Feasibility Study to Connect all African Higher Education Institutions to High-speed Internet A © 2021 The World Bank 1818 H Street NW, Washington DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved This work is a product of the staff of The World Bank. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. 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. 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. Attribution—Please cite the work as follows: “World Bank and Knowledge Consulting Ltd. 2021. “Feasibility Study to Connect all African Higher Education Institutions to High-Speed Internet”. All 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. B Contents List of Figures I List of Tables III Abbreviations IV Acknowledgements VI Executive Summary VIII 1. Introduction 1 2. Methodological Summary 4 3. Context and Connectivity Gaps in Higher Education 6 4. Cost Estimates for Connecting Higher Education 23 5. Leveraging Broadband Connectivity for Improved Learning Outcomes 39 6. Looking Ahead 44 Appendix A: Tables 50 Appendix B: Summary of Country Case Studies—Burkina Faso, Cote d’Ivoire, Mozambique and Uganda 51 I List of Figures Figure Title Page No Summary of total cost (USD, billions) of connecting all African higher Figure 1 XIII education institutions to high-speed internet Figure 2 Map of active undersea cables around Africa (proposed indicated in gray) 9 Figure 3 Global inter-regional internet bandwidth 9 Figure 4 Terrestrial fibre within and around Africa 10 Figure 5 Coverage of African regional RENs 12 Figure 6 Types of institutions served by African NRENs 13 Figure 7 Amount for 1 Mbps/month charged by different NRENs 15 Figure 8 Tangible and intangible barriers for connecting higher education 19 Figure 9 Matrix for determining bandwidth cost 33 Price drop (USD) for 2xSTM-1 10-year IRU between Nairobi and Kigali with Figure 10 34 drop in Kampala A schematic diagram to estimate the costs for connecting higher education Figure 11 38 institutions in Africa Summary of total cost (USD, billions) of connecting all African higher Figure 12 39 education institutions to high-speed internet II List of Tables Table Title Page No Table 1 Prioritizing connectivity to African higher education XV Table 2 Methodological Summary 4 Table 3 Higher education student population in Africa (rounded to nearest 100) 8 Table 4 NREN members of RRENs in Africa 11 Table 5 Stages of NREN development 14 Table 6 Traditional NREN services 15 Table 7 Comparison of bandwidth prices across NRENs 16 Table 8 Stages of the internet Exchange Ladder 18 Table 9 Major previous and current partners for university connectivity projects in Africa 20 Table 10 Recommended Progressive Bandwidth Targets for African HEIs 25 Table 11 Average device dosts in a 1:1 computing scenario 28 Table 12 Cost of equipping students and staff with access devices (2021–2025) 29 Table 13 Assumptions for calculating campus network upgrade costs in a country 29 Table 14 Summary estimates for upgrading campus networks by region (rounded to nearest 10) 31 Table 15 Projected bandwidth by region using student enrollment (2021, 2025, & 2030) 33 Table 16 Variation in unit cost price for bandwidth by region 34 Table 17 Total bandwidth cost for all African countries by region using student enrollment (2025) 35 Summary of total cost of connecting all African HEIs to upstream bandwidth for five Table 18 35 years Table 19 CapEx and OpEx elements for NREN development 35 Table 20 Support for Core Costs of African NRENs 36 Table 21 Cost drivers for UA budget 37 Table 22 Support for Core Costs of African RRENs 37 Table 23 List of key stakeholders 47 Table 24 Prioritizing connectivity to African higher education 48 Table 25 Maturity of NRENs in different countries 51 Table 26 Comparison of factors and costs across case study countries (2021–2025) 58 III Abbreviations Term Description AfCFTA African Continent Free Trade Agreement AfDB African Development Bank AfriNIC African Internet Registry ASN Autonomous System Number ASREN Arab States Research and Education Network AUP Acceptable Use Policy BYOD Bring Your Own Device CAP Country Action Plan CapEx Capital Expenditures COVID-19 Coronavirus Disease 2019 CSP Commercial Services Providers DDP Digital Development Partnership DE4A Digital Economy for Africa initiative DEA Direct Engineering Assistance DNS Domain Name System DS4DE4A Digital Skills for Digital Economy in Africa EMIS Education Management Information System FTTX Fiber to the X (home, office, school kerb etc.) Gbps Gigabits Per second GER Gross Enrollment Ratio GIS Geographical Information System HEI Higher Education Institution ICT Information and Communications Technology IDRC International Development Research Centre IFC International Finance Corporation INASP International Network for Availability of Scientific Publication IRU Indefeasible Right of Use ISCED International Standard Classification of Education ISOC Internet Society IV Term Description ISP Internet Service Provider ITU International Telecommunication Union IXP Internet Exchange Point KCL Knowledge Consulting Ltd Mbps Megabits per Second MDAs Ministries, Departments and Agencies of Government MOOC Massive Online Open Course NREN National Research and Education Network NSRC Network Startup Resource Center O&M Operations and Maintenance OpEx Operational Expenditure OSI Open Society Initiative PoP Point of Presence RCIP Regional Communications Infrastructure Program (World Bank) RREN Regional Research and Education Network SDG Sustainable Development Goal SPOC Small Private Online Course Tbps Terabits per Second TVET Technical and Vocational Education and Training UA UbuntuNet Alliance UIS UNESCO Institute of Statistics UNESCO United Nations Education Scientific and Cultural Organization WACREN West and Central African Research and Education Network WBG World Bank Group Wi-Fi Wireless – Fidelity – Local area wireless computer networking technology V Acknowledgements T he report was prepared by a team led by Samia (Practice Manager for Eastern and Southern Africa, Digital Melhem (Lead Digital Development Specialist) Development). The team is grateful to peer reviewers – and Tim Kelly (Lead Digital Development Alex Twinomugisha (Senior Education Specialist) from Specialist) and comprising: Lucine Munkyung the Education Global Practice, and from Tounwende Park (Digital Development ET Consultant), Charles Alain Sawadogo (Senior Digital Development Specialist), Hurpy (Senior Digital Development Specialist) and Xavier Stephane Decoster (Senior Digital Development Sajitha Bashir (Adviser, Office of the Global Director for Specialist), Wilson Muyenzi (Digital Development Education). Knowledge Consulting Limited (KCL) provided ET Consultant), and Casey Torgusson (Senior Digital advice, analysis, and drafting support, with special Development Specialist) from the Digital Development acknowledgement to Francis F. Tusubira (Managing Global Practice of the World Bank – for their insightful Partner, KCL) who led the team which comprised Lishan comments and inputs. The team would also like to thank Adam, Ali Ndiwalana, Jules Degila, and Fekadu Mulugeta. additional guidance and contributions provided by Javed The team is grateful for the valuable contributions of I. Khan (Consultant) and Ekua Nuama Bentil (Education Scribendi (Editorial firm) and Arnold Birungi (Graphic Specialist). Designer). We also acknowledge the contribution of the Regional The team benefited from the overall guidance provided and National Research Education Institutions (RENs) that by Boutheina Guermazi (Global Director, Digital responded to the survey and shared valuable data. We Development), Mark Williams (Practice Manager for Global would particularly like to thank the following individuals Knowledge and Expertise, Digital Development), Michel and organizations that have provided data, information, Rogy (Practice Manager for Western and Central Africa and and insights (by alphabetical order of organizations): the Middle East, Digital Development), and Isabel Neto Organization Interviewee/Respondent African NRENs 1 Benin (RBER) Adéfèmi Christelle Agossou 2 Burundi (BERNET) Grégoire Njejimana and Pierre-Claver Rutomera 3 Chad (TchadREN) Zaki Sabit 4 Côte d'Ivoire (RITER) Issa Traoré 5 Ethiopia (EthERNet) Zelalem Assefa 6 Gabon (GabonREN) Ousmane Balira Konfe and Anicet Andjouat 7 Ghana (GARNET) Lucas Chigabati and Emmanuel Togo 8 Ghana (GhREN) Benjamin Eshun 9 Guinea (Gn-REN) Kodiougou Diallo 10 Kenya (KENET) Meoli Kashorda 12 Madagascar (iRENALA) Harinaina Ravelomanantsoa 13 Malawi (MAREN) Solomon Dindi 14 Mali (MaliREN) Pierre C. B. Traoré 15 Morocco (MARWAN) Redouane Merrouch 16 Mozambique (MoRENet) Lourino Chemane 17 Nigeria (NgREN) Joshua Atah, Patricia Eromosele, Gaurav Gupta 18 Senegal (SnRER) Ibrahima Niang 19 Sierra Leone (SLREN) Thomas Philip Songu VI 20 Somalia (SomaliREN) Abdullahi Bihi Hussein 21 South Africa (TENET) Duncan Greaves 22 Tanzania (TERNET) Magreth Mushi 23 Togo (TogpRER) Eyouleki Palanga 24 Tunisia (CCK) Habib Youssef Other NRENs 1 GRENA (Georgia) Ramaz Kvatadze 2 AMRES (Serbia) Bojan Jakovljevic 3 CENIC (California) Louis Fox 4 JISC (UK) Rob Evans 5 Red CEDIA (Ecuador) Juan Pablo Carvallo 6 RNP (Brazil) Eduardo Cezar Grizendi RRENs 1 ASREN Yousef Torman 2 GÉANT Cathrin Stöver, Daniel Wustenberg, Leila Dekkar 3 Red CLARA Luis Escadenas 4 UbuntuNet Alliance Tiwonge Banda and J Kimaili 5 WACREN Boubakar Barry Major Backbone Services Providers 1 Liquid Telecom Ben Roberts 2 SEACOM Michael Otieno Other Key Informants 1 ISOC Michuki Mwangi 2 NSRC Steve Huter and Steve Song Lastly, this report could not have been produced without financial support from the members of the Digital Development Partnership (DDP), a Trust Fund administered by the WBG. The DDP offers a platform for digital innovation and development financing, bringing public and private sector partners together to advance digital solutions and drive digital transformation in developing countries, see: https://www.worldbank.org/en/programs/digital-development-partnership. VII Executive Summary B roadband connectivity is a critical enabler for Vision for Broadband Connectivity for modernizing higher education institutions (HEIs) in their mission of teaching, research, and Higher Education in Africa community outreach. Connecting African HEIs for improved learning, research collaboration, and access As a first step toward achieving sufficiency of bandwidth, to global scientific resources has been on national and a process that combined consultation with African global development agendas for many years but has never stakeholders and international benchmarking was achieved top priority policy consideration.1 The higher conducted, leading to the formulation of a “vision for education sector in Africa falls far behind the rest of the broadband connectivity” for HEIs in Africa: world in connecting to global research and education networks. The available bandwidth is generally expensive vision and limited in capacity and therefore cannot meet modern institutions’ research and education requirements. An African continent where all higher education institutions Significance of Broadband Connectivity achieveglobal parity in intellectual output and development impact for HEIs and Research and Education throughaccess to, and exploitation Networks in Africa of broadband connectivity at capacities,quality, and costs The COVID-19 pandemic highlighted the urgent need to comparable to the rest of the world. extend broadband infrastructure even further to facilitate teaching, learning, research, access to educational resources, and the attainment of effective administration in higher education. The COVID-19 pandemic led to global lockdowns and made any physical face-to-face approach To translate this to national and regional development to learning difficult. Because of the many major gaps in impact, African HEIs must develop the necessary pre- broadband connectivity, and the limited preparedness conditions to ensure that the sufficiency and affordability across the continent, most African universities were not of broadband can be seized as opportunities to improve able to participate in online teaching, which immediately learning and research outcomes as well as employability in kicked in as an alternative form of educational delivery the context of the fourth industrial revolution. in developed countries. This underscored the urgency of enhanced connectivity for HEIs, placing immediate corrective action high on the agenda of governments, educators, and development partners. Status of Higher Education Broadband Connectivity Current gaps in broadband connectivity of higher education in Africa highlight the importance of connecting over 15 million students and about 500,000 staff in higher education. Needed improvement in broadband connectivity has to be carried out across the broadband value chain—at the international level where connectivity enters the country, at cross-border and regional levels, nationally, and at the campus level.The continent is in its 1 The study uses the term higher education, also known second wave of submarine cable rollout on the western, as tertiary education in some countries, to refer to eastern, and southern coasts, presenting tremendous all post-secondary education, including both public opportunities for interconnecting the higher education and private universities, colleges, technical training sector. institutes, and vocational schools (https://www. worldbank.org/en/topic/tertiaryeducation). VIII The main barriers to utilizing this connectivity are the limitations and inadequacies in national and regional backbone networks on the supply side; and the challenges related to poor campus networks as well as the very limited individual access to computers on the demand side. The region has seen an increase in the amount of terrestrial backbone coverage. By June 2020, the size of the operational fiber-optic network had reached some 1,072,649 km compared to 622,930 km in 2015. By the same date, there was a further 119,496 km under construction, 95,057 km of planned, and at least 69,702 km of proposed fiber2. However, there are a series of challenges in cross- border connectivity. These range from different legal and commercial conditions to disparities in the quality of terrestrial fiber connections, ongoing vandalism, and fiber cuts during other construction works—especially for roads. National fiber coverage in Africa varies widely, again influenced by geography and level of competition and investment by public and private sector operators. A close examination of the fiber-optics map and population density indicates that Angola, Algeria, Cameroon, Egypt, Gambia, Ghana, Kenya, Mozambique, Senegal, Tanzania, Uganda, Zambia, and Zimbabwe have better networks that align with the populations’ geographic settlements and that could support their higher education connectivity needs. Yet, over half of African countries still need substantial investment in their terrestrial fiber backbone networks. Still, the presence of an extensive public backbone of sufficient capacity does not necessarily lead to adequate broadband connectivity to universities (Rwanda, Tanzania, and Ethiopia are examples). Bandwidth availability should be accompanied by a policy and regulatory environment that would stimulate competitive broadband ecosystems along with the growth of Research and Education Networks (RENs). 2 Hamilton Research. 2 April 2021. Africa Bandwidth Map. http://www.africabandwidthmaps.com/?p=6440. IX Role of Research and Education Model 1, exclusive connectivity via commercial service Networks as Enablers for Higher providers, is the common entry option for African countries without NRENs or those with emerging NRENs. Education Broadband Connectivity This model allows only for connecting to the internet and will not facilitate access to scientific collaboration Regional research and education networks (RRENs) available through research and education networks and national research and education networks (NRENs) or access to the tailored dimensioning available with have been coordinating the broadband connectivity REN connectivity. When countries establish NRENs, for HEIs in Africa. Three major RRENs cover the African institutions will start connecting via these, and those HEIs continent—the Arab States Research and Education not yet connected to an NREN will still connect through Network (ASREN), which covers North Africa but whose CSPs (Model 2). The third model works in countries where core members are outside Africa; the West and Central NRENs are mature and able to provide internet access African Research and Education Network (WACREN); and and global research and education connectivity. This is the UbuntuNet Alliance (UA). the model recommended for connecting African HEIs to broadband. In Model 4, institutions connect to CSPs for Some 40 of the 54 African countries are currently general internet access and to NRENs for REN traffic. associated with these RRENs, but these countries’ Model 4 works in very mature and highly competitive capacity to make the best use of regional aggregation telecommunication markets with fully functioning varies widely due to different readiness levels. RRENs internet exchange point(s) (IXP), data centers and data aggregate traffic from over 20 countries across Africa caches for popular content distribution networks, and and interconnect with the pan-European research and where HEIs are research intensive and well resourced. education network, GÉANT, to reach Europe and RRENs in other parts of the world. The different models and the widely varying states of the enabling environment and connectivity in African There are also variations in the development of NRENs countries imply that there can be no one-size-fits-all in Africa. NREN readiness is enabled when sufficient approach to connecting HEIs. Each country needs to government commitment is secured, and an organization assemble a high-level team drawn from all relevant that is recognized and supported by both public and stakeholders, including the ministries of higher education private HEIs is created. The organization needs to be and the information and communications technology adequately staffed to handle both administrative and (ICT) sector; sector regulators; HEIs; NRENs, where technical matters and to have the capacity to negotiate present; the ICT private sector; key development partners; connectivity deals on behalf of its members. Countries and other stakeholders to map out these gaps and develop that are not actively associated with one of the three a unified plan for connecting HEIs, including approaches to regional RENs face an even more significant challenge in developing and/or strengthening the national RENs. getting cheap and high-capacity broadband connectivity to universities. The reality is that only about twenty African countries currently have NRENs that deliver connectivity to HEIs, and of these, fewer than five (allowing for some progress since data was collected for Cost of Connecting HEIs to Broadband this study) can be considered as mature NRENs, points to the urgency of addressing this institutional gap in the The connectivity of higher education has the following African development ecosystem. four major components, all of which need to be addressed to complete the value chain: end-user access devices, high-quality campus networks to deliver a good broadband experience to the end-users, high-quality national networks to interconnect campuses, and regional Models for Connecting Higher Education and global networks to join national networks to the to Broadband global environment. In general, there are four models for connecting higher education, as follows: i. Model 1: Connecting exclusively via Commercial Services Providers (CSPs), ii. Model 2: Connecting via either CSPs or NRENs, iii. Model 3: Connecting exclusively via NRENs, and iv. Model 4: A hybrid model, connecting to both CSPs for general internet access and to the NREN for REN traffic. X Access to Devices need some USD 27.3 billion to upgrade campus networks and maintain them over the next five years. Access to devices is a critical enabler for higher education connectivity. Students need laptops to access learning It should be noted that the gross estimate of USD 27.3 materials around the clock and from any location; billion for campus networks is based on the broad staff need devices to conduct research, teaching, and categorizations of campuses as small (< 5,000 students), collaboration with their peers around the globe; and medium (> 5,000 and < 15,000 students), and large (> management and administrative staff need laptops to 15,000). Small campuses account for about 94% of all support the overall learning environment. The benefits HEIs and about 83% of the total cost of upgrading campus of connectivity can only be maximized when faculty and networks. Where more detailed data on campus sizes students have one-to-one access to computing devices as is available, the small category can be refined further shared access is difficult to manage and sustain. Laptops, into micro, mini, small, medium, large, and very large rather than tablets or other devices, are recommended campuses. The modeling of the case study countries at this as they possess both the functionality and attributes level (as provided for in the Cost Model) led to reductions to support teaching, learning, and research. Since in the cost of campus upgrading by 37%, 6.7%, and 32.2%, affordability is still a challenge for many students, the respectively, for Côte d’Ivoire, Mozambique and Uganda. recommended entry model is a combination of bring your This potential reduction has not been factored into the own device (BYOD) and subsidies/loan schemes, with gross estimated cost in the summary because the number owner contribution emphasized in the latter model. of countries analyzed is too small to be used as a basis for a reliable generalization across the continent—but it does The recommended approach is a phasing in by providing point to a significant potential reduction in the gross cost devices for only first-year students each successive year by taking a more detailed approach to campus size. and scaling this down year-on-year as institutions and countries take up the financing. Since most courses in HEIs have a duration of two years (TVETs) to three years (universities), this would ensure that all students have Connecting at the National and Regional laptops within three years. All staff would be equipped with fully subsidized devices over the same period. Levels Estimates based on the available data on the number of The cost of interconnecting campus networks at the higher education students and staff indicate that African national level and further regionally and globally will countries would need some USD 17.3 billion to roll out one- depend on many factors, including NREN maturity, the to-one computing devices between 2021 and 2025. level of broadband competition and market structure, and economic and geographic aspects. Data for various indicators were collected from various databases as part of the Gap Analysis phase. Duncan Greaves’ NREN Capability Upgrading Campus Networks Maturity Model3 and Michael Foley’s levels of NREN development4 provide possible mechanisms to gauge the Campus networks are crucial because all student and level of NREN maturity within a given country. NRENs staff devices must connect through a local wireless or in different African countries can be broadly categorized wired network to access the internet and other academic into “no NREN” (no NREN but varying levels of awareness and research resources. Campus networks have been about need and ongoing conversations); “emerging NREN” found often to be the main bottleneck in the connectivity (legal entity established but without or with a physical chain due to poor design. The design of campus networks network of varying coverage); “connected NREN” (physical is subject to multiple factors, which include physical network with regional/global connectivity to other characteristics such as the number of buildings, the NRENs and offering middle-ware services); and “mature distance between buildings, the skills of the engineers, NREN” (physical network with high-speed regional/ and the number of end-users and network devices. global connectivity to other NRENs and offering advanced Campus networks must be designed to meet the security, services). connectivity, and performance challenges while enabling the delivery of all critical IT applications and services. They must scale as needed and offer operational simplicity and flexibly to accommodate new computing trends. Emphasis should be placed on Wi-Fi access to create the flexibility 3 D. Greaves, An NREN Capability Maturity Model required in modern learning environments. The capital and (2009), https://repository.ubuntunet.net/ operating costs of campus digital infrastructure depend on bitstream/handle/10.20374/69/NREN_Capability_ campus size, the number of users, the quality of the pre- Maturity_Model.pdf?sequence=1&isAllowed=y. existing network, and the skillsets to design and upgrade 4 M. Foley, The Role and Status of National Research and the network. Estimates based on the available data on the Education Networks in Africa (World Bank, 2016). number and size of HEIs indicate that African countries XI ICT indicators at the country level, with a direct bearing As the cost of bandwidth falls, it will be especially on connectivity, include landlocked-ness; number of cable important to provide for such support in order to reap the landing stations; internet exchange ladder stage (reflects resulting value of the NREN to the delivery of high-speed number of IXPs and carrier-neutral data centers and their connectivity in any given country. Countries will need an interaction);5 percent population within 10 km of fiber estimated USD 513 million to support NREN core costs coverage (reflects fiber network coverage of the country); over a five-year period. Support will be provided inversely and a regulatory score, which reflects the maturity of the proportional to the level of NREN maturity. That is, mature regulatory environment (based on country scores from ITU NREN countries receive the least support, while no NREN Global Regulatory Outlook 2020). Estimates to connect countries receive the most support. campuses upstream based on these factors indicate an aggregate cost of some USD 7.3 billion for upstream Regionally, there is a need for strengthening the WACREN, connection over the next five years. UA, and ASREN, which provide regional connectivity as well as upward connectivity to other regional networks, Through negotiated procurement procedures, which such as GÉANT, RedCLARA, APAN and Internet2. RRENs benchmark the most competitive regional prices, play a critical role in NREN development and training this could be lowered even further by securing, where in internetworking technologies, such as routing and appropriate, long-term leases. The cost of USD 7.3 billion campus network development. The three RRENs require is based on final-deliver-to-campus prices, which factor in USD 25 million to accelerate NREN development and NREN and RREN costs and overhead as well as commercial capacity building. service provider costs, overheads, and profits. Estimated cost of connecting all African Supporting the Development and HEIs to high speed internet Sustainability of NRENs and RRENs The estimated total cost of connecting African HEIs Nationally, a fully functional NREN can help aggregate over the next five years (2021–2025) is thus USD 52 traffic at the national level and in turn connect to regional billion in total. This includes the expense of providing RENs, such as the UA, WACREN, and ASREN and further to devices to students and staff (USD 17.3 billion), the GÉANT (Europe), RedCLARA (South America), APAN (Asia- expense of upgrading campus networks (USD 27.3 billion), Pacific), or Internet2 (North America), so that there is full bandwidth cost for upstream connectivity (USD 7.3 billion), integration in the global research and education fabric. and NREN and RREN development/support (USD 538 NREN member institutions typically cover connectivity- million). related expenses through payments for bandwidth, but NRENs often struggle to cover core costs as well as costs All total costs assume regional demand aggregation, related to ongoing capacity building for both internal staff smart procurement strategies (e.g., benchmarking and especially member institutions where the value of regional pricing), and, where applicable, the procurement connectivity is realized. of long-term leases for bandwidth to secure the best price advantages. Potential sources of funding include A shortage of funding also means NRENs fail to retain governments, development partners, students, and HEIs. competent staff, who are attracted by the much higher Actual proportions will vary across countries depending pay within the ICT private sector; this is especially a on government funding priorities and policies as well as challenge in the development and growth stage of five to development partner funding guidelines. 