Menu
Tax Notes logo

The Impact of Unilateral Digital Taxes in Africa

Posted on Aug. 16, 2021

Nupur Jalan (jalannupur@outlook.com) is a professional and researcher specializing in corporate tax, international tax, and transfer pricing. Akshara Rao is a professional working in the tax field. Aakash Ravishankar is a professional working in accounting, audit, and taxation matters.

In this article, the authors survey Africa’s digital tax landscape, analyze various unilateral measures, and examine some digital business models to gauge how these measures might affect businesses and the continent as a whole.

Copyright 2021 Nupur Jalan, Akshara Rao, and Aakash Ravishankar. All rights reserved.

Digitalization and related technologies have given rise to a wide range of innovative business models and an endless array of new business processes. They have completely changed the way business is transacted and how economies operate, thereby disrupting traditional business models at a rapid pace and creating new challenges from a tax perspective.1 Many multinational enterprises use digital technologies such as data-mining algorithms to analyze big data, customize their products and services, increase the speed of operations, improve decision-making, become more competitive, and grow revenue. Digital technology has become a key element in the value creation process and has taken a prominent place in new business models. Also, businesses’ reliance on data — including the use of big data — has increased substantially over time. Governments and tax authorities worldwide must work to understand the digital data economy and update their rules and policies to reflect the value created by the newly emerging business models.2

African countries are ready for a “comprehensive digital transformation strategy to guide a common, coordinated response to reap the benefits of the fourth industrial revolution.”3 Digitalization is invigorating various segments of the African economy and contributing to the achievement of Agenda 2063 and the U.N.’s sustainable development goals. The continent is also seeing the production of innovative web-based applications, dynamic new business models, and an increase in global conglomerates trying to establish a presence in Africa. However, tax measures have not kept pace with these evolving business practices. Following their global counterparts, many African countries have turned to unilateral measures to implement some form of the digital tax regime.

Considering these developments, this article discusses the digital tax landscape in Africa. After analyzing some of the unilateral tax measures in the context of the African economy and considering their impact on MNEs and the region’s economy, we consider the pros and cons of these measures using specific benchmarks. We conclude with the observation that the results are not yet fully evident and how the pillar 1 and pillar 2 proposals may affect the situation remains to be seen.

I. Digitalization and Africa

Africa needs to develop a shared approach to the fourth technological revolution, a period characterized by the fusion of technologies related to the physical, digital, and biological spheres.4 The changing digital and data economy brings opportunities that can catapult Africa’s economic growth and lead to social development. Africa has numerous economic prospects in various sectors, and the continent’s youthful population presents an invaluable opportunity in this digital era. Gone are the times when African countries were dependent solely on extractive industries. MNEs in a variety of mainstream industries are willing to expand or establish their footprints in the African market for many reasons, including the promising workforce.

The African Union’s Digital Transformation Strategy for Africa looks to build on existing initiatives and frameworks — for example, the Policy and Regulatory Initiative for Digital Africa; the Programme for Infrastructure Development in Africa; the African Continental Free Trade Area; the African Union Financial Institutions (a group of three financial organs created to enhance integration); the Single African Air Transport Market; and the Protocol on the Free Movement of Persons — to support the further integration of the African Union. The Smart Africa Initiative has made the creation of a digital single market in Africa a key part of its strategic vision.5

A fundamental requirement for digital transformation is widespread access to high-speed internet. This access, however, is still a luxury — less than one-third of the African population has access to the internet. Close to 300 million Africans are more than 50 kilometers away from a fiber or cable broadband connection. Most of the population relies on mobile phones for an internet connection; access through connections installed at homes or offices is almost nil.6

The cost of internet-enabled mobile devices and access to data is another worrisome factor for African countries that have imposed high customs duties on mobile devices or excise taxes on so-called over-the-top (OTT) services (that is, technologies that can replace traditional telecommunications services) and other digital mediums, since it may negatively affect the growth of the internet value chain across the continent.7 Access to the internet remains a significant concern for the continent, and the gravity of this statement can only be fully understood by comparing Africa’s status with other regions. Figure 1 compares the percentage of individuals using the internet in Africa with several other regions.

Figure 1. Distribution of Internet Users Across the Globe

Further, conditions vary across the continent, and there are considerable differences in the various economies. Some countries on other continents are further along in the digital transformation and are better connected than many countries in the African region, according to Figure 1. For example, Mauritius ranks 58th out of the 151 economies included in the United Nations Conference on Trade and Development’s (UNCTAD) Business-to-Consumer (B2C) E-Commerce Index 2020, the highest ranking among African countries. Eight other African countries — South Africa (73), Tunisia (74), Libya (83), Kenya (89), Nigeria (88), Namibia (95), Senegal (98), and Republic of Tanzania (99) — ranked in the top 100.

The disparate economic growth among the African countries is further evident from the World Bank’s world development indicators. Figure 2 portrays the percentage of the population that has access to the internet in individual African countries.

Figure 2. Distribution of Internet Users Across African Countries

The digital landscape is expected to evolve as digital connectivity continues to explode across the continent. Over the next decade, the number of internet users in Africa will grow by 11 percent, and Africans will comprise 16 percent of all internet users globally.8 According to a report by Google and the International Finance Corporation: “By 2025, the Internet economy has the potential to contribute $180 billion to Africa’s economy growing to $712 billion by 2050.”9 Compared with 2016, mobile data traffic across Africa is expected to increase sevenfold by 2022.10 The region also has the fastest-growing tech start-up ecosystem, which plays an increasingly important role in the development of digital content and services.11

Business conditions in most African economies have developed significantly over the last 15 years,12 and the COVID-19 crisis has accentuated the reliance on digital operations. The continent’s digital economy has seen many leaps and jumps because of the group known as Africa’s KINGS — Kenya, the Ivory Coast, Nigeria, Ghana, and South Africa. Egypt, Rwanda, and Tunisia are also known for their growing digital economies, supportive tech infrastructure, and ecosystems of innovation supported by legislation (public sector) and an entrepreneurial spirit (private sector). Examples of country-specific successes include:

  • Tunisia has become a notable technology hub. This status expanded when the Start-Up Act of 2018 brought newfound investments and inspirations from innovative, youth-led technology start-ups in the private sector.13 Further, the government also initiated Tunisia Digital 2020 to promote information and communication technology (ICT). The country offers vast opportunities, including 10 competitiveness centers and technopoles as well as 18 regional cyber parks.14

  • In Nigeria, homegrown e-commerce platforms and entertainment companies are numerous. The country is also known for its fintech sector, and it receives the highest percentage of equity funding (21 percent) on the continent.15 The government has instituted several initiatives to help the ICT sector, including national ICT and broadband policies, community resource centers, a rural broadband initiative, and the Smart Nigeria Digital Economy Plan.16

  • Kenya’s fintech sector has been growing for several years, and nongovernmental organizations have been active in healthcare, agriculture, and education. The country is home to two of the continent’s major fintech companies: M-Pesa and Equitel. These companies helped boost the country’s financial inclusion from just 26 percent of the population in 2006 to 83 percent in 2021.17 The country is also popular for agricultural tech, with its startups receiving 79 percent of the continent’s equity funding in this sector.18 The government released its Digital Economy Blueprint in 2019 that aims to further expand access to the digital economy by focusing on five pillars: digital government, digital business, infrastructure, innovation-driven entrepreneurship, and digital skills and value.19

  • South Africa’s innovative digital MNEs are enhancing their footprints globally.20 The country has also experienced a rise in technology start-ups, which have raised more than $100 million. Its fintech companies, such as Yoco and Jumo, have also attracted large foreign investments.21 The country is also home to Naspers, one of the world’s largest technology investors.

  • Ghana has emerged as a key destination for building power plants. The country has 4,153 ICT establishments, which contribute $1.7 billion to the nation’s GDP (3.6 percent). The government continues to focus on this sector and plans to lead the ICT innovations in sub-Saharan Africa by 2023.22

  • Mauritius has 600 ICT companies operating within its borders.23 The government’s Digital Mauritius 2030 campaign focuses on digital government, ICT infrastructure, innovation, talent management, and cybersecurity. Mauritius leads the continent on various international indexes such as the Global Innovation Index, Global Competitiveness Index, and ICT Development Index.24

The increase in homegrown e-commerce and digital companies in Africa has coincided with efforts by other MNEs to expand their presence on the African continent. Entrepreneurs are harnessing technology to address deep-seated gaps in Africa’s markets. Examples of homegrown e-commerce companies include:

  • Naspers, founded in South Africa in 1915, is one of the largest technology investors in the world. The company founded the Naspers Fund, a ZAR 1.5 billion (about $101 million) VC fund, that supports South African businesses.25

  • IROKOTV, a Nigerian-based online platform popular all over Africa, provides paid-for Nigerian movies on demand.

  • Jumia is a vast e-commerce marketplace present in 11 African countries and three foreign jurisdictions, connects sellers with consumers and provides logistics services with shipment and delivery of packages.

  • Interswitch, a Nigerian digital payments firm with plans for a global listing could be valued at as much as $1 billion, which would make it Africa’s latest company to reach “unicorn” status.

  • Liquid Telecom, a pan-African broadband infrastructure and data services company, has more than 30,000 miles of fiber across 12 countries.

  • Loon, a sister company of Google, is going beyond fiber networks. It is building a network of high-flying balloons to connect people in rural Kenya to the internet.

  • Showmax is a leading homegrown subscription-based, video-on-demand business headquartered in South Africa. This internet TV service is a direct competitor of Netflix and offers the most original African content.26

  • Chipper Cash, an African Fintech start-up, provides peer-to-peer payment services across seven countries. It has also received global attention and backing from Jeff Bezos.27

  • Smile Identity is an ID verification and know-your-customer compliance service provider for companies in fintech (including Chipper Cash), healthcare, and other key industries. The service compliments various other e-commerce businesses and can validate more than 250 million identities across Africa.

  • Little Cab is a Kenyan app-based ride-sharing service that is a direct competitor of Uber. Many Kenyans prefer it over Uber because it accepts M-Pesa, a locally based cashless mobile payment system used by 94 percent of Kenyans.28 M-Pesa itself is a mobile-money platform that was created to send money between people but has grown into a marketplace.29

MNEs are also exploring ways to set up footprints in the African market. The following examples show that the building blocks are in place and Africa can offer a favorable climate for investors:

  • Netflix leads the African streaming market with nearly 2 million subscribers, having captured 57 percent of the subscription video-on-demand market by the end of 2020.30 The streaming platform has also increased its efforts to generate original content from the region and has rolled out lower-cost mobile-only subscription plans that cater to African-specific needs. (Discussed in detail in Section VI of this article.)

  • Uber recognized potential in the African market because of the low levels of car ownership. With more than 60,000 drivers across eight African countries — Egypt, Ghana, Kenya, Morocco, Nigeria, South Africa, Tanzania, and Uganda — the company aims to dominate the region’s transportation sector.31

  • IBM finds the continent to be a substantial market for its products and services. It currently has research labs in South Africa and Kenya. The company has recognized immense commercial potential in Africa and as a result is working toward innovative solutions in areas such as agriculture, transportation, financial inclusion, and e-governance.32

  • Alphabet Inc. has increased its digital presence on the continent. Google practically has a monopoly over the search engine market, with a share of 97.58 percent in 2020.33 Thus, Africa is an important part of the company’s advertisement market. Recognizing the enormous market, in 2019 Google opened an artificial intelligence lab in Ghana.34 The company also plans to fund start-ups and train 10 million Africans in online skills.35

  • Vodafone and Airtel — two of the world’s leading telecommunication giants — respectively hold substantial control of M-Pesa and Equitel, African fintech companies that dominate the continent’s digital payment sector and promote financial inclusion.

  • Facebook has become an essential medium for connectivity in Africa. This is evident from the increase in the number of users — from just 17 million users in 2010 to 210 million users (that is, 16 percent of the region’s population) in 2020.36 (Discussed in detail in Section VI of this article.)

At the same time, the rapid growth in digitally driven and digitally enabled companies has raised concerns for African governments in the context of their existing tax regimes. The next section offers a brief review of the digital tax landscape on the African continent.

II. The Digital Tax Landscape in Africa

International tax law and its underlying principles have not kept pace with the changes in global business practices and the transformation of the economy. Some of the salient features of digital business models include scale without mass, reliance on intangible assets, the innate role of user data and user participation, and network effects. Digital companies often operate in market jurisdictions without any physical presence. For example, Facebook opened its first and only African office in Johannesburg in 2015 — despite having 120 million users in Africa at that time.37

The fact that digital businesses often perform better than others during difficult times has placed a large target on their backs and given many governments the justification they need to impose an additional tax on the digital sector.38 The three key challenges of taxing the digital economy are illustrated in Figure 3.

