Chapter 4. Competition and infrastructure bottlenecks

This chapter focuses on two fundamental issues concerning the availability of innovative and competitive broadband access services and applications. The first is policy making and regulation to encourage competition. The second is good practices in this area intended to address such bottlenecks as access to essential facilities for deploying broadband services. Special attention is paid to dominance issues and regulation at the wholesale level, the key issues for regulators in the LAC area.


Experience in OECD countries demonstrates that competition is the key to promoting rapid broadband development. Broadband competition stimulates network rollout and higher speeds and helps to lower prices, which in turn, attracts users. Another relevant key issue, affecting both competition and investment in broadband access, is to remove infrastructure bottlenecks, because access to the existing passive infrastructure acts as a high barrier for existing operators and new entrants alike.

Broadband will only develop when the markets for fixed and mobile communication networks and services are competitive. The majority of Latin America and Caribbean (LAC) countries lag in the development of fixed networks. Many still have powerful dominant operators with large market shares and weak institutional frameworks that slow competition and block the expansion of services. Such weaknesses make it challenging to meet broadband policy objectives through market forces. Encouraging competition and network deployment of mobile broadband access is especially important, because this may be the most efficient and economical way to deliver services and competitive choice for consumers in areas where high-speed fixed networks are not well developed.

This chapter presents a set of policy and regulatory instruments to advance competition in communication networks and services in LAC countries. In addition to interconnection issues, which have already received a great deal of attention in the region, a full set of tools is proposed to advance competition in fixed and mobile markets. Removing potential bottlenecks, such as restrictive rights of way, limitations of access to poles, ducts and so forth, is of extreme importance to facilitate network deployment and encourage investment by new entrants.

Facilitating and encouraging passive infrastructure sharing is also critical, to encourage the deployment of both mobile and fixed networks. This can assist in addressing outstanding challenges to increase deployment of fibre access and backhaul infrastructure for mobile networks.

Key policy objectives for the LAC region

The main high-level policy objective for opening broadband markets in the LAC region is to promote competition along the entire value chain for all actors (including new entrants and alternative providers). This general goal can be enhanced by some specific policy objectives, such as:

  • establishing an investment-friendly environment that facilitates both national and international investment, as well as reasonable regulatory certainty

  • lowering, and where appropriate, eliminating administrative entry barriers; simplifying procedures, time and costs for obtaining licenses to deploy networks; as well as easy, quick and efficient access to scarce resources such as numbering and spectrum

  • facilitating efficient access to rights of way and passive infrastructure deployed by other actors, in particular market players with a dominant position

  • ensuring effective and efficient interconnection among the different actors

  • facilitating demand-side competition by reducing switching time and costs; promoting effective and rapid number portability in fixed and mobile markets; and monitoring and controlling retention clauses and penalties for switching operators

  • monitoring and assessing dominance in the different markets for both mobile and fixed broadband access; taking corrective measures against any dominant position in the geographical areas that lack competition

  • ensuring access to infrastructure controlled by dominant operators that is difficult to replicate (usually qualified as essential facilities), as well as encouraging network sharing among all market players, ideally through their own commercial negotiations, to reduce investment and accelerate the rollout of broadband.

Fulfilment of these policy objectives facilitates competition and investment by operators, maximises coverage of broadband services, reduces prices for the final consumer and encourages innovation in broadband services.

Tools for measurement and analysis in the LAC region

The collection and analysis of data are crucial in setting and monitoring the implementation of policy objectives, as well as for adapting specific policy or regulatory measures as necessary, and benchmarking with other countries, as addressed in Chapter 2 on regulatory frameworks and digital strategies. To inform policies aimed to promote competition, the following assessments are advised:

  • Collecting data on investment by market players. Such data, provided regularly by operators, should, when possible, show the type of investment undertaken by each competitor related to revenues and customers. Time series and comparisons with similar countries of the volume of investments can help detect trends and potential problems.

  • Monitoring competition, the typical indicators that should be produced and updated regularly are: number of operators active in each market, revenues for each operator, market shares and their evolution over time (revenues/subscriptions), concentration indexes (such as the Herfindahl-Hirschman Index [HHI]), evolution of prices, evolution of number portability, evolution of broadband speeds (both those announced and measures of real speeds), as well as coverage, quality of service and level of investments. As discussed later in this chapter, these indicators are needed to perform market analysis, which is the basis for taking regulatory reform measures aimed at encouraging competition.

  • Infrastructure bottlenecks and issues with difficulty in accessing the relevant inputs needed to provide services should also be assessed regularly. Key performance indicators (KPIs) should be identified based on data about the availability of infrastructure and other resources. For each of the potential bottlenecks, one or several KPIs should be defined. The following box lists some key performance indicators that could be used to monitor fulfilment of policy objectives, as well as references to existing reports from regulatory authorities that can be used to identify KPIs and data collected on infrastructure bottlenecks and access to relevant inputs (Box 4.1).

  • With respect to rights of way, it is difficult to define meaningful KPIs, as time frames and procedures may vary for each different right of way (i.e. ducts, poles, base station sites and so on). In addition, access to rights of way may be dependent in some cases on municipalities rather than a dominant operator. Nevertheless, it is important that the national regulatory authority have the power to facilitate access to rights of way and to monitor developments in this area, particularly since administrative authority over rights of way may be distributed among many different levels of administration in a given country. Development and publication of infrastructure maps, covered later in this chapter, is one way of informing operators about the availability of passive infrastructure that can be used in new network deployments. The aim of the regulatory authority is to lower barriers for market players to acquire new rights of way, access existing rights of way and minimise the time required to obtain approvals (e.g. access ducts, install poles, construct towers, place antennas and so on), applying simple and quick procedures in the whole country.