10 years. 5 World Bank Group, National Data Infrastructure: The Role of Internet Exchange Points, Content Delivery Networks, and Data Centers (draft WDR21 background paper, 2021). XII Cost of Di erent Components to Connect all African Higher Education Institutions to Potential Sources High-Speed Internet (USD, billions) of Funding 0.03 RREN core support and RREN Development partners, development related costs government 0.51 NREN core support and Development partners, NREN development related costs government 7.28 Cost of connecting campuses upstream Development partners, (Student Enrolment & Regional Price) institutions, students OpEx for maintaining Institutions, government, 15.51 campus networks development partners CapEx for upgrading Government, development 11.75 campus networks partners Students and Sta end-user Government, development 18.77 devices partners, students, institutions Figure 1: Summary of total cost (USD, billions) of connecting all African higher education institutions to high-speed internet Source: KCL calculations Leveraging Broadband Connectivity to Enhance Learning Outcomes Connectivity is crucial for learning, research collaboration, and access to scientific resources, yet the real benefit only occurs if it is accompanied by other efforts that expand access, equity, learning outcomes, and employability in the emerging digital economy. While curriculum, pedagogy, and assessment are key elements of the solution, the focus here is on the key ICT-related impediments to the integration of technology into learning and research and recommending a roadmap for addressing them. The integration of technology in higher education to enable better learning outcomes and academic excellence, foster research and innovation, and achieve greater operational efficiency needs to be accompanied by a digital technology integration vision, policies and strategies, a change in processes, and a shift of the mindsets of people—such as students, administrators, and faculty. A well-staffed corporate ICT department, with highly skilled engineers and a user support team, is as critical as is the presence of technology-savvy teachers and administrators that facilitate students’ success in digital technology-enriched learning environments. There are several well-known impediments to the integration of ICT in support of learning, research, and effective administration in HEIs. These include the following: i. The absence of, or deficient, ICT policies and strategies, which often arises from lack of awareness of the role of higher education connectivity ii. Limited ICT awareness and ICT literacy among faculty and administrators iii. Limited competence of campus ICT personnel iv. The poor quality of campus networks v. Poor digital learning spaces vi. Limitations in resource allocation and coordination vii. Limited individual access viii. Limited digital learning resources XIII All these have to be addressed if the opportunities Each country needs to assemble a high–level team drawn of broadband connectivity are to be exploited for from the ministries responsible for higher education, improved learning and research outcomes. Change at the ICT sector, and finance; HEIs; NREN, where present; the infrastructure/technology level is relatively easy, but the ICT private sector; key development partners; and the changes required will not be sustainable unless they other stakeholders in order to develop a national plan for are accompanied by changes in people and processes. connecting its HEIs and accelerate learning and digital Intervention should also be carried out at the regional, skills for the jobs of the future. national, and institutional levels. The following four recommended strategic interventions provide the Governments and development partners need to framework under which detailed activities would be work together to push the connectivity, learning, and undertaken: digital-skills agenda forward, which demands engaging stakeholders, cost sharing at all levels, and adhering i. Establishing and sustaining regional partnerships to a timeline for connecting HEIs in Africa. The five- with other agencies and organizations that are year timeline proposed is ambitious, and an aggressive interested and active in digital technologies for approach is required to achieve the objectives. improving research output, learning outcomes, and employability in Africa; ii. Identifying leadership and catalyzing the formation of national-level coalitions that will be responsible for spearheading change in the integration of technology in higher education; iii. Guiding national- and institutional-level approaches through toolkits; and iv. Supporting specific quick-win demonstration projects. The following key considerations are important when looking ahead at leveraging broadband connectivity to enhance learning outcomes : i. First, there is a need for increasing awareness among decision makers about the different connectivity issues and challenges. This is especially important for those in the ministries of education, ministries responsible for digital technologies, ministries responsible for finance and investment, and HEI leaders. An awareness of the magnitude of the challenges (e.g., the need for access to computing devices, the importance of the upgrading of campus networks, NREN development) is critical to ensure that adequate resources are available for higher education connectivity. ii. Second, there is a need for accelerating connectivity to facilitate teaching, research, learning, and administration in higher education in order to improve learning outcomes. iii. Third, connectivity should be accompanied by the digitalization of campuses, building ICT literacy among staff and students and enabling the individual ownership of laptops to support reforms in learning and new ways of teaching digital and soft skills to meet the demands of 21st-century jobs. iv. Fourth, coordination is an essential prerequisite for achieving higher education connectivity and attaining reforms in learning and digital skills. XIV Action plan for connecting all African HEIs to high speed internet A five-year action plan is proposed below for connecting all African HEIs to high speed internet. It is ambitious, and requires an aggressive approach to achieve the objectives. Table 1: Prioritizing connectivity to African higher education Timeline/ Year 1 Year 2 Year 3 Year 4 Year 5 priorities/ Individual Individual Individual Individual Individual Devices ownership for all ownership for all ownership for all ownership for all ownership for all first-year students first-year students first-year student first-year students students and staff Build functioning campus networks across HEIs Nationwide support for evolution of a Campus network world-class campus network All institutions progress to a goal of 2 Gbps per 1,000 Start work toward a goal of 20 Gbps Connectivity students per 1,000 students Aggressive NREN development Intensive support to upgrade countries to the NREN support to all countries; special focus Exclusive model NREN development on countries without sustainable NRENs RREN Each African country is connected though at least one regional REN, and the RRENs are interconnections interconnected at multiple points throughout the continent Developing pre-conditions Enabling policy environments related to the use of ICT in teaching, learning, and research; for leveraging curriculum; pedagogy; assessment; quality management; and industry linkages developed and connectivity for implemented in HEIs along with the necessary institutional arrangements improved learning outcomes Source: KCL. XV The implementation of the above action plan requires coordinated actions of all stakeholders, in particular: Government: Development Leaders of HEIs: Private sector: • Support the Partners: • Support the • Recognize development of • Promote development of the specific comprehensive information NRENs connectivity higher education ICT exchange among • Allocate resources requirements of policies that address countries on the to ensure sufficient HEIs and engage device, connectivity, different issues of bandwidth is in dialogue campus networking, connectivity available for higher with NRENs and capacity issues • Support projects education to optimize • Support the that enhance the • Sponsor projects service offerings, development connectivity of that leverage including through of research and higher education connectivity service level education networks and those that to enhance agreements • Promote access to leverage advanced teaching, learning, • Consider offering devices through networks to and research preferential rates national negotiation solve social collaboration or tailored service with suppliers and and economic • Serve as advocates offerings for HEIs, industry players challenges for higher education in cooperation • Allocate resources connectivity with REN. for higher education connectivity through donor funding or universal access funds XVI XVII 1 Introduction Overview Connecting Africa’s HEIs to Affordable High-Speed Internet Higher Education Institutions (HEIs)—comprising public and private universities, colleges, technical training Higher education is instrumental in fostering long-term institutes, and vocational schools—play crucial roles growth and boosting shared prosperity in all countries, in training skilled workforces, conducting research, whether high-, middle-, or low-income.8 At the individual and fostering innovation, all of which underpin level, it provides unique opportunities for advanced social development, economic growth, and national and enhanced learning to nurture skills for immediate competitiveness. However, most HEIs in Africa lack professional application. Economic returns for a tertiary access to affordable and functional quality broadband graduate are the highest in the entire educational connectivity. In addition, the available bandwidths are system – with higher employability and higher wages. expensive and limited in capacity and are thus unable to At a societal level, tertiary education provides a highly meet the research and education requirements of modern skilled workforce which is a prerequisite for a country’s institutions. Ultimately, this negatively affects national innovation and long-term growth. Furthermore, as education goals and targets as detailed in the Sustainable university-industry cooperation is becoming more and Development Goals (SDGs) and various country-specific more important, HEIs are considered the “backbone of development plans. a country’s innovation ecosystem.” A research also finds that “Tertiary schooling can also have less direct benefits The COVID-19 pandemic further highlighted the gaps for economies. By producing well-trained teachers, in the connectivity of higher education throughout it enhances the quality of primary and secondary the continent. Countries began to transition to digital education systems and gives secondary graduates technologies to address the disruption in education, greater opportunities for economic advancement. By choosing various options based on their constraints in training physicians and other health workers, it improves infrastructure and connectivity. Overall, the experience a society’s health, raising productivity at work. And by points to the acute need for investment in digital nurturing governance and leadership skills, it provides technologies and connectivity, along with reforms in the countries with the talented individuals needed to establish education systems, to accelerate digital skills and achieve a policy environment favorable to growth9.” better learning outcomes. As part of the Digital Economy for Africa (DE4A6) The university must become a initiative, the World Bank commissioned a study to primary tool for Africa’s development develop an operational roadmap to connect all African in the new century. Universities can HEIs to high-speed Internet. The initiative, in support help develop African expertise; they of the African Union Digital Transformation Strategy can enhance the analysis of African for Africa (2020-2030)7, aims to digitally enable every African individual, business, and government by 2030. problems; strengthen domestic Connecting universities and research institutions is crucial institutions; serve as a model for expanding the opportunities for teaching, learning and environment for the practice of good innovation to foster relevant digital skills on the continent. governance, conflict resolution, and respect for human rights, and enable African academics to play an active part in the global community of scholars. — Kofi Annan 8 WB Tertiary Education Position Paper: STEERing Tertiary Education – Toward Resilient Systems 6 See https://www.worldbank.org/en/programs/ that Deliver for All (forthcoming) all-africa-digital-transformation. 9 Bloom, D. E., D. Canning, K. Chan, and D. L. Luca. 7 https://au.int/en/documents/20200518/digital- 2014. “Higher Education and Economic Growth in Africa.” transformation-strategy-africa-2020-2030. International Journal of African Higher Education 1 (1): 23–57 1 With the growing significance of tertiary education in a Access to large databases, the sharing of computational country’s economic growth, technology adaptation or resources, shared access to data analysis and visualization digitalization of the sector is the present in accordance techniques further enhances the quality of research, with the general trend of the digital transformation of enabling the participation of African HEIs into the the economy. COVID-19 has also reaffirmed that distance global sciences, technology, and innovation space. The delivery for tertiary education should become a norm technology advancement and connectivity also allow to sustain the core tertiary education functions during tertiary education institutions to become a hub to the emergency or disaster situations. In this context, equip students and staff with important digital skills – broadband connectivity is a critical agenda for the African from foundational, intermediate, to advanced levels, HEIs to ensure that they also keep up with the pace and accelerating the economy-wide digital transformation. depth of transformation in the rest of the world. According to Sajitha (2020)10, the availability of bandwidth in HEIs determines: Role of Research and Education • Whether efforts to improve internet access through Networks supplying devices (laptops, tablets) are successful; • The extent to which university faculty and students While high-speed internet is prerequisite to reach the can goals of continued learning, collaborative research, and - Access international knowledge (journals, human capacity development in the tertiary education papers, databases, courses, presentations on sector, research and education networks (RENs) has been YouTube); a successful model in providing the necessary digital connectivity to HEIs. In particular, national research and education networks (NRENs) are considered a critical - Collaborate with fellow academics worldwide element in delivering sufficient and reasonably priced (including uploading papers and large data high-speed connectivity to HEIs in that they: sets) in research and teaching programs; i. Mediate between HEIs and the market, getting the - Access expensive instrumentation, such as best price offering for this closed user group through supercomputers, virtual labs and so on; economies of scale, and customer aggregation benefits for the commercial suppliers. • What online content and applications can be used for ii. Provide direct connectivity with international teaching and learning; whether online and blended research and education networks to foster research learning approaches can be introduced; collaboration and scientific resource sharing. • Whether personalized and differentiated learning iii. Offer access to digital libraries, learning management (such as adaptive learning) can be used for students systems, and scientific resources; capacity building; and also for faculty professional development; identity management; and eduroam. • Whether “cloud” services can be used for iv. Ensure high-capacity bandwidth with undisrupted administrative and teaching purposes; connection to allow spikes in usage through dedicated • The ease of updating subscriptions, managing bandwidth to each client. Connection to e-science apps, maintaining content management, learning resources such as telescopes, sensor networks, management and student information systems, as accelerators, and supercomputers requires sustained well as protection of the network by providing system high volume and quality bandwidth for short periods updates and addressing vulnerabilities. of time to transfer large research data or enable access to instruments. The implication of increased broadband connectivity on v. Create platforms for experimentation with the various learning and research outcomes is significant. Greater use aspects of network technologies, such as protocols of technology and digital resources enhances the learning and security, which have spill-over effects in other experience of students (improvements in pedagogy, networks. assessment, access to open education resources, online vi. Facilitate the formation of communities of researchers courses, etc.) and the professional development of faculty. in the areas of agriculture, bioinformatics, disaster mitigation, network development, and telemedicine, among others, among researchers in the developing world. Africa’s sustainable development challenges, such as increasing urbanization, climate change- induced crises, environmental degradation, food insecurity, and a growing load of non-communicable 10 https://openknowledge.worldbank.org/bitstream/ disease, demand extensive research and collaboration, handle/10986/34955/Connecting-Africa-s-Universities- requiring advanced research and education networks. to-Affordable-High-Speed-Broadband-Internet- What-Will-it-Take.pdf?sequence=1&isAllowed=y 2 vii. Although in some cases, commercial connectivity Chapter 4 presents a business model for connecting all pricing appears to be lower than NREN connectivity HEIs. It begins with the vision and targets for linking pricing, the above advanced qualities that NRENs HEIs to the global networks. The chapter then presents offer verify that any price comparison is fallacious. different models of access to laptops for students and Moreover, there has been “evidence that the staff. Next, the chapter reviews the necessary upgrades availability of cost-effective and cutting edge to campus computing infrastructure. This is followed by a NREN network services enables and encourages discussion of upstream connectivity options for campuses technological spillover into commercial sector, which based on an aggregation model, after which it explores ultimately benefits society as a whole11.” the necessary support to strengthen NRENS and regional RENs, which are important elements of aggregating demand across higher education. The last section brings Approach all of these components together to present the estimated cost of connecting African HEIs. There have been efforts over the last decade to assess the main challenges and opportunities of connectivity for the Chapter 5 presents a summary of the country case studies African higher education system, especially within the of Burkina Faso, Cote d’Ivoire, Mozambique, and Uganda context of the development of NRENs. Yet, there is limited to illustrate the implementation of the business model in understanding of the connectivity value chain among HEIs different contexts. and at national levels. To fill the knowledge gaps and guide action, the feasibility study was carried out through the Chapter 6 proposes a high-level strategy and roadmap for following three sub-activities, each of which is covered by leveraging connectivity to improve learning outcomes, a comprehensive report: with a specific focus on what needs to be done to ensure the full digitalization of campuses to support i. The production of a gap-analysis report addressing effective learning skills, including addressing the related connectivity, ecosystem challenges (policy, regulation, technology, processes, and people challenges. institutions, human capacity, etc.), and funding. ii. The development of a cost model. This also referenced Chapter 7 concludes the report, providing a roadmap for in-depth country case study reports for Burkina Faso, addressing the gaps and identifying potential partners to Côte d’Ivoire, Mozambique, and Uganda. the World Bank Group (WBG) Initiative for connecting all iii. The development of a high-level strategy to leverage African HEIs to broadband. campus connectivity to achieve learning outcomes in higher education. This report presents a summary of the feasibility study and establishes a roadmap for connecting all African HEIs to high-speed internet. Structure of the Report This report is organized as follows: after the Introduction in Chapter 1, Chapter 2 discusses the methodology used for gathering the data to conduct a baseline assessment for the gap analysis and to establish the business model for connecting HEIs to high-speed internet. Chapter 3 begins with setting the context by defining the universe of the target institutions. Next, the chapter sets minimum connectivity targets for the target HEIs and presents an analysis of connectivity gaps at the regional, national, and campus levels. It concludes with lessons learned from existing projects on closing the connectivity gap. 11 https://www.casefornrens.org/Resources_and_Tools/ Document_Library/Documents/Case%20for%20NRENs.pdf. 3 2 Methodological Summary T he Feasibility Study draws on extensive document analysis, and a wide range of data sources both within and outside the public domain. Surveys and interviews with key informants within and outside Africa were conducted. The study draws on extensive analysis, both qualitative and quantitative, to inform the findings and recommendations. A survey of research and education networks, operators, and key informants was conducted to establish the current connectivity environment, gaps, and challenges in higher education. Data on enrollment rates were collected from the ministries of education and cross-checked with figures available from the UNESCO Institute of Statistics (UIS). In- depth country case studies were also conducted in Burkina Faso, Cote d’Ivoire, Mozambique, and Uganda to identify connectivity gaps and challenges specific to different contexts. Table 2 provides a methodological summary of our analysis. Table 2: Methodological Summary Output/data/ Sources of data and information Analytical methods information required i. Desk study (e.g., GÉANT and TEIN compendia for 2018) • Benchmarking ii. Specific current data from NREN • Comparative examination and Target state of CEOs in Europe, USA, and South projections taking into account the 1 connectivity and utility America different views and perspectives as well (Vision) iii. Data from RREN and NREN CEOs as historical trends and technology on definition of broadband as well projections as current and projected states i. Desk study; data from multiple sources (Telegeography, ITU, Current state of Tabulation of data from multiple sources, GSMA, etc.) connectivity (user, basic statistical analysis, and data 2 ii. Questionnaires to NRENs campus, national, presentation through combination of tables iii. Interviews with NREN CEOs, large regional, global) and graphs connectivity providers, and key informants Data from UIS, complemented with recent data from ministries Estimation based on historical data and Current enrollment in 3 of education/higher education and population growth for a few countries with TVETs and universities accreditation and quality assurance data gaps bodies where available 4 • Forecasting student enrollment at a country level was based on historical figures from 2001 to date where available; model uses triple exponential smoothing forecasting technique to give more weight to recent data (less weight Data from UIS, complemented to older data) and take into account Projected student with recent data from ministries seasonality and trends in the data 4 enrollment between of education/higher education and • The average gross enrollment ratio 2021 and 2030 accreditation and quality