+ 1. Summary

Broadband networks offer high-speed Internet access. Access may be fixed, mobile or nomadic, connecting users via wired or wireless access networks overlaying an invisible fibre backbone. Over-the-top services and apps deliver valuable services and experience for users. Broadband access and adoption have a positive economic impact. Yet, of the 3.8 billion people who remain unconnected to mobile Internet, only 450 million people lived in areas without mobile broadband coverage and 3.4 billion with coverage were not using it in 2020.

Most countries promote competition, but there is no consensus on state ownership. Supply-side markets can be optimized with pro-competitive market rules and spectrum policy, an efficient mix of rivalry and sharing, fit-for-purpose infrastructure sharing models, universal service programs, and reinforced network effects. Regulatory supervision plays an essential role in supply-side competition, requiring regulatory independence, competence, and capacity. Sector regulators can cultivate competition through tailored market interventions. Policymakers can increase broadband adoption rates by addressing demand-side challenges such as digital literacy, affordability, and consumer utility.

+ 2. Considerations while reading this brief

• Which challenges related to access to broadband technology in a digital economy are most prominent in your market, both 1) in general and 2) for historically underserved groups such as women and low-income people?

• Do broadband policy and regulations in your country address:
  • Digitization: The application of broadband regulation to the digital economy?
  • Inclusivity: The specific challenges faced by women, low-income people, and/or other underserved groups in accessing broadband?

• Which entities are responsible for regulation of provision of and access to broadband? Are responsibilities clear, and are mechanisms in place to avoid regulatory arbitrage? If not, how could this be improved?

+ 3. Broadband’s characteristics and impact

High-speed Internet access

Broadband is synonymous with high-speed Internet access. While the International Telecommunication Union (ITU) defined broadband as 256 Kbps in 2003, by 2021, mobile broadband speeds averaged 40 Mbps in high-income countries, 13 Mbps in lower-income and middle-income countries, and below 10 Mbps in sub-Saharan Africa. The Internet is a global, public network of networksiv carrying voice, text, audio, images, video and other information using the Internet protocol suite (TCP/IP) to enable interconnection and interoperability. Its value is driven by strong network effects (see Competition Briefing Note), whereby existing users benefit from adding more users with whom they can communicate. Since release of the World Wide Web (WWW) content management system in 1991, the Internet has become the world’s universal data network.

Last-mile access networks

Broadband is delivered ultimately to a customer over an access network, i.e., the last-mile link to a customer. The access may be provided through a wired or wireless connection, in the latter case reaching the customer through radio waves. A network may be fixed, mobile or nomadic.

Fixed access, i.e., wired, enables connection at one location, such as an office or home, and allows particularly high speeds.

Mobile access enables connection and movement without disconnection within a wide coverage area. Each network comprises an array of radio base stations or BTS, i.e., antenna and associated equipment, spread over its coverage area. The coverage area is divided into cells, which can hand off the customer device from one to the other, thus permitting mobility (also referred to as a cellular network). A mobile operator can add new cells or increase existing cell capacity to handle increases in traffic. Mobile broadband access has become near-ubiquitous – 94% of the global population had mobile broadband coverage by 2021. Mobile access is by far the dominant means of delivering broadband in low-income and middle-income countries (by 2020, mobile access enabled 87% of all broadband connections in such countries).xiv Four generations of mobile access networks are currently used, from 2G to 5G, supporting increasingly high Internet speeds as the generations have developed. Satellite phones offer remote coverage, but only at 9.6 kbps, limiting their use to emergency response.

Nomadic access enables connection at multiple hotspots, typically using Wi-Fi, but does not offer connectivity while moving between hotspots. Most smartphones contain mobile and Wi-Fi radios, enabling connections to either network type. To reduce network congestion, mobile operators off-load data traffic to Wi-Fi networks when available. Wi-Fi is expected to serve 59% of worldwide mobile broadband traffic by 2022. Some developing countries are promoting the mobile-to-Wi-Fi trend. Since 2018, Rwanda has required all public and private institutions to offer free Wi-Fi for visitors.