Box 4.1. Key performance indicators for competition and infrastructure bottlenecks

Examples of KPIs for competition:

  • number of operators in each market

  • revenues and subscribers for the fixed and mobile markets by operator

  • market shares in terms of revenues/subscribers, concentration indexes

  • evolution of prices (typically baskets of services)

  • numbers ported among operators (fixed and mobile), portability times (average and deviation) and number failed portability requests

  • broadband speed (fixed and mobile) contracted and measures on real speeds

  • broadband subscribers by technology

  • data on volumes and growth in broadband traffic

  • data informing fixed-mobile broadband substitution.

Some examples of KPIs for infrastructure bottlenecks:

  • access supported by different technologies deployed by different providers (e.g. number and percentage of copper lines, fibre access, number of mobile base stations and coverage relative to geographic area and population density)

  • prices for termination rates (for both fixed and mobile)

  • number of unbundled and bitstream-based accesses, as well as, when possible, number/length of ducts available and used by alternative operators

  • availability of backbone infrastructure and Internet exchange points (IXPs) (relative to traffic and country dimensions)

  • number of circuits and leased lines provided to alternative operators by the incumbent (per technology and speed).

Overview of the situation in the LAC region

In the LAC region, a large variation obtains regarding competition and infrastructure bottlenecks. Different geographical areas within each country also show a wide variation, as is the case in any other region of the world.

In general, recent years have seen an increase in competition, especially in mobile markets where new entrants are competing with incumbents. However, by comparison with most OECD countries, the LAC region has room for competition both in fixed and mobile broadband markets. In many of countries, market shares are highly concentrated. Most new entrants in these markets also face infrastructure bottlenecks, such as access to rights of way, passive infrastructure and international gateways.

The reasons for this lack of competition, especially in large urban areas where there is room for several operators to compete on prices and services, are chiefly based on the following issues:

  • In a number of countries, the process for granting licences is cumbersome, protracted and costly, even when there is no associated spectrum involved. In general, except in some countries with short-term plans to adapt the licensing regime, notification-based market entry models have not been adopted.

  • Although most countries in the region do not restrict foreign investment in the telecommunications sector, a few exceptions remain, with limitations on the share of foreign ownership. This reduces the incentives for external investments. In a few countries, new entrants are still at a disadvantage by comparison with dominant publicly owned operators, which are not subject to regulatory obligations to provide access to essential facilities. Although most countries in the LAC region have liberalised telecommunication markets and, in general, operators are privately owned, some state-owned operators exist. In some countries, state-owned operators are said by some to have preferential treatment, with those players benefiting from priority on government contracts, spectrum granting or regulatory fee exemptions. As a result, in such instances, new entrants have difficulty expanding their networks and obtaining customers.

  • Termination rates for both fixed and mobile voice, although they have been reduced over the last few years, are still very high in many LAC countries compared to OECD non-LAC countries. Termination rates are still not regulated in many countries, and the drivers used to set termination rates are in many cases unclear and render prices much higher than costs. This results in higher prices for customers and a barrier for competition for new entrants that must face high costs for delivering calls to dominant operators. It also raises the issue of on-net/off-net retail price differentiation, which is a barrier to the entry of new players.

  • A minority of LAC countries still focus on retail price regulation (typically, price caps for fixed and even mobile voice), with the aim of protecting consumers. Meanwhile, not enough focus is placed on wholesale regulation facilitating access to essential facilities for new entrants. As addressed in the section on good practices, if competition is ensured by applying wholesale regulation, retail regulation may be reduced to a minimum or may even not be applicable, as prices will be disciplined by market forces and retail regulation may interfere with innovation and competition.

  • Except for a few exceptions, in the LAC region, the concept of “dominance” is considered in the regulatory framework, as well as the application of regulatory measures to solve competition issues that arise out of dominant market positions. However, in a number of cases, dominance is not analysed regularly, dominant operators are not regulated, or the regulation imposed is ineffective, due to judicial litigation by the dominant operator, lack of enforcement or the lack/inadequacy of reference offers.

  • An important regulatory tool that has been used in many OECD countries to ensure effective market access by new entrants and reduce the dominant position of incumbent fixed telecommunication network operators is local loop unbundling (LLU). Unbundling and bitstream access have helped provide a foothold to new entrants and, in particular, to Internet Service Providers (ISPs) in the broadband market. In the majority of the LAC countries, LLU and bitstream access is not imposed on dominant operators, or when it is imposed, has not been effectively used by alternative operators, due to high prices, lack of enforcement and/or inadequate reference offers. This has not allowed for competition for fixed broadband from new entrants in areas such as large cities, where LLU can help encourage competition. As discussed further in the section on good practices, LLU is one of the regulatory options that can encourage competition for fixed broadband, particularly if there are too few alternative infrastructure competitors (e.g. cable television broadband operators) to exert enough competitive pressure. However, LLU should not be viewed as an isolated regulatory tool and evaluating its applicability, together with monitoring its effects on meeting goals (e.g. encouraging infrastructure investments, innovation, competition to improve service), is critical. Promoting mobile broadband extension and competition is also important, and further regulatory measures in this area should be implemented when necessary.

  • In the mobile sector, some countries have imposed national roaming obligations to allow for the phased deployment of infrastructure by new entrants and to help overcome the market advantage of the incumbent mobile operators. This has been helpful in creating mobile competition, but often needs to be supplemented by wholesale pricing agreements that allow fair access to the incumbent’s network by new entrants. Allowing market entry by mobile virtual network operators (MVNOs) is also important for creating competitive conditions in the mobile market. Again, a key requirement for the success of MVNOs is the wholesale price for access to the networks of mobile network operators. A few LAC countries have initiated regulation permitting MVNO market access. There are, however, novel initiatives such as the Red Compartida (shared network) proposal under discussion in Mexico, where a new network infrastructure will use a public-private partnership model to provide network access to MVNOs as well as to MNOs, and even fixed network operators, in rural areas (further detail on this project appears in Box 4.2).

