently. Baraké and Le Pouhaër (2024) find that for the United Kingdom, India and Kenya a digital tax would raise more than Amount A. Legwaila (2024) reports that Kenya introduced a 1.5 per cent digital services tax in 2021, with a planned rise to 3 per cent, and that it was estimated to reach 89 companies compared with 11 under the OECD deal. South Africa has taken no such step. Banga and Beyleveld (2024) note that National Treasury has floated a digital services tax in budget reviews since 2020 without a proposal, and the tax chapter of the 2026 Budget Review does not mention one (National Treasury, 2026).

Trade commitments may narrow the options further. Banga and Beyleveld (2024) find that trade-in-services rules and limits on customs duties on electronic transmissions can directly constrain tax policy, and that South Africa’s commitments under the GATS would require any digital services tax to be non-discriminatory. They put the customs revenue forgone by not taxing digitisable imports at about US$44 million a year, around 1.2 per cent of customs revenue in 2020. That is not a large amount in its own right. But it shows how the pieces add up: a small gain from the minimum tax, a small gain from Amount A, and a small gain from customs duties, each of them sensible to chase and none of them decisive (Sengwane, 2023).

What South Africa could do instead

The Davis Tax Committee had a different idea, as Sengwane (2023) reports it. South Africa should not follow the developed countries blindly and should do its own research first. It should also add a source rule to section 9 of the Income Tax Act treating income from digital goods and services supplied by non-residents as South African in source, based on where the consumer is, and it should consider withholding tax on digital service fees, the technical services article of the UN Model, or alternative turnover-based minimum taxes. Sengwane (2023) argues that these steps would protect the base in the meantime. Nothing in the Global Minimum Tax Act (Republic of South Africa, 2024) does that job.

Administration is the obvious objection, and Legwaila (2024) makes it from the other direction, arguing that many developing countries’ main problems are administrative, such as weak enforcement, illicit financial flows, transfer pricing and the taxation of the digital economy, and that enforcing existing permanent establishment rules would help more than a minimum tax would. Sengwane (2023) adds that rules should be formulaic and hard to manipulate, partly because administrators may go easy on powerful multinationals. That argues for simple tools such as a withholding tax or a clear source rule over anything that needs a firm-by-firm profit calculation.

There is one more international tool worth a mention. The Subject to Tax Rule convention will come into legal existence on 1 January 2027 after Albania’s ratification, and it lets source states top up tax on certain intragroup payments taxed below 9 per cent in the recipient’s residence (OECD, 2026c). It is no digital tax, because it reaches only defined intragroup payments. Perry (2023) was cautious about the treaty rule, treating its benefits as ambiguous for Sub-Saharan countries, and whether it helps South Africa will depend on which treaty partners join. Even so, it is a rule written for source states, which is more than can be said for most of the rest of the package.

Conclusion

South Africa’s Global Minimum Tax Act is best judged as defence. It lets the country collect the top-up on low-taxed profits of foreign groups before a foreign parent does, it keeps an income inclusion rule for South African groups abroad, and its domestic top-up tax came through the January 2026 side-by-side deal unaffected (Republic of South Africa, 2024; OECD, 2026b). Treasury’s own estimate of R2 billion for 2026/27 says how much to expect (National Treasury, 2026). The case for it rests on protecting the base against incentive-driven erosion at low administrative cost, which is the part of the argument Perry (2023) and Mollagee (2024) can both live with.

Non-resident digital companies are a different matter, and the evidence suggests the Act barely touches them. Amount A offers South Africa a gross gain of around a quarter of one per cent of corporate income tax and possibly a net loss (Banga and Beyleveld, 2024), and the digital services taxes some peers adopted appear to raise more (Baraké and Le Pouhaër, 2024). Waiting for global consensus has not delivered a South African digital tax, and it is hard to see it doing so (Sengwane, 2023). If the aim is to tax these companies, source rules, withholding and better enforcement are the realistic route, and they sit in domestic law that Parliament can change without anyone’s permission.