Terrestrial wireless fixed access networks today often use the same technology as mobile networks, such as 4G or 5G. Satellite fixed and nomadic access networks have seen renewed interest, and investors are backing the potential of low-earth-orbit (LEO) satellites to bring broadband to remote areas that cannot be economically served by terrestrial networks.

Backbone networks

Broadband’s backbone comprises a web of domestic and international fibre links. These are vital inputs to broadband access and are largely invisible to end users. Network backhaul links carry traffic from mobile network base stations to the core network and haul traffic across long distances between urban centres and to international links. In mobile networks, backhaul links are steadily shifting from microwave to fibre because fibre enables higher access speeds and reductions in congestion.

A large volume of developing country Internet traffic is carried as IP transit, via international links, to and from Internet exchange points (IXPs) in higher-income countries. Though satellite still connects many locations, by 2015, 99% of transoceanic IP transit traffic was carried over submarine cables. Four major OTT providers – Amazon, Meta (formerly Facebook), Google and Microsoft, which accounted for 55% of global IP transit traffic in 2018 – are investing extensively in submarine cable systems and distributed data centres.

Impact of broadband access and adoption

Broadband encompasses and permits diverse functionalities, infrastructures, and technologies. Increased broadband access and adoption in developing countries boosts economic growth through gains in productivity, innovation, and efficiency. A 2018 study of 63 countries with GDP per capita less than USD 12,000 found a 10% increase in mobile broadband penetration yielded a 2% increase in GDP. A 2020 study found that an increase from 2010 to 2015 in 3G mobile broadband access and use in Nigeria improved household welfare, especially in rural areas, lifting some 2.5 million people out of extreme poverty.

Most applications (apps), services, and content are provided over-the-top (OTT) of the Internet, which operates as a general-purpose data pipe, as opposed to being provided by the telecom network operator. These enrich the broadband user’s experience, drive demand, and improve economic and social welfare. OTT services include connectivity apps, social media apps, content sources, search engines and commerce apps. OTT services are supplied by numerous firms, usually separately from network access, and sometimes free-of-charge to the end user. Many OTT apps exhibit strong user network effects.

Social media apps have become a major driver of developing country broadband adoption, sometimes through partnership with network providers such as through special pricing for using certain apps. In September 2021, Facebook (now Meta) had over 2.9 billion monthly active users worldwide, of which the vast majority accessed their accounts exclusively from mobile devices. Social media use is increasing at a similar pace as Internet use in many developing countries

+ 4. Supply-side broadband market development

Effective broadband policy provides leadership, sets stakeholder expectations, and defines the state’s role in supply-side market evolution. Policymakers have three main sets of implementation tools:

• control over state supplies to the market, such as radio spectrum and wholly or partly state-owned operators;

• regulation of competition and efficient use of infrastructure; and

• fiscal levies and expenditures, such as universal service funding.

The Broadband Commission for Sustainable Development (Broadband Commission), set up in 2010 by the ITU and the United Nations Educational, Scientific and Cultural Organization (UNESCO), promotes broadband policy on the global agenda and encourages governments to adopt national plans. By 2020, 174 countries had national broadband plans. Policymakers devising or updating broadband plans should consider in particular the roles that radio spectrum, state-owned enterprises, and competition play in supply-side broadband development.

Radio spectrum management and use

Radio spectrum is a vital input to wireless broadband technology. The transmission of signals by specific technologies depends on certain frequency bands of radio spectrum. The availability of particular radio frequencies thus directly affects market entry, market growth, innovation, and service quality. In some cases, the more radio spectrum an operator can use, the less densely it needs to build out its mobile network transmitters. Depending on the technology, there are limits to how many different systems can use the same radio frequencies, particularly because multiple users could cause interference in the signals. This may result in competing demands for the limited amounts of spectrum in the various relevant frequency bands. As a result, radio spectrum is a scarce public resource, and spectrum policy and management are critical to broadband goals. Permission to use radio spectrum is controlled by sovereign nations and coordinated by the ITU and various regional organizations. Attracting investment requires a robust and effective compliance and enforcement regime.

