GOVERNANCE AND THE DIGITAL ECONOMY IN AFRICA VOLUME 2 REGULATING THE DIGITAL ECONOMY IN AFRICA MANAGING OLD AND NEW RISKS TO ECONOMIC GOVERNANCE FOR INCLUSIVE OPPORTUNITIES © 2024 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy, completeness, or currency of the data included in this work and does not assume responsibility for any errors, omissions, or discrepancies in the information, or liability with respect to the use of or failure to use the information, methods, processes, or conclusions set forth. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Cover photo: Image generated using ImageFX from the prompt ‘Africa’, ‘regulation’ and ‘digital technology’. GOVERNANCE AND THE DIGITAL ECONOMY IN AFRICA VOLUME 2 REGULATING THE DIGITAL ECONOMY IN AFRICA MANAGING OLD AND NEW RISKS TO ECONOMIC GOVERNANCE FOR INCLUSIVE OPPORTUNITIES Tania Begazo, Clara Stinshoff, Hannelore Niesten, Georgiana Pop, Rong Chen, Gonçalo Coelho A joint report by the Digital Development Global Practice, Governance and Institutions, and Trade, Innovation and Competitiveness/Finance, Competitiveness and Innovation Practice Groups >>> Contents Acknowledgments xiii Executive Summary xv 1. Introduction: Understanding the economic governance landscape of the 2 digital economy in Africa About the report on Africa digital governance 4 1.1. The evolution of economic governance of digital 5 1.2. An overview of the characteristics and outcomes of digital markets in Africa: do they 7 mirror good economic governance? 1.2.1. International connectivity 14 1.2.2. Fiber networks for backbone and metro connectivity 20 1.2.3. Fixed broadband retail 23 1.2.4. Telecom towers 24 1.2.5. Mobile voice and broadband retail 28 1.2.6. Mobile money 34 1.2.7. Data centers and data services 37 1.2.8. Digital platforms and other digital business offerings 43 1.3. An overview of the public institutions that set rules on digital markets in Africa: what 49 actors are supporting economic governance of digital? 2. Addressing old economic governance risks for digitalization: Channeling 56 resources to the state through taxation and direct participation in market 2.1. Taxing digital infrastructure: taxes and parafiscal fees 58 2.1.1. How important are taxes and parafiscal fees for governments and operators? 60 2.1.2. Are taxes and parafiscal fees in line with good practice? 62 2.1.3. Summary of economic governance issues around taxes and parafiscal fees 73 2.2. Participating in markets as a supplier: State-owned enterprises and State-linked 76 enterprises 2.2.1. Corporate governance of state-owned and state-linked enterprises 79 2.2.2. Potential market distortions linked to SOE-SLE operation 84 3. Addressing new governance risks for digitalization: Safeguarding users 90 through data and competition rules 3.1. Regulating data: data protection, cybersecurity and enablers for data use and re-use 90 3.1.1. Personal data protection 91 3.1.2. Cybersecurity and cybercrime 95 3.1.3. Use and reuse of data and cross border data flows 98 3.1.4. Regulation and governance 100 3.2. Promoting and safeguarding competition in digital services: antitrust enforcement 102 and policy advocacy 3.2.1. Antitrust enforcement 104 3.2.2. Competition Advocacy 112 4. Strengthening economic governance of the digital economy 115 5. References 123 Annex 1: List of contributors for data collection on taxation and State- 141 owned and linked enterprises Annex 2: Summary of characteristics of segments along the digital value 143 chain 1. Submarine cables 144 2. Fiber backbone 146 3. Telecom towers 148 4. Fixed retail 150 5. Mobile retail 152 6. Data centers 154 7. Mobile money 156 Annex 3: Governance risks indices and methodological notes 158 1. Market structure 163 2. Foundational 165 3. SOE-SLEs 167 4. Taxation of digital infrastructure 169 5. Data regulation 171 6. Competition law and enforcement 173 >>> List of Tables Table 1: Examples of cable landing point owners with state involvement 18 Table 2: Top 10 companies in the data center segment in Africa (by size of data centers) 41 Table 3: Examples of government entities setting rules that affect the economic governance of 49 the digital sector Table 4: Classification of taxes and fees for digital infrastructure 62 Table 5: Average sub-saharan Africa vs. other regions (2017 – 2019) 65 Table 6: Types of taxes applied on digital infrastructure 73 Table 7: Overview of the different taxes and fees, alignment with tax and public administration 74 principles for better economic governance Table 8: Typology of competition risks in digital platforms 105 Table 9: Level of risks to economic governance 118 Table 10: Risks to economic governance and policy recommendations 120 >>> List of Figures Figure 1: Evolving risks of digitalization – first and second-generation risks 6 Figure 2: Simplified digital economy value chain and market segments 9 Figure 3: Average concentration in the digital economy across Africa, by segment 12 Figure 4: Average concentration across seven key segments, by country 12 Figure 5: Ownership in key digital infrastructure and services segments across Africa, by 13 ownership type Figure 6: Ownership in key digital infrastructure and services segments across Africa, by 13 ultimate owner Figure 7: Simplified digital value chain and key market characteristics that affect economic 13 governance. Figure 8: International internet bandwidth per inhabitant 14 Figure 9: Weighted median IP transit prices for 10 GigE and number of submarine cables 17 Figure 10: Number of market players in the international connectivity segment 19 Figure 11: Operational fiber (km) relative to country GDP and population size 20 Figure 12: Concentration in fiber broadband markets, by African sub-region, 2016 and 2020 21 Figure 13: Metro fiber networks in sub-regions relative to population 21 Figure 14: Concentration (HHI) in metro fiber in sub-regions, 2020 21 Figure 15: Ownership of backbone and metro fiber networks in African sub-regions, 2020 22 Figure 16: Vertical integration in fiber 23 Figure 17: HHI in fixed broadband retail, across African countries 24 Figure 18: Share of subscribers, by ownership 24 Figure 19: Share of towers owned by towercos 25 Figure 20: Heatmap of concentration (HHI) in terms of tower count 26 Figure 21: HHI in towers vs HHI in mobile 26 Figure 22: Tower holding by organization, by region ( percent of regional tower count) 26 Figure 23: Acquisition of tower portfolios by towercos – number of deals and towers 27 Figure 24: Number of tower acquisition deals by country, 2010-2020 27 Figure 25: Number of active mobile operators (excluding MVNO), by country 28 Figure 26: Number of MVNOs and market share, by country 29 Figure 27: Market concentration in mobile markets as indicated by HHI scores (mobile 30 subscribers), by country Figure 28: Change in HHI 2015-2020 (red = higher HHI, green = lower HHI) 30 Figure 29: HHI by region (700- 800 MHz bands) 31 Figure 30: Average HHI in terms of mobile subscribers and spectrum holding, by region 31 Figure 31: Market shares of SOEs and SLEs in mobile retail, by region (min – average – max) 32 Figure 32: Countries covered by at least one of the five largest groups 32 Figure 33: Region-wide subscriber shares of main telecom groups 33 Figure 34: Market structure upon entry in the mobiles services market (MNOs) 2008-20 34 Figure 35: Countries involved in recent M&A (2009-20) 34 Figure 36: Millions of mobile money accounts across African regions 35 Figure 37: Growth of mobile money accounts across African regions 35 Figure 38: Market structure in mobile money 36 Figure 39: Concentration of mobile money and mobile services, in terms of subscribers 36 Figure 40: Number of players in mobile money markets across Africa 36 Figure 41: Composition of firm types offering mobile money services in Africa 37 Figure 42: Ownership type of mobile money providers 37 Figure 43: Average number of internet users per datacenter, Africa vs. income groups 38 Figure 44: Mean mobile latency (in milliseconds) across African countries 38 Figure 45: Number of data centers across African countries 38 Figure 46: Average country level HHI by number of datacenters 39 Figure 47: HHI by size of datacenters (for 7 markets) 39 Figure 48: Type of ownership of data centers by subregion (2020) 40 Figure 49: Simplified data center and cloud value chain 42 Figure 50: Market shares of leading cloud hosting providers in Africa (in terms of websites) 43 Figure 51: Overview of new digital businesses and capital invested since 2010, all Africa 44 Figure 52: Total number of African start-ups and capital invested per capita since 2010, by 44 subregion Figure 53: Number of VC deals in African subregions, 2019-2021 44 Figure 54: Total VC capital invested in African HQ’d firms, 2019-2021 44 Figure 55: 2021 VC funding in Africa across verticals, in US$M 45 Figure 56: Total funding for e-Commerce (in USD million) 46 Figure 57: Website visits of largest e-commerce sites in Africa 46 Figure 58: Comparison of app usage in Africa vs HIC, MIC, and other regions, 01-12/2021 47 Figure 59: Relative importance of top ten apps in terms of daily active users (DAU) among a 48 group of top 50 apps Figure 60: ITU National Collaborative Governance 54 Figure 61: Alignment with good international regulatory practice and implementation 55 Figure 62: Tax-to-GDP ratio, 2018 60 Figure 63: General and sector-specific taxes and fees, as a proportion of market revenue 61 Figure 64: Rates of corporate income tax on mobile operators, 2021 63 Figure 65: Overview of customs duties on communication apparatus and base stations 64 Figure 66: Overview of customs duties on users’ goods for mobile services (phones and SIM cards) 65 Figure 67: Combined tax rates on use of mobile services in 2021 69 Figure 68: Share of countries with at least one SOE present in the sector or subsector in 77 advanced economies (AEs) and emerging and developing economies (EMDEs) Figure 69: Digital SOE-SLEs in Africa 78 Figure 70: SOE-SLE mandates and objectives 80 Figure 71: Ownership rights and board composition 81 Figure 72: Corporate governance and performance information disclosed (percentage of SOEs 83 and SLEs) Figure 73: Published information by SOE-SLE, by subregion 84 Figure 74: Potential SOE impacts across digital markets where competitive neutrality is lacking 85 Figure 75: Share of SOEs that are vertically integrated across more than one segment 86 Figure 76: Percentage of SOEs that are either the largest or second largest company in the 86 relevant market segment Figure 77: Data protection legal frameworks in Africa 92 Figure 78: Comprehensiveness of African data protection laws 93 Figure 79: Percent of countries per country income group that have adopted good regulatory 94 practices on personal data protection Figure 80: Cybersecurity and cybercrime frameworks in Africa 96 Figure 81: Comprehensiveness of African cybersecurity frameworks 97 Figure 82: Average scores on different data governance dimensions by income group/region 99 Figure 83: Estimated cost of a hypothetical one-day internet shutdown (USD) 102 Figure 84: Regulatory aspects important to enable competition in digital markets 103 Figure 85: Status of competition law and authorities across Africa, 2020 104 Figure 86: Number of antitrust cases and advocacy initiatives related to the digital economy 108 in Africa Figure 87: Enforcement cases by income level classification in Africa 108 Figure 88: Advocacy cases by income level classification in Africa 108 Figure 89: Types of case of antitrust cases and advocacy ( percent of total) 109 Figure 90: Number of antitrust cases and advocacy initiatives related to the digital economy 109 by segment Figure 91: Types of abuse of dominance cases 110 Figure 92: Types of firms involved in abuse of dominance cases 110 Figure 93: Risk to economic governance of digital 117 Figure 94: Risk to economic governance of digital and digitalization 119 >>> List of Boxes Box 1: Africa Digital Economy Governance 4 Box 2: Weak market outcomes in Africa 8 Box 3: Assessing competition in the digital economy 10 Box 4: Subsea cables planned or under construction 15 Box 5: Competition authorities and telecom regulators' mandate on safeguarding competition 52 Box 6: How to address risks to digital inclusion in Africa 57 Box 7: Morocco and Tunisia require operators to separate activities, which promotes a level 86 playing field Box 8: Digital SOEs in Angola: state monopolies, preferential treatment and weak market 88 outcomes Box 9: Countries where SOEs benefit from preferential access to capital from the government 89 Box 10: The road to start implementing data protection legislation in Kenya 94 Box 11: Misinformation and the COVID-19 Pandemic 101 Box 12: Anticompetitive horizontal agreements 110 Box 13: Merger control in African digital markets 112 Box 14: Competition advocacy for protecting consumers in digital markets 115 Box 15: Competition advocacy that contributed to streamlining the application of competition 115 law to digital markets >>> Acknowledgments This report was prepared by a joint team from the Digital Development, Governance and Institutions, and Trade, Innovation and Competitiveness/ Finance, Competitiveness and Innovation practice groups as part of the project ‘Africa Digital Economy Governance and Anticorruption’ (P172417), which is co-led by James Anderson, Tania Begazo, and Georgiana Pop. This report synthesizes information from five background papers on the taxation of digital infrastructure, digital state-owned enterprises and competitive neutrality, digital state-owned enterprises governance, competition enforcement and advocacy in digital markets, and data regulation, based on data collected in the period 2020-2022. This report integrates complementary analysis based on the World Bank’s Africa Digital Market Players database developed as part of this project. This report is volume 2 of the study ‘Africa: Digital Economy Governance’ and complements volume 1, ‘Digital Tools for Governance’. The preparation of this report was led by Tania Begazo (Senior Economist, IDD05) with support from Clara Stinshoff (Young Professional, IDD05) and Ana Belen Ruival (Consultant, IDD05). Clara Stinshoff, Linda Kirigi, and Antonio Neto provided inputs for chapter one. Chapter two benefited from the inputs of Hannelore Niesten (Consultant) and Tania Begazo on taxation while the section on State-owned enterprises (SOE) and State-linked enterprises (SLE) was drafted thanks to the contributions of Georgiana Pop, co-TTL (Senior Economist, ETIMT), Davida Connon, (Private Sector Specialist, ETIMT), Dennis Sanchez Navarro (Economist, ETIMT), Pascal Jaupart, (Economist, HLCSP) and Sarah Houllier (Extended Term Consultant, EAWG1). Inputs for chapter three’s section on competition policy were provided by Georgiana Pop, Davida Connon, and Gonçalo Coelho (Senior Competition Policy Consultant, ETIMT), while inputs on data regulation were provided by Rong Chen (Economist, IDD05), Lillyana Daza Jaller (Consultant), and Tania Begazo. <<< xiii James Anderson and Georgiana Pop provided valuable feedback and comments for the overall preparation of this report and through the preparation of background papers. Linda Kirigi (Consultant, IDD05) supported the preparation of this report. Clara Stinshoff, Hannelore Niesten, Linda Kirigi, Andres Sepulveda, Dennis Sanchez Navarro and Abdoulaye Singuy Ndong, as well as teams from Bowmans (see annex 1 for a list of contributors) and Progressus supported the collection and systematization of the underlying databases used for the preparation of this report, including the Africa Digital Market Players Database, the Africa Taxation of Digital Infrastructure and Parafiscal Fees database, and the Africa SOE and SLE in digital markets database, the latter prepared as a result of DD-TIC/FCI-GOV collaboration. Douglas Pearce (Practice Manager, EAEF2), Mark Dutz (Lead Economist, AFECE), and Tim Kelly (Lead Digital Development Specialist, IDD02) provided valuable peer review comments on the draft version of this report. Anat Lewin (Senior Digital Development Specialist, IDD04), David Satola (Lead Counsel, LEGOP) Jan Loeprick (Senior Economist, EMFTX), Jerome Bezzina (Senior Digital Development Specialist, IDD02), Tim Kelly (Lead Digital Development Specialist, IDD04), Henri Fortin (Lead Financial Management Specialist, EAWG2), Mariem Malouche (Senior Economist, ETIMT), and Sara Nyman (Senior Economist, ETIMT) were peer reviewers and Jana Kunicova (Senior Public Sector Specialist, EAEG1) provided comments for the preparation of background papers integrated into this report. The team thanks Manuel Vargas (Practice Manager, EAWG2), Michel Rogy (Practice Manager, IDD02), Isabel Neto (Practice Manager, IDD04), and Martha Licetti (Practice Manager, ETIMT) for their guidance and support to the project, as well as Boutheina Guermazi (Director for Regional Integration in Sub-Saharan Africa, the Middle East, and North Africa ) for her overall guidance. xiv >>> REGULATING THE DIGITAL ECONOMY IN AFRICA: MANAGING OLD AND NEW RISKS TO ECONOMIC GOVERNANCE FOR INCLUSIVE OPPORTUNITIES >>> Executive Summary While digital technologies offer enormous benefits to the economy and society, they also expose citizens, firms, and markets to risks. These include: • Exclusion and exacerbation of the digital divide, • Risks of unfair and unbalanced market outcomes, and • Risks of misuse of personal and commercial data for surveillance and illegal activities. This report focuses on understanding these risks to economic governance of the digital economy in African countries and proposes actions to mitigate them, enhancing the possibility of a virtuous cycle enabled by digital technologies. This report emphasizes risks of unfair and unbalanced market outcomes and misuse of personal and commercial data, complementing the World Bank report “Digital Africa: Technological Transformation for Jobs,” which focuses on closing the digital divide for households and enterprises. The adoption of digital technologies can support job creation and economic transformation, leading to inclusive growth in Africa. Internet availability has led to job creation and poverty reduction across Africa. For example, in Nigeria and Tanzania, internet availability increased labor force participation and reduced poverty rates by 7 percentage points, with more significant welfare effects among poorer and less-educated households.i Enterprises that use more sophisticated digital technologies are more productive and generate jobs faster than their peers in Ghana, Kenya, Malawi, and Senegal, while more intensive use of digital technologies is associated with more jobs, including for lower-skilled people.ii The increasing use of digital technologies in Africa, especially after the COVID-19 outbreak, calls for a greater understanding of the risks they create and more systematic and forward-looking actions to mitigate them in a timely manner. The number of internet users in Africa is growing rapidly, with 692 million unique mobile internet subscribers in 2021.iii The use of digital services offered over digital infrastructure has also expanded, with registered mobile money accounts growing at 12 percent a year and reaching over half a billion in 2020.iv Digital platforms are also increasingly omnipresent – Facebook, for example, has at least 250 million active users in Africa.v Governments are also progressively offering services online, with 37 African countries now offering operational portals for e-tax and 39 for e-procurement. More than 800 million African citizens now have digital IDs, and the number is growing.vi Enterprises have also increased i. Bahia et al 2021, Bahia et al 2020. ii Begazo, Dutz and Blimpo 2023 iii GSMA Intelligence 2021. iv. GSMA. 2021. State of the Industry Report on Mobile Money 2021. London: GSMA. https://www.gsma.com/mobilefordevelopment/wp-content/uploads/2021/03/GSMA_ State-of-the-Industry-Report- on-Mobile-Money-2021_Full-report.pdf. v https://www.internetworldstats.com/stats1.htm. vi Global Dataset 2018: https://id4d.worldbank.org/global-dataset. <<< xv adoption of digital technologies, with 37 percent investing in competitive markets, ultimately preventing African countries them after the COVID-19 outbreak. vii from reaping the benefits of the digital revolution. However, risks are becoming more visible: The report distinguishes traditional and new risks to robust governance of the digital sector in Africa. Traditional risks • The risks of exclusion and widening of the digital divide affect a fundamental pillar of the digital economy: digital persist in Africa, with over half a billion people (about half infrastructure that provides connectivity to access digital of the population) not using the mobile internet where services. These traditional risks are related to regulatory broadband digital connectivity is available, mostly in governance of digital infrastructure and channels through rural areas and low-income groups. viii The wealthiest 20 which the state obtains direct revenues from the sector: percent of the population are fifteen times more likely to taxation and parafiscal fees on digital infrastructure and state- have access to a computer or tablet than the poorest 20 owned and state-linked enterprises. New risks are related percent, and urban residents are over three times more to data and data-driven digital services that are essential to likely to have access to the internet than rural residents.ix digital services applications that can expand the benefits of • Concerns about noncompetitive results because of the the digital economy. These new risks are related to unfair operation of digital platforms in Africa have been reflected market outcomes because of weak competition and misuse in antitrust cases and market studies carried out by of data. competition authorities in countries such as Kenya, the Arab Republic of Egypt, and South Africa and in proposals The economic governance landscape is characterized to regulate digital markets (such as commissions by ride- by concentration of economic power in digital hailing companies). infrastructure and in certain digital services segments, • Concerns about the use of personal data are also growing. and weak regulatory power is found across entities For example, there is a lack of clarity in the financial sector with poor coordination links. concerning data ownership for credit scoring; the potential for data sharing in agriculture is not being realized; and These contrasting characteristics, together with deeper trust in digital ID systems might constrain adoption, given information asymmetry, create challenges for the balanced the potential for surveillance. economic governance of digital technologies. Market • Furthermore, the increased number of cybersecurity concentration is still high along the value chain and persistent threats and data breaches in Africa, especially in the over time, especially in key bottleneck digital infrastructure, finance sector, highlight the vulnerability of digitalization. such as international gateways and wholesale broadband In 2020, cyberattacks on online banking platforms in networks, extending to mobile retail broadband and digital Africa more than tripled.x services such as mobile money. Concentration is more limited in digital services like data centers, e-commerce, and website This report examines the governance of the digital economy in hosting. However, vertical integration and a trend to expand Africa to highlight areas where governments can develop their direct state participation in markets risk distorting the level policies to ensure that digitalization delivers benefits for all the playing field and erecting barriers to entry. population in terms of inclusion, job creation, and sustainable growth. It also aims to inform the implementation of the Digital On the regulatory side, as digital technologies permeate the Economy for Africa (DE4A) initiative in support of the African economy and new areas for government intervention arise, Union’s Digital Transformation Strategy.xi To do so, it analyzes new authorities are created, and the interconnectedness the current context of digital market players and government of sectoral and economy-wide policies directly affecting actors along the digital value chain, including characteristics digitalization becomes more important. Effective coordination that could limit the development of inclusive, efficient, and across traditional and new agencies with interlinked mandates vii. World Bank. 2021. COVID Business Pulse Survey or Enterprise Surveys. viii. GSMA, 2021. Mobile Internet Connectivity 2021: Sub-Saharan Africa Key Trends. See https://www.gsma.com/r/wp- content/uploads/2021/09/The-State-of-Mobile-Inter- net-Connectivity-2021-Sub-Saharan-Africa.pdf. ix. Seuyong et al. 2023. x. Institute for Security Studies. 2020. xi. https://www.worldbank.org/en/programs/all-africa-digital-transformation. xvi >>> REGULATING THE DIGITAL ECONOMY IN AFRICA: MANAGING OLD AND NEW RISKS TO ECONOMIC GOVERNANCE FOR INCLUSIVE OPPORTUNITIES becomes more cumbersome. Although creating new and revenue links between providers and government entities institutions or adding new mandates to regulate the digital can be weak. economy to existing ones supports technical independence and specialization, limited transparency, accountability, and State-owned and state-linked enterprises are capacity affect regulatory quality. Implementation challenges prevalent across Africa and are increasing their persist and expand as the information asymmetry on the presence and importance in key segments of the regulated service providers increases, given the rapid pace of digital value chain. technological change transforming operations and business models. New regulatory tools that embed agile regulation principles, support participatory regulation, and make better In Africa, 153 wholly or majority state-owned enterprises (SOEs) use of big data and consumer feedback are needed. and 46 enterprises with minority state shareholdings or state- linked enterprises (SLEs) operate in 52 countries. SOEs have Economic governance of digital infrastructure is essential to been important players in broadband backbone infrastructure enable the use of digital technologies. Without inclusive digital (controlling about three-quarters of fiber in Africa) and have connectivity, the risk of excluding vulnerable population groups expanded their importance in broadband retail and data centers. (low-income, women, rural, less educated, with informal jobs) However, transparency of SOE and SLE financial results and from the benefits of the digital economy is greater. Effective corporate governance arrangements is limited, with only 16 regulation of digital connectivity infrastructure is needed given percent publishing financial statements online.xii Furthermore, its economic nature: oligopolistic and concentrated markets in at least 10 countries, entities controlling SOEs still maintain given high investment and sunk costs, economies of scope, regulatory and/or policy-making powers. There is some and network effects. The small market size in some African transparency of government financial support given to SOEs or economies renders markets concentrated, making effective SLEs in only 6 out of 18 examined countries. Telecommunication digital infrastructure regulation and its implementation even regulators have implemented limited pro-competition regulation: more important. Furthermore, a good regulatory environment only one in eight SOEs or SLEs that have more than 65 percent can enable the development of African data infrastructure market share in mobile services are subject to specific obligations and give users the opportunity to choose service providers, based on significant market power. facilitating the expansion of digital services. At least 27 countries impose sector-specific taxes However, revenue links between digital connectivity on digital connectivity in Africa, with an increasing service providers and government entities (regulators, line trend of taxing digital services, such as mobile ministries, treasury) can facilitate regulatory capture and money transfers. rent seeking by government entities, reviving old risks to the economic governance of digital technologies. On the one hand, regulators, policymakers, and legislators can become The digital infrastructure sector is seen as highly profitable too favorable towards the regulated industry (especially and an easy target for revenue mobilization. At least 14 out state-owned enterprises or enterprises with minority state of 20 examined countries impose excise taxes, higher value- shareholdings) and disregard the public interest. Unchecked added tax (VAT) rates, stamp duties on digital connectivity markets risk leading to inefficient and unfair outcomes, for services, or similar additional taxes. Three countries apply example, large digital divides. On the other hand, these actors higher corporate income taxes (CIT) than on other sectors. can become too demanding on (private) operators and seek At least 20 set regulatory fees that depend on revenues and to appropriate sectoral gains through excessive taxes and are not proportionate to regulatory costs. Import duties for parafiscal fees. Either option would undermine economic equipment and devices are up to 20 percent and 25 percent, governance. In fragile contexts with limited state ownership respectively. The effective tax rate for telecommunications has in the sector, however, both effective regulatory enforcement been found to be higher than for gold mining. xiii xii. Jaupart and Begazo. 2023. xiii. Rota-Graziosi and Sawadogo. 2020. <<< xvii Sector-specific taxes (on top of regular direct and indirect The characteristics of digital business models that tend taxes), present in at least 15 out of 20 studied countries, may towards winner-takes-most dynamics create an environment create market distortions if they are not aligned with efficiency, where abuse of market power is more likely. Unlike developed equity, and simplicity tax principles and do not consider tax economies, offline markets and the need to incentivize incidence. Higher VAT and excise taxes applicable only consumer uptake still bring competitive pressure, but there to digital connectivity distort investment and consumption are signs of anticompetitive behavior and mergers that risk decisions. Excise-like taxes on connections, SIM cards, and weakening competition, putting consumers and smaller devices create barriers to digital adoption by the unconnected. suppliers at a disadvantage. Competition authorities are An array of taxes, contributions, and fees make compliance playing a role in preserving competition in new digital service and tax administration more cumbersome. Furthermore, a markets, but only in few countries. lack of predictability affects the investment environment: at least 26 changes in tax rates and rules across 14 countries Higher level of risks to economic governance of digital were observed in 2020. is associated with poorer digitalization outcomes. As digital services provided over digital infrastructure become The level of risks to good economic governance, as well as more important, new challenges to effective economic the main driver of risks, varies across countries. Among first- governance of the digital sector arise: misuse of data generation risks (related to lack of enabling market structure, and excessive market power. Surveillance and potential weak traditional digital infrastructure institutions, over- exploitation of market power affect trust in digital services, taxation of digital connectivity, and operation of digital SOEs/ which may have repercussions on the adoption of digital SLEs), market structure and operation of digital SOEs/SLEs technologies in Africa. cause the highest risks in Africa. Therefore, it is essential to strengthen regulation that allows for more competitive market structures (ex-ante pro-competition regulation), especially Although countries are adopting economy-wide in digital connectivity, and that creates a level playing field frameworks on data protection, cybersecurity, and independently of ownership, including through better corporate cybercrime, implementation is limited, and rules to governance and transparency. enable data use and reuse need further development. Second-generation risks become more relevant with higher Two key policy areas to prevent data misuse are data protection levels of digitalization. Among those, risks related to a weak and cybersecurity/cybercrime.xiv In Africa, 26 countries have competition policy framework and enforcement seem to adopted data protection frameworks applicable economy- be higher than other risks, especially given a lack of legal wide, but implementation has been limited. Provisions are frameworks or authority to boost and preserve competition generally in line with good practices except for restrictions economy-wide and oversee the conduct of digital business. related to cross-border data flows, innovations to support data Even if incipient, most countries have already included privacy and penalties.xv According to the Global Cybersecurity partial frameworks or measures related to data protection or Index (GCI), the average commitment of African countries to cybersecurity. The inherent characteristics of digital service cybersecurity is only 36 percent, indicating important gaps markets that tend to winner-takes-most and the rapid pace of regarding legal frameworks, technical rules, institutions, innovation and creation of different business models call for capacity development, and cooperation.xvi more effective antitrust enforcement and competition policy, especially in countries where digital services are substituting Although 43 countries are covered by either national offline markets. On data protection and cybersecurity, more or regional competition laws, and 36 countries effective implementation and cross-country collaboration is have adopted domestic competition laws, only 29 needed, given the opportunities from cross-border data flows competition authorities are operational, and only 10 and single digital markets in Africa. General recommendations have handled cases in digital service markets. to mitigate risks are summarized in table 0.1. xiv. World Bank. 2021. xv. Chen, Daza Jaller and Begazo. 2023. xvi. ITU. 2020. xviii>>> REGULATING THE DIGITAL ECONOMY IN AFRICA: MANAGING OLD AND NEW RISKS TO ECONOMIC GOVERNANCE FOR INCLUSIVE OPPORTUNITIES >>> Table 0.1. Risks to Economic Governance and Policy Recommendations Risk area and economic Policy recommendations to mitigate each type of risk governance issues First- generation risks Lack of enabling market structure • Introducing reforms to facilitate entry and expansion of smaller/new players • Implementing procompetition rules included in the sectoral framework Economic governance • Ensuring appropriate competition assessment in the evaluation of merger and acquisitions issue: Concentration of • Enhancing technical independence of digital infrastructure regulators and means for economic power collaboration with other authorities (competition authority when operational) that leads to regulatory capture Weak traditional digital infrastructure • Institutional design reforms and practices to enhance technical independence of regulator, institutions separating functions of various entities in the digital infrastructure sector • Capacity building in areas related to economic regulation to enable healthy digital Economic governance infrastructure issue: Weak • Deepen stakeholder consultation and transparency of decisions and motivation accountability, • More specific guidance for the consistent and predictable implementation of the legal transparency, framework independence and • Benchmarking across members of regional blocs and continental level to encourage more stakeholder engagement effective implementation leading to regulatory capture • Reassess the tax and parafiscal fees policy by identifying and systematizing current taxes and fees, quantifying tax and fees collections, and analyzing tax/fees incidence and their economic and social costs • Phase out excise-like taxes on digital connectivity that distort consumer behavior and Overtaxation of digital investment decisions across sectors and countries. Use economy-wide tax expenditure connectivity assessments to identify less-distortive measures for revenue mobilization • Ensure an equitable tax framework by minimizing over- taxation that affects the less well-off Economic governance included the unconnected, replace blanket tax exemptions by support measures targeted at issue: Appropriation of the most vulnerable sector resources and • Ensure a simple and transparent legal and regulatory framework with clear provisions on more opportunities for rent the taxes and fees applicable to all digital infrastructure operators and ensure a balanced seeking by public officials approach to sectoral taxation • Maximize stakeholder involvement and prioritize capacity building for tax and fees design and administration • Streamline the national tax framework in line with the global tax policy reforms of the digital economy (e.g., BEPS) <<< xix Risk area and economic Policy recommendations to mitigate each type of risk governance issues First- generation risks • Monitor and enhance transparency of both financial and nonfinancial disclosures by SOEs and SLEs • Mitigate conflict of interest by separating the body that exerts control and/or the board member that represents the state at the SOE/SLE from policy-making and regulatory entities and their employees Operation of digital SOEs/ • Enhance transparency and control of government support provided to SOEs and SLEs SLEs • Require SOEs to separate commercial activities from the delivery of any public service obligations (PSOs) as well as non-competitive services from services where competition Economic governance can develop, and use appropriate accounting mechanisms to minimize the risk of issue: Conflict of interest cross-subsidization of commercial activities with public funds and non-competitive with in the treatment of SOEs/ competitive services. Require digital sector SOEs to achieve commercial rates of return SLEs and rent-seeking by similar to comparable private businesses public officials • Remove any de jure monopolies granted to SOEs over the ownership and operation of digital infrastructure and the provision of digital services. • Remove preferential tax rates or forbearance on tax arrears and penalties, revise provisions in national budget laws that provide for direct on-lending to SOEs, adjust state-backed loan guarantee programs and other forms of preferential financing for SOEs Second - generation risks • Adopt economy-wide legal frameworks for data protection, cybersecurity, and cybercrime • Support the operationalization of data protection authorities (or entities with this mandate) and cybersecurity systems, including through regulatory measures, financial resources, and Data misuse capacity building • Provide clear guidelines for compliance and flexibility within the law, such as exceptions Economic governance for micro, small, and medium enterprises (MSMEs) or ex post facto liability for certain issue: Political or infringements commercially motivated • Engage in regional discussions and agreements about the interoperability of data protection surveillance rules and collaboration for the implementation of cybersecurity frameworks to support trusted cross-border data flows; adopt international conventions • Create enabling frameworks to facilitate use and reuse of data collected by public and private entities • Integrate pro-competition principles in the rules that govern competition in digital markets • Adopt competition law frameworks and support their operationalization • Build human capital and equip competition authorities with sufficient resources to pursue Anticompetitive outcomes competition enforcement and advocacy in a meaningful manner in digital markets • Bolster the de jure and de facto advocacy mandates and capacities of regional competition Economic governance authorities and engage in the implementation of competition policy under the African issue: Biased (unfair) Continental Free Trade Agreement (AfCFTA) and the Tripartite Free Trade Area, and foster outcomes for users of international cooperation digital technologies • Adjust current competition frameworks through regulation and guidelines to adapt it to the monitoring, assessment and design of remedies for anticompetitive practices and mergers involving digital markets xx >>> REGULATING THE DIGITAL ECONOMY IN AFRICA: MANAGING OLD AND NEW RISKS TO ECONOMIC GOVERNANCE FOR INCLUSIVE OPPORTUNITIES “[Information technology] lends itself to surveillance as well as liberty, to new forms of manipulation and covert control as well as new kinds of participation, to skewed, unjust market outcomes as well as greater productivity.” Benjamin Barter, Jihad vs. McWorld, The Atlantic, March 1992 <<< 1 1. >>> Introduction: Understanding the economic governance landscape of the digital economy in Africa As the use of digital technologies expands in Africa, countries are defining their social contract regarding their use and the governance of these technologies to fulfill their vision. Governments are setting priorities to reduce the digital divide, protect and use digital data, and use digital technologies to boost economic growth and social welfare. They are identifying measures to enable the development of the digital economy but also address risks that digitalization creates. This report focuses on selected topics to enhance the economic governance of the digital economy in Africa to achieve development goals more quickly.1 This means the “processes that support economic activity and economic transactions by protecting property rights, enforcing contracts, and taking collective action to provide appropriate physical and organizational infrastructure, (…) carried out within institutions, formal and informal” (Dixit 2008). More narrowly, “digital governance encompasses the norms, institutions, and standards that shape the regulation around the development and use of these technologies” (Runde and Ramanujan 2021). There is no doubt about the increasing use of digital technologies in Africa, especially after the COVID outbreak that sped up digitalization. The number of internet users has grown fast from 68 million unique mobile internet subscribers in 2010 to 692 million unique mobile internet subscribers in 2021 (GSMA Intelligence 2021). The use of digital services offered over digital infrastructure has also expanded, from 35 million active mobile money accounts in 2011 to 518 million in 2021 (GSMA 2022). The number of active digital platform users has also boomed, with over 210 million Facebook2 and 2 billion YouTube users. More than 800 million citizens have digital IDs, and governments are expanding the reach of digital ID systems. Governments have also started offering services online, with 37 and 39 African countries3 having operational portals for e-tax and e-procurement, respectively. Enterprises have also increased adoption of digital technologies with 37 percent investing in them after the COVID outbreak (World Bank COVID Business Pulse Survey, 2021). 1. This report focuses on economic areas and does not cover social, political, and civil aspects affected by digitalization. 2. https://enitiate.solutions/africas-top-10-countries-contribute-a-combined-76-to-the-continents-facebook-active-users/. 3. Dener, Cem, Hubert Nii-Aponsah, Love E. Ghunney, and Kimberly D. Johns. 2021. GovTech Maturity Index: The State of Public Sector Digital Transformation. Interna- tional Development in Focus. Washington, DC: World Bank. doi:10.1596/978-1-4648-1765-6. E-tax portals include those that allow for transactional services or those portals interconnected to other Governmental portals. E-procurement portals include those operational portals that only publish tenders and those that also include contracts. Data for both indicators is for the year 2020. 2 >>> INTRODUCTION Recent empirical evidence highlights the importance of digital Roadmap for the report technologies for more and better jobs in Africa (Begazo, Dutz, and Blimpo, 2023). Firms that use more sophisticated digital This report analyses risks related to economic governance technologies in Ghana, Kenya, Malawi, and Senegal were along the digital value chain, focusing on digital infrastructure associated with higher productivity and job growth (Cirera, and data regulation as enablers for the digital economy in Comin, and Cruz, 2022). There is a positive correlation Africa. The report is structured around four interconnected between the use of more sophisticated digital technologies by chapters. This first chapter sets the framework to assess microenterprises (often informal) and higher levels of jobs, as governance of the digital economy and provides an overview the study on Ghana, Kenya, Mozambique, Nigeria, Senegal, of the various stakeholders that interact in the African digital South Africa, and Tanzania shows. Importantly, it finds a ecosystem. This chapter also presents an overview of how positive progression between the use of more sophisticated markets are working and to what extent the objectives of digital technologies like computers or smartphones and inclusion and efficiency are being reached, given the current higher productivity, sales, and job levels (Atiyas and Dutz, governance of the digital economy. 2022). Internet availability has led to job creation and poverty reduction across Africa. For example, in Nigeria and Tanzania, The second chapter presents traditional risks to governance internet availability increased labor force participation and in foundational elements of the digital economy: digital reduced poverty rates by 7 percentage points, with more infrastructure. This chapter covers elements to address the significant welfare effects on poorer and less-educated risk of widening digital divides. To complement the World Bank households (Bahia et al. 2020, Bahia et al. 2021). The arrival report “Digital Africa: Technological Transformation for Jobs” of faster internet in Sub-Saharan Africa, facilitated by the it focuses on the taxation of digital infrastructure and direct gradual arrival of submarine cables, benefited enterprises state participation as supplier through state-owned enterprises and households located in the vicinity of faster networks. The (SOE) and state-linked enterprises (SLE). The first section probability that an individual is employed increases between maps and analyzes different contributions and tax schemes 3.1 and 13.2 percent for African countries in different samples applicable to digital infrastructure operators and highlights gaps (Hjort and Poulsen 2019). to implement more transparent, effective, and equitable taxation and parafiscal fees policies to support the development of the However, increasing use of digital technologies also implies digital economy. The second section aims to better understand higher exposure to risks: First, the risks of widening the digital governments’ direct participation in digital markets and how divide and leaving citizens outside the digital economy and the regulation and corporate governance rules can better support opportunities it brings4; second, exposing users to exploitation the performance of state-owned and state-linked enterprises of market power by providers of digital services; and third, (SOE/SLE) and sectors where SOEs participate. exposing users to misuse of data and digital technologies5. Cyberattacks and cybercrime have increased, affecting trust The third chapter highlights new (second-generation) risks, in digital technologies, for example, using transactional mobile given the digitalization of the economy and the increased apps or mobile money. Digitalization of IDs has increased importance of data-driven businesses. This chapter zooms concerns about personal data protection and demanded in on competition enforcement and policy to address market careful design of ID systems. Besides risks linked to economic power issues by new digital business models and data governance, other political and social risks also arise. The regulation to generate trust for the use of digital technologies larger digital footprint of using digital applications opens the and allow for the generation of digital technologies. possibility of political surveillance; controlling internet access becomes a means of political control, and influencing content The final chapter presents a heatmap of the various degrees of can lead to social manipulation. risks to governance faced by African countries, including first- generation and second-generation risks. First-generation risks focus on foundational digital infrastructure, including market 4. On the opportunities to close the digital divide through more affordable availability and attractive use of digital technologies in Africa, seeBegazo, Dutz and Blimpo 2023 5. On the risks and opportunities related to the use of digital technologies and data, see WBG (2021) World Development Report 2021: Data for Better Lives. <<< 3 characteristics, institutions and regulations, SOE presence and economy in Africa and the rules that govern the sector, various governance, and taxation. Second-generation risks relevant questionnaires and databases have been developed based to digital markets beyond digital infrastructure encompass on primary data collection, interviews, and proprietary data. competition policy and enforcement as well as governance Other global databases developed by the World Bank, such of data. This chapter also includes recommendations for as the Digital Business database, the Global Data Regulation different sets of countries to prioritize interventions using as database, and the Digital Antitrust database, have also been background the results from scores based on proxies for the employed for the analysis. Information presented in this report identified risks (Annex 2). was collected and systematized by March 2022, and the latest available data corresponds to 2021 unless otherwise noted. About the report on Africa digital governance This report on Africa’s economic governance of digital is part of the Africa Digital Economy Governance study (Box 1) that This report has been developed based on novel information aims at supporting the implementation of the African Union databases collected for the preparation of the report. Given Digital Transformation Strategy and the Digital Economy for limited systematized information on the functioning of the digital Africa initiative6. BOX 1: AFRICA DIGITAL ECONOMY GOVERNANCE This report is part of a broader effort to support Africa’s ambitions to take advantage of the digital revolution to improve governance. The main objectives are to document digital tools that can expand the digital economy’s development impact on governance and identify critical governance risks that can dampen the effect of digital technology adoption. Volume 1. Volume 2. Digital Tools Governance for Better of the Digital Governance Economy The present report, Volume 2: Governance of Digital, analyses risks related to economic governance along the digital value chain, focusing on digital infrastructure and data regulation as enablers for the digital economy in Africa. The companion report, Volume 1: Digital for Governance, builds on regional and global work related to the use of digital technologies by governments in order to strengthen governance and reduce the space for corruption. This report covers: • Progress in the adoption of digital tools and necessary complements for better governance in Africa • Digital tools for providing information to the public • Digital tools for streamlining the provision of government services and supporting efficiency • Digital tools for strengthening participation, accountability, and anti-corruption • Governance risks in public procurement of ICT solutions 6. https://www.worldbank.org/en/programs/all-africa-digital-transformation 4 >>> INTRODUCTION The two volumes of Africa Digital Economy Governance are based on more detailed studies of various aspects of the digital economy and governance in Africa. These background studies provide considerable details on topics such as: • Procurement of Information and Communications Technologies (ICT) in Africa. • Adoption of e-Government Procurement (e-GP) in Africa • Vulnerabilities of ICT Procurement to Fraud and Corruption • Regulating Digital Data in Africa • Taxes and Parafiscal Fees on Digital Infrastructure Services in Africa • Corporate Governance and Transparency of State-Owned and State-Linked Digital Enterprises in Africa • State-Owned Enterprises in Digital Infrastructure and Downstream Digital Markets in Africa • Antitrust and Competition Advocacy for Digital Markets in Africa • Competition Policy in Digital Markets in Africa 1.1. The evolution of economic element in the development of mobile communications worldwide. In the past, digital technologies encompassed governance of digital voice communication and simple data exchanges. Nowadays, the digitalization of voice, written documents, and transactions Economic governance of digital encompasses the norms, allows for high traceability of economic interactions and institutions, and standards that shape the regulation around individual behavior and the generation of data, enabling the development and use of digital technologies. These rules the use of new technologies such as big data analytics, determine the private sector operating environment as well as artificial intelligence, blockchain, and other distributed ledger the role of the state in the sector as a policy maker, regulator, technologies. Risks related to exclusion from and misuse and service provider. These rules are the result of the of digital technologies were more limited in the past. In the country’s economic goals: collective and individual interests past, risks of exclusion from digital technologies were mainly of the government, private actors, and civil society. In the case associated with a lack of connectivity of individuals, firms, and of digital technologies, the private sector plays a key role as governments to telecommunications (digital infrastructure) service provider along the digital value chain and as motor of networks at affordable prices and with good quality of service technological and business innovation. Given the fast pace of and a weak environment to curb market power, ensure a level technological change, adaptive economic governance -based playing field for service providers and foster private investment. on agile approaches- is becoming more critical to setting rules Risks of misuse of digital connectivity networks were mainly to govern the digital economy. circumscribed to access to private communications through wiretapping and unwanted calls. Currently, the risks of Good economic governance of digital involves managing exclusion from digital technologies are larger given their risks to enable and maximize economic opportunities for the important role as a general-purpose technology and enabler of private sector, public sector, and citizens. The stakes of poor economic activity across sectors. Risks of potential exploitation governance of digital are expanding as digital technologies of users are also larger, given the market power of providers evolve and digitalization grows and deepens across the of new technologies that enjoy economies of scale and direct economy; risks are evolving (Figure 1). Good governance is and indirect network effects. The risk of digital technology essential to the development of digital technologies; as shown misuse has become more complex given cybercrime and by Faccio and Zingales (2017), the actions of regulators and cybersecurity breaches, unlawful and unethical use of data policymakers presenting regulatory strategy that promotes for exploitation, surveillance, and misinformation, as well as competition, and lower costs for consumers was a key control of internet usage through internet shutdowns. <<< 5 >>> Figure 1: Evolving risks of digitalization – first and second-generation risks Risks of digitalization (increasing use of digital technologies) Risk of exclusion from digital technologies Risk of misuse of digital technologies Past context: voice and New context: big data, Past context: voice and New context: big data, simple data exchange AI, blockchain simple data exchange AI, blockchain Cybercrime, security Digital divide (use of Connectivity divide: breaches digital technologies): Africa-ROW, rural-urban, Widening gap of gender, income Unlawful and unethical disruptive tech adoption Access to private access/use of (personal) by individuals, firms and Market power and rent communications: data governments seeking in digital infra: surveillance/wire tapping monopolies, protected Surveillance, Market power in digital SOEs, expropriation via misinformation and markets (winner takes taxation internet shutdowns to most) restrict use of digital Source: Own elaboration Weak economic governance of digital can be reflected in North America, and ECA and less than one-fourth of the time market outcomes along the value chain. Availability and use of than subscribers in MENA, LAC, and EAP.7 Although digital voice and data communications is limited in Africa compared to technologies bring opportunities to increase productivity, other regions, with 19 percent of the population not covered by create jobs, and improve household welfare in Africa (Begazo, mobile networks that can support internet and unique internet Dutz, and Blimpo, 2023), rules and government interventions users representing only 28 percent of the population in 2021 have not been fully effective in enabling digital transformation. (GSMA 2022). The connectivity gap is more pronounced for The COVID pandemic has boosted digitalization, but risks the rural population, low-income groups, and women. People related to digital exclusion and misuse of digital technologies in the wealthiest quintile of consumption or assets are fifteen have become more visible. times more likely to have access to a computer or tablet than those in the bottom quintile, and urban residents are over Traditional economic governance risks related to digital three times more likely to have access to the internet than infrastructure are still important in Africa. How a government rural residents (Seuyong et al., 2023). Relatively low levels of designs and implements the rules of the game has an impact on use of digital technologies by individuals and firms to support market structure, market dynamics, affordability, and uptake. economic activities, as well as the supply of digital solutions, Weak competition along downstream and upstream markets are another example of gaps in the economic governance of affects digital connectivity. Both ineffective pro-competition the sector. For instance, mobile internet subscribers in SSA regulation and market distortions caused by the state’s direct spend less than half of the time on apps than those in SAR, service provision through state-owned enterprises (SOE) 7. Apptopia (2020-21). The sample includes 63 lower and upper middle- and high-income countries. African countries included are Egypt, Tunisia, Kenya, Malawi, Tanzania, South Africa, Ghana, Burkina Faso, Senegal, and Nigeria, all middle-income countries. 6 >>> INTRODUCTION and state-linked enterprises (SLI) – operators with minority economic governance of digital markets and digital business state shareholdings – determine competition dynamics. models include ministries of ICT and digital economy as well Furthermore, rules to extract rents from the sector through as line ministries across various sectors (such as industry, taxation and parafiscal fees or operation of SOE and SLE agriculture, health, transportation), regulators (in digitalized can generate economic governance risks through indirect sectors and for data protection and cybersecurity), private mechanisms for expropriation of private rents. Traditional sector from international and regional service providers to issues that affect governance are still problematic: SOE local startups as well as users of digital technologies (both corporate governance, competitive neutrality, and sectoral individuals and enterprises), civil society and international rules to prevent distortions, as well as taxes and fees that and regional organizations. Limited digital trust due to the are relatively high, complex, and unpredictable and create a initial unregulated nature of new digital markets, asymmetry dependency relationship between the government and service of information and knowledge between the private sector and providers. Key actors to define the economic governance public sector, and government interest in controlling data and of digital infrastructure include the ministry in charge of information flows affect the governance of digital markets. digital infrastructure (ministry of ICT, communications, digital economy), the regulator of digital infrastructure, the institution(s) that determine taxation of the sector (ministry 1.2. An overview of the of finance, revenue collection authority, other government institutions that set fees on the sector), the private sector characteristics and outcomes of (incumbent and entrant service providers, potential investors, digital markets in Africa: do they sectoral associations), SOE and entity that controls the state’s shareholdings (line ministry, ministry of finance, holding), mirror good economic governance? the civil society (consumer organizations, academics), and international and regional institutions (regional communities Well-functioning markets mirror the existence of good and institutions setting international standards such as economic governance that balances market power between ITU, ICANN). Regulatory capture and conflict of interest of service providers and consumers and addresses market failure government institutions influence how rights and obligations along the value chain of the digital economy. Africa is lagging are established and how differences are mediated, favoring behind other regions in terms of market outcomes: penetration certain economic interests and disregarding policy objectives of digital technologies and data consumption is relatively low, of higher adoption and use of digital connectivity. prices of internet (data connectivity) for individuals/families and small businesses are high, and quality is low (low speed New economic governance risks are appearing as digital and high latency), while use of mobile apps by unique mobile technologies evolve, and African countries have started subscriber and sophistication of digital technologies offered adopting and implementing frameworks to address issues in online is limited (Begazo, Dutz and Blimpo 2023) (see Box 2). digital markets beyond traditional digital infrastructure. Key elements for economic governance of digital markets include These market outcomes indicate that there are problems rules to protect and foster competition (for example, related related to the economic governance of the sector and the to digital platforms), rules to protect personal data, rules to effectiveness of government interventions. Having markets combat cybercrime and promote cybersecurity, and rules to that are persistently more concentrated in the hands of few enable the use and reuse of data for the development of digital players (and closed to smaller players or new entrants) makes solutions (World Bank Group, 2021). In Africa, tools to address good economic governance more difficult, given potential new challenges of digitalization are yet to be in place, given issues of regulatory capture and asymmetry of economic and nascent data regulations on privacy and cybersecurity with political power. Greater importance of state ownership to carry limited enforcement and limited enforcement of competition out commercial activities also creates risks of distorting the rules in digital markets. Stakeholders that influence the level playing field due to conflict of interest of the government <<< 7 as service provider, regulator, and policymaker. Although more challenging. This section describes the main players and challenges posed by market structure and characteristics of market features along key segments in the digital value chain market players can be addressed through robust regulation by across 54 countries in Africa to understand the market context transparent and accountable institutions, this task becomes for better economic governance of the digital economy.8 BOX 2: WEAK MARKET OUTCOMES IN AFRICA Challenges that exist in Africa, such as lack of pro-competition market rules and adequate government interventions to incentivize competition, result in unfavorable market outcomes. These include: 1. Large use gap Africa has the largest uptake gap in the world, as a share of its countries’ populations with internet availability: over two-thirds of Africa’s people with 3G+ coverage are not users, and the uptake gap is expanding. This could be because of lower levels of income or a lack of policies and investments that promote use. Mobile data traffic is very low in Africa compared to the average in low-middle-income economies. 2. Limited internet penetration While coverage has significantly expanded, this has been observed mostly in urban areas. In contrast, rural areas remain behind – almost 20 percent of the rural population has no mobile broadband coverage. Only 62 percent of the rural population has access to a mobile broadband network compared to 99 percent in urban areas (ITU, 2020). 3. High prices Prices have fallen over time; however, the cost of both data and devices remains restrictively high. For example, the average cost of 1.5 GB as a share of GNI per capita was 9 percent in Africa (9.4 percent in SSA) and more than 20 percent in DRC, CAR, Sierra Leone, Chad, and Guinea-Bissau, which is well above the Broadband Commission for Sustainable Development’s goal of 1 GB for less than 2 percent of monthly GNI per capita. (https://www. broadbandcommission.org/broadband-targets/). Further, the prices of smartphones as a percent of GNI per capita are also higher in West and Central Africa than in other LMICs. These, coupled with broadband price–discriminating strategies, further exacerbate the usage gap. 4. Poor quality of mobile internet services Low speed and high latency affect quality of service, i.e., 18 SSA countries have an average download speed below 25 Mbps compared to 78 Mbps available on average in HICs, and none have an average download speed above 100 Mbps. Slower speed and higher latency are also prevalent in rural areas, particularly due to limited data infrastructure. Further, low-quality services limit the kind of digital content and technologies that can be offered, thus affecting consumer experience and growth of digital services. Source: Begazo, Dutz and Blimpo 2023. 8. For simplification purposes, the analysis has been done at the segment level, which in some cases might coincide with the market boundaries for country-level compe- tition assessments based on supply and demand substitutability analysis. A complete market dynamics or effective competition assessment of markets is beyond the scope of this report. However, findings provide elements to identify potential market problems to be investigated further. 8 >>> INTRODUCTION There are several markets along the digital economy value fixed and mobile internet services (internet service providers chain, encompassing both digital infrastructure (connectivity and mobile network operators) purchase connectivity from and data infrastructure) and digital services (Figure 2). upstream service providers in wholesale markets. Mobile Conceptually, the first group of markets of the digital economy wholesale markets allow mobile network operations (MNOs) value chain (upstream digital connectivity segment) is the first to sell usage rights to parts of their networks through roaming mile, where the internet links up with a country.9 This includes or virtual operator (MVNO) agreements. In downstream digital markets for international connectivity, such as submarine infrastructure markets, the last mile, MNOs, and ISPs sell cables and other gateways, satellite dishes, and cross- mobile and fixed internet services to individual and business border microwave. The middle mile of digital infrastructure customers. Both wholesale and retail services are part of the comprises segments that allow internet to pass through a last mile. Data infrastructure and services (internet exchange country. This includes fiber backbone and metro networks points, colocation data centers, data exchange facilities, and as the core internet network in a country that connects cloud services that offer storage and computation capacities) international gateways and local networks, cell sites, and are part of the next segment that provides important inputs for transport connectivity across long distances in a country. the provision of digital services. Finally, digital services include Passive infrastructure, which is non-electronic components all services that can be provided over digital connectivity, of a telecom network, such as telecom towers, buildings, including mobile money (a low-barrier technology for financial ducts, and services such as security and air conditioning, transactions), digital platforms, and other services offered by are also part of the middle mile segment. Providers of retail digital businesses. >>> Figure 2: Simplified digital economy value chain and market segments Fixed wholesale Fixed Data storage Mobile money International (fiber backbone) Mobile retail connectivity wholesale Internet Exchange (cables, (roaming, Digital platforms Passive Points gateways, MVNO, satellite dishes) infrastructure (e.g. sharing) Mobile telecom towers, retail Cloud services and Other digital solutions ducts) other data services First mile Middle mile Last mile Data (Upstream (Middle (Downstream connectivity layer) connectivity layer) connectivity layer) infrastructure Digital services DIGITAL INFRASTRUCTURE Source: Elaboration by team based on World Bank Group (2021) and Competition Commission of South Africa (2019). Concentration, state ownership that affects the level playing to develop, therefore likely leading to biased market outcomes field, and vertical integration might inhibit the ability of markets (higher prices, lower access and use for consumers, higher in the digital economy to deliver low-cost, high-quality, and markups for operators, and lower entry, among others). The innovative digital services. economic characteristics of digital infrastructure markets – in particular, high fixed costs, economies of scale, and network Persistent high levels of market concentration10 means that effects – make these markets prone to higher concentration the environment is more challenging for effective competition levels. For example, in the mobile retail segment, the high 9. Adapted from WBG (2016) World Development Report 2016: Digital Dividends to show economic vertical relationships along the value chain between upstream and downstream markets. A similar value chain analysis has been used to assess data service markets in South Africa by the Competition Commission (Competition Com- mission of South Africa 2019, available at https://www.compcom.co.za/wp-content/uploads/2019/12/DSMI-Non-Confidential-Report-002.pdf ). 