  • Ensuring access to rights of way for ducts, poles and base stations is a key regulatory requirement in the LAC region, and an essential part of this Toolkit. In many cases, there are no national rules or regulation to grant right-of-way access, and municipalities have wide discretion to impose very high fees to access rights of way; to delay, or even prohibit, deployments; and to impose unreasonable conditions. This is an issue of concern for operators and discourages investment. As discussed in the next section, the LAC region and OECD countries can furnish examples of good practices aimed at reducing barriers to rights of way that can be used to improve market access and facilitate passive infrastructure deployment.

  • Infrastructure sharing of fixed networks, apart from wholesale supply obligations for dominant operators, is not developed sufficiently in the LAC, while infrastructure sharing in the mobile market is relatively more developed. In some cases, sharing of mobile infrastructure has resulted from regulation, as it has in Chile. In some LAC countries, public passive infrastructure has been made available to new entrants, for example in Peru and Brazil, where the passive infrastructure from the public electricity utility is shared by law with telecommunication operators.1

  • IXPs are widely available in the region but unevenly distributed, with the highest number of well-functioning exchanges in the South American sub-region, followed by the Central American and Caribbean sub-regions. Brazil has been a leader in increasing the number of IXPs in its territory, from a single IXP in 2004 to 23 in 2013. However, a number of countries are still without such infrastructure. This results in higher costs and lower quality of service for managing Internet traffic, which in many cases, could be exchanged at a national level, especially for alternative operators. Additionally, lack of effective IXPs is a disincentive for the development of local data centres, which should also be encouraged in the region, and will be the subject of a forthcoming Inter-American Development Bank (IDB) study of the LAC region.

  • International bandwidth high costs and the deployment of content distribution networks (CDNs) are other issues that need to be addressed to lower retail prices and increase competition and quality of service. These issues are addressed in Chapter 8 on regional integration.

Box 4.2. Promoting efficient use of numbering resources


In Spain numbers assigned to operators must be used within 12 months, counting from the moment where the assignment took place. In order to ensure an efficient use of numbering resources, operators are required to report regularly (each year) the use of numbers previously assigned, as well as a forecast on number resources that will be needed in the next three years. Once requested, number blocks must be provided within three weeks.

Source: Minetur (2009), Reglamento sobre mercados de comunicaciones electrónicas, acceso a las redes y numeración,

United Kingdom

In the United Kingdom, Ofcom has carried out several reassessments of the numbering plan. In the latest National Telephone Numbering Plan 2015, which is made public in Ofcom’s website, it is defined that withdrawal of a number allocation can take place when the communications provider has not adopted these numbers within six months from the date it was allocated.

Source: Ofcom (2015), The National Telephone Numbering Plan, Office of Communications,

Good practices for the LAC region

Authorisation/licensing models

In general, the authorisation system for operators should be as simple, prompt and inexpensive as possible. For those services not using scarce resources, such as spectrum, which cannot be allocated to a large number of players, a notification-based system can be used. Smooth access to market entry, as the experience in most OECD countries shows, facilitates competition.

Colombia offers an example of an efficient market-entry regime. After the approval of the Information and Communication Technology (ICT) Law in 2009, market-entry requirements for telecommunications operators, other than those using spectrum, have been limited to a registration process. The process is fairly rapid and not burdensome. The provision of a communication service that uses radiofrequency spectrum requires a prior authorisation or license from the Ministry of ICT (MinTIC). The ICT Registry, under the responsibility of MinTIC, can be modified and updated. Moreover, when an operator provides or stops providing a new communications service, it must inform the ICT Registry of such changes. New entrants are required to provide information on the legal and natural persons undertaking the registration, a description of the network and services to be provided and the use of scarce resources, such as spectrum.

Registration-based authorisation processes are usually not linked or specific to any service (uniform registration requirements for all services), facilitating entry for convergent operators. In general, registration should not entail high fees (registration can either be accomplished without charging a fee or involve a low payment to cover the expenses of registration). The registration process can be very simple, for example notification using a predefined format. As in many OECD countries, authorisation is granted if there is no feedback from the regulator within a short period (e.g. two weeks or a month). The registration process is also used to collect fees, send notifications to operators and for other regulatory purposes, such as data collection.

State-owned operators

Publicly owned operators have traditionally taken strategic decisions on investment, prices, coverage and innovation, based on very different drivers than private operators. In some cases, this has led to budgetary issues and potential conflicts among commercial and public policy objectives. At the same time, these state-owned operators, when operating in competition with the private sector, should also compete on an equal footing. This can be challenging for both the policy maker and the state-owned operators, since many of these state-owned operators are subject to specific regulatory frameworks.

Good practice in this area should be based on several key aspects:

  • Both private and state-owned operators must be subject to the same regulatory framework and when possible, special exemptions should not be given to state-owned operators (for example, exemptions on fees to be paid or automatic granting of licenses or contracts). However, it may be necessary to impose asymmetric regulation on a state-owned operator (usually the incumbent operator) because it is found to be dominant in one or more markets, and equally, on privately owned operators when they are in a dominant position. Universal service obligations are usually imposed only on a dominant operator (sometimes state-owned) because it has the most extensive infrastructure.

  • State-owned operators should have the flexibility to compete with the private sector. For practical purposes, they should be subject to similar conditions, whether for contracting services, budgeting and financing network deployments, or launching new services.2 A careful review may be needed to shift the setting of specific social objectives from the state-owned regulator to the scope of the governmental action, to avoid conflicts and promote healthy competition. For example, network deployments in rural areas, where there are not enough returns to cover investment but where action is needed, to ensure access, should be funded by different sources than the public operator revenues. Tenders and conditions for deployment and access should not favour or penalise public-owned operators. Chapter 5 addresses in more detail the role of publicly owned operators in the implementation of broadband extension plans.

  • It is critical to ensure the independence of the communications authority, as well as separation of functions between the ministerial agencies in charge of policy making and the ministerial agencies in charge of controlling state participation in any state-owned operator.