Figure 3. Key Tax Challenges Posed by Digitalization

There has been a lot of debate about taxing Africa’s digital economy. Various institutions including the African Tax Administration Forum (ATAF) have organized discussions on the topic. For example, the 2019 ATAF Congress focused on “Digitalization: Challenges and Opportunities — Discussing the African Tax Landscape.” The ATAF has also issued several technical guidance reports concerning tax challenges arising from digitalization on the African continent. In its guidance, the ATAF put forth a DST proposal that African countries can adopt to address these challenges. But several African countries have gone their own way, adopting different variations of unilateral measures to tax digital businesses. In this section, we discuss both the ATAF proposal and these unilateral measures.

A. ATAF Proposal

Having grown from 25 members when it was formed in 2009 to 38 by November 2019, the ATAF is an international organization that provides a platform for cooperation among African tax authorities, and “has increasingly gained legitimacy and capacity in defending African interests on taxation issues.”39 In 2015 the ATAF formed the Cross Border Taxation Technical Committee to work with global tax leaders on international tax matters. The committee released a technical note in 2019 on challenges in taxing the digital economy.40 According to the technical note, several member countries believe that the existing emphasis on nexus and profit allocation rules heavily favors the resident countries. The technical note also raised concerns regarding the lack of adequate outcomes from the OECD base erosion and profit-shifting project. After highlighting the issues, the ATAF said it intended to initiate work toward addressing the issue of artificial profit shifting.

The ATAF’s Cross Border Taxation Technical Committee released its “Suggested Approach to Drafting Digital Services Tax Legislation” in September 2020.41 The guidelines propose a DST of 1 to 3 percent on gross annual digital services revenue that would be charged at a fixed rate based on the consideration paid for eligible services — that is, gross turnover attributable to the in-scope business activities. It would be charged in addition to any income tax and would not be creditable or eligible for any other type of relief against income tax. Also, inter alia, the ATAF’s proposal:

  • indicates that DSTs are not taxes covered by double tax agreements, which means countries can introduce DSTs unilaterally without needing to renegotiate their DTAs;

  • states that DSTs are taxes levied on gross turnover, not net profit, generated from a narrowly defined set of digital business activities — in other words, they are not income taxes so they may apply to revenue that does not produce income;

  • provides that the tax base includes revenues accrued within the country, directly or indirectly, as well as revenue attributable to the country that accrued elsewhere;

  • includes a broad list of services that may be considered digital;

  • includes a method for determining the location of users; and

  • offers some exemptions and allowances, such as a minimum threshold for applying the DST and exclusions for related-party revenue.

Concerning compliance with the DST, the proposal puts the onus of filing returns on the MNE that derives the digital service revenue. The draft legislation proposes that an MNE that is a resident of or functions through a PE in the taxing country, must select a responsible member that is in charge of compliance. Alternatively, the ATAF suggests that the MNE should appoint a local representative for compliance purposes.

Unlike the European Commission’s recent initiative,42 which sought to create a uniform, one-stop taxation system for DSTs across the EU, the ATAF proposal is merely intended to function as a guideline for African countries. However, African countries have gone their own ways to adopt a variety of unilateral measures. It remains to be seen whether, if the OECD’s global proposals do not go through, the ATAF will follow in the EU’s footsteps and put forward a uniform system like the one proposed in Europe. Considering the diverse mix of countries on the African continent, most of which are still in the developing phase, it seems difficult for ATAF to achieve the sort of integration that is possible in Europe.

B. Unilateral Measures

The vast majority of DTAs in effect today were negotiated before the explosion of the internet economy, and the appropriateness of applying the prevailing tax treaties to digital transactions has been a point of debate over the last few years. There were debates about digital taxation even before the OECD introduced BEPS action 1, and certainly before the latest discussions regarding the two-pillar proposal that the OECD published in October 2019. In October 2020 the OECD released blueprints for both pillars. However, since no one global approach for taxing the digital economy exists to date, several tax jurisdictions have resorted to unilateral tax measures such as imposing a DST, VAT, or other tax on digital transactions to protect their national tax base. Though many of these unilateral tax measures emphasize the need to review some of the fundamental international tax principles, the existence of these measures has definitely made a consensus-based solution even more difficult to achieve.43

Thus, the growth and transformation of the digital economy have had significant implications for direct and indirect taxes at national and global levels. This includes broadly applicable taxes like corporate income tax and VAT as well as any special taxes on social media, digital transactions, and so forth. Table 1 offers a brief summary of direct and indirect tax measures that various African countries have introduced to increase their tax bases and include digital transactions.44

Table 1 reflects how diverse the measures are in terms of the nature of the levy, the taxable base, the definition of in-scope activities, applicable thresholds, relevant exemptions, and so forth.

In the next section, we discuss these unilateral measures in detail, and we consider the implications of the measures in the African context.

III. Policy Options for Digital Taxation

Based on an analysis of the unilateral measures discussed in Section II, it is apparent that the digital taxation measures adopted in Africa can be separated into several buckets: DSTs, significant economic presence (SEP) measures, withholding tax regime, and VAT regime. These measures and their impact on the African continent are analyzed in the following subsections.

A. DSTs

The purpose of a DST is to allow a market jurisdiction to compensate for the revenue forgone or not captured via the traditional income tax system. Accordingly, market countries are introducing DSTs on a firm’s gross revenue generated from the market country. From the firm’s perspective, it now faces two different taxes to two different countries — it will have to pay corporate taxes in its home country in addition to the DST on the same income in the market country.45

Globally, a few European countries (including Austria, France, Italy, and Spain) have enacted DSTs to tax digital and platform businesses. However, the specific wording of the legislation varies widely. A few other countries (such as Brazil, Canada, and the Czech Republic) have proposed DST legislation for taxing digital transactions.46 The basic idea behind the DST is to levy taxes on a narrow range of in-scope activities, focusing on the gross revenues of the digital tech companies providing specific digital services. Because DSTs are typically linked to the gross revenue — with specific rules including and excluding particular items — many claim they are turnover taxes. However, unlike most turnover taxes, the burden is not passed on to the service recipient. Hence, some have argued that it is possible to interpret a DST as a direct tax or corporate income tax in disguise.

Table 1. Unilateral Efforts to Tax the Digital Economy in Africa

Direct Measures

Country

Effective Date

Type

Description

Kenya

January 1, 2021

Digital Services Tax

The DST applies to income derived from the digital marketplace, at the rate of 1.5 percent. The charge is based on the gross transaction value received as consideration for the digital service. No turnover threshold is prescribed. The DST is applicable to revenue derived from the following in-scope services:

1. the offering of downloadable digital content, including mobile applications, e-books, and films;

2. over-the-top services, including streaming television shows, films, music, podcasts, and any form of digital content;

3. the sale, licensing, or any other form of monetizing of data collected about Kenyan users and generated from the users’ activities on a digital marketplace;

4. the provision of a digital marketplace;

5. subscription-based media, including news, magazines, and journals;

6. electronic data management, including website hosting, online data warehousing, file sharing, and cloud storage services;

7. electronic booking or electronic ticketing services, including the online sale of tickets;

8. the provision of search engine and automated help desk services, including the supply of customized search engine services;

9. online distance training through prerecorded media or e-learning, including online courses and training; and

10. any other service provided through a digital marketplace.

The DST shall not apply to income that is subject to withholding tax in Kenya (for example, income from management and professional fees earned through a digital marketplace).

Sierra Leone

January 1, 2021

Minimum turnover tax on online and digital transactions

Tax at 1.5 percent applies on the turnover of all digital and electronic transactions.

Withholding tax

Withholding tax of 5.5 percent on residents and 10.5 percent on nonresidents applies to contractors providing digital services.

Nigeria

February 3, 2020

Significant economic presence (SEP)

On May 29, 2020, the Nigerian government retroactively implemented the SEP concept, which provides that a nonresident company with a gross turnover above NGN 25 million has an SEP. It covers four types of digital activity:

  • providing streaming services to or enabling the downloading of digital content by persons in Nigeria;

  • transmitting data regarding Nigerian users’ activities on a digital interface;

  • providing goods and services directly or indirectly through a digital platform; or

  • providing intermediary services via a digital platform, website, or app linking suppliers to customers in the Nigerian market.

A nonresident company will also have an SEP if it uses a Nigerian domain name, registers a URL in Nigeria, or engages in a “purposeful and sustained interaction with persons in Nigeria by customizing its digital page or platform” for Nigeria. Listing prices in naira or offering payment/billing options using Nigerian currency qualifies as customizing the platform for Nigerian users.

The applicable tax rate varies between 10 and 30 percent based on how the taxpayer is categorized for corporate tax purposes.

Tunisia

December 27, 2019

DST

A DST of 3 percent applies when nonresident companies sell computer applications or digital services in the country. No turnover threshold is prescribed. A decree with detailed requirements has yet to be released.

Zimbabwe

January 1, 2019

General income tax on digital services income

A tax rate of 5 percent applies to revenue that platform or satellite broadcasting service providers domiciled outside Zimbabwe receive from persons resident in Zimbabwe, which is deemed to be income accruing in Zimbabwe. A revenue threshold of $500,000 per annum is applicable.

Table 1. Unilateral Efforts to Tax the Digital Economy in Africa (Continued)

Indirect Taxes

Country

Effective Date

Description

Algeria

January 1, 2020

The scope of the VAT has been expanded to include sales of digital services, which are subject to a tax rate of 9 percent. The law does not yet provide registration requirements for nonresident providers.

Cameroon

January 1, 2020

VAT is applicable at a rate of 19.5 percent on sales of goods and services through foreign or local e-commerce platforms, whether business-to-business (B2B) or business-to-consumer (B2C). However, further details have yet to be announced.

Ghana

January 1, 2014

Nonresidents providing digital services to consumers in Ghana are required to register and collect VAT at the rate of 12.5 percent.

Kenya

September 2, 2013

Nonresident vendors that supply digital services to Kenyan consumers must register for and collect VAT.

January 1, 2020

Through the Finance Act, 2020, Kenya imposed a standard VAT rate on certain digital services provided by nonresidents to consumers in Kenya without a PE.

Suppliers making taxable sales must register or appoint a tax representative before the transitional period ends. The scope was further expanded through the Finance Act, 2021, which included sales made over the internet or any electronic network within digital services and marketplaces to include online platforms.

Nigeria

January 1, 2020

According to the Finance Act of 2019, a nonresident supplier is required to register and collect VAT on digital services provided to consumers in Nigeria.

Sierra Leone

January 1, 2021

Goods and services tax is to be collected when facilitating the sale of digital services (including through e-platforms, downloadable content, and so forth) to consumers in Sierra Leone.a

South Africa

June 1, 2014

Nonresidents supplying digital services to consumers in South Africa, including B2C and B2B, must register and collect VAT at the standard rate of 15 percent.

April 1, 2019

The South African government expanded the VAT scope of digital services, increased the threshold for nonresidents to register for VAT from ZAR 50,000 to ZAR 1 million, and implemented intermediary rules.

Tanzania

July 1, 2015

Nonresident suppliers providing digital services to consumers in Tanzania are required to register and collect VAT at the rate of 18 percent.

Uganda

July 1, 2018

Nonresident suppliers providing digital services to consumers in Uganda must register for and collect VAT at the rate of 18 percent.

Zimbabwe

January 1, 2020

Nonresidents supplying radio, television, and digital services to consumers in Zimbabwe must register and collect VAT at the rate of 14.5 percent.

Data Sources: KPMG, “Taxation of the Digitalized Economy: Developments Summary” (July 22, 2021); EY, “Digital Services Tax Jurisdiction Activity Summary” (Apr. 1, 2021); Alvara, “Global VAT & GST on Digital Services”; and The BettsFirm, “Sierra Leone Finance Act 2021: Highlights and Analysis” (2021).

aFor the purposes of this article, we have included GST in our VAT discussion as most of the countries that have adopted a GST have included VAT under the GST regime with just the change in nomenclature. Also, in several countries these terms are used interchangeably.