Spectrum allocations for types of uses (e.g., mobile broadband, police, coastguard, etc.) determine the amount of spectrum available for mobile broadband. These determine the scope and scale of mobile broadband markets. Given that spectrum is a scarce resource, the telecom or radio spectrum regulator typically determines who may use which bands. It does so by making assignments or authorizations of specific frequencies to specific entities (or sometimes classes of entities) and determining who may enter the market.

Some frequency bands and technologies, such as for Wi-Fi, involve low-power, short-range signals that allow multiple users to use them without significant threat of interference. As a result, these can be designated for licence-exempt uses, i.e., requiring no permission to use the relevant frequency band. The choices around licence-exempt frequencies define the scope of shared public use.

Reassessments of existing allocations and reassignment of spectrum can accommodate evolving technology, needs, and uses, such as the migration from analog to digital television broadcasting. This has freed up significant blocks of spectrum for repurposing to support the introduction of 4G and 5G networks.

Approaches to spectrum licensing and pricing vary and have significant impact on market structure, competition, fiscal revenues, and consumer prices. A competitive approach, used when spectrum is scarce (as explained above), assigns spectrum through an auction to the highest bidder and/or through a beauty contest based on qualitative factors. A first-come-first-served or direct assignment approach, more often used when spectrum is not scarce, allows market entry or expansion by making more spectrum available as and when needed. Except for auctions, spectrum fees are set by administrative pricing, i.e., prices set by the spectrum regulator rather than the market using a variety of calculation methods.

The legacy and continuing role of state-owned enterprises

Where many countries’ telecommunications sectors had typically been operated by state-owned enterprises (SOEs), early telecom liberalisation efforts focused on privatisation. By 2002, incumbents in 113 countries were fully or partially owned by entities other than the State, new entrants that were not owned by the State had been licensed in another 49 countries, and full SOEs generated only 2% of global sector revenue. A 2002 study of 86 developing countries found that privatising SOEs and introducing competition improved sector performance. That said, the impact of privatisation varies.

Developing country governments often struggle to manage SOEs due to inefficiency, fiscal risks, and corruption. Competition (discussed next) is weakened if SOEs have preferential access to spectrum or rights-of-way, are exempt from competition law or fiscal levies, or face lighter enforcement efforts than private-sector rivals. SOEs can also be weakened if burdened by civil service and public procurement rules or political pressure to undertake special projects.

There can be a policy case for SOEs, and SOEs have persisted in some broadband markets. Indeed, State participation in public-private partnerships (PPPs) has gained popularity to finance particular types of new infrastructure,lxix but the case for SOEs in competitive markets is often not compelling.lxx Other potentially positive roles for the State are discussed further below in relation to infrastructure sharing models.

Competition as a driver of broadband development

The introduction of competition in broadband

After enshrining telegraph and telephone monopolies for much of the 20th Century, most countries now rely on competition to stimulate broadband development. Competition has been embraced as the most efficient means to discipline economic markets and increase consumer welfare (see Competition briefing note). In theory, competition puts pressure on firms to be efficient, innovative, and customer-focused, leading to lower prices, better coverage, higher capacity, and more service choices. The alternative – controlling behaviour through regulation (e.g., requiring minimum broadband speeds or coverage, or limiting prices) – is largely considered a second-best solution that is useful only if competition is ineffective.

By 2010, 93% of countries had introduced Internet access competition and 92% had introduced 3G competition. From 2001 to 2011, competition in 165 mobile markets resulted in 26.5% higher user penetration, leading the Broadband Commission to recommend extending competition to international gateways and fibre backbones.

Concentrated nature of supply-side broadband markets

Broadband networks require large technological investments, including large fixed costs, i.e., costs that do not vary with the usage of the network. They have economies of scale, meaning that because the fixed costs are spread out over the customers, the extra (what economists call “marginal”) cost of supplying each additional user declines as the number of users increases. And if network capacity is sufficient, it does not cost more to carry two calls or two megabits of data rather than one. Broadband networks also feature economies of scope, meaning that the cost of using the network for multiple purposes is less than it would be to build and operate a network for each purpose separately. So instead of having a network that can only carry voice calls (like an old telephone network) or television channels (like an old cable TV network), broadband networks can be used for calls, internet access, television channels, music and video streaming, and a near-infinite array of other uses.