10. Market concentration exists where a large share of the market is controlled by a small set of firms. Typical concentration indicators include concentration ratios for the largest 2, 3 or 4 firms or the Herfindahl- Hirschman Index (HHI) that accounts for the relative importance of all market players (sum of squared market shares of all market participants). <<< 9 initial investment required to build mobile infrastructure makes associated investment costs. Highly concentrated markets it difficult for new MNOs to enter the market by building can still be competitive if they are contestable. Markets are competing mobile infrastructure, while network effects create contestable if there are no sunk costs such that barriers to incentives for consumers to stay with the largest (incumbent). entry and exit are low and potential entrants have equal access While high market concentration is not necessarily indicative to the relevant technology as incumbents. In such markets, of lack of competition, other factors are important to gauge incumbent firms face a higher threat of entry, whether the the risks of market distortions, such as high structural and threat is actual or perceived. regulatory barriers to entry, including technology features and BOX 3: ASSESSING COMPETITION IN THE DIGITAL ECONOMY Assessing competition in a market is not straightforward and requires a diverse set of quantitative proxies combined with market-specific qualitative analysis. This report uses a set of quantitative indicators available for most African countries as proxies of potential risks to competition along the digital value chain. Market concentration is commonly used to proxy for the intensity of competition. This is based on insights that it is easier for firms to behave in an anticompetitive manner in more concentrated markets. As a result, competition authorities and policymakers often use structural indicators that measure the extent of market concentration, such as the Herfindahl-Hirschman Index (HHI), to screen for potential competition issues in markets. The HHI is one of the indicators that is available across most segments of the value chain and African countries. Although the analysis of the level and evolution of HHI can give insights into market dynamics, the following limitations should be considered. 1. Interpretation. Instead of competition itself, concentration indicators measure structural market outcomes but not other factors that can determine competition outcomes (see below). 2. Lack of comparability across markets. Market characteristics are important determinants of how concentration affects competition risks. Therefore, concentration indicators may have a slightly different meaning across different markets. For example, markets with high fixed costs of entry and limited demand (market size) may naturally present high concentration. 3. Data intensity and lack of data. The HHI requires complete data on market shares of all participating firms in a market. This data may not be available across markets and economies or may be inconsistently reported. The unit used to calculate market shares affects the HHI level. 4. Market definition. The definition of a relevant market is crucial to assessing its concentration. Data is often reported on a national and industry aggregate. However, relevant markets may be sub-national, regional, or international and are often more granular than pre-defined industry classification systems. 5. Partial ownership and cross-ownership. Partial and cross-ownership patterns may also reduce incentives to compete for involved firms. Concentration indicators may not capture these ownership patterns sufficiently and, therefore understate competition risk in a market. While the mapping that is used in this report tries to capture ultimate ownership as accurately as possible for firms with state participation, it does not account for private partial or cross-ownership in cases where ownership shares of private firms are less than 50%. 10 >>> INTRODUCTION The report addresses the limitations of the HHI index by considering other indicators that shed light on risks to competition on the market. Assessing the evolution of the HHI to determine whether market concentration has been stable or not over time, entry and exit in the market, and vertical integration typically provide a better overview of competition dynamics. While individually, all indicators that aim to measure competitive pressure on markets are imperfect and give an incomplete picture, together they can provide useful information (OECD, 2021). All these aspects combined provide an overview on the level of risks to competition that can be a point of first entry for a thorough analysis of competition on specific markets. Such an analysis would also take into account other structural and regulatory aspects as well as more granular and comprehensive data, as well as complementary qualitative information on firm and consumer behavior. Source: Elaboration by team based on OECD (2021) and the WBG Market and Competition Policy Assessment Toolkit (MCPAT). Often, state-owned (SOE) or state-linked enterprises (SLE) environments, accountable and transparent government are more likely to receive undue advantages from the interventions to monitor markets and regulate them more government, for example, subsidies, access to preferential effectively are necessary to ensure current investors, users, contracts and other resources (radio spectrum), and favorable and new companies share the benefits of the digital economy. regulatory treatment. Digital connectivity has a history of state involvement because of the high capital investment A newly created database on ownership patterns in digital required and strategic importance of the sector, although economy market segments in Africa shows that segments now the sector is mostly led by the private sector. Weak remain concentrated with limited entry and persistent state competition due to preferential treatment might make it easier involvement. The Africa Digital Market Players database for incumbent SOEs or SLEs to sustain higher prices or lag maps actors along eight segments in the digital economy in in network expansion and quality improvements. Finally, Africa (see Annex for methodology), including their ownership vertical integration increases the risk of market foreclosure to and activity across these segments, market shares (where entrants and other competitors in downstream markets. In the available), parent companies and state involvement, and digital economy, vertical integration is especially relevant in vertical integration across several market segments. The data key input markets such as broadband backbone, international shows concentrated segments in the digital economy, with connectivity, and mobile services. an average HHI above 5,000, which is considered extremely highly concentrated and equivalent to market structures Although digital services are also characterized by economies where only two equal-sized operators are present.11 The level of scale and scope and direct and indirect network effects of market concentration varies by segment and country, with that point to winner-takes-most dynamics, they are less extremely high concentrated markets (HHI above 6,000) in ten concentrated in most segments. However, mobile money is countries. While data center and cloud markets appear less more concentrated than mobile retail services, and general concentrated, there might be increased consolidation in the use of mobile apps is tilted toward a few providers. In these coming years. 11. Concentration is calculated at the national level for specified segments. HHI above 2,500 is considered as highly concentrated in line with the US merger guidelines and a red flag for potential competition issues. <<< 11 >>> >>> Figure 3: Average concentration in the digital economy Figure 4: Average concentration (HHI) across seven key across Africa, by segment segments of the digital Average value HHI across chain, digital value by country chain segments 10000 3,503 9000 4,186 8000 7363 5,075 6,836 4,540 7000 6161 7,362 6000 5427 4,551 6,718 5,068 4900 4919 7,638 5,504 5,052 10,000 4562 7,341 6,732 6,954 4,683 5000 7,345 4057 6,069 5,431 6,671 6,065 2,263 9,243 3,290 5,196 6,300 7,077 4000 3,844 2,921 5,772 4,015 3,250 3000 9,377 6,943 5,181 4,372 3,432 3,613 8,389 3,334 2000 2817 Average HHI 2422 2332 2,200 7,286 1000 1724 3,500 4,348 6,036 5,000 3,965 0 726 737 600 6,750 4,072 8,500 5,766 4,428 4,861 10,000 6,473 5,997 ey es il ne s rs il ta er ta No data we bl on bo re nt re 6,882 ca m to ce ck ile d xe e ba 5,554 m e ob a in il 3,086 co Fi at ob r M ar D be le M bm Te Fi Su IBRD 47442 | Average HHI Min HHI Max HHI AUGUST 2023 Note: Includes data for 54 countries in mobile retail, 38 in fixed retail, 52 in fiber backbone, 26 in telecom towers, 35 in submarine cables, 25 in data centers, and 15 in mobile money. Concentration measured using HHI. Source: World Bank, Digital Market Players Database (2022) Economic governance issues persist across digital markets 36 percent and foreign SOEs/SLEs around 13 percent (see in Africa, including high levels of concentration and vertical Figure 5). The largest individual groups in digital infrastructure integration and extensive ownership of infrastructure by SOEs are all partly owned by foreign and domestic governments: and conglomerates. The database shows a high degree of Orange Group, of which the Government of France is a vertical integration, with around 65 percent of MNOs with minority owner; MTN Group, partly owned by the Government market shares above 40 percent active in 2 or more other of South Africa; and Etisalat/Maroc Telecom, which are owned market segments (58.8 percent of ISPs). Vertical integration by the governments of UAE and Morocco, respectively (see of large operators creates risks to competition along the value Figure 6). chain, such as increasing the risk of market foreclosure to downstream rivals. Further, SOEs/SLEs are important players in fiber backbone and international connectivity but active along the entire value chain and with increasing ownership in some segments such as mobile retail and data centers. Less than half (46 percent) of African digital infrastructure operators are private firms, with national SOEs or SLEs holding around 12 >>> INTRODUCTION >>> >>> Figure 5: Ownership of operators in key digital infrastructure Figure 6: Ownership in key digital infrastructure and and services segments across Africa, by ownership type services segments across Africa, by ultimate owner Other Foreign 6.0% Orange Group 6.9% SOE/SLE MTN Group 5.5% owned Etisalat 4.1% 12.6% Bharti Airtel 3.6% Vodafone Group 3.1% Econet Global 2.5% Axian Group 1.3% Private Government of Zimbabwe 1.3% 45.5% Sudatel Telecom Group 1.3% Government of Algeria 1.2% SOE/SL Viettel Group 1.1% E-owned Government of Tunisia 1.0% 35.9% Government of Angola 0.9% Nippon Telegraph and… 0.9% Africell Holding 0.8% Note: Shows share of firm ownership of operators in seven segments of the digital value chain in Africa. Source: World Bank, Digital Market Players Database (2022) In sum, markets are still concentrated with monopolies of this section will describe market characteristics along operating in some countries, new players in a few countries, digital segments. and SOE and SLE presence (Figure 7). The remainder >>> Figure 7: Simplified digital value chain and key market characteristics that affect economic governance Source: World Bank, Digital Market Players Database (2022) <<< 13 1 . 2 . 1 . I N T E R N AT I O N A L C O N N E C T I V I T Y Equiano by Google) with additional capacity of 180Tbps and 150Tbps (Box 3). While these cables will increase connectivity to Africa is getting more and more connected. The number of a large proportion of the African population, given their extensive submarine cables landing in African countries has increased length and capacity, they will also increase the operators’ overall over the past decade, thus significantly facilitating the shift from ownership and participation in submarine cable infrastructure dependence on satellites for international internet connectivity and to more than 30 cables by 2024. This raises concerns such reducing the overall cost.12 There are more than 49 international as control and reliability due to the potential to prioritize their cables and 9 domestic cables operational in 38 coastal countries, services over other providers, especially cables wholly owned by with landlocked countries in Africa connected through terrestrial WNOs.16 Demand for connectivity continues to increase in Africa, fiber networks except for three countries.13 SAT3 was amongst especially with the advancement in internet penetration and the first cables to land in 2001, with only two other major technology; therefore, the timely increase in international internet cables (SAS-1 and SEA-ME-WE4) becoming operational until bandwidth is essential. The total international bandwidth in Africa 2009 which were largely owned by a consortium of incumbent has increased over five times in the past five years, i.e., from telecom operators in the landing countries. Thereafter, private 3827 Gbps in 2016 to over 17000 Gbps in 2020, representing the sector investment in subsea cables saw14 a large influx of major most significant growth over all other regions (Figure 8). Internet cables such as SEACOM, TEAMs, LION (2009), Main One, bandwidth per capita is growing significantly, with an average of EASSy, GLO1, I-ME-WE (2010), EIG (2011), WACS, LION2, 49 percent compound annual growth rate (CAGR) for the period SEAS, ACE (2012). Furthermore, several cables have also been 2011-2020 in the three sub-regions over the last ten years in announced and are under construction, with the most notable Africa (Figure 8). However, Sub-Saharan Africa significantly lags being those by webscale network operators (WNO)15 set to enter other regions in terms of bandwidth per inhabitant. the subsea cable market (2Africa supported by Facebook and >>> Figure 8: International internet bandwidth per inhabitant 1,400 55% 1,174 International Bandwidth per inhabitant 1,200 50% 50% 49% 48% 48% 987 1,000 44% 45% 41% 800 36% 674 35% 600 30% 359 28% (Kbps/pop) 400 207 27% 25% 134 185 200 4 11 27 28 54 20% - 15% North America East Asia & Pacific North Africa Upper-Middle Income Europe & Central Asia West & Central Africa Latin America & Lower income East & Southern Africa South Asia Lower- Middle Income Caribbean High Income 2020 CAGR Note: Yellow highlights indicate income classification, CAGR (2011-2020) Source: Compiled using international bandwidth data (2020) from Telegeography Global Internet Geography and Population data: https://statisticstimes.com/demographics (2020) World bank : World Development Indicators https://datacatalog.worldbank.org/public-licenses#cc-by (2011); Income level and regional classifications: https://datahelp- desk.worldbank.org/knowledgebase/articles/906519-world-bank-country-and-lending-groups 12. Schumann and Kende (2013) ‘Lifting barriers to internet development in Africa: suggestions for improving connectivity’ Analysys Mason, Internet Society available at https://www.intgovforum.org/multilingual/system/files/filedepot/17/Record percent2024a.pdf 13. According to TeleGeography, as of July 2021, Only three countries are not connected to international bandwidth either through submarine cables or terrestrial fiber; these are the Central African Republic, Eritrea, and South Sudan. Sudan is in the process of developing an international data gateway. https://itweb.africa/content/G98Ydq- LY3z2vX2PD 14. ITU (2017) ‘Maximising availability of international connectivity in developing countries: Strategies to ensure global digital inclusion’ Available at: https://www.itu.int/ dms_pub/itu-d/opb/pref/D-PREF-BB.GDI_01-2017-PDF-E.pdf 15. Webscale Network Operators (WNO) are “web-centric companies who own and operate large multi-country communications networks based upon hyperscale data cen- ters and submarine cables. Webscale operators may also own access networks, typically using fixed satellite or fiber. WNOs own network assets in order to support their core businesses, which comprise digital advertising, online retail, cloud, digital media, and content services, and other digital-based services.” https://www.businesswire. com/news/home/20210415005461/en/Webscale-Network-Operators-4Q20-Market-Review--- Webscale-Sector-Benefits-from-COVID-19-Pandemic-as-Annual-Reve- nues-Rise-18-to-1.71-Trillion-CAPEX-Up-28-to-133B--- ResearchAndMarkets.com 16. https://restofworld.org/2022/google-meta-underwater- cables/?campaign_id=158&emc=edit_ot_20220511&instance_id=61067&nl=on-tech-with-shira- ovide®i_ id=63413258&segment_id=91945&te=1&user_id=c0dd59226c7ce9b72f555343108c088f 14 >>> INTRODUCTION The emergence of intra-African cables has also increased have also improved connectivity in SSA, especially by also international internet bandwidth in the Africa-Africa corridor for providing landlocked countries like Uganda, Rwanda and example, the Eastern Africa Submarine Cable System (EASSy) Botswana with access to international bandwidth.19 West and which has more than 10 Tbps capacity and landing points in North Africa are lagging in regional connectivity but with plans nine Eastern and Southern African countries17. This has led to improve, such as the cross-border network planned by to an increase in the share of bandwidth flowing through the Orange and, Djoliba.20 Africa’s international internet capacity Africa-Africa corridor, from 451 Gbps to 2822 Gbps between is set to increase even further, with more than five cables 2016 and 2020.18 Cross- border terrestrial fiber networks, by planned or under construction, some of which will include private operators, such as Liquid Telecom, and governments, open access and carrier-neutral services (See Box 4). BOX 4: SUBSEA CABLES PLANNED OR UNDER CONSTRUCTION Equiano Google’s Equiano subsea cable, running from Portugal to South Africa, is due for completion in 2022. Equiano will start in western Europe and run along the west coast of Africa, between Portugal and South Africa, with branching units along the way that can be used to extend connectivity to additional African countries. Equiano will use optical switching at the fibre pair level, making it easier to allocate capacity as needed, simpler to deploy, and less expensive than other systems built to date. Fully funded by Google, the cable is expected to have 20x the capacity of the last cable built to serve the region. In partnership with Internet service providers (ISPs) and telcos, Equiano will bring capacity to more countries in Africa. Source: https://www.ifc.org/wps/wcm/connect/publications_ext_content/ifc_external_publication_site/publications_listing_page/ google-e-conomy 2 Africa 2Africa subsea cable, a partnership between China Mobile International, Facebook, MTN Global Connect, Orange, Saudi Telecom Company (stc), Telecom Egypt, Vodafone and West Indian Ocean Cable Company (WIOCC), is one of the largest subsea projects in the world, connecting 23 countries in Africa, the Middle East and Europe. At 37,000km long, 2Africa will be one of the world’s largest subsea cable projects and will interconnect Europe (eastward via Egypt), the Middle East (via Saudi Arabia), and 21 landings in 16 countries in Africa. The system is expected to go live in 2023/4, delivering more than the total combined capacity of all subsea cables serving Africa today, with a design capacity of up to 180Tbps on key parts of the system. 2Africa will deliver much needed internet capacity and reliability across large parts of Africa, supplement the fast- growing capacity demand in the Middle East and underpin the further growth of 4G, 5G and fixed broadband access. In countries where the 2Africa cable will land, service providers will obtain capacity in carrier-neutral data centers or open-access cable landing stations on a fair and equitable basis. This will support healthy internet ecosystem development by facilitating greatly improved accessibility for businesses and consumers alike. Source: https://www.2africacable.com/about 17. Telegeography International Internet Geography (2021). International Internet Bandwidth by Regional Routes, 2011 -2020 18. South Africa, Sudan, Mozambique, Madagascar, the Comoros, Tanzania, Kenya, Somalia and Djibouti. See https://www.afdb.org/en/projects-and-operations/select- ed-projects/eassy-the-eastern-africa-submarine-cable-system-156. 19. Ethiopia laid a fiber optic cable through Sudan at the end of 2005, allowing a connection to the landing station at Port Sudan, ending its dependence on satellite. On a larger scale, the World Bank has helped to fund RCIP and a fibre optic network as part of a project to interconnect electricity grids in South-East Africa. China has sup- ported government infrastructure such as a cable to Chad, along a gas pipeline, and to Uganda, along the Kenyan railway. Liquid Telecom, part of the privately-owned Econet Group, has its own optical fiber cables serving Kenya, Uganda, Rwanda, DRC, Zambia, Zimbabwe and Lesotho with links to undersea cables at Mombasa, Mtunzini and Melkbosstrand, plus plans to extend north to Egypt. https://www.econstor.eu/bitstream/10419/101381/1/795363087.pdf 20. https://www.capacitymedia.com/article/29otcq7s58uymwyj1xtds/news/orange-launches-djoliba-the-first-pan-west-african- fibre-network. <<< 15 PEACE (Pakistan & East Africa Connecting Europe) PEACE is a 15,000km long submarine cable system that provides open flexible and carrier neutral services for its customers. The system design will adopt the latest 200G technology and WSS technology, which provides the capability to transmit over 16 Tbit/s per fiber pair, servicing growing regional capacity needs. The network topology will substantially reduce network latency by adopting shortest direct route connectivity and enhancing route diversity between Asia, Africa, and Europe. The project partners include PCCW Global, Cybernet, Djibouti Telecom, Golis, Hormuud Telecom, Interxion, Telkom Kenya, HMN Tech, Orange and TelecomEgypt. It is expected to be completed in 2021. Source: http://www.peacecable.net/#about ; Africa 1 The Africa-1 submarine cable system spans 10,000km, initially landings in Kenya, Djibouti, Pakistan, United Arab Emirates, Kingdom of Saudi Arabia, Egypt and France. The Africa-1 cable system will also land in Sudan, cross Egypt through diverse new terrestrial routes on the way to France, and further connect other countries in the Mediterranean such as Algeria, Tunisia and Italy. Africa-1 consortium comprises Etisalat, G42, Mobily, Pakistan Telecommunication Company and Telecom Egypt. It is due to be ready for service in 2023 Source:https://www.submarinenetworks.com/en/systems/asia-europe-africa/africa-1; https://www.capacitymedia.com/ articles/3827601/asn-starts-construction-on-africa-1-subsea-cable EllaLink EllaLink has been in service since June 2021, providing the first-ever high-capacity direct fibers between Europe and Latin America with points of presence in Sines, Madrid, Lisbon, Marseille, Barcelona, Fortaleza, Sao Paulo & Rio de Janeiro and onward connectivity to the USA, Europe, Asia, Africa and the Middle East. Source: https://ella.link/ Other planned cables include Africa 1, Senegal Horn of Africa Regional Express (SHARE) Cable, Indian Ocean Xchange (IOX), South Asia Express (SAEx2) Africa has seen an increase in the number of cables and a prices.23 As more cables became operational, capacity prices resultant decrease of capacity prices, as much as 40 percent decreased (Figure 9), which, in conjunction with investment in some routes. Still, the majority (24 percent) of the countries in terrestrial networks, could translate to lower final prices.24 with subsea cables only have one subsea cable. The Africa However, several barriers remain in some countries, thus Coast to Europe (ACE) cable with a design capacity of 12.8 preventing benefits from international connectivity from Tbps is the only cable landing in two-thirds of the countries being passed on to domestic markets through lower capacity with one cable, i.e., Gambia, Guinea-Bissau, Guinea, Liberia, prices. Although Egypt acts as a transit hub for at least 14 Mauritania, Sierra Leone.21 22 Before the entry of private sector cables running through Africa, Asia, and Europe due to its led cables from 2009, incumbent operators who controlled the strategic positioning,25 prices for international connectivity access to capacity on SAT 3/WASC maintained high-capacity are relatively high. This could partly reflect the de facto 21. The World Bank West African Regional Communications Infrastructure Program supported the Governments of Benin, Gambia, Guinea, Liberia, São Tomé and Príncipe, and Sierra Leone. The program supported both investment in the ACE consortium and the development of an enabling policy environment for broadband infrastructure in the specific countries. ACE promoted open access in its ownership model, and therefore both Government and local operators had a stake in the Special Purpose Vehicle developed to manage the capacity and landing station in each country. https://www.itu.int/dms_pub/itu-d/opb/pref/D- PREF-BB.GDI_01-2017-PDF-E.pdf 22. ITU (2017) 23. ibid 24. Schumann and Kende (2013) 25. https://www.submarinenetworks.com/en/services/research/submarine-cables-crossing-egypt-and-cost 16 >>> INTRODUCTION concentration in the international segment, with MNOs using retail prices, poor quality, and limited coverage. Regulatory Telecom Egypt’s international gateway instead of their own barriers such as monopoly in international gateway or and the absence of a regulated standard offer for Telecom constrained competition seem to also contribute to the higher Egypt.26 Djibouti seems to have managed to extract value capacity prices in countries like Angola, Algeria, Tunisia, and from its strategic location and high level of international Cameroon despite the presence of several cables. Countries connectivity, presenting lower international connectivity such as Rwanda and Uganda benefit from efficient regional prices. Notwithstanding an abundant supply of international connections from Kenya and Tanzania and thus have lower bandwidth, a monopoly in the international gateway and capacity costs despite being landlocked. domestic broadband internet present challenges: high >>> Figure 9: Weighted median IP transit prices for 10 GigE and number of submarine cables 25 16 Weighted Median IP transit Prices for 10 14 14 Number of submarine cables 20 12 GigE, as of Q1 2021, USD 15 10 9 8 10 7 6 6 6 5 5 5 4 4 4 4 4 4 5 3 2 2 0 0 0 0 Sudan Uganda Ghana Cameroon Rwanda Nigeria South Africa Djibouti Egypt Tunisia Morocco Angola Algeria Mozambique Tanzania Cote d'Ivoire Kenya prices for 10 GigE Number of cables Note: Yellow highlight indicates a landlocked country Source: Illustration based on data from Telegeography (2021) There is an increasing trend of international carriers building content, which is expected to result in lower costs and better regional infrastructure. Operators in Africa not only purchase quality of services.29 This is the case of Microsoft Azure Cloud capacity on submarine cables but also purchase capacity in Services, Huawei Cloud Data Centres, and Amazon Web international hubs to enable transit to the global backbone, Services in South Africa,30, and planned submarine cables which is not reciprocated for operators in developed countries (Google and Facebook). requiring transit in Africa.27 Recently, the increased internet traffic has created an incentive for localization by international Cross-ownership by firms in multiple cables is common. In carriers and content delivery networks by developing Points 21 countries, operators have partial ownership of more than of Presence (PoPs), and exchange points28 in the region, one cable. In countries with more cables, there is often a thus reducing the need for international transit to access their high level of cross-ownership. In Egypt, Telkom Egypt has 26. https://enterprise.press/hardhats/egypt-make-better-use-position-subsea-internet-cable-hub/ 27. Operators in sub-Saharan Africa are thus forced to pay the cost of Internet connection as far as the global Internet provider’s point of presence (POP) and for use of the point of presence and transit. The international Internet connection model used today penalizes African operators, who pay the total costs of international Internet interconnection while users in developed countries pay no such compensation to operators in sub-Saharan Africa. https://www.itu.int/en/ITU-D/Regulatory- Market/Doc- uments/IIC_Africa_Final-en.pdf ITU 2017, MTN Consulting 28. A Point of Presence (POP) is Physical location of Internet Service Provider’s (ISP) equipment https://www.itu.int/en/ITU- D/Regional-Presence/AsiaPacific/Documents/ S04-Internet-introduction.pdf 29. Schumann and Kende (2013) 30. Reuters (2020) ‘Amazon launches data centre operations in South Africa’ available at https://www.reuters.com/article/us- safrica-amazon-com-cloud-idUSKCN2241X3 <<< 17 ownership stakes in 9 of the country’s 14 cables, and Saudi adding CLS on their shores. Countries like Burkina Faso and Telecom and Tata Communications in five each. In Djibouti, Malawi have found innovative ways to increase international Djibouti Telecom has ownership in 7 of the 9 cables, and Saudi bandwidth capacity by launching virtual landing points (dry Telecom and Telecom Egypt in four each. In Morocco, Maroc port).34 A small number of landing stations hinders increased Telecom has ownership stakes in the five cables landing in the international capacity and the resilience of network coverage country and operates all the associated cable landing stations. for operators.35 In Africa, 47 percent of coastal countries with Cross-ownership increases network capacity and the firms’ submarine cables have only one landing point. This bottleneck resilience in the event of any disruptions by preventing loss of is further exacerbated by state-owned entities that dominate coverage due to cable cuts, which have been known to take landing stations and lack of transparency in licensing of some time to repair, resulting in significant revenue losses for landing stations, resulting in a monopoly by incumbent state- companies.31 Further, partnerships are also common with firms owned operators.36 For example, in Cameroon, government- that have cable landing rights in various countries which are owned Camtel controls landing stations.37 This economic then responsible for maintaining the cable in that territory, with governance issue in one segment carries to the value chain other local firms purchasing Indefeasible Rights of Use (IRU).32 as high premiums (landing station fees) are charged, thus However, firms with cross-ownership in subsea cables that also increasing the cost of international connectivity. provide services downstream may lead to barriers to entry in those markets and coordination among service providers. State participation in markets and infrastructure necessary for international connectivity, such as submarine cables and Cable Landing Points and Stations (CLS) are also essential landing stations, raises some concerns. State-owned entities for improving international connectivity. Exclusive rights to in some countries that control access to international internet operate landing stations have historically been a preserve capacity tend to keep capacity prices high to generate revenue of incumbent operators, especially when the operators were and may not have the incentive to increase traffic by opening the only investors in the cables. Since then, open access up to new entrants.38 In 25 countries, firms that own a stake regulations have led to private investment in subsea cables,33 in submarine cables are wholly state-owned (19) or majority while countries such as Kenya, Nigeria, and Ghana have state-owned (6), and in another 14 countries, they are SLEs liberalized their landing markets, resulting in several cables present (see Table 1). >>> Table 1: Examples of cable landing point owners with state involvement Country Landing Points Investors State Ownership Senegal All Sonatel Minority Ivory Coast Abidjan Côte d’Ivoire Telecom Minority Ghana Accra Ghana Telecom (Vodafone) Minority South Africa Melkbosstrand, Mtunzini Telkom SA Majority Namibia Swakopmund Telecom Namibia Wholly Kenya Mombasa Telkom Kenya Minority Djibouti YAC A CLS; Haramous CLS Djibouti Telcom Wholly Cameroon All Camtel Wholly Egypt All Telecom Egypt Majority Source: Egypt DE4A diagnostic; Senegal DE4A diagnostic; https://www.submarinenetworks.com/en/systems/asia-europe- africa/dare1/kenya-to-invest-59m-to-the-dare1-ca- ble-system. 31. ITU 2017 32. Mukora. A, Njeru. R and Syekei. J (2021) ‘Kenya: Overview of Data Infrastructure In East Africa’ Bowmans Law. Available at: https://www.bowmanslaw.com/insights/ technology-media-and-telecommunications/data-infrastructure-in-east-africa/ 33. The ACE cable promoted open access in several countries such as Benin, Gambia, Liberia, and Sierra Leone, where government and local operators had rights to the special purpose vehicle (SPV) operating the landing station. 34. Burkina Faso RWI, https://www.ecofinagency.com/telecom/0809-41790-burkina-deploys-new-ixps-and-vlps : Malawi https://www.mwnation.com/govt-launches-virtu- al-landing-point/; both supported by the World Bank 35. Ewan Sutherland, Undersea cables and landing stations around Africa: Policy and regulatory issues, at https://www.econstor.eu/bitstream/10419/101381/1/795363087.pdf 36. Schumann and Kende (2013) 37. Schumann and Kende (2013) 38. Sunderlands, ITU 2017 18 >>> INTRODUCTION Several firms are also vertically integrated into downstream opportunities.39 For example, In Zambia, the opening of the markets. Of 75 owners of submarine cables, 73 are vertically international gateway market led to increased competition, integrated in at least one other market segment. Of these, lower data prices, and increased usage.40 While many at least 32 are active in over three market segments. For countries in Africa have liberalized their IGW, some countries example, Telecom Egypt, Telkom South Africa, and Tunisie have either partially liberalized or retained a monopoly in Telecom have ownership in more than 3 cables each, have access to international capacity in the hands of the incumbent more than 50 percent market share in fiber backbone, offer or dominant player. Countries that still maintain a monopoly IP access/transit services, own data centers, are peers in in IGW services include Algeria (Algerie Telcom41), Djibouti IXPs and are retail broadband providers. They also own (Djibouti Telecom), Equatorial Guinea, and Namibia. Other spectrum rights and towers (except Telkom South Africa) countries like Cameroon, Guinea-Bissau, Sierra Leone, and offer mobile telecommunication services (see annex for Togo, and Benin have partial competition in IGW services. more details). Further, some countries appear to have liberalized their IGW market but may not enjoy the benefits that ensue due to some Liberalization of international gateways (IGW) has promoted limiting factors that may lead to entry restrictions in the IGW the growth of international internet bandwidth in many market (such as high license fees).42 Countries that may African countries. The benefits of liberalization include a have operators with substantial importance in international significant decrease in prices for international services, gateways despite liberalization include Egypt (Telecom Egypt), increased consumer welfare, increased competition resulting South Africa (Telkom), and Senegal (Sonatel).43 in improved quality, and the potential increase in investment >>> Figure 10: Number of market players in the international connectivity segment Note: red= monopoly in international gateway, yellow=few players with some entry restrictions, green= open to free entry. Includes operational cables Source: Elaboration by authors based on RWI 2, ITU - Level of competition in international gateways, Telegeography Global Internet Geography. Latest information available as of December 2021. 39. https://www.gsma.com/mobilefordevelopment/wp-content/uploads/2016/09/GSMA-Gateway-Liberalisation.pdf 40. Zimba, Aaron & Chibuluma, Matthews & Chishimba, Mumbi & Mbale, Tozgani. (2020). Liberalisation of the International Gateway and Internet Development in Zambia: The Genesis, Opportunities, Challenges, and Future Directions.https://www.researchgate.net/publication/341600351_Liberalisation_of_the_International_Gateway_ and_Internet_D evelopment_in_Zambia_The_Genesis_Opportunities_Challenges_and_Future_Directions/link/5ec9772c458515626cc6b7ea/download 41. World Bank. 2021, Regulatory Watch Initiative (RWI) “Thorough legal, regulatory, and competitive analyses of issues related to licensing, OTTs, international gateways, spectrum management, and regulatory governance.” World Bank Group. 42. https://www.intgovforum.org/multilingual/system/files/filedepot/17/Record percent2024a.pdf 43. Decision n°2018-005 fixing for 2018 the list of powerful operators in the relevant telecommunications markets; World Bank. 2020. Senegal DE4A <<< 19 1.2.2. FIBER NETWORKS since 2016. Africa Eastern and Southern (AFE) has the largest FOR BACKBONE AND METRO number of service providers and the longest fiber network on CONNECTIVITY the continent, accounting for 45.5 percent of kilometers of fiber. On average, there tend to be 5 providers present in one As of 2020, 129 companies and organizations own 790,000 country (median), with Nigeria having 22, South Africa 21 and km of fiber networks in Africa.44 Backbone fiber on the Kenya 13 providers.45 Over the same period, the number of continent has grown markedly in the past years, adding 37 companies leaving the market has been close to zero. companies and 200,000km (31 percent increase) of fiber >>> Figure 11: Operational fiber (km) relative to country GDP and population size 120 20 Fiber (km) per th. population 18 Fiber (km) per GDP per 100 16 80 14 12 capita 60 10 8 40 6 20 4 2 0 0 Africa Eastern and Africa Western and North Africa Southern Cen tral Note: Excludes metro fiber Source: Africa Digital Economy Database, based on Africa Bandwidth Maps (2020) Despite an increase in operators and expansion of fiber, 2020, the HHI of fiber backbone markets decreased in the the fiber backbone segment remains highly concentrated. majority of countries, 46, but remain at very high levels. Only Concentration in terms of fiber owned (in km routes) remains Nigeria, Uganda, and Cote d’Ivoire have an HHI that reflects high, with an average HHI of 4563. North Africa has the a moderately concentrated market (HHI between 1500 and highest concentration levels in fiber markets, with an average 2500), while 13 countries are monopolies.47 Eleven of these HHI of 6882, while AFE and AFE have similar average 13 monopolies are at least partly state-owned. levels of concentration of around 4250. Between 2016 and 44. Based on information from Africa Bandwidth Maps (2020). 45. Owning both backbone and metro networks. 46. From the 49 countries for which we have information, only 11 countries (Algeria, Benin, Eswatini, Ethiopia, Ghana, Kenya, Mauritania, Mauritius, Nigeria, Rwanda, and Zimbabwe) have seen their HHI increase from 2016 to 2020. 47. Comoros, Djibouti, Egypt, Eritrea, Guinea, Guinea-Bissau, Libya, Madagascar, Niger, Central African Republic, Eswatini, Ethiopia and South Sudan. 20 >>> INTRODUCTION >>> Figure 12: Concentration in fiber broadband markets, by African sub-region, 2016 and 2020 10000 10000 10000 9435 9000 8000 7000 6899 6000 5000 4802 4000 4284 4232 3000 2000 1000 737 937 0 AFE AFW NA Min Average Ma x Source: Africa Bandwidth Maps http://www.africabandwidthmaps.com Almost 90 companies in Africa own fiber networks in most extensive metro fiber network among the regions per metropolitan areas (metro fiber), a total of 226,500km inhabitant, 6173km, around ten times as much as in AFW. continent-wide, with higher concentration levels than in While average concentration has decreased in AFE and AFW backbone. Metro fiber refers to networks of fiber optic cables since 2016, concentration levels remain high in all three sub- in metropolitan areas that connect the national backbone regions (see Figure 14). In particular, metro fiber markets in with last-mile facilities in cities. Africa’s metro fiber network North Africa are extremely concentrated, with an average HHI has expanded by 74 percent since 2016, comprising over of 9464 (N=35). Continent-wide, 13 of 35 metropolitan area 156,000km in AFE, 44,000km in North Africa (NA), and 27,000 markets reported only one player,48 and 13 reported two. in Africa West and Central (AFW). AFE also accounts for the >>> >>> Figure 13: Metro fiber networks in sub-regions relative Figure 14: Concentration (HHI) in metro fiber in to population sub-regions, 2020 Metro fibre (in km) per population (in million) (2019) 7000 6000 5000 4000 3000 2000 1000 0 Africa Eastern and Southern Africa Western and Central North Africa Source: Africa Bandwidth Maps http://www.africabandwidthmaps.com 48. The countries where only one player was reported by Africa Bandwidth Maps in 2020 were Algeria, Cameroon, Chad, Democratic Republic of Congo, Gabon, Cote d’Ivoire, Madagascar, Mauritania, Mozambique, Namibia, Niger, Sao Tomé e Principe, Senegal, Togo and Tunisia and Zimbabwe. <<< 21 SOEs and SLEs own the largest share of backbone and and AFE but still high, at 62 percent in AFW and 48 percent metro fiber networks across Africa. However, private telecom in AFE (see Figure 15.). In AFE, private telecom companies operators have increased their network size since 2016. State- are the largest owners of metro fiber km – owning 51 percent owned and state-linked telecom companies control around 73 of the network – whereas in AFW, minority state-owned firms percent of total kilometers in Africa – between approximately own the largest share of the network – 56 percent. Private 70 percent in AFW and 100 percent in North Africa (Figure 15). firms have slightly increased their share of new investments In North Africa, where 70 percent of fiber backbone is owned by in fiber networks compared to SLE and SOEs, with private SOEs (i.e., firms where the State owns more than 50 percent), investments accounting for more than 30 percent of current non-telecom SOEs, such as electricity and gas firms, are and planned construction and around 40 percent of proposed particularly relevant, accounting for almost 16 percent of the fiber networks. Private investments in fiber networks such backbone fiber. Metro fiber networks are also majority state- as Liquid Telecom and CSquared have shown that private owned or -linked, owning around 61 percent of metro fiber actors can viably invest in African connectivity infrastructure. networks length. In North Africa, 100 percent of metro fiber is However, with slow growth, networks will remain largely state- owned by state-linked firms, with over 75 percent being owned owned or -linked in the foreseeable future. by SOEs. The footprint of state-linked firms is lower in AFW >>> Figure 15: Ownership of backbone and metro fiber networks in African sub-regions, 2020 a. Backbone fiber networks b. Metro fiber networks 3.8% 6.1% 46.0% 46.9% 35.6% 59.2% 70.8% 56.0% 9.0% 4.3% 21.4% 20.0% 1.9% 18.3% 10.9% 50.8% 30.6% 37.6% 28.8% 22.6% 18.4% 0.0% Africa Eastern and Africa Western and North Africa AFE AFW NA Southern Central Private Minority state-owned Private Minority state-owned Majority state-owned Wholly state-owned Majority state-owned Wholly state-owned Source: Africa Digital Market Players Database, based on data by Africa Bandwidth Maps (2020) In addition to state ownership, most of the backbone and metro backbone and five in metro fiber. There are two main types fiber is owned by large operators that may be active in multiple of firms present in this group: large private- or minority state- markets and show high levels of vertical integration. These top owned telecom groups that are active on multiple markets ten firms49 and corporate groups in terms of network length and may have access to historical incumbent networks, such (km) own almost 70 percent of Africa’s backbone and metro as MTN (owns backbone in seven markets), Vodacom (three networks, respectively. They include both private and state- markets), and Telkom (two markets), as well as incumbent owned firms, with six SOEs among the top ten firms in fiber telecom SOE operators that own the vast majority of a single 49. The top ten fiber backbone owners in terms of km length of networks are Telkom Group, Algerie Telecom, Maroc Telecom, MTN Group, Movitel (Mozambique), Telecom Egypt, Vodacom Group, Globacom Group, Libyan Post Telecommunications & Information Technology Company and Unitel Angola. 22 >>> INTRODUCTION country’s network (such as Maroc Telecom, Algerie Telecom such as data centers, international connectivity, mobile retail, and Unitel in Angola). The largest fiber owners are often and towers. Ownership of a large backbone or metro network vertically integrated as well. For example, the top seven can make it easier for firms to compete in other markets as companies are present in at least 6 additional segments. These these networks are an essential input for most services. firms are often active in other key infrastructure segments, >>> Figure 16: Vertical integration in fiber a. Presence of firms with 40 percent+ (*) market share in b. Presence of firms with 40 percent+ (*) market share in mobile retail in other segments fiber in other segments 80 67 17 14 60 10 9 40 33 6 16 13 4 20 4 5 1 1 0 0 0 Two Three Four Five and Two Three Four Five and plus segments segments segments plus segments segments segments segments segments Number of active firms Number of active firms Number of firms with 40% + market share Number of firms with 40% + market share Source: Authors, based on Africa Digital Market Players Database. 1 . 2 . 3 . F I X E D B R O A D B A N D R E TA I L Fixed broadband retail markets are extremely concentrated and dominated by large incumbents or regional groups that The fixed broadband retail segment is much smaller than are often state-owned or -linked. Concentration in terms the mobile segment, with a smaller user base and less of subscribers, as measured by the HHI is 7411, which is active retailers. With African countries being ‘mobile first’,50 considered extremely highly concentrated – even compared fewer users access the internet through a fixed broadband to other infrastructure markets. Concentration is highest in connection than mobile. Nevertheless, fixed internet remains AFW and NA, with HHIs of 7700 and 7999, respectively, and vital in Africa, as it can provide fast and stable connectivity for AFE having an average HHI of 6900 (see Figure 17). In 17 out households and businesses and can be accessed via Wi-Fi of 37 countries, fixed retail markets are served almost entirely on smartphones as well. There were approximately 24 million by one provider.53 Among the relatively competitive markets fixed broadband subscriptions in 2021, up from 12.7 million in are Rwanda (HHI of 2332), Tunisia (2431), and Kenya (2766). 2016 and 4.7 million in 2010.51 Fixed subscriptions tend to be Further, markets are often dominated by incumbent ISPs. purchased on a household level, with around 14 percent of The top six broadband companies account for more than 80 households having access to the internet at home52. At least percent of subscribers on the continent. Four of them are 93 ISPs (including subsidiaries) currently provide retail fixed state-owned or -linked, including state-owned incumbents broadband services in Africa, 44 in AFE, 34 in AFW, and 15 in with more than 80 percent shareholding (Egypt Telecom, NA, with 2.5 per country on average. Algerie Telecom, and Tunisie Telecom), while the others include legacy incumbent operators Telkom South Africa and 50. In 2020, 74 percent of web traffic was generated via smartphones and roughly 24 percent via PC devices. See StatCounter (2021), accessed via: https://www.statista. com/statistics/685188/african-countries-online-traffic-channel-share/. 51. Telegeography (2021). 52. ITU (2020) Regional and global key ICT aggregates. Available at https://www.itu.int/en/ITU- D/Statistics/Documents/facts/ITU_regional_global_Key_ICT_indicator_ag- gregates_Nov_2020.xlsx. 53. They are Algeria, Botswana, Burkina Faso, Cabo Verde, Djibouti, Ethiopia, Guinea-Bissau, Liberia, Libya, Mauritania, Niger, Republic of the Congo, Mauritius, Sao Tome and Principe, Senegal, Seychelles and Zimbabwe <<< 23 private firms, most with minority state- holding. Overall, SOEs sharing 64 percent of the market in AFW. While the market comprise roughly 10 percent of the market in AFW, 67 percent share of SOEs has been on the decline in AFE since 2010, it in AFE, and 79 percent in North Africa, with SLEs alone has increased in North Africa. >>> >>> Figure 17: HHI in fixed broadband retail, across African Figure 18: Share of subscribers, by ownership countries HHI in fixed broadband 8.30% 2,431 17.39% 2.40% 7,482 27.85% 10,000 10,000 7,426 10,000 10,000 5,046 10,000 5,000 10,000 50.29% 64.09% 10,000 10,000 10,000 6,898 3,178 4,864 10,000 10,000 9,674 5,070 3,798 51.44% 10,000 2,766 4,662 10,000 2,332 6,038 10,000 4,151 HHI in fixed broadband 12.28% 2,000 4,000 3,906 6,000 8,000 25.20% 19.41% 10,000 10,000 10,000 20.04% No data 8,883 10,000 1.30% 4,050 6,591 Africa Eastern Africa Western North Africa an d Southern and Central Private Min ority state-own ed IBRD 47452 | AUGUST 2023 Majority state-own ed Wholly state-own ed Source: Telegeography (2021) Within the top 15 SOE-SLEs, all own data centers, and 12 has been uneven across countries. With the mobile economy companies own spectrum rights and provide mobile services. projected to grow by 4.3 percent per year between 2019-2025 In the top 10, three companies operate in more than one (GSMA 2020), the demand for telecom towers is expected to market. In particular, Telkom, the biggest owner of backbone grow further. Surging demand for mobile data requires adding fiber, is also the 6th biggest company in terms of subscribers. capacity and coverage to networks by building new tower sites Algerie Telecom and Maroc Telecom, the second and third and using available infrastructure more efficiently.54 There are biggest companies, are also the second and third biggest currently 106 firms that own at least 208,964 mobile towers owners of backbone fiber. across the continent, with an average of 3.6 firms per country.55 One tower site serves a relatively high number of subscribers in SSA, around 4,935 in AFE and 4,792 in AFW (see Figure 1.2.4. TELECOM TOWERS 19), compared to 1,745 in North Africa.56 The continued growth of mobile connectivity is increasing the Africa’s tower market has seen the entry of independent demand for telecom towers in Africa, but tower deployment tower companies, but MNOs still control around half of 54. The African tower market is expected to grow at 2.1 percent CAGR between 2017-2023, which is below expectations in other regions despite higher growth rates of mobile connectivity. 55. Tower count refers to tracked towers on TowerXChange. The countries included in the regional averages are Australia, Cambodia, China, Indonesia, Japan, Korea, Rep., Lao PDR, Malaysia, Mongolia, Myanmar, New Zealand, Philippines, Thailand, Vietnam (EAP), Austria, Bulgaria, Czech Republic, Denmark, France, Germany, Greece, Hungary, Ireland, Italy, Kazakhstan, Netherlands, Norway, Poland, Portugal, Russian Federation, Serbia, Spain, Sweden, Switzerland, Turkey, United Kingdom, Ukraine (ECA) and Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Ecuador, El Salvador, Guatemala, Mexico, Nicaragua, Panama, Paraguay, Peru (LAC). 56. Internationally, the number of subscribers per tower is as low as 2036 in LAC, 1326 in EAP, and 937 in ECA. Brickner (2020) Closing Africa’s infrastructure gap with sustainability at the heart of Helios Towers, see https://www.raconteur.net/sponsored/closing-africas-infrastructure-gap-with-sustainability-at-the-heart-of-helios-towers/ 24 >>> INTRODUCTION Africa’s towers. Traditionally, the majority of towers in Africa worldwide (TowerExchange, 2021). This is partly driven by a has been owned by MNOs, which owned 82 percent of SSA lack of towerco activity in North Africa (less than 1 percent towers and almost all the North African towers in 2014.57 owned by towercos), while more than half (51 percent) of the Since then, independent tower companies (towercos) have tower portfolios in SSA are owned by towercos. The presence entered African markets and built or acquired significant of towercos facilitates infrastructure sharing, which increases portions of tower sites that are then leased out to MNOs. The operational efficiency for MNOs, while also allowing for a faster independent towerco model allows MNOs to raise money for rollout of networks, avoiding congestion in densely populated network expansion and cut costs, while towercos can create areas, and decreasing parallel infrastructure. However, in economies of scale by operating large portfolios across some countries, towers are concentrated in a few or one main regions.58 In 2021, towercos owned 44 percent of towers towerco, raising concerns about competition and access to across Africa on average, compared to around 70 percent this infrastructure (for example, in Uganda). >>> Figure 19: Share of towers owned by towercos 100% 100% 100% 100% 100% 89% 79% 71% 70% 66% 60% 56% 52% 42% 32% 31% 30% 26% 22% 2% 0% 0% 0% 0% 0% 0% 0% 0% N a G C a al An o ag p . Za d a a na M n d Ke a a m co M or i Ta eria sia n d ' ia am ire So am r Bu Se n a t Al la ria e R er U bia Tu e G ia M aw yp N ca s w nd ric ut ibi qu o ny ha Zi roo R rk e g ôt zan e go op ig Fa ge oz oc an ha ab C Ivo ni on ab M o, R as Eg D al m Af ig ga bi N C hi e w n b Et m h in g e ad C C Note: Shows the percentage of towers in each country that is owned by a company also providing mobile retail services based on the Africa Digital Economy Database Source: Africa Digital Economy Database based on data from TowerXChange Tower markets are relatively concentrated, with the largest addition, three large multinational firms – American Tower number of sites owned by MNOs, governments, and three Corporation, Helios Towers, and IHS Towers – are the main multinational towercos. The concentration of ownership in investors in the independent tower market in SSA, owning 86 tower sites, as measured by HHI, is relatively high across percent of its portfolio (see Figure 22). In recent years, smaller African markets, with an average HHI score of around 4300 towercos have also established themselves in local tower (Figure 20). In SSA, average concentration levels in tower markets, with a small footprint (around 3-4 percent) overall. markets are broadly similar yet slightly lower than in mobile Through ownership shares in MNOs, African governments retail, while the reverse holds true in North Africa. Across directly or indirectly own around half of Africa’s tower portfolio, regions, MNOs constitute the largest group of owners of 102,264 sites (Figure 22).59 towers, both owned by private actors and state-owned. In 57. GSMA (2014). Tower Power Africa: Energy Challenges and Opportunities for the Mobile Industry in Africa 58. See https://www.nortonrosefulbright.com/en/knowledge/publications/48f82d8d/sharing-of-telecoms-infrastructure--- opportunities-in-2016#section2 and https://www. ft.com/content/fd9512ce-0753-11e4-81c6-00144feab7de. 59. Including firms that are minority-, majority- and wholly owned by Governments. <<< 25 >>> >>> Figure 20: Heatmap of concentration (HHI) in terms of tower Figure 21: HHI in towers vs HHI in mobile count Heatmap of HHI 4928 4638 4146 3474 3482 3413 HHI 1,500 4,000 6,000 8,000 10,000 No data Africa Eastern Africa Western North Africa and Southern and Central HHI tower HHI mo bile Note : | Tower site count by operator is incomplete for Chad and Congo, Dem. Rep. IBRD 47453 Note: Average HHI only includes those countries for which HHI in the tower market and thus excluded from this map. AUGUST 2023 is available. >>> Figure 22: Tower holding by organization, by region (percent of regional tower count) 100% 0.2% 0.9% 1.3% 1.0% 0.3% 2.2% 3.3% 7.0% 0.5% 90% 0.0% 11.5% 7.3% Other 28.5% 4.1% 80% 10.6% Small towercos 9.5% 70% 39.0% 18.6% Atlas Tower 17.3% 60% IHS Towers 50% 4.8% Helios Towers 40% 20.5% 71.5% 30% American Tower Corporation 57.8% 0.7% 54.5% 20% MNO (SOE ) 27.0% 10% MNO (fully/majo rity priva te) 0% Africa Eastern and Africa Western and North Africa Africa Southern Central Note: Shows the share of towers in regional and continent tower portfolios by type of firm / specific tower company based on the number of towers. Small towercos include Africa Mobile Networks, Anglobal, Axian Group, HOI Middle East and Africa Group, Hotspot Group, Merlin Project Services, Pula Towers, African Towers, Pan African Towers, Coast to Coast, Comco, Pro High Site Communications, BCTek Engineering, Communication Towers and International Tower Corporation. 22 MNOs are included in the MNO (private / majority private) group, and thirteen in the SOE MNO group. Source: Africa Digital Economy Database based on data from TowerXChange 26 >>> INTRODUCTION Regulation and licensing regimes play a growing role in In major tower transactions, independent towercos have safeguarding competition and facilitating entry into the acquired at least 25 percent of tower sites in SSA in ownership tower market, particularly regarding infrastructure sharing. or management contracts, highlighting the need to assess Sharing of passive or active tower infrastructure reduces the sharing to prevent market imbalances.64 Since 2010, 39 major requirements for network rollout by MNOs and thus lowers tower transactions have been finalized in 16 countries in barriers to entry for new players. While MNOs can engage in SSA, notably in Nigeria (9 times), Ghana (6), South Africa (4), bilateral infrastructure sharing, models involving independent and Kenya (4). Deals representing 45 percent of tower sites third parties have tended to achieve higher tenancy ratios and acquired were conducted in ownership contracts (Sale-Lease- therefore, more efficiency. The entry of towercos in the tower Back, SLB), while 31 percent of tower sites were transferred in market can be facilitated by appropriate licensing regimes. joint ventures, 14 percent in company acquisitions, and smaller While many HICs only require registration or notification with shares in management contracts (Managed with License to the regulatory authority, licensing regimes in LMIC are often Lease or Managed Services) and portfolio acquisitions (see stricter.60 For example, licensing might require a regulatory Figure 23). Despite interest by towercos, there have not been review of the business plan,61 collecting information on any major deals in North Africa, with an SLB deal between master-lease agreements and technical KPIs,62 or even cap Orange Egypt and Eaton Towers being cancelled in 2016 the total number of licensees altogether.63 However, even due to a devaluation of the Egyptian Pound by the Central where towercos are active, exclusive contracts in Sale-Lease- Bank and lacking regulatory approvals.65 There has been Back (SLB) deals or joint venture models between MNOs and consolidation in the towerco market to a limited extent, with towercos can restrict the ability of other MNOs to enter and American Tower Corporation’s acquisition of Eaton Towers compete. Regulatory authorities can promote competition (operating in Ghana, Uganda, Kenya, Burkina Faso, and through accessibility regulation, such as preventing Niger) and its portfolio of 5,700 towers at the start of 2020.66 anticompetitive exclusive contracts, ensuring access to right of way, and sharing mandates. >>> >>> Figure 23: Acquisition of tower portfolios by towercos – Figure 24: Number of tower acquisition deals by country, number of deals and towers 2010-2020 Number of deals 25 25000 21997 20 20000 15 14970 15000 10 22 10000 6911 5 5000 7 4150 Number of deals 4 4 1 0 2 450 0 2 4 SLB Joint Company MLL Portfolio 9 venture acquisition Acquisition No data Number of deals Number of towers Note: Summarizes the number and tower count of major deals since 2010. SLB refers to Sale-Lease-Back, IBRD 47454 | which refers to deals where the towerco buys the site from the AUGUST 2023 MNO and leases it back to them; MLL refers to Managed-with-License-to-Lease deals, which gives towercos the right to lease out space on a tower owned by a third party and managing shared resource. (Source IHS Towers). Source: TowerXChange (2021) 60. See Sousa and Heinrich (2019) Infrastructure regulation: overview and impact on TowerCos 61. In Ghana, Namibia and Egypt. 62. In Ghana, Nigeria, Egypt, and Namibia. 63. In Nigeria. 64. Ownership contracts are typically Buy-Lease-Back (BLB) contracts, and management contracts are contracts 65. The devaluation of the Egyptian Pound by the Egyptian Central Bank occurred after the agreement was reached and meant a revision to the commercial terms of the deal. See https://www.towerxchange.com/cancellation-of-menas-first-major-tower- transaction/. 66. See https://www.businesswire.com/news/home/20200102005240/en/American-Tower-Closes-Eaton-Towers-Transaction- and-Announces-Agreement-to-Acquire-Joint- Venture-Stakes-in-Ghana-and-Uganda-From-MTN. <<< 27 1.2.5. MOBILE VOICE AND network operators (MVNO)67. The number of MNOs varies, B R O A D B A N D R E TA I L with an average of 3.4 operators per country, which is in line with other regions.68 On one side, there are seventeen Around 195 firms across 57 African countries offer retail markets that are either a monopoly (Djibouti, Ethiopia, and mobile services, with 17 countries still having monopolies or Eritrea) or a duopoly (14 markets in total), while on the other duopolies. The market for mobile services comprises wireless side, six countries are home to five or more operators - Nigeria communication services (data and voice), which are offered (eight operators), Tanzania (eight), and South Africa, Ghana, by mobile network operators (MNO) and by mobile virtual Guinea, and Sierra Leone (five each).69 (Figure 25) >>> Number of operators Figure 25: Number of active mobile operators (excluding MVNO), by country 3 3 3 2 4 3 2 3 4 3 3 1 4 4 3 3 1 5 2 8 5 5 2 1 3 3 2 4 3 8 2 3 7 4 2 4 4 5 5 2 8 Number of operators 1 2 2 2 4 2 4 6 3 9 3 4 3 No data 3 2 2 5 2 Source: Africa Digital Economy Database (2021) based on Telegeography (2020) IBRD 47455 | AUGUST 2023 67. According to the International Telecommunications Unit (ITU), “MVNOs are wireless services providers that do not own the wireless network infrastructure but instead buy network capacity from existing MNOs to offer services to their users.” See https://news.itu.int/mvnos-telecom-world-value/. 68. The global average number of mobile operators (excluding MVNOs) is 3.36 as of Sep 2020. See Telegeography 2020. 69. Note that some operators in Nigeria and Tanzania are active only in selected regions, for instance, in the largest cities in Nigeria or in specific regions, such as the island of Zanzibar in Tanzania. 28 >>> INTRODUCTION MVNOs play a minor role in African mobile services markets, percent targeting the budget segment, 13.5 percent offering with only 13 countries hosting MVNOs and low market shares, value-added services for niche groups such as minorities,72 although at least 50 countries have general provisions that and smaller fractions targeted at young people (7.6 percent), allow for MVNOs. MVNOs, which offer mobile services using and those wishing to make international calls (11.5 percent).73 capacity bought from MNOs, can add competitive pressure However, Africa’s MVNOs seldom reach a market share of to the market for mobile services in specific niche segments. over 2 percent in terms of total subscribers in a country - Theoretically, 34 African jurisdictions allow for infrastructure even in a country with significant MVNO presence like South sharing by mobile operators and, thus, the presence of Africa – except for the Senegalese and Kenyan markets (15.4 MVNOs.70 However, a total of 50 MVNOs are active (including percent and 9.7 percent market shares, respectively). This is planned) across 13 markets, with South Africa being home low compared to international standards,74 suggesting that to the largest number of MVNOs, 28 firms (see Figure 26).71 MVNOs in Africa exert limited competitive pressure on MNOs. MVNOs are often active in specific market niches, with 35 >>> Figure 26: Number of MVNOs and market share, by country 30 20% 25 15% Number of MVNOs 20 Market share 15 10% 10 5% 5 0 0% a l n ya ca tte o sia ia a on e os ga bw nd oo c ny an ri oc b o or ni ni ne Af er ga Ke ay Li ba nz éu Tu om or Se am M U m h Ta M R C ut Zi C So Num ber of MVNOs Market share of MVNOs Note: Includes planned MVNOs. This is the case in Cameroon, Kenya, Mayotte, Senegal, South Africa, Tanzania and Zimbabwe. Source: Telegeography 2020 70. ITU Regulatory Tracker 71. Africa’s MVNOs are hosted by a total of 20 firms: Cell C (SA) 21, Africom, Airtel Kenya (2), Airtel Uganda (1), CamTel (1), Cell C (21), Expresso Telecom (Senegal, 1), Free Senegal (1), Libyana (1), Millicom Tanzania (1), MTN SA (3), Orange Mayotte (1), Orange Reunion (1), Orange Senegal (2), Safaricom (1), SFR Mayotte (1), SFR Reunion (2), Tangerine (Noman Communications, 1), Tunisie Telecom (1), Vodacom Group/SA (3), Wana (1). 72. Such as religious communities, occupations, hobbies, or other niche groups. See Telegeography 2020. 73. See Telegeography (2020). 74. For comparison, MVNOs have average market shares of around 20 percent in Western European and North American markets but also play a bigger role in EAP (7.4 percent market share), Middle East (8 percent), and LAC (4.5 percent). The average market share in Africa – driven by outliers Kenya and Senegal – is 3.5 percent. See Telegeography 2020. <<< 29 Mobile markets in Africa are relatively concentrated, 28).76 The HHI scores range from 2600 in Tanzania, 2846 in consolidating a trend observed in recent years. Around 190 DRC, and 2989 in Nigeria to as high as 7888 in Eswatini, million Africans live in countries where mobile retail markets 8362 in Namibia, and 8961 in Sao Tome and Principe, in are monopolies or duopolies, around ten percent of Africa’s addition to the three monopolies (HHI score of 10,000). East population. Market shares show that one MNO captures over and Southern African countries appear to have the highest 50 percent of the market in almost two-thirds of countries. In 9 concentration, 4476 on average, followed by Western and countries,75 one provider captures over 65 percent (see Figure Central Africa (4288) and North Africa (3941). While the 27). Concentration in terms of subscribers has increased in average HHI decreased in East, Southern, and North Africa recent years. Similarly, HHI scores of all markets for mobile over the past five years, it increased slightly in Western and services in Africa are above 2,500, which is commonly used as Central Africa, indicating a move towards more concentrated a threshold to identify highly concentrated markets (see Figure markets in the subregion (Figure 29). >>> >>> Figure 27: Market concentration in mobile markets as Figure 28: Change in HHI 2015-2020 (blue = higher indicated by HHI scores (mobile subscribers), by country HHI by country HHI, green = lower HHI) Change 2015-2020 HHI HHI -4,700 2,500 -3,500 4,000 -2,500 6,000 -1,500 8,000 0 10,000 1,500 No data 2,500 No data Note: Refers to the HHI in the market of mobile services. The Herfindahl-Hirschman Index is calculated IBRD 47457 | by adding the squared market shares of each operator in the market ranges andIBRD 47456 | AUGUST 2023 between 0 (not concentrated) to 10,000 (highly concentrated). Darker = higher HHI AUGUST = higher 2023 concentration. Subscriber data is aggregate for mobile voice and broadband, as separate information on mobile internet subscribers per operator was not available. Source: Authors’ calculation based on Telegeography 2020 African regulators have assigned relatively little spectrum dividend bands. However, spectrum markets in SSA appear overall, and concentration in spectrum, especially in digital less concentrated than they do in terms of mobile subscribers dividend bands, appears high – albeit lower than in mobile retail (Figure 30). The reverse holds true in North Africa, which has services. African mobile operators hold less spectrum than relatively high HHI scores across all countries in the region. A operators in other regions, from twice (LAC, Middle East) to three closer look at HHI in each country reveals that countries with times (ECA and EAP) and six times (North America). In terms higher concentration levels in spectrum holdings drive their of concentration of spectrum holdings, African markets tend to respective regional averages, such as Lesotho, Malawi, be more concentrated than other regions in the critical digital Seychelles, Chad, Benin, Gabon, and Egypt. 75. Sao Tome and Principe, Eswatini, Djibouti, Eritrea, Ethiopia, Namibia, Angola, Equatorial Guinea, Kenya 76. See Department of Justice (2010) Horizontal Merger Guidelines, available at https://www.justice.gov/atr/horizontal-merger- guidelines-08192010. 30 >>> INTRODUCTION >>> >>> Figure 29: HHI by region (700- 800 MHz bands) Figure 30: Average HHI in terms of mobile subscribers and spectrum holding, by region 6,000 5,000 4,000 3,000 2,000 1,000 - a A FW C A AP FE ca AR ic N C LA fri er E A E E S A A M m th A er or th th N or O N HHI Avg HHI HIC Avg HHI LMIC Note: The bands covered by this chart are 700-800 MHz Note: Excludes Senegal and Sao Tome e Principe due to missing data of most Source: GSMA 2021 operators, which leads to an HHI of 10,000. Source: Authors’ calculation based on Telegeography 2020 African governments hold significant stakes in mobile operators percent market share on average, while SOEs in AFW have across Africa. Overall, 91 MNOs are wholly or partially owned lower average market shares (see Figure 31). Some foreign by the Government (29.1 percent of mobile providers),77 governments also have a large footprint on the continent representing 32.8 percent of total mobile subscribers across through ownership of prominent mobile providers. Overall, the continent. Of these, 21 (11.1 percent) are wholly owned these foreign governments79 hold ownership stakes in 41 by the government, 11 (5.8 percent) majority state-owned and MNOs across Africa, which together account for around 30 23 (12.2 percent) minority state-owned. In many cases, the percent of the mobile subscribers on the continent. Notably, state-owned firm is either the leading or one of the largest the Government of South Africa owns a share in the MTN operators in the countries, with 22 state-owned MNOs having Group, which is active in 17 African countries outside of South a market share of 40 percent or more and eleven having a Africa, while the Government of France and the Government market share of 55 percent or more.78 In particular, wholly- of the United Arab Emirates own significant stakes in the owned SOEs in AFE and NA have close to or more than 50 Orange Group and Etisalat, respectively. 