Notably, DSTs give rise to numerous challenges and often leave various open questions, some of which are discussed below:

  • Coverage of DST under the scope of DTAs: In most cases, DSTs are arguably not covered by article 2 — (that is, the article titled “Taxes Covered” in the OECD model) — of most tax treaties because of various characteristics such as taxable base, taxable object, and the temporal and complementary nature of these taxes.47 From the tax levy perspective, a DST is levied on turnover or transactions — not on income — and hence it might not be covered under article 2 of the tax treaty. While counterarguments exist depending on the wording of the tax treaty, it seems like the arguments for including DSTs in the scope of article 2 are relatively weak in most cases. That said, the mere fact that the technical tax base is gross revenue does not necessarily mean the levy cannot be covered because there are other taxes that are also charged on a gross basis, too (for example, passive income taxation where withholding is done on the gross basis). Notably, Italy issued guidelines stating that the DST will not be covered by article 2.48

  • Impact on working capital: A DST can have a significant impact on an entity’s cash flow because it increases the cost of operation for the MNE when the DST is to be borne entirely by an entity, which holds true in most cases when there is almost no possibility of obtaining a double tax credit. Entities can claim a deduction of these taxes in their profit and loss accounts, but tax arbitrage and double taxation cannot be ruled out completely.

  • Impact on prices and demand: A DST can indirectly increase the price of the digital services, which can affect consumer demand patterns and the flow of cross-border capital and investments. A company that enjoys a monopoly may be able to afford to explicitly pass DSTs on to customers. For example, Google has started levying an additional charge over its prices that it labels a DST levy on operations costs/regulatory operating costs for invoices to its customers in some countries.49 Some other companies have taken similar steps to pass on these costs to the customers. That said, not all companies can afford to pass on these costs to customers. For companies that cannot pass on these costs, there will be an impact on prices and broader pricing policy because of increases in their selling and marketing expenses.

  • Administrative challenges: It is important to note that the complexity of the DST will affect both the tax administration and MNEs, increasing the administration cost for the government (in terms of requirements of additional resources) and compliance cost for the MNEs.

  • Correlation between levy and profitability: Another controversial point that MNEs are raising is that since these DSTs are linked to global revenues of the MNEs vis-à-vis the actual profits earned in the jurisdiction levying these taxes, the MNEs may have to pay DSTs even when the business is not profitable in a jurisdiction. However, from the tax administration perspective, it ensures some contribution of taxes from the MNEs every time the threshold is met because of their operation in a particular jurisdiction.

  • The possibility of DST being discriminatory: Also, multiple other questions are raised regarding these taxes being discriminatory. Reports from the Office of the U.S. Trade Representative explain this position.50 The high threshold and design element may be discriminatory under Article II of the General Agreement on Trade in Services (GATS), discussed in detail in Section IV.

Box 1 analyzes DST and some of the challenges from an African perspective.

Box 1. DST Applied in an African Context

The in-scope activities vary significantly from jurisdiction to jurisdiction, a point demonstrated in the discussion of the Kenyan, Tunisian, and Zimbabwean DSTs in Section II. Also, some companies do not have African country-specific locales/websites. In those cases, it becomes challenging to determine whether indirect forms of revenue (for example, a situation in which a user in an African country clicks or uses a website maintained by the MNE in a country other than that in which the user is located) will be covered in scope and, if it is, how to determine the revenue allocable to the African market country. The basis of calculation varies for different revenue streams and different countries. There may be situations where there is a lack of clarity in determining the revenue allocable to the operations in Africa — that is, whether factors to be used for calculation should be based on users’ click-outs, impressions, geo-sessions, and so forth. At this time, the law is not clear and further guidance with specific illustrative examples would be helpful.

Often, the tax administrations in African countries lack the resources to perform the administrative and computational work involved in DST rules, which are tedious in practice.

These taxes may increase the operational cost of MNEs in Africa. While arguments exist on both sides of the issue, often the taxpayers will not be able to claim relief under tax treaties for DSTs. The burden of the taxes will be passed on to suppliers, and the suppliers will increase their prices because the taxes affect the cost of selling into a jurisdiction. In turn, the small and medium-size enterprises based in Africa that are purchasing the cross-border services will feel the impact. DSTs will also increase compliance costs for the MNEs, which will be required to monitor any developments surrounding digital tax law to act in accordance with them.

To cover the additional expenses, MNEs are likely to shift their tax burden to consumers or regional companies. For example, Netflix revised the subscription plans it offers in Kenya and increased its charges in response to the DST applied through Kenya’s Income Tax Act and VAT Act. A standard package now costs $10.30 (increased from $8.90), and a premium package costs $13.58 (increased from $11.24).a

Further, Africa is a labor-intensive continent that depends on the flow of capital from overseas jurisdictions. The imposition of DSTs may significantly affect the flow of capital and investments into Africa, thereby affecting the entire region’s growth.

aLynet Igadwah, “Netflix Hikes Rates in Kenya on Inclusion of VAT Tax Charge,” The East African, May 7, 2021.

B. Significant Economic Presence

The SEP framework holds that a taxable presence could arise in a jurisdiction when a nonresident enterprise has an SEP based on evidence that the entity had a purposeful and sustained interaction with the jurisdiction via digital technology or other automated means. The concept widens the existing tax base by creating a new nexus, thus attempting to tackle the problems faced by market jurisdictions seeking a fair allocation of taxing rights in the digital world.

Some of the criteria that might be part of the legislative framework and support the creation of an SEP include:

  • total revenue and revenue generated through digital platforms;

  • local user interaction;

  • amount of payments made to nonresident businesses;

  • use of local promotion and marketing functions;

  • number of active users consuming the digital services; and

  • the amount of data the business collects in market jurisdictions.51

To give a practical example, India incorporated a specific provision on the SEP into its domestic income tax law that applies to the in-scope activities once the prescribed quantum of payment and number of users are exceeded. Since the SEP is introduced into the domestic law framework of a contracting state, the applicable rates will be based on the country’s corporate tax rates. However, it is important to understand how the domestic tax framework would interact and interplay with DTAs.

Needless to say, the SEP concept has been criticized on a wide range of other grounds including:

  • Coverage of SEP under the scope of DTAs: The SEP is usually covered in the countries’ income tax laws. Therefore, it may be an eligible tax covered under article 2 of the DTAs depending on the specific wordings of the relevant tax treaty. Hence, in most cases, taxpayers will be able to claim the benefits of the tax treaty if the PE’s nexus rule of PE does not cover activities covered in the scope of SEP. We can take practical examples from India, Indonesia, and Nigeria to understand the impact of the same. All three countries have implemented SEP provisions (though the actual effective date of India’s provision is deferred) in their domestic tax law. But they have not renegotiated their existing tax treaties to incorporate provisions to deal with the allocation of taxing rights to an SEP. Therefore, unilateral tax measures will have no global tax implications for the digital MNEs from countries with which the implementing countries have tax treaties in place. However, some countries have narrow treaty networks; in those cases, the SEP rule would affect the taxation of the MNEs in the jurisdiction.

  • Computation challenges: As mentioned earlier, there are numerous criteria that could be used to trigger the SEP rules. It is difficult to determine what activities should be deemed sufficient to create a nexus under these rules. Second, getting the correct data to determine whether the nexus criteria is met may be challenging for the MNEs. Further, because the computational framework for the taxes paid will be on a net basis (that is, income minus expense approach), the MNE will need to prepare a provisional profit and loss account.

  • The problem of double taxation when there is no tax treaty: When no tax treaty applies, taxability will be guided by the domestic law of the taxing jurisdiction, and MNEs will owe taxes based on SEP rules. The MNE may still be able to claim a deduction of these taxes as expenses where the residence jurisdiction does not provide for the unilateral tax credit.

  • Enforcement issues: There is no uniformity among the different implementation frameworks employed by different countries, which may give rise to interpretational issues. For example, in India’s domestic tax law, the provisions relating to the SEP do not define key terms such as “interaction” and “users,” but understanding these terms is vital to the proper interpretation and administration of the provision. All these lacunae might give rise to significant tax risks, increased tax litigations, and general tax uncertainty. In contrast, some tax jurisdictions have drafted highly technical SEP provisions. For example, Nigeria’s Companies Income Tax (Significant Economic Presence) Order, 2020, provides that a company other than a Nigerian company will be deemed to have an SEP in Nigeria if the company uses a Nigerian domain name (.ng) or registers a website address in Nigeria. It is unclear how this provision would be enforced when MNEs use virtual private networks to access the desired information. It also remains to be seen how the tax authorities will monitor non-Nigerian companies registering websites in Nigeria. The process of identifying those cases will be tedious and require the government to increase expenditures and investment in technologies.

Box 2 analyzes SEP and some of the challenges from an African perspective.

C. Withholding Taxes

Withholding taxes are not a new way of levying tax — they have been part of many countries’ tax systems for quite some time. They are a gross-based tax regime and, in most cases, are not the final taxes. One of the reasons for using withholding tax provisions is to ensure the periodic and smooth flow of tax revenues to the government at the time the payment generating taxable income occurs.

The withholding taxes are embedded in the consideration to be paid to the service provider. Although withholding taxes are charged on a gross basis, they are adjusted against the income of the deductee. Hence, depending on the exact wording and intent of the legislation in place, a withholding tax is sometimes treated as a final

Box 2. Significant Economic Presence Applied in an African Context

African countries have very limited treaty networks. For example, Nigeria, which has implemented an SEP, has a very narrow network of tax treaties. When there is a tax treaty, the SEP rules may not have an impact because of the narrower definition of PE in the tax treaty. In other cases, MNEs will be required to register and pay taxes under SEP rules, which may lead to double taxation.

Generally, the scope of the rule is too broad, and there are interpretational and computational challenges, as discussed in Section III.B. Further, the de minimis threshold is often too low — in Nigeria, it is only NGN 25 million (about $64,500). Also, the wording of SEP rules is somewhat vague and will likely increase the tax litigation risk for MNEs. For example, there are several points in Nigeria’s SEP rules that require additional guidance including issues regarding the rule’s scope.

Further, as more African countries adopt SEP rules, there may be variations in the thresholds, in-scope activities, and other details of the rules. Since the tax is calculated on a net basis, provisional profit and loss statements are required to be prepared.

A different concern is the lack of sufficient resources to undertake the administrative and computational tasks required under SEP rules. For example, Nigeria’s SEP rules will pose enforcement challenges for both businesses and the tax administration. Businesses might not even be aware of the tax obligations in Nigeria, and they might not have the required means to conduct a self-assessment. Likewise, the tax administration cannot simply rely on the information received from the businesses. Instead, they must establish a system that constantly keeps tabs on the income that is made in the country. Both parties will need to invest enormous resources to ensure compliance with and enforcement of the SEP rules.a

aOECD, “Addressing the Tax Challenges of the Digital Economy, Action 1 — 2015 Final Report” (Oct. 5, 2015).

tax. When a taxing state can comfortably exercise its enforcement jurisdiction over the taxpayer — that is, when the taxpayer is a resident or has a physical PE in the state — then the withholding tax provision may not be a final tax. In such cases, the final tax would have to be computed and paid in accordance with the domestic tax laws of the concerned jurisdiction. In other cases, the withholding tax could be a final tax; some jurisdictions follow this practice of treating it as final tax to reduce compliance burdens for nonresidents.

Withholding taxes are imposed predominantly on passive sources of income such as interest, rent, and royalties. Thus, they are often considered a form of source-based taxation, and they represent a partial shift of taxing rights. Withholding taxes are simple and relatively easy to implement. While there is no need to calculate the attribution of income, the relevant transactions need to be monitored. The tax is applied to the respective payments.52

In terms of using withholding taxes as a mechanism for taxing digital transactions, the OECD initially suggested using withholding tax provisions as a backup option for taxing digital transactions in the action 1 final report. However, later reports and public consultation documents did not refer to withholding tax provisions as an autonomous alternative.