Reference list

Banga, K. and Beyleveld, A. (2024) Are trade rules undermining taxation of the digital economy in Africa? ICTD Working Paper 181. Brighton: Institute of Development Studies. doi:https://doi.org/10.19088/ICTD.2024.007

Baraké, M. and Le Pouhaër, E. (2024) Tax revenue from Pillar One Amount A: country-by-country estimates. International Tax and Public Finance, 32(3), pp. 680-740. doi:https://doi.org/10.1007/s10797-024-09859-4

Das, R.R. (2024) Qualified domestic minimum top-up tax: should developing countries adopt it? Bulletin for International Taxation, 78(3). doi:https://doi.org/10.59403/1mx1cke

Legwaila, T. (2024) Global minimum corporate tax: developing countries beware. Obiter, 45(4), pp. 958-979. doi:https://doi.org/10.17159/x7kevp29

Mollagee, O. (2024) How appropriate are domestic minimum top-up taxes in Africa? Bulletin for International Taxation, 78(9/10). doi:https://doi.org/10.59403/21e2yz7

Oguttu, A.W. (2022) Preventing international tax competition and the race to the bottom: a critique of the OECD Pillar Two Model Rules for taxing the digital economy, a developing country perspective. Bulletin for International Taxation, 76(11). doi:https://doi.org/10.59403/zzqmwy

Perry, V. (2023) Pillar 2, tax competition, and low income Sub-Saharan African countries. Intertax, 51(2), pp. 105-117. doi:https://doi.org/10.54648/taxi2023004

Schjelderup, G. and Stähler, F. (2023) The economics of the global minimum tax. International Tax and Public Finance, 31(4), pp. 935-952. doi:https://doi.org/10.1007/s10797-023-09794-w

Sengwane, K. (2023) South Africa’s response to the digital economy’s direct tax challenges: Part 1. Obiter, 44(2), pp. 366-382. doi:https://doi.org/10.17159/obiter.v44i2.14461

Titus, A. (2022) Global minimum corporate tax: a death knell for African country tax policies? Intertax, 50(5), pp. 414-423. doi:https://doi.org/10.54648/taxi2022038

Legislation and case law

National Treasury (2026) 2026 Budget Review, Chapter 4: Revenue trends and tax proposals. Pretoria: National Treasury. Available at: https://www.treasury.gov.za/documents/national%20budget/2026/review/Chapter%204.pdf (Accessed: 18 October 2026).

Organisation for Economic Co-operation and Development (OECD) (2026a) International community agrees way forward on global minimum tax package. Press release, 5 January 2026. Available at: https://www.oecd.org/en/about/news/press-releases/2025/12/international-community-agrees-way-forward-on-global-minimum-tax-package.html (Accessed: 18 October 2026).

Organisation for Economic Co-operation and Development (OECD) (2026b) Global Anti-Base Erosion Model Rules (Pillar Two), including the Side-by-Side package published 5 January 2026. Available at: https://www.oecd.org/en/topics/sub-issues/global-minimum-tax/global-anti-base-erosion-model-rules-pillar-two.html (Accessed: 18 October 2026).

Organisation for Economic Co-operation and Development (OECD) (2026c) Albania’s ratification brings Subject to Tax Rule convention into force from 1 January 2027. Announcement, 15 September 2026. Available at: https://www.oecd.org/en/about/news/announcements/2026/09/albanias-ratification-brings-subject-to-tax-rule-convention-into-force-from-1-january-2027.html (Accessed: 18 October 2026).

Republic of South Africa (2024) Global Minimum Tax Act 46 of 2024. Government Gazette 51830, 24 December 2024. Available at: https://www.sars.gov.za/legal-lprim-act-2024-01-global-minimum-tax-act-46-of-2024-gg-51830-24-december-2024/ (Accessed: 18 October 2026).

Republic of South Africa (2025) Global Minimum Tax Administration Act 47 of 2024. Government Gazette 51884, 9 January 2025. Available at: https://www.sars.gov.za/legal-lprim-act-2025-01-global-minimum-tax-administration-act-47-of-2024-gg-51884-9-january-2025/ (Accessed: 18 October 2026).

South African Revenue Service (SARS) (2026) Global minimum tax. Available at: https://www.sars.gov.za/types-of-tax/global-minimum-tax/ (Accessed: 18 October 2026).