77. We define State-owned enterprises (SOE) as companies where the State owns more than 50 percent of the company and State-linked enterprises (SLE) as companies with State holding of less than 50 percent. This excludes Mobile Virtual Network Operators. 78. These are: Ethio Telecom, Djibouti Telecom, EriTel, CST (Sao Tome e Principe), Mobile Telecommunications (MTC) (Namibia), MTN Eswatini, Unitel (Angola), GETESA (Equatorial Guinea), Safaricom Kenya, Comores Telecom and Orange Senegal 79. Foreign governments are governments that are not the home government. <<< 31 >>> Figure 31: Market shares of SOEs and SLEs in mobile retail, by region (min – average – max) 100% 100% 100% 95% 90% 86% 80% 82% 70% 60% 60% 55% 50% 46% 40% 43% 41% 41% 37% 40% 34% 33% 30% 30% 24% 20% 19% 20% 14% 10% 10% 6% 6% 7% 5% 0% 1% 0% Minority state- Majority state- Wholly state- Minority state- Majority state- Wholly state- Minority state- Majority state- Wholly state- owned owned owned owned owned owned owned owned owned AFE AFW NA Note: Shows the minimum, maximum, and average market share of mobile operators that are minority- (SLE), majority- or wholly state-owned (SOE) across African subregions. Source: Author’s analysis based on the Africa Digital Economy Mapping (2021). Large international business groups have a significant footprint five groups are active in 42 countries80 on the continent (Figure in mobile service markets across Africa. In 2020, 48 business 32), which account for around 70 percent of Africa’s economy.81 groups were active in the provision of mobile services across The presence has remained largely the same since 201582, 57 countries. The largest pan-regional groups by country although Airtel has left Ghana, selling all AirtelTigo shares to presence are Orange Group (20 subsidiaries across Africa), the Ghanaian state.83 Across the continent, the five largest MTN (17 subsidiaries), Airtel (14 subsidiaries), Etisalat (13 groups have over 707 million mobile subscribers combined, subsidiaries) and Vodafone (8 subsidiaries). Together, these translating into 66.5 percent of all mobile subscribers in Africa. >>> Figure 32: Countries covered by at least one of the five largest groups a. 2020 Countries where at least one of the five largest groups is active b. 2015 No data Note: Highlights countries in which a subsidiary of the following groups is active: Orange Group, MTN, Airtel, Etisalat, Vodafone. IBRD 47458 | AUGUST 2023 Source: Africa Digital Economy Database (2022), WBG (2016) Breaking Down Barriers 80. Benin, Botswana, Burkina Faso, Cameroon, Central African Republic, Chad, Congo, Dem. Rep., Congo, Rep., Côte d’Ivoire, Egypt, Arab Republic of, Eswatini, Gabon, Ghana, Guinea, Guinea-Bissau, Kenya, Lesotho, Liberia, Madagascar, Malawi, Mali, Mauritania, Mauritius, Mayotte, Morocco, Mozambique, Niger, Nigeria, Réunion, Rwanda, Senegal, Seychelles, Sierra Leone, South Africa, South Sudan, Sudan, Tanzania, Togo, Tunisia, Uganda, Western Sahara, Zambia 81. In terms of 2019 GDP in current USD. Source: World Development Indicators, WBG. 82. See WBG (2016) Breaking Down Barriers 83. https://www.mobileworldlive.com/featured-content/top-three/ghana-completes-acquisition-of-airteltigo#:~:text=Bharti percent20Airtel percent20confirmed percent20full percent20ownership,aim percent20to percent20leave percent20the percent20market. 32 >>> INTRODUCTION Corporate groups have a strong presence in certain regions, operators across eleven countries85 being wholly or majority and state ownership of mobile operators is high in AFE and state-owned and a combined share of around 17 percent of North Africa. Vodafone has the largest presence of one all mobile subscribers in the region. In North Africa, SOEs operator group in AFE84 (around 24 percent of all subscribers account for 28 percent of subscribers in the region. In addition, region-wide), followed by MTN and Bharti Airtel. MTN is the Moroccan government holds a 22 percent stake in Maroc leading the number of overall subscribers in AFW (31 percent), Telecom (and Etisalat group of the United Arab Emirates holds with Bharti Airtel and Orange coming second and third. In 53 percent). In contrast to the other African subregions, SOEs North Africa, Orange, Etisalat and Vodafone are similarly large do not account for a significant share of the market in mobile with 19-20 percent of region-wide subscribers each. State- services in AFW, with the exception of foreign government owned operators play a significant role in AFE, with thirteen holdings in Etisalat. >>> Figure 33: Region-wide subscriber shares of main telecom groups a. East and Southern Africa b. Western and Central Africa c. North Africa Note: Author’s elaboration based on Telegeography 2020 Recent entry, exit, and merger figures indicate dynamism in (32 percent), firm dissolution (24 percent), or inactivity (8 the mobile services market segment. In the last ten years, 57 percent).86 In total, there were 33 M&A deals associated with firms entered the African mobile services markets, while 11 mobile operators since 2010, involving 28 companies, four87 exited. Of the new entrants, 31 were MNOs and 26 MVNOs. of which engaged in more than one deal. Figure 35 gives an Nine MNOs entered concentrated markets with one (five overview of the markets in which these deals happened. After entries) or two (four entries) existing operators (Figure 34), these exits, over half of the markets were left with only three or suggesting a potentially significant impact on competition two operators, while 40 percent had four or more players left. between operators. A fourth operator entered the market in eleven cases, while also, in eleven cases, new operators entered markets that were already rather crowded with four or more operators. Since 2010, 38 mobile operators have discontinued their operations due to mergers and acquisitions (37 percent), licenses being revoked by the regulator 84. Total of 462.27m subscribers in Africa East and Southern. See Telegeography 2020. 85. Angola, Botswana, Burundi, Comoros, Eritrea, Ethiopia, Mozambique, Namibia, Tanzania, Zambia, Zimbabwe <<< 33 >>> >>> Figure 34: Market structure upon entry in the mobile services Figure 35: Countries involved in recent M&A (2009-20) market (MNOs) 2008-20 Source: Authors’ elaboration based on data from Telegeography Mobile operators have expanded into other markets, such as almost half of Africa’s population. Having started in East Africa mobile money, cloud, and digital platforms. Half of African MNOs in 2000, Africa remains at the center of the mobile money also offer Mobile Money (MoMo) services (98 out of 196, 50 movement. Today, the continent accounts for 46.8 percent percent). Around 13 of these are dominant (40 percent+) in both of registered mobile money accounts and 64.5 percent of mobile retail and mobile money services. Further, out of the top transaction value in USD in 2020.90 With a total of 562 million ten operators in terms of subscribers in each region (30 total), accounts, around 41 percent of Africa’s population91 has 26 provide mobile money services, and 25 provide cloud and access to a mobile money account, with around 161 million hosting services. Further, MNOs increasingly move into digital (12 percent of the population) active users (see Figure 36). In services on digital platforms, such as opening an e-commerce recent years, accounts grew by 12 percent and active users by store, offering digital platforms for farmers, and, increasingly, 18 percent across Africa on average, with Southern and West digital financial services (DFS) provided by 19 of the regional Africa having the highest growth rates and East Africa adding leaders.88 Other ICT services provided by MNOs include video the most subscribers and users in absolute terms (Figure conferencing services, Digital TV, and music streaming 37). On a country level, Kenya has the highest penetration of mobile money users (73 percent in 2017), followed by other East African countries such as Uganda (51 percent), 1.2.6. MOBILE MONEY Zimbabwe (49 percent) and Tanzania (39 percent), as well as Namibia and Gabon (43 and 44 percent, respectively). By Since its beginnings in 2000, mobile money89 has grown contrast, in North African and some West African countries, consistently and has been key in delivering mobile financial less than ten percent of the population had a mobile money services to low-income customers across Africa, reaching account in 2017. 88. See more information in the digital services section. 89. According to the Poverty Action Lab, “mobile money is a digital payment platform that allows for the transfer of money between cellphone devices. The technology is installed in the SIM card of the device and can be used on regular and smartphone devices. Users can receive, withdraw, and send money without being connected to the formal banking system. The product differs from mobile banking, where users connect using internet-enabled mobile devices to manage the funds in their bank account.” See https://www.povertyactionlab.org/blog/10-22-20/rise-mobile-money-sub-saharan-africa-has-digital-technology- lived-its-promises. 90. Total transactions hit 27.5 billion (up 15 percent), valued at $495 billion (up 23 percent). See GSMA (2021) State of the Mobile Money Industry Report 91. Using https://www.worldometers.info/world-population/africa-population/ for population total. 34 >>> INTRODUCTION >>> >>> Figure 36: Millions of mobile money accounts Figure 37: Growth of mobile money accounts across regions Registered accounts Active accounts 293 198 94 46 47 16 11 3 14 1 East Central Southern West North Africa Africa Africa Africa Africa Note: Regions are defined by GSMA and may not correspond to WBG regions. Source: GSMA (2021) State of the Mobile Money Industry Report Mobile money markets show high levels of concentration and mobile money markets with innovative business models and are characterized by a large MNO presence. 164 active mobile now account for almost 20 percent of mobile money providers money services are active across 45 countries in Africa. Of (Figure 41). For example, the US/Senegalese start-up WAVE, these, 16 countries have only one92 or two providers offering funded by leading VC funds on the continent, entered the mobile money services, including seven in AFE (of 21), Senegalese market with an app-based product that could nine in AFW (of 20), and one (of four) in NA. Where mobile radically reduce the cost of using mobile money compared to money subscriber data is available, market shares based traditional technology. According to statements by WAVE, it on subscribers indicate high levels of concentration, 5334 in has already become the leading mobile money provider with AFE on average and 4789 in AFW. Further, concentration in over 4 million users in Senegal.93 Similarly, open APIs can terms of HHI is higher in mobile money than in mobile retail enable digital financial services providers to grow markets in most countries. Traditionally, MNOs and banks were the and create more useful services to customers, expanding the first to enter mobile money markets as they had the relevant market presence of the mobile money provider.94 For example, infrastructure to make mobile money transactions, including MTN Mobile Money launched open API services in 2018, access to USSD and payment infrastructure on the backend, allowing third parties to access and leverage MTN’s assets and and given regulation. In over 60 percent of countries, mobile capabilities and create new, relevant services for customers. money services are provided by MNOs that have expanded By May 2021, there were over 15,600 registered users and into mobile money markets, with only a limited number of non- more than 900 partners accessing and experimenting with the telecom firms present in these markets (Figure 41). Thirteen API, and transaction volumes overall had increased, given the percent of mobile money operators are bank-led. However, emergence of new use cases.95 start-ups, often venture capital funded, have been able to enter 93. See https://techcrunch.com/2021/09/06/sequoia-heritage-stripe-and-others-invest-200m-in-african-fintech-wave-at-1-7b- valuation/. 94. This series of blog posts provides more information on the benefits, risks and practical implementation of open APIs for Digital Financial Service providers. See https:// www.cgap.org/blog/series/open-apis-unlocking-innovation-digital-finance. 95. See https://www.cgap.org/blog/mtn-mobile-money-opened-apis-was-it-worth-it. <<< 35 >>> >>> Figure 38: Market structure in mobile money Figure 39: Concentration of mobile money and mobile services, in terms of subscribers 3 2 1 HHI mobile money HHI mobile retail 10000 5 7 9000 8000 1 7000 7 6000 6 5000 4000 2 3000 7 5 2000 1000 0 Africa Eastern and Africa Western and North Africa a G on Southern Central B n Za C ) ts n a G au n e Be a Li i K a C Rw n ia m ya (D a Es b we er a in by a - ni ne an am bi go nd at Zi e n R o s Bo h a a C m is ui on a w ba nz w Ta 5 or more 3-4 Players ui Duopoly Monopoly G >>> Figure 40: Number of players in mobile money markets across Africa 12 Benin 11 Chad Rep. of the Burkina Faso Congo Gabon Ethiopia Guinea Gambia 7 Madagascar Guinea- Malawi Botswana 5 Bissau Mali Burundi 5 CAR Lesotho Mozambique DRC Liberia Ghana Eswatini Niger Egypt Senegal Mauritania Namibia Rwanda Morocco 2 Sotuh Africa 2 Mauritius South Sudan Sierra Leone Somalia Cameroon Tanzania 1 Seychelles Sudan Zimbabwe Tunisia Cote d'Ivoire Togo Kenya Uganda Zambia Nigeria One firm Two firms Three firms Four firms Five firms Six firms Seven firms 19 firms Note: HHI figures for countries where complete subscriber data is available. Data is from 2019 and 2020 (for Zambia, data is from 2018). Source: Africa Digital Economy Database, based on GSMA (2020), BuddeCom and online research, including company reports and newspaper articles. Vertical integration between mobile retail and mobile money (Orange), and Ghana (MTN). Further, similar to mobile retail, might negatively impact competitive dynamics in mobile money the five largest telecom groups in mobile retail are also active markets. Often, these are dominant operators in mobile retail. in multiple mobile money markets on the continent. Combined, Of 67 mobile providers with a market share of 40 percent or MTN Group had 46.6 million subscribers in 202197 and M-PESA/ higher in mobile services, 46 offer mobile money services, and Safaricom 48 million98 across Africa, while Orange was the of these, at least96 12 also have a market share higher than 40 dominant mobile money provider in at least five markets and percent in mobile money services. In 7 countries, the largest MTN in at least four. About 23 percent of mobile money providers mobile operator (market share higher than 50 percent) also has are minority state-owned, but the presence of majority or wholly a market share above 50 percent in mobile money, for example, state-owned firms is low (9 percent combined) in mobile money in Kenya (Safaricom), Zimbabwe (Econet Wireless), Guinea and often tied to incumbent MNOs. 96. Given limited data availability, we cannot make a statement about market shares in the full list of countries with mobile money services. 97. MTN Annual Report 98. Vodafone annual report 36 >>> INTRODUCTION >>> >>> Figure 41: Composition of firm types offering mobile money Figure 42: Ownership type of mobile money providers services in Africa Other Majority 4.9% Wholly state- NA, 3.0% state- owned, 4.5% owned, 4.5% Bank-led 13.4% Minority state- owned, Fintech-led 22.7% 19.5% MNO-led Private, 62.2% 65.2% Note: Data is as of 2020 for the number of operators. Values were calculated for a sample of 162 (Figure 42) and 132 (Figure 43) mobile money providers operating in 45 countries. ‘Other’ firm types include a social media firm (WeChat), postal services, a grocery store conglomerate, and microfinance institutions. Source: Africa Digital Economy Database, based on GSMA (2020) and research by the team. 1.2.7. DATA CENTERS AND DATA SERVICES needed, as the African Data Centers Association (ADCA) estimates that the continent will need up to 1,000MW and 700 Data centers are key facilitators of the digital economy and are data center facilities in order to meet demand.105 growing across African countries, given increased internet use by firms and individuals. Data centers process and store data, Despite increased investment and growth, Africa’s supply of host websites and cloud-based applications, and often serve data centers is lagging other regions. As of 2021, there are at as interconnection points for data traffic.99 They are physical least 185 data centers in Africa. However, Africa accounts for facilities comprising network and storage infrastructure and only around 2 percent of data centers and less than 1 percent computing resources.100 Colocation data centers, in particular, of data center space worldwide.106 Together, the data centers on are important enablers of digital businesses and the digital the continent provide as much rack space to its 1.3 billion people economy, as they allow firms of all sizes to rent space and access as those in metropolitan areas in Denver and Kuala Lumpur, networking resources in their facilities. Increasing internet users with around 3 million and 8 million inhabitants, respectively. on the continent, the shift to more data-intensive technologies This means that one data center in Africa is serving a larger like big data, AI, and IoT, and the impact of the Covid-19 number of people,107 with one data center for about every 8400 pandemic and ensuing lockdowns are boosting demand for people on average (Figure 43). In addition, Africa’s data centers data storage and processing capacity.101 Empirically, countries are often less sophisticated in terms of their infrastructure and with more developed and open data infrastructure systems – services. Only three countries have data centers with “on-ramp” that have moved from domestic internet exchange points to facilities, which connect businesses and cloud providers and co-location facilities - have seen improvements in latency and help improve access to cloud services for enterprises and reduction in prices.102 The market has already seen increased households.108 Higher latencies are one direct result of a lack growth in the past years, roughly doubling in capacity to around of mature data infrastructure (including IXPs), as signals must 250 megawatts since 2016,103 and is expected to continuously travel to servers farther away. Latency on the continent is at grow at 12.6 percent until 2025.104 Increased investment is 44ms on average but varies significantly (see Figure 44). 99. Data center facilities often host internet exchange points, where Internet Service Providers exchange traffic, enabling African ISPs to exchange traffic locally and thereby reduce their reliance on more expensive international bandwidth (see Aviomoh et al., unpublished). 100. Cisco (2021) What is a data center? Available at https://www.cisco.com/c/en_uk/solutions/data-center-virtualization/what-is- a-data-center.html. 101. Xalam Analytics (2021). 102. Based on a conceptual framework that posits the different building blocks of a country’s data infrastructure as stages of a ladder, focusing on the openness of data infrastructure and not just its mere availability within a country. See Srinivasan, Sharada and Comini, Niccolo and Minges, Michael, The Importance of National Data Infrastructure for Low and Middle-Income Countries (March 2, 2021). Available at SSRN: http://dx.doi.org/10.2139/ssrn.3898094. 103. Xalam Analytics (2018). 104. In terms of investment, the Africa-wide CAGR will be 12.56 percent between 2020-2025, according to Aritzon (2020) Data Center Market in Africa – Industry Outlook Forecast 2020-2025. 105. Africa Data Center Association and Xalam Analytics (2021). Growing Africa’s Data Center Ecosystem: An Assessment of Utility Requirements. ADCA White Paper Series. Copyright ADCA and Xalam Analytics. 106. Refers to total data center space in sqm.based on the team’s calculation using data from Telegeography (2021). 107. With the notable exception of South Africa, which has a more mature data center market than the rest of the continent, accounting for 30 percent of data center (number) and 82 percent of data center space in Africa. 108. As of 2021, five data centers have on-ramp facilities: Teraco Johannesburg, Teraco Cape Town, Africa Data Centres Nairobi, Africa Data Centers Cape Town, and Rack Centre Lagos. See Srinivasan, Comini, and Minges (2021) The Importance of National Data Infrastructure for Low and Middle-Income Countries. http://dx.doi. org/10.2139/ssrn.3898094. <<< 37 >>> >>> Figure 43: Average number of internet users per Figure 44: Mean mobile latency (in milliseconds) datacenter, Africa vs. income groups across African countries Mean latency in African countries 27 LMIC 9,775,077 32 50 37 AFW 5,906,522 86 31 74 37 37 80 37 NA 5,466,343 33 44 53 36 47 54 51 39 44 35 40 61 35 39 AFE 3,305,772 0 53 36 43 17 34 51 46 48 28 68 UMIC 2,697,386 86 No data 47 45 37 58 37 HIC 769,659 34 Note: Morocco has ten, Rwanda three and Malawi and the Gambia one Note: Latency is the reaction time of the connection and measures how quickly a datacenter each. device gets a response after a request has been sent out. A low latency means a IBRD 47459 | AUGUST 2023 Source: Africa Digital Economy Database based on data from Telegeography more responsive connection. Latency is measured in milliseconds (ms). (2022), Datacentermap.com (2022) and Xalam Analytics (2018). Source: Ookla (Dec. 2020) About 140 companies, including local subsidiaries, own Africa - the only mature market for data centers on the Africa’s 185 data centers. Around 65 percent (119) of African continent with 57 data centers and 14680sqm average space. data centers are in AFE, 16 percent (30) in AFW, and 19 Kenya, Nigeria, and Egypt follow South Africa in terms of percent (35) in North Africa. The average size, for facilities market size, with 14-19 data centers by country (Srinivasan, with size information, is 4154sqm, 776sqm in AFW, 1275sqm Comini, and Minges, 2021). Around 150 of the data centers in in NA, and almost 8000sqm in AFE. The higher number of the data offer colocation services, at least in part. facilities and larger average size in AFE is driven by South >>> Figure 45: Number of data centers across African countries Number of data centers by country 7 10 3 14 1 1 2 1 18 9 2 1 3 3 19 1 3 2 0 6 1 5 10 9 5 1 15 57 5 11 No data 57 Source: Africa Digital Economy Database based on data from Telegeography (2022), Datacentermap.com (2022) and Xalam Analytics (2018). IBRD 47460 | AUGUST 2023 38 >>> INTRODUCTION Concentration in data centers appears to be lower than in concentration just based on the number of data centers, North mobile services and fiber networks. Assessing concentration in Africa has the lowest average HHI (2730), while markets in the market for colocation data centers is less straightforward AFE and AFW appear extremely concentrated. In both cases, than in other segments due to a lack of reliable data on the use however, this is driven by less mature markets that, to date, of data storage in the country and internationally. For in-country only have one operator present, such as Malawi, Sudan, and data center supply, only seven countries have sufficient data Ethiopia in AFE and The Gambia, Guinea, and Gabon in AFW. available to calculate shares in terms of data center space. For More mature markets tend to be less concentrated, for example, these seven countries, the HHIs range from 1700 (Kenya) to Kenya (HHI of 802 based on number of data centers), Nigeria around 4000 (South Africa), with an average of 2800. However, (HHI of 667), and South Africa (HHI of 600), indicating that data these figures are likely overestimating the HHI due to incomplete center markets may be less concentrated than other segments information on data center space for some providers. Assessing of the digital economy value chain. >>> >>> Figure 46: Average country level HHI by Figure 47: HHI by size of datacenters (for 7 markets) number of data centers 3,975 6,683 2,898 2,919 3,014 3,024 4,810 2,046 1,716 2,730 North Africa Africa Eastern Africa Western Kenya Nigeria Egypt Ghana Morocco Tunisia Sout h and Southern and Central Africa Note: Left-hand graph includes the following countries: Five in NA (Algeria, Djibouti, Egypt, Morocco and Tunisia), seven in AFW (Cameroon, Côte d’Ivoire, Gabon, The Gambia, Ghana, Guinea and Nigeria) and 14 in AFE (Angola, DRC, Ethiopia, Kenya, Malawi, Mauritius, Rwanda, South Africa, Sudan, Tanzania, Uganda, Zambia and Zimbabwe) Source: Africa Digital Marketplayers Database based on data from Telegeography (2022), Datacentermap.com (2022) and Xalam Analytics (2018). While data centers in Africa are predominantly privately owned, governments.109 Some examples of SOEs that own data centers governments have ownership in 20 percent of data center are Congo Telecom and Gamtel (both with limited capacity) in providers. Africa-wide, almost 78 percent of firms that own data AFW, TelOne and Infratel Corporation in Africa Eastern and centers are privately owned. Around 8 percent are wholly state- Southern, and Algerie Telecom and Egypt Telecom in North owned, 6.4 percent are majority state-owned, while respective Africa. Recently, governments are moving into the data center governments have minority ownership in 5.6 percent of firms. space. For example, in July 2021, the Government of Senegal For the seven countries where data is available, private firms opened a government-run data center in the city of Diamniadio own 86 percent of data center space in sqm – between 96.2 that offers hosting services to public and private sectors, where in AFW and 72 percent in NA. SOEs in data centers are more all government data and that of SOEs will be stored.110 In June common in North Africa and account for around 22 percent 2021, Togo inaugurated its first carrier-neutral data center built of firms owning data centers in the sub-region (Figure 48) by the government to host private and public sector data and and 26 percent of data center space. Among the top ten data to be managed and operated by Africa Data Centres (ADC).111 center-providing groups, six are at least partially owned by 109. NTT- one of the six – is owned by a foreign Government, Japan. 110. AllAfrica (2021) Senegal to Move All Govt Data to Huawei-Run Data Center, available at https://allafrica.com/stories/202106260013.html 111. The Data Center has 400 sqm (4,300 sq ft) of floorspace for the private sector and 100 sqm (1,076 sq ft) for the government. Data Center Dynamics (2021) https://www. datacenterdynamics.com/en/news/togo-opens-first-carrier-neutral-colocation- facility-in-lom percentC3 percentA9/ <<< 39 >>> Figure 48: Type of ownership of data centers by subregion (2020) 7.4% 3.45% 2.90% 6.9% 8.7% 3.4% 2.9% 14.8% 3.4% 5.8% 7.4% Other Wholly state-owned Major ity state-owned 82.8% 79.7% Minority state-owned 70.4% Pr ivate NA AFW AFE Source: Africa Digital Economy Database based on data from Telegeography (2022), Datacentermap.com (2022), Xalam Analytics (2018) and team’s research. Ownership of data center facilities remains relatively interest of leading international providers such as Equinix and dispersed, with a few leading parent groups. However, larger Digital Realty. The latter just announced intent to acquire a groups are expanding, and consolidation is picking up. The majority stake in Teraco and acquired Nigerian Medallion Data ten largest ultimate owners of data centers own a total of 66 Centres in 2021,113 while Equinix plans to buy MainOne. 114 facilities (36 percent) in Africa through local subsidiaries (see Table 2). Two of these are governments (Angola and Egypt) Vertical integration and monopolies in digital infrastructure with majority or whole ownership stakes in subsidiaries. The markets can slow the development of data centers in Africa, Government of South Africa partly owns two of these parent as they are key input markets for data centers. Vertical groups, MTN and Telkom; the Government of Kenya is active integration between digital infrastructure and data center through Safaricom, and the Government of Japan through markets remains relatively limited, with only 31 of the 139 Nippon Telegraph and Telephone Corporation (NTT), which data center providers being active in segments such as owns stakes in Dimension Data (formerly Internet Solutions), backbone and metro fiber and submarine cables. It is more among others. Most of the leading groups in the data center common to also offer cloud and other IT services. However, in market have recently built new sites or announced expansion some countries, incumbent telcos that preside over extensive nationally and in new countries. For example, Africa Data backbone and metro networks have established themselves Centres, owned by Econet Global, has recently entered Nigeria as data center providers as well. For example, Egypt and is building a second facility in Lagos, while Teraco Data Telecom,115 Gamtel in The Gambia116, and Sudatel117 in Sudan. Environments plans to enter the Nigerian market in the near This might make it more difficult for new private competitors future.112 The African data center market has also attracted the to enter the market and build data center facilities. Similarly, 112. https://www.prnewswire.com/news-releases/nigeria-data-center-market-size-by-investment-to-reach-usd-218-million-by- 2026--arizton-301440332.html. 113. https://www.interxion.com/news/2021/10/digital-realty-expands-coverage-and-capacity-of-platformdigital-across-africa. 114. Both: https://www.crn.com/news/data-center/digital-realty-and-equinix-fight-for-data-center-supremacy-in-africa. 115. https://www.datacenterdynamics.com/en/news/telecom-egypt-build-countrys-largest-colocation-data-center-2021/. 116. Gamtel has been commissioned to manage operations of the NBN Data Center in Banjul in 2019. See https://allafrica.com/stories/202007130604.html. 117. https://www.sudatel.sd/data-center/services-list/basic-services/colocation/ 40 >>> INTRODUCTION markets with monopolies and duopolies in fiber backbone and with monopolies in metro fiber. These market structures can metro networks indicate that wholesale markets for bandwidth impact the accessibility and diversity of bandwidth for data are less competitive and dynamic. For instance, the thirteen center operations through prices and route diversity. However, markets with monopolies in fiber backbone118 tend to have private companies have been able to enter these markets less developed data center markets.119 Similarly, there has despite State monopolies in the backbone or metro networks, not been data center development in nine of 14 markets as has been the case for Ethiopia, Tunisia, and Egypt. >>> Table 2: Top 10 companies in the data center segment in Africa (by size of data centers) Company Brand(s) Nr of DC State Linked? Countries Nippon Telegraph and Dimension Data, Minority-owned by Gov. of Kenya, Nigeria, Ghana, 12 Telephone Corporation NTT, ContinuitySA Japan South Africa Minority-owned by Tanzania, Nigeria, South Telkom South Africa Business Connexion 8 Gov. of South Africa Africa Africa Data Centers, Nigeria, Rwanda, Kenya, Econet Global 7 No Neotel (Liquid) South Africa Teraco Data Teraco 7 No South Africa Environments Wholly owned by Government of Egypt Egypt Telecom 6 Egypt Gov. of Egypt Côte d’Ivoire, Kenya, Minority-owned by Gov. of MTN Group MTN Business 6 Nigeria, Uganda, South Africa Ghana, South Africa Egypt, Kenya, Morocco, Sublime IP Sublime IP 6 No Rwanda, South Africa Safaricom, Safaricom is minority owned by Vodafone Group 5 South Africa, Kenya Vodacom Gov. of Kenya iColo, Interxion, Digital Realty Medallion 5 No Kenya, Nigeria Communications Unitel, Angola Cables, Majority- and wholly Government of Angola 4 Angola MSTelcom (incl. Net owned by Gov. of Angola One) Source: Africa Digital Economy Database based on data from Telegeography (2022), Datacentermap.com (2022), Xalam Analytics (2018) and team’s research. 118. Comoros, Djibouti, Egypt, Eritrea, Guinea, Guinea-Bissau, Libya, Madagascar, Niger, Central African Republic, Eswatini, Ethiopia and South Sudan 119. Nine of them have not seen data center development, while two have only one data center to date. <<< 41 Cloud services are a growing segment of the digital economy cloud computing and data centers can be conceptualized in in Africa, enabling businesses and individuals to access one value chain, with cloud computing services (IaaS, SaaS remote, high-capacity computing resources. Cloud computing and PaaS) moving from a high level of capital intensity to enables users to access data storage capacity and computing a service-oriented, OPEX-based, business model. African resources on a set of remote physical and virtual servers – the companies are increasingly adopting the cloud, with around cloud. In its most basic form, called infrastructure as a service 75 percent-80 percent of firms using cloud services in some (IaaS), cloud services provide infrastructure services similar fashion; in 2020 they spent around USD2.5 billion on all to those provided by data centers – but remote and dispersed forms of cloud and managed services, growing an average on a large number of servers. This allows businesses to of 20 percent in recent years (Xalam Analytics 2021). Cloud use computing resources flexibly and according to demand, infrastructure spending by firms providing cloud services has thus reducing the costs required to set up a computing also increased steadily over the past six years – reaching environment120 and allowing for resource-sharing with other USD72 billion in 2021 (IDC.com/Statista 2022), with the actors thus reducing the cost overall. Figure 49 shows that majority spent in South Africa. >>> Figure 49: Simplified data center and cloud value chain Capital Intensity Service-oriented business model Wholesale of data Colocation data Platform as a Software as a Hosting (IaaS) center space centers service (PaaS service (SaaS) • Renting of • Renting of fitted • Renting of • Renting of • Hosting services unfitted data data center servers in a hardware and delivered over center space space (racks), data center software tools to the internet (only the building in which servers environment, with users over the coupled with plus secured can be installed or without various internet middleware power, HVAC, directly management or software security services, services environment etc.) • Entire servers • Variable amount or portions of of data hosted servers (typically in a ‘pay as you go’ model) Source: Roland Berger and International Finance Corporation (2021) Markets for cloud-based web hosting services across Africa are websites attributed to 34 African countries reveals that various characterized by the presence of multiple local and international private cloud hosting providers are active in African markets. private players and remain moderately concentrated despite In total, 50 cloud hosting providers are active in Africa in 2021 showing an increasing footprint of large US-based providers. – 18 per country on average. Concentration remains moderate Cloud hosting refers to website hosting services provided on a on average in North Africa and in AFW, with an average HHI configuration of multiple dispersed virtual servers. Systematic of 1800 (AFE) and 2300 (NA), while the HHI is above 2500 data on the providers of cloud hosting services on over 95000 in AFW, indicating high levels of concentration on average.121 120. Including components such as hardware, storage, virtualization, as well as network and abstracted resources. See Xalam Analytics (2021) The Rise of the African Cloud 121. Includes data on 34 countries, 4 in NA and 15 each in AFE and AFW. Source: Builtwith.com (data was downloaded in June 2021). 42 >>> INTRODUCTION Morocco, Cameroon, and Ethiopia have extremely high levels the continent are all US-based tech conglomerates: Amazon of concentration of 5000 or more, while cloud hosting markets (responsible for roughly a fifth of African cloud hosting), IBM in Togo, Sierra Leone, and Sudan have low concentration (15 percent), Oracle (12 percent), and Microsoft (10 percent). at around 1000. The leading cloud hosting providers across >>> Figure 50: Market shares of leading cloud hosting providers in Africa (in terms of websites) Google, 6.5% Cloudflare, 4.8% IBM, 15.4% ServerCental, Internap Microsoft, 9.8% 3.7% Holding, 3.6% Others, 4.8% GigeNET, Gandi, 2.2% 1.8% The Constant A Small Quadr Company LLC, LeaseW Orange, aNet, Amazon, 18.7% Oracle, 11.6% Hivelocity, 7.0% 4.5% eb, 3.2% 1.5% 1.0% Source: Builtwith.com (data was downloaded in June 2021) 1 . 2 . 8 . D I G I TA L P L AT F O R M S shows a positive trend but has varied more over the years. A N D OT H E R D I G I TA L B U S I N E S S In 2021, African start-ups raised over USD 3 billion in equity OFFERINGS funding, which equals 310 percent YoY growth after a 41 percent decrease in 2020 - the fastest growth rate globally.126 Africa’s digital business ecosystem is growing rapidly, with Most investment and firm activity remain concentrated in more mature ecosystems in Nigeria, Kenya, South Africa, Africa’s more mature digital ecosystems like Nigeria, South and Egypt attracting the most funding and new firms. As of Africa, Kenya, and Egypt, which accounted for 87 percent 2020, there are more than 200,000 digital businesses122 in of VC investment in 2021. As a result, AFE has attracted by 190 countries across a range of sectors, according to the far the largest total sum of VC funding, followed by AFW and WBG Digital Business Database123 ; and while African firms NA (Figure 52). South Africa attracted the bulk of VC deals, represent only a small fraction of digital businesses globally accounting for 20 percent of deals and 43 percent of VC capital (1.7 percent),124 venture capital-funded start-up ecosystems invested in 2021. However, other markets are catching up, across the continent have seen continued growth in the such as Senegal, Ghana, and Morocco, which have been able past ten years (Figure 51). Since 2010, around 2,500 start- to attract more funding in recent years, diversifying Africa’s ups across 45 African countries have begun operations.125 tech-/startup landscape. The number of countries attracting Capital raised (VC, angel, or incubator/accelerator funds) also VC funding has increased steadily over the last few years.127 122. Defined as “Digital solution providers that develop and manufacture digital technology products or digital services in the narrow scope of digital economy or in the core digital (IT/ICT) sector.” Zhu, Tingting Juni; Grinsted, Philip; Song, Hangyul; Velamuri, Malathi. 2022. A Spiky Digital Business Landscape: What Can Developing Countries Do?. World Bank, Washington, DC. 123. Companies HQed in SSA+North Africa: 4997, Companies Operating in SSA+North Africa: 5047. See WBG “Zhu, Tingting Juni; Grinsted, Philip; Song, Hangyul; Vela- muri, Malathi. 2022. A Spiky Digital Business Landscape: What Can Developing Countries Do?. World Bank, Washington, DC. 124. Of around 250,000 firms on Pitchbook with VC, PE, Angel or Incubator/Accelerator backing, 4329 have a presence in Africa (HQ or office location). [Data accessed September 16, 2021] 125. Refers to firms HQ’ed in Africa that are angel-, incubator/accelerator-, or VC-backed and founded since 2010. Source: Pitchbook (2021) 126. Partech (2021) 127. In 2021 alone, the number of countries reached by VC funding increased from 26 to 29. See Partech (2021) Africa Tech VC Report 2021. <<< 43 >>> >>> Figure 51: Overview of new VC-funded businesses and VC Figure 52: Total number of African VC-funded start-ups and capital invested since 2010, all Africa VC capital invested, by subregion Capital Invested Firms founded Capital invested Firms founded 3,500 700 623 5000 1400 3,000 566 600 540 1309 4500 1200 2,500 454 500 4000 399 3500 1000 2,000 335 400 3000 800 280 2500 726 723 1,500 300 4758 600 204 2000 1,000 142 200 1500 400 108 1000 500 54 100 200 15 500 1030 678 0 0 0 AFE AFW NA 10 11 12 13 14 15 16 17 18 19 20 21 20 20 20 20 20 20 20 20 20 20 20 20 Note: Shows the number of firms founded since 2010 that are backed by angel investors, incubators/accelerators, and VC funds and total capital invested in these firms (in USD thousands). Source: Pitchbook (2021) [accessed on June 16, 2022] >>> >>> Figure 53: Number of VC deals in African subregions, 2019- Figure 54: Total VC capital invested in African HQ’d firms, 2021 2019-2021 356 1,525 308 1,423 281 246 228 200 889 181 169 172 435 409 390 194 145 42 Africa Eastern Africa Western North Africa Africa Eastern Africa Western North Africa and Southern and Central and Southern and Central 2019 2020 2021 2019 2020 2021 Note: Shows the total number of venture capital deals (all stages) in firms with headquarters in Africa and the total amount that was invested in these VC deals (equity investments). Source: Pitchbook (2022) [accessed on February 25, 2022] 44 >>> INTRODUCTION The growing VC activity has been driven by investments in a across countries, Fintech was the most significant vertical handful of sectors, particularly fintech. Fintech remained the across the six largest African countries in terms of venture most important vertical in venture funding for African startups investment, accounting for 45-95 percent of deals in Nigeria, in 2021 (Partech, 2022), attracting over USD 3.2 billion or South Africa, Egypt, Kenya, Senegal, and Ghana (Partech, 60 percent of capital invested on the continent (see Figure 2022). Among the twenty128 highest funded firms in Fintech 55). Fintech investments tended to be larger, often involving that have offices in Africa, five start-ups based in the UK, India, megadeals, which is also evidenced by the small share of and the US combine over 86 percent of total funding, with Fintech deals (32 percent) compared to their share in value. Finastra (UK) having received about 45 percent of the total. Following Fintech, the verticals logistics, edtech, E/M/S- Finally, Fintech and e-commerce verticals have produced commerce, and enterprise productivity have attracted the the first unicorns129 of the continent, Flutterwave, Interswitch, most investment in terms of equity value and number of deals, Fawry in fintech, and Jumia in e-commerce, underlining the accounting for 7-5 percent of deal value between 2019-2021. promise of these sectors for investors. While the distribution of investment across verticals differs >>> Figure 55: 2021 VC funding in Africa across verticals, in US$M Fintech 3257 Logistictech 388 Edtech 291 E/M/S/Commerce 285 Enterprise 261 Health tech 230 Clean tech 193 Agritech 121 Mo bility 105 Entertainm ent 36 Insurtech 36 Connectivity 26 HW & Electronics 13 Source: Partech 2021 E-commerce is a key sector for digital firms in Africa, driven is the leading e-commerce platform on the continent with a by rising numbers of shoppers across the continent, with a diversified product offering from e-commerce marketplaces large footprint of African platforms. While the penetration of to logistics, food, and payments across eleven economies.132 e-commerce in Africa remains relatively low at just below Among the eight highest-funded e-commerce firms in the 30 percent,130 it has been increasing at around 30 percent FCI Digital Business database,133 Jumia has attracted over YoY between 2017-19 and is expected to continue growing, half of the total amount invested between 2008 and 2020. averaging around 17.9 percent CAGR between 2017-2025. Souq, the largest e-commerce platform in the Arab world134, This is expected to increase the penetration rate of e-commerce and Takealot of South Africa are the second and third. Jumia from 28 percent (334 million online shoppers) to 40 percent, or also leads the market in terms of website visits, accounting for 520 million people, by 2025.131 Nigerian conglomerate Jumia over 370 million visits annually across South Africa, Nigeria, 128. Finastra, Paytm, You & Mr Jones, WorldRemit, Tala, MetricStream, Mozido, Branch International, Checkout.com, Interswitch, OPay, Jumo, thumbzup Innovations, Emergent Payments, Confirmation, CR2, PayJoy, Invuo Technologies, Eka, Flutterwave. 129. Firms with a valuation of USD 1 billion and more. 130. Statista (2022) The rise of e-commerce across Africa. 131. Statista (2021) Statista Digital Market Outlook. Available at https://www.statista.com/statistics/1190579/number-of-online- shoppers-in-africa/. 132. See jumia.com 133. We used the top 20 highest-funded firms in the e-commerce category and narrowed this list down to predominantly e-commerce firms. The remaining 12 firms are categorized as predominantly Fintech firms. Source: World Bank Digital Business database. 134. Souq was acquired by Amazon in 2017. See https://techcrunch.com/2017/07/03/amazon-souq-com-completed/. <<< 45 Kenya, Egypt, and Morocco alone (see Figure 57). Other for purchases made on its platform Aliexpress.137 To improve prominent players are South African conglomerate Naspers, conditions for hipping of both finished products and inputs which owns the marketplaces OLX (Egypt) and Takealot; Jiji, across the continent and given low logistics performance,138 which is active in Kenya and Nigeria and accounts for over 64 digital businesses and platforms have entered the African million website visits in the two markets; Avito in Morocco, and logistics sector. VC funding has been increasing, totaling 388 smaller outlets like Konga in Nigeria and Kilimall in Kenya.135 million in 2021. In agtech, with agriculture being central to jobs Global e-commerce giants Amazon and Alibaba have minimal and GDP in Africa, digital firms have attracted around USD operations on the continent. Amazon has purchased and 120 million in VC funding in 2021 (Partech, 2021). Compared rebranded Souq in Egypt and launched AWS services in to fintech and e-commerce, funding of the top highest funded South Africa,136 while Alibaba has introduced shipping to Africa firms in logistics and agtech appear less concentrated.139 >>> >>> Figure 56: Total funding for e-Commerce (in USD million) Figure 57: Website visits of largest e-commerce sites in Africa 57.1% 25.3% 5.2% 4.8% 3.3% 2.6% 1.7% Jumi a Souq Takealot Qoo10 We Buy Konga Copia (NG) (UAE) (SA) (SG) Cars (NG) Global (SA) (KE) Source: Pitchbook (2021), International Trade Centre (2019). With a young and mobile-savvy population, mobile app usage African subscribers only downloading an average of 2.4 apps is increasing in Africa but still significantly lagging behind per month, compared to 4.3 in middle and 4.7 in high-income other regions. Mobile devices generate around 70 percent of countries. On a regional level, SSA is lagging other regions web traffic in Africa, more than in any other region.140 Since in the amount of time people spend on their phones. Mobile Q1 2020, mobile app usage downloads usage (as measured subscribers in SSA spend less than half of the time on apps by app downloads) increased by 41 percent across Africa. than those in SAR, North America and ECA and less than a However, people in Africa still access and use mobile internet quarter of the time than subscribers in MENA, LAC, and EPA. less than in other regions. In 2021, African mobile subscribers App usage is also growing more slowly in African countries used apps less than half as often as mobile phone subscribers compared to other regions, as the number of total sessions in in other middle-income (MIC) and high-income countries (HIC, Africa has grown by 3 percent since January 2020, compared see Figure 58). This pattern is similar for app downloads, with to 17 percent in HIC and 10 percent in MIC.141 135. International Trade Centre (2019). 136. See https://aws.amazon.com/about-aws/global-infrastructure/regions_az/. 137. www.aliexpress.com. 138. SSA ranks lowest among the World regions with an average score of 2.5, compared to 2.6 in SAR, 2.7 in LAC and 2.8 in MENA in the aggregated WBG Logistics Performance Index. See WBG (2012-18) Connecting to Compete: Trade Logistics in the Global Economy. 139. Among the top ten highest funded firms in the World Bank Digital Business Database, the top three have collected less than 50% of funding (in USD million) in fintech and just slightly above 50% in agtech. 140. StatCounter 2021, see https://www.statista.com/statistics/306528/share-of-mobile-internet-traffic-in-global-regions/. 141. Begazo, Dutz and Blimpo (2023) 46 >>> INTRODUCTION >>> Figure 58: Comparison of app usage in Africa vs HIC, MIC, and other regions, 01-12/2021 a. Total app sessions by unique mobile internet subscriber b. Total number of hours spent on apps monthly, by unique mobile internet subscribers 67.6 68.1 408 60.4 376 37.7 39.0 34.6 156 14.8 Africa High income Middle SSA SAR NA ECA MENA LAC EAP income Note: Sample includes 63 lower and upper middle- and high-income countries. African countries included are Egypt, Tunisia, Kenya, Malawi, Tanzania, South Africa, Ghana, Burkina Faso, Senegal and Nigeria, all middle-income countries. Left graph shows total app sessions in a given quarter between 2016 and today. Right chart shows the top 500 apps (paid) in iOS stores, as of Dec 2020. Source: Elaborated based on data from Apptopia (2021) and GSMA (2021). App usage appears to be relatively concentrated in the four publishers in 2021 of 5254 in Kenya, 6200 in South Africa, largest markets, pointing to the importance of a few primarily 8241 in Egypt, and 8440 in Nigeria, indicating extremely foreign-owned digital firms in profitable digital services high levels of concentration. In addition, around 80 percent markets. Given the importance of attention for gathering of publishers in the sample are from foreign (non-African) and monetizing user data,142 engagement with mobile apps firms. Locally developed apps are often in the fields of finance can shed light on the importance of digital firms that publish (banking, mobile money) or e-commerce (e.g., Kilimall, apps in Africa. In South Africa, Nigeria, Egypt, and Kenya, Takealot, Jumia) and developed by both large firms such as large app publishers, in particular Meta (formerly Facebook), Safaricom and Jumia as well as smaller local publishers. The dominate app markets in terms of daily active users (DAU). importance of large US-based digital conglomerates – e.g., Among a sample of the current top 50 apps,143 Meta-owned Meta on mobile apps – is akin to other regions (including high- apps (Facebook, Facebook Messenger, Instagram, and income economies) and is accelerated by network effects and WhatsApp) averaged 82.5 percent of active users in 2021. winner-takes-most dynamics associated with digital platforms. This translates to HHIs based on the relative importance of 142. For a discussion, see for example: WBG (2021) World development Report 2021 – Data for better lives. 143. The sample consists of the top 25 free apps and the top 25 grossing apps in a given country in January 22 according to Apptopia. <<< 47 >>> Figure 59: Relative importance of top ten apps in terms of daily active users (DAU) among a group of top 50 apps a. Kenya b. Nigeria 71.5% 91.5% 6.7% 5.8% 3.8% 6.1% 2.6% 2.5% 2.8% 2.6% 2.5% 2.3% 2.0% 2.0% 1.7% 1.5% 1.3% 1.2% 1.2% 0.9% ap d ok er IA ra am LC a k i te ac et pe it t d d d il e M kT Sn ok IA ra rg m a -L M Te Lt Tw Lim i te i te JU et om ea pe .o O ob M Ti kT FZ M N Lim Lim ad g JU T di O Ti M ZE s in m Au ce g re ar la ic om a in E2 yb Ta vi gr Sh us ar er M ri c ail le Sh tM lS le Te iD fa Fi ne ta ile Sa er gi F ss nd Di er an nd Xe y Tr Pa Xe O c. Egypt 90.5% d. South Africa 90.5% 5.9% 2.3% 5.9% 2.3% 1.5% 1.1% 1.0% 0.9% 0.9% 0.8% 0.7% 1.5% 1.1% 1.0% 0.9% 0.9% 0.8% 0.7% c d es d. LC a C . r AB td c d es d. LC a C d. In Lt r le AB et Lt LL In Lt lle et am L al Lt LL Lt -L M ap am e e. -L i fy e. M ap ca e ec FZ e. Pt e i fy e. G Pt FZ Pt le Sn Pt l ot G Pt ue i u Sn Pt ot ob i gy BO Tr Sp ob m gy gy BO ok Tr Sp am M gy ok lo a M lo lo SY kT gr lo no on SY kT gr no no on le no Ti ch le Ti az Te ch ch az Te ch Te Am Te Te Am Te yo 4U yo 4U Jo Jo ia ia ed ed tM M ar t ar Sm Sm Source: Apptopia (2021) Vertical integration within digital economy markets is growing, Logistics. Konga (Nigeria) manages its shipments through its driven by leading MNOs that invest in digital services and own logistics arm KXPRESS, and payments through KongPay, platforms offering services in adjacent digital markets. While and Kilimall (Kenya) also owns proprietary payment and logistics overall, the extent of vertical integration between digital services. This is a trend also observed in other economies. platforms and owners of digital infrastructure in Africa is low, Furthermore, vertical integration of international digital service some leading MNOs have started expanding in digital services providers (Meta, Alphabet) into international connectivity could markets. For instance, Safaricom of Kenya offers a subscription- also affect digital service markets. based mobile-first, video-on-demand service- BAZE;144 an online shop for mobile phones; and a one-stop-shop app for In sum, from a market characteristics perspective, digital farmers to buy inputs, technical advice on farming practices and infrastructure markets are generally more concentrated financial services - DigiFarm145. 22 of 31 leading MNOs in Africa and face more government direct participation and vertical offer cloud and hosting services, and 16 own data centers. integration. Therefore, this increases the need for governments Further, leading African platforms (often in e-commerce) have to step up their interventions to improve economic governance entered adjacent digital markets. For example, Jumia has its and address exclusion risks (ensure balanced outcomes own payments and logistics offerings, JumiaPay and Jumia for all). Data infrastructure and digital services seem less 144. https://www.safaricom.co.ke/about/media-center/publications/press-releases/release/1052 145. Digifarm FAQs, available on Safaricom’s website: https://www.safaricom.co.ke/faqs/faq/810. 48 >>> INTRODUCTION concentrated and more dynamic. However, vertical integration that create enabling frameworks for digitalization while of MNOs into mobile payments seems to be escalating mitigating risks. Digital-related policies were once focused on concentration in mobile payments, while the operation of telecommunication regulation that focused on setting rules submarine cables by global digital service providers could for entry (licensing), access to essential infrastructure and affect the market dynamics of international connectivity. In resources (interconnection and radio spectrum), and control the case of data infrastructure and digital services, improved of operators with significant market power and universal economic governance is needed to mitigate risks related to service. With the expansion of digital services across sectors, data misuse, cybercrime, and barriers to the creation and technological innovation, and new business models, the expansion of digital business. Government interventions regulation of digital services is now broader, encompassing a would be needed to ensure digital trust and unbiased market wide array of institutions looking at digitalization issues from outcomes. For this, effective, transparent, and accountable their own sectoral or topical perspective. The coordination institutions are needed. of such an atomized group of institutions is a challenge for policy coherence and effectiveness. If not addressed, it could generate contradicting policies that prevent the development of a fully operational digital environment. Furthermore, aside 1.3. An overview of the public from national institutions and frameworks, international institutions that set rules on digital entities and rules (multilateral and bilateral) also affect the markets in Africa: what actors are economic governance environment for digitalization, such as international conventions, rules by regional communities, and supporting economic governance of bilateral trade agreements. digital? Digital technologies have slowly permeated many aspects of the daily life of citizens and usual business operations, and governments are responding through institutions >>> Table 3: Examples of government entities setting rules that affect the economic governance of the digital sector Name of the entity Information on their operation and role related to governance of digital in Africa Digital Infrastructure or ICT sector There are 54 telecommunications (digital infrastructure) regulators in Africa, but only 18 reported as independent. Telecommunications regulators in many countries apply relevant regulatory tools to ensure healthy telecom markets. One important aspect for inclusive economic governance is setting obligations on operators with significant market power, that aim at enabling competition in markets. Countries such as Senegal, Mauritania and Ghana have imposed obligations on Telecommunication dominant operators, while the process stalled in Kenya and South Africa is finalizing the process. Regulator Regulators are mostly financed by regulatory fees on operators, creating a dependency on sector’s revenues and potential regulatory capture. The transparency, accountability and financing of regulators is not well aligned with good practice as identified by the Regulatory Watch Initiative covering 23 African countries. This ministry is tasked with policymaking related to digital infrastructure and other digital matters Telecommunications/ (cybersecurity, digitalization). In some cases, it also oversees government commercial activities Communications/IC T in the sector, potentially creating conflict of interest. In 60 percent of the African digital SOE/SLEs Ministry in 20 African countries, the Telecommunications Ministry or equivalent exercises control over the state shareholding in those entities. <<< 49 Name of the entity Information on their operation and role related to governance of digital in Africa Digital Infrastructure or ICT sector Most African countries have implemented Universal Funds to make telecommunication services accessible and affordable for all. Universal funds obtain their funds from a percentage of the Universal Fund revenue of telecommunication operators and invest it to fulfil the mentioned objectives. However, Administrators their effectiveness has been limited with contributions not spent as expected. The transparency, accountability and financing of regulators is not well aligned with good practice as identified by the Regulatory Watch Initiative covering 23 African countries. Radiofrequencies are a valuable public resource that can be managed independently. In Africa, either the Telecom Ministry or equivalent or the telecom regulator manage spectrum, deciding on Radio spectrum allocation, assignment, pricing and monitoring. The transparency of assignments and pricing is not Administrators well aligned with good practice as identified by the Regulatory Watch Initiative covering 23 African countries. This can give way to governance issues due to the potential for revenue generation. Out of 44 African countries analyzed by the International Communication Union, 13 had created Cybersecurity Regulator Computer Emergency Response Team to implement a national cybersecurity strategy. Mandates on cybersecurity fall in some cases on ministries of ICT or telecom regulators. Depending on the country, the agency in charge of e-government falls under the purview of the ministry of ICT or other ministries and their agencies such as economy, industry. For example, E-government Agency Nigeria’s National Information Development Agency coordinates the implementation of IT systems at all levels of Government. Economy-wide entities 29 competition authorities operate in Africa with 10 handling antitrust cases, merger review and advocacy initiatives in digital markets. For example, the South African Competition Commission has conducted market enquiries, assessed mergers, imposed sanctions due to anticompetitive Competition Authority practices in digital infrastructure, and started cases in digital services markets. In addition, competition authorities have covered various topics related to digital infrastructure (access to essential infrastructure, transfer of spectrum, use of USSD codes for mobile money and other digital financial services). Taxation has been central to the digital regulatory debate with countries taxing digital connectivity Ministry of Finance through excise taxes (VAT overcharges, stamp duties) and others imposing taxes specific to digital services such as social media and mobile money transfers. 26 African countries have set a specific data protection framework that mandates the creation of a Data Protection Authority data protection authority. Nonetheless, these authorities are not all functional or independent. The debate surrounding intellectual property is central to the development of digital services. Authorities such as the Nigerian National Office for Technology Acquisition and Promotion has, for Intellectual Property example, the power to disseminate data of locally generated technologies, while South Africa’s Regulator Companies and Intellectual Property Commission granted a patent to a product created through artificial intelligence, being the first patent office to assure such protection to a non-human inventor. The digital environment brings new challenges to the protection of consumers from spams and exploitation. Nonetheless, many African countries lack digital-specific consumer regulations Consumer Protection making it more complex to respond to these challenges. Algeria and Tunisia are two examples Authority of countries with specific digital consumer regulations while authorities such as the South African National Consumer Commission oversee the prevention of such abuses under the general consumer protection legislation. In charge of programs to promote the adoption of technology, including digital technologies, for Innovation and productive use. Interventions are related to innovation hubs, accelerators, and incubators for digital technology agency solutions. 50 >>> INTRODUCTION Name of the entity Information on their operation and role related to governance of digital in Africa Sector-specific entities Central Banks have been key to regulate the operation of mobile money providers and fintech. Collaboration with telecom regulators and competition authorities has been key to support mobile Central Banks/ Financial money/payment systems. Open banking initiatives to facilitate data sharing in the sector can also Sector regulator benefit from collaboration with data protection authorities. Regulation of data use for algorithms (AI) (for example for credit rating) is also a new area of digital intervention for financial sector regulators. Digital technologies in the sector such as intelligent transport systems, cargo tracking systems and ride hailing apps are partly overseen by the transport regulator/ministry. For example, transportation agencies such as the Kenyan National Transport and Safety Authority (NTSA) Transportation Regulator established to harmonize the operations of different transport departments and improve road safety to that end the NTSA established the Transport Integrated Management Service that helped to improve the management of, for example, driving licenses. Rules for the creation of public good databases for data sharing fall under the ambit of the sector Agriculture Authority regulator. For example, Ethiopia’s extension of the Agriculture Sector Policy and Investment Framework on 2017, highlighted the importance of adopting ICT technologies in this sector. Source: non-comprehensive list of government entities involved with the regulation of certain aspects of digital technologies. Telecom regulators and competition authorities are the sanctions and settlements with the involved parties, helped main bodies in charge of addressing exclusion risks. The lower the prices of the services involved. main regulatory body for digital infrastructure continues Nonetheless, only 36 African countries have adopted to be the telecommunication regulator. Out of the 54 competition laws, 43 are covered by either national or regional telecommunications regulators in Africa, 18 are independent competition laws, and only 26 have operational authorities147. from the central government,146 accounting for whether the To add to this complexity, Competition Authorities do not always Central Government has direct control over their operations. count on an adequate budget or an independent structure148, Telecom regulators have had a particularly relevant role and collaboration with telecom regulators can be limited. While in promoting the affordability of digital connectivity and telecom laws allow regulators to set pro-competition rules (ex- closing the digital divide. They are generally not alone ante regulation) and, in some cases, also intervene after a in this quest, as Competition Authorities have increased competition infringement, digital services generally fall out of their relevance in digital services and digital infrastructure the scope of the law. The absence of a competition authority markets. Competition authorities have tools to promote may result in a lack of pro-competition policies in digital service competition through policy advocacy, impose remedies and markets. Concurrent mandates on antitrust enforcement of sanctions in cases of agreements between competitors or competition authorities and telecom regulators (Box 5) require abuse of dominance, and impose conditions for the approval close collaboration for increased effectiveness but support a of mergers. The South African and Kenyan authorities are broader ecosystem to address competition issues that can examples of Competition Authorities whose interventions in counter potential regulatory capture of the sectoral regulator. telecommunication markets, ranging from market studies to 146. Team’s analysis based on the Country Profiles Provided by Teleography‘s Global Comms, analyzed on March 12, 2022 (https://www.telegeography.com/products/global- comms/); the RWI was also used as a source of the analysis. 147. Pop, G. and G. Coelho (2023) “Competition Policy in Digital Markets in Africa” 148. Nyman and Falco (2020) <<< 51 BOX 5: COMPETITION AUTHORITIES AND TELECOM REGULATORS’ MANDATE ON SAFEGUARDING COMPETITION In many countries, sector regulators — either alone, or concurrently with competition agencies —are granted enforcement mandates around competition issues in the telecom sector, a key enabler of competition in digital markets.149 Effective competition enforcement in digital infrastructure is paramount for the development of downstream digital markets. An effective enforcement of competition law can contribute to alleviate the market power concerns that may stem from telecom-related bottlenecks. The countries where both sector regulators and competition agencies can enforce competition rules in the telecoms sector are: Cameroon, Egypt, Ethiopia, Kenya, Madagascar, Mauritius, Namibia, Seychelles, South Africa, Eswatini, Tanzania, Zambia, Zimbabwe, Tunisia and Gabon.150 Various African countries attribute competition enforcement powers in the telecoms sector to the telecom regulators. This is the case of those countries: (i) without a functioning competition agency (Cabo Verde, Guinea, Liberia, Mauritania, Rwanda, São Tomé e Príncipe and Chad); (ii) where the national competition agencies do not have enforcement powers (Burkina Faso, Côte d’Ivoire, Senegal,150 Central Africa Republic, Republic of Congo, Democratic Republic of Congo); and (iii) where the telecoms regulator has exclusive competition law enforcement powers over its sector (Morocco). It is worth noting that telecom regulators have generally a mandate to promote competition but not antitrust powers, for example in Democratic Republic of Congo. Source: Pop and Coelho, (2023), “Competition Policy in Digital Markets in Africa” Taxation has also been a part of regulating digital ecosystems The evolution of digital business models, the use of online to ensure appropriate contributions to the central budget for services, and the increasing capture, use and re-use of digital government provision of complementary services and more data have increased concern on how safe those environments efficient government functioning. Despite the relevance of are and how to prevent cybersecurity threats and data misuse. taxation in promoting the development of the economy in In this sense, many countries have created offices dedicated any country, it is necessary to align the taxation of digital to safeguarding such environments, including data protection connectivity and digital services with policy objectives such and cybersecurity agencies. Data protection authorities as affordability and diffusion. Preventing from doing so (DPAs) are an essential component of the 26 specific laws could slow the expansion of the use and availability of digital issued to enable data regulation in the African continent. services. This is the case of taxation of digital platforms in, for Nonetheless, in several of these countries, including Angola example, Uganda, which imposes a daily tax on social media, and Egypt, the creation of the DPA is required by law but has or examples of taxation on mobile payments in Malawi, DRC, yet to be established in practice. Although the preservation and Kenya. Promoting coordination between agencies would of the security of data, consumers, and the digital ecosystem help generate taxation structures that are efficient, equitable, as a whole is essential to the promotion of a sustainable and simple. Another example of taxation that can affect the digital ecosystem, it is essential to coordinate such policies development of the digital economy is excessive taxation of with competition policies so that the implementation of data digital infrastructure, particularly sector-specific excises, VAT protection-related policies does not constitute an unnecessary overcharges, and higher CIT. Appropriate policy coordination burden for new market entrants. to balance short-term revenue collection needs and medium- term effects on the digital economy is key for a better Inter-institutional collaboration between financial regulators, economic governance environment where governments do telecommunications regulators, and competition authorities, not expropriate business through excessive taxation. among others, is paramount to promoting the adoption of digital 149. See, for example, Kenya (the Communications Authority), Rwanda (the Utilities Regulatory Authority), Cape Verde (the Multi-Sector Regulatory Agency for the Economy). 150. In the following countries, a MoU exists between the competition agency and telecoms regulators: Eswatini, Mauritius, and Namibia. 