assurance (GER) at the regional level was used to bodies where available estimate enrollment for some countries that lacked historical student enrollment data but have population projections for the 20–24 age category; GER was derived from forecast of student enrollment and population in age category 20–24 at a country level • Forecasting student enrollment at a country level was based on historical figures from 2001 to date where available; model uses triple exponential Data from UIS, complemented smoothing forecasting technique to give with recent data from ministries more weight to recent data (less weight Projected staff numbers 5 of education/higher education and to older data) and take into account between 2021 and 2030 accreditation and quality assurance seasonality and trends in the data bodies where available • Determined regional average for student-staff ratio; this was used to estimate number of staff for some countries that lacked data about the number of higher education staff In Scenario 1, forecast student enrollment was multiplied by the progressive targets in Table 10 to estimate bandwidth requirements at the country-level. Quantifying bandwidth Desk study; data from multiple sources 6 In Scenario 2, minimum connection port gaps (Telegeography, ITU and UIS) sizes for each campus size (small, medium & large) were used to multiply by the number of institutions in each campus size category to estimate bandwidth requirements • For the Local Price, the unit cost of bandwidth (per Mbps) was derived by adding the cost of IP transit for the cheapest provider and the cost of local access to deliver the bandwidth in a metro area for the cheapest provider in the country Determining the unit Desk study; data from Telegeography • For the regional price, the unit cost 7 cost price of bandwidth and other sources of bandwidth (per Mbps) was derived by adding the cost of IP transit for the cheapest provider and the cost of local access to deliver the bandwidth in a metro area for the cheapest provider in the region where the country is located • Calculated regional12 averages to cater to countries that lack data 12 https://www.worldbank.org/en/region/afr. 5 Derived aggregation savings from NREN Identifying savings i. Desk study; data from multiple maturity at the country level as well as that can accrue from sources (Telegeography, ITU, 8 the country’s performance based on aggregating bandwidth GSMA, etc.) various national ICT indicators pertinent to demand ii. Questionnaires to NRENs connectivity i. Desk study ii. Interviews with selected RREN and NREN CEOs as well as Identifying connectivity multinational connectivity Qualitative analysis to identify key issues and 9 gaps and challenges providers and key sector common threads informants iii. Questionnaires to NRENs iv. Deep-dive country cases i. Desk study • Qualitative and quantitative analysis to ii. Interviews with selected in-country identify specific country issues 10 Deep-dive country case stakeholders • Testing the cost estimates using country- iii. Data collection guide for in-country specific parameters team to facilitate data collection 6 3 Context and Connectivity Gaps in Higher Education T he analysis summarized in this chapter looked at iii. Institutions that are responsible for policy, regulation, both the demand and supply sides. The demand standards, and sector management—making side comprises the universe of target institutions, them critical players that need to be considered for down to the users; and the supply side includes broadband connectivity. the entire bi-directional connectivity value chain, starting from the end-user and extending to the global research While categories (ii) and (iii) have been included as part of and education community (and vice versa). The gaps are the universe of target institutions in so far as connectivity the situational factors that impede or block the delivery of is concerned, the relative numbers in each country value along the supply chain. compared to the main body of HEIs is very small and has no significant impact on bandwidth needs. They also have Development partners are enablers, but they do not their independent budgets to fund networks within the always act in harmony in order to create the synergy of buildings and computers and are therefore not included in interventions, and this also needs to be recognized as a the cost estimates for these elements. gap. The chapter therefore concludes with a review of previous and current development partner interventions Table 3 shows the student population based on the UIS to address connectivity gaps in HEIs. 2020 report (derived from data collected during the COVID-19 school lockdowns).15 Coupled with 500,000 The Universe of Target Institutions and staff, the higher education population totaling close to 16 million students, faculty, and researchers represented Student Numbers about 1.3% of the African population in 2019. There is no universal agreement on what constitutes higher education across Africa. Higher education in this document covers all post-secondary education, including public and private universities, colleges, technical training institutes, and vocational schools. UNESCO’s International Standard Classification of Education (ISCED)13 maps national education systems in a way that facilitates a comparison of programs across countries.14 All institutions from UNESCO–ISCED Level 5 and above are considered part of higher education and taken into account for this report. The private HEIs, which are growing fast and catering to about a third of higher education students and staff, are also included in the analysis and projections. The universe of target institutions in this study comprises the following: i. HEIs; ii. Other institutions that are closely allied with HEIs and play a critical direct or collaborative role in promoting research and education as well as their direct benefit to national development outcomes. These include research centers, the training hospitals used by the schools of medicine, and establishments such as libraries, whose resources support TVET and/or university-level training and research; and, 13 http://uis.unesco.org/en/topic/international- standard-classification-education-isced. 14 http://uis.unesco.org/en/isced-mappings. 15 https://en.unesco.org/covid19/educationresponse. 7 Table 3: Higher education student population in Africa (rounded to nearest 100) Country HEI students Country HEI students Country HEI students Algeria 1,600,700 Eswatini 1,798,100 Namibia 56,000 Angola 253,300 Ethiopia 757,200 Niger 80,100 Benin 126,200 Gabon 10,100 Nigeria 1,513,400 Botswana 49,400 Gambia 5,000 Rwanda 80,800 Burkina Faso 117,800 Ghana 443,700 São Tomé and Príncipe 2,300 Burundi 61,700 Guinea 118,000 Senegal 184,900 Cabo Verde 11,700 Guinea-Bissau 3,700 Seychelles 1,300 Cameroon 290,300 Kenya 562,500 Sierra Leone 9,000 Central African 12,600 Lesotho 22,600 Somalia 196,800 Republic Chad 42,500 Liberia 43,900 South Africa 1,116,000 Comoros 6,500 Libya 375,000 South Sudan 11,300 Congo 54,800 Madagascar 143,800 Sudan 204,100 Congo, Dem. Rep. 464,700 Malawi 12,200 Tanzania 178,600 Côte d'Ivoire 217,900 Mali 72,600 Togo 101,900 Djibouti 4,700 Mauritania 19,400 Tunisia 282,200 Egypt 2,914,500 Mauritius 38,900 Uganda 258,500 Equatorial Guinea 1,000 Morocco 1,056,300 Zambia 56,700 Eritrea 10,200 Mozambique 213,900 Zimbabwe 135,600 Source: UNESCO Institute for Statistics, 2020. Connectivity Gaps in Higher Education The following major dimensions of the connectivity gaps are discussed in this section: i. Availability and sufficiency of global connectivity ii. Availability and sufficiency of regional connectivity iii. Existence and effectiveness of regional RENs iv. Availability and sufficiency of national connectivity v. Existence and effectiveness of NRENs vi. Existence and sufficiency of campus networksvxii. vii. Availability and sufficiency of last-inch connectivity: individual access level Availability and Sufficiency of Global Connectivity The sufficiency and competitiveness of global connectivity influence the lowest price that users can get. Outside exceptional circumstances created by sound policy and regulation, effective competition requires that there are at least three providers and that the total available capacity is much higher than what is needed by the market to avoid scarcity effects on pricing. Africa has seen substantial growth in international connectivity in recent years with the landing and upgrading of eight submarine cables (ACE, WACS, Main One, GLO-1, AST3, NCSIS, SAIL, and SACS), constituting 127 Tbps in the west coast, and five cables (EASSy, SEACOM, LION, TEAMS, and SEAS) that brought 25 Tbps capacity to the east coast of the continent. Figure 2 shows the different undersea cables that currently serve Africa (gray indicates under construction and planned). The availability of landing stations in all coastal countries (except Eritrea) has spurred fiber-optic links between undersea cable landing stations and the capital cities and national fiber-optic backbones connecting major towns. Non-coastal countries are also able to connect to submarine cables albeit at often much higher cost. 8 Figure 3: Global inter-regional internet bandwidth Source: Telegeography, 2019 The absence or presence and number of cable landing stations in a country introduces the first global connectivity divide. Among the 54 African countries recognized by the United Nations, 38 have access to the sea, while 16 are land-locked. Of the 38 countries with access to the sea, 37 had at least one submarine cable landing by the end of 2019, Eritrea being the only exception. Eleven countries have one cable landing, 10 countries have two cable landings, six have three cable landings, and 10 have more than three cable landings. The study also established the following: i. Although the cost of bandwidth has come down, Figure 2: Map of active undersea cables around Africa internet access in Africa is still much more expensive (proposed indicated in gray) compared to other regions of the world and often less Source: Submarine Cable Map, TeleGeography reliable, especially inland. https://www.submarinecablemap.com/ ii. Africa still largely consumes internet content Figure 3 shows the global inter-regional bandwidth. While, produced and/or hosted in other parts of the world, on the face of it, Oceania (with 5,563 Gbps) has an inter- which requires expensive international transit. Big regional bandwidth that is less than Africa (with 12,240 content and cloud service providers have started to Gbps), the population of the two is currently estimated as move to Africa, and some, such as Facebook, have 42.1 million and 1.31 billion, respectively,16 resulting in 132 started to invest in infrastructure, but that is yet to Mbps per 1,000 people in Oceania compared to a meagre have significant continental impact. 9 Mbps per 1,000 people in Africa. On a comparative global basis, the African continent is clearly still severely iii. The regional/local exchange of internet traffic due underserved in terms of available global capacity. to the still-limited penetration of IXPs (34 out of 54 African countries have at least one in-country IXP) means that many countries must rely on expensive transit capacity to Europe to exchange traffic that would have otherwise been exchanged on the continent. iv. In 2019, the five largest carriers operated 41% of all international connectivity to Africa compared to a world average of 29%. This highlights the high degree of market concentration that persists compared to other parts of the world.17 16 UN Data Population Estimates (2019). 17 Telegeography, Global Internet Analysis (2019). 9 Availability and Sufficiency of Regional Various factors contribute to the gaps in terrestrial connectivity across Africa, as follows: Connectivity The existence of competitive regional connectivity i. Despite multiple efforts at policy and regulatory is vital for all countries as it enables access to global harmonization,18,19,20 countries in Africa, even within connectivity for those countries that have no direct the same economic blocks, still have different ICT- access to submarine cables, and it facilitates the sector policy and regulatory environments and various development of infrastructure to exchange internet traffic financial and taxation policies with which licensed regionally, improving performance and saving expensive providers must be compliant. This makes it difficult international bandwidth. Specific to this study, it also for providers to build and operate regional (multi- allows regional research and education collaboration. country) infrastructure and to offer similar prices Africa has also seen an increase in the amount of across countries for the same service. terrestrial backbone coverage. By June 2020, the size of ii. Limited competition in backbone infrastructure the operational fiber-optic network had reached 1,072,649 (de facto monopoly, either private or public) in the km compared to 622,930 km in 2015. By the same date, countries through which a cable must transit, leading there was a further 119,496 km under construction, 95,057 to high transit costs. km of planned, and 69,702 km of proposed fiber. iii. Disparities in the quality of terrestrial fiber connections, ongoing vandalism, and fiber cuts during Figure 4 shows the terrestrial fiber (both operational and other construction works—especially roads. under construction) around Africa. A close examination As a result of the absence, and/or the high costs of shows that while some countries, such as Rwanda, terrestrial east-west and north-south cables, the routing Ethiopia, Zimbabwe, Nigeria, and Ghana, have a fairly of traffic from east to west or south to north has tended extensive fiber coverage at the national level, the number to rely on marine fiber, which, while much cheaper, also of regional links, especially east to west, is very limited. introduces high levels of latency. Regional connectivity is therefore one of the significant macro-level gaps that will need to be addressed if the objective of high-quality broadband to each institution is to be achieved. Existence and Effectiveness of RRENs The connectivity of HEIs to international submarine cables is in most cases coordinated through RRENs. With the possible exception of TENET in South Africa, which has been supported by the government for a long time through research funding, all other NRENs that have made substantial progress in Africa have benefited from working through the regional model.21 RRENs have made access to lower-cost international and regional bandwidth possible because they leverage demand aggregation at both national and regional levels. 18 African Information Society Initiative (AISI), https:// www.uneca.org/publications/african-information- society-initiative-aisi-decade%E2%80%99s-perspective. 19 Programme for Harmonisation of ICT Policies in Sub Saharan Africa (HIPSSA), supported by ITU and European Commission, https://www.itu.int/en/ITU-D/ Projects/ITU-EC-ACP/HIPSSA/Pages/default.aspx 20 Revised AU/NEPAD African Action Plan, https://nepad.org/. 21 The African Bandwidth Consortium (http://www. Figure 4: Terrestrial fibre within and around Africa foundation- partnership.org/pubs/pdf/more_bandwidth. Source: www.africabandwidthmaps.com/fibrereach/ pdf) that was funded by the Partnership for Higher Education in Africa (https://www.iie.org/en/Programs/ PHEA) was the first functional regional aggregation model. 10 Three major RRENs cover Africa: ASREN,22 which connects North Africa but whose core members are outside Africa; WACREN,23 and UA.24 Table 4 gives the NREN membership in each of these regional RENs and also shows countries in each region that do not have any NRENs. ASREN, WACREN, and UA aggregate traffic from over 20 countries across Africa, as shown in Figure 5, and interconnect with the pan-European GÉANT network to reach Europe as well as RRENs in other parts of the world. Table 4: NREN members of RRENs in Africa ASREN Members UbuntuNet Members WACREN Members Algeria: ARN Botswana: BotsREN Benin: RBER Comoros: Burundi: BERNET Burkina Faso: FasoREN Djibouti: Dem. Rep. of the Congo: Eb@le Cameroun: RIC Egypt: EUN & ENSTINET Ethiopia: EthERNet Chad: TchadREN Libya: LibREN Kenya: KENET Côte d'Ivoire: RITER Mauritania: Madagascar: iRENALA Gabon: GabonREN Morocco: MARWAN Malawi: MAREN Ghana: GARNET Somalia: SomaliREN Mozambique: MoRENet Guinea: Gn-REN Sudan: SudREN Namibia: Xnet Liberia: LRREN Tunisia: RNU & RNRT Rwanda: RwEdNet Mali: MaliREN Somalia: SomaliREN Niger: NigerREN South Africa: TENET Nigeria: NgREN Sudan: SudREN Senegal: SenRER Tanzania: TERNET Sierra Leone: SLREN Uganda: RENU Togo: TogoRER Zambia: ZAMREN Zimbabwe: ZARNet Countries without an NREN Angola Cape Verde Eswatini Central African Republic Eritrea Republic of Congo Lesotho Equatorial Guinea Mauritius Gambia South Sudan Guinea Bissau Seychelles São Tomé and Príncipe Source: KCL using data from ASREN, UbuntuNet, and WACREN25 22 ASREN website, http://asrenorg.net. 23 WACREN website, https://www.wacren.net. 24 UA website, https://www.ubuntunet.net. 25 Other Arab countries that are members of ASREN but not part of Africa have been omitted from this list. 11 Availability and Sufficiency of National Connectivity Irrespective of the cost estimates to connect any country, there needs to be sufficient national backbone coverage for national transport and adequate network points of presence (PoPs) to enable last-mile connectivity to HEI campuses. National fiber coverage in Africa varies widely, influenced by geography and level of competition and investment by public and private sector operators, among others. While many countries have seen significant deployment of their national backbones, which has enabled connection to major cities, where most of the HEIs are located, last-mile connections to institutions outside major cities are still a big challenge. The data indicate that small geographic-sized nations, such as Burundi, Cape Verde, Mauritius, Rwanda, and São Tomé and Príncipe, and digitally advanced countries, Figure 5: Coverage of African regional RENs such as Morocco, South Africa, and Tunisia, have made Source: AfricaConnect3 good progress in building terrestrial backbone networks. Angola, Algeria, Cameroon, Egypt, Gambia, Ghana, Kenya, Mozambique, Senegal, Tanzania, Uganda, Zambia, and Significant connectivity gaps in Africa are associated Zimbabwe have networks that could support most of with the maturity and effectiveness of the RRENs that, in their higher education connectivity needs. Some of these the context used here, is related to the ability to roll out, countries, such as Egypt, have a very high concentration operate, and maintain high-capacity regional connectivity of the population in urban areas, where fiber networks based on NREN demands; provide sufficient global are well-built. Over half of African countries, however, still connectivity and a range of services, and enable effective need substantial investment into their terrestrial fiber regional collaboration among regional communities backbone networks. of practice. From this perspective, the UA is the most advanced RREN in Africa, and it is not surprising that Thirty-four out of 54 African countries have at least one IXP universities in this region (even with the exclusion of South in the country to help facilitate local traffic exchange and Africa) generally have much higher bandwidths at much save expensive international transit.26 In 2010, the Internet lower prices. RRENs or similar models must, therefore, be Society’s (ISOC’s) team in Africa set a target endorsed by recognized as critical success factors for the availability the African IXP Association to localize 80% of internet of sufficient connectivity, especially in the early stages of traffic at both national and regional levels by 2020. At NREN development. the country level, this has only been achieved by South Africa, with Nigeria and Kenya at about 70% of their traffic exchanged locally—the latter two make annual savings of approximately USD 40 million and USD 6 million, respectively.27 Thirteen out of 54 African countries have at least one carrier-neutral data center, with South Africa having 21, followed by Nigeria and Mauritius with 10 each, and Kenya with 7. 26 Africa IXP Association, http://www.af-ix.net/. 27 Internet Society, Anchoring the African Internet Ecosystem: Lessons from Kenya and Nigeria’s Internet Exchange Point Growth (2020). 12 Existence and Effectiveness of NRENs The current experience within sub-Saharan Africa has to date shown NRENs to be the most effective approach in accelerating a reduction in HEI connectivity costs through aggregation and economies of scale. The NREN survey results collected during the gap analysis phase highlight a number of findings summarized in this section. Figure 6 indicates that all NRENs (100%) serve universities, 85% serve research institutions, while 75% serve TVETs. Other types of institutions served include bodies associated with the educational sector, such as examination bodies and education sector regulators. Figure 6: Types of institutions served by African NRENs Source: NREN survey, 2020 Survey results show that African countries are at different levels of NREN development or maturity based on a number of indicators relevant to higher education connectivity. The most pertinent ones integrated into the cost model include the following (see Table 25): i. Presence of NREN (1 point); ii. NREN governance structure (1 point); iii. Government recognition of NREN/NREN relationships (1 point); iv. Variety of funding sources for NREN (1 point each for membership fees, government grants, and sale of bandwidth); v. Whether has a network (virtual 1 point, physical 2 points); vi. Whether NREN has an Autonomous System Number (ASN) (1 point).28,29 This facilitates routing within the NREN network, the exchange of routing information with other network operators, and the ability to directly peer with an IXP; vii. Whether at least one university has its own ASN that can facilitate multi-homing (1 point); viii. Whether any ASN peers with any other networks (1 point); ix. NREN regional/global connectivity (transit in Africa 1 point, transit in Europe 2 points); x. Middle-ware services offered by NREN (1 point each for ICT training, DNS, NOC services); xi. Advanced services offered by NREN (1 point each for identity and access management, data center services, video conferencing, research management tools). 28 AfriNIC is the regional internet registry that allocates these for the African region, https://afrinic.net/asn. 29 AfriNic ASN Statistics, https://stats.afrinic.net/asn/. 13 Using a combination of Duncan Greaves’ NREN Capability Maturity Model30 and Mike Foley’s levels of NREN development,31 different African countries can be categorized into four broad groups in terms of NREN maturity, summarized in Table 5. Table 5: Stages of NREN development Contribution Name Status Countries to savings via aggregation Angola, Cape Verde, Central • No established NREN African Republic, Comoros, • Varying levels of awareness about Republic of Congo, Eritrea, NREN need Equatorial Guinea, Eswatini, No NREN 0% savings • Ongoing conversations Gambia, Guinea-Bissau, • HEIs buy bandwidth directly from Lesotho, Mauritius, São Tomé ISPs and Príncipe, Seychelles, South Sudan • Formal NREN established as legal Botswana, Burkina Faso, entity Burundi, Cameroon, Chad, • Formal commitment from HEIs Djibouti, Guinea, Liberia, Emerging NREN 30% savings • Formal NREN organizational Libya, Mali, Mauritania, structure Namibia, Niger, Sierra Leone, • Without a network Sudan, Zimbabwe • Coherent operations of NREN Algeria, Benin, Cote d’Ivoire, • With a network of varying coverage DRC, Egypt, Ethiopia, Gabon, • NREN has ASN and peers with Ghana, Madagascar, Malawi, Connected NREN other networks Morocco, Mozambique, 60% savings • Members may also have own ASNs Nigeria, Rwanda, Senegal, to support multi-homing Somalia, Tanzania, Togo • Offers middleware services Tunisia, Zambia • Network with high-speed regional or global connectivity to other NRENs Mature NREN Kenya, South Africa, Uganda 90% savings • Offers advanced NREN services • International collaboration and access to advanced services Source: NREN survey, 2020 30 Greaves, NREN Model. 31 Foley, National Research and Education Networks in Africa. 14 Mature NRENs additionally offer a range of traditional services beyond bandwidth services, as detailed in Table 6, that other approaches to aggregation (e.g., using commercial service providers) do not provide. The reality that only about twenty African countries have NRENs that deliver connectivity to HEIs, and of these, less than five (allowing for some progress since data was collected for this study) can be considered as mature NRENs, points to the urgency of addressing this institutional gap in the African development ecosystem. Table 6: Traditional NREN services Traditional NREN services Examples Connectivity (ALL), eduroam, IPV6, Network Monitoring, troubleshooting, Network services disaster recover, QoS, managed router services CERT/CSIRT, vulnerability scanning, anti-spam solution, intrusion detection Security services services Identity services Identity federation, eduroam, eduGAIN Collaboration services Journal access, mailing list, e-mail hosting, content management services Multimedia services Web conferences, events recording Storage services DNS hosting, cloud storage, file sender, virtual machine, web hosting Professional services Training and capacity building services Source: NREN survey, 2020 NRENs charge their members a wide range of prices for bandwidth, from zero in countries such as Côte d’Ivoire, Ethiopia, Gabon, Senegal, and Tunisia (government covers the cost) to as high as $900 per Mbps/month in Chad due to the market structure and regulatory environment. Figure 7 shows that most NRENs charge between $25 and $99 per Mbps/month to account for the high cost of distributing bandwidth to members. Figure 7: Amount for 1 Mbps/month charged by different NRENs Source: NREN survey, 2020 15 Table 7 compares bandwidth pricing among the selected African NRENs that participated in the survey. Table 7: Comparison of bandwidth prices across NRENs NREN/Country Cheapest (USD) Bandwidth (Mbps) Most Expensive Bandwidth (Mbps) (USD) KENET 5 ≥ 4,000 80 ≤5 (Kenya) MAREN Does not vary with - - 85 (Malawi) amount MoRENet Does not vary with - - 60 (Mozambique) amount MARWAN 3 ≥ 5,000 21 ≤ 100 (Morocco) NgREN - - 25.5 Varies with amount (Nigeria) RENU 10 ≥ 5,000 50 ≤ 99 (Uganda) SomaliREN 92 ≥ 50 115 ≤ 10 (Somalia) TERNET 15 (in capital) 85 (in capital) ≥ 1,000 ≤5 (Tanzania) 35 (outside capital) 100 (outside capital) Source: NREN Survey and Interviews with CEOs, 2020 The level of maturity affects the potential savings through demand aggregation. For example, 40% of NRENs own no physical network that can be used to distribute bandwidth to members. This would potentially increase the cost of bandwidth because NRENs have to rely on other providers to deliver bandwidth to member institutions. This highlights the need to invest in better infrastructure that can be used to distribute high-speed connectivity to universities and TVETs in various countries at cheaper cost. NRENs face many challenges, including the following: i. Lack of awareness among both members and other stakeholders including decision makers in the ministries of education, finance, and ICT and heads of HEIs about what an NREN is and how it benefits HEIs; ii. Constraining policy and regulatory environments that impact NREN operations and also lead to high costs of national and international bandwidth; iii. Limited or no distribution network for last-mile access to peri-urban and rural institutions that are also likely to be more challenged in terms technical capacity and funding; iv. Competition with commercial service providers, creating a confrontational rather than a collaborative environment; v. Lack of sufficient funding among member institutions that also have many competing priorities, leading to defaulting on payments or late payments for services; vi. Poor campus networks at most member institutions so that that more bandwidth does not directly result in visible changes in speed or user experience; and vii. Low levels of technical expertise among ICT staff, along with retention challenges, within NRENs and at member institutions. 