Together with network effects (mentioned above), economies of scale and scope create structural barriers to entry into the market. The very nature of broadband network economics makes it difficult for new competitors to succeed in building networks, attracting customers, and competing. Such barriers lead to highly concentrated markets, i.e., where there is a small number of broadband providers, and in the case of fixed broadband, often only one. These economics of broadband make policy-led efforts to strengthen competition and attract investment in networks and services particularly challenging.

Regulating for competition

When broadband policy seeks to rely on market forces, regulators must adopt and enforce pro-competitive market rules to ensure fair play and a level playing field. Public service licensing and other market entry authorisations are critical to ensuring contestable markets. Regulation needs to avoid erecting artificial entry barriers while implementing policy goals. Regulators may regulate certain non-price behaviour of all firms to ensure an orderly and functioning marketplace – as well as to empower and protect consumers – on topics such as interconnection, telephone numbering, SIM registration, quality of service, customer care, cybersecurity, privacy and data protection, and consumer dispute resolution.

Enforcing market rules intended to strengthen competition requires proactive supervision, investigation, and remedies. Regulators may have to decide difficult cases. For example, some non-cooperative behaviour (such as not offering to share facilities) could be a sign of healthy competition, while other non-cooperative behaviour (such as denial of an important input to a competitor that needs those facilities to compete) could prevent competition, depending on the circumstances. Conversely, cooperative behaviour, such as network sharing, could also be pro- or anti-competitive. Sharing would eliminate competition in the development of the network, yet may make it possible for a service provider to use a network without which it would not have been able to provide a competing service.

Periodic market assessments may be employed to discern if a provider has market dominance or significant market power that enables it to unilaterally increase prices and profits (see Competition briefing note). Many telecom laws empower sector regulators to impose regulatory obligations on such firms. Typically, wholesale remedies that require the dominant firm to make its facilities or wholesale services available to service providers that can compete with it in retail markets are considered more effective than regulation of the dominant firm’s retail services. For instance, rather than regulating the prices a dominant firm charges consumers, requiring the dominant firm to permit rivals to use parts of its network so that they can compete with it in retail services at least introduces some competitive pressure in the market. Regulatory remedies involving passive infrastructure such as towers and dark fibre are often more practical and beneficial than those involving active services such as wholesale capacity (see next section for a discussion of optimisation of infrastructure sharing).

As discussed in this briefing note, competition is a key tool for advancing broadband development. National competition authorities also often have jurisdiction over telecom competition matters. Appropriate exercise of the concurrent jurisdiction conferred on sector regulators and national competition authorities requires cooperation and coordination to ensure well-functioning and competitive telecom markets (see Competition briefing note).

Optimising competition and infrastructure sharing

A key question that broadband policy makers and regulators face is how to foster the optimal level of competition. Eventually, competition yields diminishing returns in driving rollout and quality of networks and services and imposing discipline on costs, and consolidating resources through infrastructure sharing would be more effective. Understanding market structure options and pitfalls is vital for effective policy interventions.

Facilities- and services-based competition

Competition involves rivalry among providers. Providers may compete up and down the full value chain with each operating its own network facilities, i.e., facilities-based competition. However, economies of scale and scope (discussed above) may make it more efficient to have fewer networks. One way to address this tension is to allow for or require some sharing of infrastructure, particularly infrastructure that is costly to duplicate, while seeking competition in the provision of services, i.e., services-based competition. A key question in broadband policy is how much sharing and consolidation among providers to aim for or allow in the underlying infrastructure.

Past telecom sector liberalisation experience offers lessons. Policymakers introducing mobile competition in the 1990s focused on spectrum licensing and facilities-based competition, not infrastructure sharing. Both last mile and backhaul links were wireless in emerging markets, while backhaul in developed markets used either wireless or leased copper lines. The end-to-end configuration of 2G networks thus allowed operators to scale-up fixed costs as penetration and use grew, enabling multiple operators to achieve economies of scale in the same coverage area.