Putting these recommendations into practice is challenging and politically complex, which is one reason why, in OECD countries, state-owned operators have been privatised.

Foreign ownership restrictions

In general, foreign ownership restrictions in the telecommunications sector have been lifted in most of the LAC region, although some countries limit the share of foreign participation. However, in the broadcasting sector, foreign ownership restrictions are widely used.

Telecommunications markets have large associated infrastructure costs, requiring a long recovery period. This means that market players with large funds to invest and with easy access to financial markets are better placed to invest the necessary sums. Most of these actors are transnational, rather than based in the country invested in. This is especially challenging for small countries, but it also applies to large countries, and for example, in most of OECD countries, where many operators are not national, but transnational.

Administration of scarce resources

The availability of spectrum is a key issue for encouraging competition in mobile markets. As noted in Chapter 3 on spectrum policy, development of broadband in the LAC region will depend on more spectrum being made available for mobile broadband services. This is also relevant in terms of accelerating the digital switch-over for television services, so that the digital dividend can help expand the use of mobile broadband services. Authorities should ensure that this spectrum is fully available, to speed this process.

Well-defined policies and procedures for number allocation and management are also important. Numbering plans for different services must be periodically reviewed, to anticipate future needs and allow for the provision of emerging services needing public numbering (for example for machine-to-machine, or M2M, services, as discussed in Chapter 8). It is important to set up well-defined procedures allowing market players to request numbers and to ensure that they are assigned rapidly by the relevant authority, verifying that numbers have been efficiently used (Box 4.2). One-off fees for number assignment should in general be oriented to cover the cost of numbering management, and any recurrent fee should be aimed to encourage efficient management by the operator. High fees should be avoided, since they may deter or disincentivise growth of services and/or entry by new players.


Providing telecommunication services requires interconnection among operators, since each operator’s customers need to connect to customers of other operators. Ensuring that operators interconnect effectively is a key requirement of regulatory policy. All the countries in the LAC region have established obligations to interconnect. However, as there are no incentives for operators with large market shares to facilitate interconnection, regulatory authorities must ensure that interconnection is prompt and effective. The following good practices help to ensure that interconnection for fixed and mobile voice services is not a barrier for competition:

  • Requiring the publication of a reference offer by dominant operators. This should specify: all the technical issues; prices for interconnection and other ancillary services; procedures for requesting interconnection; time for provision; procedures for incident management; standards for quality of services; penalties for not meeting published offers, and any other relevant issues. The regulator must review the reference offer and consider input from all operators (both dominant operators and new entrants), to ensure it is adequate. Reference offers should be issued and reviewed regularly to address any problems, and should be adapted to changing costs and technological developments.

  • Ensuring that prices for interconnection are cost-oriented. In many cases, and especially when there is an asymmetric situation in terms of market share, dominant carriers have a clear incentive to increase costs for rivals by setting high termination rates. This can result in significant costs for new entrants, in particular as interconnection is crucial in the early stages as they build up their customer base. Although in certain cases, negotiation among operators may lead to low prices for termination, prices for interconnection should not only be monitored but, in most cases, directly regulated. Regulated prices for termination rates can be based on costs, in order to facilitate low rates, while ensuring that the cost for providing the termination service is covered.3

Interconnection will also remain important for all-IP networks that replace the public switched telecommunication network (PSTN) for fixed services, including voice services. 4G Mobile Networks are also moving to use IP for voice. In this context, reference offers for interconnection should facilitate a smooth transition to managed VoIP interconnection among operators. When technically feasible, the implementation of IP interconnection protocols and procedures should be ensured to improve the efficiency and effectiveness of interconnection among operators that are using IP networks for managing voice services.

The provision of Internet services also relies on interconnection with other operators. Many, if not most, of the content and applications accessed by consumers are hosted by an operator different from their service provider. However, there are important differences between voice and Internet services, justifying a different regulatory treatment:

  • A much larger proportion of Internet traffic consists of connection to resources provided by content and application providers in foreign countries. Although more and more of this traffic is being managed by CDNs hosting content at a national level, a significant part of the traffic is international, and the interconnection model is different and more complex than in voice services.

  • Internet interconnection has largely been based on peering (direct traffic exchange between two operators) using bill and keep models (no payments passing between operators, provided that traffic is in general symmetrical). The Internet interconnection model is, however, evolving, due to the rapid growth of Internet traffic for access to audiovisual content. New actors, such as CDNs, have appeared, and access operators would like to charge large content providers to send content to end consumers.

Prices for Internet interconnection have not typically been regulated, and the market has provided excellent results for Internet interconnection in most countries with high penetration and use of Internet services. There is therefore no need to apply in the LAC region models similar to those applied for voice (regulation of termination prices and publication of reference offers). That does not mean that regulators should not pay attention to Internet interconnection. In fact, prices for Internet interconnection are higher in the LAC region than elsewhere in the world, and active intervention is needed in some areas to ensure sufficient competition. Recommended good practices include:

  • Monitoring by regulators of prices for interconnection and evolution of interconnection models. Data from the relevant players should be collected and processed to make informed decisions and take action when needed. The evolution of the interconnection models, driven by the increased use of Internet for accessing content, may present challenges in the future, and it is important to detect trends and potential bottlenecks as they emerge.

  • Promotion of the deployment of IXPs. IXPs are platforms that enable new market entrants to compete in the Internet service market in an efficient and cost-effective way. By using IXPs to interconnect with other networks, service providers can reduce their operational costs, becoming more competitive. A better quality of service is achieved, as networks can generally interconnect directly with more efficiency without a third-party network connection. In some contexts and depending on the concrete infrastructure bottlenecks, IXPs can increase the available bandwidth to the service providers, which, in turn, can offer better connectivity packages to their customers. A more detailed discussion on the role and benefits of Internet exchange points and local data centres, as well as best practice, is discussed in Chapter 8 on regional and international integration.