In some tax jurisdictions, withholding tax principles are embedded within other measures for taxing the digital economy. Even if a measure is not referred to as a withholding tax, it may still be implemented as one. For instance, although its scope was later broadened, India’s equalization levy initially taxed online advertisements and the provision of digital advertising space. While it was introduced outside the ambit of the country’s Income Tax Act and as a non-descriptive tax, the levy was similar to a withholding tax mechanism — that is, a specific portion of payments made in accordance with nonresident companies’ invoices was withheld in the name of the equalization levy. In a few other countries (for example, Malaysia, Mexico, Pakistan, and Slovakia), the withholding tax provisions are embedded in the domestic income tax laws and used to collect taxes on various types of e-commerce and digital transactions. It is also worth noting that Thailand had proposed using a 5 percent withholding tax to tax e-commerce businesses, specifically the supply of goods and services into its territory.53

When used to tax digital transactions, withholding taxes raise several fundamental concerns, including:

  • Issues with enforcement and collection: Collecting withholding taxes presents immense costs and administrative challenges for a country. These issues might not be pertinent in B2B transactions, but they will exist in B2C transactions. With B2C transactions, the obligation to collect withholding tax falls on individuals. Collecting such small amounts from countless individuals will be challenging for a government and, in most cases, will not be feasible. In the action 1 final report, the OECD suggests that one possible solution is to place the compliance burden on the intermediaries that process the cross-border payments, an alternative that a few countries have implemented in practice. However, this also introduces other concerns. For example, the intermediary might not have the ability to identify every transaction, accurately characterize its nature, and determine whether withholding taxes should apply. Another alternative that the OECD suggests is to have a mandatory registration system for nonresidents that requires them to create a special bank account to keep track of every payment received from residents. The withholding tax could be collected directly from these accounts.

  • Issues with gross-based taxation: Platform-based start-ups and other platform companies spend substantial amounts of capital on product development, and they continuously update their service offerings to maintain a competitive advantage. A withholding tax can become burdensome because it is imposed on gross revenues. In the action 1 final report, the OECD suggests that one possible solution is to fix the tax rate at a relatively low amount to reflect the typical profit margins. For example, the margins could be determined based on a statistical analysis of profit margins gained by domestic taxpayers in similar industries. Further, if applied solely on revenue before expense deductions, a withholding tax could hamper cross-border trade, increase compliance costs, and impede the growth of start-ups.

  • Mismatch with value creation: It is a well-established international principle, recognized in the action 1 final report and elsewhere, that profits should be taxed in the jurisdiction where the economic activity that generates the profits is carried out and where value is created. However, this principle might not always hold true when withholding tax provisions are used to collect taxes on digital transactions.

  • Treaties and double taxation: Even when tax treaties are in place, sometimes withholding taxes may not be covered and double taxation concerns may arise. For example, if a withholding tax is part of an SEP regime, then it may fall under the purview of article 2 in most DTAs. Once DTAs are amended to specifically provide for the allocation of taxing rights in the context of an SEP, the taxpayer would be eligible for a double tax credit for the tax suffered in the market jurisdiction. However, if a digital withholding tax takes the form of an equalization levy, it would not be eligible for coverage under article 2 and might lead to double taxation with no credit available for the tax suffered in the market jurisdiction. For example, India’s equalization levy is not part of Income Tax Act, 1961, and is not covered in its tax treaties. Separately, there may be situations in which an overall loss position in the parent jurisdiction means an MNE cannot utilize a tax credit in full. Further, many jurisdictions usually do not provide for the carryforward of such foreign withholding tax credits or in some cases jurisdictions put some limitations on carryforwards. In those cases, the MNE can only deduct the taxes paid in foreign jurisdictions as expenses in the profit and loss statement, which may not give them a benefit equivalent to that of a situation in which would have been available if they had been able to claim credit. Imposing these costs on growing but not yet profitable businesses can have a detrimental effect on growth, impede job creation, and ultimately limit tax revenues. Notably, the U.N.’s model treaty offers the new (and optional) article 12B — often referred to as the U.N. digital tax proposal — which seeks to ensure that digital tax levies are credited.

  • Issues with the distribution of credit: Sometimes an MNE’s profit is allocated among different group entities based on some transfer pricing guidelines or adjustments; however, in such cases, withholding tax liability remains with the payer entity. Suppose 15 percent of a 25 percent profit margin remains with the recipient entity after paying appropriate intragroup charges. The other 10 percent of the group profit margin will still be taxed elsewhere without credit, as typically the law restricts the availment of credit to the same taxpayer — that is, credit can be claimed by the taxpayer in whose name the withholding tax was withheld.

Box 3 analyzes withholding taxes and some of the challenges from an African perspective.

Box 3. Withholding Taxes Applied in an African Context

Withholding taxes can provide tax collection certainty for African countries that have limited administrative capacity. It may be an easy and straightforward solution in some situations that does not require MNEs to register separately, but it creates double taxation risks when withholding tax is non-creditable. Notably, this is true when there is no tax treaty, and many African countries have narrow tax treaty networks.

At times, the rules on calculating withholding tax credits are very complex. The documentation and compliance requirements for obtaining a withholding tax credit can be very burdensome, especially when withholding is used as way to tax digital transactions. Most young and growing companies do not have the resources to comply with these requirements and will not obtain a tax credit. Further, levying these taxes on a gross basis may restrict the start-up ecosystem and limit companies’ ability to expand their operations on the continent. They may worry that their working capital will be limited by these taxes if they cannot generate profit from the continent.

If African countries use withholding taxes to tax digital transactions, thresholds and in-scope activities should be defined clearly to avoid interpretational challenges and litigation issues at later stages. It would be helpful for countries to periodically issue FAQs addressing taxpayers’ concerns.

Notably, withholding taxes can be ineffective when cryptocurrencies like Bitcoin, which are gaining popularity in many African countries, are used to make cross-border payments.

Collecting small amounts of withholding taxes from large numbers of individual consumers would be administratively challenging. Thus, withholding taxes may work better in B2B situations.

D. VAT

VAT is an indirect, largely consumption-based tax. Origin-based VAT mechanisms tax a transaction partly in the state where it is created or originates, while a destination-based VAT is partly applied in the state of consumption. Most VATs are levied in the market jurisdiction, applying the destination principle, especially when the VAT is imposed as a means of tax collection on digital services. Some states impose indirect taxes on the consumption of digital services to raise revenue while also equalizing the conditions of competition between local and foreign suppliers of digital services.54

The modalities of these rules may vary. In the context of digital transactions, specific VAT rules are usually made to cover B2B and B2C transactions because they differ substantially in nature, volume, characteristics, and the appropriate degree of enforcement. When VAT is imposed on the consumption of digital services in the B2B sphere — that is, when the recipient is registered for VAT — the tax liability is often discharged through a reverse charge mechanism applied by the service recipient; in turn, the business entity that receives the digital service takes a credit for the tax payment. B2C business models pose more difficulty, especially in the internet era when customers are scattered across jurisdictions. It would be incredibly demanding for the government to exercise its enforcement jurisdiction on each customer. Thus, when nonresidents supply goods or services to consumers or unregistered businesses, many VAT systems require the nonresident to register for and dispose of the VAT liability in the jurisdiction. For example, in the B2C context, the VAT systems in Ghana and Nigeria require the nonresident digital service provider to register for, collect, and remit the tax to the government. Notably, some countries like South Africa place the responsibility for discharging VAT on the nonresident supplier in all cases.

If imposed correctly, a destination-based tax mechanism can be a good approach that avoids affecting the existing system.

It’s an approach that the OECD supports in its 2017 international VAT/GST tax guidelines, which cover the intersection of VAT and the digital economy and which have been endorsed by more than 100 jurisdictions.55 The OECD suggests that tax authorities, in close coordination with suppliers and their financial intermediaries, set up a separate, simplified VAT registration system to facilitate remote registration.

However, criticism does persist. Concerns regarding the use of VAT in the digital context include the following:

  • The possibility of double economic taxation: Implementing VAT on digital transactions along with another form of tax targeting similar transactions would result in the same transaction getting taxed twice. For example, India levies GST on online information database access and retrieval (OIDAR) services, a category of services provided by the supplier through the internet with no physical interface and received by the service recipient online.56 India also imposes a direct tax in the form of an equalization levy that specifically targets digital transactions when payment is made to a nonresident e-commerce operator. Occasionally, the GST on OIDAR and the equivalization levy overlap, potentially giving rise to economic double taxation. Even though the government introduced the two laws for different reasons, the overlap is evident and complying with both could be a costly affair for the digital MNEs.

  • Issues with income characterization and possibility of tax disputes: Characterization of income is extremely important for VAT and GST. Often VAT systems are built so that the applicable tax rate for a transaction depends on its nature or character. Incorrect characterization can result in erroneous tax rates, which could lead to disputes and tax-related litigation.

  • Administrative and enforcement challenges versus simplicity and flexibility: VAT legislation varies widely in different jurisdictions, resulting in a system that lacks uniformity and comparability. Also, there are no tax treaties to provide VAT relief when there is an issue of double taxation. These factors could contribute to tax uncertainty, increase administrative burdens, and create double taxation risks.57 Unless reverse charges cover the transactions, imposing VAT on digitized business models could mean MNEs have to register, charge, collect, and account for VAT in numerous jurisdictions and also file periodic tax returns with many tax administrations.

Meeting VAT compliance obligations and registering with multiple jurisdictions can be quite burdensome. Hence, simplified registration and compliance regimes would be helpful and welcomed.

To simplify multiple tax obligations, the Indian government implemented a single Goods and Services Tax Act, 2017, which took effect July 1, 2017, and subsumed most of the indirect tax laws that existed in India. Similarly, Australia implemented a single GST law, with effect from July 1, 2000.

The EU’s electronically supplied service rules provide another useful example. The European Commission amended the VAT regulation and changed the point-of-service rules to focus on the place of consumption. The rules allow offshore companies to register with a single member state that collects taxes and remits the taxes due to all member states. Different tax administrations have invested substantial resources in this project, but it has eased collection and compliance obligations for those doing business in EU member countries.

There are other steps that can be taken to administer the VAT in a better manner and make it more widely acceptable.

In a VAT regime, businesses act as VAT collectors, which means their knowledge and understanding of models, systems, and processes are key to delivering an efficient and cost-effective regime. Thus, it would be helpful for the government to consult with businesses on these issues. Also, providing sufficient lead time between the date a regime is announced and the date of implementation is key.

There is an interactive relationship between trade and taxation. Therefore, it is generally recognized that trade agreements also affect tax measures to the extent that both affect the free circulation of goods, services, capital, labor, and technology. The best practice is to ensure that unilateral tax measures are compatible with the other domestic tax regulations, DTA obligations, and generally accepted rules of international commerce. Discussions have arisen regarding the interaction of unilateral digital tax levies with various other international laws principles.

Box 4 analyzes VAT and some of the challenges from an African perspective.

Box 4. VAT Applied in an African Context

In Africa, the VAT forms the most significant share of the continent’s tax revenue.a This is evident by analyzing the strong and positive correlation between VAT and tax revenue in a few sample countries in the year 2017, for which the average correlation coefficient is around 0.83. This means these taxes are not new to the region, and there are some systems in place for it.

Figure 4. Composition of Tax Revenues Across Africa

Nonetheless, applying VAT to digital services may be challenging for MNEs. MNEs will be required to register for, charge, and remit VAT. They will also need to comply with reporting obligations. Because some countries in Africa offer small markets, some MNEs may choose to avoid these responsibilities and simply not operate in jurisdictions if they cannot turn a profit from these jurisdictions.

In South Africa, the Taxation Laws Amendment Act (Act No. 31 of 2013), requires nonresident entities that supply electronic services to recipients in South Africa to register and account for VAT if their supplies in the country exceed ZAR 50,000 in a month. South Africa was only the second country in the world to implement this model of VAT reform for digital transactions, and few relevant statistics are available for the period June 2014 to September 2017.b Similarly, in August 2019 the Kenya Revenue Authority announced that the provision of online platforms for use by third parties is a taxable supply under the 2013 VAT Act.

The administrative procedures associated with VAT could pose a big challenge for emerging economies, including many in Africa, since they lack resources for strict, proper enforcement of these tax measures. A 2018 study by the United Nations Economic Commission for Africa reviewed VAT data from 24 African countries and found a tax gap of 50 percent or more in 12 of them.c The main reasons cited

for the gaps were compliance challenges, limited enforceability, and policy issues like the large volume of exemptions. Hence, if VAT is adopted as a means of digital taxation, there should be proper systems in place to administer these taxes.

VAT registration will impose challenges on suppliers as well. Each jurisdiction may have its own rules and procedures for registration. There may be differences in filing requirements, deadlines, tax rates, invoicing requirements, and even languages. Every jurisdiction will impose these burdens, which will significantly increase the compliance burdens for the suppliers.d

To ease the compliance burden, it would be helpful for the African countries to act together through a regional body like the ATAF to implement a single, unified approach for collecting VAT on digital transactions in Africa. In this regard, it would be useful to examine the EU’s VAT regulations concerning the definition of a taxable supply of electronic services and the implementation of a mini one-stop-shop (MOSS) procedure. The MOSS system simplifies practices in the EU since — rather than requiring individual VAT registrations for each country — VAT declaration and payment can be made to the tax office in the member state in which the company is established. However, having a system like MOSS in Africa would be much more difficult in practice since the continent is not bound by one overarching legal system; it would likely require the enactment of new legislation or treaties between the individual African countries concerned. Still, the possibilities of such a framework are worth analyzing.