This facilities-based competition improved dynamic efficiency by driving operators to expand coverage and upgrade networks. Across 200 countries from 2001 to 2014, facilities-based competition using 1G, 2G or 3G technologies achieved wider mobile coverage more quickly. Facilities-based competition also improved static efficiency, pushing prices to their lowest sustainable levels. In the United States, the price per minute of mobile voice calls declined steadily from USD 0.44 in 1993 to USD 0.05 in 2011.

In contrast, infrastructure sharing was a centrepiece of opening wired telephone access markets to competitive entry. Fixed costs of wired infrastructure were sunk costs, i.e., they had been fully incurred and could not be recovered. This advantage over new entrants operated as a moat, i.e., a protective barrier to entry in the market against new entrants seeking to build networks and compete with the incumbent. This established or embedded incumbents’ natural monopoly characteristics, as their average costs would be inherently lower than those of new entrants.

Therefore, policymakers seeking to use competition to improve market performance required incumbents to share infrastructure with new entrants. This was typically a one-way obligation applying only to the incumbents and not new entrants, i.e., it was asymmetric regulation. One approach required them to share essential facilities that could not be viably replicated, such as access to the local loop, i.e., the last-mile connection to the customer. New entrants combined these wholesale inputs with their own facilities to compete with incumbents. Another approach introduced service-based competition, where rivals shared most infrastructure, forcing incumbents to offer end-to-end wholesale services that rivals could resell. Both approaches required rate regulation of wholesale inputs, improving static efficiency through price reductions, but reducing dynamic efficiency by suppressing investment incentives for facilities owners and access seekers. Efforts to address these disincentives largely failed.

The limits of mobile broadband facilities-based competition

Meanwhile, as mobile broadband has been deployed, the sustainable level of network rivalry in many countries has plateaued and begun to decline due to two evolving cost factors. First, customers use increasing volumes of data for a variety of business and personal purposes. The growing volumes of data on the radio access network require more spectrum, towers, and radios; this trend, which will increase significantly with the introduction of 5G, increases per-customer costs. Second, replacing microwave backhaul with fibre that can handle larger traffic volumes has increased costs and rendered them less scalable. At the same time, telecom operators’ revenues per customer have remained flat due to competitive pressure from OTT providers in voice and text services (e.g., people calling and texting over WhatsApp using the operator’s data connection instead of its voice and SMS services).

These trends have led to consolidation in many markets. Among 30 European Economic Area states, the number of countries with 4 or more mobile network operators declined from 17 in 2012 to 12 in 2017. The United States went from 4 to 3 mobile network operators in April 2020. India had 15 2G operators in 1999, but only eight 3G or 4G operators in 2019. Similar consolidation has occurred in West Africa, where active national mobile network operators have declined from 6 to 3 in Côte d’Ivoire, 6 to 4 in Ghana, and 5 to 2 in Liberia.

Passive and active infrastructure sharing

Underscoring the need for more infrastructure sharing in all broadband markets, the Broadband Commission, ITU, and World Bank recommend blanket policies to promote sharing of passive infrastructure, i.e., facilities that do not include the electronics, such as rights-of-way, towers, ducts, dark fibre, equipment rooms, and power supplies. At the same time, greater caution is required when it comes to sharing of active infrastructure, i.e., the electronics such as antennas, transmission equipment, and software. Such sharing involves complex transactions and can undermine the incentives to compete that lead to greater investment in networks and lower prices for consumers. Therefore, policymakers may wish to take a case-by-case approach to permitting or requiring active infrastructure sharing.

Infrastructure sharing models

Broadband infrastructure sharing requires fit-for-purpose institutions designed around the sharing model and role of the State, taking into account the type of infrastructure.

Sharing models include (1) asymmetric, where an operator with market power provides access to rivals; (2) wholesale, where an upstream party provides access to downstream operators; and (3) cooperative, where a multi-operator-owned entity provides access to members.

States may choose to:

• permit voluntary sharing that might otherwise have been prohibited, e.g., on grounds that it reduces facilities-based competition;

• induce operators to share;

• participate in the investment alongside operators sharing in it; or

• compel operators to share when they might otherwise have chosen not to do so.