16 Sustainable NRENs must be recognized as a critical Existence and Sufficiency of Campus Networks element in delivering sufficient and reasonably priced connectivity to universities at the national level. The quality of campus networks plays a critical role in the higher education connectivity value chain. Assessments NRENs owned and operated by universities are the most by the Network Startup Resource Center (NSRC) and successful; such NRENs are responsive to the needs of the the International Network for Availability of Scientific users but still need support and funding from government Publication (INASP) indicate that most of the campuses in or governmental agencies.32 In this model, there is mutual the African HEIs suffer from poor design and fragmented recognition of the roles of government on one hand and institutional management, with campus ICT services often research and education institutions on the other hand in seen as not critical to the institutions—which leads to lack creating an effective NREN. What has emerged since the of funding. publication cited is that Availability and Sufficiency of Last-inch Connectivity: Individual Access Level “this cooperative model, if well handled, The current modern research and education environment produces the best of both worlds: requires that all students, researchers, and faculty have availability of funding from government; individual and full-time access to online resources and and management and control by collaboration opportunities. The historical approach for the universities through a Board and universal access to computers for students at HEIs was executive arrangement selected by computer labs. However, in addition to the challenges of members (with possible government sustainability and the growing numbers of students not representation). This model permits the resident on campus, the closing of universities during NREN to be managed and operated with the COVID-19 lockdowns has underscored the reality that computer labs are neither sustainable nor versatile. This the structures and efficiency of a private has accelerated the shift toward one-to-one computing. sector organization while maintaining This shift brings in the challenge of the affordability of accountability to both government end-user devices for students that come from the poorer and the members. The best example of sections of society as a gap to be addressed. this the cooperative approach on the African continent is South Africa: While The ability to own a decent laptop, which should be the TENET started and is operated as a minimum entry level for students at the higher education grassroots NREN, SANReN – owned by level in view of the applications they run and the work the government of South Africa – later they do, is dictated by family wealth. Previous World Bank brought on board high-capacity national studies have confirmed that students from wealthier and international connectivity that has families dominate higher education enrollment in Africa been entrusted to TENET to manage and and are more likely to successfully complete their studies on time.34 But there are few families that can afford to operate. TENET and SANReN operate buy computers (only about 15% of the population in Africa under a cooperative umbrella called currently lives on more than $5.50 a day)35 and the model the South Africa NREN: SANREN. It will work for only a small proportion of students in Africa, should be noted that this cooperative raising equity challenges. It is imperative that access arrangement can be transitioned to, to end-user devices does not create another rich-poor regardless of whether the NREN was divide, regardless of whether it is a rural or urban campus, started by HEIs or by government, compounding the divide that exists among students though the process can be long. A throughout their educational life. key element in achieving this is trust, which enables the strengths of each side to create the synergy required for sustainability33.” 34 Darvas, Peter, Shang Gao, Yijun Shen, and Bilal Bawany. 2017. Sharing Higher Education’s Promise beyond the Few in 32 Foley, National Research and Education Networks in Africa Sub-Saharan Africa. Directions in Development.Washington, 33 See GÉANT Policy Paper: “Breaking the Final Connectivity DC: World Bank. doi:10.1596/978-1-4648-1050-3. Barriers for Higher Education Institutions in Africa: 35 https://blogs.worldbank.org/opendata/85- The Next Steps and A Call to Action” (2021). africans-live-less-550-day 17 The associated challenge, originating from lower levels of education, is the absence of, or very limited, digital literacy, which again affects the poorer sections of the population most. Summarizing the Dimensions of Connectivity Gaps and Divides The environment in higher education connectivity is as diverse as the countries involved; therefore, the challenges are often complicated, specific, and contextual. To these must be added the divides within the same country, which include urban versus rural locations; rich versus poor families; and level of digital literacy, which is linked to family wealth. National ICT indicators collected during the gap analysis phase also have a direct bearing on connectivity for higher education institutors at the country level. The most pertinent integrated into the cost model include the following: i. Whether the country is landlocked or has access to the ocean, the latter of which allows direct access to submarine cables. Direct access to submarine cables reduces base bandwidth cost. ii. Number of submarine cable landing stations. Landlocked countries have none, while Egypt has the most with 15. More landings improve competition amongst cable providers, resulting in competitive pricing. Internet exchange ladder stage. Countries were categorized into four stages (see Table 8) depending on the number of IXPs and carrier-neutral data centers they have as well as the interaction between these two important facilities.36 Table 8: Stages of the internet Exchange Ladder Stage Status Countries Algeria, Cabo Verde, Central African Republic, Chad, Comoros, No IXP, internet traffic Equatorial Guinea, Eritrea, Ethiopia, Guinea-Bissau, Lesotho, Libya, Stage 0 exchanged overseas Mauritania, Niger, São Tomé and Príncipe, Seychelles, Sierra Leone, Somalia, South Sudan Benin Domestic internet traffic Botswana, Burkina Faso, Cameroon, Congo, Côte d'Ivoire, Egypt, Stage 1 between ISPs exchanged at Eswatini, Gabon, Gambia, Guinea, Liberia, Madagascar, Malawi, Mali, IXP Namibia, Rwanda, Senegal, Sudan, Tanzania, Togo, Tunisia, Zambia, Zimbabwe Diversity of participants at IXP, presence of global Angola, Burundi, Democratic Republic of Congo, Mauritius, Morocco, Stage 2 Content Distribution Mozambique, Uganda Networks (CDNs) IXP located alongside carrier- Stage 3 neutral co-location data Djibouti, Ghana, Kenya, Nigeria, South Africa center Source: NREN survey and Interviews with CEOs, 2020 iii. Percent of population within 10 km of fiber coverage (reflects fiber network coverage of the country). This has a direct bearing on the cost of connecting especially rural campuses. iv. Regulatory score, which reflects the maturity of regulatory environment. It is based on individual country scores from ITU Global Regulatory Outlook 2020. A good regulatory environment leads to more competitive offerings. 36 World Bank Group, 2020. National Data Infrastructure The Role of Internet Exchange Points, Content Delivery Networks, and Data Centres (was still in draft form) 18 Global Presence of submarine Presence of cables and effective number ofaa RRENs landings Enabling Cross-border policy and fibre Regional regulation networks Number of Carrier- Number of Neutral IXPs Penetration Data Centers Human Taxation on of national Availability Presence of Capacity to ICT fibre of reliable effective design/ eSuipment National backbones power NRENs operate and (public & supply services connectivity private) Local (rural vs. urban) Institutional State of Availability of capacity to Tangible Barriers/Divides Campus Last Mile exploit Campus networks access broadband connection Ownership of computers and laptops Level of Intangible Family wealth digital Barriers/Divides User (rich vs. poor) Literacy Figure 8: Tangible and intangible barriers for connecting higher education Source: KCL A summary of tangible and intangible barriers is given in Figure 8. It is evident from the figure that multiple variables interact and that there can be no one-size-fits-all approach to addressing the gaps. While a general intervention approach can be developed, as is done in this report, it would need to be adapted to the specific gaps and challenges in each country. A Review of Previous and Existing Programs on the University Connectivity Agenda Multiple multilateral development organizations, foundations, and multinational companies have been, or continue to be, actively engaged in supporting the advancement of higher education connectivity in Africa, especially over the last 20 years, underscoring the broad interest in advancing higher education in Africa. These have included, among others, the Partnership for Higher Education in Africa (PHEA), which supported the African Bandwidth Consortium; the Leland Initiative, funded through USAID and which provided wireless backbones to campuses in selected countries; and Fostering Research and Education Networking in Africa (FRENIA), funded by the Andrew W. Mellon Foundation, which provided funding for the startup of NRENs in Africa. The International Development Research Center (IDRC) provided initial funding for NREN development and especially fostered the creation of RRENs, including the UA and the WACREN. The European Commission, through the AfricaConnect37 Projects, has to date deployed the highest level of resources to support and grow both connectivity and other aspects of NRENs and RRENs in Africa. The World Bank has provided direct funding—Burundi, Mozambique, Tanzania, Somalia, Malawi, and Nigerian NRENs38 are among the beneficiaries. While not planned directly for university connectivity, the World Bank Regional Communications Infrastructure Project39 (RCIP) has benefited universities in Rwanda and, much earlier on, Kenya. The Open Society Foundation and Open Society Initiative (OSI) of South Africa have provided support to the UA. The African Development Bank (AfDB) has funded a number of higher education projects to improve science, technology, and innovation (STI). Guided by the New Education Model in Africa (NEMA), which emphasizes ICT-based delivery as a central component that is adapted to different country contexts, the AfDB has funded ICT infrastructure (networks and computers) and training to help improve the delivery and quality of STI programs in HEIs. 37 https://africaconnect3.net 38 The support to Tanzanian universities was directed through a government Ministry. It subsequently ran into challenges of sustainability. 39 https://www.worldbank.org/en/search?q=Regional+Communications+Infrastructure+Program. 19 Foundations such as the Bill & Melinda Gates Foundation, Carnegie Corporation, Rockefeller Foundation, Ford Foundation, John D. and Catherine T. MacArthur Foundation, William and Flora Hewlett Foundation, and Andrew W. Mellon Foundation have been playing roles in the development of the capacity of HEIs in Africa. These foundations have not only provided funding but also carried out considerable analyses into the problems of connectivity, content, and knowledge sharing between academic and research institutions and libraries in Africa. Other partners of higher education connectivity include the following: i. Multinational companies, including CISCO, Google, Intel, Juniper Networks, and Microsoft which provide tools and equipment at concessionary prices along with training opportunities; ii. Research networks that provide technical assistance and experiential support—GÉANT Association, RedCLARA, and Internet2; iii. The NSRC at the University of Oregon and ISOC, both of which play a major role in training NREN engineers on network operations and management; and iv. The African Internet Registry (AfriNIC) in the delivery of IP numbers and other resources. Table 9 shows the major partners, both previous and current, for university connectivity projects in Africa. Focus Area Major Partners Support Remarks/Lessons While NSRC, ISOC, and CISCO provide direct training opportunities, the GÉANT Association Direct technical and more advanced NRENs open opportunities training and for bilateral collaboration or twinning, which NSRC, ISOC, GÉANT support, financial have been a major source of learning at the Association (formerly support for management and operational levels, especially Capacity building DANTE and TERENA), capacity building, through attachments and secondments that are RedCLARA, OSI, CISCO, sponsorship for needs driven. The Cisco academies established AfDB, European Union participants, around Africa have provided a lot of training equipment (CCNA and CCNP) for networking professionals. A sizeable portion of EU funding is dedicated to capacity building. Foundations such as It should be noted that the initial major drive for Carnegie, Hewlett, Bill connectivity was driven by the need for easier & Melinda Gates, Ford Support to specific Content access to global information resources and that Foundation, PHEA, and research institutions the Carnegie Corporation of New York was a other institutions such major player in driving the formation of the PHEA. as INASP AfriNIC, through negotiations led by the Research and Education Networking Unit of the Internet resources AfriNIC ASN, IP numbers Association of African Universities (AAU), agreed on a discount of 50% on the costs of ASNs and IP addresses for the REN community in Africa. Development partner funding normally covers costs of travel and board. The actual knowledge Andrew W. Mellon and experience support has been donated Foundation, European by more advanced NRENs or RRENs through NREN development Commission, AfDB, Bilateral funding discussions, attachments, and secondments. World Bank, Canada While a significant part of this has been from (IDRC) outside Africa, the major part has been intra- Africa. 20 Until the funding of AfricaConnect, the EC had never funded IRUs. Earlier initiatives to support REN growth and connectivity (South America, Asia, Northern Africa, and the Middle-East) Financial resources, focused on recurrent payment for bandwidth, Regional network European Commission long-term leases for which meant that funds were exhausted development fiber without creating sustainability. The UbuntuNet insisted on IRUs, leading to a delay of almost two years before the EC gave consent. IRUs are now a common feature of GÉANT connectivity procurements. Sources: various A review of the different initiatives, especially the larger funding to connect higher education, such as the EU-funded AfricaConnect project, brings out the following key lessons: i. The need for collaboration: All the efforts around access, capacity, and content focus on the same end beneficiaries. Most of them are, however, isolated from each other, losing potential synergy. Future initiatives should make an effort to bring all stakeholders around an integrated plan of intervention. The PHEA is noteworthy for bringing key American private foundations together around common causes; and the AfricaConnect programs work collaboratively among the GÉANT Association (formerly DANTE and TERENA), NSRC, and ISOC. ii. The need for beneficiary contribution and driving direction: The AfricaConnect initiatives have demonstrated beneficiary contribution as a key aspect of sustainability. They have also been responsive to beneficiary needs and direction, with outside expertise bringing on board especially procurement and communication skills. Networks in the Alliance region have been implemented and are operated by the owners. iii. Government support: Many of the NRENs are challenged by the need to contribute to any initiative, and the smaller ones much more so. This does not reduce the necessity for such contribution but rather points to the need to get government commitment to counterpart contribution before any intervention. African governments need to step up and take ownership of the national and regional RENs as a critical development necessity. Where governments are not committed, investments will not be sustainable. 21 Addressing the Gaps protection of critical infrastructure, including ensuring sufficient compensation for fiber cuts; eliminating or There are challenges and gaps at the five levels of reducing taxes on communication and communication connectivity—global, regional, national, campus, and equipment; and deploying universal access funds to user. An awareness gap among decision makers was also enable broadband in remote and sparsely populated areas found as an important bottleneck to the development of are all action areas under the direct control of national connectivity in higher education. All the following need to governments. be considered as part of the business and implementation plans to achieve the goal of connecting all African HEIs to Campus networks as well as individual ownership of high-speed internet. A gap that cannot be integrated in laptops remain the weakest links in the connectivity value the business and implementation plans, but is a necessary chain. It needs to be noted that campus networks must be part of any major implementation is increasing the approached in the wider context of national-level access coordination among the various partners who are either to enable non-resident student access and online distance already on board or should be brought on board in order to learning and to create resilience of access in periods such maximize synergies. as the COVID-19 lockdowns. Campus networks will be the most expensive area of investment. The required quality The development of toolkits, policy briefs, and ongoing of campus networks demands the presence of a well- awareness workshops for decision makers will help motivated expert ICT technical resource. address the awareness challenge. Ongoing platforms that bring university and TVET leaders together to discuss Successful interventions point to the following: device access, campus networking, and NREN issues i. The need for collaboration: All the efforts around will be crucial. Development partners could also provide access, capacity, and content focus on the same platforms for experience sharing among countries and end beneficiaries. Most of them were, however, promote higher ICT strategies for higher education that isolated from each other, losing potential synergy. can serve as avenues for improved understanding of the The WBG initiative should make an effort to bring importance of higher education connectivity. all stakeholders around an integrated plan of intervention. The promotion of regional and continental collaboration ii. The need for beneficiary contribution and driving to consolidate markets will attract private sector players direction: The AfricaConnect initiatives have to invest in more marine cables, transcontinental cables, demonstrated beneficiary contribution as a key aspect and major data centers. The opportunities created by the of sustainability. They have also been responsive African Continent Free Trade Agreement (AfCFTA) should to beneficiary needs and direction, with outside be seized to this end. This goes along with regional and expertise bringing on board especially procurement continental approaches to enabling policy and regulation. and communication skills. iii. Government support: Many of the NRENs are National governments, once they appreciate the causal challenged by the need to contribute to any initiative, linkages between broadband connectivity and individual and the smaller ones much more so. This does not access in higher education on one hand and development reduce the necessity of such contributions but rather on the other hand need to own and lead efforts at the points to the need to get government commitment national level to address the gaps created by policy, to counterpart contributions before any intervention. regulation, and limited funding for the HEIs sector. This Where governments are not interested, investments includes support to NRENs, including enabling their start- will not be sustainable. off where they do not yet exist. The interest of government does not take away the onus on heads of HEIs to work together to set up and exploit the opportunities of NRENs, which in turn are the foundation for the RRENs. National governments must also lead policy and regulation that will attract major private sector player to establish data centers. Moreover, governments need to lead in ensuring a sufficiency of IXPs to keep local traffic local. The sufficiency and penetration of national backbones, along with a competitive carrier environment, will lead to cost-based pricing and affordability. Eliminating monopolies; reducing the cost of licenses; enforcing the shared use of telecom infrastructure, civil-works, and access to the alternative infrastructure provided by transport and energy operators; legislating for the 22 The WBG and the partnerships they create are in position ii. Demand Side Challenges to bring influence to bear on, and/or provide funding Absence of ICT policies and strategies that link support for, the following supply side and demand side investments in ICT to learning and research: There are challenges: many institutions where the approach to rolling out ICT i. Supply Side Challenges (affect all service providers services and systems is handled casually and piecemeal, including RRENs and NRENs) without any overarching policy and strategy grounded in “the why” of learning and research. The participatory formulation of institutional ICT policies and strategies is a Taxation: Taxation impacts all segments of the delivery foundational gap that will need to be addressed for most chain. Heavy taxation leads to reduced investment capital institutions. for network improvement and expansion, higher prices to achieve good returns on investment, and a lower uptake of Sustainability: Education generally and higher education services. in Africa, right from TVET levels, is severely underfunded, and institutions always struggle to meet costs. Sufficiency Enabling national policies, laws, and regulations: of funding maintenance and expansion is, therefore, National policy, laws, and regulations impact the national an aspect that needs to be carefully examined before segments of the delivery chain. This includes investment interventions are implemented. Unfortunately, it is not policy; ICT sector policy, laws, and regulations, especially just insufficiency of funding that leads to neglect of ICT regarding the availability of class licenses; policy infrastructure—many institutions still suffer from limited inconsistencies that originate mainly from the desire to high-level awareness of the potential benefits of excellent increase tax revenue, arising from the finance sector, and ICT services and systems, which places these among the the desire to reduce the cost of devices and services from bottom priorities. the ICT sector; and the sometimes-aggressive competition between NREN and operators, which operators always Shortage of computers and laptops compounded by win. limited digital literacy: The only sustainable solution to end-user access is universal personal ownership of Regional barriers: Regional barriers originate from laptops that have the capability to handle the applications inconsistencies in policy, laws, and regulations across used and the work done at the higher education level. national borders, even within the same economic blocks. It is therefore evident that this is one of the key areas The African Continent Free Trade Agreement (AfCFTA) is a to be addressed if all the upstream investments are to key piece in addressing this. achieve the desired outcomes. Related to the shortage of computers and laptops is digital literacy. The shutdown of services by governments: The partial or total shutdown of selected services and quite often the The challenges around this and how it should be addressed internet severely disrupts operations and, where it occurs are discussed in another report.40 periodically, is a disincentive for investment—it leads to a loss in revenue that cannot necessarily be recovered Lack of technical competence to implement, without taking governments to court. maintain, and expand services and systems: A major underlying cause for the absence of competent human Insecurity: This is really just a reality to be recognized. resources is the lack of appreciation for ICT expertise. This Insecurity due to internal conflict, regional conflict, or is compounded by insufficient budgets, as discussed under terrorism, wherever it occurs on the continent, hinders sustainability to hire and retain competent personnel. or makes it more expensive, or completely blocks, the rollout of the high-capacity infrastructure, especially fiber, required to deliver universal broadband. The Sahel region, the Horn of Africa, and parts of the Great Lakes region have been particularly prone to continuing armed conflict, terrorism, or both, and will pose a significant challenge in implementation. 