Like in India, where both the OIDAR rules under GST and the equalization levy may apply to some transactions, some African countries such as Kenya and Zimbabwe also impose different tax measures — for example, both a DST and VAT — on the same digital transactions. This might lead to double taxation.

MNEs may shift the tax burden to consumers when a reverse charge is permissible. For instance, Facebook has already announced that it will be charging VAT to Zimbabwean advertisers because of Zimbabwe’s decision to impose VAT on digital companies.e Similarly, one of the reasons why Netflix revised the cost of some subscription packages was Kenya’s application of VAT to digital services.f

aATAF, African Tax Outlook (2019).

bSouth Africa Department of National Treasury, “VAT on E-Services: South Africa’s Experience” (2018).

cWhy Tax Collection Remains a Challenge in Sub-Saharan Africa,” EY Global, Aug. 22, 2020.

dOECD Centre for Tax Policy and Administration, “Transfer Pricing Aspects of Business Restructurings: Discussion Draft for Public Comment, 19 September 2008 to 19 February 2009” (Sept. 19, 2008).

eNyasha Nigel Machiri, “Facebook to Charge Zimbabwe Advertisers VAT,” MNE Tax, Aug. 25, 2020.

fLynet Igadwah, “Netflix Hikes Rates in Kenya on Inclusion of VAT Tax Charge,” The East African, May 7, 2021.

The next section briefly considers the relationship between the new tax measures and WTO policy.

IV. Relationship to WTO Policy

A. Tax Disputes and the WTO Generally

Member states agree to curtail some elements of their sovereignty when they join the WTO and sign its agreements to promote free trade. For reasons including the desire to maintain tax sovereignty, taxation matters are usually kept outside the ambit of the WTO. Still, it is not uncommon for concerns regarding taxation to be raised at the WTO.

Under the General Agreement on Tariffs and Trade 1947, 22 out of 124 initiated disputes concerned tax measures of member states. Before the dispute settlement body of the WTO, 38 out of 402 initiated disputes were tax disputes.58 Taken together, 39 of the disputes involved indirect taxation measures, and 21 were about direct taxation measures. In the case of “United States — Tax Treatment for ‘Foreign Sales Corporations,’” the dispute settlement undertaking clearly stated that “a Member of the WTO may choose any kind of tax system it wishes — so long as, in so choosing, that Member applies that system in a way that is consistent with its WTO obligations.”59 Further, the line of the WTO rulings in the cases involving the U.S. incentives for exporters contributed toward the consensus that the WTO’s prohibition of export subsidies can be extended to find that the use of income taxes to create a subsidy violates WTO rules.

To invoke the WTO’s jurisdiction in tax matters, member states can claim that there is a violation of WTO obligations, such as a violation of the most favored nation clause, nondiscrimination rules, or the national treatment principle.

B. Unilateral Measures Under WTO

Multilateralism and cooperation are considered the fundamental bases of the WTO principles. Unilateral measures are inconsistent with the spirit of the WTO and inconsistent with the WTO obligations.60 Under article 23 (strengthening of the multilateral system) of the WTO’s Understanding on Rules and Procedures Governing the Settlement of Disputes, member states shall not adopt any unilateral measures and seek redress only in accordance with the dispute settlement undertaking. When interpreting this article, a WTO panel report observed that “by its very nature, the WTO Agreement favours a multilateral approach to trade issues.”61

Because of the delay in the global consensus on taxing the digital economy, the measures taken thus far are unilateral measures. While these measures were implemented in the context of the international tax framework, the results of these measures could still violate the WTO obligations and be brought within the WTO ambit.

The GATS is the only WTO agreement that refers to DTAs, and Article XXII thereof limits its application to matters that are not covered by DTAs. However, this might not protect the digital tax measures from WTO scrutiny because the majority of the unilateral actions are not part of the DTAs. The WTO has not been approached to decide whether the unilateral measures are acceptable. Regardless, it is essential to be aware of the WTO’s potential perspective on digital taxation and the possibility of its intervention. The following sections discuss the temporary moratorium that the WTO has imposed regarding tariffs on digital services and African countries’ opposition to the same. They then proceed to analyze the unilateral measures introduced by African countries and the possible application of the WTO rules.

C. Moratorium on Digital Customs Duties

In 1998 the members of the WTO committed to a temporary moratorium according to which the member states shall not impose any customs duties on electronic transmissions. This temporary moratorium has been continuously extended by the member states. However, several disputes have emerged recently as a result of the rapid growth of the digital economy. Notably, developed countries tend to interpret electronic transmissions to include every digital product and service. In contrast, developing countries restrict the interpretation to only electronic carriers — for example, physical CDs and hard drives are included but not the content itself.62

There has been an ongoing debate between the developed and the developing countries regarding whether the moratorium should be adopted permanently, with the former pressing for a permanent ban and the latter fighting for the removal of the ban. The WTO estimates the moratorium is responsible for a global tariff loss of $756 million each year, with developing countries facing 92 percent of that loss.

India and South Africa have strongly opposed the temporary moratorium. In 2018 the countries submitted a joint communication opposing the ban and raising concerns from the perspective of developing countries.63 The countries stated that developing countries are bearing most of the losses and noted that disagreement regarding what constitutes electronic transmission has made it yet more difficult for developing countries to impose any tariffs on the digital economy. In 2019 the two countries submitted another communication reiterating their concerns.64 The 2019 communication highlighted the concern of revenue loss, referring to 2017 when developing countries lost $10 billion and sub-Saharan Africa lost $2.6 billion as a result of the moratorium. Nigeria is one of the top six affected countries with a loss of $580 million.

D. Possible WTO Disputes

The treatment of unilateral digital tax measures under the WTO rules depends on various factors. For instance, the WTO rules applicable to direct and indirect taxes differ. Typically, indirect taxes such as VAT are considered to fall within the ambit of the General Agreement on Tariffs and Trade, and they are generally considered accepted practice when it comes to taxing e-trade.65 Ten African countries — Algeria, Cameroon, Ghana, Kenya, Nigeria, Sierra Leone, South Africa, Tanzania, Uganda, and Zimbabwe — have extended their VATs to nonresidents importing goods or services into their countries.

However, in terms of the application of WTO rules, the uncertainties mainly concern the DSTs imposed by Kenya and Tunisia. The treatment accorded by these DSTs differentiates between residents and nonresidents and because they are not included in the DTAs, they appear to be excluded from the tax carveout available under Article XXII:3 of the GATS.

Some of the questions that may arise when analyzing digital tax measures — especially DSTs — from the WTO perspective are discussed below.

1. Direct or Indirect Tax

The applicability of WTO agreements varies based on the nature of the tax. In the case of indirect taxes, the relevant WTO agreement is the GATT. Indirect taxes such as VAT that apply to both domestic and imported products will be subject to Article III of the GATT (“National Treatment on Internal Taxation and Regulation”). Essentially, the country must ensure that like products are taxed similarly. The GATT may apply to direct taxes in some circumstances, such as when higher direct taxes or internal charges are applied on like imported products.66 The temporary moratorium applies to custom duties, but none of the digital taxes introduced by the African countries have been challenged to be like custom duties.

Several countries have not yet classified the nature of DSTs. While these taxes are used as a substitute for the non-application of corporate tax on digital MNEs, they share some features with indirect taxation. Nonetheless, in addition to GATT, the Agreement for Trade in Services becomes more relevant in SEP and DST systems.67 Article 1:2(b) of the GATS defines services as the supply of service “by a service supplier of one Member, through commercial presence in the territory of any other Member.” Therefore, both measures can be brought within the ambit of the GATS since they rely on either an economic presence or a commercial extraction from the country. Also, since these taxes are only levied on nonresidents, Article II (“Most Favored Nation Treatment”) and Article XVII (“National Treatment”) of the GATS can be invoked. However, countries can still plead for an exception under Article XIV (“General Exemptions”) of the GATS to prove that the differential treatment was necessary for the effective and equitable imposition or collection of direct taxes.

2. Coverage of DTAs

Another issue specific to DSTs is whether the measures are covered by DTAs. As noted, the GATS allows a tax carveout under Article XXII (“Binding Arbitration”) if the measures are covered under DTAs, and those measures cannot be challenged by the member states under either Article XXII or Article XXIII (“Dispute Settlement Undertaking”). Alternatively, measures cannot be challenged if they fall outside the commitments made by a country under Article XVIII (“Additional Commitments”).

All the African countries with DSTs have imposed them unilaterally, and none of them have amended their DTAs to cover DSTs. This means the WTO could adjudicate the propriety of DSTs, which could cause an overlap between the different frameworks.

3. Like Services and Discrimination

WTO agreements apply and can be invoked when like imported services are treated discriminatorily. The DSTs can be challenged because they only apply to like services provided by nonresidents and the application does not fit any of the allowed exceptions. For instance, the DST will apply to Uber and not Little Cab, even though they provide like services and are direct competitors.

When like services are treated differently, it results in discrimination. The DSTs have been criticized, especially by the United States, for allegedly being discriminatory. Although it does not operate under the ambit of the WTO, the U.S. Trade Representative (USTR) initiated investigations into DSTs imposed by several countries and found them to be discriminatory in nature. While the USTR has not initiated any investigations against DSTs put in place by African countries, the taxes are similar to those it has investigated, so similar treatment can be anticipated. Though the WTO prohibits the USTR from initiating unilateral investigations under its agreement,68 the grounds raised in USTR investigations are relevant to the WTO’s work. Further, DSTs imposed on a gross basis operate like tariffs and negatively affect cross-border trade. The implications of violating the WTO’s rules of discrimination are broader than those under article 24 of DTAs based on the OECD model, which only forbid discrimination against foreigners compared with domestic persons.

While this section focuses on WTO agreements, countries can also challenge unilateral digital tax measures using other trade agreements, such as preferential trade agreements.69 Discussions regarding taxation are not restricted to the international tax regime — overlap with international trade practice is likely. Especially when the measures are not included in DTAs, concerned countries may invoke the jurisdiction of trade forums that can make binding decisions. Though African countries’ digital regimes have still not been subjected to scrutiny by developed countries, such as the United States, the similarity between their measures and the measures that have been challenged makes scrutiny likely. Hence, the above proposals need to be analyzed according to the WTO agreements. A broader and deeper examination than that attempted here is necessary to ensure that any law that seeks to tax nonresidents complies with obligations of WTO member countries under their trade agreements.

V. Implications for MNEs

While investment decisions are based on various tax and nontax factors, in general MNEs want to expand in jurisdictions where the laws provide them with an atmosphere of transparency, simplicity, reliability, and stability rather than jurisdictions marked by economic and political instability. In contrast, complex and unavailable procedures; differences or inconsistencies in the tax administration’s instructions, opinions, and procedures; a lack of transparency; and inequitable tax case law and tax dispute resolution mechanisms are among the tax risks and costs that MNEs are least prepared to tolerate when choosing where they will expand operations.

The points discussed in Section III concerning digital taxes have a considerable impact on MNEs operating in Africa as unilateral tax measures often lead to tax uncertainty and may discourage investment,70 especially when the continent’s market may not offer MNEs substantial revenue at first. Companies may be required to incur additional out-of-pocket tax costs as a result of some unilateral tax regimes (for example, DSTs). Thus, it might affect their routine working capital and fund flow and eventually affect the transaction prices. MNEs may have to incur huge operating expenses to set up a footprint in a jurisdiction, the digital taxes may add to that burden, and MNEs may decide to quit the market. There can also be a substantial risk of double taxation with these unilateral tax measures unless the MNE’s home country provides for unilateral tax relief through its domestic tax law, which is not the case in most instances.71

Ultimately, some MNEs may wish to gross up the taxes with the total service payments, which will then be borne by the customers — that is, by African companies or individual African customers. This increases the cost of digital services and might have a negative effect on the MNE’s global market, business, and consumers.

Further, companies need to consider the interaction between digital taxation reforms and transfer pricing provisions when an MNE renders a digital service to an associated enterprise in Africa. Performing transfer pricing studies for digital transactions might be a tedious exercise for MNEs because of the inherent uncertainties on a range of factors such as the nature of digital services; identification of the core value drivers; valuation of data; and the identification and analysis of development, enhancement, maintenance, protection, and exploitation for intangibles.