52 >>> INTRODUCTION financial services. This entails interaction between ex-post the Competition Directorate of the WAEMU Commission has competition enforcement, ex-ante merger control, and rules to blocked a proposed joint venture to create an interoperability enable entry and expansion of new players (such as regulation platform between two mobile operators also active in digital of access to USSD, payments interoperability, or access to financial services due to strong competition concerns.151 clearinghouses). As described in Pazarbasioglu et al. (2020), Competition authorities may enter into memorandums of policy enablers of DFS can be divided into three categories: understanding (MoU), with telecommunications and financial conducive legal and regulatory frameworks, enabling financial service regulators (World Bank, 2022). These MoUs can and digital infrastructure, and ancillary government support cover a wide variety of matters, including: (i) management of systems. The categories imply, among others, promoting complaints, (ii) designation of technical focal points/working mobile phone penetration, rolling out digital and biometric ID groups, (iii) development of joint action plans/platforms for systems, and ensuring competition between incumbents and sharing know-how and expertise, and (iv) development Fintech providers. Financial regulators also need to consider of implementing procedures for consultations on cases, the challenges that could arise from the development of regular information exchanges, and comments to laws and DFS, particularly those related to users’ privacy and safety— regulations, which may impact competition in the sector. requiring cross-sector coordination with the relevant authorities in data and consumer protection, and cybersecurity. A level Sectoral ministries have also been involved in regulating digital playing field in access to data technologies and infrastructure environments. Transport regulators in Nairobi, for example, is essential to promote the development and usage of DFS. had long aimed at reducing the usage of cash when paying As mentioned in the World Bank’s Fintech and the Future of for public transportation. Despite some attempts to implement Finance Overview Paper, connectivity through mobile internet card payments, it was through the use of mobile payments and low-cost computing and data storage are essential for through mobile phones that they accomplished the objective. the development of Fintech companies. Regulation of digital Digital platforms for ride-hailing services disrupted traditional infrastructure is, consequently, essential for the development urban transportation markets, and ministries/regulators of of DFS. Furthermore, the growth of DFS depends partly on transport intervened to enable the operation of such players. regulators being able to set rules and regulations that would Sectoral agencies or ministries – such as those in charge of allow for access to essential infrastructure. regulation in agriculture, health, education, and energy - are also introducing rules related to digitalization, data protection, Mobile money providers in Africa have been found to engage cybersecurity, and data sharing, among others, needing in unilateral practices that may require further competition coordination with the respective authorities. scrutiny, such as entering into exclusivity agreements with a network of agents. For instance, in Kenya, the Competition As more government entities introduce digital-related Authority (CAK) investigated Safaricom (M-Pesa) for alleged regulations to either safeguard a particular aspect of the abuse of dominance through exclusivity contracts that digital environment or promote its usage, coordination excluded rival money transfer providers from M-PESA’s mobile becomes essential. ITU measures the level of collaboration agents’ services, resulting in around 85,000 agents being between different regulators on digital matters, covering two barred from engaging in business with other mobile operators. instances of cooperation: (i) the regulatory collaboration Safaricom committed to allowing other mobile payment firms in digital core areas, including the collaboration of core to work with its mobile money transfer agents. The package digital regulators such as telecommunication regulators and of remedies was created in close collaboration between the broadcasting regulators, and (ii) cross-sector institutional competition agency and telecoms regulator and designed to cooperation including a broader array of regulators such as open up the main bottlenecks of Safaricom’s mobile money health, environment, or education ministries or competition transfer service (M-Pesa). African competition authorities have authorities. Lower levels of connection between agencies will also dealt with merger notifications regarding the creation of make it more complex for an integrated approach to digital interoperability platforms by mobile money platforms with regulation, risking isolated actions that do not accomplish potentially exclusionary effects for third parties. For example, policy goals. Out of the 57 African countries included in the 151. Opinions 01/2020, of July 7, 2020, and 003/2000 of June 20, 2000, of the Court of Justice of WAEMU: In the West African Economic and Monetary Union (WAEMU), competition enforcement is exclusively entrusted to the regional WAEMU Commission. <<< 53 analysis, 16 are below the average of low-income countries. high levels of connection among different authorities. Although While Côte d’Ivoire leads the ranking with the highest score in the ITU analysis shows greater interaction between various the sample with 25 points. Nigeria, Malawi, Ghana, Botswana, authorities in these countries, those with lower levels of The Gambia, Eswatini, Rwanda, Morocco, and Cameroon are coordination need to promote collaboration across the board also above the average of high-income countries, showing for a more efficient digital environment. >>> Figure 60: Degree of progress towards National Collaborative Governance Note: Country classification data according to the World Bank’s Country and Lending Groups presented in July 2021. Source: ITU 5G Benchmark. Last reviewed April 5, 2022. Besides coordination, more effective regulatory frameworks market power (SMP), spectrum management, and rules on and implementation are key. For example, in the case international connectivity. of digital infrastructure regulation, both the adoption of international practices on the books and implementation lag Additionally, based on the analysis presented in the in some countries. Policies that restrict competition or do not “Technological Transformation for Jobs in Africa: How Digital enable competition in the digital-relevant markets affect how Development Can Support Good Jobs for All” World Banks these markets develop and increase risks of unfair or biased report, implementation is not consistent across African market outcomes that do not favor consumers. As shown by countries: Out of the eight operators with more than 65 percent indicators collected by the World Bank’s Regulatory Watch market share in mobile services, only one is subject to specific Initiative, the level of alignment with international best practice SMP obligations. This weak implementation of regulation of rules and implementation is only 60 percent on average digital infrastructure, combined with the underenforcement (Figure 6X). These indicators cover five key areas relevant of competition and data rules -that will be further developed to good economic governance: licensing, interconnection and in this report -makes it more complex for the development of access to infrastructure, regulation of operators with significant digital services. 54 >>> INTRODUCTION >>> Figure 61: Alignment with good international regulatory practice and implementation 5 4 3 2 1 0 Egypt Sierra Leone Guinea-Bissau Algeria Tanzania Uganda Burkina Faso Ghana Guinea Liberia Senegal Kenya Rwanda Benin Cabo Verde Togo Mo ro cco Tunisia Côte d'Ivoire Gambia Ma li Niger Nigeria Licensing Interconnection SMP Spectrum Management International Connectivity Source: Technological Transformation for Jobs in Africa: How Digital Development Can Support Good Jobs for All flagship report based on World Bank, Regulatory Watch Initiative (RWI) Phase 2, 2021. Note: Information as of October 2020 covering 23 African countries in North, West, and East Africa. Full alignment with good practice = 1 for each area. Having guiding principles that rule over digital regulation and allow for fluid interactions between different agencies could prevent unintended negative consequences that impact the development of digital services while promoting diffusion and affordability. In general, three principles for more effective economic governance in the sector apply: (1) embedding competition principles to curb persistent, excessive market power, (2) allowing for agile regulation152 given the rapid pace of technological change, and (3) integrating collaborative policy making across government entities, with private sector and civil society and taking into consideration supranational rules. 152. Agility implies “an action or method of nimbleness, fluidity, flexibility or adaptiveness.” The principles of agile regulation are: inclusive and collaborative, adaptive and iterative, proactive and future-facing, decentralized and experimental, activity- based and risk-weighted, technology-neutral and outcome-based, and contextual and human-centered. <<< 55 2 >>> Addressing old economic governance risks for digitalization: channeling resources to the state through taxation and direct participation in markets Addressing governance risks in digital infrastructure is key to minimizing risks of exclusion from digitalization through affordable, available, and accessible connectivity, but the institutional design can create incentives for the government or government officials to protect certain operators and bias market results. Besides having accountable, transparent, and technical regulators that address market failures and promote social goals, there are two important aspects that create risks due to goals that might conflict with regulation for better functioning digital infrastructure markets. Governments can see the sector as a source of revenue generation and therefore impose excessive taxes and parafiscal fees disregarding or discounting negative effects on the inclusive adoption and use of digital technologies for economic transformation and growth. Governments and government officials can also see state ownership of enterprises as a means to extract rents for the government (especially through SOE or SLE placed at an advantageous position) or personal rents (especially through board membership at SLE or not wholly owned SOE). Furthermore, state ownership of enterprises confers the government with the possibility of influencing more directly (from the operational perspective) the performance and direction of the sector. In both cases, incentives can foster regulatory protection of certain firms, either to ensure revenue collection or a comfortable environment for SOE or SLE. Spectrum management is another area where the government or government officials can extract rents from the sector, especially when deciding 153. For an overview of spectrum pricing and licensing in Africa, see GSMA (2021), available at https://www.gsma.com/spectrum/resources/effective-spectrum-pricing-africa/ 56 >>> ADDRESSING OLD ECONOMIC GOVERNANCE RISKS FOR DIGITALIZATION on spectrum assignments and spectrum pricing. However, they pose. A broader set of interventions is needed to fully this topic is outside the scope of this report.153 This section address the risks of digital exclusion (Box 6), and they are zooms in on taxation and enterprises with (full, majority, covered more extensively in the “Digital Africa: Technological and minority) state shareholdings and the governance risks Transformation for Jobs” World Bank’s flagship report. BOX 6: HOW TO ADDRESS RISKS TO DIGITAL INCLUSION IN AFRICA Digital technologies (DTs) have the potential to drive the creation of more and better jobs in Africa. Africa’s jobs challenge is to create good, new jobs for the large number of young people entering the workforce and better jobs for existing workers. Greater adoption of improved technologies is a critical and under-emphasized requirement to meet this goal. Therefore, Africa’s technology challenge is to produce and promote better technologies and products that all Africans want and can afford to buy. New research provides robust new evidence that internet availability has led to job creation and poverty reduction across Africa, discusses challenges to the inclusive impacts of DTs in Africa, and proposes policy options for addressing these challenges: 1. Mounting evidence suggests that internet use has significant inclusive job impacts. In Nigeria and Tanzania, labor force participation increased by 3 and 8 percentage points, respectively, after 3 or more years of exposure to Internet availability. Similarly, poverty rates fell by 7 percentage points. Welfare impacts were higher among poorer and less-educated households, underlining the inclusive impact of the internet. 2. The key developmental challenge is the low productive use of DTs, which lags significantly relative to availability. Less than one-third of those with internet availability are using it across Africa. Similarly, enterprise digitalization is low. To address low productive use, enterprises and households need greater ability to pay on the supply side as well as (2) willingness to productively use these technologies on the demand side, through more attractive and appropriate DTs and skills development. 3. To address the simultaneous jobs and technology challenges, Africa needs more activist productive use policies to close the digital divide and boost adoption by individuals and enterprises: • Affordable availability policies on the supply side must focus more on implementation of good rules and go beyond the internet to include data infrastructure & regulation – with a greater focus on competition and regional integration. • Attractiveness & capabilities policies on the demand side require demonstrations of the productive use of digital and complementary technologies for households and microenterprises, innovation/entrepreneurship policies to generate and scale attractive apps for non-users, and capabilities support programs. • National strategies should be focused on promoting the productive use of technologies by households/ microenterprises and larger firms, with an emphasis on complementary factors such as access to credit, electricity, and urban-rural feeder roads – while managing downside risks, including the potential for increasing digital divides (especially displaced low-income people unable to adapt & adjust) and the potential for misuse by business (data protection, cybersecurity, and consumer protection) and government (surveillance, misinformation). <<< 57 Two sets of “productive use for jobs” policies: (1) ability to pay; (2) willingness to use new technologies Licensing & regulation of dominance Pro-competition regulation Access & sharing of essential infrastructure Policies to ensure ability to pay Affordability of internet Management of radio frequencies Cost-reducing regulation More effective SOEs & private participation Taxation Rationalization of sectoral surtaxes & fees Availability of internet More effective, innovative universal access/ & complementary Government interventions to service technologies (analog complement markets Climate adaptation & resilience infrastructure) Affordable availability of Enabling environment for Enablers for IXPs, data centers, and cloud data data infrastructure computing, including crossborder rules Reduction of barriers and support of drivers of Digital entrepreneurship entry & expansion of firms, national/regional willingness to use Attractiveness and Policies to elicit Generation of more attractive (skills- appropriate) capabilities for Technology and innovation DTs for productive use by firms productive use of digital Enablers of new DTs & trust; safeguards against & complementary Data policies & regulation business and government misuse technologies Addressing digital divides and supporting workers Social inclusion to adapt and adjust Source: Begazo, Dutz and Blimpo (2023) Digital Africa: Technological transformation for jobs. 2.1. Taxing digital infrastructure: mechanism to boost the efficiency of the sector or achieve specific sector goals.155 taxes and parafiscal fees Despite the importance of an efficient, fair, and simple African countries have adopted different approaches to tax fiscal framework for boosting digital transformation and digital infrastructure, resulting in the application of additional mitigating digital exclusion risks, current tax policies and tax taxes compared to other economic activities, the imposition administration challenge the sector. In terms of tax policy, of fees that are not proportionate to regulatory objectives, African countries’ rising public debt levels, particularly in the and a complex array of taxes and parafiscal fees. Most taxes aftermath of the Covid-19 crisis, call for increased revenue and fees154 are not aligned with tax principles and public mobilization to help governments to fund basic services administration principles and affect the economic governance and infrastructure. The relatively few operators, the degree of the digital infrastructure sector. In many countries, taxes of formality, and the ability to trace revenues in the telecom discriminate among economic activities and are higher for sector make the telecom industry an easier channel for tax digital connectivity; some are regressive, and many are collection to generate fiscal space. Besides the cost burden, difficult to administer. Parafiscal fees are not proportionate this might create a dependency relationship between the to the cost of providing the service or are not used as a telecom sector and the government and create incentives to 154. General taxes include monetary transfers from consumers and firms to the general government (direct and indirect taxes) and are not linked to a specific use, while parafiscal fees or contributions are linked to specific services provided by the government or specific objectives. According to the OECD classification, the term “taxes” is limited to compulsory, unrequited payments to the government or a supranational authority. “Non-tax” revenues are all other government revenues that are not classified as taxes. OECD (2020), Revenue Statistics 1965-2019: interpretative guide, p. 319. Available at: http://www.oecd.org/tax/tax- policy/oecd-classification-taxes- interpretative-guide.pdf. 155. Based on a high-level comparative assessment of key legislative, tax, and regulatory policy instruments in more than 20 African countries, the background note prepared for this chapter systematizes information on the current state of play of telecom taxes and parafiscal fees. Detailed information from these 20 countries is complemented with information for all countries in Africa (when available) and compared to other regions. 58 >>> ADDRESSING OLD ECONOMIC GOVERNANCE RISKS FOR DIGITALIZATION protect the major taxpayers and keep revenues unchanged or At the same time, issues with tax administration and operators’ rising instead of opening markets or boosting competition in behavior can undermine good governance of taxation and favor of consumers. fees policy in the sector and tilt government choices to more administrable taxes, although less efficient or equitable. In While telecom taxation helps government in plugging budget terms of tax administration, African countries face difficulties in spending deficits, excessive and multiple taxes and fees collecting direct taxes on income (e.g., personal income tax) for African telecom operators and users may impede digital and indirect taxes (e.g., VAT, sales tax).156 Lower average tax economy transformation, resulting in a loss of the digital revenues as a proportion of GDP in African countries compared economy’s economic and social benefits. Telecom taxation to OECD countries (Figure 61) enhance the need for African provides governments with income to compensate for budget countries to rely on taxes in a handful of large, formal-sector spending deficits. COVID-19 exacerbates the effect that corporations, including telecom operators. Particularly fragile developing countries’ tax-to-GDP ratios lag well behind those states (DRC, Nigeria) display low tax-to-GDP ratios. Several of wealthy countries, as well as falling commodity prices fragile states with low tax collections have been able to make and rising debt. Therefore, increasing domestic revenue up for the revenue shortfall by relying on nontax revenues from mobilization is especially a priority for low-income countries. extractive industries.157 Having a tax-to-GDP ratio of 13 percent Multiple taxes and fees in many African countries may have to 15 percent is linked to a greater possibility of rebuilding significant consequences for both companies providing the institutions to overcome fragility, which implies that some services, but especially the consumers for whom such taxes states should boost tax revenue by more than 30 percent to may account for a sizable portion of their income. Excessive cross the threshold.158 In Sub-Saharan Africa, just 6.3 percent taxation potentially raises the costs of digital delivery of of large taxpayers (corporations) account for 78 percent of products and services and makes them less accessible, total tax collections.159 However, cross-border profit-shifting by particularly for those living in rural areas or belonging to low- multinationals complicates the collection of profit-based taxes.160 income groups. Excessive taxation and fees exacerbate digital exclusion risks. Due to a lack of resources, African countries experience difficulties in effectively combatting profit-shifting, and one African countries should balance the trade-offs in tax policy policy response has been more reliance on indirect (excise- between raising revenues for short-term benefits and like) taxes. Lower performance in terms of tax administration encouraging faster and broader digitalization for longer-term may correlate with less efficient but more administrable benefits. Taxation of the digital economy enables African policy choices.161 These policy choices have their drawbacks. governments to efficiently mobilize domestic resources At the individual level, in the past, telecom services were for funding public goods such as law and order and public overwhelmingly purchased by wealthier taxpayers in infrastructure and to ensure an equitable business level developed countries and thus considered a luxury product playing field. However, policymakers and regulators need to be that is attractive to be taxed.162 However, this situation has careful not to negatively affect users, particularly the poorest changed radically, and now internet use is considered an in rural and remote areas, and the spread of digitalization enabler of inclusion and poverty reduction, as recent impact benefits. While these trade-offs are universal, especially in the studies for Nigeria and Tanzania show.163 Taxing economic aftermath of the COVID-19 crisis, new financial challenges activity through transactional or consumption taxes (VAT or of higher spending, fiscal constraints, and lower revenue excises) rather than individual income results in potential generation are more acute in Africa. harm to the less well-off and industries that are highly taxed. 156. In developing countries, the tax structure only represents 16.3 percent PIT, 14.6 percent CIT, and 31.5 percent VAT. In developed countries, the tax structure amounts to 24.2 percent, 8.9 percent CIT, and 20.0 percent VAT. See OECD (2018). G20 Policy Guide: Digitisation and informality: Harnessing digital financial inclusion for individuals and MSMEs in the informal economy, available at http://www.oecd.org/g20/G20-Policy-Guide-Digitisation-and-Informality.pdf. See also: M. Carnahan (2015), Taxation Challenges in Developing Countries Asia & the Pacific Policy Studies. 157. E. Crivelli and S. Gupta (2014), “Resource Blessing, Revenue Curse? Domestic Revenue Effort in Resource-Rich Countries”, IMF Working Paper 14/5, IMF, Washington DC; J.-F. Brun, G. Chambas, and M. Mansour (2015), “Tax Efforts of Developing countries: An Alternative Measure” in M. Boussichas and P. Guillaumont, Financial Sustainable Development, Paris, Economica. 158. M. Mansour and J.-L. Schneider (2019), “How to Design Tax Policy in Fragile States,” IMF, Fiscal Affairs Department, p. 4. Available at: https://www.imf.org/~/media/Files/ Publications/HowToNotes/HowToNote1904.ashx 159. ATAF (2019). African Tax Outlook 2019. Available at: https://events.ataftax.org/index.php?page=documents&func=view&document_id=49 160. The OECD/G20 Inclusive Framework on Base Erosion and profit shifting (BEPS) aims to ensure that profits are taxed where the economic activities that generate those profits are performed and where value is created. https://www.oecd.org/tax/beps/ 161. See, for instance, the Tax Administration Diagnostic Assessment Tool (TADAT) reports. https://www.tadat.org/home 162. T. Matheson, P. Petit, “Taxing telecommunications in developing countries,” International Tax and Public Finance (2021), Vol. 28, issue 1, No. 9. Available at: https://link. springer.com/content/pdf/10.1007/s10797-020-09621-6.pdf 163. Bahia et al, 2020; Bahia et al, 2021 <<< 59 >>> Figure 62: Tax-to-GDP ratio, 2018 40 34.3 35 32.1 29.1 27.8 30 25 20.4 19.4 20 18.3 17.4 17.1 16.8 16.7 16.5 14.6 14.1 13.1 15 11.8 11.4 10 7.5 6.3 5 0 a l da ca so o i n re pt na CD go sia ia a ar ) us ga aw RC nd c oo ny y oi er sc ri oc an Fa ha i To ni ne rit Eg al E Af ga Iv ig er Ke (D ga Tu or O w au M G d' Se a N am U h R M da in ep M ut e rk C ot a So .R M Bu C em ,D go on C Source: OECD tax database Finally, it is worth noting that other core reforms are needed to of economic footprint, Kenya’s mobile sector contributes a ensure that a more favorable taxation system delivers benefits significant amount of taxes and fees. While the mobile market to consumers. Boosting competition in digital infrastructure revenue accounts for 3 percent of Kenya’s GDP,164 the sector’s markets is essential to ensure that lower taxes are passed on tax and fee payments accounted for around 6.5 percent of to consumers and that tax incidence is not borne by the most the government’s total tax revenue165 As a result, the tax vulnerable. Furthermore, more efficient tax administration contribution of the mobile sector is 2.2 times its economic and more robust technical capacity to administer taxes in size. On the other hand, operators in Somalia appear to be the sector, including a better understanding of the business paying well below the practice in the region.166 IMF estimated model and enhanced monitoring to avoid profit shifting, that the telecom sector contribution amounted to 5.6 would be needed. percent of 2016 projected tax revenue.167 Lack of regulation further complicated by the weak institutional capacity and inconsistent reporting by telecom operators contribute to 2 .1 .1 . H O W I M P O RTA N T A R E practical challenges in calculating and collecting taxes and TA X E S A N D PA R A F I S C A L F E E S F O R fees on the telecom sector and in enforcing payment.168 G O V E R N M E N T S A N D O P E R AT O R S ? Increased revenue generation could require a more formal tax system and regulation that replace the de facto purely Taxes from digital infrastructure are important for overall tax voluntary telecom contributions to the public finances. collection, but there are cases where telecom operators’ contribution to fiscal revenues is less than expected. In terms 164. Oxford Economics (2018) cited in GSMA and EY (2020). 165. Kenyan National Bureau of Statistics, 2019. Available at: https://www.knbs.or.ke/?wpdmpro=statistical-abstract-2019 166. Federal Government of Somalia telecoms revenues (in USD) as a percent of estimated operator revenues. See, also: IMF Country Report No. 17/61 for Somalia (2017): https://centralbank.gov.so/wp-content/uploads/2020/08/Article-IV-Feb-2017.pdf 167. IMF Country Report No. 17/61 for Somalia (2017), p. 18. 168. World Bank (2018), Federal Republic of Somalia: Systematic Country Diagnostic, p. 36, available at: https://documents1.worldbank.org/curated/en/554051534791806400/ pdf/SOMALIA-SCD-08152018.pdf 60 >>> ADDRESSING OLD ECONOMIC GOVERNANCE RISKS FOR DIGITALIZATION In Sub-Saharan Africa, sector-specific taxes and fees account instance, the tax contribution of the mobile sector in Tanzania for a significant portion of the mobile sector revenue. In the was estimated at USD 404 million, representing 34 percent mobile telecom segment, Sub-Saharan Africa has the highest of total market revenue.172 In 2018, the Kenyan mobile average sector-specific taxes and fees169, and the mobile sector’s overall tax contribution was estimated at USD 954 sector in Sub-Saharan Africa was subject to higher taxes than million, accounting for 37 percent of total market revenue.173 any other region (Figure 63).170 In Sub-Saharan Africa, 60 In Côte d’Ivoire, the total tax contribution of the mobile sector percent of annual tax and fee revenue was generated from was estimated at USD 358 million in 2016, accounting for economy-wide taxes on operators and users and 40 percent 31 percent of the mobile sector’s total market revenue.174 In from sector-specific taxes and fees.171 Sector-specific taxation Guinea, tax contributions in 2020 were reportedly 64 percent can be so high that the mobile sector’s contribution to sector- of market revenues.175 These numbers are above the level of specific taxation is greater than that of general taxation. For other regions, including Sub-Saharan Africa (25 percent).176 >>> Figure 63: General and sector-specific taxes and fees, as a proportion of market revenue Latin-Amer ica 14% 4% Europe 17% 4% Asia Pacific 14% 10% Middle East and North Africa 14% 10% Sub-Saharan Africa 15% 10% 0% 5% 10% 15% 20% 25% 30% General taxes and fees Sector -specific taxes and fees Source: GSMA (2019), Rethinking mobile taxation to improve connectivity 169. ODI (2020), “How tax officials in lower-income countries can respond to the coronavirus pandemic,” Coronavirus Briefing Note, London: https://odi.org/en/publications/ how-tax-officials-in-lower-income-countries-can-respond-to-the-coronavirus-pandemic/ 170. GSMA (2019), “Rethinking mobile taxation to improve connectivity”, p. 14: https://www.gsma.com/publicpolicy/wp- content/uploads/2019/02/Rethinking-mobile-taxation- to-improve-connectivity_Feb19.pdf 171. GSMA (2019), “Rethinking mobile taxation to improve connectivity”, p. 15: https://www.gsma.com/publicpolicy/wp- content/uploads/2019/02/Rethinking-mobile-taxation- to-improve-connectivity_Feb19.pdf 172. GSMA (2020), “Tanzania: Driving social and economic value through mobile-sector tax reform”, p. 4: https://www.gsma.com/publicpolicy/wp-content/uploads/2021/04/ GSMA_Mobile_taxation_in_Tanzania_2021.pdf 173. GSMA (2020), “Mobile taxation in Kenya: Accelerating digital development,” p. 5: https://www.gsma.com/publicpolicy/wp- content/uploads/2020/03/GSMA_Mobile_ taxation_in_Kenya_2020.pdf 174. GSMA (2018), “Reforming mobile sector taxation in the Democratic Republic of the Congo: Enabling economic growth through a supportive tax system,” p. 6: https:// www.gsma.com/subsaharanafrica/wp-content/uploads/2018/11/GSMA_DRC- report_ENGLISH_72pp_WEB.pdf 175. As of May 2021, total mobile sector revenues were estimated at GNF 4,731 billion ($505 million) in 2020, which corresponded to 3.3 percent of Guinea’s GDP. The total fiscal contribution of the Guinean mobile sector is estimated at GNF 3,037 billion ($324 million) in 2020. This represents 64 percent of total mobile sector revenues. 176. GSMA (2020), “Tanzania: Driving social and economic value through mobile-sector tax reform”, p. 4: https://www.gsma.com/publicpolicy/wp-content/uploads/2021/04/ GSMA_Mobile_taxation_in_Tanzania_2021.pdf <<< 61 A high dependency on the sector for taxation revenue may costs, or distorting the level playing field in digital infrastructure weaken government incentives to change the status quo and markets. However, taxes and parafiscal fees in Africa are not reform taxation and parafiscal fees policy or even support generally aligned with these principles. other sectoral reforms that can change market dynamics and potentially generate more revenues from more users and In Africa, governments and regulators subject telecom operators in the future. The importance of sector-specific operators and users to a wide variety of taxes and parafiscal taxes further highlights weak economic governance of taxation fees, which are often specific to the digital infrastructure and fees policy since it signals rent extraction from the sector sector (Table 4). African telecom operators and users are compared to other sectors. subject to general taxes, such as corporate income tax (CIT) and value-added tax (VAT), but in some countries at higher rates than other sectors. Some taxes also apply exclusively 2 .1 . 2 . A R E TA X E S A N D PA R A F I S C A L to the telecom sector. Sector-specific taxes encompass ad FEES IN LINE WITH GOOD valorem (based on turnover) and specific (nominal amount PRACTICE? based on some activity measures such as minutes, data, and SMS) excise taxes or a differentiated VAT or sales tax Good economic governance of the sector involves following on handsets and other devices, activations and connections overarching fundamental tax principles of sound fiscal policy and use of telecommunications services. Sub-Saharan Africa design and principles for public administration that deliver a has the highest average sector-specific taxes and fees in balanced system of transfers from users and infrastructure the mobile telecom segment.177 In some countries, elevated service providers to the government to fund necessary import duties on devices and telecom equipment are also public investments and services. Fundamental principles present. Additionally, regulatory fees include compensation for assessing Africa’s taxation system are its efficiency for operational licenses or annual regulatory contributions, (i.e., minimum impact on economic behavior), equity (i.e., use of state resources such as spectrum, and fees for other proportional tax burden to or increasing with income), and regulatory services (such as numbering and equipment simplicity (i.e., low administrative and compliance costs). In homologation). In most cases, there is no clear policy on the case of regulatory fees, public administration principles setting regulatory fees or transparency on all the applicable call for fees that are proportionate to the resources needed fees. Mandatory contributions for operators also apply, for to regulate and that consider effects on market functioning, example, to universal service funds to finance universal that is, to avoid generating entry barriers, raising operational access and service projects. >>> Table 4: Classification of taxes and fees for digital infrastructure Applied directly on telecom consumers Applied on telecom operators (and indirectly on consumers) Taxes on handsets and other devices Corporate Income Tax (CIT) and additional CIT (on profits) Customs duties • VAT at normal rates or higher VAT on • Handsets, infrastructure inputs, network equipment handsets Regulatory fees for the sector • Handset excise tax • General regulatory fees (on revenues) • Some countries grant VAT exemptions • Spectrum fees (one-off/variable) Taxes on activations or connections • Licensing/authorization fees (one-off/variable) • VAT at normal rates • Other regulatory fees (numbering fees, homologation and registration • Specific or ad valorem tax on SIM, initial of equipment fees) connection Sector contributions Taxes on use • Universal service contributions (revenue) • VAT at normal rates and higher VAT on Other taxes with earmarked revenues: education tax, health tax, cultural usage tax, environment tax, vocational training tax • Use excise tax (specific or ad valorem) Local taxes Source: The data collection on taxes and parafiscal fees of the telecom sector is sourced from various sources including Bowmans’ data, GSMA’s country reports, Deloitte and PwC Tax Summaries & reports, OECD’s and IBFD’s tax databases and from desktop research (e.g., countries tax legislations). Custom duties were collected from the World Trade Organisation. 177. ODI (2020), “How tax officials in lower-income countries can respond to the coronavirus pandemic,” Coronavirus Briefing Note, London, www.odi.org/sites/odi.org.uk/ files/resource-documents/200304_tax_corona_lt_ec_0.pdf. 62 >>> ADDRESSING OLD ECONOMIC GOVERNANCE RISKS FOR DIGITALIZATION TAXES AND FEES ON DIGITAL INFRASTRUCTURE operators in Angola179, Guinea180, and Tunisia181 are subject to a OPERATORS 35 percent CIT rate, which is higher compared to the standard rate of 25 percent (or even 15 percent in Tunisia) for other In at least four countries, operators are subject to higher CIT companies.182 Côte d’Ivoire183 subjects mobile operators to a 30 rates (an additional 5 percent to 10 percent) than in other percent CIT rate, which is higher than the standard 25 percent sectors. General CIT rates applicable across sectors vary from CIT rate applied to other sectors (Figure 64). Note that Zambia 15 percent in Mauritius to 40 percent in Zambia178. In some levies a higher CIT rate of 40 percent for income over 9,069 African countries, telecom operators face higher CIT rates USD (ZMW 250,000) – but this higher CIT rate is not limited to that typically apply to other industries. For example, telecom telecom operators – instead of the general 35 percent. >>> Figure 64: Rates of corporate income tax on mobile operators, 2021 45% 40% 35% 40% 30% 10% 10% 33% 30% 30% 5% 30% 30% 30% 30% 30% 30% 25% 20% 28% 28% 28% 27% 27% 25% 25% 25% 25% 20% 23% 20% 15% 15% 15% 10% 5% 0% a am i a co na a na pt ) ep ni n a a re a ia ia da on go a s r ya al G o C c ca ne bi nd iu bi s ol s gy er oc an eg oi fri R en ha Fa an ro ni To e ng ri t m am as ig ui Iv ga (D A B E en nz or Tu e G au Za K w G d' N ag A th U M R Ta S M e ou ad ki ot C .R ur S M C B em ,D go on C General CIT Rate Additional CIT rate for mobile operators Source: World Bank, Digital Infrastructure Taxes and Fees database Setting higher CIT rates for telecommunication can distort investments, lower output and higher prices, creating incentives allocation of resources across industries. Imposing CIT rates for oligopolistic collusion at the country level to raise prices for telecom operators above the average headline rate in the by limiting output.186 Imposing a CIT surcharge increases the Africa region (of 26.5 percent184 in 2021) is commonly justified incentive for multinational companies to transfer revenues by because the telecom industry is perceived to be profitable across borders by more proactive tax planning aimed at shifting above average, but this instrument is distortive and does not profits from higher to lower-tax jurisdictions.187 The OECD/ address root causes of the concern.185 However, higher CIT G20 Inclusive Framework on Base Erosion and Profit Shifting rates could result in inefficient cross-sectoral allocation of (BEPS) aims to strengthen poor nations’ ability to combat tax 178. The CIT rate is 40 percent for income over USD 11324,88 (ZMW 250.000). 179. Presidential Legislative Decree 3/12 of 16 March 2012. 180. Article 229 of the GTC. 181. Law 2002-1 of 15 January 2001. 182. Although a higher CIT rate applies on telecom companies, this is not considered as a mobile-specific tax since the same rate applies on banking and insurance activities (Angola), banks and insurance companies and companies importing, storing and distributing petroleum products (Guinea), oil service companies, banks, financial institutions, car dealers, large commercial enterprises and franchisees of foreign brands (Tunisia). 183. Article 51 of the GTC (30 percent for telecommunications and IT companies) and article 64 of the GTC (standard 25 percent). 184. Author’s own calculation of 21 selected African countries. 185. T. Matheson, P. Petit, “Taxing telecommunications in developing countries”, International Tax and Public Finance (2021), Vol. 28, issue 1, No. 9, p. 264. 186. C. Davidson, and L. Martin, General Equilibrium Tax Incidence under Imperfect Competition: A Quantity-setting Supergame Analysis, Journal of Political Economy 93(6), 1212-1223 (1985). 187. By the payment of e.g., excessive royalties from firm-specific technology, excessive management or service fees, over-invoicing for inputs through offshore cost centers (particularly capital goods), and excessive leverage (thin capitalization). See R. Schatan, Tax Minimizing Strategies and the Arm’s-Length Principle, Tax Notes International p. 121-126 (Jan. 9, 2012). <<< 63 base erosion. On 8 October 2021, 136 nations agreed to a apparatus may be subject to import duties and other surcharges, global two-pillar approach on a new revised international tax which may potentially be higher than other goods and services. framework to address digital challenges for large businesses Based on International Trade Centre’s data for the period 2014 (not just those in the ‘digital economy’).188 Besides digital – 2019 for 37 low-income and middle-income Sub-Saharan service providers (like multinational digital platforms), digital African countries, average import duties on base stations infrastructure operators would be subject to a minimum are 2.4 percent higher than the average country of all other global corporate income tax of 15 percent (Pillar II). Although regions and income groups (excluding African countries) and implementation would be challenging, and some countries 2.5 percent higher on switches and transmission equipment.189 have not agreed to the approach (notably, Nigeria and Kenya), Countries apply different import duty rates within the sector. For this would be an instrument to improve the taxation system. example, Ghana subjects imported telecom network equipment Finally, the most effective instrument to address concerns to customs duty (e.g., base stations are subject to a 10 percent related to the high profitability of certain operators due to duty), while customs and VAT exemptions apply to machinery weak competition would be to use sectoral rules to boost and apparatus used in other industries (e.g., agriculture, mining, competition rather than imposing higher CIT rates on the and transportation). Angola imposes a 2 percent levy on base digital infrastructure (telecom) sector. stations, but import duties or other specific taxes do not apply to infrastructure inputs, notably antennas, towers, and network African countries levy customs duties for the import of mobile equipment. Additional VAT rates may also apply to telecom equipment goods from 0 percent to 10 percent, although, infrastructure. Telecom infrastructure in Egypt is subject to an internationally, goods for capital investments and productive additional 5 percent VAT rate to “develop State resources and inputs are generally subject to low or no import duties. Network to aim for social justice”.190 equipment such as antennas, base stations, and communication >>> Figure 65: Overview of customs duties on communication apparatus and base stations 12% 10% 8% 6% 4% 2% 0% na ia em am o ep on Se a ia ut ga l go au r d' ) la M us ad nya co Ta ica n sia re a t a ca a yp RC s bi nd d ni an er ne go oi ha Fa . R ero i To oc an ni r rit am as e Bu Be Eg ig Ke Af Iv (D nz ga n ui An Tu or G w a ag N G G U h R in M e rk C ot So M C ,D go on C Base stations (HS 851761) Communica tion appa ra tus (HS 851762) Source: Tariff lines of the WTO Customs duties on base stations and communication apparatus were collected from the World Trade Organisation (WTO) website. These refer to the Harmonised System (HS) code 851761 (‘Base stations for transmission or reception of voice, images or other data, incl. apparatus for communication in a wired/wireless network (such as a local/wide area network)’), HS code 851762 (‘Communication apparatus (excluding telephone sets or base stations); machines for the reception, conversion and transmission or regeneration of voice, images or other data, including switching and routing apparatus’). 188. OECD/G20 Base Erosion and Profit Shifting Project, Addressing the tax challenges arising from the digitalization of the economy, 8 October 2021. Available at: https:// www.oecd.org/tax/beps/statement-on-a-two-pillar-solution-to-address-the-tax- challenges-arising-from-the-digitalisation-of-the-economy-october-2021.pdf 189. Latest year available at Market Access Map, International Trade Centre (ITC), www.macpmap.org. 190. Article 2 of the Egyptian Added Value Tax Law No. 67/2016. 64 >>> ADDRESSING OLD ECONOMIC GOVERNANCE RISKS FOR DIGITALIZATION Handsets and SIM cards are also subject to customs duties, are relatively stable with a general range between 10 percent adding up to the cost of devices. Operators initially bear (including Angola, Benin, Burkina Faso, Guinea, Nigeria, these expenses but are passed on to end users. Based on Senegal, Togo, and Uganda) and 0 percent (Egypt, Kenya, ITC data for the period 2014 – 2019 for 37 low-income and Mauritius, South Africa, Tanzania, and Uganda). Import duties middle-income Sub-Saharan African countries (Table 5), on SIM cards range from 0 percent to 30 percent (Figure 65). average import duties on handsets are 1.6 percent higher Import duties on SIM cards are the highest in Cameroon (30 than the average country of all other regions and income percent), while Egypt and Tunisia do not levy any import duties groups (excluding African countries). Import duties on devices on SIM cards. >>> Table 5: Average Import Duty for Handsets by Income Group, Sub-Saharan Africa vs. Other Regions (2017 – 2019) Income Group Region Upper middle Lower middle High income Low income Total Average 2017 – 2019 average income income Sub-Saharan Africa 0.0% 1.0% 6.8% 6.8% 5.8% All Other Regions 2.3% 6.7% 4.7% 8.3% 4.2% Grand Total 2.2% 6.1% 5.5% 7.0% 4.5% Note: Considering 340 observations for the selected period. Source: ITC, Import duties for HS Product Code 851712 (Telephones for cellular networks or for other wireless networks) >>> Figure 66: Overview of customs duties on users’ goods for mobile services (phones and SIM cards) 35% 30% 25% 20% 15% 10% 5% 0% em am o ep on n re t la na a a al go ia sia ia ad nya au r Se a Ta ca d' ) s co yp ca a RC s ni iu bi nd d go oi an So ne g er ne Fa ro ha To ri oc an ni rit Be Eg am as Iv ig Ke Af ga (D nz An e ui Tu or G w a ag N G G U h R in M M e ut rk C ot .R M Bu C ,D go on C Mob ile phones (851712) SIM cards (HS 852321) Customs duties on telephones for cellular networks and SIM cards were collected from the World Trade Organisation (WTO) website for the latest data available in August 2021. These refer to the HS Code 851712 (‘Telephones for cellular networks’); and HS Code 852321 (‘Cards incorporating a magnetic stripe’) <<< 65 In some countries, operators are subject to socio-economic fees, these costs may be indirectly passed on to end users. taxes aimed at funding certain activities (i.e., education, culture, Excessive regulatory fees and taxes on revenues and turnover sports, etc.). These taxes include education tax, health tax, should be avoided. For example, Guinea imposes a Telephone cultural tax, environment tax, vocational training tax, gaming Network Access Tax (TARTEL) of 3 percent on the turnover tax, and advertisement tax. In some countries, socio-economic of telecom operators.196 In Senegal, telecom operators are taxes are targeted at the telecom sector. For example, Côte subject to the tax for access to or use of the public telecom d’Ivoire imposes a culture contribution of 0.2 percent on telecom network (RUTEL), which is levied at the rate of 5 percent of companies and companies carrying out mobile transfers to the turnover, excluding taxes, on the amount paid by users to promote cultural activities.191 Telecom operators in Guinea the operator.197 In Côte d’Ivoire, a 5 percent tax on telecom contribute 1 percent of net annual revenue to a Research and enterprises198 applies to the company’s monthly turnover Development Fund.192 In Egypt, all telecom licensees contribute exclusive of VAT.199 A 2 percent surtax for the development 0.5 percent of their annual gross income in scientific research, of new technologies in rural zones applies on a monthly education, and training programs in ICT. Despite the economic turnover, excluding VAT, from prepaid and post-paid telecom and social benefits of socio-economic contributions, multiple services.200 In Mali, telecom companies are also subject to a taxation is not healthy; it raises operational costs and creates tax on telecom as of January 1, 2013, at the rate of 5 percent space for governance issues.193 on turnover derived from phone calls.201 Besides taxes, as in other regulated industries, African Telecom operators may be subject to initial licensing fees for telecom operators pay parafiscal fees. These include general obtaining the licenses and annual license fees. Many African regulatory fees (to cover the cost of regulatory activities and countries, including Uganda, Kenya, Tanzania, Madagascar, provide revenues to ensure the financial independence of the DRC, South Africa, and Egypt, charge annual licensing regulator), licensing and authorization fees, spectrum fees, fees as a percentage (ranging between 0.4 percent and 2 and other regulatory fees (e.g., numbering and homologation percent) of the telecom operator’s annual gross turnover or of equipment), in addition to Universal Service Fund (USF) revenue. For instance, the licensing fees in Madagascar and contributions. The design of these fees does not always DRC represent an annual regulatory payment for the license support healthy and competitive markets. arrangement of 2 percent on revenues. In Tanzania, annual licensing fees of 1 percent of the gross annual turnover apply. General regulatory fees, also known as administrative or South-Africa applies variable annual licensing fees between sectorial fees, on revenue and turnover, are present in 0.15 percent and 0.35 percent. Only a few countries (e.g., most developing economies. General regulatory fees aim to Rwanda and Togo) levy a lump sum for annual licensing guarantee the financial independence of the regulator and fees. In Rwanda, operators pay a one-off annual licensing compensate the regulator for its costs of regulation.195 Fees fee (between USD 184,000 and USD 275,000). In Togo, based on revenues rather than profits require telecom operators licensees are subject to an annual licensing fee of USD to pay the same amount, irrespective of whether telecom 37,000. In Somalia, telecom operators are required to pay operators keep the profits or invest in new infrastructure and license fees at a flat administration fee of USD 50,000 per mobile services. Although telecom operators bear regulatory year for 20 years.202 Some countries still restrict the number 191. See Art. 1129 of Code Général des Impôts (http://www.dgi.cgici.com/indexs.htm). 192. World Bank (2019), Guinea: Opportunities for Enhanced Domestic Revenue Mobilization: Value-Added Tax and Excise Taxes, p. 57. Available at: http://documents1. worldbank.org/curated/pt/472021561614678154/pdf/Guinea-Opportunities-for- Enhanced-Domestic-Revenue-Mobilization-Value-Added-Tax-and-Excise-Taxes.pdf 193 . M. Oseni, Multiple Taxation as a Bane of Business Development in Nigeria, p. 123. Available at: https://pdfs.semanticscholar.org/3e0f/6cf138393b766fc6c5d7c24dcd3a58c493a3.pdf. 194. Given the economic nature of setting spectrum fees to allow for the efficient use of spectrum resources and foster competition in the market. The analysis of the level and method for charging spectrum fees is out of the scope of this report. For a general discussion on good practices on spectrum management, see ITU & World Bank (2021), Digital Regulation Handbook. 195. C. Blackman and L. Srivastav (2011). Telecommunications Regulation Handbook: Tenth Anniversary Edition. World Bank and the International Telecommunication Union, Washington, DC. https://openknowledge.worldbank.org/handle/10986/13278 License: CC BY 3.0 IGO. 196. Finance Law 2015. GSMA, Reforming mobile sector taxation in Guinea: Unlocking socio-economic gains from mobile connectivity, 2018, p. 28. https://www.gsma.com/ publicpolicy/wp-content/uploads/2018/05/GSMA-Guinea-taxation-report- EN.pdf 197. Law no. 2010-14 of 23 June 2010. This tax is imposed by Law 2008-46 of 3 September 2008, and the rate was increased from 2 percent by Article 20 of Law no. 2010-14 of 23 June 2010. 198. I.e., tax on telecommunications, IT and communication enterprises (taxe sur les entreprises de télécommunication et des technologies de l’information et de la communication). 199. Article 1130 GTC. 200. Article 1127 of the GTC (i.e., taxe pour le développement des nouvelles technologies en zones rurales). 201. Article 253 AD of the GTC (i.e., taxe sur l’accès au réseau des télécommunications ouvert au public; TARTOP). Internet services and sales of mobile phones and telecom equipment are exempt from this tax. IBFD database, 2021. 202. P. Lange (2015), Telecommunications Contribution to Public Finance in Somalia, ICT Regulatory Technical Assistance. 66 >>> ADDRESSING OLD ECONOMIC GOVERNANCE RISKS FOR DIGITALIZATION of operators and subject entry and licensing fees to results allocation and use of USF collections including annual audits from auctions (e.g., Senegal). High fees, especially if based and publication of annual reports are essential for greater on revenues, can potentially create regulatory barriers to entry effectiveness205, as well as increased capacity to design and affect operators with lower margins. A lump-sum fee that is appropriate projects and implement them. If USF funds proportionate to the resources needed to monitor and regulate remain unused, USF contributions become an additional tax licensees would be the most advisable practice. This would on digital infrastructure. also be simpler to facilitate compliance. The lump-sum fees could be adjusted periodically based on a transparent formula. Other regulatory fees may add to the complexity of fees structure and increase compliance costs. Several African Operators in many African countries contribute a portion countries impose additional regulatory fees. The two most of their revenues to a universal service fund (USF) or their common are numbering resources fees,206 and homologation equivalent to finance the development of networks in and registration of equipment fees. Numbering fees generally unconnected areas and support the adoption and use of apply due to the relative scarcity of numbering resources and digital technologies. USF levies can vary between 0.3 percent to facilitate effective control and supervision of the numbers (Tanzania) to 5 percent (Tunisia) of operators’ revenue. For due to the rising number of electronic communication users example, operators in Guinea pay a USF contribution of 1.5 and services. Homologation fees are also implemented by percent of net annual revenue into a fund intended to support several countries to protect users from counterfeit mobile expanded coverage. In Egypt, telecom licensees are not devices.207 Numbering fees are observed across various Sub- directly responsible for contributing to the USF but instead Saharan African countries.208 In 2018, the average numbering contribute indirectly through license fees, which form part of fee within the Sub-Saharan African countries was USD the National Telecom Regulatory Authority (NTRA) budget.203 0.26 per assigned/booked phone number.209 In developed Other countries, such as The Gambia or Angola,204 do not countries, charges for numbering are not very common.210 The have any USF. Even if there is an USF in place in African DRC has introduced a registration levy consisting of an annual countries, the funds do not always sufficiently support greater payment of USD 1 for 2G handsets and USD 7 for 3G and 4G connectivity. In some countries, financial reporting amongst handsets.211 The DRC also imposes a new monthly tax (RAM the existing funds remains absent. For instance, all telecom – Registre des Appareils Mobiles) on all mobile devices on operators in Uganda licensed by the UCC are required to Congolese territory “ranging from the equivalent of US$0.17 contribute 2 percent of their annual gross revenue to the for 2G devices to more than US$1.17 for 3G and 4G devices USF for the purpose of developing rural communications, over a six-month period.” According to the government, this information and communication technology. But there is no tax would allow them to “limit the market of counterfeit mobile publicly available information on how the UCC utilizes the devices, combat mobile device theft, and improve the quality of funds or how much it has collected in the last three years. the mobile phone network by blocking non-compliant devices The operation of USF can help bridge the digital divide, but with international standards”.212 The RAM has been scrapped more effective use is needed. More transparency on the for second-generation (2G) devices in October 2021. 203. Article 9 of Egyptian Telecoms Laws. 204. Previous to March 2020, network operators and public use electronic communications service providers were subject to 1 percent of gross revenues contributions to the USF. 205. For more details on the governance of USF in selected Africa countries, see World Bank (2021) Regulatory Watch Initiative Phase 2. 206. Overall, there are three main categories of numbering fees: a) standard telephone numbers E.164 (for the subscribers directly connected to the operator); b) carrier selection code (to select the operator); and c) signaling point codes (for interconnection with other networks at national (NSPC) and international (ISPC) level). Available at: https://ec.europa.eu/digital-single-market/en/news/5th-report-implementation-telecommunications-regulatory- package-1999 207. GSMA (2019c), Mobile Policy Handbook: An insider’s guide to the issues, available at https://www.gsma.com/publicpolicy/mobilepolicyhandbook/wp-content/ uploads/2019/01/MPH7_ENG_web_spreads.pdf. 208. GSMA (2019), Rethinking mobile taxation to improve connectivity, available at https://www.gsma.com/publicpolicy/resources/rethinking-mobile-taxation-to-improve- connectivity. 209. Rota-Graziosi, G. and F. Sawadogo (2020), The tax burden on mobile network operators in Africa, Foundation pour les Études et Recherches sur le Développement International; hal-03109370. 210. Antelope Consulting (2018), Discussion on the numbering regulation for Somalia, available at http://public.antelopeweb.fmail.co.uk/publications/Discussion percent20on percent20the percent20numbering percent20regulation percent20for percent20Somalia.pdf. 211. Decree no. 20/005 of 9 March 2020 (Décret n° 20/005 du 09 mars 2020 modifiait et complétant le Décret n° 012/15 du 20 février 2012 fixant les modalités de calcul et les taux des revenus des prestations de l’Autorité de Régulation de la Poste et des Télécommunications, « A.R.P.T.C. » en sigle). See, for instance: https://www.financialafrik. com/2020/04/28/rdc-les-appareils- mobiles-desormais-taxes-de-1-a-7-dollars-usd/#:~:text=C’est percent20le percent20Journal percent20officiel,la percent20Poste percent20et percent20des percent20T percentC3 percentA9l percentC3 percentA9communications; https://deskeco.com/2020/04/27/rdc-la-nouvelle-taxe-de-larptc-sur- la-certification-des-smartphones-vient-frapper-le; https://www.digitalbusiness.africa/rd-congo-un-decret-introduit-une-taxe-de-1-a-7-dollars-sur-les-telephones-mobiles/ 212. P. Baraka (2020), “New tax on mobile devices threatens digital inclusion in the Democratic Republic of Congo”, available at: https://globalvoices.org/2020/12/30/new-tax- on-mobile-devices-threatens-digital-inclusion-in-the-democratic-republic-of- congo/ <<< 67 TAXES AND FEES ON TELECOM USERS rates: 12 percent on airtime217 and value-added services.218 Starting July 1, 2021, a 12 percent excise duty on mobile Surcharges to general VAT on handsets and SIM cards data replaced the social media tax. Consequently, the total increase the cost of device acquisition and accessing mobile tax on internet use is 30 percent after factoring in the existing services. In Sudan, an increased 30 percent VAT rate (as 18 percent VAT.219 In Madagascar, excise duty of 8 percent opposed to the regular 17 percent) applies to the sales of (previously 10 percent in 2020) applies to telecom services goods from telecom companies.213 Users in some countries such as internet and data services, placing and receiving calls are subject to sector-specific taxes and fees for the activation and text message services, and mobile money services.220 and connection (e.g., activation fees for SIM cards and Mobile phone services are subject to a 10 percent excise duty numbering taxes). While these types of charges are rarely in Gambia. In Kenya, airtime and telephone services (SMS, observed internationally, our research illustrates that some voice, and mobile data services) are subject to excise duty African countries levy such charges. These are usually fixed at 20 percent of their excisable value.221 Rwanda features an amounts, which disproportionately affect people with lower excise duty of 10 percent on telephone communications.222 incomes, discouraging the adoption of digital connectivity. Telecom services, such as the provision of data and telephone Removing these activation and connection taxes can help calls, are generally subject to the general rate of VAT, ranging lower affordability barriers, encouraging more people to enter from 12.5 percent (Ghana) to 19.25 percent (Cameroon). the mobile market and use digital services, thus expanding the Ghana imposes a 5 percent (previously 9 percent) government’s tax base. communications service tax on charges payable to users of an electronic service, which includes placing and receiving After the purchase of a handset and paying activation fees to voice calls, SMS/text messages, and internet/data services.223 gain access to a telecom network, many African countries levy Burundi levies a tax on local calls at the rate of BIF 52 per sector-specific taxes (including excise duties and other taxes) minute (approx. USD 0.026).224 In Côte d‘Ivoire, a specific tax on the use of connectivity services such as calls, SMS, and on phone calls, information technologies, and communication data. Zambia charges the highest excise duty at 17.5 percent services is levied at a rate of 3 percent on the VAT exclusive to telecom services such as internet and data services, placing invoice amount in respect of communication services provided and receiving calls and text message services, and mobile by mobile companies and internet service providers.225 money services. In Tanzania, mobile services, including calls, Mauritius levies 10 cents per SMS on every message that is SMS, and data are subject to an airtime excise tax of 17 percent, sent by an operator.226 Benin levies a consumption contribution in addition to the 18 percent VAT rate.214 Algeria imposes a tax on all electronic communication services (voice, SMS, and on prepaid mobile phone services of 7 percent (previously 5 internet) at 5 percent of the service price.227 In Guinea, the tax percent).215 In Cameroon, an excise duty of 2 percent applies on telephone consumption applies to telephone usage at the to mobile telephone communications and internet services. In rate of 1 GNF (USD 0.0001) per second on telephone calls the DRC, 10 percent excise duties apply to mobile services in (outgoing and incoming), 10 GNF (USD 0.001) per connection addition to the standard VAT rate.216 In Uganda, excise duty for SMS messages, and 5 percent of the price of any internet is chargeable on telecom goods and services at the following services.228 The tax of 1 GNF per second raised the average 213. Section 14(2) of the VAT Act (since 1 Jan. 2012). 214. Excise (Management and Tariff) Act, Sec. 2(1) and 124(4). 215. Article 76 of Finance Law 2017. 216. Ordinance-Law No. 007/2012 of 21 September 2012 on the Excise Code, and Ordinance Law No. 011/2012 of 21 September 2012 establishing a new Tariff of Import Duties and Taxes. 217. 12 percent for fiscal year 2021/2022 (previously 20 percent). Value-added services include internet. See also: https://www.ucc.co.ug/wp-content/uploads/2021/09/2Q21- MARKET-PERFOMANCE-REPORT-compressed.pdf 218. Section 2 of the Excise Tariff Act, amended by the Excise Duty (Amendment) Bill 2021. 219. https://qz.com/africa/2028653/uganda-replaces-ott-social-media-tax-with-tax-on-internet-bundles/ 220. Article 03.01.01 with its appendix of the General Tax Code. See also: https://www.commsupdate.com/articles/2020/03/09/telecoms-excise-duty-restored-to-10-in- madagascars-2020-finance- law/#:~:text=The percent20excise percent20duty percent20levied percent20on,burden percent20faced percent20by percent20the percent20sector. 221. Increased from 15 percent to 20 percent, effective 1 July 2021 (Finance Act 2021). Sec 32(b)(i) Kenya’s Finance Act 10 of 2018. 222. Article 4 on Products and corresponding rates of the 2019 Law Establishing the Excise Duty. 223. Introduced by the communications Service Tax (Amendment) Act 2019. 224. Budget Law 2018/19 (taxe spécifique de téléphonie mobile sur le trafic national) (article 66). 225. Article 1141 of the GTC (i.e., taxe sur les communication téléphoniques et technologies de l’information et de la communication). See also IBFD database, 2021. 226. Value Added Tax Act, Sec. 506. Available at: https://www.mra.mu/download/VATAct.pdf 227. Decree No. 2018-341 of July 25, 2018. Available at: https://sgg.gouv.bj/doc/decret-2018-341/ 228. http://documents1.worldbank.org/curated/pt/472021561614678154/pdf/Guinea-Opportunities-for-Enhanced-Domestic- Revenue-Mobilization-Value-Added-Tax-and- Excise-Taxes.pdf 68 >>> ADDRESSING OLD ECONOMIC GOVERNANCE RISKS FOR DIGITALIZATION price per minute by 14.2 percent, resulting in an immediate government to include texting in the excise base in late 2015, 16.1 percent fall in voice traffic.229 The decline in voice traffic resulting in an even steeper drop in SMS traffic. was offset by an increase in SMS traffic, which prompted the >>> Figure 67: Combined tax rates on use of mobile services in 2021 35% 30% 25% 20% 15% 10% 5% 0% na a a ut ga l go ia Se a Ta i c a d' ) M us em am o co ep on n re t la U ia ia ad nya au r yp RC ca bi nd s d ni er s go oi an ha i To an oc Fa ro r rit am ni So ne Be Eg as ig Af ga Iv (D Ke nz An e Tu or G w N a ag G h R in M e rk C ot .R M Bu C ,D go VAT on usage Mobile-specific tax on usage on C Source: Authors’ own calculation using Telecom Taxes and Fees database and national legislations in 2021 International incoming calls may also be subject to additional of international connectivity, creating obstacles to regional and sector-specific taxes in the form of ‘surtaxes on incoming international trade for local and regional businesses. Excises international traffic’ (SIIT). Among the 20 countries included in on international calls could also trigger a shift to untaxed the taxation analysis for this report, incoming calls are subject platforms, both legal and illegal.232 Cross-border callers have to fees ranging between USD 0.08 per minute and USD been particularly inclined to use voice-over Internet protocol 0.19 per minute. According to Deloitte and GSMA’s report (VoIP) to avoid these taxes. An illegal alternative is the usage on surtaxes on international incoming traffic in Africa, the tax of ‘SIM boxes’, which employ local SIM cards to convert in the countries where SIIT is applied has caused the price international calls made over the Internet into local phone of terminating international incoming calls to increase by an calls.233 There are countries that are considering taxes on average of 97 percent.230 Kenya, Rwanda, Burundi, Uganda, international calls or fees charged for the use of international and South Sudan agreed in May 2014 to waive the SIIT for clearinghouses that monitor international traffic (Begazo, Dutz calls originating in these countries as part of the agreement and Blimpo 2023). for the One Network Area, but Burundi reimplemented the SIIT in 2019. ECOWAS countries are also subject to policies to eliminate such fees, but with limited implementation. Benin TAXES AND FEES ON USERS OF DIGITAL SERVICES stopped applying the specific tax, which was 53 CFA (approx. USD 0.098)/min in 2016.231 Surtaxes on incoming international A trend in several African countries is taxing the usage of traffic may impact users in the African region and raise the cost certain digital media services, commonly referred to as 229. T. Matheson, P. Petit, “Taxing telecommunications in developing countries”, International Tax and Public Finance (2021), Vol. 28, issue 1, No. 9, p. 267. 230. GSMA (2014), “Surtaxes on International Incoming Traffic in Africa”, available at https://www.gsma.com/publicpolicy/wp- content/uploads/2012/03/Surtaxes_on_ International_Incoming_Traffic_in_Africa_FULL-REPORT_WEB.pdf 231. Regulatory Watch Initiative (RWI) phase 2 (March 2021), “Thorough legal, regulatory and competitive analyses of issues related to licensing, OTTs, international gateways, spectrum management and regulatory governance”, p. 103. 232. T. Matheson, P. Petit, “Taxing telecommunications in developing countries”, International Tax and Public Finance (2021), Vol. 28, issue 1, No. 9, p. 269. 233. While estimating the magnitude of the illegal market is difficult, the GSMA (2014) reported that 10 percent of international calls in Ghana were rerouted via illegal SIM boxes. As a result, authorities have employed sophisticated tools to track SIM boxes and their owners. GSMA (2014), Surtaxes on International Incoming Traffic in Africa, available at https://www.gsma.com/publicpolicy/wp- content/uploads/2012/03/Surtaxes_on_International_Incoming_Traffic_in_Africa_FULL-REPORT_WEB.pdf <<< 69 over-the-top (OTT) services. Online social networking and may have a greater impact on poor consumers because, communication services provided on online platforms have due to price discrimination, the fees are likely to be higher in been singled out as taxation targets. For example, on July proportion when the transaction amount is lower. Transaction 1, 2018, Uganda introduced a social media tax of UGX taxes on the underlying amount are gaining importance in 200 (USD 0.055) per day or UGX 1,400 (USD 0.38) for a Sub-Saharan African countries, with the recent approval of week on July 1, 2018, accessing digital platforms, including the e-levy in Ghana.240 In Uganda, a digital financial services WhatsApp, Facebook, and Twitter.234 The OTT tax in Uganda tax went into effect in July 2018, which was the first time a was intended to reduce the time citizens spent online and transaction tax (at an initial rate of 1 percent on mobile supposedly curb the spread of false information. Interestingly, money deposits, withdrawals, transfers, and payments the OTT tax only generated 0.27 percent of overall revenues between persons) had been enacted.241 It was reported in fiscal year 2019,235 less than 20 percent of the target due.236 that the measure significantly impacted Ugandan demand The low tax collection was mainly due to poor policy design for mobile money services, with over half of mobile money encouraging high avoidance (e.g., by using virtual private users abandoning mobile money transactions in favor of cash networks) and administrative difficulties.237 The Excise Duty and bank transfers.242 Weaknesses in the tax policymaking (Amendment) Act 2021 repealed the OTT tax. Starting July process, particularly minimal research and incident analysis, 1, 2021, all internet subscribers in Uganda will have to pay a and limited stakeholder consultation, undermined the new 12 percent on internet bundles. Benin repealed a social quality of policy design.243 In response to public outcry and media tax of USD 0.0089 (5 FCFA) fee per MB for data used the drop in usage of mobile money services, the tax law to access social media to raise revenue for public services was revised in November 2018, imposing a 0.5 percent after online and street protests.238 Niger was considering a tax tax on the value of withdrawals only (but not at the time of on internet to indirectly capture OTTs. actual transfer).244 In Tanzania, in the face of the Covid-19 pandemic, the Tanzanian government introduced new levies More recently, mobile money has attracted the interest of on ‘mobile money transfers and withdrawal transactions levy’ tax authorities looking to close budget deficits. Taxation has in July 2021.245 On August 31, 2021, the Finance Minister taken a variety of forms, ranging from excise duties on the informed that the levies would be reduced by 30 percent, fees or charges levied by mobile money providers (including with the mobile operators also revising their fees downwards telecom operators) to transaction taxes on the underlying by 10 percent. The tax has been modified multiple times amount. Kenya was one of the first countries in Africa to since its introduction, with the most recent changes applying introduce a tax on fees charged for mobile phone-based exclusively to electronic money withdrawals. The tax rate financial transactions (which is currently 12 percent).239 Excise varies based on the transaction size. In Zimbabwe, an duty on digital transaction fees also applies in Uganda (15 intermediated money transfer tax of 2 percent applies on the percent) and Tanzania (10 percent). Taxing transaction fees transfer of money between persons mediated by financial 234. 