40 Refer to Report 3: High-level Strategy to Leverage the Campus Connectivity to Achieve Learning Outcomes in Higher Education. 23 4 Context and Connectivity Gaps in Higher Education B ased on the numbers of end-users and campuses, Setting progressive targets took into account several this chapter quantifies the gaps at each of the factors, including the following: major elements of the connectivity value chain i. The need to factor in the challenges of sustaining and develops a model to provide cost estimates major investment that must, in the medium to long of the required interventions at national, regional, and term, be taken up by African governments—this pulls continental levels. Starting with the vision and the down the level of investment. progressive targets, the model looks at funding needed for ii. The current severely suppressed demand due to i. Access devices very high costs. Volume bandwidth procurements ii. Campus networks using long-term leases where applicable have been iii. Upstream connectivity demonstrated to drive down prices sharply and iv. Regional and global-level connectivity therefore increase demand, leading to a virtuous cycle. This pushes up the investment. iii. The opportunities of next-generation technologies The Vision and Progressive Targets for that are coming increasingly into play, leading to Connecting HEIs much greater bandwidth volumes over existing fiber networks. This maximizes opportunities from the These cost estimates are based on a high-level vision investment. of connecting African HEIs to networks and services iv. The bandwidth requirements of commonly used comparable to the rest of the world and the associated applications, technologies, and resources that can progressive targets. support increased access to and the quality of higher education.41 Vision: v. The current spend on bandwidth in many HEIs around Africa would be sufficient for the volumes they need to be at par with the rest of the world if the cost of bandwidth was at par. This assures sustainability. An African continent where all higher education institutions The progressive bandwidth targets in Table 10 with achieveglobal parity in intellectual projections for 2025 and 2030 take into account these and output and development impact also include comparators from around the world. throughaccess to, and exploitation of broadband connectivity at capacities,quality, and costs comparable to the rest of the world. To link this to national and regional development benefit requires that these African institutions simultaneously develop the necessary pre-conditions to ensure that the sufficiency and affordability of broadband can be seized as opportunities to improve learning and research outcomes as well as employability in the context of the fourth industrial revolution. This vision is not limited to the capabilities or funding from any initiative; it is rather focused entirely on where Africa desires to be. From interviews with various key stakeholders, there is a common aspiration that African HEIs should, at a minimum, be at levels comparable with the rest around the world in terms of connectivity if parity 41 S. Bashir, Connecting Africa’s Universities to Affordable High-Speed Broadband Internet: What at a global level in both intellectual property output and Will it Take? (Washington, DC: World Bank, 2020). development benefits is to be achieved. 24 Table 10: Recommended Progressive Bandwidth Targets for African HEIs Minimum Year Remarks Bandwidth 2021 (target 0.2 Gbps @ 1,000 Translates to 1 Gbps for a campus of 5,000 and 10 Gbps for a campus of 50,000 minimum) Translates to 10 Gbps for a campus of 5,000 and 100 Gbps for a campus of 2025 2 Gbps @ 1,000 50,000. This should be the minimum entry level for the WBG intervention. Translates to 100 Gbps for a campus of 5,000 and 1 Tbps for a campus of 50,000. Actual size for any campus to be based on the TENET approach: “sufficient 2030 20 Gbps @ 1,000 bandwidth to be able to use the prevailing applications of the day” with port sizes twice the normal usage. Source: Interviews and discussions with NREN CEOs In addition to the bandwidth targets, it is also recommended that: Computing devices: One-to-one for both faculty and students. All faculty should own a laptop within the first year. Ownership by students should be phased in through ensuring that all first-year students can secure personal laptops over a successive number of years. Since most courses have a three-year duration, this means that all students would own a laptop by 2023—this assumes 2021 as the first year of implementation. Section 4.2 discusses the various acquisition models. Campus networks: Functioning campus networks with external connectivity of at least 2 Gbps per 1,000 students. Moreover, ubiquitous and properly dimensioned broadband Wi-Fi should be achieved by 2025. Immediate focus should be placed on promoting reliable and functioning campus networks over the next two years. Section 4.3 presents the costing for upgrades to campus networks. NRENs: All countries should have functioning NRENs by 2023. Capacity building along with connecting more HEIs should create opportunities for full NREN maturity by 2025. RRENs: The three regional RENs that cover Africa have extended connectivity to all African countries and also interconnected their networks at multiple points within the continent by 2025. Access to Laptops While smartphones and to a greater extent tablets can support teaching and learning, they still lack the full range of attributes and functionality to serve as full-fledged individual learning platforms—the focus here is therefore transitioning to individual laptops. Universities and TVETs have adopted many models to ensure that devices are available to students and teachers. The most common models are as follows: i. Allowing students and faculty to BYOD ii. Enabling ownership through an institutional scheme 25 The BYOD Approach The first seeming challenge of the BYOD approach is that, based on a global population approach, there are few families that can afford to buy computers (only about 15% of the population in Africa currently lives on more than $5.50 a day),42 and the model will work for only a small proportion of students in Africa, raising equity challenges. The reality, however, is that the nature of the education systems in most African countries is such that it is the richer families that can afford the schools that produce the overwhelming majority of students who get into HEIs through the competitive selection processes. Where a BYOD approach is adopted, a robust acceptable use policy (AUP) that defines the code of conduct for the ethical and safe use of campus network resources, digital resources, and data in compliance with national and global privacy regulations, intellectual property rights, security, etc. should accompany the scheme. Universities will need to enforce requirements such as minimum specifications of personal devices and security measures. Box I: New Brunswick, Canada, Education and Early Childhood Development—Laptop Subsidy Scheme The New Brunswick Department of Education and Early Childhood Development Bring Your Own Device (BYOD) program is aimed at students in grades 9 to 12. The program is designed to allow for more personalized learning opportunities to help better prepare students for post-secondary education and the workplace. The Department understands that purchasing a new device may be a financial pressure for families. In response, the Department launched a financial assistance program for low- to middle-income families. Under the program, parents and guardians have two options to help provide their student with a device: i. Buy a device that meets the minimum device requirements (see Related Links section) and apply for reimbursement within six months of the purchase. Reimbursements can be provided by e-transfer or check. ii. Apply for a subsidy online. Once the application is approved, families will receive a subsidy code to apply toward a laptop’s online purchase. Families that receive the full subsidy are entitled to CAN$600 toward the cost of the new laptop. Source: https://www2.gnb.ca/content/gnb/en/services/services_renderer.201514.Laptop_Subsidy_Program. 42 https://blogs.worldbank.org/opendata/85-africans-live-less-550-day. 26 Institutional Schemes Institutional schemes involve bulk educational purchase and delivery of the laptops that are then given to students either at no cost or through a partial subsidy (owner pays a portion) or through loan programs. Bulk purchase, if properly managed and negotiated, exploits economies of scale fully and could be linked to local startups for assembly and service. The management of such a scheme can also be outsourced to suppliers based on negotiated prices to reduce the management and administrative load on the institutions. Box II gives a specific example from Burkina Faso, and Box III gives examples of access and subsidy models for staff and students based on institutional schemes from around the world. Box II: Burkina Faso Partial Subsidy of Student’s Computing Device Staff Device Subsidy Schemes The Government of Burkina Faso aims to provide computing devices to students using a subsidy scheme. The “One Student, One Computer” scheme aims to invest US$3.8 million to provide 10,013 undergraduate students with laptops. The government intends to cover 50% of the cost with support from the World Bank. Students must cover the remaining 40% either via direct cash payment or via loan through the Coris money platform. Loans are available from the National Fund for Education and Research (FONER) or the National Centre for Information, Educational and Vocational Guidance and Scholarships (CIOSPB) for scholarship holders. A local company, Horizon Informatique SA, will distribute the computers. Ultimately the government wants to distribute computers to 50,000 students. The supplier is expected to deliver 8000 computers in 2020. Source: https://www.ecofinagency.com/telecom/0605-41327-burkina-faso-s-govt-officially-launches-its-one-student-one-computer-program. Box III: Subsidy Schemes for Access to Laptops for Students and Staff Staff Device Subsidy Schemes i. Full subsidy for devices for teaching and learning purpose by the HEI ii. 50% from the university budget and 50% from research grant and personal funds iii. Subsidized access to laptops paid by staff through loans Student Laptop Subsidy Schemes i. Free laptop scheme—free rotating laptops for students for an academic year and returned at the end of the year or when students complete a semester. University or government pays for the laptops, which are bought on a discount basis. ii. Laptop checkout program—laptops made available in the university library. Students use a library card to check out a laptop for short- and long-term use, with a nominal fee applied (e.g., North Western University) iii. Laptop loaner program—many universities in developed countries provide laptops as an integral part of students’ loans—a laptop loan is applied to a student loan account Government partial subsidy (contribution) to the price of the equipment—Seychelles, for example, provides 3,000 Rupees (USD 170) toward the purchase of a computer, with the rest covered by students.43 43 http://www.egov.sc/documents/ICT_LaptopScheme_20140827.pdf 27 Implementation and sustainability of institutional schemes can be a challenge, pointing to the need for good advance planning including stakeholder consultation. Box IV summarizes the experience and challenges in Kenya, one of the African countries that has demonstrated long-term commitment to the development of digital skills and innovation. Box IV: Kenya’s Quest towards One-to-One computing for Students The Kenyan government has been pursuing different models for enabling students’ access to computers. The COVID-19 pandemic especially underscored the need for student devices. The Higher Education Loans Board is seeking 2.5 trillion shillings (US$23 million) to distribute laptops to first-year students via loan. The board wants to provide at least 60,000 laptops to government-sponsored students in public universities. The scheme draws on the earlier experience of the government in subsidizing computers for university students: The Wezesha scheme was launched by the Kenya ICT Board and the Ministry of Information and Communication (MoIC) in 2012. It was aimed at providing subsidized laptops for 15,667 university students. The scheme provided 9000 Kenyan shillings (US$120) towards the purchase of a laptop by university students. There were two models of laptops sold by five pre-qualified retailers. The $120 subsidy aimed to reduce the laptop price by 15% and 33%, depending on the laptop model. Source: https://www.ictworks.org/subsidized-laptops-15667-kenyan-university-students-real-ict4edu-investment/#.YM8hay1h29Y), https:// techweez.com/2020/12/08/helb-seeks-billions-to-loan-60000-laptops-to-students/ The Jomo Kenyatta University of Agriculture and Technology has also experimented with a Taifa Laptop program to assemble laptops locally, aimed to be sold to students. While this has faced some challenges, Kenya’s quest for different models indicates that countries may need to adopt multiple strategies to realize the one computer per student target. Source: https://techweez.com/2017/09/04/taifa-laptops-jkuat/, https://kiruik.medium.com/the-taifa-laptop-saga-could-jkuat-have-designed- the-program-better-9ac1d3665dcc Estimating Laptop Costs In order to estimate the costs of computing devices, classroom devices, user software, and data storage, a forecast of the number of university students and staff over the next five years (2021 to 2025) based on available data on the number of higher education students for all African countries is used. Since the time series unit costs for computing devices are not readily available in Africa, average values of the unit costs for these devices (laptops and tablets and other devices such as classroom devices, data storage, software, and graphics) are used. Generally speaking, prices of devices are high in Africa due to high tax rates, often compounded by restrictive trade policies. Table 11 gives the range of costs of laptops and related software across African countries. Table 11: Average device costs in a 1:1 computing scenario Estimated 1:1 computing per student Type of device Description (USD) Portable devices Entry-level laptop for educational purposes 270–45044 Software Word processing, spreadsheet, presentation software 36–108 Total 306–558 Source: Calculations draw on European 2nd Survey of IT in Schools, Objective 2: Model for a “highly equipped and connected classroom” https:// ec.europa.eu/information_society/newsroom/image/document/2019-10/ictineducation_objective_2_report_final_4688F777-CDED-C240- 613EE517B793385C_57736.pdf, see page 84. 44 Estimates are made on a laptop price that ranges between (USD) $270 and $450 used by three students for three years. 28 Aggregate Laptop Costs for Higher Education The model gives total costs for the one-to-one ownership of laptops. How this is phased in or out, and over what period of time, will vary from country to country according to national policies and levels of poverty. It will also be influenced by the funding policies of development partners. Table 12 provides a summary of the aggregate cost of equipping students and staff with access devices. The estimates indicate the cost of equipping first-year students and all staff would be USD 17.3 billion between 2021 and 2025. Table 12: Cost of equipping students and staff with access devices (2021–2025) Estimated Average cost of Forecast student Year Forecast staff numbers devices enrollment (USD, million) 2021 18,741,000 848,000 3,478 2022 19,485,000 882,000 3,142 2023 20,190,000 918,000 3,256 2024 20,949,000 949,000 3,888 2025 21,659,000 985,000 3,521 Total 101,024,000 4,582,000 17,285 Source: KCL calculations Upgrading Campus Networks All campus networks need to be optimized for the intensive use of wireless services, which also lend themselves more easily to the modern learning and research environments. This does not eliminate the need for wired connections where high performance and improved stability are mission-critical. It should be noted that large numbers of non-resident students, the need for student access during holidays, and addressing the demand for off-campus students (distance education) mean that national-level eduroam needs to be considered as an extension of the campus services. Staffing and advanced skills are vital requirements for advanced campus networks. This will include ICT engineers and application developers as well as communication and financial specialists. There is no globally agreed yardstick on the number of skilled ICT staff required for the operation of campus networks. Staff number scales based on the number of users, network devices, number of networks, security issues, complexity of routing.45 In the private sector, one skilled ICT professional typically serves 200 to 300 users.46 HEIs need at least one skilled ICT professional per 300 to 450 users. Table 13 summarizes the assumptions used for calculating the cost of upgrading campus networks across all African HEIs. Table 13: Assumptions for calculating campus network upgrade costs in a country Area Assumptions Average number of students per institution is defined as 3,000 for a small campus, 9,000 Student enrollment students for a medium campus and 24,000 students for a large campus. A small campus has 2 medium and 3 large buildings, a medium campus has 4 medium and Number of buildings 6 large buildings while a large campus has 8 medium and 12 large buildings. A small campus needs a 5-km fiber network backbone, while medium and large campuses Length of back-haul fiber need a 10-km and a 20-km backbone, respectively. Assumed a unit cost of $20 per meter of laying fiber, including civil works. 45 https://verber.com/it-staffing/. 46 https://www.auvik.com/franklyit/blog/tech-user-ratio/. 29 Small campuses have a simple network with 1 switching center (with a core router and layer-3 switch), while medium campuses 2 switching centers (each with a core router) and one border router. A large campus has 3 switching centers (each with a core router) and Switching centers two border routers, giving the network the ability to support 2 independent connections. Given the poor reliability of power in many African countries, each switching center will have a standby generator. A campus needs a small data center (tier I) with racks, centralized UPS, and some servers. We budgeted 3 servers for a small campus, 9 servers for a medium campus, and 12 servers Data center for a large campus. The data center, switching center, and network operations center (NOC) should be co-located to save on costs. A small campus has 1 fully integrated smart classroom with different technologies, Multimedia classroom including smartboards, projectors, cameras, speakers, audio equipment, lighting, etc. A medium campus has 3 of these, while a large campus has 5 smart classrooms. Support to institutional A local area network and PCs in the main library that are connected to an online public library access catalogue (OPAC). At least 1 ICT skilled professional for every 450 students that earns at least 1.5K per Skilled staff month to compete with the private sector. The staff should have access to one training opportunity per year in line with the needs of their institution. Institutions should be able to access technical support to help the technical team Consulting and design implement various solutions that address their institution's needs. This can start with support campus network design and span to other areas, including installing and maintaining various systems and equipment. Equipment supplies and Institutions should be able to undertake corrective and preventive maintenance to extend maintenance the campus network's life and operation. Source: KCL calculations Table 14 shows a summary of the estimates of the capital expenditure (CapEx) and operating expenses (OpEx), given at a regional level, for upgrading all university and TVET campus networks across Africa. The OpEx cost covers five years (a default period that can be modified in the model) and includes a maintenance component for the campus networks (15% for hardware and software costs). It should be noted that the gross estimate of USD 27.3 billion for campus networks is based on the broad categorizations of campuses as small (< 5,000 students), medium (> 5,000 and < 15,000 students), and large (> 15,000). Small campuses account for about 94% of all HEIs and about 83% of the total cost of upgrading campus networks. Where more detailed data on campus sizes are available, the small category can be refined further into micro, mini, small, medium, large, and very large campuses. The modeling of the case study countries at this level (as provided for in the Cost Model) led to reductions in the cost of campus upgrading by 37%, 6.7%, and 32.2%, respectively, for Côte d’Ivoire, Mozambique and Uganda. This potential reduction has not been factored into the gross estimated cost in the summary because the number of countries analyzed is too small to be used as a basis for a reliable generalization across the continent—but it does point to a significant potential reduction in the gross cost. 30 Table 14: Summary estimates for upgrading campus networks by region (rounded to nearest 10) CapEx (USD, OpEx (USD, Region Total (USD, millions) millions) millions) Eastern and Southern Africa 6,470 8,450 14,920 Northern Africa 1,680 2,330 4,010 Western and Central Africa 3,600 4,730 8,330 Total 11,750 15,510 27,260 Source: KCL calculations Box VII: Norwegian GigaCampus Project In 2005, in response to the Norwegian Ministry of Education and Research, the Norwegian Association of HEIs and the higher education sector, the Norwegian NREN UNINETT launched a four-year project entitled GigaCampus 2006–2009. The project was granted financial support amounting to NOK 45.8 million (USD 5 million) to coordinate the evolution of a world-class campus ICT infrastructure around Norway. A key objective was to strengthen the community of network engineers from the various universities and colleges around the country through working groups, seminars, and workshops. They were encouraged to share their experiences on campus network development for the benefit of the whole higher education sector. GigaCampus worked within seven areas of focus that were identified as critical for campus networking, including physical infrastructure, networking and design, mobility, real-time communications, security, network operations, and monitoring. The project held a total of 47 seminars and workshops during the four-year period. The working groups produced a total of 22 best-practice documents. The knowledge exchange encompassed several issues—wireless setups with eduroam; core campus network upgrades with increased capacity, functionality and resilience; IPv6 implementations; security architecture design; network monitoring setup; etc. GigaCampus was also involved in building campus projects, giving recommendations to the design of the data center and communication rooms. This includes cabling, power, and cooling as well as fire protection systems. GigaCampus ran several national-level procurement processes for ICT equipment for campus networks. Thirty agreements within 10 principal fields were signed during the four years. The coordination of these purchasing operations resulted in substantial economy of scale advantages for ICT equipment in terms of price and contractual terms. The coordination and standardization of infrastructure, bringing network engineers together and agreeing on joint best practices through technical specifications, generated a long-lasting benefit for the higher education ICT community in Norway. Source: https://services.GÉANT.net/sites/cbp/Knowledge_Base/Reports/Documents/gigacampus_final_report.pdf 31 Connecting Campuses Upstream This is important for a developing country context because, based on the surveys as well as specific discussions with CEOs of selected NRENs, a major NREN as a Model of Connectivity for HEI percentage of the traffic from most HEIs is to non-REN users, largely because research is still limited in most HEIs. Following are four models for the connectivity of any HEI Any attempt to limit access to only users and resources depending on the maturity of the NREN, the maturity within the REN community would eliminate the early of the telecommunications markets, the extent of development of the value proposition of lower-cost government support, and the financial capacity of the connectivity. institutions: i. Model 1: Connecting exclusively via CSPs; Model 4, emphasizes the importance of CSP and NRENs ii. Model 2: Connecting via either CSP or NREN; working together to provide cheaper rates for HEIs. This iii. Model 3: Connecting exclusively via NRENs; requires various conditions to have been satisfied in the iv. Model 4: An emerging advanced and hybrid model, country, as follows: connecting to both CSPs and NRENs. In this model, i. The telecommunications market is fully mature— CSPs provide commodity internet47 and exchange at providing sufficient and competitive cost-based a far cheaper rate, while NRENs and RRENs handle choices for all services and to all users—including research and education traffic. universal broadband service; Model 1 is the predominant model in countries without ii. Competition in the provision of broadband services NRENs or with emerging NRENs. This model assumes is high with open and fair spectrum access to make that the CSP provisions networking infrastructure all the such access affordable at both the individual and way to the institutional campus. The CSP provides mainly institutional levels; internet connectivity. The model has a major drawback iii. There is sufficient penetration of IXPs and local in that HEIs lack easy access to the abundant resources hosting of major databases along with low peering and collaborations available through NRENs. The costs; opportunities for taking advantage of economies of scale iv. All institutions can afford and sustain the technical are limited, making this model potentially more expensive. environment and human resources; and v. The income from pure REN traffic and other services is Model 2 is a transitional model as the NREN grows but sufficient to sustain the NREN. has not reached all institutions and therefore combines No single African country is anywhere close to meeting bandwidth aggregation for those not connected, working these pre-conditions yet. While the existence of Model 4 is with CSPs, and provides all connectivity services to those recognized, it is not included in the cost projections. connected to the NREN. NRENs at this stage are normally focused on connectivity services, support for technical capacity building among member institutions, and Computing Bandwidth Required and Unit Prices improvement of campus networks. To estimate bandwidth requirements for each country, forecast student enrollment is multiplied by the As NRENs grow out of model 2 to model 3, they offer an progressive targets (see Table 10). While the bandwidth increasing range of services to members, increasing the requirement can also be estimated based on the number value proposition. of institutions in a given country, it is difficult to predict the change in the number of institutions while the number It should be noted that in models 1, 2, and 3, NRENs and of students is expected to increase exponentially over the RRENs do not make any distinction between traffic next decade. As such, although the approach is presented exchanged among users within the closed REN and traffic in the supplementary Excel model, this report will focus from such users to customers connected to commercial only on the student enrollment-based estimation. networks. This enables each institution to connect to an NREN for all types of traffic. Based on this, Table 15 shows the projected bandwidth requirements for the African continent on a regional basis using student enrollment, giving a total of 43.3 Tbps by 2025 and 506.8 Tbps by 2030. 