MNEs must also be wary of the possible interplay between direct and indirect tax provisions when the MNE sells goods online. This interplay might pose challenges including in terms of the event triggering taxation, the object and purpose of the legislation, valuation of goods, and the possibility of double taxation. Failure to pay attention to these interplays might lead to tax disputes. Compliance costs are also a key concern when MNEs are attempting to transact business in Africa. Even concerning withholding taxes, liability may be imposed directly on the MNEs if the private consumers do not comply with the tax regulations. MNEs will bear enormous administrative costs if they are mandated to ensure compliance by private consumers. To efficiently monitor compliance, MNEs may need to incur huge costs in terms of staff training and be forced to appoint representatives in every jurisdiction, which is evidenced by the ATAF’s proposal to have MNEs appoint local representatives.72 Complex tax rules such as SEP regulations might force MNEs to invest substantial amounts of money in information technology infrastructure and the human resources needed to monitor their legal obligation and compliance.

VI. Emerging Business Models

As stated in Section I, business models are rapidly evolving with the emerging internet of things, artificial intelligence, blockchain, collaborative economy, and other technological advancements. In particular, the growth of “as a service” models in which companies sell the outcome of a product rather than the product itself highlights the greater demand for value propositions and changes in service offerings over time.

In its interim report on the tax challenges of the digital economy, the OECD highlights four business models: multi-sided platforms, resellers, vertically integrated firms, and input suppliers.73 However, with many established digital companies having multiple business lines, it becomes difficult to differentiate between multi-sided platforms and other digitized businesses.74

A. Collaborative Platform and Sharing Economy

Digitalization has given rise to a new era marked by the collaborative platform and sharing economy. Most new businesses use platforms and technology to connect people, organizations, resources, and various actors within their ecosystem. These businesses have overtaken the pipelines model because of their ability to add new sources of value creation, supply needed tech solutions, and invest heavily in data analytics.

Platform businesses amplify positive network effects that create value and try to mitigate negative ones that can diminish the value of the network to others and slow the system down. For a platform to succeed, cross-side effects must be taken into account and managed accordingly. A wide variety of platforms may be trying to solve users’ different pain points, including online shopping for goods, online ordering for restaurants, online rental services, logistics, and courier services.

In the following subsections, we briefly discuss two platform- and data-based business models and the relationship between varying structures and digital taxes when businesses have footprints in African countries that have adopted unilateral digital taxes.

B. Digital Media Streaming

1. The Business Model and African Presence

Digital media streaming service providers allow users to stream videos on a wide range of internet-connected devices. Their revenue models differ in terms of rental and subscription plans. These companies give their users legal access to a large movie and television database, offer personalized suggestions based on proprietary algorithms, and provide seamless streaming services without the interruption of advertisements in a wide range of jurisdictions. Revenues from consumers are directly or indirectly earned in different jurisdictions, but the revenue is booked in another jurisdiction. This raises questions regarding how the MNEs classify and account for revenue locally and highlights the challenges of taxing the digital economy.

Consider Netflix, a company founded in 1997 that offers some of the most popular streaming services and has the largest global subscription pool. The company has expanded its footprint and is now accessible in over 190 countries worldwide. In 2013 Netflix diversified and entered the content production industry.

This gave them the edge that paved the way to becoming the leader in the streaming market, moving from a DVD-by-mail and online streaming platform to a company producing its own highly successful content as shown in Figure 5. Netflix became more popular than competitors Amazon Prime and Hulu. The company is estimated to have some 203.67 million subscribers worldwide.75

Figure 5. Netflix and Its Revenue Streams

Netflix entered Africa in 2016; by the end of 2020, it dominated the streaming services market in the region. It has a dominating 57 percent share of the African subscription video-on-demand market. In 2017 the number of Netflix subscribers in Kenya was just 3,600, and it ended 2020 with a 700 percent increase in the number of subscribers — that is, a total of 29,500 subscribers. Netflix is one of the most visited websites in South Africa, with nearly 23.09 million visits per month. An analysis of the company’s business model is essential if Netflix wishes to maximize its continued growth in Africa.

In some countries, Netflix operates on the following model76: The entity in a particular jurisdiction is regarded as a local distribution entity that operates primarily as a limited risk distributor. This entity enters subscription contracts with end-user subscribers and strategic partners. Usually, another entity is responsible for the content to be streamed online. This other entity usually purchases the license for content. Once the license is purchased, it is immediately transferred to an offshore entity. The entity in the local jurisdiction never takes ownership of the licensing rights, but it receives an arm’s-length remuneration for its functions. A business model canvas of Netflix is shown in Figure 7. (Reference can also be made to the company’s Form 10-K.)

Figure 6. Business Model Canvas of Netflix

Thus, from the value chain perspective, the company’s input is content that is either self-produced or purchased from third parties. The final product is the personalized platform for intangible products. Technology development is part of the group of activities — for example, research and development, data mining algorithm, data center cost, and media design — that provides the platform’s customized recommendation system and improves user experience. Technology development is also very important for the decision-making process for new original content creation.

Value creators for Netflix and similar entities include77:

  • The platform interface: The Netflix platform that enables subscribers to watch content online.

  • Technology: The technology used by the company to manage the uninterrupted provision of services.

  • Data: User data plays an important role for these businesses. For example, Netflix generates and analyzes data from the viewing patterns of end-users, and these data assist the service provider in identifying end-user preferences and, accordingly, modifying or adding content to enhance customer viewing satisfaction. Thus, Netflix uses big data to customize offerings for its streaming services, and it is also used in the production of content. Data is a core component of the business models.

  • Content: Different digital media streaming providers stream different content. Developing and suggesting user-specific content based on an analysis of user data helps make a particular streaming platform popular. Localized content is also provided depending on the user’s location and language.

2. Applying Unilateral Digital Tax Measures

Consider taxation based on SEP. If the company has a digital presence in Africa, the assets of the digital PE in Africa would include user data and some portion the software and hardware (for example, the website) that enables the MNE to collect user data and provide a customized service. In these situations, it will become too difficult to determine the amount that should be allocated to the revenue and cost when preparing a profit and loss account for this digital PE. The other challenge is what factors should be used to determine the amount of revenue to be allocated to the PE because the criteria specified in the definition are wide, and the law is not that clear.

Similarly, there may be multiple issues when it comes to applying VAT. These MNEs will be required to comply with multiple registration requirements in each of the operating jurisdictions for the B2B and B2C segments. Even in cases when VAT in the B2B sector is applied using a reverse charge, the company would still need to register when engaging with individual customers. Further, VAT is usually determined based on the address of the service recipient — that is, the place of supply rule — and these companies have to rely on the address that the user provides.

Under the DST regime, it may be challenging to determine the basis on which digital taxes will be levied. Most likely, it would seem to be subscribers. However, subscribers may subscribe to the product in one jurisdiction and use it in another jurisdiction. Also, having the option to buy multiple user licenses with a subscription would make it more challenging to determine how revenue should be allocated to a particular jurisdiction if the users are spread over multiple jurisdictions. Similar problems may arise because of the use of virtual private networks.

Because the number of individual subscribers of such online streaming is high, levying withholding taxes would not be a workable option for the reasons discussed in Section III.C.

The biggest question that remains is what would be the optimal method for levying taxes on businesses like Netflix. In the past, different countries have raised concerns about the distance-selling model used by Netflix. According to some reports, Netflix billed all international customers from the Netherlands and did not declare any income in other European countries.78 Later, customers in some places received emails stating that they would be billed from local service entities going forward.

C. Social Media Networks

1. The Business Model and African Presence

Social networking sites are online platforms that provide new venues for users to create a public profile and connect with other users on the websites. They typically provide a variety of ways for users to interact, such as chat sessions, messaging, email, video, voice chat, file sharing, blogging, and discussion groups. Advertisers often target these social media platforms and buy advertising spots on them because of the variety of audiences.

Consider Facebook, which is among the biggest social networking platforms. The company’s offerings include its eponymous social networking site; a pair of messaging apps, Messenger and WhatsApp; and the photo and video-sharing tool Instagram. Facebook also operates Oculus, which allows users to connect through a virtual reality ecosystem.79 Overall, the platform helps connect people around the world, and users themselves create a lot of content.

Facebook has become an essential medium for connectivity in Africa. This is evident from its user base, which grew from 17 million users in 2010 to 210 million users in 2020, the latter figure constituting 16 percent of the region’s population. The company is building a 37,000-kilometer underwater cable around the continent to increase internet connectivity and attract more users.80 In South Africa, Facebook.com is the second most visited website (following Google.com) with 195.6 million visits every month.

Its significant presence in the region makes Facebook a suitable platform for advertisers, which can target users on the platform and get a direct conversion on sales with less cost. Facebook monetizes big data — for example, it analyzes user data to reveal patterns, trends, and associations, especially relating to human behavior and interactions — to sell to goods and service providers and enables targeted advertising. User data and user engagement are critical to this model. The platform owner sells data, typically organized based on specific user interests, and hosts advertisements for advertisers on its website. Thus, the platform provides different ways to advertise, including by promoting content in news feeds and promoting trends and commercial user accounts.81 In return, the platform owner would receive advertising fees from such third-party advertisers. (See Figure 7.)

Figure 7. Business Model of Facebook

Figure 8 presents the business model canvas for Facebook.

Figure 8. Facebook’s Business Model Canvas

In its financials, Facebook segregates its revenue into two segments: advertising and other. Advertisers pay based on the number of impressions delivered or other measurable user actions such as clicks. Further, the price per ad depends on factors including but not limited to the marketer’s interest in information about the user and information shared by users, geography, ad relevance for the user, and other aspects as determined by Facebook’s matching algorithms. Facebook delivers ads to users based on user profiles, using its algorithms to target the marketers’ desired audience and deliver a measurable return on interest for the company’s marketing spend. Advertisements are often shown differently depending on geography, demography, events, interests, keywords of the content, and device type. The cost of the platform consists primarily of R&D cost, cost of revenue, and marketing costs.

The other segment of revenue usually includes income from the delivery of consumer hardware devices, fees that the company receives from developers that use its payments infrastructure, and various additional sources such as the Oculus business.

Facebook has subsidiaries in different jurisdictions. Figure 9 illustrates how a typical local subsidiary in the social network sector operates.82

Figure 9. Example of How a Social Network Business Model Works

Key value creators in social network businesses include83:

  • Platform network effects from users: Facebook, an internet-enabled value network, brings individuals together in a social capacity and allows advertisers to target specific user groups. In a value network, value is generated through the action of linking.84 The value grows as the platform continuously collects, analyzes, and stores big data and performs data analysis. Facebook takes advantage of positive externalities, making internal use of the data to enhance network effects and lock users into the platform. Network effects, in a way, make social networks more attractive, creating self-reinforcing feedback loops that grow the user base without much additional effort by the platform businesses.

  • Technology: Technological development is significant to value creation activities to maintain the platform and improve the user experience to lock in the users with the platform while collecting and analyzing the big data to monetize by providing targeting advertising. Facebook spends much on the R&D side to keep on track with the evolving technological advancements and beat its competitors.

  • User preference data: Big data, data mining, and process mining are the central features of the business.85 The more users that participate in the network and provide their data, the more valuable and interesting the service is to advertisers. User-generated content also increases with an increase in big data. Investment in advanced technology is required for efficient data analysis, leading to refined user preferences and better-targeted advertisement. Key activities include platform development, content management, and data center management.

2. Applying Unilateral Digital Tax Measures

If social networks were to pay taxes under an SEP regime in African countries where they don’t have a physical presence, the first step in allocating profit to the SEP is applying a separate-entity approach and preparing profit and loss accounts. As with streaming services, challenges will arise in determining the appropriate basis for revenue and cost allocation.

Because the companies have advertisers located in multiple jurisdictions, challenges would also arise in complying with the VAT laws of multiple jurisdictions. Sometimes one company in the group may buy an advertising slot but another company in the group may actually use it. In those cases, VAT will be levied based on the address provided.

If a DST were applied to the social network business model, problems might arise in terms of determining the right data to rely on — that is, using clicks or impressions based on geographic areas or other factors. Another issue is that data generated by user interaction would not be of much use without using data mining, process mining, or analytics to reveal potential patterns, trends, and associations and selling the conclusions to specific advertisers. Does this warrant a higher allocation of profits to the jurisdiction in which the technological tools are used to analyze data?