234 Ivory Coast, FY19 Financial Law. See PWC, Ivory Coast: Corporate - Other Taxes, available at https://taxsummaries.pwc.com/ivory-coast/corporate/other-taxes. 235. Although the telecom sector is less than 2 percent of Uganda’s GDP, ICT contributes a significant share of domestic revenues – climbing to 10.5 percent of overall revenues (3 percent consists of excise duties) in 2019. See World Bank, “Digital solutions in a time of crisis”, July 2020, p. 7, http://documents1.worldbank.org/curated/ en/775621594292073824/pdf/Uganda-Economic- Update-Fifteenth-Edition.pdf. See also Pollicy.org (2019). “Offline and Out of Pocket: The Impact of the Social Media Tax in Uganda on Access, Usage, Income and Productivity”. Available at https://www.genderit.org/resources/offline-and-out-pocket- impact-social-media-tax-uganda. 236. Kavuma, S. N., et al (2020), “An analysis of the distributional impact of excise duty in Uganda using a tax-benefit microsimulation model”, WIDER Working Paper, No. 2020/70, ISBN 978-92-9256-827-6, The United Nations University World Institute for Development Economics Research (UNU-WIDER), Helsinki, p. 5. Available at: https://www.econstor.eu/bitstream/10419/229294/1/wp2020-070.pdf [last accessed on March 23, 2021]. 237. Kavuma, S.N., et al (2020), p. 5. 238. The Decree 218-34 of July 25, 2018. 239. Sec 32(b)(iii) Kenya’s Finance Act 10 of 2018. See also Excise Duty Act (First Schedule, part II – excisable services, as amended). 240. An injunction is pending, and the levy is expected to be introduced in May. 241. GSMA, “The causes and consequences of mobile money taxation: An examination of mobile money transaction taxes in Sub- Saharan Africa”, June 2020, p. 21. Available at: https://www.gsma.com/mobilefordevelopment/wp- content/uploads/2020/06/GSMA_The-causes-and-consequences-of-mobile-money-taxation.pdf 242. GSMA, “The causes and consequences of mobile money taxation: An examination of mobile money transaction taxes in Sub- Saharan Africa”, June 2020, p. 22. Available at: https://www.gsma.com/mobilefordevelopment/wp- content/uploads/2020/06/GSMA_The-causes-and-consequences-of-mobile-money-taxation.pdf 243. A. Lees and D. Akol (2021), “There and Back Again: The Making of Uganda’s Mobile Money Tax”, ICTD Working Paper 123. Available at: https://opendocs.ids.ac.uk/ opendocs/bitstream/handle/20.500.12413/16727/ICTD_WP123.pdf?sequence=1&isAllowed=y 244. Excise Duty (Amendment) (No. 2) Bill, 2018. See: A. Lees and D. Akol (2021), “There and Back again: The Making of Uganda’s Mobile Money Tax”, ICTD Working Paper 123, Brighton, Institute of Development Studies. 245. These regulations have been issued as follows: The National Payment Systems (Electronic Mobile Money Transfer and Withdrawal Transactions Levy) Regulations, 2021: Government Notice No 496A of 2021, published on 30 June 2021, and The Electronic and Postal Communications (Airtime Levy) Regulations 2021: Government Notice No 496M of 2021, published on 30 June 2021. Available at: http://www.parliament.go.tz/polis/uploads/bills/1623832428-13.06.2021 percent20THE percent20FINANCE percent20BILL percent202021.pdf 70 >>> ADDRESSING OLD ECONOMIC GOVERNANCE RISKS FOR DIGITALIZATION institutions other than by cheque.246 Taxation on mobile companies.252 For instance, the Kenyan Finance Act 2020 money penalizes payment or transfer using mobile money introduced the digital service tax at the rate of 1.5 percent of through a digital medium as opposed to traditional commercial the gross transaction value (effective 1 January 2021). The banking services (banks or cash transactions), although this tax is payable by a person whose income from the provision of is seen as a quick way of raising revenues and capturing services is derived from or accrues in Kenya through a digital informal transactions. Progress on electronic payments and marketplace. However, it is important to avoid the negative financial inclusion slows down when taxes on mobile money effects of DST on the digital sector growth in African countries, are imposed without significantly expanding the tax base.247 particularly startups and SMEs, for example, through de minimis thresholds to target established and profitable digital As more transactions are conducted online, governments are businesses (ATAF, 2020). rethinking the application of indirect taxes on these transactions to keep or expand the tax base. Application of VAT/GST on online purchases is a good practice to level the playing field TAX EXEMPTIONS AND INCENTIVES between online and offline transactions.248 Monitoring and identifying digital activity are extremely challenging in case of Digital infrastructure operators are eligible for tax incentives lack of physical presence and the intangible character of the that apply either across all sectors or in the telecom sector. goods and services provided in the region, particularly in a cross- The incentives frequently aim to attract foreign investment by border context. By 2021, more than 80 countries worldwide, 10 promoting the country as a location for investors or fostering in Sub-Saharan Africa, have introduced consumption taxes local investment in specific economic sectors or economic on digitally provided goods and services. 249 In Kenya, while zones with low-income or high unemployment rates.253 This is electronic services delivered to a person in Kenya were already also an instrument used in fragile environments. For example, within the scope of the VAT, the Finance Act 2019 specified that in Burkina Faso, a company in all sectors that invests between supplies made through a digital marketplace are also subject 2,000 million XOF (approx. USD 3.69 million) and 25 million to VAT.250 However, the current administrative challenges of XOF (approx. USD 46,17) and creates at least 40 jobs enjoys collecting VAT from digital goods and services provided by non- an exemption of income tax for a 4-year period and reduction of residents remain unaddressed.251 income tax of 50 percent on years 5 to 7 during their operational period.254 Nigeria offers investment incentives and license Tax authorities face challenges in collecting revenues from incentives specific to telecom operators. Telecom operators profits generated by digital businesses (such as digital can periodically recover their investment and have access platforms) that offer services without a physical presence in to a guarantee of long-term loans at minimal interest rates. their country, and that can shift tax liabilities to other jurisdictions In some countries, telecom operators enjoy tax incentives (World Bank, 2021). Although a multilateral approach would as a result of special agreements with the government, or be useful, progress has been slow as part of the G20/ OECD some operators enjoy differential treatment. MTN Congo S.A. Inclusive Framework, and some African countries have (MRN Congo-Brazzaville) was granted a five-year 50 percent not supported the framework (Nigeria, Kenya). In Africa, reduction on its corporate tax rate due to such an investment Kenya, Nigeria, Sierra Leone, Tunisia, and Zimbabwe have agreement. In the 2019 financial year, following the enactment implemented alternative tax measures aiming at taxing digital of new local legislation, the tax incentive granted was reduced 246. Finance Act [Chapter 23:04], Sec. 22 G. http://www.zimtreasury.gov.zw/index.php?option=com_phocadownload&view=category&id=14:finance-act&Itemid=790. See also: Income Tax Act, Sec. 36G; Statutory Instrument 205 of 2018. https://www.veritaszim.net/sites/veritas_d/files/SI percent202018-205 percent20- percent20Amendments percent20to percent20Finance percent20Act percent20 percent26 percent20Income percent20Tax percent20Act.pdf ; http://www.zimtreasury. gov.zw/index.php?option=com_phocadownload&view=category&id=14:finance-act&Itemid=790 247. N. Ndung’u (2019), Taxing Mobile Phone Transactions in Africa: Lessons from Kenya, Brookings Institution, Africa Growth Initiative Policy Brief. 248. See for example, OECD (2019), https://www.oecd.org/tax/the-role-of-digital-platforms-in-the-collection-of-vat-gst-on-online- sales-e0e2dd2d-en.htm 249. World Bank, 2022, Presentation on digital taxation, Webinar on Digital Taxation. 250. VAT Act, No. 35, Sec. 5(7) and 5(9). Available at: http://www.kenyalaw.org:8181/exist/kenyalex/actview.xql?actid=No. percent2035 percent20of percent202013 251. For a more detailed discussion on taxation of digital services around the world see World Bank (2021), World Development Report: Data for better lives, chapter 7. 252. For additional information on digital service taxes, see C. Clavey, J.L. Pemberton, J. Loeprick and M. Verhoeven (2019), “International Tax Reform, Digitization, and Developing Economies”, World Bank, Washington DC. Available at: https://openknowledge.worldbank.org/handle/10986/32530 ; A. Cebreiro-Gómez, C. Clavey, M. Estevão, J. Leigh-Pemberton, and B. Stewart (2020), ”Digital services tax: country practices and technical challenges”, World Bank, Washington DC. Available at: https:// openknowledge.worldbank.org/bitstream/handle/10986/36840/P169976002e89a07209ae40d48d6ebb7154.pdf?sequen ce=1 253. T. Matheson, P. Petit, “Taxing telecommunications in developing countries”, International Tax and Public Finance (2021), Vol. 28, issue 1, No. 9, p. 274-275. 254. Investment Code: Loi 038-2018/AN. 255. See, MTN Group Limited, Annual Financial Statements for the year ended 31 December 2020, p. 54. Available at: https://www.mtn.com/wp-content/uploads/2021/03/2020- MTN-Annual-Financial-Statements.pdf <<< 71 to 25 percent for the remainder of the agreement.255 Deferred Other countries implement tax exemptions or reduced VAT for payments for spectrum fees have also been observed for handsets and other devices. For instance, Rwanda exempts selected types of operators. The licensing framework in South mobile handsets and SIM cards from all taxes (including Africa provided a payment holiday for a period of three to VAT) as an incentive to promote the ICT sector. Senegal also five years for the proposed wireless open-access network exempts VAT in respect of mobile and landline phone devices (WOAN), while in Kenya, operators with more than 50 percent and SIM cards used by individuals and legal entities liable to local ownership can pay spectrum fees in installments during the tax on telecom services.259 In Tunisia, a favorable VAT a 5-year period. In these cases, the differential treatment of rate of 7 percent (instead of the general 19 percent) applies operators regarding taxes and fee payments can affect the to mobile phones and accessories in a bid to promote the level playing field and market dynamics. ICT sector. Kenya decided to implement tax exemptions on handsets in 2009, removing the 16 percent VAT rate on mobile This raises the question of whether tax incentives are handsets encouraging their widespread adoption.260 This shift appropriate and part of a good governance environment. made handsets more affordable, followed by a more than Some authors argue that allocating investment incentives to 200 percent increase in handset purchases and a 50 percent the telecom sector is unjustifiable, given that telecom operator to 70 percent increase in penetration rates.261 The VAT Act licenses confer exclusive rights to exploit locational rents 2013 allowed the taxation of previously exempted products from selling services to the domestic market.256 However, in such as mobile phones, computer hardware, and software, a fragile context, investment incentives might be necessary to and now mobile phone users in Kenya pay a 16 percent VAT kick off the sector. Tax incentives, such as exemptions or lower on the purchase of a mobile telephone.262 From a tax policy rates, could make mobile phone equipment and services more perspective, the question arises whether handsets, SIM affordable257, but alternative policies, such as targeted subsidies cards, and other devices deserve to be subject to a reduced to the demand or supply side, could be more effective. It is, VAT rate to encourage consumption. Reduced VAT rates therefore, important that tax incentives are well designed and are generally seen as an inefficient subsidy to the poor and administered to maximize their effectiveness and efficiency, can significantly complicate VAT administration.263 Single- minimizing market distortions.258 Fiscal transparency and good rate consumption tax systems are generally considered the governance are required for accountability and consistency, least distortive and the most desirable.264 Also, rather than as well as to decrease opportunities for rent-seeking and granting VAT exemptions, the focus should be on minimizing corruption. To the extent possible, tax incentives should distortions and eliminating any higher VAT on telecom goods also be granted based on rules rather than discretion. Tax and services than in other sectors. Removing high taxes on incentives should be subject to legislative oversight unified small, portable goods like mobile phones may also lower the in the tax legislation, and their fiscal costs on the broader risk of smuggling.265 Targeted supply or demand subsidies as economy assessed periodically as part of a tax expenditure part of universal service policies in the sector would be more review. Countries should prioritize establishing adequate efficient than blanket VAT exemptions or lower VAT rates on systems for monitoring the implementation of projects that access devices. receive discretionary fiscal benefits based on performance criteria, at the very least through random audits. 256. T. Matheson, P. Petit, “Taxing telecommunications in developing countries”, International Tax and Public Finance (2021), Vol. 28, issue 1, No. 9, p. 271. 257. J. Mistry (2005), “A conceptual framework for the role of government in bridging the digital divide”, Journal of Global Information Technology Management 8(3), 28-47. 258. IMF, OECD, UN and World Bank (2015), “Options for Low Income Countries’ Effective and Efficient Use of Tax Incentives for Investment”, report to the G-20 Development Working Group, p. 19. Available at: https://www.oecd.org/tax/options-for-low- income-countries-effective-and-efficient-use-of-tax-incentives-for-investment.pdf. For guidance on cost-benefit analysis of incentives, see also Kronfol, Hania; Steenbergen, Victor. 2020. Evaluating the Costs and Benefits of Corporate Tax Incentives: Methodological Approaches and Policy Considerations. Finance, Competitiveness and Innovation in Focus. World Bank, Washington, DC. © World Bank. https:// openknowledge.worldbank.org/handle/10986/33433 License: CC BY 3.0 IGO. 259. See also art. 361 of the GTC. IBFD database, 2021. 260. N. Ndung’u (2019), “Taxing mobile phone transactions in Africa; Lessons from Kenya”, Brookings Africa Growth Initiative. 261. D. Strusani and G. Solomon (2011), “Mobile telephony and taxation in Kenya”, GSMA and Deloitte LLP, United Kingdom. Available at: https://www.gsma.com/publicpolicy/ wp-content/ 262. N. Ndung’u, “Taxing mobile phone transactions in Africa: Lessons from Kenya”, Africa Growth Initiative, Policy Brief, august 2019, p. 2-3. 263. T. Matheson, P. Petit, “Taxing telecommunications in developing countries”, International Tax and Public Finance (2021), Vol. 28, issue 1, No. 9, p. 270. 264. OECD (2017), International VAT/GST Guidelines, OECD Publishing, Paris. Available at: https://www.oecd.org/ctp/international- vat-gst-guidelines-9789264271401-en.htm 265. T. Matheson, P. Petit, “Taxing telecommunications in developing countries”, International Tax and Public Finance (2021), Vol. 28, issue 1, No. 9, p. 270. 72 >>> ADDRESSING OLD ECONOMIC GOVERNANCE RISKS FOR DIGITALIZATION >>> Table 6: Types of taxes applied on digital infrastructure Excise Duties CIT Surcharges VAT Reductions and VAT Stamp Duties Incentives/ Exemptions Surcharges Angola Yes No No Yes Investment incentives Benin No No No Yes (*) No Burkina Faso No No No No No Tax credits, customs Cameroon No No Yes No waivers and investment incentives Congo, Dem. Rep (DRC) No No Yes No No Cote d'Ivoire Yes No No No No Egypt No Yes No Yes No Ghana No No Yes No No Capital allowance, Kenya No Yes Yes No deductibility of capital expenditure Madagascar No No Yes No No Mauritius No No No No No Morocco No No No No No Nigeria No No No No Investment incentives Rwanda No Yes Yes No Investment incentives Senegal No No No Yes (*) Investment incentives South Africa No No No No Investment incentives Tanzania No No Yes No No Togo No No No Yes (*) No Tunisia Yes Yes No Yes Investment incentives Accelerated capital Uganda No No Yes No allowance deductions (*) Other specific and ad valorem taxes applied on telecommunication service use (prices). Source: Niesten and Begazo (2023), based on World Bank, Digital Infrastructure Taxes and Parafiscal Fees Database. 2 .1 . 3 . S U M M A RY O F EC O N O M I C economic governance and digitalization outcomes. Table 7 GOVERNANCE ISSUES AROUND provides an overview of the different taxes and fees, which TA X E S A N D PA R A F I S C A L F E E S are discussed individually in the previous subsections, as well as illustrative country examples, a summary of their alignment Taxes and fees applied on telecom services in Africa are with the three tax principles, and their potential impact on the not always aligned with good tax and public administration sector’s performance. principles, generating undesirable negative impacts that affect <<< 73 >>> Table 7: Overview of the different taxes and fees, alignment with tax and public administration principles for better economic governance Simplicity Efficiency Potential impact Selected country examples Taxes on Users Equity By potentially reducing affordability of handsets General VAT: Kenya, Uganda, Côte Taxes on (including smartphones), these taxes may act as a d’Ivoire, Ghana, Zimbabwe handsets and X X X barrier to those who do not have access to digital VAT exemption: Rwanda, Tanzania other devices connectivity, negatively impacting adoption. Favorable VAT: Tunisia Taxes on connections may negatively impact adoption, lowering the positive economic impact Taxes on of connectivity. These taxes risk being regressive Benin, Côte d’Ivoire, Ghana, Egypt, activations or X X X when they are expressed as fixed amounts, as they Tunisia connections represent a larger share of income for the least well off. Taxes on use distort consumption choices, reduce Excise duty: Zambia, Tanzania, DRC, Taxes on use X X X efficiency, and reduce affordability of telecom Ghana, Madagascar, Uganda, Gambia, services for the poorest mobile services users. Kenya, Rwanda, Cameroon Excise duties on transaction fees of mobile money: Kenya, Tanzania and Taxes on new services such as mobile money Uganda Taxes on and other digital financial services can create emerging digital X X X inefficiencies, potentially leading to lower Specific taxes on the value of the services consumption of digital connectivity and reducing actual transfer of mobile money: the positive externality linked with these services. Tanzania, Uganda, Zimbabwe, Cameroon, Nigeria, Ghana Taxes on Operators Regulatory fees, often based on revenue, may distort pricing decisions and lower economic General X - X efficiency. Regulatory fees may also generate Guinea, Gambia, Senegal regulatory fees complexity through the sheer number of fees, and potentially discourage investment in the sector. Licensing and authorization fees can be inefficient Annual licensing fees in Kenya, if they represent double taxation of the service, Tanzania, Madagascar, DRC, South- Licensing and create a barrier to entry/expansion, and are Africa, and Egypt (ranging between 0.4 authorization X - X not proportional to the level needed to finance percent to 2 percent) fees regulatory services. High licensing fees can prevent entry and limit the investment and innovation in the Lump sum annual licensing fees sector. (Rwanda and Togo) Varying between 0.3 percent (e.g., Universal When underused, if based on revenues, universal Tanzania) to 5 percent (e.g., Tunisia) of service service funds leave capital unemployed which operators’ revenue contributions could otherwise be productively invested to roll out X - X the network and increase mobile penetration. 0 percent in Angola since March 2020 (If fund If raised based on revenues, they may distort underutilized) pricing decisions and lower economic efficiency. No USF: e.g., The Gambia 74 >>> ADDRESSING OLD ECONOMIC GOVERNANCE RISKS FOR DIGITALIZATION Simplicity Efficiency Potential impact Selected country examples Equity Taxes on Operators Regulatory fees should be high enough to cover administrative tasks to offer the service (such as homologation/equipment registration, maintenance of the numbering system). Disproportionate high Numbering resources fees applicable Other regulatory X - X fees can hamper market expansion and end in various countries including Rwanda, fees up introducing another layer of taxation, which Côte d’Ivoire and Senegal limits the transparency of the tax system and, in some cases, leads to double taxation of the same service. Corporation tax is usually payable on profits, but commonly on revenue as minimum alternative tax where profits are negative. Higher taxation in the sector can affect resource allocation/investments Some African countries apply a higher Corporation tax X ✓ across sectors. CIT rate on the telecom sector: Angola, Guinea, Tunisia, Côte d’Ivoire However, from an equity perspective, corporation tax may be partly borne by shareholders, which could make it more desirable than indirect taxes. Since they act as barriers to trade, import duties tend to be detrimental to economic efficiency. Countries tend to put in place lower duties on ✓ capital inputs (equipment) to support national High import duties: Ghana, Tunisia, Import duties X X economic activities. Their impact on equity depends Madagascar, and Cameroon on the type of goods taxed; duties on necessity goods may be socially regressive while high duties on inputs will distort production decisions Local taxes have various bases and forms across countries; their consistency with the principles of efficiency, equity, and transparency should be Local economic contribution (Senegal), Local taxes - - X considered case by case. They tend however Tunisia (local authority tax), Nigeria to induce some level of complexity for firms that (infrastructure maintenance fee) operate across the entire country, such as mobile operators. Non-regulatory but specific earmarked fees can have various bases and forms. They are only Côte d’Ivoire (culture contribution), Other taxes applied to the telecom sector, or to a range of Guinea (research and development (e.g.: education different sectors, including the telecom sector. Their - - X fund), Egypt (scientific research, tax; cultural tax, consistency with the principles of efficiency, equity, education, and training programs in etc.) and transparency should be considered case by ICT) case. They tend, however, to induce some level of complexity. Source: Niesten and Begazo (2023) based on analysis by Deloitte, GSMA Association (2016), OECD (2014) and ITU (2013), and information on African countries based on World Bank Digital Infrastructure Taxes and Fees database. <<< 75 2.2. Participating in markets as a The presence of SOE in digital infrastructure segments is not uncommon. SOEs are found predominantly in natural supplier: State-owned enterprises monopoly sectors, including in digital infrastructure segments, and State-linked enterprises and with slightly greater frequency in emerging and developing economies (EMDEs) (Figure 68). This occurrence plays There are two main avenues for State-owned enterprises out in upstream digital fixed infrastructure and downstream (SOEs)266 and State-linked enterprises (SLEs)267 to affect digital communications and data services sectors across the governance of the digital economy. First, weak corporate both advanced economies (AEs) and EMDEs.268 In governance of these enterprises can reduce the accountability E-communications – fixed line networks (i.e., ownership and and efficiency of SOEs and affect their effectiveness in operation of backbone fixed infrastructure), SOEs were present achieving public policy goals. Second, the favorable treatment in over 55 percent of EMDEs and AEs.269 SOEs are also found of SOE and SLE vis-à-vis other market players in the sector downstream in more contestable digital communications can distort markets and protect SOE-SLE from market forces services sectors, including fixed line and mobile retail services. or pro-competition regulation to ensure balanced outcomes SOEs are slightly more common in EMDEs in retail fixed line for consumers and investors. This section highlights issues services (58 percent in EMDEs versus 48 percent in AEs) and that weaken the corporate governance of SOE-SLE and retail mobile services (46 percent in EMDEs versus 42 percent exacerbate adverse effects on market dynamics and in AEs).270 performance, especially in digital infrastructure. 266. SOE comprises any corporate entity recognized by national law as an enterprise and in which the national or subnational government exercises ownership, including joint stock companies, limited liability companies, and partnerships limited by shares. Statutory corporations, with their legal personality established through specific legislation, should be considered as SOEs if their purpose and activities are of a largely economic nature (i.e., the entity operates in a market for goods or services that could, in theory, be provided by a private company). SOEs also include government entities not organized as companies but operating in business or market activities. Source: World Bank (2019), Integrated SOE Framework; OECD Regulatory Questionnaire 2018. 267. Companies where the state holds a minority shareholding but does not hold, directly or indirectly through another company, the largest single share of the company’s equity capital. 268. “Upstream” refers to the following: wholesale activities (maintenance and operation of mobile and fixed line communications infrastructure), international gateway/landing stations (e.g., submarine cable, terrestrial cable, satellite), passive infrastructure services (e.g., ducts, towers). “Downstream” refers to retail services (mobile and fixed line communications, including international); mobile payment services; cloud and hosting services (e.g., website/email, data center collocation, cloud storage); other data services (e.g., mobility analytics, cloud computing, IoT solutions); digital services on digital platforms (e.g., e-commerce, ride-hailing, e-commerce for farmers); other communications services (e.g., videoconferencing, contact centers). 269. According to OECD-World Bank Group (WBG) Product Market Regulation (PMR) data (2018-2019), 270. Contestable sectors in EMDEs are marginally more likely to have an SOE, as is the case in financial services (~85 percent in EMDEs as compared to ~75 percent in AEs) and retail fixed line services (~60 percent in EMDEs compared to ~50 percent in AEs). However, EMDEs are much more likely to have SOEs in traditionally commercial sectors, such as accommodation, food and beverage (~60 percent for EMDEs, compared to ~30 percent for AEs), construction (~50 percent for EMDEs, and <20 percent for AEs), manufacture of refined petroleum products (>60 percent for EMDEs, compared to <20 percent for AEs), and wholesale trade, including motor vehicles (~45 percent for EMDEs, compared to ~15 percent for AEs). 76 >>> ADDRESSING OLD ECONOMIC GOVERNANCE RISKS FOR DIGITALIZATION >>> Figure 68: Share of countries with at least one SOE present in the sector or subsector in advanced economies and emerging and developing economies EMDEs avg. (26) AEs avg. (33) 0% 20% 40% 60% 80% 100% Water collection, treatment and supply Air transport - air-traffic-control activities Natural monopolies Railways - operation of railroad infrastructure Electricity transmission Electricity distribution Air transport - operation of airports Operation of terminal facilities (such as harbors and piers) Gas transmission E-Communications - fixed-line networks Gas distribution Railways - passenger transport Electricity generation Electricity retail supply Financial service activities, except central banking,… Other urban, suburban and interurban passenger transport Railways - freight transport Gas retail supply Contestable sectors E-Communications – retail fixed line services (voice,… E-Communications – retail mobile services (voice, video… Gas import E-Communications - mobile networks Electricity export Electricity import Sea, coastal and inland passenger water transport Gas storage Air transport - international passenger transport Air transport - domestic passenger transport Gas export Gas production Sea, coastal and inland freight water transport Gambling and betting activities Motion picture distribution and projection Accommodation, food and beverage service activities Manufacture of fabricated metal products, machinery and… Potentially competitive sectors Transport by Coach Manufacture of motor vehicles and their parts and… Building and repairing of ships and boats Construction Manufacture of aircraft and spacecraft Manufacture of railway and tramway locomotives and… Manufacture of refined petroleum products Manufacture of chemicals and chemical products Manufacture of pharmaceuticals, medicinal chemical and… Wholesale trade, incl. of motor vehicles Manufacture of basic metals Manufacture of computer, electronic and optical products Road Freight Transport Manufacture of tobacco products Note: AEs: Advanced Economies. EMDEs: Emerging Markets and Developing Economies. The percentage reflects averages across 25 AEs and 33 EMDEs indicated in paren- theses as being covered by the 2018 PMR database. ‘Potentially competitive sectors’ understood as markets where competition can be developed more easily without need of specific pro-competition ex ante regulation. Source: WBG-OECD Product Market Regulation Data (2018). <<< 77 In Africa, there are over 160 SOEs and SLEs operating in the SOE to SLE can be seen as a step towards a more favorable digital value chain, with approximately half being SOEs. SOE environment for private investment, bringing the possibility are important market players in international connectivity and of better economic governance by subjecting enterprises to fiber backbone (see section 1.2 for more details), while SLE market forces. In the following sections, the analysis focuses are relatively more important in mobile services. There is a fair on a sample of 37 SOE- SLE in 18 countries where more amount of spatial variation across the continent, too. With SLE detailed information on corporate governance and competitive more present in more liberalized countries and SOE more neutrality was collected through surveys. present in more closed markets. In general, transition from >>> Figure 69: Digital SOE-SLEs in Africa Number of SOEs Number of SLEs a. Number of SOEs b. Number of SLEs 1 3 1 2 3 0 6 0 2 2 2 0 1 0 2 0 1 1 0 1 2 1 1 0 0 2 3 0 1 0 1 4 0 1 2 2 1 0 3 5 2 0 1 1 1 0 1 1 3 4 2 0 3 2 1 1 0 2 0 1 3 1 1 0 2 0 0 2 0 2 2 1 1 3 2 1 0 2 0 1 0 2 0 3 1 1 2 2 2 0 3 5 0 3 1 2 4 3 1 4 6 8 1 1 5 0 2 0 No data No data 0 3 4 2 1 3 1 1 3 0 8 1 IBRD 47461 | AUGUST: Source World Bank authors based on corporate websites of SOE-SLEs. IBRD 47462 | 2023 AUGUST 2023 In terms of their activities, the majority of SOE-SLEs hosting services (49 percent), mobile payment services (43 surveyed were active predominantly in the following market percent), and digital platform services such as e-commerce, segments: international communications infrastructure, fixed ride-hailing, or e-commerce for farmers (35 percent). Across infrastructure such as backbone, and the provision of fixed the 37 SOEs and the 13 market segments that were studied, communications services to end users. These SOE-SLEs some 41 percent of SOE-SLEs (15) were the sole operator in are typically active in more than one market segment (i.e., at least one of the market segments in which they operate— they are vertically integrated to some degree along the value predominantly in fixed communications retail services chain). Around 73 percent (27) of the 37 SOE-SLEs analyzed (including voice and data), and fixed infrastructure and related own and provide access to fixed backbone infrastructure. wholesale services (backbone infrastructure, leased lines, Approximately 65 percent (24) are mobile network providers in metropolitan networks). Most studied SOE-SLEs are vertically the African region. SOE-SLEs are also quite active downstream integrated. For example, Kenya’s Safaricom, Mauritius in data infrastructure and services markets, such as cloud and Telecom, MTN South Africa, Telkom South Africa, and Maroc 78 >>> ADDRESSING OLD ECONOMIC GOVERNANCE RISKS FOR DIGITALIZATION Telecom participate in all digital market segments studied Compared to private companies, SOE-SLEs face some for purposes of this paper, as well as other non-ICT services distinct governance challenges. On the one hand, they may (e.g., real estate). When SOE-SLEs operate in traditionally suffer from undue politically motivated interference, leading commercial or contestable sectors that are more conducive to unclear lines of responsibility, a lack of accountability, and to private sector participation, and they are sole operators in efficiency losses. On the other hand, a lack of any oversight due one segment, there is a need to ensure a level playing field to passive ownership by the State can weaken the incentives between all operators and appropriate access to services in of SOE-SLEs and their staff to perform in the best interest of the monopoly segment. This will help to ensure that SOEs or the enterprise and the public. This can raise the likelihood of private firms are not benefiting from preferential treatment, self-serving behavior by corporate insiders and increased debt that they are subjected to competitive pressure and do not liability for the state. Additional governance issues arise when deter entry, and that they are properly incentivized towards SOEs have the dual goals of carrying out economic activities greater productivity and higher quality service delivery. and fulfilling a public policy role. Good corporate governance of SOE-SLEs is critical to ensure their positive contribution to economic efficiency and competitiveness. Establishing and 2 . 2 .1 . C O R P O R AT E G O V E R N A N C E O F enforcing corporate governance standards for SOE-SLEs S TAT E - O W N E D A N D S TAT E - L I N K E D in line with good international practice would help them to ENTERPRISES become more efficient and ensure positive contributions by SOE-SLEs. Digital technologies hold great promise for economic development, and in many countries in Africa, the state carries Transparency regarding financial and non-financial out commercial activities in the digital sector to influence performance is crucial for strengthening the accountability sector growth. State involvement in digital infrastructure and of SOE-SLEs’ boards and management and for enabling the services remains widespread in Africa through State-owned state to act as an informed owner. Disclosure refers to the enterprises (SOEs) and State-linked enterprises (SLEs). release of financial and non-financial information on the state SOE-SLEs have historically been present in telecom markets of affairs of a company. Disclosure requirements are both and often enjoyed monopoly rights, extending from submarine an incentive and a means for the board and management cable landing stations, fixed line and wireless broadband to perform their duties professionally. Disclosure is highly services to international gateway services and mobile internet. valuable for SOE-SLEs pursuing important public policy Almost all countries have liberalized the sector through objectives with a large impact on the state budget and on the various steps, including moving from majority shareholding risks carried by the state. However, disclosure requirements in SOEs to minority shareholding in SLEs. However, in many should not compromise essential corporate confidentiality countries, SOE-SLEs remain important service providers. The and should not put SOE-SLEs at a competitive disadvantage presence and performance of SOE-SLEs in the provision of against other market players. digital infrastructure and services can impact the quality and affordability of digital connectivity and the growth of the digital While each SOE-SLE operates within a specific country economy overall – boosting it or restraining its potential. and business environment, and any effort to strengthen Ensuring a level playing field in markets where SOE-SLEs its governance should be tailored to those circumstances, participate and focusing government direct participation on several broad principles of transparency and disclosure are solving market failures is vital to crowd in private participation relevant. Based on their national institutional and economic and increase the economy’s efficiency as a whole. Therefore, backgrounds, SOE-SLEs should strive to observe high appropriate management for SOE and SLE – to avoid biased standards of transparency and report financial and non- market results - is necessary to ensure good economic financial information on the enterprise in line with high-quality governance of digital infrastructure. internationally recognized corporate disclosure standards. Information that should readily be made available to the <<< 79 public notably include (but is not limited to) a clear statement practices in line with their own national requirements. Most of enterprise objectives; financial and operating results; the companies in the sample operate both under the company governance, ownership, and voting structure of the enterprise; law and a general public enterprise law (or SOE law). They remuneration of board members and high executives; also tend to be subject to many other public sector laws board member qualifications, selection process, roles, and and regulations. Working on equally applying all laws and independent status; material foreseeable risk factors and regulations for SOE-SLEs and the private sector is necessary risk management measures taken; financial assistance, to level the playing field. The line ministry executes ownership including guarantees received from the government; related rights in more than half of the companies covered. For these party transactions; and any other relevant issues relating companies, a clear legal separation between policymaking to employees, other stakeholders, and the general public. and ownership should be established so that ministers Websites can also be a useful tool to allow the general public, will only be able to influence the sector and its participants policymakers, and investors easy access to information on through transparent and fair regulation. Around 40 percent of SOE-SLEs. the companies studied lack clear mandates and objectives, affecting their effectiveness and accountability. Clarifying The evidence collected via the surveys of digital SOE-SLEs271 mandates and objectives will allow to design performance in Africa indicates that the countries and companies in the agreements, define key performance indicators (KPIs), and sample have room to improve their corporate governance have a more comprehensive performance oversight system. frameworks and practices to better fit best international >>> Figure 70: SOE-SLE mandates and objectives No clear objectives 40.5% Comme rcial, Social/non-commercial, and Regulatory functions 2.4% Comme rcial and Regulatory functions 2.4% Comme rcial and Social/non-commercial 14.3% Social/non-commercial 11.9% Comme rcial 28.6% 0% 10% 20% 30% 40% 50% Source: Jaupart and Begazo (2023) based on responses to the Digital SOE-SLE questionnaire. 271. The analysis is based on the information collected through a survey of 44 digital companies with state participation from 18 countries using desk research. The countries were selected to ensure fair representation of all Africa’s sub-regions, language groups, and market structures. Within those groups, the selection was based on the extent of publicly available data at the national level and the significance of SOEs in the sector. The objective of the survey is to identify common issues across these countries and pilot the survey for future use in other countries. The performance of the companies considered is assessed relative to the OECD Guidelines on Corporate Governance of State-Owned Enterprises with the objective to support corporate governance framework improvements. 80 >>> ADDRESSING OLD ECONOMIC GOVERNANCE RISKS FOR DIGITALIZATION Several issues were identified with respect to boards’ the enterprises without a transparent appointment process are composition, autonomy, functioning, and execution of SOEs (as opposed to SLEs). Board directors can be removed ownership rights. In almost 60 percent of the sample, the before the completion of their terms at the sole discretion line ministry executes ownership rights, creating a conflict of of the appointing authority in 91 percent of the SOEs in the interest between the government’s role as policy maker and sample and 82 percent of surveyed SLEs. The board Chair competitor in the market. Furthermore, not all boards of the and CEO positions are generally held by different individuals studied companies have a balanced composition with limited in most companies in the sample. The same person acts as government representation and the presence of independent Chair and CEO in only three companies (2 SOEs and 1 SLE), directors. For example, in 78 percent of the sampled SOE- two of which operate in Angola. In some other countries in SLEs, the Minister or high-level government official is present the sample, the Chair is a minister or other senior politician. on the board. Public disclosure of information on the nomination Conflicts of interest arise when a board member’s personal process of board members and the final appointments can help interests are contrary to those of the company; however, ensure professionalism and transparency. There is a formal members of the board of directors are not required to declare and transparent process to appoint board members in 80 conflicts of interest in 23 percent of the surveyed companies. percent of the surveyed companies but with one exception: all >>> Figure 71: Ownership rights and board composition What ministry, government agency or Minister or high-level Government Minister or high-level Government institution executes the ownership rights? official present in the Board? official present in the Board? 90% 90% 78% 80% 78% Specialized agency 8.1% 80% 70% 70% 60% 60% 50% 50% Tre asury/Ministry of Finance 32.4% 40% 40% 30% 22% 30% 22% 20% 20% Line minis try 59.5% 10% 10% 0% 0% No Yes No Yes 0% 20% 40% 60% 80% Source: Jaupart and Begazo (2023) based on responses to the Digital SOE-SLE questionnaire. The independent audit of annual financial statements by percent), but the external audits are conducted by private a professional audit firm is good corporate governance audit firms in only about two-thirds of the cases (64 percent). practice but not fully implemented for SOE-SLEs in Africa. The remaining third of companies for which information is Truly independent external audits strengthen the credibility available is audited by a public auditor, either the Supreme of companies’ financial reporting and reasonably assure the Audit Institution (SAI) or another relevant public body. These owner, investors, and the general public that the financial findings indicate that the practice of having independent statements correctly represent the company’s performance. external audits has not been adopted universally yet by digital Almost all companies surveyed conduct external audits (97 SOEs and SLEs. <<< 81 Board efficiency and effectiveness could be improved in African continent.273 Approximately 70 percent of companies several companies through specialized board committees and provide information on three or fewer categories among the robust management of conflict of interest. However, enacting 14 categories considered. About 11 percent of companies rules aligned with best practice does not always guarantee provide information on 10 or more items, and only three effective implementation in practice. Consequently, improving companies disclose data on all 14 items covered. Most the legal and regulatory framework to promote better corporate companies do not publish much non-financial information governance of SOE-SLEs does not always translate into more about their management, policies, and corporate governance efficient and transparent management of SOE-SLEs, notably arrangements. Company commercial and social objectives where there is a lack of capabilities or political will. are the most frequently reported corporate governance items. On the other hand, details of major transactions and Corruption remains a serious problem in many SOEs and performance contracts are seldom disclosed. Fewer than can have a sizable influence on the financial performance 17 percent of SOE-SLEs publish their annual reports or of companies. Strong criminalization of bribery can have financial statements on their websites. SOEs and SLEs tend important deterrent effects on the practice. It can also to disclose the same type of information to the public, with contribute to communicating that companies prohibit all SOEs usually being slightly more transparent. One important sorts of bribery payments and hereby improving behaviors. exception concerns sustainability and environmental In 92 percent of the digital companies in the sample (SOEs policies. The share of SLEs providing details on those is and SLEs combined), public officials and employees are almost twice as high as that of SOEs (43 percent and 22 subject to antibribery laws (Figure 72). The other companies percent, respectively). Because private shareholders own a representing 8 percent of the sample are all SLEs with larger fraction of the company capital, they might demand minority government ownership. better environmental and social practices. As SOE-SLEs can benefit from various types of financial support from governments, it is recommended as good practice for companies to disclose all support received from the state in a transparent manner. Among the SOEs covered, only 38 percent of companies disclose consistently if they have benefited from government support. Therefore, there is significant scope to improve practices in this specific area as laws and regulations often provide for such disclosure while practices are not always consistent. It is crucial that financial relations with the government be disclosed. This can reduce the likelihood that SOEs will be used as vehicles for off-budget spending and borrowing, political patronage, or corruption.272 If controls on financial support to SOEs are not effective, SOEs can become not only a source of distortion of the playing field but also an important liability risk. For example, in Comoros, Comores Telecom accounts for 40 percent of national debt, and to some extent, these debts were taken on without full government knowledge. Several findings emerge from the empirical analysis of corporate governance related information published on the webpages of all digital SOE-SLEs operating on the 272. International Monetary Fund (2020). Fiscal Monitor. Chapter 3: State-owned enterprises: the other government. April 2020. 273. Primary data was collected via desk research with the objective to create a dataset capturing the amount of information African digital SOE-SLEs make publicly available online. The geographic focus was on the whole African continent, including North Africa. Data collection took place between the months of October and November 2020. The effectiveness of financial and non-financial reporting can be measured across a number of different dimensions, including completeness, accuracy, timeliness, and relevance. Here the analysis focuses mainly on completeness and relevance. A total of 162 companies were covered by the desk research exercise, with 79 SOEs and 83 SLEs. 82 >>> ADDRESSING OLD ECONOMIC GOVERNANCE RISKS FOR DIGITALIZATION >>> Figure 72: Corporate governance and performance information disclosed (percentage of SOEs and SLEs) Information about shareholders Information about management team Major transactions and/or material events Key performance indicators, targets, results Performance Contracts Sustainability, e nvironment, social policie s Governance structures and policies Foreseeable risk factors Related party transactions Information about Board members Remuneration of key executives Remuneration of members of the Board Social objectives Comme rcial obje ctive s 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% SLE SOE Source: Jaupart and Begazo (2023) based on corporate websites of SOE-SLEs. Looking at the average country transparency of SOEs and SOE-SLEs have a functioning website. In East, Southern, and SLEs,274 East and Southern African countries SOE-SLEs tend West Africa, that percentage is above 80 percent. In Central to be more transparent (Figure 73).275 SOE-SLEs are the most Africa, however, only 63 percent of companies have a website. transparent in Kenya, followed by Egypt, Cabo Verde, Senegal, Corporate governance and performance information is most Sudan, and South Africa. In Kenya, for example, SOE-SLEs accessible in Southern Africa, with information on 33 percent provide information on 54 percent of the corporate governance of categories being available to the public on average. In the items tracked on average. In contrast, the Central Africa region other four regions, between 15 percent and 24 percent of the does not perform very well in terms of transparency overall. tracked corporate governance items are disclosed. Finally, Comparing transparency and disclosure outcomes across annual reports and financial statements are seldom shared African regions reveals interesting patterns. Southern Africa with the public. Only 5 percent of Central Africa SOE-SLEs is the best-performing region overall, followed by East Africa publish those on their websites. Fewer than 30 percent of (Figure 72). Central Africa underperforms all the other regions Southern African companies do so, though this is the highest in every category. In North Africa, more than 90 percent of regional figure for the continent. 274. The country average is calculated by taking the average number of corporate governance categories made available by SOE- SLEs per country and dividing it by 14 (the total number of categories). The average is then expressed in percentage to ease interpretation. 275. No digital company covered in Madagascar and Somalia. <<< 83 >>> Figure 73: Published information by SOE-SLE, by subregion 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 Central Africa East Africa North Africa Southern Af rica We st Af rica We bsite (%) SOE-SLE Corporate Governance Inf ormation (%) Annual report (%) Financial statements (%) Source: Jaupart and Begazo (2023) based on responses to the Digital SOE-SLE questionnaire. SOE-SLE transparency is positively correlated with economic various protections that may inhibit competition with and entry development and higher-quality regulatory institutions. by private sector actors. While SOEs are not a problem per Econometric results (Jaupart and Begazo, 2023) suggest se for competition and market dynamics, the benefits that that better regulatory quality encourages SOEs and SLEs SOEs may commonly receive as compared to private actors to publish more financial information, but not policy and risk distorting the functioning of markets and have significant governance related information. More populous countries implications for the viability and profitability of private tend to have more transparent SLEs, while countries with companies. Overall, the policy and regulatory environment higher GDP per capita tend to have more transparent SOEs. has evolved towards providing a level playing field to SOEs Furthermore, the level of transparency is not correlated with and private sector operators through both better corporate market structure in retail markets. These findings indicate that governance and better regulation that is neutral to ownership. African digital SOE-SLEs have room to improve transparency. SOE-SLE can affect the market in which they operate but also Strengthening regulation-making institutions might contribute upstream and downstream segments and ancillary markets to achieving this objective. having impacts in other sectors, including overall digitalization (Figure 74). 2.2.2. POTENTIAL MARKET DISTORTIONS LINKED TO SOE-SLE O P E R AT I O N 276 SOEs play a significant role in specific digital infrastructure sectors and downstream digital markets, but they benefit from 276. This section is based on Pop, Connor and Coelho (2022 2023). 84 >>> ADDRESSING OLD ECONOMIC GOVERNANCE RISKS FOR DIGITALIZATION >>> Figure 74: Potential SOE impacts across digital markets where competitive neutrality is lacking Digital Infrastructure Upstream SOE Markets Downstream CIT Surcharges • SOEs control of essential infrastructure • Government benefits (e.g., subsidies, • Barriers to entry and expansion may can: preferential treatment) to SOEs can protect SOEs • Increase risk of anticompetitive input crowd-out private investment exclusion • Vertical integration of SOEs can • Facilitate inefficient hoarding • The security Governments provide to increase risk of anticompetitive • Limit innovation downstream SOEs (benefits and bailouts) reduces behavior (e.g., margin squeeze, self- • Limit service-based competition due incentives to increase efficiency, preferencing of SOEs downstream) to refusal to share infrastructure productivity, and quality in the goods or services delivered • Regulatory capture by SOEs can distort • Vertical integration of SOEs can lead rule-making and enforcement of an ex to suboptimal market outcomes, e.g., • Cross-subsidization by SOEs of ante and ex post regulation in SOEs’ high wholesale prices, limited network commercial activities with government favor development funds for non-commercial activities distorts markets Ancillary Markets • SOEs may influence rule-making in other markets (e.g., data markets) through lobbying • SOEs may monopolize financial flows and influence financial access in other markets • SOEs may limit and/or delay the emergence and growth of new tech-based markets, to preserve their market position elsewhere Source: Pop, Connon and Coelho (2023), “State-Owned Enterprises in Digital infrastructure and downstream digital markets in Africa” The degree of government ownership of SOEs in the digital on the GSMA Connectivity Index both in the aggregate and infrastructure (mobile, international connectivity and data in terms of network performance and affordability (the level markets) varies across countries, but where state ownership of mobile tariffs). For example, Comoros, Libera, and Sierra is higher, it appears that market outcomes may generally be Leone all have at least two wholly owned SOEs in upstream lagging. A closer examination of the data revealed that 16 digital infrastructure sectors (backbone, international landing (around 43 percent) are wholly owned by the state, 6 (16 stations, etc.)) and they are the weakest performers overall percent) are majority-owned, and 15 (41 percent) are minority and exhibit the weakest scores amongst these countries for owned (10-25 percent). Additionally, 41 percent of SOE-SLEs mobile tariffs, which is supported by findings on affordability, tend to be vertically integrated across different segments with each ranking 156th, 149th, and 129th out of 170 of the digital infrastructure chain (Figure 74). Countries economies globally. with more than one wholly state-owned SOE in the digital infrastructure sector also tend to exhibit the lowest scores <<< 85 >>> >>> Figure 75: Share of SOE/SLEs that are vertically integrated Figure 76: Percentage of SOEs/SLEs that are either the largest across more than one segment or second largest company in the relevant market segment 11 41 54 89 Vertically Integrated Non-Integrated No Yes Source: Pop, Connon and Coelho (2023) “State-Owned Enterprises in Digital infra- Note: Market share is based on the number of subscribers according to GSMA data structure and downstream digital markets in Africa“, based on WBG, Database of as of Q4 2019. SOEs in Digital Sectors across Africa Source: Pop, Connon and Coelho (2023) “State-Owned Enterprises in Digital infra- structure and downstream digital markets in Africa“, based on WBG, Database of SOEs in Digital Sector across Africa Of the 37 SOE-SLEs analyzed in detail across 18 countries, 8 their financial accounting, which fosters the identification of (approximately 20 percent) perform both commercial and non- costs associated with public service obligations (PSOs) and commercial functions. However, in almost 90 percent of the competitive and non-competitive services and monitoring of countries analyzed, there are no requirements for the SOEs revenues (or losses) associated with other activities (Box 7). to separate commercial and non-commercial (i.e., universal If the legal framework fails to define commercial versus non- and public service obligations) functions. Account separation commercial activities and subsequently also fails to impose is also useful to avoid cross-subsidization between services an obligation on operators to separate such activities to some provided in markets with limited or absent competition (e.g., degree, this can facilitate cross-subsidization, create a lack broadband backbone) and those provided in markets where of transparency regarding financial flows, and potentially competition is possible (e.g., retail services). Morocco and compromise market-based decision-making by telecom Tunisia are the only two countries that legally oblige operators operators, including SOEs. to separate commercial from non-commercial activities in BOX 7: MOROCCO AND TUNISIA REQUIRE OPERATORS TO SEPARATE ACCOUNTS, WHICH PROMOTES A LEVEL PLAYING FIELD Morocco: The Décret n° 2-97-1025 du 27 chaoual 1418 (25 février 1998) relatif à l’interconnexion des réseaux de télécommunications tel qu’il a été modifié et complété par le décret n°2-05-770 du 6 joumada II 1426 (13 juillet 2005) requires companies operating in the telecommunications sector and declared dominant by the telecom regulator to maintain separate accounts (account unbundling) for their activities in order to identify the following categories of costs: (i) general network costs, i.e. the costs relating to the network elements used at both by the operator for services intended for its own users and for interconnection services; (ii) costs specific to interconnection services; (iii) costs specific to other services provided by the operator; (iii) all other costs. This provision also applied to SLEs. 86 >>> ADDRESSING OLD ECONOMIC GOVERNANCE RISKS FOR DIGITALIZATION Tunisia: The Code des Télécommunications at Article 26bis (Ajouté par art. 2 de la loi n°2008-1 du 8 janvier 2008) requires that Operators of public telecommunications networks and access networks keep separate accounts so as to distinguish between each network and each service, and to avoid any cross-subsidization between their operations. Operators send every year to the telecom regulator a financial statement that allows to distinguish between services. These statements are audited by an independent firm designated by the telecom regulator. Source: Pop, Connon and Coelho (2023), “State-Owned Enterprises in Digital infrastructure and downstream digital markets in Africa“, based on WBG, Database of SOEs in Digital Sectors across Africa Of the 37 SOEs (fully or majority state-owned) surveyed of digital infrastructure, fixed and mobile communications across 18 countries, only one of the host countries, Tanzania, services providers, and downstream data services providers, systematically required SOEs in the digital sector to achieve a with or without state ownership, and to different legal classes commercial rate of return in their operations. SOEs operating of businesses, with no differences in coverage, applicability, in a commercial and competitive environment should be transparency or implementation. Regulatory neutrality is expected to earn rates of return similar to comparable private paramount to ensure a level playing field in digital markets, businesses over a reasonable period of time; otherwise, particularly where there is vertical integration in the market, private actors can be undercut and crowded out from the or ownership of essential infrastructure, and a higher risk of market. In Ghana and South Africa, two SLEs were ultimately self-preferencing. However, lack of regulatory neutrality can expected to achieve a commercial rate of return, not due be compounded by the conflict of interest of the government to an overarching legal obligation but rather because the as SOE owner and regulator, as in the case of Angola (Box 8). government held only a minority stake. When SOEs are not required to achieve a commercial rate of return, they Furthermore, in order to promote competition, government operate within soft budget constraints, which reduces the intervention is needed to set ex-ante regulations to prevent incentives to increase efficiency, productivity, and quality of abuse by operators that control essential facilities, including the services delivered and can consequently lead to poor SOE-SLEs, and to enable competition. SOE-SLEs in Angola, market outcomes, as seems to be the case of SOE in Sierra Benin, Comoros, Gabon, Morocco, South Africa, and Tunisia Leone. SOEs operating within soft budget constraints could manage essential facilities infrastructure, even though they factor their low profit margins into their pricing. Further, SOEs also operate in the retail segments of the value chain—this could also exclude competitors by pursuing aggressive pricing creates risks of self-preferencing if governments do not also policies financed by the low profit, if not below-cost pricing. have in place ex-ante regulation. Lack of effective ex-ante regulation can end up protecting SOE-SLEs. Protection of To the extent possible, public and private businesses should SOE-SLE can also be indirect, for example, by preventing conduct their activities under the same regulatory conditions entry to markets where SOE-SLEs operate. In Morocco in order to avoid SOEs receiving advantages that distort and Senegal, the licenses for providing telecommunication competition in the marketplace. Where this is not feasible, services are awarded through an auction, and the government appropriate adjustments should be made to neutralize the has discretion on when to launch this process. Lack of remaining advantages or disadvantages. The word “regulatory” transparent, objective, and non-discriminatory assignment of is interpreted broadly as referring to the legal and regulatory radio spectrum can favor SOE-SLEs. SOE-SLEs and private frameworks in which businesses operate (e.g., the general firms should bid on an equal footing for spectrum, meaning business environment dealing with business laws, licensing and governments should avoid engaging in bilateral negotiations regulations, bankruptcy, and antitrust) and the enforcement of with SOE-SLEs that may result in a below-market price for product market regulations in their relevant sector. At a general spectrum – especially in the absence of independent spectrum level, regulation should be non-discriminatory – i.e., it should valuation to provide a reference point to the government - or apply equally to companies involved in the management in a competitive time advantage in relation to private players. <<< 87 Direct negotiation for 4G spectrum has been experienced in of a discriminatory assignment is especially high whenever Senegal (Sonatel’s 17-year 4G license (1.5GHz and 800Mhz) spectrum management is not entrusted to an independent and Kenya (Safaricom’s 4G spectrum license in exchange for authority that can prevent the state from favoring a SOE-SLEs the delivery of a national police security network). The risk through the preferential award of spectrum rights. BOX 8: DIGITAL SOES IN ANGOLA: STATE MONOPOLIES, PREFERENTIAL TREATMENT AND WEAK MARKET OUTCOMES Angola exhibits weak performance in its digitial infrastructure sector overall. It ranks 130th out of 170 economies in the overall GSMA Connectivity Index. Network coverage is approximately 57 percent, which is just below the continental average, and places it 128th out of 170 economies globally. Affordability is marginally better, as Angola ranks 96th on the GSMA affordability ranking (as compared to 129th for Sierra Leone), but mobile penetration is lower at 46 percent. Angola is identified as having highly concentrated markets, with a Telecom HHI score of 6100 (one of the 10 most concentrated in the Africa region). Market structure and regulatory protections for state-owned incumbents may at least in part be one of the causes for these market outcomes. Three of the SOEs studied are the largest market players with respect to ownership of backbone infrastructure, and the provision of fixed and mobile communication services. Angola Telecom, which provides wholesale and retail fixed communications services is the sole operator with respect to wholesale, but competes with private sector operators with respect to retail mobile and fixed communications services. Indeed, by law, the ownership and operation of digital infrastructure in Angola is reserved to the State, and although activities that are not part of the basic network can be provided by companies with no government participation, it is only possible through concession contracts. Additional institutional structures and preferential treatment may also be a problem, however, compounding the perceived risks and low returns associated with private investment. Various SOEs sit as members of the Technical Council of the telecommunications sector regulator, which suggests prices and fees for the sector and consequently creates a conflict of interest for SOEs in terms of the commercial and non-commercial activities of SOEs in the sector. Furthermore, in downstream segments where Angola Telecom competes with private operators in the provision of retail mobile and fixed line services, Angola Telecom has preferential access to finance through capital injections and other operating subsidies from the government, priority loans through state-owned banks, and government guarantees. More generally, SOEs are exempted from the bankruptcy law and are not required to achieve a commercial rate of return. Source: Pop, Connon and Coelho (2023), “State-Owned Enterprises in Digital infrastructure and downstream digital markets in Africa“, based on WBG, Database of SOEs in Digital Sectors across Africa. SOEs benefit from regulatory protections in many countries Taxation is also applied favorably to some SOEs: 84 percent studied, which typically limit private sector participation and of the SOEs surveyed (in 72 percent of countries surveyed) disincentivize investment. Entry can be restricted if licenses are subject to full tax liability in their home countries (i.e., are issued only at the government’s discretion through tender subject to the same rate as private sector actors in the same processes, for example, in Angola, Ethiopia, and Morocco. market segment). However, 11 percent of SOEs are granted SOEs in Angola, Benin, Comoros, Gabon, Morocco, South preferential treatment with respect to tax credits or when tax Africa, and Tunisia manage essential facilities infrastructure arrears exist - in Angola, Sierra Leone, and Tanzania. and also operate in other segments of the value chain—this creates risks of self-preferencing if governments do not also SOE-SLEs benefit from preferential access to finance. Of the have in place ex-ante regulation mandating non-discriminatory SOE-SLEs surveyed, it was reported that 19 percent benefit and transparent access to infrastructure. from preferential access to finance from the government, 88 >>> ADDRESSING OLD ECONOMIC GOVERNANCE RISKS FOR DIGITALIZATION such as through reduced interest rates, government-backed risks particularly when SOE-SLEs are vertically integrated along loans, debt guarantees, or capital injections, either because of value chains, with downstream service providers potentially express legal permissions or in practice. At the country level, able to access networks at reduced rates or even free of charge. this amounts to 39 percent of the countries studied offering Such benefits that are not available to private actors give SOEs SOE-SLEs preferential access to finance from the government, a competitive advantage and could allow them to expand or for example, in Angola, Ghana, Liberia, Mauritius, Mozambique, upgrade networks or price more aggressively and potentially South Africa, and Tanzania (Box 9). Financial benefits can raise undercut private competitors. BOX 9: COUNTRIES WHERE SOES BENEFIT FROM PREFERENTIAL ACCESS TO CAPITAL FROM THE GOVERNMENT Angola: Angola Telecom, which is wholly state-owned, receives operating subsidies and capital injections from the government, as well as priority loans through state-owned banks and government-backed loans. Ghana: The government can provide reduced interest rates for SOEs, and according to the Public Corporations Act, the president can waive payment of interests. Liberia: Sources indicate that, in practice, the government frequently covers Liberia Telecom’s liabilities. Mauritius: The National Computer Board, which engages in various downstream data services, including mobility analytics, cloud computing, digital platform services and is wholly state-owned, is legally entitled to receive financial contributions from the consolidated funds of the Government. Mozambique: Moçambique Telecom, which is 90 percent state-owned and active in all upstream aspects of digital infrastructure ownership and management, including international landing stations and communications, and passive infrastructure services, as well as downstream aspects, including mobile and fixed retail services, receives preferential access to government-backed loans as a form of financing. South Africa: Broadband Infraco, another wholly-owned SOE that is engaged in the ownership and management of fixed communications infrastructure, international gateways, and cloud and hosting services, has preferential access to government-backed loans and receives capital injections and grants directly from the government. In addition, the governing law provides that Broadband Infraco may borrow money, issue a guarantee, indemnity, or security, or enter into any transaction necessary in order to achieve its objects. Tanzania: The Tanzania Telecommunications Corporation, also wholly state-owned and active in fixed and mobile communications retail services, international landing infrastructure and communications services, passive infrastructure services (ducts, towers), and mobile payment services, has access to government-backed loans and reduced interest rates through the operation of various provisions of the Tanzania Telecommunications Corporation Act, 2017. Source: World Bank, Database of Digital SOE-SLEs in Africa None of the countries studied maintain de jure asymmetric, preference in the law in favor of domestic entities generally, preferential conditions with respect to procurement which would benefit both domestic SOEs and private processes for SOEs. Only Sierra Leone maintains a de jure entities over foreign bidders. <<< 89 3 >>> Addressing new governance risks for digitalization: safeguarding users through data and competition rules Rapid development of digital technologies in recent years has shown its great potential for Africa to promote job creation, improve the delivery of public services, and enhance individual welfare. For instance, it is estimated that e-commerce platforms, such as Jumia, could create about three million new jobs in Africa by 2025 (BCG 2019). Mobile money, exemplified by the global household name—M-Pesa, contributes to poverty reduction in many African countries (Suri 2017). The COVID-19 global pandemic has led to an accelerated rise in the use of digital technologies around the world, increasing innovation but also leading to various governance challenges and risks: risks of data misuse and risks of abuse of market power by digital service providers. 