  • Fostering Internet openness. This issue, addressed in Chapter 7 on convergence, may have implications for interconnection among players. Regulatory initiatives in this area are focused on retail service provided to customers, and in general, there is no need to set any specific regulation for Internet interconnection. However, disputes may occur among operators that require intervention from the regulatory authority to ensure that interconnection agreements are consistent with regulation of Internet openness.

  • Development of backbone infrastructure and gateways. Lack of backbone infrastructure and alternatives for connecting to foreign networks is an issue in a number of LAC countries, leading to higher costs for Internet interconnection. Chapter 8, on regional integration, addresses the situation of the LAC region and good practices in this area.

Increasing mobile broadband competition with the entry of MVNOs

Mobile providers play a key role in the provision of mobile broadband access, and in particular in most of the LAC countries, where fixed networks have limited geographical coverage. Taking into account trends in the reduction of prices for mobile terminals, technological evolution and the new spectrum made available for mobile broadband, mobile providers’ role is likely to increase in the next few years. Encouraging competition in this market and facilitating entry of new mobile operators is thus essential.

Spectrum is a scarce resource, and deploying mobile networks demands high sunk costs. In practical terms, the market for mobile network operators with a national footprint is limited to a small set of operators, typically from three to five. Given this limit on the number of Mobile Network Operators (MNOs) able to compete for mobile services, it is important to facilitate competition from other mobile providers using mobile wholesale access from MNOs. These Mobile Virtual Network Operators (MVNOs) do not have their own mobile access infrastructure, and use the mobile access network deployed by MNOs. Depending on the availability of their own core network infrastructure and numbering, MVNOs can be classified in several different categories, ranging from full MVNOs to resellers whose business model is based on marketing services completely supported by MNOs.

MVNOs account for a small market share in many OECD countries, including Chile and Mexico, and also operate in some other LAC area countries, such as Argentina and Colombia. Although the role they play in competition is not as important as that of new MNOs entering a market, MVNOs may increase competitive pressure on MNOs, and provide customised products for specific niche markets, for retail segments MNOs do not cover.

A first set of good practices should ensure that there are no regulatory barriers impeding MVNO entry, specifically:

  • Ensuring that the regulatory framework allows for simple and prompt authorisations for MVNOs, addressing specifically the issue of number allocation for these players, and that the rights and obligations for mobile providers, such as portability requirements and consumer protection, also apply to them.

  • Monitoring the wholesale market for the use of MNOs’ mobile access infrastructure by MVNOs, and maintaining regular contacts with both MVNOs and MNOs to identify any refusal to provide wholesale access, or any other barrier MNOs may impose.

  • Ensuring that MNOs do not set unreasonable technical or pricing conditions or establish unfair limitations that disrupt the business model for the MVNOs. In principle, there is no need for ex ante regulation if MNOs are willing to provide wholesale services to MVNOs, but if potential problems are detected, at least in a first phase of MVNO introduction, regular meetings with both MNOs and MVNOs are recommended, and, if necessary, specific regulations may be set out to avoid obstacles to market entry by MVNOs.

MVNOs use wholesale access from MNOs to provide their services. This means that they must have an agreement and interconnection with at least one MNO to use their network. Experience shows that in many countries, these agreements take place without the need for imposing regulatory obligations on MNOs. However, often where there are few MNOs with similar market shares and incentives to tacitly collude on deterring the entry of MVNOs, MVNOs may not be able to reach an agreement, or wholesale prices may be too high to allow for a sustainable business model that allows them to compete with MNOs. In such situations, national regulatory authorities may use the following expedients:

  • Undertaking an official enquiry into the reasons for lack of agreement, by obtaining data from MNOs and MVNOs and if necessary, due to refusal from MNOs, for example, imposing regulation of wholesale access for MNOs, setting obligations to define a reference offer for MNOs, or even regulating prices. The obligation to provide wholesale access could be imposed on all MNOs or those in a dominant position. In certain cases, the credible threat of regulation is enough to ensure that MVNOs can obtain wholesale access at reasonable terms.

  • Including clauses aimed at ensuring provision of wholesale access for MVNOs in licenses for new MNOs, as part of the requirement to authorise mergers/acquisitions, after careful analysis of the expected effects on competition, or in new bands of the spectrum aimed to ensure provision of wholesale access for MVNOs.

An interesting new initiative in the LAC area is Mexico’s Red Compartida (“shared network”), promoted by the Secretary of Communications and Transportation (SCT) (Box 4.3). This mobile network is intended to support any MVNO as well as MNOs and fixed operators, particularly to improve coverage in rural areas, by acting as a neutral wholesale platform. As the Red Compartida had not been rolled out at the time of writing, it was too early to assess the effect on competition with respect to MVNOs.

Box 4.3. Some examples encouraging competition from MVNOs

Chile: regulation of wholesale access offers for MVNOs

The Chilean market for mobile services is concentrated in three operators that accounted for just over 95% of the market in December 2014. Several MVNOs (full and mainly resellers) offer retail services accounting for less than 5% of the market. The Chilean regulator, Subsecretaría de Telecomunicaciones, or SUBTEL, is preparing regulation to encourage competition by MVNOs, aimed to set reasonable, transparent and non-discriminatory conditions for wholesale access supply to MVNOs.

Additionally, SUBTEL has also included the obligation to publish reference offers for wholesale access use by MVNOs in the tender process for spectrum in the 2 600 megahertz (MHz) and 700 MHz bands.