Levying withholding taxes may be a more workable option in this business sector because most of the advertisers are businesses rather than retail customers. The other challenges of withholding taxes discussed in Section III would nevertheless subsist.

VII. Digital Taxes in a Growing Market

Africa is a debt-loaded continent, and COVID-19 has made this debt burden an even bigger concern.86 In April 2020 the African Union released research on the economic impact of COVID-19, forecasting in April 2020 that the continent could lose up to $500 billion and that countries may need to borrow heavily to survive after the pandemic.87 IMF research indicated that debt-to-GDP ratios in the sub-Saharan region have doubled over the last decade to 57 percent.88 If the pandemic reduces global bond investors’ interest in refinancing deficits, some states may struggle to refinance. Also, demand in African countries is limited because of low international standards and limited technological advancements, which affect the proper deployment of the available resources. Taken as a whole, these economic conditions indicate that the African economy relies heavily on foreign direct investments (FDI) and is in dire need of innovative technological solutions to ensure optimum utilization of its rich and abundant natural resources.

Regardless of these economic issues, Africa’s digital economy is sustaining and is rapidly evolving. Africa is a developing continent with an average national budget deficit of 4 percent as of 2020 (the recommended norm is 3 percent). From a fiscal policy standpoint, problems with domestic resource mobilization can, to an extent, be cushioned by the rising tide of rapid digitalization.89 Technology can provide a range of solutions to various development problems, and new offerings can increase access to services, enhance productivity, and improve livelihoods.90 This increases the pressure to enhance digital innovations on the continent, which requires a substantial amount of work. Harnessing the power of digital developments and the region’s demographic dividends will help the continent advance along a path of accelerated economic development and accompanying social welfare. If Africa is to compete in the digital age, there must be a shift in traditional thinking. The collaboration between government, business (local and international), labor, and academia can change views, advance policies, and create an environment that favors knowledge-sharing and execution. For example, Jumia (a large e-commerce platform mentioned in Section I) found that online sales account for less than 1 percent of total retail sales in Africa, compared with 24 percent in China.91 The continent needs to find an effective niche in the global digital economy to enhance technology-led, inclusive social and economic growth. Accordingly, Africa needs policies backed by concerted nationwide efforts and public-private partnerships.92

Taxes and other fiscal determinants also play a key role in influencing potential investors who prefer to participate in an economy with ample resources, a stable business environment, and a good governance system in place.93 Striking a balance between ensuring the digital sphere is investment-friendly and leveraging maximum revenue for financing sustainable development goals will be a key tax policy challenge for African governments.94 Though the unilateral digital tax measures were introduced to address a universally accepted concern — tax evasion — the range of measures introduced by African countries and the limitations of these measures will affect digital growth. The digital economy can benefit the continent greatly, and it is important to ensure that these various measures do not impede this digital growth. Digital MNEs provide a beneficial platform for regional SMEs and start-ups. Any tax measures targeting them should consider the impact that the tax may have on the supply of beneficial services or the internet in general. Empirical research should be conducted periodically to better understand the potential impact, and there should be adequate public consultation for new tax policies to ensure that stakeholders’ concerns are recorded.

The continent should not sacrifice long-term growth potential to collect a small amount of digital taxes, which can affect the FDI flow and may act as a barrier for MNEs that are considering establishing a footprint in Africa. For example, the social media tax in Uganda led to a drop in the number of internet users from 18.5 million (47.4 percent penetration) to 13.5 million (35 percent) in its first six months alone. Similarly, it has been estimated that a short-lived social media tax in Benin would have resulted in forgone GDP growth of $260 million and taxes of $40 million.95

As policymakers continue to evaluate ways to tax digital businesses, it will be necessary to avoid creating distortive tax policies that generate negative externalities at different points in the internet value chain. Ultimately, fiscal policy that contradicts the public policy objectives of ensuring meaningful, affordable access to digital technology and advancing human rights can limit innovation and productivity. The following subsections examine some measures that can be undertaken to improve the effectiveness of various policy measures.

A. Invest in Tax Administration and Technology

Tax administrations need sufficient resources because limitations may impede the effective enforcement of the tax policy measures in place. As was highlighted in a report prepared for the IMF, “retaining the newly recruited tax specialists would be extremely challenging due to their high market demand and due to the divergent pay gap between the public sector and advisory firms.”96 In many African countries, tax authorities urgently need training, resources, and capacity. Tax authorities must also consider investing in the latest technology, including AI and data analytics tools, to gather and analyze information about digital businesses conducted with African consumers to ensure effective tax enforcement. The tax authorities can consider borrowing best practices from developed countries and specialized bodies like Tax Inspectors Without Borders for help putting these technological measures into practice.97 Further, Africa can also make greater use of e-governance to create path-breaking reforms to address various digitalization challenges and work toward running cost-effective governments.98

B. Regional Collaboration

Problems with tax enforcement could be battled by strengthening regional collaboration. For African states, collective action through a unified digital tax regime is likely to provide better compliance since individual African countries offer relatively insignificant markets for digital MNEs. A single point of registration, return filing, and payment would facilitate compliance by digital MNEs and be much simpler than having to do all that for each of Africa’s 54 countries. This is illustrated by the EU’s MOSS for VAT, an optional scheme that allows digital MNEs to account for VAT in one EU country rather than each of the 28 individual EU member states.99 A coordinated tax regime would ensure better alignment with regional coordination efforts under the African Continental Free Trade Agreement. Some steps have already been taken in this direction — for example, the East African countries announced that they intend to develop a joint strategy to tax digital firms — but these efforts are still being developed.100 African states acting in concert through regional bodies like ATAF are also better able to leverage their combined market potential to compel compliance from digital MNEs.

C. Improve Tax Certainty and Ensure Fairness

To reduce tax uncertainty and ensure fair allocation of taxing rights, policymakers must harness policies in a manner that ensures benefits are shared fairly and equitably.101 Proper and sound economic tax impact assessments must be conducted before implementing any tax proposal. This assessment must include public consultations with key stakeholders.

Understanding the potential effects and unintended consequences of a particular action by conducting a detailed assessment of implications and creating a rigorous economic rationale for digital taxes can facilitate better tax policy design, prevent adverse effects on revenue generation, and mitigate the digital divide, especially for marginalized groups including women, youth, and those in the informal sector who already face higher barriers to internet access.

D. Data Protection

One of Africa’s biggest resources is its large and youthful population. This factor is helping the African economy by attracting digital MNEs that depend on the data extracted from such masses. While this is an attractive incentive for MNEs, governments must introduce measures that ensure data confidentiality and protection. In this regard, the African countries should borrow from EU and U.K. data protection laws and consider the need to regulate different players in the online world.102

E. Participation in International Platforms

Thus far, African states have been rule takers in the global tax landscape. Instead, they must actively participate in various multilateral discussions about international tax and trade reform — including discussions about the OECD BEPS inclusive framework and the WTO’s e-moratorium — to ensure their unique challenges are addressed in the global tax reset and considered when new rules are drafted for cross-border digital flows. While participation is important, African states must bear in mind that their challenges are different from those of developed countries. Therefore, their final solutions will have to be fit for purpose and unique to the realities of their local economies.

Needless to say, there is heavy political pressure to introduce some mechanism for taxing digital services because of the large revenue gaps. However, such a taxation levy can have both positive and negative impacts. Earlier sections of this article discuss the concern that digital taxes are likely to result in double taxation if the MNE’s country of residence does not provide credit for taxes suffered in the market jurisdiction. This may result in different treatment of different taxpayers and distort business decisions.

The ongoing international and unilateral efforts to tax digital transactions, though useful, are of limited application to Africa’s unique situation. Developed countries, however, have the advantage and can take some unilateral steps. They can leverage a substantial share of digital MNEs’ market. Further, the digital MNEs will often have a presence in these countries beyond just a website, which can also be leveraged. Most African countries, however, lack these advantages.

To encourage growth and safeguard tax revenue, digital taxation rules must be simple and flexible to allow businesses to comply easily today and accommodate new digital business activities tomorrow. Simple and flexible rules can still result in complexity if there is limited coordination between different countries.103 A successful approach would consider the still limited (though growing) digital market in Africa, the benefits that African SMEs garner from the use of digital platforms, the peculiar administrative challenges faced by African tax administrations, and more. This approach should be multilateral but focused on the concerns of developing economies.

VIII. Conclusion

African countries have implemented a variety of unilateral tax measures that affect MNEs and the countries implementing them in both positive and negative ways. Every tax measure has its own policy objectives, advantages, and drawbacks; no single factor can be analyzed to decide the practical appropriateness of a tax measure for the continent as a whole.

Table 2 attempts to present the key impacts that various unilateral tax measures would have for the continent as a whole.

Table 2. Impact of Unilateral Digital Tax Measures in Africa

Unilateral Measure

Impact on Prices

Investment in Technology

Computational Complexity

Capacity Requirements for Tax Administration

Possible Impact on FDIs

DST

Will increase the price of digital products and services for both businesses and consumers, directly or indirectly.

May require investment in technology to streamline implementation (that is, how the tax authorities administer the levy and how MNEs can get data for computational purposes).

Computations are complex. It is especially challenging to get data points and ensure that data points are completely accurate.

Tax administrations need more resources to administer.

May significantly affect the flow of FDI.

SEP

May not affect prices, considering the option for nonresident MNEs to avail the benefits of DTAs or unilateral tax relief.

See above.

See above.

See above.

Moderate impact when there is an option for nonresidents to benefit from DTAs or unilateral tax relief.

Withholding Taxes

Might indirectly affect prices if prices are grossed up.

Requires low investment on the technology side because the process of complying with the levy is not complex.

Relatively simple computations.

Requires comparatively fewer administrative resources.

See above.

VAT

High impact on prices for end consumers because the tax burden is passed through the supply chain and borne by ultimate consumers. However, the price will not be affected in B2B situations when businesses can claim input credit for the taxes paid.

Requires moderate investment because a lot of tax technology solutions for VAT are already available.

Relatively simple computations.

Tax administrations need more resources to administer.

Moderate impact because the tax burden is passed through the supply chain.

According to the table, VAT looks like an easy solution in terms of keeping the jurisdiction attractive to FDI. However, the burden of VAT will fall directly on consumers where they will not be able to claim input tax credit. Considering that there are wide income disparities on the continent and the vast majority of the population has limited disposable income, this may affect the consumption of goods and services. Further, though administering VAT collections would demand high resources, African countries could effectively work with the ATAF to formulate a single unified approach for enforcing VAT collections, thereby improving the efficiency of the available resources. Africa could also take a practical lead from the EU’s MOSS system to simplify procedural compliance and reduce administrative complexities.

Nonetheless, the future of these diverse unilateral measures is contingent on the progress toward global consensus. Also, it is likely VAT levied to cover digital services would continue even if the pillar 1 and 2 proposals are enacted.

In its response to the pillar 1 and 2 proposals, ATAF raised several concerns on the ability of the proposals to address the imbalance of taxing rights. Regarding pillar 1, ATAF supported the objective of the unified approach and the use of country-specific thresholds under the new nexus rules, but it raised concerns regarding the definition of routine activities, adoption of transfer pricing rules, valuation of sales and user participation, use of a mandatory safe harbor, and mandatory binding arbitration rules.104 Regarding pillar 2, the ATAF responded that the minimum effective tax rate must be set at a high rate to prevent profit shifting and that imbalances can only be addressed if the subject-to-tax rule is applied first.

The two-pillar proposal received a further global push on July 1 when 130 inclusive framework members agreed on the proposal to reform the international tax rules.105 However, Africa’s two most prominent economies — Kenya and Nigeria — have not agreed to the proposal. Also on July 1, the ATAF issued a statement expressing its concerns, mainly regarding the quantum of residual profits to be allocated to market jurisdictions, the use of binding dispute resolution, the appropriate global minimum tax rate, and the scope of subject to tax rules.106 The forum called for an equitable tax allocation and measures to disrupt illicit financial flows from the continent.

In light of global developments, resistance from some African countries, and the concerns raised through the ATAF statement, uncertainty prevails regarding how digital taxes will evolve over time on the African continent. Only 25 of the African countries are participating in global discussions — and several of the countries that are nominally participating are inactive.107 Hence it seems likely that even after a global solution is offered, unilateral measures in some form or another will be here to stay in some African countries, at least until they are specifically revoked.

Ultimately, even the few African countries that are playing active roles in the global discussion feel their voices are not being heard by the developed countries. A shift from unilateral measures to a unified approach might not favor African countries if their concerns are not addressed.