This section examines infrastructure sharing models for different types of infrastructure through these four roles of the State.

1. Voluntary sharing arrangements

Sharing towers, the primary last-mile broadband infrastructure in emerging markets, can reduce costs and mitigate health or environmental concerns. Some operators share towers asymmetrically on an ad hoc basis, but most sharing occurs through wholesale tower companies (TowerCos), formed either by independent investors or structural separation within operator-led groups. TowerCos manage half of the global tower inventory outside China. By 2020, 30% of emerging markets had two or more TowerCos and another 30% had one. Early evidence suggests that TowerCo market entry helps drive down retail prices. Policymakers can stimulate further market entry by removing entry barriers while maintaining regulatory oversight to prevent potential abuse of dominance.

Similarly, because fibre is now a preferred middle-mile and last-mile wired solution, sharing benefits operators by reducing costs. Wholesale fibre companies (which are sometimes referred to as FibreCos, Infracos, or Netcos) have entered many emerging markets to offer dark fibre or transport over intercity, metro, and access infrastructure. These include utilities (often state-owned) seeking to commercialize excess capacity on internal networks, such as Lesotho Electricity Company; spinoffs of incumbent operators, such as Malawi’s Open Connect Limited (which was separated from the privatised fixed-line incumbent) or MTN Global Connect (which is being separated from MTN Group’s mobile business); and standalone new entrants, such as South Africa’s Dark Fibre Africa and CSquared, which operates in Ghana, Liberia, and Uganda. In Africa alone, as of 2019, over 250,000 km of new fibre must be deployed to achieve universal mobile broadband. Policymakers can attract further entry and investment by ensuring open licensing and equal access to state-owned rights of way, poles, and ducts.

Voluntary infrastructure sharing is common for submarine cables, i.e., cables laid on the seabed between land-based stations to carry telecom traffic across ocean and sea, which are the principal first mile in most emerging markets. Some 426 submarine cables connect nearly all coastal countries worldwide,cxxx including 37 of 38 coastal African countriescxxxi and 12 Pacific island territories (with plans to connect the remaining seven).cxxxii Most submarine cables are owned by groups of operators under a cooperative model, but some function as wholesalers or internal OTT networks. Operators with submarine cable capacity typically self-provision, i.e., use the capacity for their own purposes, and offer wholesale service, i.e., provide services to other operators. Pricing depends on the competitive landscape, but regulators can counter anticompetitive behaviour. In 2010, Fiji’s regulator found landing station owner FINTEL had significant market power, introduced rate regulation, and afforded local operators rights to purchase capacity directly from cable owner Southern Cross.
Voluntary sharing of active radio access networks (RANs), i.e., the part of the telecom system that connects customer devices to the network through radio connection to the base station, has also become common. Such sharing first gained a foothold mainly in developed economies during 3G and 4G deployments. Facing costly radio spectrum pricing, smaller operators struggling in crowded markets embraced sharing as an alternative to merger or exit, with regulators closely scrutinizing and conditioning the arrangements to address competition concerns. As costs of technology upgrades increase, interest in 5G RAN-sharing, such as the shared 5G network to be deployed by mobile operators M1 and StarHub in Singapore, has increased as a means to improve business cases.

However, in many markets, operators tend not to share even if it might be more efficient to do so. For example, Africa and the Middle East together had only 10 active network sharing agreements out of 98 worldwide by 2017. It is unlikely that simply permitting RAN-sharing will suffice to foster its development in many markets where the necessary business culture or trust is lacking. Developing country governments may have to induce or compel sharing (or participate alongside operator investors) in order to introduce RAN sharing.

Some higher-income markets, such as Europe and the United States, have experienced rapid growth of voluntary partnerships between mobile operators and mobile virtual network operators (MVNOs). MVNOs offer mobile telecommunications services to customers using the mobile networks of MNOs with which they have agreed upon wholesale arrangements. Typically, MVNOs serve niche market segments (e.g., particular population groups). This benefits MNOs by increasing penetration through service differentiation to such niche market segments and using network capacity that might otherwise have gone unused, hence leading to voluntary arrangements. By 2019, about 1,500 MVNOs served 337 million customers (4% of all mobile customers worldwide). MVNOs have entered some developing markets, with one in Fijicxlii and nearly 20 in South Africa, but they have not entered other countries such as Nigeria, which does not permit MVNO market entry.