3.1. Regulating data: Data protection, cybersecurity and enablers for data use and re-use There is a growing concern about data protection and cybersecurity risks associated with various digital economic activities. Data protection is at the core of this apprehension for individuals around the world. From social media to mobile payments to telehealth appointments, our personal information is stored in databases on an unprecedented scale. While these innovations make our lives easier and keep us connected, unless the data are adequately protected it can be misused for all kinds of purposes, from harassment to fraud. Therefore, government interventions are needed to generate trust in digital technologies. 90 >>> ADDRESSING NEW GOVERNANCE RISKS FOR DIGITALIZATION: SAFEGUARDING USERS THROUGH DATA AND COMPETITION RULES The increased use of the Internet for both personal and collected information on data regulations across 80 countries professional needs has created opportunities for dangerous globally, including 24 Sub-Saharan African countries and players seeking to take advantage of vulnerabilities for 3 North African countries.282 Additional information for personal gain. In 2020, Kenyan internet users faced 14 selected countries was collected through desk research and million malware attacks between January and August. The interviews to understand challenges in the implementation of number of cyberattacks in Zimbabwe grew five times during rules captured by the GDRD. the same period.277 In August 2020, Experian, a global consumer credit reporting company, sold the personal data An environment that enables data use while safeguarding of about 24 million South Africans to a fraudster posing as a privacy and cybersecurity contributes to shaping public trust. legitimate client.278 In December 2020, personally identifiable A comprehensive data protection framework is generally information belonging to Absa Bank’s customers -who are associated with higher confidence in the regulatory quality, spread throughout twelve African countries- was leaked. while regulations on cybersecurity and cybercrime are According to the African Union Convention on Cyber Security particularly positively related to more confidence in the quality and Personal Data Protection, cybercrime “constitutes a of public services and the credibility of the government’s real threat to the security of computer networks and the commitment. Nonetheless, implementation is key to fostering development of the Information Society in Africa”.279 adequate digital environments that promote efficiency and productivity. Inefficient implementation can damage public Adequate legal and regulatory frameworks are key for countries perception and prevent a country from reaping the benefits of to be able to fully reap the benefits of emerging technologies a comprehensive data framework. while minimizing the associated risks (World Bank Group, 2021). The international nature of the use and impact of these technologies complicate their regulation. Concerns about how 3 .1 .1 . P E R S O N A L D ATA P R O T E C T I O N data is acquired, handled, used, shared, and reused have led to governments establishing heterogeneous approaches The protection of personal data is essential to data governance. for data governance. Data are the building blocks of these Personal data refers to data that conveys information that revolutionary technologies and restricting their flow can is specific to a known or knowable individual. Lack of trust hamper trade, innovation, and economic growth (Cory and in the way personal data is managed makes individuals Dascoli 2021). Governments have a difficult task: ensuring uncomfortable about sharing such data, which could limit the adequate flow of data across borders and within a country growth of digital markets. to allow novel technologies to operate adequately while safeguarding individual rights. A rights-based approach can As of December 2021, twenty-six African countries have lead to increased trust, which can, in turn, foster data flows introduced a general data protection regulation or have and data-based digital solutions for development.280 adopted general data protection laws. Tunisia, Mauritius, and Burkina Faso were regional pioneers in this regard, This section of the report focuses on a few key regulatory introducing data protection laws as early as 2004. During the aspects: data protection, cybersecurity, and cybercrime following decade, several countries followed suit, and as of to boost digital trust; and rules on the use, transfer, and December 2021, twenty-six African countries have adopted re-use of data to enable new digital technologies. Other general data protection laws. Notably, Mauritius passed its aspects of the data ecosystem, as described in the Word Data Protection Act, which is closely aligned with the EU Development Report ‘Data for Better Lives’ (World Bank General Data Protection Regulation (GDPR), five months 2021), are not covered. The analysis is based on data from before the EU regulation was implemented in May 2018. The the Global Data Regulation Diagnostic (GDRD)281, which year 2020 was one of significant advances for data protection 277. Makoni, Cyberattack surge highlights Africa security risk 2020 278. Corfield, Personal data from Experian on 40% of South Africa’s population has been bundled onto a file-sharing website 2020 279. See Preamble, African Union Convention on Cyber Security and Personal Data Protection. 280. World Bank Group (2021). 281. Chen (WDR21 background paper, 2021). The World Development Report 2021 conducted a Global Data Regulation Diagnostic. The Diagnostic is based on a detailed assessment of domestic laws, regulations, and administrative requirements in 80 countries. Data are collected through standard questionnaires which are completed mainly by lawyers specializing in ICT and data governance. Data are further verified through detailed desk research of legal texts, reflecting the regulatory status of each country as of June 1, 2020. 282 . The African countries included in the Global Data Regulation Diagnostic analysis are Angola, Benin, Burkina Faso, Cameroon, Democratic Republic of Congo, Cote d’Ivoire, Egypt, Ethiopia, Gabon, The Gambia, Ghana, Kenya, Liberia, Madagascar, Malawi, Mali, Mauritius, Morocco, Nigeria, Rwanda, Senegal, Sierra Leone, South Africa, Tanzania, Togo, Tunisia, and Uganda. <<< 91 legislation in Africa. Egypt’s Law on the Protection of Personal today’s world. For example, although Cameroon has no general Data came into force in October 2020. Prior to the approval of data protection law, its Law on Cybersecurity and Cybercrime this law, Egypt had sector-specific legislation that addressed –applicable to electronic communications networks and data protection issues, such as the Labor Law and the Banking information systems— includes provisions on data privacy. Law. In South Africa, although the Protection of Personal Although the Democratic Republic of Congo and Liberia have Information Act was signed into law in 2013, most of the not introduced any laws addressing the issue of data protection, relevant provisions were not operational until July 2020. Most both countries’ Constitutions include provisions regarding the recently, Rwanda published a data protection law in its Official individual right to privacy. Furthermore, eight African countries gazette in October 2021. Other economies have reportedly (denoted in gray on the map below) lack any mention of data engaged in discussions to introduce general data protection protection or privacy in their legal frameworks. It is worth laws, including Ethiopia, Malawi, and Tanzania; however, no noting that some countries have joined binding international public drafts of those laws are currently available. conventions on data protection. In 2016-2017, Mauritius, Senegal, and Tunisia became the first African countries to join Other countries have opted for a more targeted approach, the Council of Europe Convention 108 on data protection283; introducing sector-specific laws or relying on constitutional others have followed, and some are adopting the modernized provisions. These strategies are insufficient to address the 108+ convention. risk that data subjects and data processors are exposed to in >>> Data protection legal framework Figure 77: Data protection legal frameworks in Africa 0 0.2 0.4 0.6 0.8 1 No data Note: a top score of 1 (dark blue) indicates the existence of a general data protection law, a score of 0.5 indicates the existence of only a sector-specific personal data protection legislation; a score of 0.25 (lightest blue) indicates that there are privacy and/or data protection rights protected in the country’s constitution. IBRD 47463 | AUGUST 2023 (2023) based on Global Data Regulation Diagnostic and Data Guidance (2021) Source: Chen, Daza Jaller and Begazo 283 . Council of Europe Convention No. 108 on data protection: The Convention for the Protection of Individuals with regard to Automatic Processing of Personal Data. https://www.coe.int/en/web/conventions/full-list/-/conventions/treaty/108 Morocco became a Member State of the Council of Europe Convention 108 earlier in 2019. https://www.coe.int/en/web/conventions/full- list/-/conventions/treaty/108/signatures 92 >>> ADDRESSING NEW GOVERNANCE RISKS FOR DIGITALIZATION: SAFEGUARDING USERS THROUGH DATA AND COMPETITION RULES The existing data protection frameworks in African countries also included data protection by design and data protection included in the Global Data Regulation Diagnostic284 are by default, considered more novel and inclusive regulations. largely comprehensive (Figure 78). The analyzed regulations Data protection by design means that data protection should tend to include prescriptions that are considered good be considered from the initial design stages of products and international practice, including purpose limitation, data systems and throughout the lifecycle of the data collected, and minimization, and data subject rights. Kenya and Benin have not as an afterthought. >>> Figure 78: Comprehensiveness of African data protection laws Data subject's right to redress 25 Data subject's right to challenge accuracy 25 Data sharing restrictions 25 Storage limitation 25 Data minimization 25 Purpose limitation 25 Data Protection Authority 24 Safeguards on automated decisions 19 Necessity & proportionality test 9 Data protection by design/default, PETs 2 0 5 10 15 20 25 # of national laws that adopted the provision Source: Chen, Daza Jaller and Begazo (2023) based on Global Data Regulation Diagnostic and desktop research Additionally, the countries that opted for a general data Other countries in the region envision an alternative approach, protection law included the mandatory creation of a data including Burundi and Somalia, where the DPA will be part of protection authority (DPA) as part of its provisions. An the telecommunications regulator. This phased evolution, as independent DPA is regarded as a critical element of an part of an existing regulator, can help developing countries set effective data protection regulatory framework (Bennett and up their DPAs, focusing on building the agency’s resources Raab 2006); however, many African countries cannot afford before it becomes fully independent. While some countries like to establish an independent DPA and, therefore, establish it Angola and Egypt have mandated the creation of a DPA, the as part of an existing agency as a first phase. For example, DPAs have not been established. In other cases, it has taken this is the case in Nigeria, Rwanda, and Uganda, where the a few years to establish the authority and issue implementing data protection authorities are not separate from the ministry. regulations (Box 10). 284 . In addition to the data obtained from the Global Data Regulation Diagnostic, desktop research was conducted to analyze the existing data protection legal frameworks in every African country. <<< 93 BOX 10: THE ROAD TO START IMPLEMENTING DATA PROTECTION LEGISLATION IN KENYA The Data Protection Act was passed in 2019 and Kenya appointed its first Data Protection Commissioner’s (DPC) in November 2020, and since then there has been significant progress towards implementation of the Act. The DPC launched a new official website, which will provide the public with data protection resources and a platform to access important information, report breaches, or seek redress. Additionally, a Taskforce on Development of Data Protection Regulations was created in January 2021, to review the existing data protection legislation and propose an implementation framework, including the Data Protection Regulation. The DPC published several guidelines, open to public participation, on data subject consent and impact assessments. More recently, the Regulations of the Office of the Data Protection Commissioner (“ODPC”) were gazetted in January 2022 and took effect on 11 February 2022. The Registration Regulations provide a grace period of 6 months for compliance with the Registration Regulations; applications for registration must be submitted by 14 July 2022. However, the Commissioner must still negotiate a budget to be able to adequately fulfill the Office’s mandate. When compared to countries in different income groups, the with regard to data sharing restrictions, data processing data protection legal frameworks of African countries are requirements, and data subject rights. Finally, the region comprehensive. African countries perform closely to low-and- outperforms all other groups in the sample on a few indicators. middle-income economies in other regions. Nonetheless, More than half of the African countries provide rights to limit regulations that include data protection by design are less the making of decisions about individuals solely as a result frequent in Africa compared to other countries across different of automated processing of personal data, and 65 percent of income groups. Africa is in line with or slightly below the the existing data protection laws in the continent mandate the adoption rate of high-income countries (HICs) in the sample creation of a data protection authority. >>> Figure 79: Percent of countries per country income group that have adopted good regulatory practices on personal data protection Necessity & proportionality test Data Protection Authority Purpose limitation Data subjects' right to Storage limitation redress Safeguards on automated Data minimization decisions Data subjects' right to Privacy by design/default, challenge accuracy PETs Data sharing restrictions Low income Lower middle Upper middle High income Africa Source: Chen, Daza Jaller and Begazo (2023) based on Global Data Regulation Diagnostic and desktop research 94 >>> ADDRESSING NEW GOVERNANCE RISKS FOR DIGITALIZATION: SAFEGUARDING USERS THROUGH DATA AND COMPETITION RULES Although the legal and regulatory framework of personal data data misuse or breaches. Security requirements may include protection in Africa is generally comprehensive, adequate mandatory encryption of personal data, implementation implementation is essential to the safeguarding of individuals of rigorous internal policies, or the appointment of a data in the digital economy. Implementing such requirements is manager. Data breach notification requirements keep data particularly complex in countries with a strong presence processors and controllers accountable for notifying data of micro, small, and medium-sized enterprises (MSMEs). subjects and/or authorities of data breaches. Additionally, The cost of compliance may lead these companies to measures must be in place to criminalize certain online prefer risking a potential sanction than investing in the activities, such as unauthorized data to a system. The creation necessary upfront expenses, endangering their customers’ of a national cybersecurity strategy (NCS), infrastructure, personal data (Chander, et al. 2021). Governments that face and institutions –such as Computer Emergency Response budgetary constraints may not devote adequate attention to Teams (CERTs) or Computer Security Incident Response this personal data protection, failing to enforce existing rules Teams (CSIRTs)— is also of key importance, as it can help to (Chander, et al. 2021). Experiences in Morocco, Nigeria, identify, investigate, and address cyber-security threats and and Kenya exemplify the complexities of implementing protect critical national infrastructure. data privacy protection. Morocco has seen a disparity in compliance as multinationals subjected to foreign data Africa has fallen behind regarding measures to improve protection laws and companies in sectors like finance and cybersecurity and combat cybercrime. While some countries in health show higher levels of implementation, while local the region have shown moderate or high commitment in these companies are reportedly not in full compliance. In most areas, most African countries’ cybersecurity and cybercrime cases, the implementation authority has not issued any initiatives are in the early stages (ITU, 2019). Cybercrime sanctions to date to test enforcement powers. rules are found in different pieces of legislation throughout the continent. In a set of countries, such as Tunisia, cybercrime rules are enshrined within the criminal code. Benin’s Digital Code, 3.1.2. CYBERSECURITY AND which came into force in 2018, is divided into seven chapters CYBERCRIME addressing different aspects of digital activities, including a chapter dedicated to “cyber criminality and cybersecurity.” Although personal data protection is central to promoting Ghana took a similar approach, inserting cybercrime rules digital competitiveness, it is not enough. Comprehensive into its Electronic Transactions Act. While some countries, like cybersecurity and cybercrime regulations must be in place to Burkina Faso, cover cybercrime within their data protection ensure digital trust and boost the effectiveness of personal laws, others, including Cote d’Ivoire, Egypt, Madagascar, data protection regulations. These measures are essential Nigeria, Senegal, and Tanzania, introduced laws focusing to protect personal data “against accidental or unauthorized solely on cybercrime. Finally, cybercrime regulation is found destruction of accidental loss as well as against unauthorized within the telecommunications law for some countries, like access, alteration or dissemination”.285 Misuse or breach of Sierra Leone. The Democratic Republic of Congo stands out as data can be costly to individuals, firms, and society as a whole. the only African country in the sample that has not updated its legislation to address cybercrime. Every other country studied A regulatory framework that imposes cybersecurity in the region criminalizes at least the unauthorized damage, requirements on data processors and controllers and deletion, deterioration, alteration, or suppression of data criminalizes illegal access or use of infrastructure, systems, collected or stored as part of databases holding personal data. and data is needed to address the mounting concerns about 285 . Convention 108 of the Council of Europe <<< 95 >>> Global Rank Figure 80: Cybersecurity and cybercrime frameworks in Africa 0 0.176 0.352 0.528 0.704 0.88 No data Note: The Global Cybersecurity Index (GCI) measures the commitment of countries to cybersecurity. A score of 1 indicates the highest level of commitment. Scores in African to 47464 countries range from 0.008IBRD 0.88. | AUGUST 2023 Source: ITU (2019) Most cybersecurity laws broadly compel controllers and generate personal data, and restoring data and systems processors to ensure the integrity of the data and systems and after a physical or technical incident are some of the the confidentiality of personal data (Figure 80). Additionally, recommendations that could be implemented to improve and related to the previous section on data protection, half cybersecurity standards. Mauritius and Nigeria stand out as of the countries that adopted a general data protection law the only countries that require data protection awareness require the implementation of an internal policy that includes programs among employees and the encryption of personal the appointment of a data protection officer and procedures data. The lack of a comprehensive cybersecurity strategy for preventing and detecting violations. Nevertheless, these becomes clearer when compared to other countries around measures are not enough to ensure cybersecurity and the world. African countries and many low- and middle- personal data protection according to international standards. income countries have not implemented a broad regulatory framework to ensure cybersecurity. A comprehensive cybersecurity strategy should include ongoing security assessments of systems that use or 96 >>> ADDRESSING NEW GOVERNANCE RISKS FOR DIGITALIZATION: SAFEGUARDING USERS THROUGH DATA AND COMPETITION RULES >>> Figure 81: Comprehensiveness of African cybersecurity frameworks 25 20 19 # of national laws that adopted the provision 20 17 16 15 10 10 8 8 7 5 6 5 3 3 2 0 Performance of internal Encryption Routine assessments Risk assessment Cyber-security plan National CERT Integrity of data & Personal data protection Employee awareness Confidentiality Internal policy Anonymization/pseudony Ability to restore data & systems programs systems controls mization officer Security requirements for automated Cybersecurity requirements for data National processing of personal data processors/controllers cybersecurity strategy requirements Source: Chen, Daza Jaller and Begazo (2023) based on Global Data Regulation Diagnostic and desktop research African countries in the sample have adopted strong a lack of critical infrastructure protection, has left the country provisions for preventing cybercrime. One example is that exposed to breaches and misuse of data. On the other hand, most of the countries in the sample consider unauthorized, Nigeria established its Cybersecurity Policy and Strategy in damaging deletion, deterioration, alteration, or suppression 2014 (Toka Cyber Builders 2020). This strategy mandated of data collected or stored as part of databases holding the creation of the ngCERT to coordinate its implementation. personal data as a cybercrime. The fact that these provisions Nonetheless, lack of resources and capacity to guide and are comprehensive and consistent across countries not support sectoral CERTs have made the tasks of the ngCERT only provides clear rules for digital actors but could also more complex and have delayed the implementation of the facilitate coordination among countries to prevent and combat regulation. These countries exemplify the complexity of cybercrime activities across the region. Half of the sampled designing and implementing a cybersecurity strategy. countries mandate the creation of a national cybersecurity strategy, infrastructure, and institutions to identify, investigate, Safeguarding cybersecurity and combatting cybercrime and address cybersecurity threats. activities are on the working agenda of a few regional organizations. The African Union Commission organized the Implementation of cybersecurity initiatives varies greatly from first African Forum on Cybercrime in 2018, seeking to promote one country to another. Out of 44 African countries surveyed the adoption of cybercrime policies and legislation, improve by the International Telecommunications Union, only 13 international cooperation in the fight against cybercrime, and countries had a national CERT, and 14 had an NCS, making strengthen African criminal justice systems (African Union the region the worst global performer in both respects (ITU 2018). The AU Convention on Cyber Security and Personal 2019). While having strong laws in place is important, an Data Protection (Malabo Convention) also includes a chapter adequate institutional framework is crucial to ensure proper on cybersecurity and cybercrime. The Convention has not implementation of the strategies and enforcement of the been adopted by the minimum number of members and rules. While Benin announced the development of a National is not in force. Initiatives to identify gaps in the Convention Cybersecurity Strategy in 2018, it has not been adopted to and other options to establish a regional framework are date,286 which, together with an absence of know-how in the underway. ECOWAS appears to be one of the most active judiciary system as to how to deal with cybersecurity issues and African organizations on cybersecurity and cybercrime. In 286 . World Bank, Cybersecurity Capacity Review: Benin (2020) <<< 97 2011, it introduced the Directive on Fighting Cybercrime. Compared to other countries divided by income groups, Africa Although implementation is required by all member states, performs poorly in facilitating the use of public intent data. most member states have no relevant legislation or are still Less than one in five of the sampled countries demand that in the process of adopting it. Senegal stands out as the only public institutions use common technical standards that allow country that had introduced cybercrime legislation prior to for the interoperability of systems, registries, and databases. the Act, and Cote d’Ivoire, The Gambia, and Ghana have When compared to low and middle-income countries, Africa incorporated the Act. In 2012, the SADC published its Model has the lowest adoption rate on mandatory use of common Law on Computer Crime and Cybercrime, while in 2008, data classification categories across all government database the EAC promulgated the Framework for Cyberlaws, which applications. Finally, only half of the analyzed countries passed includes recommendations to adapt criminal laws and criminal an Open Data Act or an open data policy applicable across the procedure rules to address the issues presented by the use entire public sector. of ICTs and recommended that the Partner States accede to the Council of Europe Convention on Cybercrime. In 2016, When analyzing reuse of private data, the picture is even starker. the Economic Community of Central African States (ECCAS) Only one out of ten of the sampled countries allow private adopted the Declaration of Brazzaville, seeking to harmonize sector service providers to digitally verify or authenticate a national policies and regulations in the region, including person’s identity against data stored in the ID system. Similarly, cybersecurity laws and capacity building. Finally, a few African in more than 85 percent of the African countries studied, countries have joined the Convention on Cybercrime, also individuals do not have the right to data portability, which known as the Budapest Convention -an international treaty implies the ability to obtain their data processed by a controller that addresses crimes committed via the internet and other in a structured, commonly used, and machine-readable format computer networks, requiring parties to criminalize certain and to request the data transferred to another service or offenses (Council of Europe 2001). The Council of Europe product provider. Limiting data portability could be particularly drafted the Convention in 2001, and it has been ratified by damaging to the development of the digital economy as it 65 countries. Mauritius was the first African country to ratify could potentially result in the concentration of data by existing the convention in 2013, followed by Senegal in 2016, and players, diminishing innovation and generating further market finally Ghana and Morocco in 2018. South Africa signed the concentration. Finally, African countries fall behind upper- Convention in 2011 but has not ratified it yet. International middle and high-income countries when it comes to voluntary conventions offer the possibility to align domestic frameworks licensing of patent or intellectual property rights. None of the and complement regional approaches. African countries studied allow standard-setting organizations to mandate patent or intellectual property rights (IPR) holders to provide voluntary licensing access to “standard essential” 3 .1 . 3 . U S E A N D R E U S E O F D ATA A N D data or applications on FRAND (fair, reasonable, and non- C R O S S - B O R D E R D ATA F L O W S discriminatory) terms. A clearer regulation that enables data sharing while protecting users’ privacy could help grow the Facilitating a system in which data can help build a more digital environment in African countries. efficient digital economy requires enablers to promote the use and reuse of public and private intent data. Enablers The sampled African countries show strong and comprehensive of public and private data allow users to access and reuse regulatory frameworks on non-personal and personal data that data. Enablers include implementing open data laws or protection and e-commerce transactions performing at similar policies on public sector data, granting citizens access to levels to upper-middle income countries. Nonetheless, more information rights to request access to government records needs to be done to strengthen the regulatory framework in or data, facilitating data portability, and adopting an open enabling the reuse of public and private intent data and cross- licensing regime. border data flows (Figure 82). 98 >>> ADDRESSING NEW GOVERNANCE RISKS FOR DIGITALIZATION: SAFEGUARDING USERS THROUGH DATA AND COMPETITION RULES >>> Figure 82: Average scores on different data governance dimensions by income group/region High income Upper middle income Lower middle income Low income Sub-Saharan Africa E-comme rce/transactions 90 80 70 Cro ss border data 60 Enabling use of public intent data tra nsactions/flows 50 40 30 20 10 0 Cybersecurity and cybercrime Enabling use of private intent data Non-personal data protection Personal data protection Source: Chen, Daza Jaller and Begazo (2023) based on Global Data Regulation Diagnostic CROSS-BORDER DATA FLOWS Despite the concerns related to privacy, national security, and cross-border law enforcement, data localization and state data Although cross-border flow of data can bring benefits to infrastructure are not necessarily the most effective policy societies, promoting new business models and fostering the instruments and raise risks of negative impact on a country’s growth of the data-driven economy, countries are setting ability to participate in the global economy, attract private limitations due to security, economic, and political concerns. (including foreign) investment and boost cross-border trade. As data traffic grows, it also brings challenges such as Cory and Dascoli (2021) estimate that restricting data flows hacking, data breaches, and the possibility of citizens being reduces total trade volume, lowers productivity, and increases exposed to surveillance. To foster the positive externalities of prices for data-based downstream industries. Furthermore, data flows while preventing its negative ones, governments locating servers in one specific country is the opposite of a have implemented different data sovereignty regulations cybersecurity measure, as the data may be better protected aiming to protect the privacy of their citizens, foster economic when distributed in servers in different jurisdictions. Keeping development, and protect national security. These policies the data static negatively impacts its resilience, making it more can take many forms, including data localization requirements vulnerable to attacks. Cross-border data flow restrictions can imposing on companies the burden of storing the data, or a also affect the viability of regional data centers in Africa. copy of it, in the territory of that country or subjecting the use and transfer of data to certain regulations. In some cases, these Countries across the world have implemented diverse rules can also facilitate surveillance and censorship, therefore approaches to cross-border data flows in order to foster data the need for accountability for data policy implementation. exchanges while protecting personal data and digital security. Globally, these policies are reshaping how states relate to The two main strategies involve conditioning the movement each other and internally with their own societies. of data across borders or mandating that data is stored locally (Casalini and López-González, 2019). The diversity of <<< 99 cross-border data flow systems from country to country has be needed. In Africa, initiatives such as the African Union Data generated uncertainty and undermined the free flow of data. Policy, the African Network of Data Protection Authorities, The main systems through which data flows are implemented and the Smart Africa Data Protection Program provide fora worldwide include unilateral mechanisms, trade agreements, to discuss approaches to facilitate and regulate cross-border standards-setting, and plurilateral agreements (Casalini, data flows, but discussions are just starting. López-González & Nemoto, 2021). Unilateral mechanisms are some of the most common 3 .1 . 4 . R E G U L AT I O N A N D approaches in African countries. Among these, adequacy and GOVERNANCE accountability are two typical ways for cross-border transfers of personal data. Half of the studied African countries adopted Laws and regulations are relevant to determine the cost the adequacy approach. This strategy, also known as pre- of transactions in economic activities but also shape the authorized safeguards, allows transfers to countries that offer public’s perception of governments. African countries with an equivalent level of data protection. Kenya opted for the more comprehensive and robust data governance regulatory adequacy approach and allows for personal data transfers frameworks have a higher level of public perception of regulatory only if the controller or processor has proved that the laws in quality (Chen, Daza Jaller and Begazo, 2023). In countries the destination country are equivalent and that the transfer is that followed recommended practices on cybersecurity and necessary. The accountability approach has been implemented cybercrime, the public is more confident in the quality of the in other countries, including Ghana. Under this approach, the civil service and the degree of its independence from political controller is responsible for ensuring that the recipient of the pressures, the quality of public services, the quality of policy data complies with the relevant personal data laws. A different formulation and implementation, and the credibility of the approach allows the DPA to authorize a controller to transfer government’s commitment to such policies. Countries with a the data even if the destination country does not afford an healthier rule of law tend to have stronger cybersecurity and adequate level of data protection if the controller assures cybercrime frameworks.287 the protection of the data through contracts or internal rules. Madagascar, Togo, and Senegal are some examples of As digital technologies gain relevance and become online countries that have implemented such an approach. communications channels, countries face the challenge of promoting transparency and preventing the diffusion of false Mechanisms for cross-border collaboration and coordination information. In this sense, several African countries, including are needed, given the transnational nature of digital services. South Africa, Eswatini, Zimbabwe, Mauritius, Kenya, Nigeria, Interoperability, as a step towards harmonizing criteria and Tanzania, Egypt, and Rwanda, have instituted cybercrime laws approaches across countries, is critical to facilitate data flows or regulations that aim at preventing the spread of false news, and simplify the implementation of data regulations both from particularly amid the disinformation surrounding the Covid-19 a private and public perspective. Even with data localization pandemic (Box 11). Although there is a need for a more rules, collaboration is needed given the transnational nature comprehensive strategy to prevent misinformation288, the lack of cybercrime, but also considering that electronic evidence is of clarity in the terminology of current regulations, together with relevant for various investigations, including on antitrust, taxation, the broadness of their scope and relatively high penalties, could and money laundering involving non-digital economic activities. threaten freedom of speech and user’s privacy. For this, more effective mutual legal assistance treaties would 287 . Data on “Government Effectiveness”, “Rule of Law”, and “Regulatory Quality” are from the World Governance Indicators. The correlation coefficients between cybersecurity and cybercrime score and Government Effectiveness score is 0.50; between cybersecurity and cybercrime score and Regulatory Quality is 0.55; between cybersecurity and cybercrime score and Rule of Law score is 0.44; between Personal data protection score and Regulatory Quality score is 0.25. 288. Proposed initial steps include: “1. promote digital literacy initiatives; 2. Develop content moderation policies in a multistakeholder process and with independent oversight; 3. Integrate humans and technology in the fight against untruths online; 4. Increase transparency in spending on political advertisements online; and 5. Design a measurement agenda to improve the evidence base and inform more targeted policies to stop the creators and spreaders of untruths.” Lesher, M., H. Pawelec and A. Desai (2022), “Disentangling untruths online: Creators, spreaders and how to stop them”, OECD Going Digital Toolkit Note, No. 23, https:// goingdigital.oecd.org/data/notes/No23_ToolkitNote_UntruthsOnline.pdf. 100>>> ADDRESSING NEW GOVERNANCE RISKS FOR DIGITALIZATION: SAFEGUARDING USERS THROUGH DATA AND COMPETITION RULES BOX 11: MISINFORMATION AND THE COVID-19 PANDEMIC The Data Protection Act was passed in 2019 and Kenya appointed its first Data Protection Commissioner’s (DPC) in November 2020, and since then there has been significant progress towards implementation of the Act. The DPC launched a new official website, which will provide the public with data protection resources and a platform to access important information, report breaches, or seek redress. Additionally, a Taskforce on Development of Data Protection Regulations was created in January 2021, to review the existing data protection legislation and propose an implementation framework, including the Data Protection Regulation. The DPC published several guidelines, open to public participation, on data subject consent and impact assessments. More recently, the Regulations of the Office of the Data Protection Commissioner (“ODPC”) were gazetted in January 2022 and took effect on 11 February 2022. The Registration Regulations provide a grace period of 6 months for compliance with the Registration Regulations; applications for registration must be submitted by 14 July 2022. However, by March 2022, the Commissioner had to still negotiate a budget to be able to adequately fulfill the Office’s mandate. While proper regulatory frameworks for data protection Shutdowns have happened with particular intensity in times and cybersecurity foster trust in the government’s ability to of crisis, elections, or protest, and several governments safeguard a safe and thriving digital environment, internet consider they are necessary to protect public safety and the shutdowns undermine not only the development of the digital spread of misinformation. Nonetheless, Internet shutdowns economy but also public trust. Internet shutdowns have, have sweeping economic, social, and human rights-related globally, grown significantly in the last 10 years; according to consequences. In 2016, the Brookings Institute published a Statista there were 3 internet shutdowns in 2011 compared report which concluded that shutting down the internet costed to 188 in 2018 (Feldman 2019). India has been the country countries over $2.4 billion in 2015 (West 2016). Netblocks. with more internet shutdowns, accounting for 134 shutdowns org -an organization in charge of mapping internet freedom in 2018, while African countries like Ethiopia (5 shutdowns - estimates the impact on GDP that a one-day nationwide in 2018) and the Democratic Republic of Congo (3) are internet shutdown could have in different countries around also included among the list of top 10 countries that have the world; figure presents the impact for some of the African interrupted internet services the most (McCarthy 2020). countries that blocked internet access in the last years. Several shutdown cases have been brought to the courts. Africa has seen a particularly steep increase in internet Judicial cases in Zimbabwe and Sudan have been particularly shutdowns; according to a BBC report, in 2019 there were relevant as courts have upheld the right of internet access, 25 instances of shutdowns in the continent, compared to 20 ordering internet service providers to restore their services. in 2018 and 12 in 2017. Seven out of the fourteen countries that blocked access to the internet in 2019 had never done it in previous years: Benin, Gabon, Eritrea, Liberia, Malawi, Mauritania, and Zimbabwe. Most of the shutdowns registered were nationwide, and only Ethiopia and Sudan had targeted shutdowns. During the first semester of 2021 Access Now, an organization dedicated to promoting internet freedom, registered shutdowns in Uganda, Chad, Niger, Senegal, Ethiopia, Congo, and Gabon (Díaz Hernández et al. 2021). <<< 101 >>> Figure 83: Estimated cost of a hypothetical one-day internet shutdown (USD) Source: Netblocks.org/cost/ Note: Using Brookings Institution (2016) method which relies on development indicators. Based on 2020 data. 3.2. Promoting and safeguarding Effective competition in digital markets can be facilitated by regulations that cover diverse aspects, competition law competition in digital services: enforcement being one of them. Various policies can impact antitrust enforcement and policy market dynamics in the digital ecosystem - notably the upstream digital infrastructure, digital market operators, and advocacy data, which can be used both as an input and an output in digitalized value chains. For digital services markets (e.g., Despite the benefits that digital technologies can bring to the digital platforms, data, and cloud services, and mobile economy, specific risks are associated with their expansion. money) to be fully functioning, it is paramount to have in Growing dominance has been a long-standing concern in place a regulatory framework that can effectively tackle advanced economies, and countries need to be well-prepared existing bottlenecks and promote competitive outcomes in the to address this potential risk. Large direct and indirect network upstream digital infrastructure, an efficient use of data and effects and economies of scale foster market concentration encourage competition among digital operators. In addition, (Guellec & Paunov, 2017). Indirect network effects make it is essential to have in place equally effective competition market entry more difficult, protecting incumbents through laws comprising: (i) ex-ante merger review that can prevent first-mover advantage (Coale, 2018). However, the growing mergers that substantially reduce market competition, (ii) dominance of digital service providers has not been a major ex-post enforcement against anticompetitive agreements concern across African countries, as digital technologies and abuse of dominance to deter firms from engaging (including digital platforms) capture a relatively small portion in anticompetitive conduct; and (iii) effective competition of the offline market in most sectors, but their importance has advocacy tools to promote competition principles across all increased due to COVID-19 mobility restrictions, and countries policies and regulations that may affect the digital economy.289 need to be well-prepared to set a path that minimizes risks to effective competition. 289 . WBG WDR, Chapter 3 (2021). 102 >>> ADDRESSING NEW GOVERNANCE RISKS FOR DIGITALIZATION: SAFEGUARDING USERS THROUGH DATA AND COMPETITION RULES >>> Figure 84: Regulatory aspects important to enable competition in digital markets Digital infrastructure Digital markets International International Active Mobile Money Data & Cloud Digital Platforms gateways gateways infrastructure Services • MNOs • Communications • Submarine • Fixed wholesale • Radio spectrum • Banks • IXPs • Content cables • Mobile wholesale • Telecom • Data services • Search • Landing stations • Passive operators • Entertainment infrastructure; • Transactions masts, poles, towers, etc Data Effective data protection Effective spectrum regulations management • Data portability • Multi-homing Effective ex ante and ex post regulation of operators with a Effective ex ante regulation ‘Gatekeeper’ or a ‘Strategic Market Status’ of operators with Significant Market Power and ex post rule enforcement Effective consumer protection against aggressive and misleading commercial practice Effective sharing of active and passive infrastructure Effective competition law enforcement and advocacy for pro-competitive regulation Source: Pop and Coelho (2023), “Competition Policy in Digital Markets in Africa”. Competition law enforcement and advocacy are key to ensuring (ii) protecting competition within digital markets by deterring that the operation of digital business (digital platforms) delivers anticompetitive behavior vis-à-vis consumers, business users the best deals for consumers and generates greater economic and providers of complementary services.290 This section impact. Promoting an effective competition policy in digital describes issues related to antitrust enforcement and product markets is essential to ensure smaller firms and consumers do market regulations that affect competition in the markets not miss out on the economic benefits brought by digitalization. where digital business (including multisided transactional Due to their market characteristics – especially direct and digital platforms) operate, centering at the role of competition indirect network effects and economies of scale – digital authorities. Competition policy and law enforcement entails markets show a propensity towards increased consolidation investigating and punishing anticompetitive conduct in the and entrenched significant market power. This gives rise market, notably anticompetitive agreements, and abuse of to two types of competition concerns: (i) preserving market dominance, reviewing mergers to ensure competition is not contestability by lowering barriers to entry and expansion, and lessened, and promoting competition advocacy. 290. World Bank. 2021. World Development Report 2021: Data for Better Lives. Washington, DC: World Bank (“WBG WDR 2021”), Chapter 3; Jacques Crémer, Yves- Alexandre de Montjoye and Heike Schweitzer, “Competition policy for the digital era”, Final Report for the European Commission (2019); Stigler Center for the Study of the Economy and the State, Stigler Committee on Digital Platforms (2019). <<< 103 3.2.1. ANTITRUST ENFORCEMENT of November 2020, at least 43 African countries had a national competition law or were members of a regional agreement In Africa, the number of national and regional competition establishing antitrust regulations, accounting for more than 88 authorities has increased significantly in the last decade, percent of the continent’s population and 93 percent of the boosting the potential for competition law enforcement against GDP (see Figure 85 below). Out of the 43 African countries anticompetitive conduct. According to recent research by the surveyed, at least 26 have a functional competition authority World Bank (WBG) and Africa Competition Forum (ACF), as in place, less than 50 percent of African countries. >>> Figure 85: Status of competition law and authorities across Africa, 2020 Note: Additional countries with operational authorities include Mali, Ethiopia and Rwanda. Source: ACF-World Bank (2022 unpublished), The Institutional Gauge: Ensuring Sound Outcomes for Competition Policy in Africa. Nevertheless, there is still room for improving the African competition agencies still rely on a Minister to send independence safeguards of African competition agencies through their budget request to Parliament. This overall in a way that effectively shields them from undue private situation of fragile independence is confirmed by the fact and public influence. The 2019/20 WBG- ACF survey that only 4 of the 19 agencies combine both structural and indicates that out of 19 agencies surveyed, only 42 percent financial independence: Mauritius and Botswana.292 identified themselves as structurally independent from the executive branch (Algeria, Botswana, COMESA, Egypt, While countries have competition law frameworks, ample Kenya, Mauritius, Morocco, Nigeria, Eswatini, Zimbabwe, grounds for granting exemptions and suboptimal merger control and Tunisia).291 In terms of financial independence, only 57 powers can affect the ability of authorities to discipline digital percent of the surveyed agencies considered themselves business market behavior. Although laws punish collusion financially independent from the executive branch. In fact, and abuse of dominance, exemptions from the application even though their budgets are allocated by Parliament, most of the law can be granted in some jurisdictions not only to 291 . ACF-World Bank (2022), The Institutional Gauge: Ensuring Sound Outcomes for Competition Policy in Africa. 292 . ibid 104 >>> ADDRESSING NEW GOVERNANCE RISKS FOR DIGITALIZATION: SAFEGUARDING USERS THROUGH DATA AND COMPETITION RULES anticompetitive agreements but also to abusive unilateral New issues, such as dealing with collusive algorithms or conduct – an aspect particularly relevant for digital platforms. the treatment of (the power of) data in antitrust and merger Similarly, where merger control exists, ample grounds to cases, pose challenges that have not yet been developed by reverse technical decisions by the minister on public interest primary and secondary legislation on competition. In many grounds (e.g., Tunisia) or to review mergers based on public countries, the definition of a dominant firm does not account interest regardless of their impact on competition – for example, for the existence of multi-sided markets. All these features of in Morocco - can hinder efficient market dynamics involving competition frameworks are important to address competition digital platforms. In addition, general competition laws in most risks (Table 8). countries do not address specific concerns of digital markets. >>> Table 8: Typology of competition risks in digital platforms Type of case Behavior Specificities in digital platforms Examples Anti-competitive Collusion: • Data and algorithms can improve pricing, customization • Algorithms in e- agreements Agreement and market trends prediction which generates commerce can be between efficiencies, but they could also facilitate collusive trained to independently competitors agreements without human interaction.293 collude (e.g., by on market following the behavior parameters • Hub and spoke cartels. When firms outsource pricing of a price leader) which such as price, algorithms to the same third party this might create increases the risk of quantities, market coordination as competitors would be using the same “hub” collusion. segmentation, for developing their pricing algorithms and strategies. etc. • Algorithms could lead to tacit collusion. In most cases this is not prohibited by competition laws but its outcomes could lead firms to suppress output and increase prices in the same ways as explicit collusion. 294 Vertical • Resale price maintenance (RPM) An upstream supplier • Tourism and travel restraints: controls or restricts the retail price of its products or digital platforms Agreement services downstream by specifying a retail price. An imposing MFNs on between firms in incumbent may use RPM to protect its market position hotels not allowing them different levels from competition by, for example, placing restrictions to sell their rooms at of a on low cost online retailers’ ability to introduce price lower prices than the value chain that discounts. This might facilitate exclusion of lower price on the tourism reduce the ability cost distribution methods, like direct online sales platform (Even in offline of to consumers and can facilitate collusion between channels). (downstream) downstream retailers. 295 firms to compete • Most favored nation clauses (MFN). Restrictions imposed by platforms that require that a seller on the platform will not sell the product at a lower price through another platform. MFNs tend to raise prices charged by sellers and fees charged by platforms, discourage entry and distort innovation. 296 293. OECD (2017). “Algorithms and Collusion: Competition Policy in the Digital Age.” Available at: http://www.oecd.org/daf/competition/Algorithms-and-colllusion- competition-policy-in-the-digital-age.pdf 294. Kovacic, W., Marshall, R., Marx, L. and H. White (2011). “Plus Factors and Agreement in Antitrust Law”, Michigan Law Review 110(3), 393—436. Available at: https:// repository.law.umich.edu/cgi/viewcontent.cgi?article=1118&context=mlr 295. ICN (2015). “Online Vertical Restraints Special Project Report.” Available at: https://www.internationalcompetitionnetwork.org/wp-content/uploads/2019/11/SP_ OnlineVR2015.pdf 296. Boik, Andre & Corts, Kenneth. (2016). “The Effects of Platform MFNs on Competition and Entry.” Journal of Law and Economics. 59. Available at: https://www. researchgate.net/publication/305220937_The_Effects_of_Platform_MFNs_on_Competition_and_Entry <<< 105 Type of case Behavior Specificities in digital platforms Examples Abuse of Abusing a • Digital platforms have a greater tendency to tip towards • Dominance in search dominance dominant dominance given the prevalence of network effects and markets can manifest in position by strong economies of scale and scope that arise with abuse in the shopping excluding rivals high fixed costs/low variable cost structures and the market (e.g., through reliance on data and data-intensive technologies to gain refusal to deal, a competitive advantage. • A platform refuses discrimination, access to information tying, predatory • Given their multisided nature, dominance on one side of that would allow a third pricing, the platform can influence anticompetitive behavior on part to interoperate with exclusivity, another side. it and restricting access to • Where a platform controls technology or assets that are • A platform may rank essential inputs) essential to compete, there may be an exclusive abuse its own products or by exploiting through refusal to deal or through exclusivity higher when returning consumers a response to a • Digital platforms that are vertically integrated into consumer’s search product lines where they compete with third parties who sell on their platforms, may engage in self • Supplier of OS obliges preferencing including through algorithms that are biased device manufacturer to (intentionally or not) towards the platform • install supplier’s suite of apps as a condition to • Digital platforms that have presence in adjacent markets licensing the OS. (as is relatively common given the economies of scope present), can also abuse their dominance in the form of • Ride hailing apps forced tying or bundling of products. accused of predatory pricing to drive taxis out • Predatory pricing can be part of the business strategy of of the market. platforms seeking to quickly leverage network effects to dominate the market. • There has been debate over whether excessive collection of personal data on users could be considered an exploitative abuse. 106>>> ADDRESSING NEW GOVERNANCE RISKS FOR DIGITALIZATION: SAFEGUARDING USERS THROUGH DATA AND COMPETITION RULES Type of case Behavior Specificities in digital platforms Examples Mergers Mergers and • Identifying anticompetitive mergers in digital markets • Social media site acquisitions is more challenging since these mergers are less acquires a messaging are not an likely to meet the turnover or asset thresholds typically service and other social anticompetitive established in competition legal frameworks that signify media sites and merges practice like the when a merger must be notified to a competition data sets to acquire a other categories authority. This is because digital firms, by their nature, broader set of data above but are less likely to hold tangible assets and may not they may be generate significant revenues especially in their start up • Platform with a map prohibited by phases. service acquires a a competition smaller maps app authority if • “Small” mergers or vertical mergers that are typically partially to eliminate an they appear not considered to pose a risk to competition may in independent source of likely to reduce fact be damaging to competitive dynamics of digital mapping software competition by markets where the target firm holds data or intellectual strengthening property that may provide a competitive advantage to the market power of acquiror. the merged entity or by creating • Emergence of killer or zombie acquisitions as a potential conditions theory of harm where (typically cash rich) digital under which platforms acquire smaller firms and put their innovations coordination on hold before they can become a competitive threat. between firms become easier. Source: World Bank, 2021, Antitrust and Digital Platforms: An analysis of global patterns and approaches by competition authorities. EFI Insight-Trade, Investment and Com- petitiveness. Washington, DC: World Bank. In Africa, competition authorities on the continent have the digital economy are still emerging in their practice. It might engaged in relatively few cases and advocacy initiatives also result from resource and capacity constraints for antitrust involving both digital platforms and the broader digital investigations and advocacy initiatives on the continent. economy thus far. According to the World Bank’s Global Digital Among the authorities that have engaged in digital economy Antitrust Database and subsequent research that broadened cases and initiatives, the Competition Commission of South the type of sectors to include other digital economy markets Africa (CCSA) appears as the most active authority on the (including digital platforms),297 eleven African authorities298 continent (16 cases overall or 46 percent of cases), followed have investigated 29 antitrust cases and engaged in 6 by the Competition Authority of Kenya (CAK) that has finalized advocacy initiatives between 2008-2021. The relatively low five cases and the COMESA Competition Commission with engagement of African competition authorities (relative to the three cases (see Figure 86 below). number of countries) could indicate that competition issues in 297. In addition to digital platforms, antitrust cases from the following sectors were included: Digital advertising, Digital financial services, Software & data services, Wearables and All. 298. Algeria, Botswana, COMESA, Egypt, Kenya, Mauritius, Nigeria, Seychelles, South Africa, Tanzania, Zimbabwe. <<< 107 >>> Figure 86: Number of antitrust cases and advocacy initiatives related to the digital economy in Africa South Africa 6 2 1 7 Kenya 2 3 COMESA 3 Tanzania 2 Egypt 1 1 Botswana 2 Zimbabwe 1 Seychelles 1 Nigeria 1 Mauritius 1 Alger ia 1 Abuse of dominance Advocacy Anticompetitive agr eement Mer ger Source: Cases include enforcement activities and advocacy initiatives, including in Digital platforms, Digital advertising, Digital financial services, Software & data services, Wearables, or on the digital economy as a whole (without singling out sectors). At the time of writing (April 15, 2021), the team was not able to access all documents on the website of the Seychelles Fair Trading Commission and may thus be missing cases. Note: Pop and Coelho (2023),“Competition Policy in Digital Markets in Africa” The majority of antitrust cases and advocacy initiatives in Africa middle-income African countries have largely surpassed upper- have taken place in upper-middle-income countries, even though middle income, with 67 percent of the initiatives analyzed. This low-middle-income countries appear to be catching up. Whilst appears to suggest that low-middle income countries have not in high and upper- and upper-middle income and high-income neglected analyzing the competition issues in digital markets, African countries represent over 60 percent of competition especially through the use of competition advocacy tools (see enforcement cases, in terms of advocacy initiatives, lower- Figure 87 and Figure 88). >>> >>> Figure 87: Enforcement cases by income level Figure 88: Advocacy cases by income level classification classification in Africa in Africa High- Upper- income 3% middle income Lower- middle 33% income 38% Lower- Upper- middle middle income income 67% 59% Source: Pop and Coelho (2023),“Competition Policy in Digital Markets in Africa” African competition authorities’ digital economy activities enforcement and competition advocacy initiatives for the are often merger reviews and frequently address sectors period 2008 to 2021, the most common form of engagement most relevant to African economies – digital platforms and by African competition authorities remains merger review mobile money. Based on the WBG desk research on antitrust (46 percent of cases), followed by investigations of abuse 108>>> ADDRESSING NEW GOVERNANCE RISKS FOR DIGITALIZATION: SAFEGUARDING USERS THROUGH DATA AND COMPETITION RULES of dominance cases (29 percent of cases) and competition percent) and software and data services (14 percent) (See advocacy (17 percent of cases, see Figure 89 below). In Figure 90 below). Both the prevalence of merger and abuse of terms of market segments, activities of competition authorities dominance cases and the sectoral distribution of cases might in African jurisdictions tend to focus on digital platforms (37 be explained by the structure of the economies affected as percent of cases), followed by digital financial services (21 well as the capacity of authorities.299 >>> >>> Figure 89: Types of case of antitrust cases and advocacy Figure 90: Number of antitrust cases and advocacy initiatives (% of total) related to the digital economy by segment Advocacy Merger Abuse of dominance Anticompetitive agreement 29% 2 Abuse of dominance Merger 46% 8 2 Advocacy 2 4 17% 3 2 5 1 1 2 2 1 Digital Digital Software & Other Anticompetitive agreement platforms financial data services 8% services Note: ‘Other’ includes cases that relate to all segments of the digital economy, digital advertising and wearables. Source: Pop and Coelho (2023),“Competition Policy in Digital Markets in Africa” ANTICOMPETITIVE PRACTICES through algorithms that are biased towards the platform. Digital platforms with presence in adjacent markets (as is Enforcement of abuse of dominance rules is particularly relatively common given the economies of scope present) important in digital markets since these are often pursued by a can also abuse their dominance in the form of forced tying or small number of vertically integrated firms that control access bundling of products. Furthermore, where a platform controls to the digital ecosystem. Digital platforms have a greater technology or assets that are essential to compete, abuse tendency to tip towards dominance given the prevalence of may arise through refusal to deal or exclusivity conditions.300 network effects and strong economies of scale and scope that arise with high fixed costs/low variable cost structures Among the anticompetitive practices that authorities have and the reliance on data and data-intensive technologies investigated in digital markets, exclusionary conduct301 against to gain a competitive advantage. Given their multisided rival firms is most common in Africa (6 of 12 cases; see Figure nature, dominance on one side of the platform can influence 90 below). Most abuse of dominant cases were brought against anticompetitive behavior on another side of the platform. digital platforms, both domestic and international (Facebook, Digital operators also have the incentive to abuse their market Mediatiz, Bluespec, Computicket, Uber – see figure below), power, especially those that are vertically integrated into and, to a smaller extent, against foreign tech conglomerates product lines where they compete with third parties who sell on (Microsoft, Google – see figure below). their platforms and may engage in self-preferencing, including 299. Nyman & Barajas, Antitrust and Digital Platforms: An analysis of global patterns and approaches by competition authorities. EFI Insight-Trade, Investment and Competitiveness (2021). 300. Nyman & Barajas, Antitrust and Digital Platforms: An analysis of global patterns and approaches by competition authorities. EFI Insight-Trade, Investment and Competitiveness (2021). 301. Standard abuse of dominance analysis focuses on pricing and non-pricing exploitative and exclusionary conduct by firms with market power. Exploitative abuses refer to practices whereby the dominant firm takes advantage of its market power to extract rents from consumers in a way that could not have been obtained by a non-dominant firm (e.g., excessive prices or imposition of unfair terms and conditions). Examples of exploitative abuses of dominance include excessive pricing and/ or imposing unfair prices, terms or conditions to consumers or business partners. Exclusionary abuses, on the other hand, concerns practices directed against rivals that indirectly cause a loss to consumer welfare by limiting the rivals’ ability to compete. Examples of exclusionary abuses of dominance include: (i) single branding (i.e., requiring the buyer on a particular market to concentrate its purchases to a large extent with one supplier) and rebates (conditional rebates that differentiate the purchase price for each customer depending on its behavior), (ii) tying (i.e. making the sale of one product conditional upon the purchase of another distinct product) and bundling (i.e. offering a package of two or more goods either without possibility of buying them separately – pure bundle – or without benefiting from the discount at which the bundle of products is sold – mixed bundling), (iii) refusal to supply (i.e. denying a buyer access to an input in order to exclude that buyer from participating in an economic activity, and (iv) predatory pricing (i.e. lowering the price, in a way that makes the company incurs losses or forego profits in the short run so as to enable it to deter entry or exclude competitors from the market. <<< 109 >>> >>> Figure 91: Types of abuse of dominance cases Figure 92: Types of firms involved in abuse of dominance cases Violation of obligations of firm with SMP, 1 Tech Collusion, conglomerate Digital 2 (foreign), 2 platform Refusal to (domestic), 3 deal, 2 Price gouging, 1 Exclusionary conduct, 5 Digital Predatory platform pricing, 2 Incumbent (foreign), telecom provider 3 (domestic), 2 Source: Pop and Coelho (2023),“Competition Policy in Digital Markets in Africa Exclusivity agreements by digital platforms and mobile money digital markets. Aside from investigating exclusivity obligations, operators – that result in the anticompetitive foreclosure competition agencies have also investigated self-preferencing of competitors – represent the core of abuse of dominance by vertically integrated digital platforms through the use of cases investigated by competition agencies in Africa.302 In fact, algorithms that are biased towards the dominant platform the issue of supplier protection can be especially relevant in (case Bluespec, South Africa). Africa due to the prevalence of SMEs across the value chain, which possibly explains the high number of exclusivity cases. The investigation of potentially anticompetitive agreements Several cases investigated by African competition authorities by African competition agencies has been far rarer thus concern the imposition of exclusivity obligations by dominant far. Collusion scenarios are particularly common in digital firms on their business partners, including real estate agents markets when the pricing is set by the digital platform through that are barred from advertising in rival real estate platforms algorithms. An exception to this trend concerned the acquisition (L’Express Property, Mauritius); or money transfer services of a minority shareholding in a dominant firm in Egypt with the (Safaricom, Kenya), as well as blocking the entertainment aim of eliminating competition in the market (see case Glovo/ industry from using the ticketing services of competitors Delivery Hero, Egypt, in Box 12 below). This case shows how of the dominant firm to competitors of the dominant firm digital platforms can engage in traditional forms of collusion (Computicket, South Africa). All these cases reveal that, digital such as the acquisition of minority shareholdings with the market incumbents often limited the commercial freedom of aim of accessing commercially sensitive information of a their business partners to engage in commercial relations with competitor and of influencing its business behavior. its competitors, thus raising barriers to entry or expansion in BOX 12: ANTICOMPETITIVE HORIZONTAL AGREEMENTS Egypt In 2018, the Egyptian Competition Authority (ECA) accused Glovo and Delivery Hero – two food delivery digital platforms – of agreeing on Glovo’s exit from the Egyptian market to cease competition with Delivery Hero’s brands, thus eliminating any competition between the two companies. 