FOOTNOTES

1 Simbarashe Hamudi, “Digitalisation and the Challenges for African Administrations,” 1(2) J. on Financing for Dev. 177 (2020).

2 Yiming Zhang Beirão, “Does Big Data Generate Value on Netflix and Facebook Platforms and How Should They Tax These Values?” Tilburg University Master’s Thesis (2019).

4 Honest Prosper Ngowi, “Reflections on the Implications of the Fourth Industrial Revolution (FIR) on Taxation,” 1(2) J. on Financing for Dev. 204 (2020).

5 See African Union, supra note 3.

6 United Nations Conference on Trade and Development, “The UNCTAD B2C E-Commerce Index 2020,” 7 (2020).

7 Shamira Ahmed and Alison Gillwald, “Multifaceted Challenges of Digital Taxation in Africa,” Research ICT Policy Brief 7 (Nov. 2020).

8 International Telecommunication Union, “Economic Contribution of Broadband, Digitization, and ICT Regulation: Econometric Modelling for Africa” (2019).

9 Google and International Finance Corporation, “e-Conomy Africa 2020: Africa’s $180 Billion Internet Economy Future” (Nov. 11, 2020).

10 Acha Leke and Tawanda Sibanda, “The Rapid Growth of Digital Business in Africa,” Harv. Bus. Rev. (Apr. 22, 2019).

11 Alice Munyua, “5 Policy Hotspots That Are Key to Africa’s Digital Transformation,” World Economic Forum (Aug. 26, 2019).

13 Isabella Wilkinson, Niousha Roshani, and Carolina Rossini, “Africa’s Digital Economy: Opportunities and Obstacles During COVID-19,” Portulans Institute blog, May 21, 2020.

14 Tunisia Investment Authority, “ICT Sector” (2019).

15 “Tech in Africa,” The Generalist, June 20, 2021.

16 World Bank Group, “Nigeria: Digital Economy Diagnostic Report” (2019).

17 Mike Chitavi, Lauren Cohen, and Spencer C.N. Hagist, “Kenya Is Becoming a Global Hub of FinTech Innovation,” Harv. Bus. Rev. (Feb. 18, 2021).

18 “Tech in Africa,” supra note 15.

20 Amatka (Pty) Ltd., “The Future of Digital Business Models in Sub-Saharan Africa” (Sept. 2016).

21 World Bank Group, “South Africa Digital Economy Assessment” (2018).

22 World Bank Group, “Ghana Digital Economy Diagnostic” (2019).

23 National Computer Board (Mauritius), “ICT Industry in Mauritius” (last accessed July 9, 2021).

24 Mauritius Ministry of Technology, Communication & Innovation, “Digital Mauritius 2030” (2018).

25 Jake Bright, “Naspers Foundry Is Open for South African Startup Pitches, CEO Says,” TechCrunch, Oct. 24, 2019.

26 Alexandra Wexler, “Netflix Ups the Ante in Africa,” The Wall Street Journal, Dec. 22, 2020.

27 Jake Bright, “African Fintech Startup Chipper Cash Raises $30M Backed by Jeff Bezos,” TechCrunch, Nov. 19, 2020.

28 Eleni Mourdoukoutas, “Africa’s App-Based Taxis Battle Uber Over Market Share,” Africa Renewal (Aug.-Nov. 2017).

29 Jonas Koch, Kai Rövenich, and Kalle Dunkel, “How Africa’s Growing Mobile Money Market Is Evolving,” EY, Mar. 13, 2020.

30 Digital TV Research, “African OTT to Generate $1.7 Billion” (2020).

31 See Mourdoukoutas, supra note 28.

32 IBM, “IBM Research: Africa” (last accessed July 24, 2021).

33 “Search Engine Market Share Africa, Jan.-Dec. 2020,” StatCounter (last accessed May 15, 2021).

38 Mukesh Butani and Vidushi Gupta, “De-Escalating the Digital Tax Debate,” Bloomberg Quint, June 21, 2020.

39 UNCTAD, “Chapter 7: Recommendations,” in Economic Development in Africa (2020).

40 ATAF, “The Tax Challenges Arising in Africa From the Digitalisation of the Economy,” ATAF Technical Note CBT/TN/01/19 (2019).

41 ATAF, “Suggested Approach to Drafting Digital Services Tax Legislation” (Sept. 30, 2020).

42 European Commission, “A Fair & Competitive Digital Economy — Digital Levy” (last accessed July 12, 2021) (summarizing the initiative and providing links to related documents including an impact assessment and comments received on the proposal).

43 Hamudi, supra note 1.

44 Notably, there has been controversy regarding whether DSTs are direct or indirect taxes because they have features of both. For purposes of this article, we have categorized DSTs as direct taxes.

45 Young Ran (Christine) Kim, “Digital Services Tax: A Cross-Border Variation of the Consumption Tax Debate,” 72(1) Ala. L. Rev. 131 (2020).

47 Diana Padilla, “Criteria to Distinguish DSTs and Other Taxes Qualifying as Income Tax for Article 2 Purposes,” University of Amsterdam Master’s Thesis (2020).

49 George Nguyen, “Google Passes on 2 Percent ‘Regulatory Operating Cost’ for Ads Served in India and Italy,” Search Engine Land, July 27, 2021.

50 Office of the U.S. Trade Representative, “USTR Releases Findings in DST Investigations” (Jan. 6, 2021).

51 OECD, “Addressing the Tax Challenges of the Digital Economy, Action 1 — 2015 Final Report” (Oct. 5, 2015) (action 1 final report).

52 Kilian Heller, “Beat the Bit and Here Is How! How to Deal With the Digital Economy in International Taxation,” Maastricht University Master’s Thesis, at 37 (2015).

53 KPMG, supra note 46.

54 Chris Noonan and Victoria Plekhanova, “Taxation of Digital Services Under Trade Agreements,” 23(4) J. Int’l Econ. L. 1015 (2020).

55 See Nana Ama Sarfo, “Digital Taxation in Africa: Safety in Numbers?” Tax Notes Int’l, Sept. 7, 2020, p. 1261.

56 Indian Central Board of Excise and Customs, “Online Information Data Base Access and Retrieval (OIDAR): GST (Goods & Services Tax)” (last accessed May 15, 2021).

57 Eli Hadzhieva, “Impact of Digitalisation on International Tax Matters — Challenges and Remedies,” European Parliament Committee on Financial Crimes, Tax Evasion, and Tax Avoidance, PE 626.078 (Feb. 2019).

58 Jennifer Farrell, “Chapter 3: The WTO and Taxation: Preliminary Observations,” in The Interface of International Trade Law and Taxation (2011).

59 WTO Appellate Body Report, “United States — Tax Treatment for ‘Foreign Sales Corporations,’” WT/DS108/AB/R (Feb. 24, 2000).

60 Ministry of Economy, Trade and Industry (Japan), “Chapter 15: Unilateral Measures,” in 2015 Report on Compliance by Major Trading Partners With Trade Agreements, Part II (2015).

61 WTO Panel Report, “United States — Import Prohibition of Certain Shrimp and Shrimp Products,” WT/DS58/R, para. 7.43 (May 15, 1998).

62 Wallace Cheng and Clara Brandi, “Governing Digital Trade — A New Role for the WTO,” German Development Institute Briefing Paper (2019).

63 India and South Africa, “Moratorium on Custom Duties on Electronic Transmissions: Need for a Re-Think,” Work Programme on Electronic Commerce (July 12, 2018).

64 India and South Africa, “Growing Trade in Electronic Transmissions: Implications for the South,” Work Programme on Electronic Commerce (June 3, 2019).

65 Alexander Pogorletskiy and Sergei Sutyrin, “Chapter 8. Taxation of International E-Trade: Russian Particularities,” in Adapting to the Digital Trade Era: Challenges and Opportunities (2021).

66 Christian L. Neufeldt, “The WTO and Direct Taxation: Direct Tax Measures and Free Trade,” 59 Harv. Int'l L.J. 5 (2018).

67 Patrick Low, “Digital Services Taxes, Trade and Development,” Trade & Investment in Services Associates Working Paper No. 2020-07 (2020).

68 WTO, “WTO Dispute Settlement: Resolving Trade Disputes Between WTO Members.” See also article 23 of the WTO’s “Understanding on Rules and Procedures Governing the Settlement of Disputes”; and WTO Panel Report, “United States — Import Measures on Certain Products From the European Communities,” WT/DS165/R (Jan. 10, 2001).

69 Noonan and Plekhanova, supra note 54.

70 Colin John Clavey et al., “International Tax Reform, Digitalization, and Developing Economies,” World Bank Working Paper 142256 (2019).

71 Patrick T.F. Schrievers, “Digital Services Tax Emerging in Africa,” NovioTax blog, Dec. 17, 2020.

72 See supra Section II.A.

74 Hadzhieva, supra note 57.

75 Julia Stoll, “Netflix Subscribers Count Worldwide 2013-2021,” Statista, Apr. 21, 2021.

76 Suchint Majmudar, Nupur Jalan, and Elvira Misquith Tigdi, “Digital Economy: New Profit Allocation and Nexus Approach,” 30 J. Int’l Tax’n 10 (2019).

77 Id.

78 George Turner, “Netflix, Tax Reform and the Unreal Nature of Digital Taxation,” Tax Watch UK, Dec. 8, 2020.

79 Mathew Johnston, “How Facebook Makes Money,” Investopedia, Jan. 30, 2021.

81 OECD, supra note 73.

82 Majmudar, Jalan, and Tigdi, supra note 76, at footnote 92.

83 Kristopher B. Jones, Search Engine Optimization: Your Visual Blueprint for Effective Internet Marketing (2013).

84 Daniel McCarthy, Peter Fader, and Bruce Hardie, “Valuing Subscription-Based Businesses Using Publicly Disclosed Customer Data,” 81 J. of Marketing 17 (2016).

85 Charu C. Aggarwal, “Chapter 1. An Introduction to Social Network Data Analytics,” in Social Network Data Analytics (2011).

86 EY Global, supra note 12.

87 Mickaël Sallent, “External Debt Complicates Africa’s COVID-19 Recovery, Debt Relief Needed,” U.N. Africa Renewal (July 30, 2020).

88 Karin Strohecker and Joe Bavier, “As New Debt Crisis Looms, Africa Needs More Than World Is Offering,” Reuters, Nov. 19, 2020.

89 Shamira Ahmed and Alison Gillwald, “Multifaceted Challenges of Digital Taxation in Africa,” Research ICT Africa Policy Brief 7 (Nov. 2020).

91 Leke and Sibanda, supra note 10.

92 Vladimir V. Korovkin, “National Digital Economy Strategies: A Survey of Africa,” ORF Issue Brief No. 303 (July 2019).

93 UNCTAD, “Tax Incentives and Foreign Direct Investment,” ASIT Advisory Studies No. 16 (2000).

94 Hamudi, supra note 1.

95 Christoph Stork and Steve Esselaar, “Regulatory and Tax Treatment of OTTs in Africa,” Mozilla and the African Union Commission (2019).

96 Vito Tanzi and Howell Zee, “Tax Policy for Developing Countries,” IMF Economic Issues No. 27 (Mar. 2001).

97 Ivan K. Mugabi, “A Commentary on Airbnb as a Digital Business Model: Regulation and Tax Challenges for East Africa,” 1(2) J. on Financing for Dev. 218 (2020).

98 Nimmo Elmi, “The Colonial Aftermath in Digitalising Tax,” 1(1) J. on Financing for Dev. 86 (2019).

100 Julie Martin, “East Africa Nations Pledge to Jointly Address Taxation of Digital Firms,” MNE Tax, Nov. 13, 2020.

101 Wawan Juswanto and Rebecca Simms, “Fair Taxation in the Digital Economy,” ADB Institute Policy Brief No. 2017-5 (Dec. 2017).

102 Mugabi, supra note 97.

104 ATAF, “ATAF’s Opinion on the Inclusive Framework Pillar One (Including the Unified Approach) and Pillar Two Proposals to Address the Tax Challenges Arising From the Digitalisation of the Economy” (Nov. 2019).

105 OECD, “130 Countries and Jurisdictions Join Bold New Framework for International Tax Reform” (July 1, 2021). The number of countries joining the OECD proposal has since risen to 133.

107 African Union, “Briefing for the Ministers on Taxing the Digital Economy and the Global Tax Debate,” Briefing Paper for the Extraordinary Specialised Technical Committee on Finance, Monetary Affairs, Economic Planning and Integration (Dec. 2020).

END FOOTNOTES

Copy RID