2. State-induced sharing arrangements

Some countries have sought to induce additional infrastructure sharing through asymmetric, cooperative, or wholesale models by introducing regulatory or financial incentives. In the 2018 European Electronic Communications Code, the EU adopted regulatory incentives, encouraging rivals to co-invest in shared fibre or make commercial offers to rivals by relieving those who do from asymmetric regulatory obligations to provide access. New Zealand provided financial incentives to develop fibre access networks. Its Ultra-Fast Broadband Initiative involved paying a one-time government subsidy to the bidder in each region proposing the lowest amount of subsidy it requires to receive for it to build a network and establish a fully independent wholesale provider (often referred to as a reverse auction). Malawi employed an anchor tenant purchase in 2015. This involved using connectivity procurement by the Government to provide sufficient scale and certainty of funding to induce new private investment in wholesale IP transit that could also be made available to private sector operators. Some countries, such as Malaysia, have used spectrum licensing to incentivize mobile operators to make wholesale offers for MVNOs, while others have extracted commitments to host MVNOs when approving mergers between mobile operators.

3. State equity participation in wholesale PPPs or SOEs

Some countries have instead set up PPPs by investing directly alongside private partners in the equity of wholesale or cooperative operators. In 2013, Rwanda established a combined 4G/LTE and fibre wholesale operator in partnership with KT Corp of South Korea. Nine West African governments have since 2011 partnered with local operators to introduce cooperatives as submarine cable system landing parties, including Benin, Burkina Faso (via terrestrial route), The Gambia, Guinea, Guinea-Bissau (pending), Liberia, Mauritania, São Tomé & Príncipe, and Sierra Leone.

Other countries have retained or re-established fully state-owned enterprises to serve as wholesale infrastructure and service providers. Broadband Infraco was established by the South African Government in 2007 as a state-owned wholesale FibreCo, acquiring existing fibre assets from state-owned national railway Transnet and electric utility Eskom. Broadband Infraco subsequently has invested in additional fibre assets, is mandated to reach unserved or under-served areas, and encourages infrastructure sharing to avoid duplication of fibre network roll-outs. Botswana Fiber Networks (BoFiNet) was established by the Botswana Government in 2012 as a state-owned wholesale provider. It acquired state-owned Botswana Telecom’s existing domestic fibre infrastructure and capacity in two submarine cables, and it is mandated to drive connectivity and economic growth.

4. State-compelled asymmetric sharing arrangements

By 2019, at least 110 countries had embraced asymmetric access regulation. Such regulations covered a broad range of sharable facilities, often limited to passive infrastructure and sometimes only to owners with market power. Ghana’s 2008 telecom law requires every licensed operator, without regard to market power, to share passive infrastructure with any requesting operator. Some countries compel mobile operators with market power to make wholesale offers to MVNOs, although Canada and South Africa have recently considered and rejected such moves based on findings that retail markets were competitive and no wholesale remedy was necessary. Many countries also compel public utilities and authorities to share land, poles, ducts, and fibre with operators, as this typically has no adverse investment impact. Informed policymakers and regulators exercise caution to ensure that compelled asymmetric access between rivals does not undermine investment.

Universal service frameworks and programs

In 2021, 450 million people – 93% of whom lived in low- and middle-income countries – had no mobile broadband coverage and were beyond the viable reach of any fibre access network. Satellite/Wi-Fi access offers the potential to close 10-20% of this coverage gap with fixed or nomadic broadband, but mobile broadband remains the most cost-effective solution to reach the remaining 80-90%.

Policymakers in market-based economies have four main options to close the coverage gap. The first is to foster more infrastructure sharing to improve viability. The second is to impose minimum coverage obligations under service or spectrum licenses. This forces operators to cross-subsidize rural service internally from their urban profits and is thus limited by overall network profitability. It may also reduce the government’s radio spectrum revenue and typically cannot achieve universal coverage. The third option is for the government to procure services for public-sector connectivity needs (for schools, health facilities, and other remote sites) in unserved areas to stimulate coverage expansion.