Source: SUBTEL (2014a), Licitación 700 MHz: SUBTEL define frecuencias para Entel, Movistar y Claro,

Germany: MVNOs’ wholesale agreements as remedies for clearing an acquisition

Telefónica acquired E-plus in 2014. The European Commission had concerns that the merger would result in a reduction of competition between MNOs and also weaken the position of MVNOs, at the expense of consumers. The merging parties had been close competitors, in particular for prepaid customers, by offering a lesser network quality for less expensive rates than Deutsche Telekom and Vodafone. The merger was thus considered likely to constitute a significant impediment to effective competition in the market. As part of the remedies imposed to clear the acquisition, Telefónica agreed to enter into capacity-based wholesale agreements with up to three “Upfront Mobile Bitstream Access MVNOs”. These agreements allow MVNO(s) to purchase, in return for a payment up front, up to 30% of the total capacity of the merged company’s network, for up to ten years after the completion of the merger.

Spain: Obligations to provide wholesale access to MVNOs based on market dominance

In 2006, the Spanish regulator Comisión Mercado del Telecomunicaciones (CMT, now CNMC) declared joint dominance among the three main mobile operators at that time, noting that wholesale network access to third parties had been blocked as the focal point for tacit collusion. As CMT noted, despite the evidence of pent-up demand from service providers seeking wholesale access, none of the established MNOs voluntarily granted access on a commercial basis. In other European markets, meanwhile, MVNOs were able to obtain wholesale access from MNOs. By denying such access, the MNOs managed to maintain a high level of profitability in the retail market, deterring the entry of MVNOs. The three jointly dominant MNOs were required to meet the following obligations: i) the obligation to grant access to their networks on reasonable requests; ii) the obligation to charge reasonable prices. Once the regulatory obligations took effect, MVNOs started to obtain wholesale agreements and compete in the retail market. At the end of 2013, around 30 MVNOs were operating in the country, accounting for 10% of the total market share in terms of revenues.

Mexico: Wholesale shared nationwide 4G network with access for MVNOs and MNOs

The government of Mexico, through the Secretary of Communications and Transport (SCT) is promoting the Shared Network (Red Compartida) to accelerate availability and access to broadband services. This will be a mobile network using at least 90 MHz in the 700 MHz band. It aims to provide wholesale capacity, infrastructure and telecommunication services to other companies and not to end users. It will be designed, deployed and operated using private investment. Service is scheduled to begin in early 2018. The tendering process opened at the end of 2015 with the publication of the bid specifications, and the concession is expected to be awarded in the second semester of 2016.

Source: SCT (2016), Red compartida, (accessed on 23 February 2016).

National roaming agreements

National roaming agreements allow for full coverage at the national level for operators with a limited footprint. It takes a substantial period for any new entrant to establish national coverage. In the interim, it makes sense to utilise a network from another operator in areas where the new entrant has not yet deployed its network. Ensuring national roaming agreements under specific regulations can help new entrants deploy their networks faster and compete against established players. As the costs of deploying mobile infrastructure are in general substantially lower than those for fixed access networks, it is advisable to establish sunset clauses to ensure that such national roaming obligations only remain in force for a reasonable period (typically four to six years) to encourage investment from new entrants to deploy their own infrastructure.

National roaming can also be useful when licences for mobile services are provided on a regional basis. In this case, national roaming agreements allow services to be provided at a national level for regional operators once they have entered into a national agreement with other MNOs. National roaming agreements may also help provide service in rural areas where an MNO believes the financial basis for deployment is lacking.

National roaming agreements are in many cases a natural outcome of the market, and policy makers should ensure that no legal barriers prevent such agreements (Box 4.4). In specific cases, especially in large countries, where nation-wide network deployment can take a long time for new entrants, obligations must be imposed on the dominant operator to provide national roaming for new entrants at a reasonable price in specific areas. Such obligations can be limited in duration, providing an incentive for new entrants to extend their infrastructure. National roaming can also be set as a condition for spectrum licenses, as analysed in the next chapter.

Box 4.4. Selected national roaming cases


The general competition plan (Plano Geral de Metas de Competição [PGMC]), is a comprehensive regulatory package aimed at encouraging competition in relevant telecommunication markets, enacted in November 2012.

Among other issues, Article 43 of the PGMC stipulates that mobile operators in a dominant position publish a reference offer for national roaming, including voice, data and SMS services. These reference offers must include all the technical and commercial information relevant for obtaining national roaming access, including prices and geographical coverage.

Sources: ANATEL (2012), Plano Geral de Metas de Competição (PMGC) – Resolução No. 600 de 8 de novembro de 2012,; Vivo (2013), “Oferta Pública de Referencia de Roaming Nacional”,


The Chilean communications authority, SUBTEL, launched a spectrum tender for the provision of LTE services in October 2013. Obligations in the tender included connecting a number of isolated regions and schools, and the tender set specific obligations for providing national roaming to other MNOs that did not have coverage in specific areas, as well as to provide wholesale access for any MVNO.

Source: SUBTEL (2014b), Bases del concurso público para otorgar concesiones de servicio público de transmission de datos en las bandas de frecuencias 713-748 y768-803MHz,


According to Articles 119-120 of the Federal Law on Telecommunications and Broadcasting (Ley Federal de Telecomunicaciones y Radiodifusión), enacted in July 2014, the preponderant operator or the one with relevant market power has the obligation to provide national roaming on a temporary basis in the geographical areas where the other operator has no available infrastructure. Prices for national roaming services are to be determined by the communications regulator (Instituto Federal de Telecomunicaciones [IFT]), based on a cost model aimed to encourage effective competition and the best international practices.

Source: Mexico (2014), Ley Federal de Telecomunicaciones,


In February 2013, the CRC issued Resolution 4112, which forced operators to provide national roaming as part of their license conditions. National roaming obligations, as laid down by the resolution, include:

  • Availability of a national roaming public offering that includes essential technical requirements, deadlines and service fees, coverage areas and so on.

  • Clear obligations for home and visited network providers.

  • Regulated prices. For voice and SMS services, actual prices must remain below certain cap values.1

1. Actual prices must remain below the caps, which are fixed at the same level as the termination rates (for voice and SMS), and national data roaming prices were set at around USD 0.014 (COP 25.63) per megabyte (MB) in 2013, USD 0.01 (COP 19.36) per MB in in 2014, and USD 0.007 (COP 13.09) per MB from 2015 on (with USD-COP exchange rates of February 2013).