47 A general, commercially available connection to the “regular” internet as opposed to a special-purpose restricted network such as Internet2 or the US military's NIPRNet or some other specialized backbone network. Generally, a commodity internet connection offers no content, application protocol, or destination restrictions or quality- of-service controls; see, for example, https://serverfault. com/questions/206557/what-is-commodity-internet. 32 Table 15: Projected bandwidth by region using student enrollment (2021, 2025, & 2030) Projected higher education students Projected bandwidth requirements Region (thousands) (Gbps) 2021 2025 2030 2021 2025 2030 Eastern and Southern 7,470 8,716 10,314 7,470 17,432 206,281 Africa Northern Africa 6,314 6,897 7,607 6,314 13,795 152,147 Western and Central 4,957 6,046 7,416 4,957 12,091 148,326 Africa Total 18,741 21,659 25,337 18,741 43,318 506,754 Source: KCL calculations Figure 9 shows two ways to determine the unit price (USD/Mbps/month). The unit price of bandwidth varies widely depending on the distance from the fiber network, local access and transit costs, the maturity of the NREN, the national ICT situation, and regulatory score. The local price comprises the cheapest cost of IP transit and the cheapest cost of local access to deliver the bandwidth in a metro area within a country. IP transit is calculated based on 10 GigE volume or more from the cheapest provider in the country. Local metro access costs to deliver bandwidth to HEIs are calculated based on gigabit Ethernet (GigE) circuits where available and smaller circuits in locations without big capacities, assuming that HEIs are located at most 15 km from a provider’s PoP in a metro/urban area. There are a number of countries for which there are no such data. Consequently, average costs were taken at the regional level and applied to all the countries according to location (see Table 16—Local Price). Instead of using the cheapest provider prices in-country, procurement negotiations that use the cheapest rates from within the region/vicinity can alter the upstream connectivity cost dramatically, and this is the Regional Price (see Table 16—Regional Price). Figure 9: Matrix for determining bandwidth cost Source: KCL 33 Table 16: Variation in unit cost price for bandwidth by region Local Price Regional Price (Average rate derived from IP transit (Average rate derived from IP transit and and local metro access rates of local metro access rates of cheapest provider cheapest provider in country) in region) IP transit Metro access Total (USD/ IP transit Metro access Total (USD/Mbps/ (USD/Mbps/ (USD/Mbps/ Mbps/ (USD/Mbps/ (USD/Mbps/ month) month) Month) month) month) Month) Eastern and Southern 48.8 3.8 52.5 10.5 3.8 14.3 Africa Northern Africa 49.5 4.3 53.8 15.9 4.3 20.2 Western and Central 50.3 1.9 52.2 2.4 1.9 4.3 Africa Source: KCL calculations Impact of Aggregation and Total Cost of Connecting Campuses Upstream Aggregation refers to pooling demand before procurement in order not only to exploit economies of scale but to bring other factors into play, including market maturity (linked to national ICT indicators) and NREN maturity (linked to the capacity and competence of the NREN to connect and deliver services and the cohesiveness of the in-country NREN community). This approach is informed by the experience of the UA and NRENs in the Alliance’s region working with GÉANT, where a combination of procurement based on capitalized long-term leases, smart procurement using a negotiated procedure (suitable for markets that are not yet mature), global and regional benchmarking as a basis for successive rounds of price negotiation, and regional demand aggregation (for the Alliance) as well as national demand aggregation (for the member NRENs) had the following two key outcomes: i. The quotations in the first round of bidding went down by factors of more than 100 by the end of procurement. Figure 10 shows the drop in the best price for a protected 2xSTM-1 10-year IRU between Nairobi and Kigali, with a drop in Kampala. This very sharp drop was an outcome of the procurement process.48 100000 10000 1000 NOV - 11 100 SEPT-12 10 1 Best Average Figure 10: Price drop (USD) for 2xSTM-1 10-year IRU between Nairobi and Kigali with drop in ii. The procurement, along with increasing sector liberalization in Eastern and Southern Africa, contributed to a major trend of market prices coming down closer to cost-related pricing. Aggregation as well as the method of procurement are both critical in ensuring the best prices for the HEIs. Table 17 compares the total bandwidth costs for all African countries based on student enrollment using both local and regional prices discussed in section 4.4.2. The projected high impact of aggregation on cost is evident in the savings. 48 Cathrin Stöver, AfricaConnect Update, euroafrica-ict, Nov 2012, Lisbon 34 Table 17: Total bandwidth cost for all African countries by region using student enrollment (2025) Total (USD, millions) Total (USD, millions) Region Using Local Prices Using Regional Prices % Savings Total Total From Total Total No No Aggregation With Aggregation With Aggregation Aggregation Aggregation Eastern and Southern Africa 3,790 1,484 1,381 541 60.8% Northern Africa 5,386 1,899 1,929 680 64.7% Western and Central Africa 5,143 1,973 614 235 61.7% Total 14,319 5,356 3,924 1,456 62.4% Source: KCL calculations Table 18 provides a summary of the total costs for bandwidth indicated in Table 17 for five years starting with the first year of implementation, the initial assumed period being 2021–2025 inclusive. It is also assumed that smart procurement strategies (e.g., benchmarking regional pricing) combined with the procurement of long-term leases will be used to secure the best price advantage. Table 18: Summary of total cost of connecting all African HEIs to upstream bandwidth for five years Cost (USD, millions) Category rounded to nearest 10 With Aggregation No Aggregation Using Student Enrollment & Local Prices 26,780 71,595 Using Student Enrollment & Regional Prices 7,280 19,620 Source: KCL calculations Strengthening and Sustaining NRENs Creating, strengthening, and sustaining NRENs is important to ensure that demand is aggregated and that HEIs have access to shared capacity building initiatives and global research and education resources. The CapEx of NRENs includes the initial network design cost, the cost for the purchase and upgrade of optical switching equipment, and the cost of connecting to university campus network nodes, as summarized in Table 19. Where NRENs are used as the basis for aggregating bandwidth demand for HEIs, investment made in long-term leases for bandwidth will provide NRENs with the necessary resources for CapEx. Table 19: CapEx and OpEx elements for NREN development CapEx Elements OpEx Elements Initial network design cost Training and skills Cost of connectivity lease, IRU or trade, internet connection Staffing, management, oversight, and governance Initial optical and switching equipment Network O&M Engineering and contracts Communication and outreach Physical installation and inter-campus connection Network services Overheads Source: KCL 35 NREN OpEx includes staffing, management and oversight expense, training and skills development both for NREN managers and member universities, network operations and management, and communication and outreach. Discussions with NREN CEOs in Africa indicate that they spend about 60% of their OpEx on connectivity-related expenses (e.g., network services, network O&M) and the remaining 40% on human resources and related costs. While member institutions cover connectivity-related expenses through payments for bandwidth, NRENs often face major challenges in trying to cover core costs as well as costs related to ongoing capacity building for both internal staff and especially member institutions. Supporting NRENs until both their income combined with ongoing government support can sustain them is therefore a critical part of the connectivity initiative. Support of USD 1 million covering core costs for five years was assumed for each mature NREN, increasing USD 1.5, 2, and 2.5 million, respectively, for each connected NREN, emerging NREN, and no NREN country. The list of countries and levels of support are presented in Table 20. Table 20: Support for Core Costs of African NRENs State of NREN development Countries Proposed support per year USD 2.5 million per country, USD Angola, Cape Verde, Central African 12.5 million per country over 5 Republic, Comoros, Republic of Congo, years: Eritrea, Equatorial Guinea, Eswatini, No NREN Gambia, Guinea-Bissau, Lesotho, Mauritius, São Tomé and Príncipe, Seychelles, South Sudan (15) USD 187.5 million Botswana, Burkina Faso, Burundi, USD 2 million per country, USD 10 Cameroon, Chad, Djibouti, Guinea, million per country over 5 years: Emerging NREN Liberia, Libya, Mali, Mauritania, Namibia, Niger, Sierra Leone, Sudan, Zimbabwe (16) USD 160 million Algeria, Benin, Cote d’Ivoire, DRC, Egypt, USD 1.5 million per country, USD 7.5 Ethiopia, Gabon, Ghana, Madagascar, million per country over 5 years: Connected NREN Malawi, Morocco, Mozambique, Nigeria, Rwanda, Senegal, Somalia, Tanzania, Togo Tunisia, Zambia (20) USD 150 million USD 1 million per country, USD 5 Mature NREN Kenya, South Africa, Uganda (3) million per country over 5 years: USD 15 million Source: KCL calculations In aggregate, countries need about USD 513 million to accelerate NREN growth and promote network capacity building and training within each African country over the next five years. 36 Strengthening and Sustaining Regional RENs The three regional RENs, WACREN, UA, and ASREN, aggregate national traffic and connect to international networks. The experience of the UA, summarized in Table 21, shows that 60% of the REN budget is spent on connectivity-related expenses, with about 40% spent on core costs (e.g., human resources and the promotion of coordination among national research and education networks). Table 21: Cost drivers for UA budget Regional Research and Education Network Cost Items (2020) CapEx investment in network costs (USD 1.1 million) UA49 Connectivity and network operation cost (USD 1.3 million) Human resources costs (USD 0.6 million) Number of countries connected: 10 Coordination and other costs (USD 1 million) Total = USD 4 million Source: UA50 In Africa, RREN transmission costs are expected to be covered by member NRENs and through international support such as the EU-funded AfricaConnect project; however, the three RRENs do need resources to continue promoting NREN development as well as training and capacity building—for member NRENs and the RRENs themselves. Assumptions: i. RRENs spend USD 25,000 per country for NREN development over the next five years. This includes sensitization workshops, short term consulting, and NREN business plan design. ii. RRENs spend USD 25,000 per country for training and promotion of direct engineering assistance (DEA) over the next five years. iii. RRENs spend USD 700,000 each, annually, to support the critical mass of human resources, including ICT engineers and application developers as well as communication and financial specialists for the next five years. Table 22: Support for Core Costs of African RRENs RREN UA WACREN ASREN Number of countries with membership region 24 22 10 Number of institutions 4,094 2,329 977 NREN development and promotion costs (USD 3,000,000 2,750,000 1,250,000 millions) Training and capacity building 3,000,000 2,750,000 1,250,000 (USD millions) Human resource costs (USD millions) 3,500,000 3,500,000 3,500,000 Total (USD millions) 9,500,000 9,000,000 6,000,000 Source: KCL calculations In aggregate, RRENs need about USD 25 million to accelerate NREN growth and promote network capacity building and training over the next five years. 49 UbuntuNet Alliance members, https://ubuntunet.net/about/council-of-members/ 50 Interviews with UbuntuNet Alliance management team 37 Determine Characteristics of Higher Forecast Staff Forecast Student Education Institutions Numbers Enrolment Numbers of Higher Education Institutions Historical Staff Numbers Data Historical Student Enrolment Data by Country by Country by Country Type of Higher Education Institutions Projected Student-Staff Ratio Popn. Est. for 20-24 year by Country by Country Age Cagtegory by Country Campus Size of Higher Education Institutions Projected GER for Higher Education by Country by Country Projected Staff Numbers Projected Projected Student Student Enrolment Projected Staff Numbers Enrolment (2021-2030) by Country Enrolment (2021-2030) by (2021-2030) by Country (2021-2030) Country byCountry Calculate Cost of Upgrading Campus Networks Calculate Cost of Access Devices eam Bandwidth Calculate Cost of Upstre Calculate ICT Indicator Calculate NREN Maturity Score Existence of NREN CapEx OpEx Cost of Staff Devices Cost of Student Devices Scenario 1 Scenario 2 Landlocked/Sea-Access Proportion of First -Year NREN Governance Structure Quantity of Network Average # of Students Proportion of Staff subsidised Forecast of student N Number/size of campuses Students Landing Stations (number) Equipment per Campus Size (% of total number) enrolment by country (small, medium & large) (% of enrolment) Government Recognition IXP Ladder Stage Unit Cost of Network # of Students per ICT Subsidy for Staff Devices Progressive bandwidth M Minimum connection port Subsidy for Student Devices NREN Funding Sources Equipment Skilled Professional (% of device+software cost) targets per student s size for each campus size (% of device+software cost) Fibre Coverage (% of popn 10km radius) NREN Physical Network Size of Campus Monthly Salary of ICT Unit Cost of Staff Device Higher Education GER by (small, medium, large) Professional ITU Regulatory Score Regional/Global Connectivity Wealth Quintiles per Region 38 Annual ICT Trainings Local Price of Bandwidth gional Price of Bandwidth Reg Weight attrbuted to each ICT Middle-ware Services Unit Cost of Staff Software Proportion of Device Support per ICT Professional Indicator (% of total score) for each Wealth Quintile Price of Cheapest IP Transit Price of Cheapest IP Transit Advanced Services Cost per ICT Training Provider in Country Provider in Region Unit Cost of Student Device Price of Cheapest Metro Price of Cheapest Metro Maintenance as Unit Cost of Student Access Provider in Country A Access Provider in Region ICT Indicator Score Percentage of Capex Software by Country NREN Maturity Level Number of Years to by Country Calculate Opex avings Aggregation Sa Weight attributed to Weight attributed to Calculate Core Support Core Support for 3 RRENs for NREN NREN Status National ICT Indicators (ASREN, UbuntuNet Alliance & WACREN) Assigned Score per NREN Level Support per NREN Level B Cost of Upstream Bandwidth Cost of Upgrading Campus Cost of End-user Access Devices Student Enrolment & Student Enrolment & Core Support for RRENs/NRENs Networks Local Price Regional Price t High-Speed Internet Four Major Cost Components of Connecting All African Higher Education Institutions to Figure 11 Connecting All African Higher Education Institutions to High–speed Internet Cost of Connecting African Higher Education A schematic diagram for arriving at the cost of connecting HEIs is shown in Figure 11. Figure 12 sums up the cost elements for the different components that make up the total cost of connecting all HEIs in Africa. These include the cost of equipping students and staff with access devices, designing and upgrading campus networks and upstream bandwidth, and providing support to RRENs/NRENs. The total costs are for five years starting with the first year of implementation, the initial assumed period being 2021 – 2025 inclusive. It is also assumed that the procurement of long-term leases combined with demand aggregation will be used to secure the best price advantages. Potential sources of funding include governments, development partners, students, and HEIs. Actual proportions will vary across countries depending on government funding priorities, development partner funding guidelines, and the means of different stakeholders. Cost of Di erent Components to Connect all African Higher Education Institutions to Potential Sources High-Speed Internet (USD, billions) of Funding 0.03 RREN core support and RREN Development partners, development related costs government 0.51 NREN core support and Development partners, NREN development related costs government 7.28 Cost of connecting campuses upstream Development partners, (Student Enrolment & Regional Price) institutions, students OpEx for maintaining Institutions, government, 15.51 campus networks development partners CapEx for upgrading Government, development 11.75 campus networks partners Students and Sta end-user Government, development 18.77 devices partners, students, institutions Figure 12: Summary of total cost (USD, billions) of connecting all African higher education institutions to high-speed internet Source: KCL calculations It is evident from the above that African countries need about USD 52 billion to connect all HEIs to high-speed internet. This cost can be reduced further if bandwidth prices are reduced to reasonable levels of USD 10 per Mbps/month or less. 39 5 Leveraging Broadband Connectivity for Improved Learning Outcomes W hile getting connected to the internet is vii. Limited individual access crucial, it requires a much broader effort to viii. Limited digital learning resources shift teaching practices to approaches that All these have to be addressed if the opportunities of enhance students’ learning ability to meet broadband connectivity are to be exploited for improved the new demands of the job market. Learning requires learning and research outcomes. students that are ready to learn, effective teaching that is supported by inputs such as digital technologies, and skilled higher education management that pulls Deficient ICT Policy and Strategy52 everything together.51 A major challenge in many HEIs is the absence of, or While curriculum, pedagogy, and assessment are key deficiencies in the policies that provide for the strategic elements of the solution, the focus here is therefore on vision and plan for application of ICT in learning, research, the key ICT-related impediments to the integration of and administrative effectiveness, often because ICT is technology into learning and research and recommending considered a technology add-on to other policies rather a roadmap for addressing them. than a priority strategic intervention. A good institutional ICT policy should do the following: The integration of technology in higher education i. Define “the why” and objectives of ICT with respect to enable better learning outcomes and academic to the HEI’s mission of learning, research, and excellence, foster research and innovation, and achieve community outreach, which must also be supported greater operational efficiency needs to be accompanied by administrative effectiveness. The policy and by a digital technology integration vision, policies and strategy should be guided by the strategic priorities of strategies, a change in processes, and a shift of the the university or TVET institution. mindsets of people—such as students, administrators, ii. Define the institutional positioning and organizational and faculty. A well-staffed corporate ICT department, arrangements related to ICT in all aspects of the HEI’s with highly skilled engineers and a user support team, is activities. It is particularly important to note that the as critical as is the presence of technology-savvy teachers direction of ICT in the organization is: and administrators that facilitate students’ success in digital technology-enriched learning environments. The a) Guided by the aspirations of the business units heads of the HEIs must lead the way in the integration (learning, research, library, management, and of digital technologies in the instructional, research, and administration), not by ICT personnel. administrative realms. b) Requires consensus among stakeholders to There are several well-known impediments to the minimize the likelihood of system failure even integration of ICT in support of learning, research, when technical success is achieved. and effective administration in HEIs. These include the iii. Define high-level priorities for the key information following: systems and corporate databases, infrastructure i. The absence of, or deficient, ICT policies and policy, security policy, and AUP. All these would be strategies, which often arises from lack of awareness expanded into planning and operational documents of the role of higher education connectivity at the implementation level. ii. Limited ICT awareness and ICT literacy among faculty iv. Define a transformation enterprise architecture and and administrators roadmap for higher education. iii. Limited competence of campus ICT personnel v. Address burning issues, such as foundational and iv. The poor quality of campus networks capability gaps. v. Poor digital learning spaces vi. Limitations in resource allocation and coordination 52 See, e.g., F. F. Tusubira and Nora Mulira, “Integration of ICT in Organisations: Challenges and Best Practice 51 World Bank, World Development Report 2018 (WDR Recommendations Based on the Experience of Makerere 2018)—LEARNING to Realize Education’s Promise, University and Other Organisations,” http://ahero.uwc.ac.za/ https://www.worldbank.org/en/publication/wdr2018. index.php?module=cshe&action=viewtitle&id=cshe_426. 