302. Exclusivity agreements generally do not give rise to competition concerns, unless they lead to or contribute to a (cumulative) foreclosure effect on the relevant market where the contract goods or services are sold or purchased. 110 >>> ADDRESSING NEW GOVERNANCE RISKS FOR DIGITALIZATION: SAFEGUARDING USERS THROUGH DATA AND COMPETITION RULES After Glovo entered the Egyptian market in 2018, its competitor, Delivery Hero, acquired 16 percent of Glovo’s shares in the same year. The Shareholders’ Agreement between the parties confers Delivery Hero with certain corporate rights that allow it to access commercially sensitive information about its competitor operations in the Egyptian market and influence Glovo’s strategic business decisions. Hence, the acquisition of a minority shareholding was used as an instrument to collude with Glovo and reach an agreement on the latter’s exit from the market. The ECA issued a cease-and-desist order demanding Glovo to resume its operations in Egypt immediately. Notwithstanding this intervention, in 2018, Delivery Hero acquired 16 percent of Glovo’s shares and influenced the strategic decision, leading to the latter’s market exit. Source: Pop and Coelho (2023) ,“Competition Policy in Digital Markets in Africa”, based on http://www.eca.org.eg/ECA/upload/News/Attachment_E/8285/Glovo%e2%80%99s%20 exit%20from%20the%20Egyptian%20ma rket.pdf The issue of essential facilities303 is key in antitrust cases in digital MERGER CONTROL markets, especially in the African context, as lack of access to essential facilities may prevent smaller firms from competing African competition agencies have assessed the potentially in downstream markets. Access to USSD is a key piece of anticompetitive effects of multiple mergers in digital infrastructure used to provide mobile financial services on nearly markets, and designed procompetitive remedies to offset any phone, at low cost, and without requiring access to the user’s the competition issues identified.306 Through merger review, SIM card. Essential facilities cases may concern refusal to grant African competition agencies have addressed the challenges unstructured supplementary service data (USSD) interoperability posed by: (i) product market definition in digital markets, in (see case Bankers Association, Zimbabwe)304. USSD technology particular, the integration of offline and online businesses in is currently the best available option to serve low-income the same relevant market (MIH eCommerce/Takealot merger, customers, as it consists of a communications service controlled South Africa); (ii) the role of data and algorithms as a barrier by MNOs; hence, it is critical to provide mobile money on most to market entry, which can hinder third-parties’ ability to phones, at low cost, and without requiring access to the user’s compete downstream (see Box 13 below on the Uber/Careem SIM card. A digital platform can also be considered to be an merger, Egypt and Google/Fitbit merger, South Africa); (iii) the essential facility from the perspective of the final user that is importance of protecting users’ privacy in a digital context, blocked from using the dominant platform (see case GovChat, namely through the protection of users’ consent and by South Africa in Box 6 below). enforcing forms of data separation within the merging firm (see Box 13 below on the Google/Fitbit merger, South Africa); Even though most abuses of dominance focused on and the risk of self-preferencing by vertically-integrated digital exclusionary conduct, some competition agencies have platforms operating in two- sided markets (Naspers/We Buy been active in tackling online exploitative conduct, such as Cars merger, South Africa). price gouging, during the COVID-19 pandemic. In Nigeria, for instance, Jumia had to delist multiple protective and hygiene products, including hand sanitizers, where prices had been raised during the pandemic.305 303. An essential facility can consist of an asset, infrastructure, data or technology to which a third party needs to gain access in order to compete on a downstream market. A facility is essential if there are no reasonable alternatives are available and duplication of the facility is not feasible due to legal, economic or technical obstacles. 304. USSD enables customers to send instructions to the mobile money provider along with their personal identification number (PIN) for authentication, while enabling the mobile money provider to send responses to clients and confirm transactions. 305. https://www.fccpc.gov.ng/news-events/releases/2020/03/13/jumia-delists-390-products-on-account-of-fccpc-warning-over- hike-in-prices-of-protective-and-hygiene- products-assures-commission-of-cooperation/. 306. Merger control relates to the procedure of reviewing a transaction that result in a durable combination of previously independent units and have a reasonable likelihood of generating significant harmful impact on the level of market competition. A merger or concentration is typically notifiable to a competition authority ex ante whenever certain thresholds pertaining to the turnover or assets of the parties to the merger are met. In terms of administrative procedure, best practice suggests merger review should include a fast-track procedure or a two-phase procedure with different information requirements. Mergers should be blocked – unless the merging parties submit remedies capable of removing the competition concerns – whenever they result in a substantial lessening of competition in the market that is not offset by efficiency gains that can be passed to consumers. <<< 111 BOX 13: MERGER CONTROL IN AFRICAN DIGITAL MARKETS South Africa, Data Privacy The issue of data privacy was analyzed in South Africa in the context of the Google (big tech international conglomerate active in the online search and advertising markets) /Fitbit (manufacturer of wrist-worn wireless- enabled wearable technology devices that measure fitness) merger. The South African Competition Commission considered the transaction was likely to substantially lessen competition in the market, namely by restricting access to health data collected by Fitbit (Google could have the incentive to limit Fitbit’s interoperability with other operating systems). To alleviate the competition concerns identified by the Competition Commission, Google agreed to a set of commitments for ten years, which led to the merger’s approval in December 2020. The merger commitments included (i) maintaining data separation between Fitbit data and Google’s existing data; (ii) not using any Fitbit body, health, or fitness activity location data in or for Google Ads; and (iii) allowing third-party access to Fitbit data through the Fitbit Web API, without charging and subject to the users’ consent. Egypt, The role of data and algorithms in two-sided markets In the Uber/Careem merger (2019), the Egyptian Competition Authority (ECA) assessed the potentially anticompetitive impact stemming from data control in the ride-hailing market, where both merging parties operate. According to the ECA, data and algorithms for ride-hailing – especially region-specific mapping data owned by Careem – may act as an entry barrier as new market entrants in Egypt would be unable to replicate or access it. Control over Careem’s data would also further strengthen the network effects of the ride-hailing platform the more data the platform can collect. The merger would also consolidate key algorithms in the hands of the merged entity, which new entrants may not be able to match. In addition, the merger would also limit the chances for ‘multi-homing,’ with drivers having fewer ride-hailing platforms available to work for, leaving consumers with fewer options to change platforms should prices increase or quality reduce. Despite the competition risks identified, the ECA cleared Uber’s acquisition of rival ride-hailing app Careem, subject to a package of commitments, which included a commitment to grant future competitors access to Careem’s “static points of interest map data” on a one-time basis, as well as access to trip data, including rider and driver information, subject to data protection law. Source: Pop and Coelho (2023) ,“Competition Policy in Digital Markets in Africa”, based on https://www.docdroid.net/GXSIQ7c/ecas-assessment-of-the-acquisition-of-ca- reem-inc-by-uber-technologies-incnon- confidential1.pdf / http://www.compcom.co.za/wp-content/uploads/2020/12/Competition-Commission-conditionally-approves- the-Goo- gle-Fitbit-merger.pdf / 3.2.2. COMPETITION ADVOCACY and buy-in. The overarching goal is to promote a competition culture across government, market actors, and consumers. Competition advocacy is one of the most effective and important functions of a competition agency. Competition Competition agencies, sector regulators, and other advocacy refers to all of those activities of a competition governmental (and non-governmental) bodies at the national authority that are intended to promote competition but that do and regional levels—acting alone and in collaboration with each not relate to antitrust law enforcement.307 Advocacy activities other—can all play an important part in competition advocacy and tools typically fall into four major categories: (i) provision in Africa. Across a sample of national competition authorities of regulatory advice and opinions (regarding current or draft in Africa surveyed by the African Competition Forum,308 the laws and regulations); (ii) preparation of market studies/ majority are equipped with at least a general mandate on inquiries; (iii) collaboration with sector regulators, often through competition advocacy, in addition to specific advocacy tools memoranda of understanding (MoUs); and (iv) awareness or obligations such as reviewing current or draft laws for their raising activities and capacity building to promote compliance potential competition effects or develop market studies.309 307. International Competition Network (2002), Advocacy and Competition Policy, at 31. 308. The ACF-World Bank report, Institutional Design for Independent and Efficient Competition Authorities. Ensuring Sound Outcomes for Competition Policy in Africa (“ACF-World Bank II”) reviews the status, institutional design, and activities of African competition authorities, with a view to assessing their independence and ability to engage in efficient implementation of competition policy. 112 >>> ADDRESSING NEW GOVERNANCE RISKS FOR DIGITALIZATION: SAFEGUARDING USERS THROUGH DATA AND COMPETITION RULES Implementation of advocacy mandates at the national level is are still nascent, although the Africa Competition Forum has generally lacking, however, and varies significantly. In addition, launched an analysis of roaming and digital markets across only around half of the 20 authorities surveyed keep a public their members. repository of their advocacy initiatives. Regional competition authorities, specifically COMESA, CEMAC, EAC, ECOWAS, Awareness raising and capacity building target sector SADC, and WAEMU, also have competition advocacy regulators - telecommunications sector regulators and others mandates, although only the mandates of the authorities in involved in digital markets, such as data protection agencies, COMESA, CEMAC, and the WAEMU specify that they should typically seek to increase the policymakers’ understanding of review current or draft laws and regulations for their impacts the benefits of competition for their sectors and explain how on competition. In terms of implementation, it appears that to regulate in the least distortive manner. When such activities regional competition authorities play an important role in target the private sector, these efforts can encourage voluntary the region but are generally less active than several of their compliance and reduce risks to consumer welfare. This can national-level counterparts. potentially reduce costs with respect to enforcement or market studies/inquiries in the future and allow competition authorities Regulatory advice and opinions on current or draft laws and to direct their resources elsewhere. Advocacy initiatives intend regulations is one of the key aspects of competition advocacy, to raise awareness and increase transparency on the economic as it aims to eliminate or minimize potential market distortions characteristics of digital markets and emerging technologies that laws and regulations may engender. At the national level, and to better streamline the application of competition rules to such advocacy efforts in the context of digital infrastructure digital markets (e.g., how to better capture aspects such as non- and/or digital markets have been somewhat limited and price dimensions of competition and how to assess dominance focused on mobile money. At the regional level, evidence of in multi-sided markets). At the national level, competition specific regulatory advice and opinions being provided by authorities have a strong track record in awareness-raising regional competition authorities (either for regional or national activities, and increasingly so around the digital markets and level laws/regulations) is limited. However, regional bodies, digital infrastructure. At the regional level, regional competition including telecommunications bodies, have prepared model authorities tend to play an important awareness-raising and laws and guidelines for regulating digital infrastructure and capacity-building function regarding competition law and digital markets, reportedly addressing competition issues, enforcement generally, but their focus on digital markets and access to infrastructure, and spectrum allocation to some digital infrastructure can be further increased. extent, among other issues.310 Model laws and guidelines aim to assist member states as they develop national governance Collaboration between competition authorities and frameworks for digital. telecommunications regulators, or any other government agency with powers over digital markets, is critical to enhance Market studies/inquiries aim to look at market and/or sector the effectiveness and efficiency of their respective actions dynamics to assist with detecting and removing market for the benefit of consumers. This is particularly important for distortions and entry restrictions, thus helping regulators to digital infrastructure and digital markets because, in many pursue pro-competition reforms and set appropriate access countries, telecommunications regulators—not competition conditions, notably in the context of the new challenges authorities—are granted competition enforcement mandates brought about by the digital economy. At the national level, in addition to regulatory power for their sector. At the national such efforts as they relate to digital infrastructure and digital level, collaboration between national competition authorities markets have occurred predominantly in active jurisdictions in and sector regulators is commonly achieved through MoUs, Africa and focus on mobile money services, data portability and and almost 60 percent of African national competition interoperability for financial services, data services, and online authorities have signed MoUs with sector regulators—most of intermediation platforms. At the regional level, such studies/ which have been in the telecommunications, banking, energy, inquiries pertaining to digital infrastructure or digital markets and transport sectors.311 Direct collaboration on specific 309. Countries that responded to the survey that was used to inform the analysis in the ACF-World Bank II report include: Algeria, Angola, Botswana, Burkina Faso, Cameroon, Congo, Egypt, The Gambia, Kenya, Malawi, Mauritius, Morocco, Namibia, Nigeria, Seychelles, South Africa, Eswatini, Tunisia, Zambia, Zimbabwe. 310. See Pop and Coelho (2020) “Getting the Competition Game Right in Mobile Communications and Radio Spectrum in West Africa: An Assessment of Regulatory Restrictions to Competition”, World Bank Group, mimeo; Pop and Coelho (2021), “Competition Across the Radio Spectrum in West Africa: Regional and National Aspects”, Competition Policy International. 311. ACF-World Bank (2016), Breaking Down Barriers, Unlocking Africa’s Potential through Vigorous Competition Policy. <<< 113 digital infrastructure and digital market issues between the competition agencies’ competition law mandate.315 national competition authorities and sector regulators is also Digital businesses’ power to behave unilaterally vis-à-vis the emerging. At the regional level, cooperation between regional consumers by dictating terms and extracting concessions is competition authorities is also possible, though it appears to further accentuated by the information asymmetry between be nascent. Cross-border collaboration, specifically on digital sellers and consumers that characterizes e-commerce and the infrastructure, occurs through regional associations of national web more generally, with consumers constantly presenting with telecommunications regulators, specifically ARICEA, CRASA, complex contractual terms and conditions. This information and WATRA. National telecommunications regulators are also asymmetry within a context of existing market power makes collaborating through MoUs with other national-level entities consumers particularly vulnerable to misleading, aggressive, active in digital markets (such as data protection authorities), or purposefully deceptive commercial conduct.316 It is in this although explicit treatment and consideration of competition context that competition advocacy initiatives assume a key issues appears to be limited. role in re-balancing the position of consumers in relation to online sellers. This can be achieved by identifying the causes Various competition advocacy initiatives undertaken by of potentially abusive behavior and by proposing remedies that African competition agencies show that competition issues in can mitigate the exercise of market power by digital operators. the digital economy are on the radar of African authorities – One particular area of concern is that of mobile money, even where actual enforcement may be lacking. Competition including mobile lending, where, in addition to dealing online, advocacy refers to activities that promote a competitive consumers also face high information asymmetry in relation environment through non-enforcement mechanisms, such to financial firms, which leaves them especially unprotected. as building relationships with government entities, increasing public awareness of competition’s benefits, and identifying African competition authorities use advocacy initiatives to and removing anti-competitive policies and regulations.312 boost competition through consumer behavior (switching Three African competition authorities have engaged in a total and more informed decision-making). In addition to bridging of six advocacy initiatives related to the digital economy,313 information asymmetry between digital platforms and according to recent data covering from 2008- 2021. Of these, consumers, consumer protection initiatives can be embedded two initiatives each target digital platforms, mobile money with a competition angle and result in the adoption of pro- services, and broader aspects of the digital economy. Three competition recommendations. Competition can be weak initiatives are market inquiries, a strategy often used to better when customers find it difficult to switch providers due to lack understand competition-related dynamics and potential issues of information or behavioral biases. Competition authorities in a given sector or market.314 In Africa, market studies have can propose Informational remedies to make it easier for explored the markets for digital platforms (South Africa), customers to find out about products or services in the market e-commerce (Algeria) or digital financial services (Kenya). and to facilitate comparisons. They can also address barriers to switching, such as consumer inertia or lack of familiarity Advocacy initiatives in African digital markets have focused with the switching process and encourage upfront disclosure on issues of consumer protection, which often go beyond of ‘hidden’ charges (Box 14). 312. Advocacy activities and tools typically fall into four major categories: (i) provision of regulatory advice and opinions (regarding current or draft laws and regulations); (ii) preparation of market studies/inquiries; (iii) collaboration with sector regulators, often through memoranda of understanding (MoUs); and (iv) general awareness raising activities and capacity building to promote compliance and buy-in. See: International Competition Network, Advocacy and Competition Policy. A Report Prepared by the Advocacy Working Group. Presented at the ICN Annual Conference, Naples, Italy, 2002. Available: http://www.internationalcompetitionnetwork.org/ uploads/library/doc358.pdf 313. Excluding advocacy initiatives related to telecommunications services. 314. Both the definition and the goal(s) of market studies may differ between competition authorities and cases. For a discussion of the use of market studies by competition authorities globally, see OECD (2016) The Role of Market Studies as a Tool to Promote Competition. Available at https://one.oecd.org/document/DAF/ COMP/GF(2016)4/en/pdf. 315. Consumer protection seeks to protect consumers within the context of individual business to consumer (B2C) transactions, and its rules are applicable to all firms regardless of the existence of market power. Competition law focuses on the protection of the competitive market process, and protects consumers only indirectly. 316. This refers to the ability to by an operator with market power to dictate terms and extract concessions that no one would reasonably consent to in a competitive market: see Subcommittee on Antitrust, Commercial and Administrative Law of the Committee on the Judiciary, Investigation of Competition in Digital Markets, Majority Staff Report and Recommendations (2020). 114 >>> ADDRESSING NEW GOVERNANCE RISKS FOR DIGITALIZATION: SAFEGUARDING USERS THROUGH DATA AND COMPETITION RULES BOX 14: COMPETITION ADVOCACY FOR PROTECTING CONSUMERS IN DIGITAL MARKETS Kenya The Competition Authority of Kenya’s (CAK) study on the digital lending markets in Kenya aimed at ensuring digital lenders increase transparency of the products they offer consumers (2019). In the long run, CAK hopes to develop policies that will ensure digital loanees are protected across regulated and unregulated lenders. CAK has also developed an advocacy program to encourage disclosure and transparency in digital financial services through the engagement of mobile phone operators, banking institutions, and micro-finance institutions (2018). Furthermore, it carried out an end-line survey on the price awareness levels of consumers of digital financial services. Source: Pop and Coelho (2023) based on WBG Advocacy contest submission (2018) / https://www.the- star.co.ke/business/kenya/2020-02-24-competition-authority-of-ken- ya-seeks-to-regulate-vast-digital-lending-sector/ Other initiatives sought to raise awareness and increase market features of digital platforms (see Online Intermediation transparency on the economic characteristics of digital Platforms Market Inquiry, South Africa, in Box 15 below), or markets and emerging technologies and how they might (iii) the overall approach by the competition agency to the impact the application of competition law. These initiatives enforcement of competition and consumer rules in the digital aimed to upgrade the methodology applicable to (i) market economy (see report on Competition in the Digital Economy, definition in two-sided and free markets in order to capture South Africa in Box 15). These initiatives clearly show the dominance in such markets and design a theory of harm importance of tailoring the enforcement of the competition that takes into consideration the non-price dimensions of rules to the digital economy without having necessarily to competition (see Guidelines on Market Definition in Box 10 amend the existing regulatory framework and to convey the below), (ii) online intermediation platforms, bearing in mind the agency’s approach to the business community. BOX 15: COMPETITION ADVOCACY THAT CONTRIBUTED TO STREAMLINING THE APPLICATION OF COMPETITION LAW TO DIGITAL MARKETS South Africa The South African Competition Commission announced in 2020 that it will conduct a market inquiry into online intermediation platform services (the “Online Intermediation Platforms Market Inquiry” or OIPMI). The market inquiry is driven by the existence of market features of online intermediation platforms, such as first-mover advantages, network effects of two-sided markets, and product development advantages from data accumulation; in addition, firm behavior, such as most-favored-nation(MFN) pricing rules317, acquire competitive threats (so- called ‘killer acquisitions318) and leverage dominance in some areas to exclude or limit rivals in others (such as self- preferencing of data and platform access). These market features require competition law to be streamlined in order to consider new theories of harm and to act against entrenchment strategies to ensure concentration does not impair market contestability. The South African Competition Commission elaborated a report on “Competition in the Digital Economy,” providing a review of the emerging competition issues in e-commerce, consumer empowerment, and guiding businesses on the approach adopted by the Commission. The Report analyzes how South Africa’s competition laws can be implemented to achieve equitable outcomes in the digital economy and our intentions in this regard. It also closely examines the leading digital disrupters and their impact on established industries. The publishing of the report was accompanied by a virtual dissemination event in November, 2020. Source: Pop and Coelho (2023) based on https://www.cak.go.ke/sites/default/files/Guidelines percent20on percent20Relevant percent20Market percent20Definition percent20(1). pdf / http://www.compcom.co.za/online-intermediation- platforms-market-inquiry/ / http://www.compcom.co.za/wp-content/uploads/2020/11/COMPETITION-COMMISSIONER-TO- LAUNCH-THE-COMMISSION percentE2 percent80 percent99S-COMPETITIVE-STRATEGY-IN-SOUTH-AFRICA percentE2 percent80 percent99S-DIGITAL-ECONOMY.pdf 317. MFN are vertical restraints where the terms and condition of supply are set by reference to the terms and conditions applicable to other supply relationships. ‘Wide’ MFN requires retailers to publish on a price comparison tool or online marketplace the same or better price and conditions as those published on any other sales channel. ‘Narrow’ MFN, on the other hand, require retailers to publish on a price comparison or online marketplace the same or better price and conditions as those published on its own platform. 318. Shutting down competition even before there is a marketable product <<< 115 4 Regulatory initiatives with competition implications can emerge as a follow-up from enforcement or advocacy by the competition agencies, even though it is important to assess the pros and cons of Government interventions in the market and their effects. Even in the absence of a decision finding an infringement of the competition rules, Governments have intervened to attempt to create a level playing field in the digital market. For example, in 2016 the South African Metered Taxi Industry brought a complaint before the Competition Commission, alleging Uber abused its dominant position by engaging in predatory pricing practices. The South African Competition Commission did not ascertain the existence of an abuse by Uber and decided not to prosecute it. Meanwhile, South Africa created an independent body to regulate the electronic ride hailing sector in South Africa. In sum, governments are using competition policy tools to address issues regarding market power in African economies, but this is a nascent trend. Less than half of African countries have an operating competition authority, limiting the potential for antitrust to address market power issues brought by digital technologies. 116 >>> STRENGTHENING ADDRESSING NEWECONOMIC GOVERNANCE GOVERNANCE RISKS FOR DIGITALIZATION: OF SAFEGUARDING USERS THROUGH DATA AND COMPETITION RULES THE DIGITAL ECONOMY 4 >>> Strengthening economic governance of the digital economy There are countries that are at higher risk of weak economic governance of the digital economy based on the factors analyzed in previous chapters. This section provides a heatmap of risks considering first-generation risks - market structure, traditional digital institutions (digital infrastructure regulation and interinstitutional collaboration), taxation of digital infrastructure, SOE-SLEs operation and their transparency, and second-generation risks - data regulation and cybersecurity, competition enforcement, and advocacy. The risk levels were calculated using proxy variables, and therefore, the results are intended to provide initial insights on areas of attention for further assessment based on the country conditions. The methodology for the risk assessment and underlying country variables are included in Annex 3. >>> Figure 93: Risk to economic governance of 49% 47% 49% 66% Average governance risk digital 63% 68% 61% 54% 43% 45% 44% 34% 43% 51% 49% 55% 22% 49% 57% 62% 84% 62% 76% 28% 69% 73% 82% 67% 76% 48% 64% 54% 56% 29% 59% 53% 39% 55% 68% 44% 22% 32% 34% 47% 68% 59% 50% 49% 45% 72% 84% 62% 47% 49% No data 33% 48% 34% 52% 26% 60% Source: Authors based on a series of indicators that proxy risks, including the World Bank Africa databases on digital market players, digital infrastructure taxes IBRD 47465 | and parafiscal fees, and digital SOE-SLEs, among others. AUGUST 2023 <<< 117 The level of risks to good economic governance varies across framework and enforcement seem to be higher, especially countries, as well as the main driver of risks. Among first- given a lack of legal frameworks or authority to boost and generation risks - related to lack of enabling market structure, preserve competition economy-wide. Even if incipient, weak traditional digital infrastructure institutions, over taxation most countries have already included partial frameworks or of digital connectivity and operation of digital SOE-SLEs), measures related to data protection or cybersecurity. The market structure and operation of digital SOE-SLEs place the inherent characteristics of digital service markets that tend highest risks on average in Africa. Therefore, it is important to to winner-takes most and the rapid pace of innovation and strengthen regulations that allow for more competitive market creation of different business models call for more effective structures, especially in digital infrastructure, and that create antitrust enforcement and competition policy, especially a level playing field independently of ownership, including in countries where digital services as substituting offline through better corporate governance and transparency. markets. More effective implementation and cross-country Among second-generation risks – related to data regulation collaboration is needed on data protection and cybersecurity, and competition - that become more relevant with a higher given the gains from cross-border data flows and the single level of digitalization, risks related to weak competition policy digital market in Africa. >>> Table 9: Level of risks to economic governance 118 >>> STRENGTHENING ECONOMIC GOVERNANCE OF THE DIGITAL ECONOMY Source: Authors based on a series of indicators that proxy risks, including the World Bank Africa databases on digital market players, digital infrastructure taxes and parafiscal fees, and digital SOE-SLEs, among others. Higher level of risks to economic governance of digital is -0.48 and -0.42, respectively. Correlation with the uptake gap associated with poorer digitalization outcomes (Figure 94). (difference between availability and adoption) is lower (-0.21), Unconditional correlations of the level of economic governance suggesting that demand-side factors such as disposable risk with digital technologies availability (proxied by population income and skills/capabilities are conditioning the importance covered by 3G or 4G networks) and digitalization (proxied of economic governance risks for uptake. by unique mobile internet subscriber per 100 inhabitant) are >>> Figure 94: Risk to economic governance of digital and digitalization Source: Authors based on a series of indicators that proxy risks (see Annex 3) and GSMA (2021) <<< 119 The key recommendations to address the regulatory risks of countries relatively higher risk in a given area compared to identified are presented in the table below, including examples other risks in the same country. >>> Table 10: Risks to economic governance and policy recommendations Examples of countries with Risk area and relatively higher economic governance Policy recommendations to mitigate each type of risks risk in this area issues compared to other risks First-generation risks Lack of enabling • Introducing reforms to facilitate entry and expansion of smaller/new players Cabo Verde, market structure • Implementing pro-competition rules included in the sectoral framework Chad, Congo • Ensuring appropriate competition assessment in the evaluation of merger and Rep, Ghana, Economic governance acquisitions Mauritania, issue: Concentration • Enhancing technical independence of digital infrastructure regulators and Namibia, of economic power means for collaboration with other authorities Rwanda, that can lead to • (competition authority when operational) Togo, Uganda, regulatory capture Zimbabwe Weak traditional • Institutional design reforms and practices to enhance technical independence Burundi, Central digital infrastructure of regulator, separating functions of various entities in the digital infrastructure African Republic, institutions sector Djibouti, • Capacity building on areas related to economic regulation to enable healthy Eritrea, Libya, Economic governance digital infrastructure as foundation for the digital economy Madagascar, issue: Weak • Deepen stakeholder consultation and transparency of decisions and motivation Mozambique, accountability, • More specific guidance for the consistent and predictable implementation of the Sao Tome transparency, legal framework and Principe, independence, • Benchmarking across members of regional blocs and continental level to Seychelles, and stakeholder encourage more effective implementation Sudan engagement leading to regulatory capture Over taxation of digital • Re-assess the tax and parafiscal fees policy by identifying and systematizing Cameroon, connectivity current taxes and fees, quantifying tax, and fees collections, and analyzing tax/ Cote d’Ivoire, fees incidence and their economic and social costs. Madagascar, Economic governance • Phase out excise-like taxes on digital connectivity that distort consumer Senegal, issue: Appropriation of behavior and investment decisions across sectors and countries. Use Zimbabwe sector resources and economy-wide tax expenditure assessments to identify less-distortive more opportunities for measures for revenue mobilization. rent seeking by public • Ensure an equitable tax framework by minimizing over-taxation that affects the officials less well-off including the unconnected, and replace blanket tax exemptions by support measures targeted at the most vulnerable. • Ensure a simple and transparent legal and regulatory framework with clear provisions on the taxes and fees applicable to all digital infrastructure operators • Maximize stakeholder involvement and prioritize capacity building for tax and fees design and administration • Streamline the national tax framework in line with the global tax policy reforms of the digital economy (e.g., BEPS) 120 >>> STRENGTHENING ECONOMIC GOVERNANCE OF THE DIGITAL ECONOMY Examples of countries with Risk area and relatively higher economic governance Policy recommendations to mitigate each type of risks risk in this area issues compared to other risks First-generation risks Operation of digital • Monitor and enhance transparency of both financial and non-financial Algeria, Angola, SOE-SLEs disclosures by SOE and SLEs Benin, Botswana, • Mitigate conflict of interest by separating the body that exerts control and/or Cameroon, Economic governance the board member that represents the state at the SOE-SLE from policymaking Chad, Comoros, issue: Conflict and regulatory entities and their employees. Congo DR., of interest in the • Enhance transparency and control of government support provided to SOE and SLE Egypt, Equatorial treatment of SOE-SLE • Require SOEs to separate commercial activities from the delivery of any public Guinea, Eswatini, and rent-seeking by service obligations (PSOs), and competitive and non-competitive services, and use Ethiopia, The public officials appropriate accounting mechanisms to minimize the risk of cross-subsidization. Gambia, Guinea, • Require digital sector SOEs to achieve commercial rates of return similar to Kenya, Libya, comparable private businesses. Morocco, • Remove any de jure monopolies granted to SOEs over the ownership and Mozambique, operation of digital infrastructure and the provision of digital services. Namibia, Niger, • Remove preferential tax rates or forbearance on tax arrears and penalties, Rwanda, Sierra revise provisions in national budget laws that provide for direct on-lending Leone, Somalia, to SOEs, adjust state-backed loan guarantee programs and other forms of South Africa, preferential financing for SOE Sudan, Tanzania, Togo, Tunisia Second-generation risks Data misuse • Adopt economy-wide legal frameworks for data protection and cybersecurity Botswana, and cybercrime Ethiopia, Economic governance • Support the operationalization of data protection authorities (or entities with this Eswatini, issue:Political mandate) and cybersecurity systems, including through regulatory measures, The Gambia, or commercially financial resources, and capacity building DRC, Namibia, motivated surveillance • Provide clear guidelines for compliance and flexibilities within the law, such as Tanzania, South exceptions for MSMEs or ex post facto liability for certain infringements Africa (among • Engage in regional discussions and agreements about the interoperability of data countries with protection rules and collaboration for the implementation of cybersecurity frameworks more than 20 to support trusted cross-border data flows; adopt international conventions unique mobile • Create enabling frameworks to facilitate use and re-use of data collected by subscribers per public and private entities 100 inhabitants) Anticompetitive • Integrate pro-competition principles in the rules that govern competition in Benin, Ghana, outcomes digital markets Rwanda, Uganda Economic governance • Adopt or implement competition law frameworks more effectively (among countries issue: Biased (unfair) • Build human capital and equip competition authorities with sufficient resources with more than outcomes for users of to pursue competition enforcement and advocacy in a meaningful manner in 20 unique mobile digital technologies digital markets subscribers per • Bolster de jure and de facto advocacy mandates and capacities of regional 100 inhabitants) competition authorities and engage in implemenation under the African Continental Free Trade Agreement (AfCFTA) and the Tripartite Free Trade Area, and foster international cooperation • Adjust current competition frameworks through regulation and guidelines to adapt it to the monitoring, assessment and design of remedies for anticompetitive practices and mergers involving digital markets. Source: Authors <<< 121 Looking into the future, additional research and analysis is needed to highlight the importance of better economic governance of the digital economy. More in-depth research is required into what would be the best practices to promote open and competitive environments and more effective implementation of regulation. For example, limited empirical evidence is available on how changes in competition intensity, implementation of certain regulations on digital infrastructure, levels of taxation and parafiscal fees, the presence of SOEs and SLEs, enforcement of competition rules in digital markets and data regulations would impact the performance of digital markets, the adoption of digital technologies and the effects of digitalization. Conducting empirical analyses and in-depth case studies on these areas would give policymakers and regulators a better understanding of valuable strategies to promote a more dynamic digital environment and overcome political economy issues. Expanding data collection and dissemination of databases is essential to enable this research and inform better policy and regulatory decisions. Important variables include characteristics of regulations that affect digital markets and their degree of implementation, market characteristics indicators, levels of digital taxation and the importance of taxation for revenue mobilization, and corporate governance, financial and performance indicators for SOEs and SLEs, among others. 122 >>> REFERENCES STRENGTHENING ECONOMIC GOVERNANCE OF THE DIGITAL ECONOMY >>> References ACF-World Bank. 2022. “The Institutional Gauge: Ensuring Sound Outcomes for Competition Policy in Africa”. World Bank. Unpublished Act of the Parliament of the Republic of Ghana. 2019. “Communication Service Tax (Amendment) Act, 2019”. Act 998. August 19, 2019. http://elibrary.jsg.gov.gh/fg/laws percent20of percent20ghana/2 percent20REP/Communications percent20Service percent20Tax percent20Amendment percent20Act,2019 percent20Act percent20998.pdf Acts Supplement. 2018. “Excise Duty(Amendment) No. 2, 2018”. Uganda Gazzete N° 58. https://old.ulii.org/system/files/ legislation/act/2018/12/Scan percent204.pdf Adepoju, Paul. 2019. “South Sudan launches first international voice gateway”. IT Web Africa. Last modified May 29, 2019. https://itweb.africa/content/G98YdqLY3z2vX2PD Advocacy Working Group. 2002. “Advocacy and Competition Policy”. International Competition Network. Naples. Italy. https://www.internationalcompetitionnetwork.org/wp- content/uploads/2018/09/AWG_AdvocacyReport2002.pdf Africa Bandwidth Maps. 2021. “Africa’s International Bandwidth Reaches 15.289 Tbps”. Africa Telecom Transmission Map Published. 2020/ 2021. http://www.africabandwidthmaps.com African Development Bank Group. 2017. “EASSy: The Eastern Africa Submarine Cable System”. October 20, 2017. https:// www.afdb.org/en/projects-and-operations/selected-projects/eassy-the-eastern-africa- submarine-cable-system-156 African Union. 2014. “Legal Framework for Cyber-security and Personal Data protection” https://au.int/sites/default/files/ treaties/29560-treaty-0048_-_african_union_convention_on_cyber_security_and_personal_data_protection_e.pdf African Union. 2020. “African Union Convention on Cyber Security and Personal Data Protection”. May 11, 2020. https://au.int/ en/treaties/african-union-convention-cyber-security-and-personal-data-protection African Union. 2021. “The African Union Convenes The Second African Cybercrime Forum”. Last updated July 08, 2021. https:// au.int/en/pressreleases/20210708/african-union-convenes-second-african- cybercrime-forum Agência Nacional das Comunicações (ANAC) of Cabo Verde. n. d. Accessed 24.04.2022. http://www.anac.cv/index. php?option=com_content&view=article&id=87&Itemid=56&lang=en Akol, Doris and Lees Adrienne. 2021. “There and Back Again: The Making of Uganda’s Mobile Money Tax”. ICTD Working Paper 123. July, 2021. https://opendocs.ids.ac.uk/opendocs/bitstream/handle/20.500.12413/16727/ICTD_WP123.pdf?sequenc e=1&isAllowed=y Antelope Consulting. 2018. “Discussion on the numbering regulation for Somalia”. February 20, 2018. http://public. antelopeweb.fmail.co.uk/publications/Discussionpercent20onpercent20the percent20numbering percent20regulation percent20for percent20Somalia.pdf. <<< 123 Apptopia. 2021. “Worldwide and US Download Leaders 2021”. https://blog.apptopia.com/worldwide-and- us-download- leaders-2021 Arizton Advisory & Intelligence. 2021. “Nigeria Data Center Market Size by Investment to Reach USD 218 Million by 2026 - Arizton”. Cision PR Newswire. December 08, 2021. https://www.prnewswire.com/news-releases/nigeria-data-center- market-size-by-investment-to-reach- usd-218-million-by-2026--arizton-301440332.html ATAF. 2019. “Africa Tax Outlook 2019”. September, 2019. https://events.ataftax.org/includes/preview.php?file_ id=49&language=en_US Atiyas, Izak and Mark A. Dutz (2022). “Digital Technology Uses among Microenterprises: Productivity and Jobs Outcomes across sub-Saharan Africa”, mimeo. Australian Competition and Consumer Commission. 2015. “Online Vertical Restraints Special Project Report”. 2015 International Competition Network Annual Meeting (ICN 2015). https://www.internationalcompetitionnetwork.org/wp-content/ uploads/2019/11/SP_OnlineVR2015.pdf. Authority of the Republic of Kenya. 2018. “The Kenya Gazette”. Supplement No. 116 (Acts No. 10)“. Nairobi. September 21st, 2018. https://www.kra.go.ke/images/publications/Finance-Act-2018.pdf Authority of the Republic of Kenya. 2021. “The Kenya Gazette”. Vol. CXXIII—No. 141. Nairobi. July 2nd, 2021“. http://kenyalaw. org/kenya_gazette/gazette/volume/MjM4NA--/Vol.CXXIII-No percent20141/ Aviomoh, Blimpo, 2022 unpublished, Datacenters and ISP in Africa Bahia Kalvin, Castells Pau and Pedros, Xaver. 2020. “Mobile technology: two decades driving economic growth.” Working Paper. GSMA Intelligence. November, 2020. https://data.gsmaintelligence.com/api- web/v2/research-file- download?id=54165922&file=121120-working-paper.pdf Bahia, Kalvin; Castells, Pau; Cruz, Genaro; Masaki, Takaaki; Rodriguez-Castelan, Carlos; Sanfelice, Viviane. 2021. Mobile Broadband Internet, Poverty and Labor Outcomes in Tanzania. Policy Research Working Paper;No. 9749. World Bank, Washington, DC. http://hdl.handle.net/10986/36172. Bannerman, Natalie. 2021. “ASN starts construction on Africa-1 subsea cable”. Capacity, February 03, 2021. https://www. capacitymedia.com/article/29otcvon80h8p2ay7rnr4/news/asn-starts-construction-on- africa-1-subsea-cable Bannerman, Natalie. 2020. “Orange launches Djoliba: the first pan-West African fibre network”. Capacity, November 10, 2020. https://www.capacitymedia.com/article/29otcq7s58uymwyj1xtds/news/orange- launches-djoliba-the-first-pan-west- african-fibre-network. Baraka, Providence. 2020. “New tax on mobile devices threatens digital inclusion in the Democratic Republic of Congo”. Global Voices. Last updated December 30, 2020. https://globalvoices.org/2020/12/30/new-tax-on-mobile-devices-threatens- digital-inclusion-in-the- democratic-republic-of-congo/ Begazo T., Dutz M. A., and Blimpo M. P. 2023. “Digital Africa: Technological Transformation for Jobs”. https://www.worldbank. org/en/region/afr/publication/digital-africa Ben Ali, Zine El Abidine. 2001. “Loi n° 1-2001 du 15 janvier 2001 portant promulgation du code des télécommunications”. Le Code des Télécommunications. January 15, 2001. WIPO Lex No.TN049. https://wipolex.wipo.int/en/text/204159 124 >>> REFERENCES Bennett, Colin J. and Raab, Charles. 2006. “The Governance of Privacy Policy Instruments in Global Perspective”. May, 2006. https://mitpress.mit.edu/books/governance-privacy Blackman, Colin and Srivastav, Lara. 2011. z’Telecommunications Regulation Handbook : Tenth Anniversary Edition”. World Bank and the International Telecommunication Union. 2011. https://openknowledge.worldbank.org/handle/10986/13278 Boik Andre and Corts Kenneth. 2016. “The Effects of Platform MFNs on Competition and Entry”. Research Gate. February, 2016 https://www.researchgate.net/publication/305220937_The_Effects_of_Platform_MFNs_ onCompetition_and_Entry Brickner, Taryn. 2020. “Closing Africa´s Ingrastructure gap with sustainability at the heart of Helios Towers”. Racounter. March 17, 2020. https://www.raconteur.net/sponsored/closing-africas-infrastructure-gap- with-sustainability-at-the-heart-of- helios-towers Brown Felicity, Metcalfe Daniel, Stacey Oliver and Taylor, Martyn. 2016. “Sharing of Telecoms Infrastructure – opportunities in 2016”. May, 2016. https://www.nortonrosefulbright.com/en/knowledge/publications/48f82d8d/sharing-of-telecoms- infrastructure---opportunities-in-2016#section2 Brun Jean - Francois, Chambas Gerard and Mansour Mario. 2015. “Tax Effort of Developing Countries: An Alternative Measure”. Chapter 11 in Financing Sustainable Development - Addressing Vulnerabilities, Boussichas, M., Guillaumont, P. (eds.), Ferdi, Economica, Paris. https://ferdi.fr/en/publications/tax-effort-of-developing-countries-an-alternative-measure BSA. 2017. “Cross-border data flows”. https://www.bsa.org/files/policy-filings/BSA_2017CrossBorderDataFlows.pdf Builtwith.com . “Find out what websites are Built With”. https://builtwith.com/ Burkina Faso Investment Code. 2018. “Burkina Faso 2018-10-30 loi 038 - 2018 portant code des investissements”. https:// investmentpolicy.unctad.org/investment-laws/laws/276/burkina-faso-burkina- faso-investment-code-2018- Business Wire. 2020. “American Tower Closes Eaton Towers Transaction and Announces Agreement to Acquire Joint Venture Stakes in Ghana and Uganda From MTN”. January 02, 2020. Boston. https://www.businesswire.com/news/ home/20200102005240/en/American-Tower-Closes-Eaton- Towers-Transaction-and-Announces-Agreement-to- Acquire-Joint-Venture-Stakes-in-Ghana-and-Uganda- From-MTN Business Wire. 2021. “Webscale Network Operators, 4Q20 Market Review…”. April 15, 2021. Dublin. https://www. businesswire.com/news/home/20210415005461/en/Webscale-Network-Operators-4Q20- Market-Review---Webscale- Sector-Benefits-from-COVID-19-Pandemic-as-Annual-Revenues-Rise-18-to- 1.71-Trillion-CAPEX-Up-28-to-133B--- ResearchAndMarkets.com Cebreiro-Gomez Ana, Clavey Colin, Estevao Marcello, Leigh-Pemberton Jonathan and Stewart, Benjamin. 2022. “Digital Services Tax : Country Practice and Technical Challenges”. Equitable Growth, Finance and Institutions Insight. World Bank, Washington, DC. https://openknowledge.worldbank.org/handle/10986/36840 Chen, Rong. 2021. Mapping Data Governance Legal Frameworks around the World: Findings from the Global Data Regulation Diagnostic. Policy Research Working Paper 9615. https://doi.org/10.1596/1813-9450-9615 Chen, Daza Jaller and Begazo. 2023. Data protection regulation in Africa. <<< 125 Chibuluma, Matthews; Chihimba, Mumbi; Fainess, Mbale; Tozgani, Fainess Mbale and Zimba, Aaron. 2020. ResearchGate. January,2020. “Liberalisation of the International Gateway and Internet Development in Zambia: The Genesis, Opportunities, Challenges, and Future Directions”. January,2020. https://www.researchgate.net/publication/341600351_Liberalisation_ of_the_International_Gateway_an d_Internet_Development_in_Zambia_The_Genesis_Opportunities_Challenges_ and_Future_Directions/lin k/5ec9772c458515626cc6b7ea/download Cirera, X., Comin, D., and Cruz, M. (2022). Bridging the Technological Divide: Firm Technology Adoption in Developing Countries. Washington, DC: World Bank. http://hdl.handle.net/10986/37527. Cisco.2021. “What Is a Data Center”. 2021. https://www.cisco.com/c/en_uk/solutions/data-center- virtualization/what-is-a- data-center.html/ Clavey Colin, Pemberton Jonathan Leigh, Loeprick Jan and Verhoeven Marinus. 2019. “International Tax Reform, Digitization, and Developing Economies”. Open Knowledge Repository. October, 2019. https://openknowledge.worldbank.org/ handle/10986/32530 Clement, Jessica. 2021. “Mobile internet traffic as percentage of total web traffic in August 2021, by region”. Statista. September 24, 2021. https://www.statista.com/statistics/306528/share-of-mobile- internet-traffic-in-global-regions/ Clement, Jessica. 2021. “Distribution of web traffic in selected African countries as of August 2021, by device”. Statista. Last updated September 30, 2021. https://www.statista.com/statistics/685188/african- countries-online-traffic-channel-share/ Comini Niccolo, Minges Michael and Srinivasan Sharada. 2021. “The Importance of National Data Infrastructure for Low and Middle-Income Countries”. August 2, 2021. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3898094 Communications Authority of Kenya, n. d. Accessed 24.04.2022. https://www.ca.go.ke/ Competition Authority of Kenya, undated. “Competition Authority of Kenya Revised Guidelines on Relevant Market Definition” https://www.cak.go.ke/sites/default/files/Guidelines percent20on percent20Relevant percent20Market percent20Definition percent20(1).pdf Competition Commission of South Africa. 2020. “Competition Commissioner to launch the Commission´s Competitive Strategy in South Africa´s Digital Economy”. November 04, 2020 http://www.compcom.co.za/wp-content/uploads/2020/11/ COMPETITION-COMMISSIONER-TO-LAUNCH- THE-COMMISSIONpercentE2percent80percent99S-COMPETITIVE- STRATEGY-IN-SOUTH-AFRICApercentE2 percent80 percent99S-DIGITAL-ECONOMY.pdf Competition Commission of South Africa. 2019. “Data Services Market Inquiry”. December 1 2019. https://www.compcom. co.za/wp-content/uploads/2019/12/DSMI-Non-Confidential-Report-002.pdf. Corfield, Gareth. 2020. “Personal data from Experian on 40 percent of South Africa’s population has been bundled onto a file- sharing website”. The Register. Last updated September 14, 2020. https://www.theregister.com/2020/09/14/south_africa_ experian_data_breach_wesendit/ Cory, Nigel und Dascoli, Luke. 2021. “How Barriers to Cross-Border Data Flows Are Spreading Globally, What They Cost, and How to Address Them”. In Information Technology and Innovation Foundation. July 19, 2021. https://itif.org/ publications/2021/07/19/how-barriers-cross-border-data-flows-are-spreading-globally- what-they-cost 126 >>> REFERENCES Council of Europe. 1985. “Convention for the Protection of Individuals with regard to Automatic Processing of Personal Data (ETS No. 108) ”. Council of Europe. October 10, 1985. https://www.coe.int/en/web/conventions/full-list?module=treaty- detail&treatynum=108 Crivelli, Ernesto. and Gupta, Sanjeev. 2014. “Resource Blessing, Revenue Curse? Domestic Revenue Effort in Resource-Rich Countries”, IMF Working Paper 14/5, IMF, Washington DC https://www.imf.org/external/pubs/ft/wp/2014/wp1405.pdf Daily News Egypt. “Egyptian Competition Authority accuses Glovo, Delivery Hero of restricting competition”. Daily News Egypt. May 28, 2019. https://dailynewsegypt.com/2019/05/28/egyptian- competition-authority-accuses-glovo-delivery-hero-of- restricting-competition/ Data Center Map. 2022. “Data center”. Last updated on 06. 02.2022. https://www.datacentermap.com/germany/frankfurt/ lyonerstrasse-28_connectivity.html Davidson, Carl and W. Martin, Lawrence. 1985. “General Equilibrium Tax Incidence under Imperfect Competition: A Quantity- setting Supergame Analysis”. In Journal of Political Economy Vol. 93, No. 6 (December, 1985), pp. 1212-1223. University of Chicago Press. https://www.jstor.org/stable/i331953 Deloitte. 2022.“ Tax guides and highlights“. https://dits.deloitte.com/#TaxGuides Dener Cem, Nii-Aponsah Hubert, Ghunney Love. E., and Johns Kimberly. D. 2021. GovTech Maturity Index: The State of Public Sector Digital Transformation. International Development in Focus. Washington, DC: World Bank. doi:10.1596/978-1-4648- 1765-6: Deskeco. 2020. “RDC : la nouvelle taxe de l’ARPTC sur la certification des Smartphones vient frapper le faible pouvoir d’achat de la majorité des Congolais”. Desk Eco. Last updated April 27, 2020. https://deskeco.com/2020/04/27/rdc-la-nouvelle- taxe-de-larptc-sur-la-certification-des-smartphones- vient-frapper-le Díaz Hernández, Marianne, Rafael Nunes, Felicia Anthonio, and Sage Cheng. 2021. “#KeepItOn update: who is shutting down the internet in 2021?” Accessnow. June 7. Accessed August 5, 2021. https://www.accessnow.org/who-is-shutting-down- the-internet-in-2021/. Digital Africa. 2021. “Djibouti-Crossroads of Submarine Cables”. Last updated February 3, 2021. https://resilient.digital-africa. co/en/blog/2021/02/03/djibouti-crossroads-of-submarine-cables/ 2021 Report”. https://www.wcoesarpsg.org/ Digital Council. 2021. “Africa Digital Infrastructure Market Analysis wp-content/uploads/2021/11/juanita-clark-africas-evolving-digital- landscape.pdf Direction générale des Impôts (DGI). 2021. “Impots Et Taxes En Cote D´Ivoire”. Ministere Du Budget Et Du Portefeuille De L´etat. Les Publications de la DGI –Edition 2021. https://www.dgi.gouv.ci/images/PDF/impot_et_taxes.pdf Dixit, Avinash K. 2008. “Economic Governance”. In: Palgrave Macmillan (eds) The New Palgrave Dictionary of Economics. Palgrave Macmillan, London. https://doi.org/10.1057/978-1-349-95121-5_2611-1 Donkin, Chris. 2021. “Ghana completes acquisition of AirtelTigo”. Mobile World Live. November 03, 2021. https:// www.mobileworldlive.com/featured-content/top-three/ghana-completes-acquisition-of- airteltigo#:~:text=Bharti percent20Airtel percent20confirmed percent20full percent20ownership,aim percent20to percent20leave percent20the percent20market <<< 127 Egyptian Competition Authority. 2019 “ECA’s Assessment of the Acquisition of Careem, Inc. by Uber Technologies, Inc.”. December 19, 2019. Cairo. https://www.docdroid.net/GXSIQ7c/ecas-assessment-of- the-acquisition-of-careem-inc-by- uber-technologies-incnon-confidential1-pdf Emejo, James. 2020. “COVID-19: Jumia Delists 390 Products over Price Manipulations”. In This Day News, March 12, 2020. https://www.thisdaylive.com/index.php/2020/03/12/covid-19-jumia-delists-390- products-over-price-manipulations/ Enterprise. 2020. „Could Egypt make better use of its position as a subsea internet cable hub?“. July 15, 2020. https:// enterprise.press/hardhats/egypt-make-better-use-position-subsea-internet-cable-hub/ Ernst & Young LLP. 2018. “Reforming mobile sector taxation in the Democratic Republic of the Congo: Enabling economic growth through a supportive tax system”. 2018. https://www.gsma.com/subsaharanafrica/wp-content/uploads/2018/11/ GSMA_DRC- report_ENGLISH_72pp_WEB.pdf Ernst & Young LLP. 2020. “Mobile taxation in Kenya: Accelerating digital development”. GSMA. March, 2020. https://odi.org/en/ publications/how-tax-officials-in-lower-income-countries-can-respond-to-the- coronavirus-pandemic/ EY Global Tax Alerts Egypt issued VAT Law no. 67, 2016 on 7 September 2016 in the Official Gazette to be effectively applied from 8 September 2016.1. https://www.flandersinvestmentandtrade.com/export/sites/trade/files/attachments/EY percent20- percent20Egypt percent20issues percent20VAT percent20Law percent20no. percent2067 percent20of percent202016.pdf EY organization. 2020. „Worldwide Corporate Tax Guide 2020“. https://assets.ey.com/content/dam/ey- sites/ey-com/en_am/ tax-and-law/ey-worldwide-corporate-tax-guide-20-july-2020.pdf Federal Competition and Consumer Protection Commission (FCCPC). 2020. “Jumia delists 390 products on account of FCCPC warning over hike in prices of protective & hygiene products, assures Commission of cooperation”. March 13, 2020. https:// www.fccpc.gov.ng/news-events/releases/2020/03/13/jumia- delists-390-products-on-account-of-fccpc-warning-over- hike-in-prices-of-protective-and-hygiene- products-assures-commission-of-cooperation/ Feldman, Sarah. 2019. “Governments Are Shutting Down the Internet More INTERNET ACCESS.” Statista.com. April 24. Accessed August 5, 2021. https://www.statista.com/chart/17794/number-of- internet-shutdowns-by-government/. Financial Times. 2020. “Towers of power: European telcos find value in masts”. August 2, 2020 https://www.ft.com/content/ ffe9c3d8-f28c-4508-8c24-f18a7e44dd00 Global Cyber Security Capacity Centre. 2022. “CMM Reviews around the World”. Last updated March 2022. https://gcscc. ox.ac.uk/cmm-reviews Google and IFC, a member of the World Bank Group. Undated “e-Conomy Africa 2020”. Accessed 23.04.2022. https://www. ifc.org/wps/wcm/connect/publications_ext_content/ifc_external_publication_site/publicati ons_listing_page/google-e- conomy GovChat. Undated .Accessed 24.04.2022. https://www.govchat.org/ GPFI and G20 Argentina. 2018. “G20 Policy Guide Digitisation and informality: Harnessing digital financial inclusion for individuals and MSMEs in the informal economy”. 2018. https://www.oecd.org/g20/G20- Policy-Guide-Digitisation-and- Informality.pdf 128 >>> REFERENCES Graves, Laura. 2016. “Cancellation of MENA’s first major tower transaction”. TowerXchange. August 12, 2016. https://www. towerxchange.com/data/news/documents/b1qxwqp9dgwq9q/cancellation-of-mena- s-first-major-tower-transaction GSMA. 2021. “Tanzania: Driving social and economic value through mobile-sector tax reform”. GSMA. March, 2021. https:// www.gsma.com/publicpolicy/wp-content/uploads/2021/04/GSMA_Mobile_taxation_in_Tanzania_2021.pdf GSMA. 2021. “Roadmaps for 5G Spectrum: Sub-Saharan Africa”. August, 2021. https://www.gsma.com/spectrum/wp-content/ uploads/2021/09/spec_ssa_5g_iot_report_09_21.pdf GSMA. 2021. „Mobile Internet Connectivity 2021: Sub-Saharan Africa Key Trends“. https://www.gsma.com/r/wp-content/ uploads/2021/09/The-State-of-Mobile-Internet-Connectivity- 2021-Sub-Saharan-Africa.pdf. GSMA. 2021. “Spectrum Pricing and Licensing in Africa – Driving Mobile Broadband”. July 21, 2021. https://www.gsma.com/ spectrum/resources/effective-spectrum-pricing-africa/ GSMA. 2019. “Mobile Policy Handbook: An insider’s guide to the issues”. https://www.gsma.com/publicpolicy/ mobilepolicyhandbook/wp- content/uploads/2019/01/MPH7_ENG_web_spreads.pdf GSMA. 2019. “Rethinking mobile taxation to improve connectivity”. Last updated February 19, 2019. https://www.gsma.com/ publicpolicy/resources/rethinking-mobile-taxation-to-improve-connectivity GSMA. 2016. “Benefits of Gateway Liberalisation- Stimulating Economic Growth”. 2016. https://www.gsma.com/ mobilefordevelopment/wp-content/uploads/2016/09/GSMA-Gateway- Liberalisation.pdf GSMA. 2014. “Surtaxes on International Incoming Traffic in Africa”. September, 2014. https://www.gsma.com/publicpolicy/wp- content/uploads/2012/03/Surtaxes_on_International_Incoming_Traffic_in_Africa_FULL- REPORT_WEB.pdf GSMA and EY. 2020. “Mobile taxation in Kenya: Accelerating digital development.” March, 2020. https://www.gsma.com/ publicpolicy/wp-content/uploads/2020/03/GSMA_Mobile_taxation_in_Kenya_2020.pdf GSMA and EY. 2018. “Reforming mobile sector taxation in Guinea: Unlocking socio-economic gains from mobile connectivity”. GSMA. March, 2018. https://www.gsma.com/publicpolicy/wp-content/uploads/2018/05/GSMA-Guinea-taxation-report- EN.pdf GSMA and IFC. 2014. Tower Power Africa: Energy Challenges and Opportunities for the Mobile Industry in Africa. September, 2014. https://www.gsma.com/mobilefordevelopment/wp-content/uploads/2014/11/Africa-Market-Report- GPM-final.pdf Hanouch, Michel and Chen, Gregory. 2015. “Promoting Competition in Mobile”. February 2015 https://www.cgap.org/research/ publication/promoting-competition-mobile-payments-role-ussd Haranas, Mark. 2022. “Digital Realty, Equinix Fight For Data Center Supremacy In Africa”. CRN. Last updated January 04, 2022. https://www.crn.com/news/data-center/digital-realty-and-equinix-fight-for-data- center-supremacy-in-africa/ Huseby Jim, Stewart John. 2021. “Digital Realty Expands Coverage and Capacity of PlatformDigital® Across Africa”. Interxion. October 28, 2021. https://www.interxion.com/news/2021/10/digital-realty-expands- coverage-and-capacity-of- platformdigital-across-africa/ IBFD. 2021. “IBFD Africa Tax Symposium Hosts a Webinar Edition”. April 22, 2021. https://www.ibfd.org/news/ibfd-africa-tax- symposium-hosts-webinar-edition-0 <<< 129 IBFD. Undated. “IBFD - Tax Research Platform”. Accessed 25.04.2022. https://research.ibfd.org/#/ IFO institute. 2018. „Global Cybersecurity Index (GCI)“. Last updated 2018. https:// dice.ifo.de/en/node/317818#:~:text=Thepercent20Globalpercent20Cybersecurity percent20Indexpercent20(GCI,orderpercent20topercent20raise percent20cybersecurity percent20awareness. IHS Towers. 2021. “IHS Towers at a glance”. Last updated December 31, 2021. https://www.ihstowers.com/about-ihs/ihs- towers-at-a-glance IMF, OECD, UN and World Bank. 2015. „Options for Low Income Countries’ Effective and Efficient Use of Tax Incentives for Investment ”. October 15, 2015. https://www.oecd.org/tax/options-for-low-income-countries-effective-and-efficient-use-of- tax- incentives-for-investment.pdf International Monetary Fund (IMF). “Somalia 2016 Article IV Consultation—Press Release; Staff Report; and Statement by The Executive Director for Somalia”. February, 2017. https://centralbank.gov.so/wp- content/uploads/2020/08/Article-IV- Feb-2017.pdf International Monetary Fund (IMF). 2020. “Fiscal Monitor”. April 2020. https://www.imf.org/en/Publications/FM/ Issues/2020/04/06/fiscal-monitor-april-2020 International Telecommunication Union (ITU), Telecommunication Development Bureau (BDT). 2016.”Maximising availability of international connectivity in developing countries: Strategies to ensure global digital inclusion”. Geneva. https://www.itu.int/ dms_pub/itu-d/opb/pref/D-PREF-BB.GDI_01-2017-PDF- E.pdf International Trade Center. 2019. “Joining forces for e-commerce: How small African firms succeed with collaborative business models”. April 01, 2019. https://intracen.org/resources/publications/joining-forces- for-e-commerce-how-small-african- firms-succeed-with International Trade Centre (ITC). 2022. “Customs Tariffs For products: 8517110000 – Line telephone sets with cordless handsets, 8517130000 – Smartphones for wireless networks, 8523210000 – Cards incorporating a magnetic stripe for the recording of sound or of other phenomena”. https://www.macmap.org/en/query/customs- duties?reporter=276&year=2022&partner=392&product=852321&level=6 International Trade Centre, 2019. “Regiones y zonas de disponibilidad”. AWS Amazon. 2019. https://aws.amazon.com/es/ about-aws/global-infrastructure/regions_az/ International Trade Center (ITC). Undated “Market Access Map”. Accessed 25.04.2022. https://m.macmap.org/ ITU News. 2019. “How mobile virtual network operators (MVNOs) drive value and improve lives”. May 22, 2019. https://news. itu.int/mvnos-telecom-world-value/ ITU Publications. 2021. “Digital trends in Africa 2021. Information and communication technology trends and developments in the Africa region 2017-2020”. ITU, 2021. https://www.itu.int/dms_pub/itu- d/opb/ind/D-IND-DIG_TRENDS_AFR.01-2021- PDF-E.pdf ITU, Telecommunication Development Bureau (BDT). 2013. “Study on international Internet connectivity in sub-Saharan Africa”. March, 2013. https://www.itu.int/en/ITU-D/Regulatory-Market/Documents/IIC_Africa_Final-en.pdf 130>>> REFERENCES ITU. 2013. “Introduction to The Internet”. IPv6 Transition Strategies and Technologies Workshop, 5th – 6th August 2014. Vientiane. https://www.itu.int/en/ITU-D/Regional-Presence/AsiaPacific/Documents/S04- Internet-introduction.pdf ITU. 2016. “Maximising availability of international connectivity in developing countries: Strategies to ensure global digital inclusion”. Geneva. https://www.itu.int/dms_pub/itu-d/opb/pref/D-PREF-BB.GDI_01- 2017-PDF-E.pdf ITU. 2020. “ICT Regulatory Tracker”. Last updated 2020. https://app.gen5.digital/tracker/metrics?_ga=2.66812398.191 7600628.1650804633- 2062661382.1650804633&_gl=1*16avhm8*_ga*MjA2MjY2MTM4Mi4xNjUwODA0NjMz*_ ga_27GW57N RWK*MTY1MDgwNDYzMi4xLjEuMTY1MDgwNDY0Ny4w ITU. 2020. “Digital Regulation Handbook”. http://handle.itu.int/11.1002/pub/81630556-en ITU. 2020. Regional and global key ICT aggregates. https://www.itu.int/en/ITU-D/Statistics/Documents/facts/ITU_regional_ global_Key_ICT_indicator_aggregates_Nov_2020.xlsx. ITU. 2021. “Cybersecurity initiatives”. Last updated April 2021. https://www.itu.int/ en/action/cybersecurity/Pages/default.aspx#:~:text=percentE2percent80 percent8BApercent20fundamentalpercent20rolepercent20of,andpercent20Communication percent20Technologies percent20(ICTs). ITU. 2021. “The Benchmark of Fifth Generation Collaborative Regulation”. June 21, 2021. https://digitalregulation.org/wp- content/uploads/G5Benchmark_ReviewBoardReport_21062021.pdf Jaupart, Pascal and Begazo, Tania. 2023. Corporate Governance and Transparency of State-Owned and State-Linked Digital Enterprises in Africa Jarju, Momodou. 2020. “Gambia: D173. 4 Million NBN Data Center Investment Yet to Accrue Income” https://allafrica.com/ stories/202007130604.html. Jumia.com. n. d. Accessed 24.04.2022. https://group.jumia.com/ Kafeero, Stephen. 2021. “To control speech, Uganda is taxing internet usage by 30 percent”. Quartz Africa. July 3, 2021. https://qz.com/africa/2028653/uganda-replaces-ott-social-media-tax-with-tax-on-internet-bundles/ Kavuma, Susan Namirembe et al. 2020. “An analysis of the distributional impact of excise duty in Uganda using a tax- benefit microsimulation model”. WIDER Working Paper, No. 2020/70, ISBN 978-92-9256-827- 6, The United Nations University World Institute for Development Economics Research (UNU-WIDER), Helsinki. https://doi.org/10.35188/UNU- WIDER/2020/827-6 Kavuma, Susan. 2020. “An analysis of the distributional impact of excise duty in Uganda using a tax-benefit microsimulation model”. Last access on March 23, 2021. https://www.econstor.eu/bitstream/10419/229294/1/wp2020-070.pdf Kene-Okafor, Tage. 2021. “Sequoia Heritage, Stripe and others invest $200M in African fintech Wave at $1.7B valuation”. TechCrunch. September 2, 2021. https://techcrunch.com/2021/09/06/sequoia-heritage- stripe-and-others-invest-200m- in-african-fintech-wave-at-1-7b-valuation/?guccounter=1. Kene-Okafor, Tage. 2021. “Facebook-backed 2Africa set to be the longest subsea cable upon completion”. TechCrunch. September 30, 2021. https://techcrunch.com/2021/09/29/facebook-backed-2africa-set-to- be-the-longest-subsea- cable-upon- completion/?guccounter=1&guce_referrer=aHR0cHM6Ly93d3cuZ29vZ2xlLmNvbS8&guce_referrer_ <<< 131 sig=A QAAAMPPSx1BshyFbdmNUjkSFa0edqmR79zt3J2cIt_wtSR2ncdyUoUt1LMtG9I41GiLo2m5wRQBDn3aPzjH mYk3tbYLSJ7Epmz4ymjSkf3H-1UWTD-OIWLNCABcugaOHVXPXK5D9QxCDN-0Wcu9- 2zqZQ6JNmGX076BL18BzZ6oVJYp Kenya Gazette Supplement. 2018. “Acts 2018”. Kenya Gazette Supplement No. 121 (Acts No. 10). September 21st, 2018, Nairobi. https://www.kra.go.ke/images/publications/Finance-Act-2018.pdf Kenyan National Bureau of Statistics. 2019. “Statistical Abstract 2019”. KNBS. August 28, 2020. https://www.knbs. or.ke/?wpdmpro=statistical-abstract-2019 Kieron Osmotherly. 