Source: CRC (2013), Resolución 4112 de 2013,

Emerging services

The ICT sector has been one of the most innovative areas in recent years, with many new services and technologies emerging, increasing consumer welfare and reducing the cost of providing better services. It is very important to ensure that an adequate framework exists for innovation and experimentation with new services. ICT policies in general and broadband services in particular must ensure that there are no regulatory barriers for new emerging services, establishing an innovation-friendly market open to new technological advances and service experimentation.

Regulatory authorities should also as a general rule refrain from regulating emerging services too early except where justified, for example, in the case of consumer protection, national security or other specific issues. When new products or services are launched, it is often unclear what the barriers to competition will be. It is thus advisable to let the market stabilise before imposing obligations aimed at increasing competition. This will also help to avoid picking winners or restricting technological developments. That being said, authorities should monitor emerging markets to take action when and if required.

Number portability

Number portability is one of the key tools for competition for fixed and mobile services. It is also a crucial component related to the switching of providers phase within consumer protection frameworks (further addressed in Chapter 13 on consumer protection). A prompt, effective and simple number portability process allows consumers to switch from one operator to another, encouraging competition. When no number portability process is available or the process is slow, cumbersome or fails in a relevant number of cases, customers refrain from switching operators, rendering other pro-competitive measures ineffective.

Although the use of broadband services does not entail the use of fixed or mobile numbering, broadband services are often contracted bundled with fixed and/or mobile voice services, meaning that number portability is also relevant in the context of creating competition for broadband services.

The experience and tools in implementing number portability have advanced during the last ten years. As an example, in many OECD countries, porting numbers can take less than a day. In many of these countries, number portability does not incur costs for customers. The main success factors for number portability are shown in Box 4.5.

Box 4.5. Success factors for number portability

To enhance competition and empower consumers, many regulators have introduced number portability, which allows consumers to keep their number when they switch to another network provider. This is widely regarded as a fundamental prerequisite of open competition and choice in telecommunication markets. For this to be successfully implemented and for transition costs to be brought to a minimum, however, regulators should take into consideration some key success factors:

  • Application of international experience. Regulators should make the most of the extensive existing international experiences on the implementation of number portability, and use, when possible, well-established processes and technical solutions.

  • Involvement of operators. Although regulators define the requirements and key issues to be considered in implementing number portability, the operation of number portability falls on service providers. It is thus crucial that regulators make sure service providers work with them to define and implement number portability processes.

  • Porting time. Long porting times translate into lower numbers of ports. In the initial stages, to ensure a smooth introduction, longer periods of maximum portability time may be acceptable, but regulators should require that in the medium term, the shortest possible time is taken to complete number portability. The shorter the porting process, the better it is for competition and consumers.

  • Porting fees. In many of the LAC countries and other emerging countries currently implementing number porting, a vast majority of subscribers are prepaid users extremely sensitive to prices. High porting fees hinder porting opportunities. If they exist, fees should be minimal for consumers or fall on the operator that receives the numeration. Ideally, no onus should fall on consumers once they have expressed a preference to change provider.

  • Simplified processes for consumers. Number porting should be done in a simple and rapid manner for consumers. The creation of one-stop-shop to request the new operators of choice to implement portability is a good practice.

  • Consumer awareness. Awareness that alternative suppliers and mechanisms for number portability are available is critically important. In some countries, regulators have gone to significant lengths to make consumers aware of number portability, such as through publications in their websites, social media and marketing campaigns. Additionally, accurate information on portability should be given to consumers, before and during portability processes by service providers, and especially immediately after the switching is concluded, and information should be presented in an easily accessible format.

  • Clear guidance. Regulators should ensure that service providers are aware of, understand and comply with all obligations relating to national legislation on number portability and to the best practices that apply to each service.

Customer retention and SIM locks

Some commercial strategies, from mobile and fixed operators, are based on less expensive up-front offers for a service or a bundle of services and/or handheld subsidies, locking in customers by preventing them from switching provider for a fixed duration. In these cases, users can sometimes face high penalties if they switch providers before the end of this period. Although bundling of services and handheld subsidies are not per se poor outcomes for some customers, the practice of imposing lengthy customer retention periods, which often range from 24 to 36 months, can be harmful for competition in the short term.

Good regulatory practice in this area can include:

  • Reducing the time that customers are locked into a contract for services and ensuring that once the initial contract is over, no further lock-in period is enforced, and that customers are allowed to switch provider whenever they wish to do so. Any penalties imposed on customers who change providers during the lock-in period should reflect the time left in the contract period.

  • Where handheld subsidies are provided, the price for the terminal should be transparent on the bill and customers should have the choice, during the contract period, to purchase the terminal outright.

Locking of SIM cards is another practice used by mobile operators subsidising terminals. When SIM lock is in use, the terminal can only be used for the operator providing the subsidy, and consumers cannot use it to access service from another operator. To encourage competition, it is important that regulations require an obligation to provide SIM unlock codes when requested by the customer, once the contract is finished, or after a reasonable period allowing for cost recovery for the operator. Unlocked handsets are another means by which users can select another operator when roaming in another country, such as through purchasing a local SIM card. In the section about consumer protection (Chapter 13), issues related to these issues are also addressed.

In-building wiring

Deployment of broadband services in multi-dwelling buildings is usually undertaken by providing ducts inside the building to access each home or directly deploying cable on the façade of the building. This is usually done by the operator deploying the access infrastructure, although there can be regulations in force to include the corresponding passive infrastructure in new dwellings. The inability of new entrants to access in-building wiring can be a barrier to competition in the broadband market. Where buildings are already wired, it is often difficult for a new entrant to install a competing wire, either because of opposition by the homeowners’ association or because of engineering difficulties.