40 vi. Define the ICT management, control, and i. Leveraging the continuous flow of engineering and maintenance unit, ensuring that it reports to the top other ICT students on campuses; these are always level of the HEI as its role is cross-cutting. looking for practical training opportunities as part of vii. Provide an implementation Master Plan and Budget their courses, are innovative, and offer free services that guides prioritization and integration in the HEI in exchange for training. They are also able to work as expenditure plans. interns at modest costs. viii. Provides for annual review and adaptability in ii. Sharing a pool of professionals so that cost is recognition of the rapid evolution of the opportunities distributed among different institutions and/or created by technology. outsourcing services, where it is more cost-effective to do that than having full-time staff in particular disciplines. Just as faculty need to understand and appreciate the Limited ICT Awareness and ICT Literacy importance of, and how to use, ICT in learning, research, among Faculty administration, and management, it is critical for ICT personnel to understand and appreciate the importance of In HEIs, faculty are always part of any key decision. While letting business owners lead in defining functionality. ICT the younger faculty are largely ICT literate, the decision personnel particularly need to understand enough about levels are still dominated by the older generation, who learning, research, library services, administration, and went through their schooling to postgraduate level other campus business processes to provide support that without experiencing ICT-enabled learning and research is empathetic and relevant. environments. This drives appreciation down, fuels resistance, and impedes key decisions around ICT that would lead to major changes in all aspects of achieving the HEI mission. Across higher education, the improvement of faculty skills in the latest tools and technologies and Poor Quality of Campus Networks ability in leveraging digital tools to conduct teaching, The GÉANT Association has, through the AfricaConnect assessment, and research should therefore be given high project and working with ISOC and NSRC (Box III), priority. funded a great deal of the DEA in the design of campus networks across Africa, bringing out the importance The shortage of skills is often acute in the computer of ongoing technical support and training in the latest sciences and other ICT-intensive fields. Digital skills are campus network design techniques. Further, drawing inherently practical, and staff who can teach by example on the European experience, there is a need for national and understand business applications for these skills can initiatives that facilitate the sharing of experience on- ensure their relevance to market needs. campus networking design and technology environments. Such initiatives should allow HEIs to organize workshops It is particularly important to develop and follow to share campus network experiences; set up working comprehensive change management strategies to groups around design, infrastructure, mobility, security, bring faculty on board in both formulating enterprise identity management, etc. issues; coordinate the purchase architecture and ICT policy and strategy and leading their of network equipment to benefit from economies of scale; own individual transformation. and develop a shared national campus best practices repository.53 Infrastructure is another important aspect that Limited Competence of Campus ICT determines the quality of campus network. Putting campus connectivity infrastructure, such as cable, copper, Personnel and fiber optics, in place presents a significant challenge because most of the buildings in higher education on the The development of campus ICT environments depends continent were designed for the traditional teaching and greatly on the availability of skills and resources. One learning environment. The alteration of these buildings of the biggest challenges to HEIs with respect to hiring, is often complicated—in some cases, the presence of developing, and retaining competent ICT human resources asbestos makes this hazardous for IT experts in wiring the in African HEIs is direct competition with the rapidly building. growing ICT private sector. With the exception of very large and well-endowed institutions, HEIs cannot hope to compete sustainably with the private sector for expert ICT human resources and will need to adopt different 53. GÉANT, Campus Best Practices, https://services. approaches, which include the following: geant.net/sites/cbp/Knowledge_Base/Reports/ Documents/geant-campus-best-practices.pdf. 41 Resource Allocation and Coordination Limited ICT Literacy and Individual Resource constraints, including the funding of HEIs, are Ownership of Laptops considered challenges to the under-resourced campus Limited ICT literacy among students and the limited networks. This is compounded where senior management individual ownership of laptops create a vicious cycle— and faculty are not familiar with the critical importance unfortunately, a combination of poverty and school of well-resourced, well-designed, and functioning campus systems where ICT literacy is not addressed as a key area networks to achieve organizational objectives. of learning mean that a major section of the student population needs remedial intervention as a first-year Most HEIs have small and fragmented ICT support requirement. Some HEIs have instituted remedial learning, departments. Consequently, there is limited coordination such as based on the International Computer Driving in providing resources and centralizing the design and License (ICDL)54 structure or other, but the success of this is implementation of the different information systems, constrained by the limited ownership of laptops. often with some departments acquiring donated equipment that may not be compatible with the network, It is evident that programs to ensure access to owned raising disputes between departments and central IT computers55 and address ICT literacy for first-year students management. must be part of the interventions aimed at addressing the remaining ICT-related challenges. Resource allocation and coordination challenges will need to be addressed through enterprise architecture that serves as a blueprint for higher education business processes, the information needed and supporting technologies, and ICT policy and strategy and change Limited Digital Learning Resources management. Related to limited skills among faculty is the shortage of digital learning resources. While there is a growing trend toward massive open online courses (MOOCs) and small private open courses (SPOCs), efforts to develop African Digitally-Enriched Learning Spaces MOOCs and SPOCs are yet to catch on.56,57 The MOOCs space is currently dominated by the big four—Coursera, Digital and physical learning space is increasingly edX, Futurelearn, and Udacity, but the proportion of becoming an important aspect of student-centered African students signing up for these is expected to be low, learning. Higher education spaces are typically designed particularly if studying for certificates, because individual for traditional teaching methods and therefore need students must meet their own costs. While efforts have to be upgraded to meet the growing blended learning started,58,59 these are still limited to South Africa60 and requirement. African higher education needs to develop Northern African countries such as Egypt, Morocco and and adapt its physical learning spaces to facilitate the Tunisia. integration of technology into the classroom (e.g., through the use of smart classrooms), campus, and The development of Study Webs of Active Learning for residences and also to ensure that virtual learning spaces Young Aspiring Minds (SWAYAM)61 in India provides a are safe. great example of how such a MOOC can be launched and maintained with access, equity, and quality in mind. The ministries of higher education and those responsible for infrastructure need to develop new building codes that help create modern institutions with blended learning. The physical and architectural design of new educational institutions should take the need for interactive classrooms and digital-enabled learning into 54 https://icdlafrica.org. consideration. 55 Refer to Report 2: Cost Estimates for Connecting All African HEIs. 56 MOOCs in Africa, https://blogs.worldbank. To ensure the safety of the virtual environment, org/edutech/moocs-in-africa universities and TVETs need to develop risk-based 57 https://en.unesco.org/news/unesco- security strategies that keep pace with security threats supports-open-moocs-africa and challenges and ensure adherence to acceptable 58 https://trueafrica.co/lists/e-learning-platforms- technology use policies by students and faculty. This will africa-tutor-ng-mest-school-education-startups/ increase safe learning and caring in increasingly digital 59 https://www.atingi.org/en/tool 60 https://www.news.uct.ac.za/article/-2020-04- environments. 29-massive-uptake-in-mooc-participation 61 Majumder, C., 2019. SWAYAM: The Dream Initiative of India and its uses in Education. 42 Strategy and Roadmap for Change Formation of National-level Leadership and Coalitions Change at the infrastructure/technology level is easy, but the changes required will not be sustainable unless Policy is vested in the ministries responsible for higher accompanied by changes in people and processes. education, and that is a necessary starting point for Intervention should also be carried out at the regional, establishing national leadership. The ministry will, in national, and institutional levels. The following four most cases, have direct control of policy for TVETs with recommended strategic interventions provide the respect to pedagogy but will not have direct control of framework under which detailed activities would be such policy in universities. The coalition should include undertaken: the ministries responsible for ICT, science, and industry; associations of vice-chancellors, presidents, and rectors; i. Establishing and sustaining regional partnerships the national bodies responsible for HEI quality, standards, with other agencies and organizations that are and accreditation; HEI academic and business leaders and interested and active in digital technologies for the ICT directors of HEIs; the NREN; and the ICT service improving learning outcomes and employability in providers. Africa; ii. Identifying leadership and catalyzing the formation The formation of such coalitions should be followed of national-level coalitions that will be responsible for by a dialogue among these players at the national spearheading change in the integration of technology level on the integration of technology in education. A in higher education; nationwide higher education ICT integration master iii. Guiding national- and institutional-level approaches plan that is sponsored by the relevant ministry can be through toolkits; and a starting point for such discussion. The plan could, iv. Supporting specific quick-win demonstration projects. among others, outline strategies and steps to upgrade These are expanded on in the following sections. campus networks, improve access to devices by students and faculty, enhance the development and exchange of learning content, promote the digital skills of faculty Establishing and Sustaining Partnerships across the country, and provide approaches to funding and sustainability. The success of the recommended interventions will require significant funding as well as expertise in the relevant Guiding National and Institutional-level Approaches areas of improvement or transformation. Partnerships will bring together key agencies and organizations that bring for ICT Integration in Education on board the following: Many of the key changes required to deal with the ICT- i. Funding (for example, European Commission, AfDB, related gaps are not context specific and can therefore bilateral and multilateral development agencies, be guided by toolkits that provide menus of choices and private charities, the private sector, NGOs, and ICT also point to resources. Toolkits and/or guidelines can be industry both within and outside Africa); developed for the following specific areas: ii. Policy leadership (for example, the African Union, the AAU, regional groupings of universities in the different i. Institutional enterprise architecture and ICT policy regional communities, RRENs, and NRENs); and and strategy development and implementation iii. Expertise (for example, communities of practice, ii. The roles, membership, and structure of the National learning associations, RRENs and NRENs, NSRC, and Level Leadership Team ISOC). iii. Laptop ownership strategies that ensure sustainable interventions The partnerships create resources that can work together iv. Sustainability strategies for major initiatives to create synergy and generate push for the integration of v. Campus network design principles and blueprints ICT in higher education in the region. The PHEA62, which vi. Effective methods for building faculty digital skills63 brought together seven private foundations for 10 years vii. The development of MOOCs and SPOCs (2000–2010), could be examined for experiential lessons viii. The development of digital technology integration about working toward a common goal through joint self-assessment tools. funding. While it might be harder for public funders to do the same, an integrated intervention plan would create the required synergy. 62 See Report on a Decade of Collaborative Foundation Investment, https://www.fordfoundation.org/ media/1760/2010-accomplishments-of-the- 63 See, for example, the all aboard initiative in partnership-for-higher-education-in-africa.pdf. Ireland, https://www.allaboardhe.ie/. 43 Support for Specific High-impact Interventions and b) Designing and implementing a robust campus Quick-win Demonstrators digital infrastructure; In addition to any major systemic interventions, there are c) The design of digital services with various opportunities for specific interventions that would consideration for accessibility to all devices accelerate the pace of change or support countries and (mobile phones, tablets, computers, and future institutions that come up with quick-win proposals. The devices); following attributes should be satisfied by each: d) Evaluating physical spaces in relation to i. Should be in one of the key areas where change is changing pedagogical models, accessibility required; needs, and emerging instructional ii. Should be bedded in a supportive policy and strategy technologies and fostering innovation in environment; learning spaces; iii. Should bring on board a good number of the key stakeholders around the initiative; e) Staffing IT operation departments and iv. Should be based on a strong theory of change with motivating and retaining engineers and a measurable outcome indicators; and support team; v. Should have significant internal funding to show ownership along with a realistic sustainability plan. f) Developing, implementing, adapting, and Possible high-impact interventions and demonstrators sharing lessons on AUPs; and include the following: g) Sharing ICT expertise with other institutions. i. Working with regional RENs, the identification of universities in Africa and other developing regions, as well as those than can provide good examples in more advanced countries, as an aide to peer-learning and twinning in creating digitalized campuses. Several Implementation African NRENs were able to develop through peer- learning and twinning arrangements, and this is an Leveraging connectivity, as outlined in this section, approach that can produce visible high impact in a involves major initiatives, with multiple elements and short period of time. inter-relationships. The recommendations, therefore, ii. Working with regional university associations, need to be examined in considerable depth as part of MOOCs and SPOCs providers rolling out implementation planning. Everything proposed above comprehensive ICT-enabled learning in selected pilot is doable and achievable within clear time-frames, courses across selected locations as demonstrators depending on policy-level commitment and the availability and program learning opportunities. of sufficient funding. iii. Working with NRENs in developing national blueprints for digital campuses, and leading their The impediments discussed above provide the starting- rollout and implementation, along with capacity state of digitalization in higher education. The end- building for campus ICT staff; developing/updating state would be fully digitalized campuses where all and rolling out training programs to develop categories of users, both on and off campus, can exploit awareness among HEI governance, management, and the opportunities of digital technologies for their work— faculty levels and skills for faculty, administrative staff, with competent ICT support staff enabled by sustainable and students; and supporting the development and funding to maintain the resources in optimum working rollout of shared skills and resources, including human condition. The rapid evolution of technology, however, expertise. means that change and improvement are open-ended iv. While regional- and national-level initiatives processes. While funding from external sources remains are essential, real progress can only be made at critical for the achievement of progress for up to five years institutional levels. This leads to the support of or more and then phasing out over another five years, selected HEIs to measure progress and facilitate consistent funding from HEIs and national governments knowledge exchange on how best to create digitalized should be available to ensure that national economies higher education environments with a focus on the continue benefiting from the returns from digitalized HEI following: campuses. a) The development/updating of institutional ICT policies to ensure digital technologies are fully integrated into teaching, learning, and administration; 44 6 Looking Ahead The following key considerations are important when exchange or minimizing the licenses required for looking ahead at implementation: multi-country fiber projects. i. First, there is a need for increasing awareness among There are a number of policies that can help to improve the decision makers about the different connectivity availability and sufficiency of national backbone networks issues and challenges. This is especially important by increasing competition and reducing prices through for those in the ministries of education, ministries increased investment. These include the following: responsible for digital technologies, ministries i. Eliminating monopoly provisions from the market responsible for finance and investment, and HEI structure; leaders. An awareness of the magnitude of the ii. Reducing the cost of operator and spectrum licenses, challenges (e.g., the need for access to computing hence the barriers to entry and costs to the end-user; devices, the importance of the upgrading of campus iii. Enforcing the shared use of telecom infrastructure, networks, NREN development) is critical to ensure civil-works, and access to the alternative that adequate resources are available for higher infrastructure provided by transport and energy education connectivity. operators; ii. Second, there is a need for accelerating connectivity iv. Legislating for the protection of critical infrastructure, to facilitate teaching, research, learning, and including ensuring sufficient compensation for fiber administration in higher education in order to improve cuts; learning outcomes. v. Eliminating or reducing taxes on communication and iii. Third, connectivity should be accompanied by the communication equipment, which will reduce end- digitalization of campuses, building ICT literacy user costs and drive-up demand; and among staff and students and enabling the individual vi. Deploying universal access funds that have provided ownership of laptops to support reforms in learning funds for infrastructure in remote and sparsely and new ways of teaching digital and soft skills to populated areas of the country. meet the demands of 21st-century jobs. iv. Fourth, coordination is an essential prerequisite for The development of the campus ICT environment depends achieving higher education connectivity and attaining mainly on the availability of skills and resources. In reforms in learning and digital skills. addition to increasing higher education leaders’ awareness of the importance of the digital environment for improved Government and development partners need to work teaching, learning, and research, there is a need for: together to push the connectivity, learning, and digital- skills agenda forward, which demands i. Designing campus networks for changing learning and teaching requirements, including advanced i. Engaging stakeholders; research, innovation, and digital entrepreneurship; ii. Cost sharing at all levels; and ii. Improving the skills of the technical teams in HEIs iii. Adhering to and implementing a timeline for through DEA; connecting HEIs in Africa. iii. Investing in engineers and IT staff to maintain and Key required policy interventions to address the availability grow the campus network facilities and services; and sufficiency of regional connectivity include iv. Investing in campus-wide digital infrastructure, in particular local area networks, network monitoring i. Harmonizing ICT policy and regulation across and management systems, identity management neighboring countries or within regional blocks. systems, wireless services, data storage, and high- Some of these can be addressed through the now- performance computing; and operational African Continent Free Trade Agreement v. Developing AUPs, which should be communicated (AfCFTA). to end-users to make them aware of the services ii. Creating incentives to attract investment in inland available to them and should state the end-users’ data centers, which will attract major carriers to roll rights and obligations regarding how they use those out more inland capacity and facilitate more exchange services. of local traffic. iii. Minimizing barriers for cross-border infrastructure, such as by eliminating the need for national licenses to exchange traffic at IXPs to enable regional traffic 45 Engaging Stakeholders While the specific approach to each stakeholder is different, there must be common key messages. The following are recommended as key messages: i. Ensuring that all African HEIs are connected to broadband will lead to global economic benefit; ii. Getting connectivity and individual laptop access to all students and staff in African HEIs will require USD 52 billion over the next five years; iii. There is a need for coordination among development agencies to create synergy; and iv. African countries must take ownership and make contributions to the initiative. Table 23 lists the key stakeholders that need to be brought on board, and their potential roles. The stakeholders may be beneficiaries—the primary target of the interventions or implementers—possess the technical know-how to implement ICT in HEIs at the campus, national, and regional levels. The roles entail a range of different activities, which include the following: i. Planning, policy, and regulation—positioned to ensure both policy and regulation as well as the integration of access to broadband as key elements of national and regional development initiatives. ii. Funding—positioned to directly fund and, because of the nature of the organization, to be able to mobilize additional funding from other partners. iii. Knowledge—possessing documented experience, research findings, best practices, and toolkits that can guide all the different roles on different aspects of ICT integration in higher education. iv. Capacity building—provide direct training opportunities, opportunities for bilateral collaboration or twinning, and attachments and secondments for staff. v. Content—positioned to provide or guide the development and distribution of learning content for HEIs. vi. Publicity—positioned to disseminate the key messages effectively as both a mobilization and partnership sustainability tool. 46 Table 23: List of key stakeholders Planning, Policy, Capacity Organization/ Country/Region Implementer Beneficiary Funding Knowledge Content Publicity and Regulation Building AU, AUC, regional economic communities (RECs), UNECA, 1 X X X X X African countries Ministers responsible for higher education, ICT, Finance and 2 X X X X Economic Development and national planning Agencies AAU and Regional University Associations (IUCEA, SARUA, 3 X X X X AWAU) 4 Staff and Student Associations X X X 5 African RRENs and NRENs X X X X X X African national regulatory agencies and regional regulatory 6 X X X organizations (CRASA, EACO, WATRA) 7 The World Bank X X X X 8 European Union X X 9 AfDB X X X X Other multilateral agencies (e.g., UNESCO, ITU, UNDP, 10 X X X X X UNCTAD) Major private foundations (e.g., PHEA, Bill & Melinda Gates, 11 X X X 47 MasterCard Foundation) Bilateral donors (e.g., USAID, Sida, SDC, GIZ, IDRC/CIDA, 12 X X X DANIDA, FDCO/UK, JICA, KOICA) Advanced RRENs—APAN, GÉANT Association, RedCLARA, 13 X X X X Internet2, TEIN 14 Agencies from new economic power centers X Major connectivity and equipment providers (e.g., WIOCC, 15 Seacom, Hurricane, Liquid Telecom, WACS, Huawei, Cisco, X X Juniper) Multi-national providers operating on the African continent— 16 X X Bharti Airtel, MTN Group, Orange Major content providers and content delivery networks (e.g., 17 X X X X Akamai, Amazon, Cloudflare, Google, Facebook, Microsoft) 18 Major training and knowledge providers (NSRC, INASP) X X X X Major policymaking and civil society organizations working on connectivity-related issues—World Economic Forum 19 (WEF), New Partnership for Africa's Development (NEPAD), X X United Nations ICT Task Force, Association for Progressive Communications (APC), Alliance for Affordable Internet (A4AI) 20 Major public and private online learning content providers X 21 Media houses and journalists X Cost-sharing Development experience has shown that sustainability requires ownership of the initiative by beneficiaries, which means having a say in the formulation and planning as well as implementation and that they need to make a contribution. An approach akin to what EUC has used for AfricaConnect is therefore recommended—beneficiaries (in this case RREN) that are still emerging contribute 20%, increasing to higher percentages as maturity increases. The actual starting percentage and how it grows will call for negotiations between funders and beneficiaries, and this would increase in steps at predefined levels of maturity. An associated and important consideration is that beneficiaries should be prepared to meet recurring costs that, as a minimum, would be about 15% of capital investment. It is recommended that the cost of long-term leases as well as the annual O&M associated with IRUs are paid upfront for the duration of the leases and therefore capitalized upon, as has been done under the AfricaConnect project. Implementation Timeline Table 24 summarizes the timeline for reaching different levels of individual access, broadband, and NREN maturity. Projections beyond 2025 are not given as the rapid evolution of technology, combined with the current suppressed demand for higher education in African countries, makes projections of required connectivity and costs beyond five years unrealistic. Parallel to these activities and also starting from the first year is developing the necessary pre-conditions in HEIs to ensure that improved connectivity translates to improved learning outcomes, a national-level activity around educational policy and regulation. Other than the right campus network environment, which is included in the cost projections, the broad range of soft activities required to achieve this are not part of the cost projections. Table 24: Prioritizing connectivity to African higher education Timeline/ priorities/ Year 1 Year 2 Year 3 Year 4 Year 5 Individual Individual Individual Individual Individual ownership ownership for ownership for ownership for ownership for Devices for all students and all first-year all first-year all first-year all first-year staff students students students students Nationwide support for evolution of a Campus network Build functioning campus networks across HEIs world-class campus network “All institutions progress to a goal of 2 Gbps per 1,000 “Start work toward a goal of 20 Gbps per Connectivity students” 1,000 students” Aggressive NREN development support to all countries; special Intensive support to upgrade countries to the NREN NREN development focus on countries without Exclusive model sustainable NRENs RREN Each African country is connected though at least one regional REN, and the RRENs are interconnections interconnected at multiple points throughout the continent Developing pre-conditions Enabling policy environments related to the use of ICT in teaching, learning, and research; for leveraging curriculum; pedagogy; assessment; quality management; and industry linkages developed and connectivity for implemented in HEIs along with the necessary institutional arrangements improved learning outcomes Source: KCL 48 Ultimately, connectivity for higher education rests on the national governments. Each country needs a concerted national effort to expand connectivity, accelerate online learning, improve campus networks, and promote NREN development. Each country needs to assemble a high-level team drawn from the ministries responsible for higher education, the ICT sector, and finance; HEIs; NREN, where present; the ICT private sector; key development partners; and other stakeholders in order to develop a national plan for connecting its HEIs and accelerate learning and digital skills for the jobs of the future. The key strategic interventions to move this plan ahead include: For African For Development For Leaders of HEIs: For Private sector: Governments: Partners: • Support the • Recognize • Support the • Promote development of the specific development of information NRENs connectivity comprehensive exchange among • Allocate resources requirements of higher education ICT countries on the to ensure sufficient HEIs and engage policies that address different issues of bandwidth is in dialogue device, connectivity, connectivity available for higher with NRENs campus networking, • Support projects education to optimize and capacity issues that enhance the • Sponsor projects service offerings, • Support the connectivity of that leverage including through development of higher education connectivity service level NRENs and those that to enhance agreements • Promote access to leverage advanced teaching, learning, • Consider offering devices through networks to and research preferential rates national negotiation solve social collaboration or tailored service with suppliers and and economic • Serve as advocates offerings for HEIs, industry players challenges. for higher education in cooperation • Allocate resources connectivity. with REN. for higher education connectivity through universal access funds and donor funding. 