2019. “The new shape of the global telecom tower industry”. TowerXChange. December 09, 2019. https:// www.towerxchange.com/data/news/documents/b1qxws204w3sh1/the-new- shape-of-the-global-telecom-tower-industry Killian, Clifford. “The causes and consequences of mobile money taxation: An examination of mobile money transaction taxes in Sub-Saharan Africa”. GSMA. June, 2020. https://www.gsma.com/mobilefordevelopment/wp-content/uploads/2020/06/ GSMA_The-causes-and- consequences-of-mobile-money-taxation.pdf Killian, Clifford. 2020. “The causes and consequences of mobile money taxation: An examination of mobile money transaction taxes in Sub-Saharan Africa”. GSMA. June, 2020. https://www.gsma.com/mobilefordevelopment/wp-content/ uploads/2020/06/GSMA_The-causes-and- consequences-of-mobile-money-taxation.pdf Kovacic William E, Marshall Robert C, Marx Leslie M and White Halbert L. 2011. “Plus Factors and Agreement in Antitrust Law”. 2011. https://repository.law.umich.edu/cgi/viewcontent.cgi?article=1118&context=mlr Lange, Peter. 2015. “Telecommunications Contribution to Public Finance in Somalia”. World Bank Memorandum. Laws of Kenya. 2013. Act No. 35 of 2013. Act Title: VALUE ADDED TAX. http://www.kenyalaw.org:8181/exist/kenyalex/ actview.xql?actid=No. percent2035 percent20of percent202013 Lesher Molly, Pawelec Hanna and Desai Arpitha. 2022. “Disentangling untruths online: Creators, spreaders and how to stop them”. Going Digital Toolkit Note, No. 23. https://goingdigital.oecd.org/data/notes/No23_ToolkitNote_UntruthsOnline.pdf. Makoni, Munyaradzi. 2020. “Cyberattack surge highlights Africa security risk. Scidev”. October 10, 2020. https://www.scidev. net/sub-saharan-africa/news/cyberattack-surge-highlights-africa-security-risk/ Malawi News Agency. 2016. “Govt launches virtual landing point”. May 18, 2016. https://www.mwnation.com/govt-launches- virtual-landing-point/ Mansour, Mario and Schneider, Jean-Luc. 2019. “How to Design Tax Policy in Fragile States”. September, 2019. https://www. oecd.org/g20/G20-Policy-Guide-Digitisation-and-Informality.pdf Matheson, Thornton and Petit Patrick. 2021. “Taxing telecommunications in developing countries”. International Tax and Public Finance, Springer; International Institute of Public Finance, vol. 28(1), pages 248-280, February. https://ideas.repec.org/a/ kap/itaxpf/v28y2021i1d10.1007_s10797-020-09621-6.html Mathur, Priyamvada & Rubio, Jordan. 2022. “Mapping Africa’s growing venture capital landscape”. Pitchbook. Last updated April 22, 2022. https://pitchbook.com/?utm_source=google&utm_medium=cpc&utm_ campaign=Brand-Other- EU&adgroup=Brand-Other- General&utm_term=pitchbook&device=c&utm_ content=&_bk=pitchbook&_bt=474498043830&_bm=p&_bn=g&_bg=112740451580&kwdaud=kwd- 132 >>> REFERENCES 299395290400&sfid=VsVUWlvk- dc_pcrid_474498043830_pkw_pitchbook_pmt_p_slid productid pgrid_112740451580_ptaid_kwd-299395290400&gclid=Cj0KCQjw6pOTBhCTARIsAHF23fIWMa6Avfua4cPzjAbdchh6fzeklH_ vsGH8cXL8bcBx 7 ZXLn8SYsaAlkKEALw_wcB Mauritius. MRA. 2021. “Value Added Tax Act (Consolidated Version up to Finance Act 2021)”. https://www.mra.mu/download/ VATAct.pdf McCarthy, Niall. 2020. “The Countries Shutting Down The Internet The Most.” Statista.com. January 22. Accessed August 5, 2021. https://www.statista.com/chart/15250/the-number-of-internet-shutdowns-by- country/. McKay, Claudia and Michel Hanouch (2021). “MTN Mobile Money Opened APIs – Was It Worth It?”. October 28, 2021. CGAP. https://www.cgap.org/blog/mtn-mobile-money-opened-apis-was-it-worth-it. Meta. 2022. “Meta Earnings Presentation Q4 2021”. Last updated: February 2, 2022. https://investor.fb.com/investor-events/ event-details/2022/Meta-Q4-2021-Earnings/default.aspx Miller, Mark. 2021. “Public finance and development: top things to read on fiscal responses to Covid-19”. ODI. https://odi.org/en/ insights/public-finance-and-development-top-things-to-read-on-fiscal-responses- to-covid-19/ Ministere Du Budget et Du Portefeuille de L´etat . 2022. “Code Général des Impôts 2022”. Direction Generale des Impots. http:// www.dgi.cgici.com/indexs.htm Ministry of Finance and Economic Development. 2021. “Finance Act 2021”. http://www.zimtreasury.gov.zw/index. php?option=com_phocadownload&view=category&id=14:finance- act&Itemid=790 Ministry of Finance. .2018. “Income Tax Act”, Sec. 36G; Statutory Instrument 205 of 2018. https://www.veritaszim.net/ sites/veritas_d/files/SI percent202018-205 percent20- percent20Amendments percent20to percent20Finance percent20Act percent20 percent26 percent20Income percent20Tax percent20Act.pdf Mistry, Jamshed. 2005. “A conceptual framework for the role of government in bridging the digital divide”. Journal of Global Information Technology Management 8(3), 28-47. July 2005 DOI: 10.1080/1097198X.2005.10856401 MTN. 2020. Annual reports. https://www.mtn.com/annual-reports/ MTN. 2021. “Audited summary group financial statements for the year ended 31 December 2020”. https://www.mtn-investor. com/reporting/annuals-2020/index.html National Cybersecurity Policy of Nigeria. 2014. Last updated December 2014. https://www.cert.gov.ng/ngcert/resources/ NATIONAL_CYBESECURITY_STRATEGY.pdf National Telecommunication Regulatory Authority of Egypt. 2013. “Egypt Telecommunication Regulation Law”. February 2003. https://www.tra.gov.eg/wp-content/uploads/2020/11/Law-No-10-of-2003.pdf Ndung’u, Njuguna. 2019. “Taxing Mobile Phone Transactions in Africa: Lessons from Kenya”. Brookings Institution, Africa Growth Initiative Policy Brief. August 5, 2019 https://www.brookings.edu/research/taxing-mobile-phone-transactions-in-africa- lessons-from-kenya/ NetBlocks.2022. NetBlocks Cost of Shutdown Tool (COST). https://netblocks.org/cost/ <<< 133 Niesten, Hannelore and Begazo, Tania. 2023 „World Bank, Digital Infrastructure Taxes and Parafiscal Fees in Africa“ Njeru, Rose and Syekei, John. 2021. “Kenya: Overwie of Data Infrastructure in East Africa”. April 06, 2021. https://www. bowmanslaw.com/insights/technology-media-and-telecommunications/data-infrastructure- in-east-africa/ Nyawira, Susan. 2020. “Competition Authority of Kenya seeks to regulate vast digital lending sector”. The Star. February 24, 2020. https://www.the-star.co.ke/business/kenya/2020-02-24-competition-authority-of-kenya-seeks-to- regulate-vast- digital-lending-sector/ Nyman, Sara and Barajas, Rodrigo. 2021. “Antitrust and Digital Platforms: An analysis of global patterns and approaches by competition authorities”. EFI Insight-Trade, Investment and Competitiveness. https://documents1.worldbank.org/curated/ en/893381632736476155/pdf/Antitrust-and-Digital- Platforms-An-Analysis-of-Global-Patterns-and-Approaches-by- Competition-Authorities.pdf OECD. 2016. “The role of market studies as a tool to promote competition”. December 01, 2016. https://www.oecd.org/daf/ competition/the-role-of-market-studies-as-a-tool-to-promote- competition.htm OECD. 2017. “Algorithms and Collusion: Competition Policy in the Digital Age.” http://www.oecd.org/daf/competition/ Algorithms-and-colllusion-competition-policy-in-the-digital- age.pdf OECD. 2017. “International VAT/GST Guidelines, OECD Publishing”. OECD. April 12, 2017. https://www.oecd.org/ctp/ international-vat-gst-guidelines-9789264271401-en.htm OECD. 2018. “Indicators of Product Market Regulation”. https://www.oecd.org/economy/reform/indicators-of-product-market- regulation/ OECD. 2019. “The Role of Digital Platforms in the Collection of VAT/GST on Online Sales”. June 20, 2019. https://www.oecd. org/tax/the-role-of-digital-platforms-in-the-collection-of-vat-gst-on-online-sales- e0e2dd2d-en.htm OECD. 2020.“Transparency and Disclosure Practices of State-Owned Enterprises and their Owners”. http://www.oecd.org/ corporate/transparency-disclosure-practices-soes.pdf OECD. 2021. “International collaboration to end tax avoidance”. Last updated November, 2021. https://www.oecd.org/tax/beps/ OECD. 2021. “Revenue Statistics”. 2021. https://www.oecd.org/tax/tax-policy/oecd-classification-taxes- interpretative-guide.pdf OECD. 2021. “Addressing the taxchallenges arising from the digitalisation of the economy”. OECD/G20 Base Erosion and Profit Shifting Project. July 2021. https://www.oecd.org/tax/beps/brochure-addressing-the- tax-challenges-arising-from-the- digitalisation-of-the-economy-july-2021.pdf OECD. 2021. “Statement on a Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy”. OECD/G20 Base Erosion and Profit Shifting Project. October 8, 2021. https://www.oecd.org/tax/beps/statement-on-a-two- pillar-solution-to-address-the-tax-challenges- arising-from-the-digitalisation-of-the-economy-october-2021.pdf OECD. 2022.“OECD Tax Database“. https://www.oecd.org/ctp/tax-policy/tax-database/ OECD. 2017. “Green growth indicators framework”. Accessed April 22, 2022. https://www.oecd.org/greengrowth/green- growth-indicators/ 134 >>> REFERENCES OGEL. 2012. “Presidential Legislative Decree 3-12 on incentives for Angolan petroleum companies 2012”. March 16, 2012. https://www.ogel.org/legal-and-regulatory-detail.asp?key=24407 OOKLA. 2020. “Speedtest Global Index”. December, 2020. https://www.ookla.com/articles/icymi-ookla- data-and-analysis- from-december-2020 Oseni, Michael. 2014. “Multiple Taxation as a Bane of Business Development in Nigeria”. March, 2014. https://pdfs. semanticscholar.org/3e0f/6cf138393b766fc6c5d7c24dcd3a58c493a3.pdf Parekh, Nidhi and Hare, Aimee. 2020. “The rise of mobile money in sub-Saharan Africa: Has this digital technology lived up to its promises?”. October 22, 2020. https://www.povertyactionlab.org/blog/10-22- 20/rise-mobile-money-sub-saharan-africa- has-digital-technology-lived-its-promises Partech. 2021. “2021 Africa Tech Venture Capital”. https://partechpartners.com/2021-africa-tech-venture- capital-report/ Partech. 2021. “2021 Partech Africa Report”. Feb. 3, 2022. https://partechpartners.com/news/2021- partech-africa-report/ Payments: The Role of USSD”. CGAP. https://documents1.worldbank.org/curated/ en/586781468127790413/pdf/952010BRI0Box30ole0of0USS D0Feb02015.pdf Peace Cable, n.d., PEACE Cable System. Accessed 23.04.2022. http://www.peacecable.net/#about Pedros, Xavier, Sivakumaran, Mayuran. 2019. “Rethinking mobile taxation to improve connectivity”. GSMA. 2019. https://odi. org/en/publications/how-tax-officials-in-lower-income-countries-can-respond-to-the- coronavirus-pandemic/ Pollicy. 2019. “Offline and Out of Pocket: The Impact of the Social Media Tax in Uganda on Access, Usage, Income and Productivity”. Genderit.org. February 13, 2019. https://www.genderit.org/resources/offline- and-out-pocket-impact-social- media-tax-uganda Pop Georgiana, Coelho Gonçalo and Connon, Davida. 2023. “State-Owned Enterprises in Digital infrastructure and downstream digital markets in Africa”. Pop, Georgiana and Coelho Gonçalo. 2023. “Competition Policy in Digital Markets in Africa” Pop, Georgiana and Coelho, Gonçalo. 2021.“Competition Across the Radio Spectrum in West Africa: Regional and National Aspects”, Competition Policy International. January 6, 2021. https://www.competitionpolicyinternational.com/ competition-across-the-radio-spectrum-in-west-africa- regional-and-national-aspects/ Président de la République Du Senegal. 2008. “LOI n° 2008-46 du 3 septembre 2008 instituant une redevance sur l’accès ou l’utilisation du réseau des télécommunications publiques (RUTEL)”. Journal Officiel. J.O. N° 6441 du Samedi 6 DECEMBRE 2008. http://www.jo.gouv.sn/spip.php?article7199 Président de la République Du Senegal. 2010. “Loi n° 2010-14 du 23 juin 2010 portant loi de finances rectificative pour l’année 2010”. Journal Officiel. http://www.jo.gouv.sn/spip.php?page=recherche&recherche=Law+no.+2010-14 PWC. 2022. „Worldwide Tax Summaries Online. Last updated 07 April 07, 2022. https://taxsummaries.pwc.com/ Qiu, Winston. 2020. “A Study on Submarine Cables Crossing Egypt and their Costs“. Submarine Cable Networks. 22 April 2020. https://www.submarinenetworks.com/en/services/research/submarine-cables- crossing-egypt-and-cost <<< 135 Rédaction. 2020. „RDC : les appareils mobiles désormais taxés de 1 à 7 dollars USD“. Financial Afrik. April 28, 2020. https:// www.financialafrik.com/2020/04/28/rdc-les-appareils-mobiles-desormais-taxes-de-1-a- 7-dollars-usd/#:~:text=C’est percent20le percent20Journal percent20officiel,la percent20Poste percent20et percent20des percent20T percentC3 percentA9l percentC3 percentA9communications Regulatory Watch Initiative (RWI) phase 2. March 2021. „Thorough legal, regulatory and competitive analyses of issues related to licensing, OTTs, international gateways, spectrum management and regulatory governance“. World Bank, Washington, DC. https://documents1.worldbank.org/curated/en/692551602223646754/text/Main-Report.txt Republique Du Burundi. 2018/2019. „LOI N°1/ 013 du 30 juin 2018 portant fixation du budget general de la Republique de Burundi Pour L´Exercice 2018/2019“. Cabinet Du President. June 30, 2018. https://www.cabri- sbo.org/uploads/bia/ Burundi_2019_Approval_External_BudgetProposal_MinFin_COMESAEACECCAS_Fre nch.PDF Reuters Staff. 2020. „Amazon launches data centre operations in South Africa“. Johannesburg. April 22, 2020. https://www. reuters.com/article/us-safrica-amazon-com-cloud-idUSKCN2241X3 Roland Berger and International Finance Corporation. 2021. Unpublished. Rota-Graziosi, Grégoire and Sawadogo, Fayçal. 2003. “The tax burden on mobile network operators in Africa”. Fondation pour les études et recherches sur le développement international (FERDI). https://ferdi.fr/dl/df-FR6QBHUxA1kbft3GHpctTupj/ ferdi-p278-the-tax-burden-on-mobile-network- operators-in-africa.pdf Runde, Daniel. F. and Ramanujam, Sundar. R. 2021. “Digital Governance. It Is Time for the United States to Lead Again”. In: Center for Strategic & International Studies, CSIS Briefs. Washington, D.C. https://csis- website-prod.s3.amazonaws.com/ s3fs- public/publication/210802_Runde_U.S._Digital_Governance.pdf?0HOwy9OU_m5ypT.HwAiNthoEBjXzYV MA Rusell Jon, 2017. “Amazon completes its acquistion of Midle Eastern e-commerce firm Souq”. Techcrunch. Last updated July 03, 2017. https://techcrunch.com/2017/07/03/amazon-souq-com-completed/ Rwanda Utilities Regulatory Authority. n. d. Accessed 24.04.2022. https://rura.rw/index.php?id=23 Rwanda Revenue Authority. 2019. “Law No 025/2019 of 13/09/2019 establishing The Excise Duty”. Official Gazette n° 37 bis of 23/09/2019. https://www.rra.gov.rw/fileadmin/user_upload/new_excise_duty_law official_gazette 1.pdf Safaricom. 2021. “Safaricom unveils baze digital content platform”. May 28, 2021. https://www.safaricom.co.ke/about/media- center/publications/press-releases/release/1052 Safaricom. 2021. “Spectrum Pricing and Licensing in Africa – Driving Mobile Broadband”. GSMA. Jul 21, 2021. https://www. gsma.com/spectrum/resources/effective-spectrum-pricing-africa/ Safaricom undated “DigiFarm”. Accessed 24.04.222. https://www.safaricom.co.ke/faqs/faq/810 Schatan, Roberto. 2012. “Tax-Minimizing Strategies and the Arm’s-Length Principle”. Tax notes. https://www.taxnotes. com/tax-notes-international/transfer-pricing-and-apportionment-issues/tax- minimizing-strategies-and-arms-length- principle/2012/01/09/s9m1 Schumann, Robert and Kende, Michael. 2013. “Lifting barriers to internet development in Africa: suggestions for improving connectivity”. Analysys Mason, Internet Society. May 2013. https://www.intgovforum.org/multilingual/system/files/ filedepot/17/Record percent2024a.pdf 136 >>> REFERENCES Secretariat Democratic Republic oft he Congo. Trade Policy Review. WT TPR/S/339 https://docsonline.wto.org/dol2fe/Pages/FE_Search/FE_S_S009- DP.aspx?language=E&CatalogueIdList=228700,106572,100383&CurrentCatalogueIdIndex=0&FullTextHas h=&HasEnglishRecord=True&HasFrenchRecord=True&HasSpanishRecord=True Secretariat General Du Gouvernement. 2018.„Decree No. 2018-341 of July 25, 2018. https://sgg.gouv.bj/doc/decret-2018-341/ Seyama, Bra Willy. 2020. “Africa’s top 10 countries on Facebook”. eNitiate. March 13, 2020. https://enitiate.solutions/africas- top-10-countries-contribute-a-combined-76-to-the-continents- facebook-active-users/ Seuyong, Feraud, Silwal, Ani, Rudra, Begazo, Tania, Newhouse, David and N’Ghauran, Alain. 2023, “The Size and Distribution of Digital Connectivity Gaps in Sub-Saharan Africa.” Equitable Growth, Finance & Institutions Insight. Situ, Klétus. 2021. “Faure Gnassingbé inaugurates Lomé Data Centre, Togo’s first carrier hotel”. Togo First. Last updated June 07, 2021. https://www.togofirst.com/en/itc/0706-7979-faure-gnassingbe-inaugurates- lome-data-centre-togo-s-first- carrier-hotel/ Solomon Gabriel and Strusani, Davide.2011. “Mobile telephony and taxation in Kenya”. GMSA and Doloitte LLP. 2011. https:// www.gsma.com/publicpolicy/wp- content/uploads/2012/03/mobiletelephoneandtaxationinkenya.pdf Sousa, Joao and Heinrich, Diego. n. d. “Infrastructure regulation: overview and impact on TowerCos”. Delta Partners. Last accessed April 23. 2022. https://deltapartnersgroup.com/2019/04/24/infrastructure-regulation-overview-and-impact-on- towercos/ Special Gazette of the United Republic of Tanzania. 2021. “Special Bill Supplement”. Excise Duty (Amendment) (No. 2) Bill, 2018. http://www.parliament.go.tz/polis/uploads/bills/1623832428-13.06.2021 percent20THE percent20FINANCE percent20BILL percent202021.pdf Srinivasan, Sharada and Comini, Niccolo and Minges, Michael. 2021. “The Importance of National Data Infrastructure for Low and Middle-Income Countries”. March 2, 2021. Available at SSRN: https://ssrn.com/abstract=3898094. Statista Research Department, 2021. “Number of online shoppers in Africa from 2017 to 2025”. Statista. September 23, 2021. https://www.statista.com/statistics/1190579/number-of-online-shoppers-in-africa/ Statista. 2022.“The rise of e-commerce across Africa“. Accessed 24.04.2022. https://www.statista.com/study/105869/the-rise- of-e-commerce-across-africa/ Statistics Times. 2022.. “List of continents by population 2021”. Accessed April 22, 2022. https://statisticstimes.com/ demographics/continents-by-population.php Steel Iain. 2020. “How tax officials in lower-income countries can respond to the coronavirus pandemic”. Odi. April 03, 2020. https://odi.org/en/publications/how-tax-officials-in-lower-income-countries-can- respond-to-the-coronavirus-pandemic/ Strusani Davide. 2014. “Surtaxes on International Incoming Traffic in Africa”. GSMA. September, 2014. https://www.gsma.com/ publicpolicy/wp- content/uploads/2012/03/Surtaxes_on_International_Incoming_Traffic_in_Africa_FULL- REPORT_ WEB.pdf Strusani, Davide. 2015. “Digital inclusion and mobile sector taxation in Tanzania”. February, 2015. https://www.gsma.com/ mobilefordevelopment/wp-content/uploads/2015/01/Digital-inclusion-mobile- sector-tax-Tanzania.pdf <<< 137 Submarine Cable Networks. n. d. „Africa-1“. Accessed 23.04.2022. https://www.submarinenetworks.com/en/systems/asia- europe-africa/africa-1 Sudan Taxation Chamber. Laws and Regulations. https://www.tax.gov.sd/?page_id=1225 Sutherland, Ewan. 2014. “Undersea cables and landing stations around Africa: Policy and regulatory issues”. 25th European Regional Conference of the International Telecommunications Society (ITS): Disruptive Innovation in the ICT Industries: Challenges for European Policy and Business. Brussels. June 22nd-25th, 2014. http://hdl.handle.net/10419/101381 Swinhoe, Dan. 2021. “Togo opens first carrier-neutral colocation facility in Lomé”. dana Center Dynamics (DCD). June 08, 2021. https://www.datacenterdynamics.com/en/news/togo-opens-first-carrier-neutral- colocation-facility-in-lom percentC3 percentA9/ Taty, Dominique; Gnamien, Jean Claude. 2022. “Ivory Coast: Corpor–te - Other Taxes”. PWC Worldwide Tax Summaries. Last updated March 21, 2022. https://taxsummaries.pwc.com/ivory-coast/corporate/other-taxes Tax Administration Diagnostic Assessment Tool (TADAT). Undated. Accessed 25. 04.2022. https://www.oecd.org/tax/beps/ Taxania Revenue Authority. 2017. Tax laws. https://www.tra.go.tz/index.php/laws TCHEMEDIE Jephté. 2020. “RD Congo : Un décret introduit une taxe de 1 à 7 dollars sur les téléphones mobiles”. Digital Business Africa. Last updated May 1, 2020. https://www.digitalbusiness.africa/rd-congo- un-decret-introduit-une-taxe-de- 1-a-7-dollars-sur-les-telephones-mobiles/ Telecom Malagasy. 2020. “Telecoms excise duty restored to 10 percent in Madagascar’s 2020 finance law”. Comms Update. March 9, 2020. https://www.commsupdate.com/articles/2020/03/09/telecoms-excise-duty-restored-to-10- in- madagascars-2020-finance-law/#:~:text=The percent20excise percent20duty percent20levied percent20on,burden percent20faced percent20by percent20the percent20sector TeleGeography. 2020. The State of the Network. 2020 Edition. https://www2.telegeography.com/hubfs/assets/Ebooks/state- of-the-network-2020.pdf TeleGeography 2022 “Submarine Cable Map”. Last updated April 12, 2022. https://www.submarinecablemap.com/submarine- cable/africa-1 TeleGeography 2022. “Global Internet Geography Database”. Accessed April 22, 2022. https://statisticstimes.com/demographics Teleography‘s Global Comms. 2022. March 12, 2022 https://www.telegeography.com/products/globalcomms/ The Republic of Uganda. 2014. Excise Duty Act 2014. https://www.parliament.go.ug/cmis/views/918a2bcd- 824f-40b3- b3ab-8132fef02e77 percent253B1.0#:~:text=2 percent20 percent3A-,THE percent20REPUBLIC percent20OF percent20UGANDA percent20THE percent20EXCISE percent20DUTY percent20ACT percent2C percent202014.,to percent20provide percent20for percent20related percent20matters. Uganda Communications Commission, 2021. “Market Performance Report 2Q21”. April - June, 2021. https://www.ucc.co.ug/ wp-content/uploads/2021/09/2Q21-MARKET-PERFOMANCE-REPORT- compressed.pdf UNCTAD. 2020. “Preparatory Report fort he Ex Post Review oft he Competition Policy oft he West African Economic and Monetary Union”. 2020. Geneva. https://unctad.org/system/files/official- document/ditcclp2020d2_en.pdf. 138 >>> REFERENCES United Nations - Department of Economic and Social Affairs Population Dynamics. 2018. “World Urbanization Prospects 2018”. Accessed April 22, 2022. https://population.un.org/wup/Download/ United Nations Conference On Trade and Development (UNCTAD). 2021. “Digital Economy Report 2021”. Geneva, 2021. https://unctad.org/system/files/official-document/der2021_en.pdf United States Department of Justice. 2010. “Horizontal Merger Guidelines”. Last updated August 19, 2010. https://www.justice. gov/atr/horizontal-merger-guidelines-08192010 Van Der Made, Jan. 2021. “Senegal to Move All Govt Data to Huawei-Run Data Center”. All africa. June 25, 2021. https:// allafrica.com/stories/202106260013.html/ Vodafone Group Plc. 2021. Annual Report 2021. https://investors.vodafone.com/sites/vodafone- ir/files/2021-05/vodafone- annual-report-2021.pdf West, M. Darrell. 2016. Internet shutdowns cost countries $2.4 billion last year. Governance Studies, Washington DC: Center for Technology Innovation at Brookings Center World Bank Group and OECD. 2018. „TC Data 360. Product Market Regulation, Index. https://tcdata360.worldbank.org/ indicators/gim.prod.mkt.reg?country=BRA&indicator=3300&viz=line_ch art&years=1998,2018#comparison-link World Bank Group (WBG). 2016. “Breaking Down Barriers: Unlocking Africa’s Potential through Vigorous Competition Policy”. July 27, 2016. Washington, DC. https://documents1.worldbank.org/curated/en/243171467232051787/pdf/106717- REVISED-PUBLIC- WBG-ACF-Report-Printers-Version-21092016.pdf World Bank Group. 2018. “Federal Republic 0f Somalia Systematic Country Diagnostic”. May 01, 2018. https://documents1. worldbank.org/curated/en/554051534791806400/pdf/SOMALIA-SCD-08152018.pdf World Bank Group. 2019. „GUINEA Opportunities for Enhanced Domestic Revenue Mobilization: Value-Added Tax and Excise Taxes“. https://documents1.worldbank.org/curated/pt/472021561614678154/pdf/Guinea-Opportunities-for- Enhanced- Domestic-Revenue-Mobilization-Value-Added-Tax-and-Excise-Taxes.pdf World Bank Group. 2019. “Integrated SOE Framework (iSOEF): Corporate Governance of State-owned Enterprises”. https://olc. worldbank.org/content/integrated-soe-framework-isoef-corporate-governance- of-state-owned-enterprises World Bank Group. 2020. “Digital solutions in a time of crisis”. July, 2020. https://documents1.worldbank.org/curated/ en/775621594292073824/pdf/Uganda-Economic-Update- Fifteenth-Edition.pdf World Bank. 2019. “Digital Business Indicators—Digital-Market-Regulations”. Last updated: June 25, 2019. https://www. worldbank.org/en/research/brief/digital-business-indicators-digital-market-regulations World Bank. 2019. “Guinea: Opportunities for Enhanced Domestic Revenue Mobilization: Value-Added Tax and Excise Taxes”. 2019. http://documents1.worldbank.org/curated/pt/472021561614678154/pdf/Guinea-Opportunities-for- Enhanced- Domestic-Revenue-Mobilization-Value-Added-Tax-and-Excise-Taxes.pdf World Bank. 2020. “Global Cyber Security Capacity Program Phase I and II: Strengthening national Cyber Security Environment of Selected Developing Countries”. June 1St. 2020. https://www.worldbank.org/en/news/feature/2020/06/01/kwpfgscp World Bank. 2020. Data Bank Worldwide Governance Indicators. Last updated: September 27, 2021. https://databank. worldbank.org/source/worldwide-governance-indicators#. <<< 139 World Bank. 2021. Digiwatch. https://dig.watch/newsletter World Bank. 2021. “Identification for Development (ID4D). ID4D Global Dataset 2018”. https://datacatalog.worldbank.org/ search/dataset/0040787 World Bank. 2021. “Worldwide Governance Indicators (WGI)”. http://info.worldbank.org/governance/wgi/ World Bank. 2021. World Development Report 2021: Data for better lives. https://www.worldbank.org/en/publication/wdr2021 World Bank. 2022. “COVID-19 Business Pulse Survey Dashboard”. Last updated: March 7, 2022 https://www.worldbank.org/ en/data/interactive/2021/01/19/covid-19-business-pulse-survey-dashboard World Bank. 2022. “Explore databases”. https://databank.worldbank.org/databases/infrastructure/page/1?qterm=Digital World Bank. 2022.“Innovations in Tax Compliance: Building Trust, Navigating Politics, and Tailoring Reform“. February , 2022. https://www.worldbank.org/en/events/2022/02/17/innovations-in-tax- compliance-building-trust-navigating-politics-and- tailoring-reform World Bank undated. “The World by Income and Region”. Accessed April 22, 2022. https://datatopics.worldbank.org/world- development-indicators/the-world-by-income-and-region.html World Bank undated. “World Development Indicators (WDI)”. https://datatopics.worldbank.org/world- development-indicators/ World Bank 2021. “Income level and regional classifications”. Accessed April 22, 2022. https://datahelpdesk.worldbank.org/ knowledgebase/articles/906519-world-bank-country-and-lending- groups World Bank undated. Digital Economy for Africa Initiative (DE4A). Last accessed April 23. 2022. https://www.worldbank.org/en/ programs/all-africa-digital-transformation World Trade Organization (WTO). 2019. “Tariff Line Duties“. https://www.wto.org/english/tratop_e/tariffs_e/tariff_data_e.htm Worldometers. 2022. Africa Population (LIVE). Last updated 24.04.2022. https://www.worldometers.info/world-population/ africa-population/ Xalam Analytics, and Africa Data Centres Association (ADCA). 2021. “Growing Africa’s Data Center Ecosystem: An Assessment of Utility Requirements. ADCA White Paper Series”. February 2, 2021. http://africadca.org/en/white-paper-what-utilities- are-required-for-africas-digital-infrastructure-needs Xalam Analytics. 2018. “The African Data Center Boom”. https://xalamanalytics.com/research/investor- reports/the-african- data-center-boom-2018/ Xalam Analytics. 2021. “The Rise of the African Cloud 2021“. https://xalamanalytics.com/product/african-cloud-market-2021/ Youtube, undated. “YouTube by the Numbers”, Official Blog. Accessed April 19, 2022. https://blog.youtube/press/. Yves-Alexandre de Montjoye and Schweitze, Heike r. 2019. “Competition policy for the digital era”. 2019. https://ec.europa.eu/ competition/publications/reports/kd0419345enn.pdf Zhu, Tingting Juni; Grinsted, Philip; Song, Hangyul; Velamuri, Malathi. 2022. A Spiky Digital Business Landscape: What Can Developing Countries Do?. World Bank, Washington, DC. https://doi.org/10.1596/39437 140 >>> ANNEX 1 REFERENCES >>> Annex 1: List of contributors for data collection on taxation and State-owned and linked enterprises >>> Digital infrastructure taxes and parafiscal fees ANGOLA Soliman, Hashish & Partners Emmanuel Gbahabo Rua Joaquim Kapango Frederic Soliman Templars MC Jurist Advogados Soliman, Hashish & Partners Igonikon Adekunle Saifallah Kadry Templars BENIN Soliman, Hashish & Partners Nicolin Assogba Kimathi Kuenyehia RWANDA Cabinet d’Avocats Olga Anasside & Kimathi and Partners Arthur Rugango Nicolin Assogba Kafui Quashigah Cedar Ark Law Délalie Hélène Paty Kounake Kimathi and Partners Consolate Ndagire Attorney at Law /Certified Compliance Cedar Ark Law Officer KENYA John Syekei SENEGAL BURKINA FASO Bowmans Abou Faye SARR Cheikh Fall Denis Magonga Scp D’avocats Geni Et Kebe Cabinet Maitre Cheikh Fall Bowmans Mamadou Moctar Faye Rose Njeru SOUTH AFRICA Cabinet Maitre Cheikh Fall Bowmans Livia Dyer Angela Mukora Bowmans CAMEROON Bowmans Jing & Partners Law Firm TANZANIA MADAGASCAR Wilbert Kapinga CONGO, DEM. REP. Hanna Keyserlingk Bowmans Tanzania Benoit Kadima HK Jurifisc Law Firm Aisha Ally Sinda John W FFooks & Co Benjamina Rasoanaivo Bowmans Tanzania HK Jurifisc Law Firm CÔTE D’IVOIRE Sariaka Randrianarisoa TOGO Khaled Abou El Houda HK Jurifisc Law Firm Koffi Sylvain Mensah Attoh Houda Law Firm Cabinet Attoh-Mensah Malick Lo MAURITIUS Houda Law Firm Devarshi Ramburn TUNISIA Assane Dieye Bowmans Omar Ferchiou Houda Law Firm Vartika Sahai Ferchiou & Associés Elodie Dagneaux Bowmans Sonia El Mabrouk Houda Law Firm Ferchiou & Associés Vanessa Bah Leroux MOROCCO Houda Law Firm Mehdi Benouna UGANDA Lilian Cadel Biassaly Ecovis Brian Kalule Houda Law Firm Bowmans NIGERIA Sophie Nyombi EGYPT Dipo Komolafe Bowmans Mohamed Hashish Templars <<< 141 >>> SOE and SLE in digital markets ANGOLA Bowmans (Mauritius) André Maurício Shianee Calcutteea MC Jurist Advogados Bowmans (Mauritius) Nuno de Miranda Catanas Anu Matikola MC Jurist Advogados Bowmans (Mauritius) Laura Maia Lucena Devarshi Ramburn MC Jurist Advogados Bowmans (Mauritius) EGYPT SIERRA LEONE Rana Abdelaty Henrietta Cole Soliman Hashish & Partners Law Firm Basma & Macaulay Mohamed Hashish Oluwagbemileke Joy Jegede Soliman Hashish & Partners Law Firm Basma & Macaulay Marwa AlSherif Soliman Hashish & Partners Law Firm SOUTH AFRICA Saifallah Kadry Livia Dyer Soliman Hashish & Partners Law Firm Bowman Gilfillan John Paul Ongeso ETHIOPIA Bowman Gilfillan Ermias Ayalew Aman Assefa & Associates TANZANIA Gelila Haile Aisha Ally Sinda Aman Assefa & Associates (Bowmans Alliance Firm) Bowmans Tanzania Wilbert Kapinga ESWATINI Bowmans Tanzania Derrick Ndo Jele Robinson Bertram OTHERS Olivier Jacquinot KENYA Progressus Corporation John Syekei Bowmans - Coulson Harney LLP Mike Jensen Rose Njeru Association for Progressive Communications Bowmans - Coulson Harney LLP Angela Mukora Bowmans - Coulson Harney LLP MAURITIUS Nafiisah Jeehoo Bowmans (Mauritius) Fariis Jahangeer Bowmans (Mauritius) Fazil Hossenkhan 142 >>> 2 ANNEX 1 >>> Annex 2: Summary of characteristics of segments along the digital value chain The Africa Digital Market Players Database aims to provide a comprehensive overview of actors that own or operate digital infrastructure, provide digital connectivity, or deliver other digital services along a set of pre-defined segments along the digital value chain. To create the database, the following steps were followed: • The first step was identifying and accessing 1-3 datasets for each market segment that would provide information on relevant actors and, if possible, market shares in the respective segment.319 Market share data could be provided by counting the number of infrastructure facilities owned in a country and segment, the number of subscribers, or the length of fiber networks. The final data used includes proprietary datasets (7), datasets scraped from web sources (3), and online reports and blogs (2); see the summary in the table below. • Following the compilation of data, various datasets together were merged into one master database and cleaned. Merging meant that the same firms that were present in more than one dataset were integrated into one line if they belonged to the same subsidiary on a country level. Information on individual segments was recorded in columns. For each segment, it was recorded at a minimum whether a firm was active in that segment (yes/no), the size of their ownership, networks or customers/subscribers, and their market share in this segment in this country. Recognizing the complex structure of internationally active corporations, one subsidiary was created for each country if the infrastructure was owned or services offered in this respective country. • As a final step, ownership information was researched and added to the database. For each firm, where available, it was recorded whether any government owned any part of it, or whether it was privately owned. Then, information on the extent of government ownership (wholly government-owned, majority government-owned owned and minority government- owned) and the presence of a foreign government was added. For both private and state-owned companies, an ultimate owner was denoted in a respective column, which was designated as the firm or government that owned the largest, and in most cases over 50 percent, of the subsidiary. See below table for a listing of all sources: Segment Sources Ownership of firms Orbis, Telegeography Global Comms database, desk research International connectivity Telegeography Global Internet Geography, desk research Wholesale fixed markets / fibre backbone Africa Bandwidth Maps, Telegeography Global Internet Geography, Afterfibre. com Spectrum Telegeography Global Comms, PolicyTracker Telecom towers TowerXChange Data centers Datacentermap.com, Xalam Analytics, Telegeography Data Center Research Service, PeeringDB, desk research Internet exchange points PeeringDB Mobile retail services Telegeography Global Comms Fixed retail services Telegeography Global Comms, Telegeography Global Internet Geography Mobile money GSMA, Budde.com, desk research 319. Disclaimer: This dataset was created with the most updated information that was accessible to the team. In cases of contradictions between data sources, the team consulted with colleagues working in the region and/or applied best judgment. <<< 143 >>> 1. Submarine cables Number of SOEs/ Number of SLEs with partial New Entrant Country Name HHI Level of Competition in IGW cables ownership 2017-2021 in a cable Algeria 5 1 3433 monopoly 1 Angola 3 2 4869 monopoly 1 Benin 2 1 8556 Restricted to few operators Botswana 1 - Burkina Faso 0 - Burundi 0 - Cabo Verde 4 1 5222 Restricted to few operators 2 Cameroon 6 2 2339 Restricted to few operators 3 Central Republic 0 - African Chad 0 - Comoros 3 2 4067 Restricted to few operators 1 Congo, Dem. Rep. 1 1 10000 Restricted to few operators Congo, Rep. 1 1 10000 Open to competition Côte d'Ivoire 4 1 3835 Open to competition Djibouti 9 1 1835 monopoly 2 Egypt, Arab Rep. 14 1 1179 Restricted to few operators 1 Equatorial Guinea 2 2 5918 monopoly 1 Eritrea - 0 Eswatini - 0 Ethiopia - - Gabon 2 1 8556 Open to competition Gambia, The 1 1 10000 Open to competition Ghana 5 1 2832 Open to competition Guinea 1 1 10000 Restricted to few operators Guinea-Bissau 1 1 10000 Restricted to few operators Kenya 5 3 3687 Open to competition 1 Lesotho - 0 - Liberia 1 1 10000 Monopoly Libya 4 2 4476 Monopoly Madagascar 3 1 5214 n/a 1 Malawi - 0 - 144 >>> ANNEX 2 Number of SOEs/ Number of SLEs with partial New Entrant Country Name HHI Level of Competition in IGW cables ownership 2017-2021 in a cable Mali - 1 - Mauritania 1 1 10000 Monopoly Mauritius 4 2 4567 Open to competition 2 Morocco 5 1 9223 Open to competition Mozambique 2 0 5531 Open to competition Namibia 1 1 10000 Monopoly Niger - 0 - Nigeria 6 1 2281 Open to competition Rwanda - 0 - Sao Tome and 1 0 10000 Restricted to few operators Principe Senegal 5 1 3437 Restricted to few operators 1 Seychelles 1 0 10000 Open to competition Sierra Leone 1 1 10000 Restricted to few operators Somalia 3 0 3858 n/a 2 South Africa 7 4 2279 Restricted to few operators 1 South Sudan 0 - Sudan 4 1 3223 Restricted to few operators Tanzania 3 1 5166 Open to competition Togo 1 1 10000 Restricted to few operators Tunisia 4 3 3372 Restricted to few operators Uganda 0 - Western Sahara 0 - Zambia - 1 - Zimbabwe 0 - Note: Reflects operational international submarine cables and their landing points; HHI based on maximum capacity of the operational cables per country as listed on Telegeography Global Internet Geography and supplemented with desk research <<< 145 >>> 2. Fiber backbone Number of Number of New entrants Country Name HHI Monopoly Duopoly firms SOEs/SLEs 2017-2021 Algeria 7 7 6464 no no 1 Angola 5 3 3166 no no 1 Benin 6 2 4782 no no 0 Botswana 5 3 4224 no no 2 Burkina Faso 8 5 2158 no no 2 Burundi 2 1 737 no yes 0 Cabo Verde 2 1 5669 no yes 0 Cameroon 6 4 3896 no no 1 Central African Republic 1 1 6366 yes no 1 Chad 2 2 9435 no yes 0 Comoros 1 1 10000 yes no 0 Congo, Dem. Rep. 4 2 2585 no no 1 Congo, Rep. 3 2 2397 no no 0 Côte d'Ivoire 9 4 1266 no no 0 Djibouti 1 1 10000 yes no 0 Egypt, Arab Rep. 2 2 6797 no yes 0 Equatorial Guinea 2 2 5104 no yes 0 Eritrea 1 1 10000 yes no 0 Eswatini 1 1 10000 yes no 1 Ethiopia 2 2 6216 no yes 0 Gabon 2 2 7354 no yes 0 Gambia, The 2 2 5006 no yes 0 Ghana 7 4 2675 no no 0 Guinea 10 5 3265 no no 0 Guinea-Bissau 6 4 1875 no no 0 Kenya 13 7 1356 no no 0 Lesotho 3 2 4982 no no 0 Liberia 5 4 3305 no no 0 Libya 2 2 7352 no yes 0 Madagascar 3 0 5156 no no 0 Malawi 4 2 3057 no no 1 Mali 4 3 3956 no no 0 Mauritania 4 3 5465 no no 40 Mauritius 5 2 3153 no no 2 146 >>> ANNEX 2 Number of Number of New entrants Country Name HHI Monopoly Duopoly firms SOEs/SLEs 2017-2021 Morocco 6 4 4802 no no 1 Mozambique 5 2 3126 no no 0 Namibia 3 2 5269 no no 1 Niger 4 1 6734 no no 1 Nigeria 22 7 937 no no 2 Rwanda 4 2 3959 no no 0 Sao Tome and Principe 0 no no Senegal 7 5 3493 no no 1 Seychelles 0 no no Sierra Leone 5 5 3583 no no 1 Somalia 9 0 1764 no no 2 South Africa 21 9 3283 no no 1 South Sudan 4 2 7318 no no 0 Sudan 5 2 3740 no no 2 Tanzania 9 6 2709 no no 3 Togo 5 2 4384 no no 1 Tunisia 5 5 5880 no no 2 Uganda 11 4 1480 no no 3 Western Sahara Zambia 5 2 3101 no no 0 Zimbabwe 7 5 2432 no no 0 Note: HHI based on km of operational fiber <<< 147 >>> 3. Telecom towers Number of Number of M&A deals Country Name HHI Monopoly Duopoly firms SOEs/SLEs since 2010 Algeria 4 3 3135 no no Angola 4 3 3946 no no Benin 0 no no Botswana 2 0 no yes Burkina Faso 4 1 2700 no no 2 Burundi 0 no no Cabo Verde 0 no no Cameroon 5 1 3988 no no 2 Central African Republic 0 no no Chad 2 0 726 no yes Comoros 0 no no Congo, Dem. Rep. 2 0 1643 no yes 2 Congo, Rep. 4 0 3418 no no 1 Côte d'Ivoire 3 1 4467 no no Djibouti 0 no no Egypt, Arab Rep. 5 2 2869 no no Equatorial Guinea 0 no no Eritrea 0 no no Eswatini 0 no no Ethiopia 1 1 10000 yes no Gabon 2 1 5022 no yes Gambia, The 0 no no Ghana 5 0 3489 no no 6 Guinea 0 no no Guinea-Bissau 0 no no Kenya 3 2 5307 no no 4 Lesotho 0 no no Liberia 0 no no Libya 0 no no Madagascar 4 0 3645 no no Malawi 2 1 no yes Mali 0 no no Mauritania 0 no no Mauritius 0 no no 148 >>> ANNEX 2 Number of Number of M&A deals Country Name HHI Monopoly Duopoly firms SOEs/SLEs since 2010 Morocco 3 2 3744 no no Mozambique 3 1 3698 no no Namibia 2 2 5144 no yes Niger 4 1 no no 2 Nigeria 9 0 3349 no no 9 Rwanda 1 0 5318 yes no 2 Sao Tome and Principe 0 no no Senegal 3 1 4106 no no 1 Seychelles 0 no no Sierra Leone 0 no no Somalia 0 no no South Africa 13 3 2633 no no 4 South Sudan 0 no no Sudan 0 no no Tanzania 5 2 3196 no no 3 Togo 0 no no Tunisia 3 3 4181 no no Uganda 2 1 6635 no yes 5 Western Sahara 3 0 Zambia 3 1 4733 no no 2 Zimbabwe 3 2 4400 no no Note: HHI based on number of towers <<< 149 >>> 4. Fixed retail Number of Number of New entrants Country Name HHI Monopoly Duopoly firms SOEs/SLEs 2015-2021 Algeria 1 1 10000 yes no 0 Angola 3 2 3932 no no 3 Benin 3 1 6898 no no 0 Botswana 1 1 10000 yes no 0 Burkina Faso 1 1 10000 yes no 0 Burundi 3 1 6038 no no 2 Cabo Verde 1 1 10000 yes no Cameroon 2 1 3798 no yes 0 Central African Republic no no Chad 1 726 no yes Comoros 0 no no Congo, Dem. Rep. yes 1643 no yes 2 Congo, Rep. no 3418 no no 1 Côte d'Ivoire 0 1 4467 no no Djibouti 0 no no Egypt, Arab Rep. 0 0 5882 no no 1 Equatorial Guinea 0 0 no no Eritrea 1 1 10000 yes no 0 Eswatini 4 1 9621 no no 2 Ethiopia 1 1 10000 yes no 0 Gabon 3 2 7426 no no Gambia, The 0 no no Ghana 0 0 no no Guinea 4 3 4050 no no 2 Guinea-Bissau 1 1 10000 yes no 0 Kenya 3 1 4445 no no 1 Lesotho 0 no no Liberia 4 1 5070 no no 0 Libya 0 no no Madagascar 1 0 10000 yes no 0 Malawi 7 2 2766 no no 2 Mali 0 no no Mauritania 1 1 10000 yes no 0 Mauritius 1 1 10000 yes no 0 150>>> ANNEX 2 Number of Number of New entrants Country Name HHI Monopoly Duopoly firms SOEs/SLEs 2015-2021 Morocco 0 no no Mozambique 0 no no 3 Namibia 2 1 5046 no yes 0 Niger 1 1 10000 yes no 0 Nigeria 1 1 10000 yes no 0 Rwanda 3 2 7418 no no 1 Sao Tome and Principe 1 0 yes no Senegal 1 1 10000 yes no 0 Seychelles 1 1 10000 yes no 0 Sierra Leone 0 no no Somalia 0 no no South Africa 4 1 6591 no no 1 South Sudan 0 no no Sudan 2 1 5000 no yes 0 Tanzania 4 1 4151 no no 0 Togo 4 1 4864 no no 2 Tunisia 6 4 2431 no no 0 Uganda 0 no no Western Sahara 3 0 no no Zambia 0 no no Zimbabwe 1 1 10000 yes no 0 Note: HHI based on the number of subscribers <<< 151 >>> 5. Mobile retail Number of Number of New entrants Country Name HHI Monopoly Duopoly firms SOEs/SLEs 2010-2021 Algeria 3 2 3454 no no 0 Angola 2 2 8200 no yes 0 Benin 2 0 5189 no yes 0 Botswana 3 1 3766 no no 0 Burkina Faso 3 1 3875 no no 0 Burundi 4 1 4064 no no 1 Cabo Verde 3 1 5099 no no 0 Cameroon 6 1 3856 no no 1 Central African Republic no no Chad 4 726 no yes Comoros 2 1 5976 no yes 1 Congo, Dem. Rep. 6 1 2912 no no 1 Congo, Rep. 4 1 5071 no no 1 Côte d'Ivoire 3 1 3432 no no 0 Djibouti 1 1 10000 yes no 0 Egypt, Arab Rep. 4 2 3162 no no 1 Equatorial Guinea 3 2 6945 no no 1 Eritrea 1 1 10000 yes no 0 Eswatini 2 2 6596 no yes 1 Ethiopia 1 1 10000 yes no 0 Gabon 2 1 5022 no yes 0 Gambia, The 4 1 4359 no no 0 Ghana 5 2 4065 no no 1 Guinea 5 1 4549 no no 0 Guinea-Bissau 3 1 5053 no no 0 Kenya 7 2 5144 no no 1 Lesotho 2 1 6125 no yes 0 Liberia 2 0 5005 no yes 0 Libya 3 3 5045 no no 0 Madagascar 4 0 3697 no no 1 Malawi 2 1 5051 no yes 0 Mali 3 1 4651 no no 1 Mauritania 3 1 3984 no no 0 Mauritius 3 1 3802 no no 0 152 >>> ANNEX 2 Number of Number of New entrants Country Name HHI Monopoly Duopoly firms SOEs/SLEs 2010-2021 Morocco 4 2 3388 no no 0 Mozambique 3 1 3935 no no 1 Namibia 4 2 8060 no no 0 Niger 4 1 3421 no no 0 Nigeria 8 1 2960 no no 2 Rwanda 4 1 5250 no no 2 Sao Tome and Principe 2 1 8132 no yes 1 Senegal 8 1 4130 no no 0 Seychelles 2 1 5166 no yes 0 Sierra Leone 4 1 4625 no no 1 Somalia 7 0 3007 no no 2 South Africa 34 2 3069 no no 1 South Sudan 3 0 5282 no no 3 Sudan 3 1 3633 no no 0 Tanzania 10 3 2422 no no 3 Togo 2 1 5012 no yes 0 Tunisia 4 3 3501 no no 0 Uganda 8 1 4612 no no 2 Western Sahara 3 0 no no 0 Zambia 3 1 3625 no no 0 Zimbabwe 5 3 5200 no no 0 Note: HHI based on the number of subscribers <<< 153 >>> 6. Data centers Number of Number of Country Name HHI Monopoly Duopoly firms SOEs/SLEs Algeria 5 1 3333 no no Angola 5 2 2500 no no Benin 0 no no Botswana 0 no no Burkina Faso 0 no no Burundi 0 no no Cabo Verde 0 no no Cameroon 2 0 5000 no yes Central African Republic 0 no no Chad 0 no no Comoros 0 no no Congo, Dem. Rep. 2 0 5000 no yes Congo, Rep. 0 no no Côte d'Ivoire 4 1 10000 no no Djibouti 2 1 5000 no yes Egypt, Arab Rep. 7 1 2449 no no Equatorial Guinea 0 no no Eritrea 0 no no Eswatini 0 no no Ethiopia 1 0 10000 yes no Gabon 1 0 10000 yes no Gambia, The 1 1 10000 yes no Ghana 9 1 1111 no no Guinea 1 1 10000 yes no Guinea-Bissau 0 no no Kenya 14 3 802 no no Lesotho 0 no no Liberia 0 no no Libya 0 no no Madagascar 0 no no Malawi 1 1 10000 yes no Mali 0 no no Mauritania 0 no no Mauritius 8 1 1600 no no 154 >>> ANNEX 2 Number of Number of Country Name HHI Monopoly Duopoly firms SOEs/SLEs Morocco 9 2 1200 no no Mozambique 1 0 yes no Namibia 0 no no Niger 0 no no Nigeria 16 1 667 no no Rwanda 2 0 5000 no yes Sao Tome and Principe 0 no no Senegal 0 no no Seychelles 0 no no Sierra Leone 0 no no Somalia 0 no no South Africa 30 3 600 no no South Sudan 0 no no Sudan 1 1 10000 yes no Tanzania 4 0 2500 no no Togo 0 no no Tunisia 6 3 1667 no no Uganda 3 1 3333 no no Western Sahara 3 0 no no Zambia 3 1 4400 no no Zimbabwe 2 1 6800 no yes Note: HHI based on number of facilities <<< 155 >>> 7. Mobile money Number of Number of New entrants Country Name HHI Monopoly Duopoly firms SOEs/SLEs 2017-2021 Algeria 0 1 no no Angola 0 2 no no Benin 2 2 5473 no yes 0 Botswana 4 1 4676 no no 2 Burkina Faso 3 0 no no 1 Burundi 4 0 no no 0 Cabo Verde 0 1 no no Cameroon 5 2 3353 no no 1 Central African Republic 2 0 yes no 0 Chad 2 0 no yes 0 Comoros 0 2 no no Congo, Dem. Rep. 4 1 3183 no no 0 Congo, Rep. 2 1 no yes 0 Côte d'Ivoire 7 1 3591 no no 1 Djibouti 0 1 no no Egypt, Arab Rep. 4 1 no no 0 Equatorial Guinea 0 2 no no Eritrea 0 0 no no Eswatini 1 0 10000 yes no 0 Ethiopia 2 1 no yes 1 Gabon 3 1 no no 0 Gambia, The 2 1 no yes 0 Ghana 6 1 4358 no no 2 Guinea 3 1 6458 no no 1 Guinea-Bissau 2 1 5503 no yes 1 Kenya 5 3 6955 no no 1 Lesotho 2 0 no yes 0 Liberia 2 1 no yes 1 Libya 0 2 10000 no no Madagascar 3 0 no no 0 Malawi 3 0 no no 0 Mali 3 0 no no 1 Mauritania 1 1 yes no 0 Mauritius 1 2 yes no 1 156 >>> ANNEX 2 Number of Number of New entrants Country Name HHI Monopoly Duopoly firms SOEs/SLEs 2017-2021 Morocco 4 1 no no 3 Mozambique 3 0 no no 0 Namibia 2 1 no yes 0 Niger 3 0 no no 0 Nigeria 18 0 no no 3 Rwanda 3 0 3152 no no 0 Sao Tome and Principe 0 0 no no Senegal 6 1 no no 3 Seychelles 1 1 yes no 0 Sierra Leone 3 1 no no 0 Somalia 4 0 no no 0 South Africa 6 5 no no 4 South Sudan 2 0 no yes 2 Sudan 2 1 no yes 1 Tanzania 6 1 2817 no no 1 Togo 2 1 no yes 0 Tunisia 4 3 no no 0 Uganda 6 0 no no 0 Western Sahara 3 0 no no 0 Zambia 7 1 3216 no no 2 Zimbabwe 3 0 8672 no no 1 Note: HHI based on the number of subscribers <<< 157 >>> Annex 3: Governance risks indices and methodological notes The governance risk index encompasses the risks to economic governance of digital markets identified in the report: • Market characteristics • Foundational digital infrastructure institutions • Presence of SOE-SLEs • Taxation of digital infrastructure • Data regulation • Competition law and enforcement For each of these elements, the team identified key metrics that were then weighted to capture the level of risks to economic governance of the digital economy. These metrics should not be read as isolated risk factors but considered as compounded risk elements. A market structure that presents high levels of concentration together with a weak institutional framework for digital infrastructure, presence of dominant SOE or SLEs that are not transparent, high levels of excise-like taxes, absent or weak data protection frameworks, and absent or weak competition regulation and enforcement in digital markets will generate the highest risk to the economic governance of the digital economy. Analysis based on these digital economy governance risk indices should be conducted considering some caveats. The index and subindices are constructed based on the data available for each analyzed country. Data was not always available for all countries or, in other cases, for the same period. The latest information available was considered. Additionally, the analyzed metrics are proxies for the elements that the report analyzes. There is no general metric to measure market structure or data regulation. The team gathered available databases and metrics that could provide a clearer understanding of the levels of risks, but this inevitably leaves other relevant aspects out of the analysis. MARKET STRUCTURE The market structure risk index aims at capturing concentration (see Section 1 for a discussion of market concentration and indicators) across the value chain and vertical integration of the largest operators in key segments and degree of entry in retail markets. To provide a clearer view of concentration, the team identified seven segments in which concentration could affect the provision of digital services. These include submarine cables, telecom towers, backbone fiber, mobile retail, fixed broadband retail, data centers, and mobile money. The HHI on these markets was retained as a relevant proxy to the capacity to enable competition indigital infrastructure. Annex 2 lists the sources of HHI data (See Section 1 for a discussion of analytical underpinnings) in these segments. With these data, the team identified the percentage of segments in which the HHI was higher than 5000 points (equivalent to only having two operators with similar market share) and the percentage of those that fell between 2500-5000 points. The percentage of segments that fall into these two categories represents how concentrated segments are along the value chain. The higher the percentage, the higher the concentration across sectors and the risk to economic governance. More concentrated markets need more active regulation to enable contestability or mimic competition through regulatory interventions. 158 >>> ANNEX 3 Another relevant proxy for market structure is the entry and exit of participants in mobile markets. Mobile broadband in Africa has grown exponentially in the last ten years, and how the market structure reacted to this growth is relevant to understanding market dynamics. Based on the data on a number of mobile operators provided by Telegeoography (2021), the team analyzed whether the mobile service segment had exits or entries in the last ten years. The team coded the data, giving it 1 point if net exits had occurred in the last ten years, 0.5 if no change in the number of players had taken place, and 0 if entry had occurred. Net exits indicate potential issues in the sector, while net entries is a sign of a more attractive enabling environment and better economic governance. Finally, vertical integration was also taken into consideration as a structural indicator. Understanding whether large (and potentially dominant) providers are vertically integrated and can exercise this power across the value chain is relevant to understanding how markets operate. The team identified the existence of an MNO or fiber backbone provider with more than 40 percent of the market share that is vertically integrated into two or more of the identified segments. This element was given a value of 1 if an integrated large provider existed and 0 if no such provider was found. Data was not available for all the countries included in the report. In this sense, the index may not be as accurate for countries where data was unavailable for one of the analyzed segments. Additionally, whether new companies joined or exited mobile markets does not capture the nuance of the causes of why companies enter or exit a market. The combination of data on HHI, market entrants, and vertical integration gives a general overview of whether the market structure is more prone to competition or not. The development of competition would need more attention from regulators to ensure an enabling environment in case the risks associated with the current market structure are significant. F O U N D AT I O N A L D I G I TA L I N F R A S T R U C T U R E I N S T I T U T I O N S Generating an enabling environment that allows for the development of digital technologies while preventing the risk associated with this growth requires independent and efficient regulators. The role of regulators of digital infrastructure is particularly relevant since, as described in this report, digital connectivity infrastructure is the base of digital technologies. Consequently, understanding whether the country has the regulatory structure to promote competitive and efficient telecommunication services is essential to analyzing the risks of economic governance of digital technologies. In consequence, the traditional institutions’ risk index is calculated as the simple average of four variables: (i) the age of the telecommunication regulator; (ii) the independence of that regulator; (iii) the level of collaboration between different regulators; (iv) whether the country has instituted ex-ante pro-competition regulation. The age of the telecommunications regulator is a relevant proxy for the experience of both the organization and its staff handling such topics. The 2021 Teleography’s Country Profiles were used as a base to retrieve the year in which telecommunication regulators were founded. From that year, the team calculated how many years had passed up to the present. The higher the number of years, the more experienced the regulator. Whether the telecommunications regulator is independent from the central government is also relevant to understanding the risks of developing digital technologies. A more independent authority could make technical decisions without the pressures of political officials. Based on Annex X of the “Background Paper: Competition Policy in Digital Markets in Africa,” the team gathered data on whether the telecommunications regulator was a unit within a ministry or an independent organization. One point was awarded if the authority was not independent (part of the ministry), and 0 if it was. <<< 159 Digital technologies, such as those presented in this report, involve the participation of several regulators. Whether these organizations collaborate is relevant to providing a unified and cohesive regulatory framework. The team gathered the data provided by ITU’s G5 regulation country profiles for 2021. Pillar I of this report measures National Collaborative Governance, including cooperation among ICT regulators and cooperation with other sector agencies. This index goes from 0 to 29.63. The team estimated the gap in collaboration by calculating the percentage out of the maximum possible points for the countries included in this report. A higher percentage represented a lower level of collaboration. The tools that the regulator has to promote competitive and efficient markets are also a relevant proxy for a country’s institutional framework. The team used the data provided by ITU’s Regulatory Tracker for 2018, particularly whether ex-ante pro-competition regulations were in place. The index goes from 1 to 20. The team calculated the percentage of the gap on the ex-ante index; a higher percentage means that the regulator has fewer ex-ante tools. Ex-ante pro-competition regulations include SMP determination, interconnection, number portability, infrastructure sharing, spectrum trading, and neutrality. As a caveat for the selected proxies, the age of the regulator does not capture the quality and depth of the work conducted by the regulator. In some cases, a regulator that has been in operation for many years may not have effectively promoted efficient digital infrastructure markets, while a younger one may have been very effective in impacting the market. Nonetheless, the existence of a telecommunications regulator shows that a country has dedicated resources and efforts toward this task and will, in general, be helpful in providing a proxy for institutional strength. Indicators on transparency and more detailed information on independence factors (election of the head of the authority, budget, etc.) were not readily available for all countries, so they were not considered. These indicators together provide a relevant proxy on the experience of the regulator, its ability to make independent technical decisions, and the tools that the regulator has should promote more competitive markets for the development of digital technologies. Strong and independent regulators are necessary to prevent economic governance risks in digital infrastructure markets. SOE-SLES The presence of SOEs and SLEs in digital markets can pose a risk to the economic governance of the sector through distortions on the level playing field or weak performance. As described in the report, SOEs and SLEs320,321 can raise the risk of preferential treatment that undermines competition on the merits, especially when there is lack of transparency and accountability in the operation of these enterprises. To analyze the risk of potential differential treatment due to ownership, five proxies were identified: (i) number of digital SOEs and number of digital SLEs (SOE-SLE index), (iii) importance of SOE in digital sectors; (iv) opacity of non-financial information; (v) opacity of financial information. The number of digital SOEs and digital SLEs was calculated based on TeleGeography information complemented with desk research. The SOE-SLE index was calculated based on the number of SOEs and SLEs with the following formula: number of SOEs + number of SLE/2)/number of segments. This proxy allows for a better understanding of the presence and influence of SOEs and SLEs across segments, giving a lower weight to SLEs – where private players exert management control. The importance of SOEs in digital sectors is proxied by the largest market share of an SOE in a market segment. A larger market share in a segment shows the relevance of SOEs and their influence on the development of digital technologies. Transparency is relevant for accountability. Allowing stakeholders to analyze how SOEs operate and perform is relevant to promoting better governance. Consequently, the team analyzed the number of items disclosed for financial and non-financial 320. For the purpose of the index, an SOE is defined as a firm with State ownership of at least 50 percent and an SLE where the government owns any stake less than 50 percent. 321. As described earlier in this report and in the ”Digital SOE-SLE Governance in Africa” that served as a background paper for this report. SOE includes all corporate entities recognized by national law as an enterprise, in which the national or subnational government exercises ownership. In turn, SLEs are those companies in which the state holds a minority shareholding. 160>>> ANNEX 3 data (See Jaupart and Begazo 2023 for more details). The team calculated the percentage of the opacity of the published data. Out of 14 possible elements in the case of non-financial data, a higher percentage indicates a lower level of publication. In the case of financial data, the team analyzed the lack of publication of financial statements or annual reports. A score of 100% implies no publication, while a score of 0% indicates that reports are published. Additional indicators on the differential treatment of SOE/SLE were not available for all countries and, therefore, were not considered in the index. The number of SOEs and SLEs, combined with the market shares they hold and their operations’ transparency, allows for a clearer understanding of how these companies may impact digital markets. The presence of SOEs and SLEs in relevant digital segments does not necessarily impede the development of digital services. Promoting competition in infrastructure markets and increasing transparency for all SOEs and SLEs is essential to prevent state ownership of service providers from negatively impacting the sector. TA X AT I O N O F D I G I TA L I N F R A S T R U C T U R E As explained in this report, over-taxation of digital-related services and infrastructure negatively impacts the governance of the sector through the dependency it creates between the government and service providers. To calculate the risk of over-taxation, the team analyzed the simple average of two variables: (i) number of segments with sector- specific taxes and (ii) average import duties for handsets and base stations. Data was obtained from ITU and complemented with data provided by the contributors listed in Annex 1 and desk research. Sector-specific taxes in eight segments were considered, including incoming international voice services, international data services, international mobile roaming, national data services, national mobile roaming, national voice services, outgoing international voice services (IDD), and internet services. Out of these eight possible segments, the team calculated the percentage in which each country had instituted sector-specific (excise-like) taxes. The data for this analysis was retrieved from ITU’s 2020 database, complemented by the World Bank’s Digital Infrastructure and Taxation database. As developed in this report, handsets and base stations are particularly relevant to the development of digital infrastructure and adoption. In this sense, the team analyzed the average import duty of each country as a percentage of the maximum import duty in Africa, which is 25% for handsets and 20% for base stations. Data from WTO was used to construct this indicator. Whether sector-specific taxes are in place on top of regular direct and indirect taxes and in how many segments, combined with the level of import duties on the infrastructure necessary for digital services, are relevant proxies to understand the risk that over-taxation could have in digital connectivity. D ATA R E G U L AT I O N Data regulation is essential to both safeguard users’ safety and privacy and to enable the reuse of data. Preventing cybercrime and promoting cybersecurity is also essential to avoid online threats. In this sense, the data regulation risk index is calculated as the simple average of three variables: (i) absence of data protection framework, (ii) weakness of cybersecurity framework, (iii) gaps in the data policy framework. <<< 161 Countries have decided to regulate data protection through different mediums. These include general frameworks, sectoral rules, or general provisions in the constitutions. The team collected data on which of the analyzed countries had opted for each option and those with no protection framework. The team then granted each a score from 0 if the country had a general framework, 0.25 if there were sectoral rules, 0.5 if protection was included in the constitution, and 1 if there was no framework (See Chen, Daza Jaller and Begazo 2023). The chosen system is a proxy for the relevance of data protection in that particular country and the efforts made to protect and promote data use. The weakness in the cybersecurity framework was calculated based on ITU (2020). This index’s score has a maximum of 100 points. The team then calculated the percentage the country was missing to reach that perfect score, with a higher percentage implying a weaker system. Based on the WDR Global Data Regulation Index 2020 described in “Mapping Data Governance Legal Frameworks Around the World Findings from the Global Data Regulation Diagnostic” working paper (Rong 2020), the team proxied the gaps in data regulation frameworks. A higher percentage implies a more significant gap. A caveat to these proxies is that the existence of a data protection framework does not imply that the system is effectively protecting data. The enforcement and the resources available for that enforcement are equally relevant for the adequate protection of data and the prevention of cybercrime. Understanding how the country regulates data protection, cybersecurity, and the gaps in regulation is a relevant proxy to understand whether the system has minimum conditions to be able to protect data as digital services develop or if they would pose a risk to the economic governance of digital markets. COMPETITION AUTHORITIES AND COMPETITION LAW AND ENFORCEMENT The enforcement of competition rules is essential for the development of a safe and vibrant digital environment. The low or non-existent enforcement of competition rules would potentially imply that sub-optimal market outcomes associated with anticompetitive behavior, mergers that raise competition concerns, and rules that prevent competition may affect the economic governance of digital markets and result in biased outcomes that favor infringers. Consequently, the competition law and enforcement risk index is calculated as the simple average of two variables used to proxy this element: (i) absence of a competition authority; (ii) absence of enforcement or advocacy actions on digital markets. The presence of a competition authority is a relevant proxy to the institutional framework to enforce competition policies in a country. This element of the risk index was weighted 0 if there was a competition authority in operation and one if there was no agency322 in charge of competition law. Data was retrieved from the “Background Paper: Competition Policy in Digital Markets in Africa” prepared as a background paper for this report and complemented with desk research. Digital markets pose challenges to competition authorities. In this sense, analyzing whether the country has not taken actions to address these challenges is a proxy for how the country has reacted to the policies and behavioral challenges related to the development of digital technologies. Zero points were granted to those countries that had taken action through both advocacy and enforcement actions, 0.5 if only one of those was taken, and 1 if no action had been taken. Data was extracted from the “Competition Advocacy for Digital Markets in Africa” Background Paper. Having a competition authority that at least has handled 322. As described in the “Competition Policy in Digital Markets in Africa” background paper, in some cases the Telecommunication Authority is in charge of implementing competition rules, while in others both the Competition Authority and the Telecommunication Regulator have authority over competition-related matters. Examples of the first one includes countries in which there is no functioning competition agency, such as Cabo Verde, Guinea, Liberia, Mauritania, Rwanda (before 2021), São Tomé e Príncipe and Tchad, those where the competition regulator has no enforcement powers, including Burkina Faso, Côte d’Ivoire, Senegal, Central Africa Republic, Republic of Congo, Democratic Republic of Congo, and Morocco where the telecom agency is the only one in charge of enforcing competition law. 162 >>> ANNEX 3 a case or inquiry in digital markets puts the country in a better situation to address potential competition problems in digital markets in the future, reducing the risk on economic governance. Two caveats should be introduced. First, the absence of enforcement or advocacy actions in digital-related markets does not imply underenforcement of competition law. As described in the “Competition Policy in Digital Markets in Africa” paper that served as a background paper to this report, the lack of action by competition authorities may just show that digital markets are still nascent and do not generate any competition-related concern or that there are other enforcement priorities. Second, and similar to the case of the data regulation risk index, the existence of a Competition Authority does not automatically imply that the regulator is effective and efficient and has enough tools to deal with issues in digital markets. The absence of more detailed indicators of agency effectiveness across all African countries does not allow to fully capture nuances of the challenges faced by these organizations in terms of ensuring an effective enforcement of competition policy. >>> 1. Market Structure <<< 163 The market structure risk index is calculated as the simple average of three variables: • average concentration along 7 digital segments (6 when mobile money information is not available in the Digital market players database) (as measured by HHI based on subscribers, capacity, km, accounts, or data sq meters depending on the data available for each segment), • indicator on exits in mobile services (1 if net exits have occurred in the last 10 years, 0.5 if no change in number of players, and 0 if entry has occurred), and • indicator on potentially problematic vertical integration (existence of a MNO or fiber backbone provider with more than 40 percent of market share vertically integrated into two or more other segments). 164 >>> ANNEX 3 >>> 2. Foundational Digital Infrastructure Institutions <<< 165 The traditional institutions risk index is calculated as the simple average of three variables: • Lack of independent regulator (1 if regulator is a department within the ministry or there is no regulator) • Gaps on collaboratively policy making between telecom regulator and other authorities (1- ITU G5 index’ pillar on collaboration (2020) • Lack of ex ante pro-competition regulation (based on ITU regulatory tracker capturing regulations such as SMP, interconnection, number portability, infrastructure sharing, spectrum trading and neutrality (2018)). 166>>> ANNEX 3 >>> 3. SOE-SLEs <<< 167 The SOE-SLE risk index is calculated as the simple average of three variables for 2020: • Existence of SOE and SLE ((number of SOE + number of SLE/2)/number of segments) • Importance of SOE in the sector (proxied by the largest market share of a SOE in a segment) • Opacity of non-financial information (number of items not disclosed on the SOE-SLE website as proportion of 14 items, avg. by country) and opacity of financial information (financial statements or annual report not disclosed on the SOE-SLE website, avg. by country) 168>>> ANNEX 3 >>> 4. Taxation of digital infrastructure <<< 169 The digital infrastructure over-taxation risk index is calculated as the simple average of two variables: • Number of segments with sector-specific taxes (higher VAT, excise tax or other on consumption) among 8 segments (based on ITU database, 2020 and complemented with World Bank’s Digital Infrastructure and Taxation database) • Average import duties for handsets and base stations as percentage of maximum import duty in Africa (25 percent, 20 percent) (WTO data 2021) 170 >>> ANNEX 3 >>> 5. Data regulation <<< 171 The data regulation risk index is calculated as the simple average of three variables: • Absence of data protection framework (0 if there is a general framework, 0.5 if there are sectoral rules, 0.25 if there are general provision in the constitution, 1 if the countries doesn’t have a framework). Data as of 2021. • Weakness of cybersecurity framework (1- GCI index calculated by ITU (2020)) • Gaps in the data policy framework (1- WB WDR data regulation index (2020)). This variable is excluded when not available for a country. 172 >>> ANNEX 3 >>> 6. Competition law and enforcement <<< 173 The competition law and enforcement risk index is calculated as the simple average of two variables: • Absence of a competition authority (0 if there is a competition authority in operation) by end 2021 • Absence of enforcement and/or advocacy actions on digital markets (1 if no action, 0.5 if either an antitrust case or advocacy action, 0 if both enforcement and advocacy actions were issued by the competition authority) by end 2021 174 >>> ANNEX 3 GOVERNANCE AND THE DIGITAL ECONOMY IN AFRICA MAIN REPORTS VOLUME 1 DIGITAL FOR GOVERNANCE: REACHING THE POTENTIAL FOR THE DIGITAL ECONOMY IN AFRICA — DIGITAL TOOLS FOR BETTER GOVERNANCE VOLUME 2 GOVERNANCE OF DIGITAL: REGULATING THE DIGITAL ECONOMY IN AFRICA — MANAGING OLD AND NEW RISKS TECHNICAL BACKGROUND PAPERS • ICT Procurement In Africa • Adoption of EGP in Africa • Vulnerabilities of ICT Procurement to Fraud and Corruption • Regulating Digital Data in Africa • Taxes and Parafiscal Fees on Digital Infrastructure Services in Africa • Corporate Governance and Transparency of State-Owned and State-Linked Digital Enterprises in Africa • State-Owned Enterprises in Digital Infrastructure And Downstream Digital Markets in Africa • Competition Advocacy for Digital Markets in Africa • Competition Policy in Digital Markets in Africa