The fourth option establishes universal access subsidies to reach areas remaining after exhausting the first three options. Contributions may come from levies on telecom revenue, fiscal appropriations, international donor grants, or other sources. Held in a universal service fund, deposits are disbursed to operators selected to extend coverage in designated areas. Contracts can be awarded through reverse auctions or other mechanisms. As in Pakistan, the successful operator may be required to provide wholesale open access to the subsidized facilities. Economic challenges faced by remote and dispersed countries, such as small island developing states, require proportionately larger external subsidies.

Politics and regulation

As discussed above, regulatory frameworks can play an important role in optimising broadband policy. Telecom regulation, once limited to managing spectrum and monopolies, now requires strong institutions and an evolving focus on fostering and strengthening competition. This is all the more important given the continued role of SOEs in the sector alongside and often in competition with privately-owned entities.

Sector regulators in liberalized markets are typically entrusted with expansive duties and powers that are vital to performing their functions. Because regulatory independence typically leads to more objective, reasoned, competitively neutral, and predictable decisions, a majority of countries have separated telecom regulator structures, financing, and decision-making from ministries and SOEs (increasing from 14 in 1990 to 153 in 2018). In reality, regulators in many countries often attempt to align to the political will of a country’s government. This can lead them to be influenced by political pressure to protect an SOE incumbent that may be an important employer or provider of revenue through its control over fibre or the international gateway. Regulators may also find that they are circumvented by actions of governments that may issue orders or new licenses without lawful powers or respecting requisite procedure.

The pressures on regulatory authorities make it particularly important to establish the institutional framework and regulatory processes in primary legislation. A sector law typically sets out the framework governing the regulator’s composition, selection, decision-making, powers, funding, and accountability. The legal framework may establish principles and procedures to ensure stability, long-term orientation, transparency, consultation, evidence-based and non-discriminatory decisions, judicial oversight, and dispute resolution. It may also authorize regulators to acquire sufficient competence and capacity through authority to hire qualified staff and source external services.

+ 5. Demand-side broadband policy

In 2020, some 3.4 billion people in areas with mobile broadband coverage did not use the Internet. Responding to the usage gap, the Broadband Commission and policymakers in many countries have stepped up their focus on demand-side measures in consumer markets, where penetration levels are impacted by skills, affordability, and consumer utility.

Lack of digital and literacy skills is the predominant broadband adoption barrier. Many developing countries and regions, such as Kenya and the Pacific Islands, increasingly recognise the need for educational, training, and certification measures to provide their populations with the requisite skills to participate in the digital ecosystem.

Affordability of broadband devices and services is the next most important adoption barrier. In 2019, the average cost of entry-level smartphones in developing countries was 34% of GDP/capita, an unattainable investment for many consumers. Airtel, Orange, Safaricom, and Vodacom offer smartphone financing to facilitate broadband adoption in some African markets. Tax burdens on consumers contribute to the affordability gap. Some countries impose higher levies on broadband than other goods and services. In 2017, Guinea levied taxes equivalent to 61% of mobile operators’ market revenue, of which over half was sector-specific.

Limited utility to consumers also suppresses adoption in some developing countries. For example, some globally popular OTT apps have limited localization of content and language in some countries. OTT providers, particularly in Asia, are now increasingly localizing their applications and content. Some governments are also adding content. In 2015, Rwanda established Irembo, an online e-government platform that operates in English, French, and Kinyarwanda over the Internet (as well as on 2G USSD channels) and through support agents. Irembo aims to make citizen-government transactions easier, faster, and less costly.ccv

+ 7. Additional resources

Further reading

• Digital Regulation Platform (ITU and World Bank)
• OECD (Organisation for Economic Co-operation and Development) broadband
• European Union broadband

Organisations

• Broadband Commission
• ITU

+ 8. References

For the full list of references, please download the PDF of the brief in English or in French.