49 50 Appendix A Table 25: Maturity of NRENs in different countries Has government Has governance Has variety of Has physical Regional/global Regional REN Country NREN name recognition/ Maturity stage structure funding sources network connectivity relationships Botswana BotsREN No No No No No Emerging NREN Burundi BERNET No Yes No No Yes Emerging NREN Congo, Democratic Eb@le Yes Yes Yes No Yes Emerging NREN Republic of Ethiopia EtherNet Yes Yes Yes Yes Yes Connected NREN Kenya KENET Yes Yes 3 Yes Yes Mature NREN Madagascar iRENALA Yes Yes Yes No No Connected NREN UA Malawi MAREN Yes Yes Yes Yes Yes Connected NREN Mozambique MoRENet Yes Yes 2 Yes Yes Connected NREN Eastern and Namibia Xnet No Emerging NREN Southern Africa Rwanda RwEdNet Yes Yes Yes Yes Yes Connected NREN Somalia SomaliREN Yes Yes 3 No Yes Connected NREN South Africa TENET Yes Yes 3 Yes Yes Mature NREN Sudan SudREN No No Emerging NREN Tanzania TERNET Yes No 2 Yes Yes Connected NREN Uganda RENU Yes Yes 2 Yes Yes Mature NREN Zambia ZAMREN Yes Yes 3 Yes Yes Connected NREN 51 Zimbabwe ZARNet No Yes Emerging NREN Benin RerBenin Yes Yes Yes Yes Yes Connected NREN Burkina Faso FasoREN No Emerging NREN Cameroon RIC No No Emerging NREN Chad TchadREN Yes Yes 2 No No Emerging NREN Côte d'Ivoire RITER Yes Yes 2 Yes Yes Connected NREN Gabon GabonREN Yes Yes 2 No No Connected NREN WACREN Ghana GARNET Yes Yes 2 No Emerging NREN Guinea Gn-REN Yes Yes 2 No Yes Emerging NREN West and Central Africa Liberia LRREN No Emerging NREN Mali MaliREN Yes Yes Yes No Yes Emerging NREN Niger NigerREN No Emerging NREN Nigeria NgREN Yes Yes Yes Yes Yes Connected NREN Senegal SenRER Yes No Yes Yes No Connected NREN Sierra Leone SLREN Yes Yes Yes No No Emerging NREN Togo TogoRER Yes Yes Yes Yes Yes Connected NREN Algeria ARN Yes Yes Yes Yes Yes Connected NREN Djibouti No Emerging NREN Egypt, Arab Republic ASREN EUN & ENSTINET Yes Yes Yes Yes Yes Connected NREN of Libya LibREN No Emerging NREN Northern Africa Mauritania No Emerging NREN Morocco MARWAN Yes Yes Yes Yes Yes Connected NREN Tunisia RNU & RNRT Yes Yes Yes Yes Yes Connected NREN Appendix B Summary of Country Case Studies—Burkina Faso, Cote d’Ivoire, Mozambique and Uganda T he four case study countries, Burkina Faso, Côte The Education Sector d’Ivoire, Mozambique, and Uganda, represent divergent higher education and connectivity environments. While Burkina Faso and Côte d’Ivoire Burkina Faso are French-speaking with an education system influenced The higher education sector in Burkina Faso comprises by the French model, Mozambique’s and Uganda’s 18 accredited universities (10 public and 8 private) and 75 education systems are modelled after the Portuguese and Grandes Écoles (“Great Schools” or elite HEIs with highly British models respectively. All four countries have made competitive admission requirements) (23 public and 52 progress with enrollment in lower levels of education, but private). Student enrollment in higher education in 2020 are still struggling with students’ transition from lower was estimated at 145,000 students with the University of education to higher education. Gross Enrolment Ratio Ouagadougou, the largest HEI, accounting for over 50% of (GER) for higher education in Burkina Faso was reported total enrollment. Private sector participation in providing at 7.1% (2019), Côte d’Ivoire at 9.3% (2017), Mozambique at higher education has been increasing, currently accounting 7.3% (2018) and Uganda at 6.8% (2018). All are below the for about 21% of the total enrollment. Student enrollment is estimated average for sub-Saharan Africa of 9.4% (2018), projected to grow to 297,000 students by 2030. and far below the world average of 38.8% (2019). Based on the forecast of student enrollment and population Burkina Faso provided very useful insights into and an in the age category 20–24, Burkina’s GER for higher extreme experience of the situation in a landlocked education is estimated to reach 9.8% in 2025 and 11.2% in country in West Africa. Connectivity in Burkina Faso HEIs 2030. While this will eclipse the estimated sub-Saharan is only available for staff, not for students. In Côte d’Ivoire, Africa average of 9.1% in 2018, it is still far below the world the private higher education sector caters to 50% of the average of 38.7 % at the time. enrollment and relies on government subsidies, just like the national research and education network RITER, which is heavily dependent on its government subsidy to provide Connectivity at all levels of education is still a major connectivity. The Government of Mozambique is the main challenge. Based on a recent survey University provider of Mozambique’s education and connectivity, thus Ouagadougou I Professeur Joseph Ki-Zerbo, with about investment by the government is critical for its enhanced 70,000 students, offers only 34 Mbps to its users. Only the connectivity. The Mozambique Research and Education academic and administrative staff at the University use Network (MoRENet) is still a project under the Ministry of the connectivity provided by the university. Students have Science and Technology, Higher Education and Vocational to buy their connectivity from commercial providers with Training and provides subsidised connectivity to HEIs. coverage around the campus. Uganda, on the other hand, has a relatively matured NREN, the Research and Education Network of Uganda (RENU), owned by the HEIs, and a competitive ICT environment that has paved the way for the improved connectivity of the higher education in the country although this connectivity is yet to reach all institutions. 52 Côte d’Ivoire There are currently 403 entities providing higher Education Education Sector Challenges in Côte d’Ivoire. These include 7 public and 33 private universities, as well as 35 public and 328 private specialized The four countries exhibit similar challenges in relation to schools (Grandes Écoles). Due to limited capacity, the leveraging ICT for learning. The higher education sector Government has been directing new tertiary students to challenges across the four countries include: subsidized private institutions, which accounted for over 50% of all enrollment in 2018. i. Lack of an ICT policy for the sector to promote digital literacy and e-learning as an avenue to improve Côte d’Ivoire’s literacy rate for 2018 was 58.4%. The primary learning outcomes, particularly in higher education; and secondary schools’ enrollment rates are increasing ii. Generally low digital literacy among both lecturers/ with GERs reported at 100.3% and 54.6% for primary and tutors and students on how to leverage ICT for secondary education respectively in 2019. Like other African teaching and learning; countries, Côte d’Ivoire is struggling with the transition iii. Lack of knowledge and capacity on how to leverage of students from lower levels of education into higher ICT to improve teaching and learning; education. The gross enrollment ratio (GER) for higher iv. Very limited investment in campus ICT resources and education was 9.3% compared to a sub-Saharan average of infrastructure compounded by inadequate public 9.4% and a world average of 38.8% in 2018. The population funding for higher education that has not kept pace in the HEIs was 253,955 in 2019 and is projected to reach with growth in enrollment; 396,000 students in 2025. v. Lack of mechanisms to evaluate and identify relevant and/or appropriate digital content and applications for different levels and programs of education; Mozambique vi. Lack of an adequate pool of high-level ICT champions within higher education that can promote the There are currently 53 HEIs—19 accredited universities, adoption and use of ICTs within their institutions. 27 institutes, four schools, and three academies. Student enrollment in higher education in 2020 stood at 240,000 students and is projected to grow to 380,000 by 2030. While the private sector accounts for 58.5% of all HEIs, the public sector still accounts for most student enrollment (61.5%). Despite the growth in student enrollment, the GER for higher education in Mozambique was only 7.3% in 2018, below the estimated sub-Saharan Africa average of 9.4% and far below the world average of 38.8% for the same year. Uganda At the higher education level, there are currently 52 accredited universities and 184 TVETs (classified into “other degree awarding institutions” and “other tertiary institutions”). With about 259,000 students, universities have higher enrollment (72%) compared to TVETs (24%). Lower enrollment in TVETs is attributed to the negative image associated with TVETs, the general perception being that those pursuing TVET courses are failures that did not make it to university. Based on the latest data from the National Council for Higher Education (NCHE), Uganda’s GER for higher education was only 6.9% in 2018.1 This is below the sub- Saharan Africa average of 9.4% and far below the world average of 38.8% in 2018.2 1 NCHE, The State of Higher Education Report 2017/18 (2019). 2 World Bank data, https://data.worldbank.org/indicator/ SE.TER.ENRR?end=2020&locations=ZG-1W-UG&start=2010. 53 Communication Sector Environment Eighteen million Ivorians (79%) use mobile broadband to connect to the internet. Burkina Faso Côte d’Ivoire is connected to four different submarine The ICT environment has been improving in Burkina Faso cables—SAT-3/WASC, ACE, WACS, and MainOne—and has over the last 10 years, yet there is a still a significant market one of the most extensive national fiber backbones in West concentration that makes affordable network pricing a Africa, with 15,750 km of operational fiber as of the end of challenge—the duopoly between Onatel, the incumbent 2019. The cost of interconnection between any of the four operator, and Orange controls the market. Broadband landing stations is, however, prohibitive, making a mix- prices remain very high, and broadband access is still and-match approach to get the best of available pricing restricted to the major urban centers of Bobo-Dioulasso very difficult. The high cost of access and devices and lack and Ouagadougou. of confidence or skill to use the internet were cited as the main reasons hindering internet access at the household The cost of connectivity in Burkina Faso is prohibitive. On and individual levels in the country. Onatel’s website, the cost of 2 Mbps per month is 2 million XOF (about USD 1,800), while the price of 20 Mbps per The Côte d’Ivoire Internet Exchange Point (CIVIX) connects month is USD 10,000. Moreover, 11 members, including all licensed telecommunication providers, ISPs, and some international internet actors. i. Broadband prices remain very high, and broadband CIVIX hosts nine Autonomous System Numbers (ASNs), access is still restricted to the major urban centers of terminates 60 optic fibers, and provides 96 Ethernet ports. Bobo-Dioulasso and Ouagadougou; The IXP has 2 PoPs—one hosted by Orange Côte d’Ivoire and ii. Students are not allowed to access the limited the other by MTN Côte d’Ivoire. In terms of data centers, connectivity available at HEIs. Even in the capital city, MainOne has begun collaborating with Orange to build a none of the universities interviewed offers internet data center that is co-located with MainOne’s submarine access to their students in Burkina Faso. Universities cable landing station in Abidjan. provide the limited connectivity to administrative staff and sometimes to academic staff and researchers; The country is working to address its poor cybersecurity iii. There is a lack of ICT infrastructure in HEIs (electricity, image, having been reported to host 45% of the crime computers, multimedia rooms, networks), and cybercrime servers in Africa, ahead of South Africa particularly in rural areas, compounded by a lack of (19%) and Morocco (17%). It has been working to address ICT procurement strategy and specialists; and cybersecurity issues and crack down on cybercrime by iv. The nascent national research and education network, developing a cybersecurity strategy framework and setting FasoREN, cannot as of yet provide any services. Each up specialized agencies and departments to tackle the vice. university/institution still buys its internet capacity directly from the ISPs, mainly Onatel, the incumbent operator. Mozambique Burkina Faso invested in creating Burkina Faso Internet Mozambique’s ICT sector is competitive, with different Exchange Point (IXP) in Bobo-Dioulasso to provide a players in the market, but the cost of broadband remains mechanism for local providers to aggregate and exchange high. Mozambique adopted a technology-neutral licensing traffic. It has also created two Virtual Landing Points (VLPs), framework in 2016 and currently has 42 licensed operators. one in Ouagadougou and a more recent one in Bobo- Despite this, the market is dominated by three major Dioulasso. But it still lacks a carrier-neutral data center to providers—Movitel, Tmcel, and Vodacom. Mozambique is facilitate its digital ecosystem. connected to two different submarine cables—SEACOM and EASSy—located in the capital, Maputo. GoM operates Côte d’Ivoire a data center at the Maluana Science and Technology Park, in Manica, which supports e-government services provided The ICT sector in Côte d’Ivoire has seen sustained growth by different MDAs. The data center hosts the Mozambique led by an expansion of the mobile sector and the increasing Internet Exchange Point (MOZIX), which facilitates the adoption of digital technologies by the government and the exchange of local traffic among 18 local networks, including private sector. The country adopted a unified technology- MoRENet. neutral licensing framework in 2016; still, the market remains dominated by three providers—Orange, Moov Uganda Telecom, and MTN. Cote d’Ivoire has seen an expansion of advanced services. Data from the regulator, ARTCI, indicates Uganda’s ICT sector is highly competitive. There are 33 a 3G coverage of 94% and 58.5% 4G coverage in 2020. There telecommunication service providers, but the market were 37 million mobile subscribers, representing 150% SIM remains highly concentrated, rendering competition card penetration due to the use of multiple SIM cards. suboptimal. Mobile voice is the predominant service, while 54 the use of data services is still limited, with access largely via mobile phones due to the wide coverage of the mobile ICT Sector Challenges networks. Recent data from UCC indicates that 85% of the population lives within coverage of a 3G mobile network From an ICT sector perspective, the four countries share and 61% within coverage of 4G. However, the proportion similar challenges that inhibit the use of ICTs in higher of individuals using the internet is still low, at only 24%, education and better connectivity, including the following: compared to an African average of 28% and a world average I. Inadequate development and deployment of ICT of 54%. infrastructure to cover the whole country, particularly rural under/unserved areas; Although ICT policy and regulation encourage infrastructure Ii. Poor-quality and limited geographical reach of the sharing, this has yet to be fully embraced. Uganda now has national electricity grid, particularly in rural and peri- about 21,472 km of fiber-optic cable laid by both public and urban areas; private licensed providers. This covers 49% of all districts Iii. High cost of broadband services, which cannot be and 24% of sub-counties, but the duplication among afforded by the majority of the population due to licensed operators on certain routes means that effective poverty; coverage is limited and mainly covers major urban centers. Iv. Lack of awareness among leadership of government Total international bandwidth grew to 175 Gbps in 2020. agencies about the importance and potential benefits of ICT in developing their institutions as well as the The Uganda Internet Exchange Point (UIXP) has 29 socio-economic development of the country; networks that peer at the exchange and provide access V. Insufficient coordination and alignment among public to content caches from Google, Facebook, and Akamai. institutions in relation to ICT projects and initiatives, Besides the National Data Center built and operated by resulting in duplication as well inefficiencies in public NITA-U, and largely used by government MDAs, First Brick service delivery; Holdings is building Uganda’s first tier-III carrier-neutral VI. Lack of sufficient numbers of qualified human data center—Raxio Data Center. Namanve Industrial resources to serve the ICT sector both in private Park, along Jinja road, the main fiber route between Kenya and public institutions to ensure successful and Uganda, hosts the new data center. The country also implementation of different ICT projects and has two major national computer emergence response initiatives; teams (CERTs)—CERT-UG, under NITA-U, and a telecoms VIi. Poor integration of the ICT component within the sector CERT under UCC. These are complemented by the objectives and strategic documents designed to guide cybercrimes unit under the Directorate of Forensic Services the development of the country; and of the Uganda Police Force. However, cybersecurity VIii. Lack of a vibrant and competitive local ICT sector that awareness and investigative capacity for computer-related nurtures innovation and entrepreneurship in various crimes are still low. technology areas. 55 National Research and Education Following the COVID-19 lockdowns of educational Networks institutions, MCTESTP negotiated with the three major licensed operators to provide special rates that allow unlimited internet access for the registered students and FasoREN, Burkina Faso staff of HEIs to designated academic systems and content via their regular mobile phones. FasoREN, the Burkina Faso research and education network, is still under development. Using resources from the Education Access and Quality Improvement (EAQIP)3 project financed by the World Bank, FasoREN has been working on a business plan to develop a fully functioning RENU, Uganda NREN that can deliver applications, services and high- speed connectivity to HEIs across the country. The RENU network is the most advanced among the case study countries. It consists of dark fiber and a managed bandwidth backbone at speeds of up to 20 Gbps in metro areas and 1 Gbps access links that connects 204 campuses across the country, including 30 universities (out of 52) and Riter, Côte d’Ivoire 16 TVETs (out 184). Other connected institutions include 57 research institutions, seven teaching hospitals, 52 The Réseau Ivoirien de Télécommunication pou secondary schools, and 28 institutions affiliated with the Enseignement et la Recherche (RITER), the Ivorian Research education sector. and Education Network, was created in September 2012 by the seven public universities of Côte d’Ivoire to federate RENU charges member institutions the same unit rate telecommunications infrastructure. It is an entity under for bandwidth irrespective of location, but there is a the supervision of the Ministry of Higher Education and requirement to procure a minimum of 10 Mbps to justify Scientific Research. It has 2Gbps internet that costs USD the distribution costs. For connected institutions, current 55,000 monthly and is supported by the ministry. bandwidth amounts are inadequate for teaching and research purposes, but they lack resources to buy more RITER offers 10, 20, or 30 Mbps access to internet to its bandwidth. It is a kind of “chicken-and-egg” problem— members depending on their size. A plan is underway to institutions need to consume more bandwidth in order to connect RITER to the international research and education lower the unit cost, but RENU and its members currently network through the WACREN network. This new link is lack sufficient resources to commit to larger volumes of expected to add another 1 Gbps internet bandwidth to the bandwidth through long-term leases, which could help RITER community. RITER is also unique in the sense that the lower the unit cost of bandwidth, allowing them to get network also connects both public and private universities. more bandwidth even within their current budgets. MoreNET, Mozambique Challenges at the NREN-level MoRENet is a project under MCTESTP that is meant to Despite the different levels of NREN maturity across the address the research and education networking needs of four countries—FasoREN is categorized as an emerging- the higher education sector in Mozambique. The network NREN, RITER and MoRENet as connected-NRENs, and currently provides broadband connectivity to 18 universities RENU as a mature-NREN—they face a number of similar and 29 TVETs although the amount of bandwidth is still challenges that hinder better performance. These include far below the recommended minimum. The bandwidth the following: requirement for the current enrollment of 240,000 higher i. HEIs have inadequate budgets for bandwidth. education students is about 240 Gbps (based on a minimum Institutions currently only provide bandwidth for staff. of 1 Gbps per 1,000 students), but MoRENet has only 5.9 ii. Campus networks at HEIs are still in a poor state. Gbps. iii. Institutions have low levels of technical expertise among their ICT teams to manage their networks and provide the requisite services. iv. Awareness about the potential benefits of an NREN among HEIs and other stakeholders is still very low. v. The cost of last-mile connectivity to facilitate the 3 https://projects.worldbank.org/en/projects-operations/ distribution of bandwidth is high, especially to the project-detail/P148062. HEIs in rural areas. 56 Cost of Connecting Higher Education in Mozambique Burkina Faso, Côte d’Ivoire, Mozambique The overall total estimated cost of connecting HEIs in and Uganda Mozambique to high-speed broadband for a period of five years (2021–2025) is USD 452 million. This includes the Calculations were made to connect all HEIs in the four expense of providing devices to students and staff (USD 109 countries based on the cost estimates discussed in Chapter million), the cost of upgrading and maintaining campus 4. networks (USD 238 million), support to MoRENet (USD 10 million), and bandwidth cost for upstream connectivity Burkina Faso (USD 95 million). The potential savings on bandwidth cost in Mozambique from demand aggregation, smart The overall total cost of connecting HEIs in Burkina Faso procurement strategies (e.g., benchmarking regional to high-speed broadband is USD 455 million over a period pricing), and procuring long-term leases are estimated at of five years (2021–2025). This includes the expense of 61%. providing laptops to students and staff (USD 67 million) and the cost of upgrading and maintaining campus Uganda networks (USD 343 million)—the major costs being core support to FasoREN (USD 10 million) and the bandwidth The estimated overall total cost of connecting HEIs in cost for upstream connectivity (USD 35 million). The Uganda to high-speed broadband is USD 730 million potential savings on bandwidth cost in Burkina Faso from over a period of five years (2021–2025). This includes demand aggregation, smart procurement strategies (e.g., providing devices to students and staff (USD 110 million) benchmarking regional pricing), and procuring long-term and upgrading campus networks (USD 574 million)—the leases are estimated at 39%. major cost being core support to RENU (USD 6 million) and bandwidth cost for upstream connectivity (USD 41 million). Côte d’Ivoire The potential savings on bandwidth cost in Uganda from demand aggregation, smart procurement strategies (e.g., The overall total cost of connecting HEIs in Côte d’Ivoire to benchmarking regional pricing), and procuring long-term high-speed broadband is USD 876 million over a period of five leases are estimated at 73%. years (2021–2025). This includes the expense of providing laptops to students and staff (USD 125 million) and the cost of upgrading and maintaining campus networks (USD 711 million)—the major costs being core support to RITER (USD 10 million) and bandwidth cost for upstream connectivity (USD 30 million). The potential savings on bandwidth cost in Côte d’Ivoire from demand aggregation, smart procurement strategies (e.g., benchmarking regional pricing), and procuring long-term leases are estimated at 64%. 57 Comparison of Country Costs Table 26: Comparison of factors and costs across case study countries (2021–2025) Burkina Faso Cote d’Ivoire Mozambique Uganda Projected population (2025) 23,352,000 29,160,000 35,000,000 50,976,000 Forecast HEI Student and Staff 230,000 339,000 313,000 352,000 population (2025) GER for higher education 11.2% 9.3% 7.3% 6.9% Cost of connecting campus upstream using Regional Price with 35 30 95 41 Aggregation saving (USD, millions) Cost of laptops for first-year students and staff (2021–2025) 67 125 109 110 (USD, millions) Number of HEIs 93 330 53 236 CapEx + OpEx 343 711 238 574 (USD, millions) NREN core support and 10 10 10 6 development-related costs Total Cost Estimate 455 876 452 731 (USD, millions) Source: KCL calculations It is evident from the table above that Burkina Faso currently has a higher enrollment rate for higher education and more connectivity challenges than the rest of the countries, which increases its cost of upstream connectivity. Burkina Faso needs extensive support in the development of its nascent FasoREN. Uganda’s low GER for higher education weighs on its upstream and devices cost. Cote d’Ivoire and Mozambique have large student populations in public and private sectors subsidized by the government, increasing their overall cost of connecting students and staff by 2025. These subtle differences among the countries indicate the importance of national contexts that influence the cost of connectivity. The cost estimates have been developed to accommodate these differences between the countries. 58