To encourage infrastructure-based competition and reduce high market entry costs for new entrants, facilitating broadband availability and minimising civil works in buildings, regulatory authorities may:

  • Require building developers to deploy vertical wiring passive infrastructure in new buildings, setting rules on the size of ducts to allow for the deployment of networks from several operators and using different technologies (e.g. fibre, copper), as well as providing chambers for operators’ connection to the in-building wiring (Box 4.6).

  • Regulators should consider promoting in-building infrastructure sharing for fibre cabling for individual apartments, and avoid exclusivity agreements. In this context, some European Union countries have imposed symmetrical obligations for the party deploying in-house wiring based on national law, namely Spain, Portugal and France (BEREC, 2011). In Finland, all in-house wiring belongs to the house owners and is therefore not included in the wholesale market definition. Such arrangements are also in force in Sweden and Korea, which also has an effective framework for labelling new buildings depending on their fibre connectivity. This framework, in Korea, has played a major role in transitioning to “fibre-ready” multi-dwelling buildings. In the LAC region, promoting and regulating in-building infrastructure should focus mainly in urban areas with high-rise buildings. In smaller villages and suburban areas, the emphasis of regulatory action should be on simplifying management of rights of way by municipalities. A summary of some applications of these practices in selected countries are provided below (Box 4.7).

Box 4.6. Passive telecommunications infrastructure in buildings


Korea is one of the leaders in the deployment and take-up of fibre-optic networks in the OECD area. The percentage of fibre connections in total broadband subscriptions reached 66% in June 2014, close to Japan, with 71%. This successful penetration of fibre networks is, however, underpinned by a decade-long effort to enhance the in-building wiring framework for multi-dwelling buildings in Korea, which was seen as one of the main ways to facilitate greater competition in fibre infrastructure deployment. The Building Certification Programme (BCP) certifies that an apartment building complex is equipped with suitable communication infrastructure for fibre-based broadband services. For instance, when every apartment building is connected to at least four optical cables (i.e. in-building facilities such as main telecommunication rooms, ducts and wiring for fibre-to-the-home, or FTTH, services for residents), it qualifies for BCP’s “supreme grade”. BCP has two other grades, first and second, both of which should ensure FTTB connectivity. Unshielded twisted pair (UTP) cables are used for in-building wiring for first- and second-grade buildings and apartments.

The programme can be applied to most of the apartments and major buildings in the country. If a building is compliant, the applicant is awarded a certificate and given permission to publish and advertise the award. Korea’s Building Certification Programme’s certificates, introduced in July 1999, is now the de facto standard for in-building wiring, especially in multi-dwelling residential units, with 28.8% of Korean households complying with BCP (13.2% ranked as “supreme grade”, 60.9% as “first grade”, and 25.9% as “second grade”). The most striking feature of BCP is that it is not based on the regulation of in-house wiring but on competition in the housing market. Current regulations for in-building wiring in Korea do not require facilities for FTTH or fibre-to-the-building (FTTB). The Korean government, assisted by a growing housing market, succeeded in upgrading in-building wiring in most new apartments to the “superior” fibre-based system. Bearing in mind that 58.6% of Koreans live in apartment buildings, it may be noted that this programme has significantly increased competitive access to in-house facilities and prevented apartments from being locked into a single provider, without imposing high costs on residents to change operators.


The Ministry of Transportation and Telecommunications published Decree 18/2004 in 2014, establishing a new regulation for telecommunications infrastructure in new buildings, and setting out technical norms to be followed. Among these norms, telecommunications passive infrastructures in buildings must allow for use from multiple telecommunications operators, enabling co-owners to freely choose their telecommunications provider.


In 2011, the Spanish government updated the regulation defining technical norms for Common Telecommunications Infrastructure (CTI) to be deployed in new buildings. This was intended to support access for telecommunications services, to ensure that all types of access infrastructures were supported (including fibre). The regulation aims to ensure that the passive infrastructure deployed in each new building can accommodate in-house building from several operators.

Source: Minetur (2011), Real Decreto 346/2011,

Box 4.7. Examples of application of in-building wiring sharing practices


The Law on Modernising the Economy (LME, dated 4 August 2008) introduced a system of rights and obligations for operators deploying very high-speed broadband solutions. It sets out specific rules for very high-speed broadband in order to: i) facilitate the deployment of fibre in private premises (inclusion in the agenda of meetings of buildings’ co-owners, recommendations for the agreements between buildings’ co-owners and operators, individuals’ “right to fibre”, and so on); ii) reduce the risk of a local monopoly in the building, through the sharing of terminals, the implementation of which is ensured by ARCEP; and iii) equip new buildings with optical fibres. In particular, the process of installing fibre in buildings is facilitated for operators and imposed on property developers in greenfield housing. The party that installs the fibre in the building (i.e. the building operator) is responsible to the property owner for all operations performed on the network on the private property, and must fulfil the obligation to share its infrastructure, allowing other operators to provide ultrafast broadband services to the residents of the building under fair and non-discriminatory conditions. Furthermore, Article L. 34-8-3 created by the LME stipulates that the concentration point must be located outside of private property, “except in instances defined by [ARCEP]”.


New rules were laid down not only to promote the installation of fibre-optic cable in new buildings but also to avoid monopolisation by the first operator of vertical communications infrastructures in existing buildings by the time the new regime went into force. In this case, the first-building operator to reach an existing building is required to install at least two fibres per home (apartment) and the associated infrastructure is to be shared by other operators (e.g. vertical infrastructure and ODF). For new buildings, the same rule applies, but the responsibility for the installation of the infrastructure and cabling remains with the owner/builder.


The regulation establishes that operators that install in-building fibre cabling shall meet all reasonable access requests, and are obliged to conform with third-party procedures, technical constraints, prices and timing with respect to the provision of access to the fibre facilities installed. Such wholesale agreements must provide for the establishment of technical implementations, so that other operators can share fibre resources under reasonable conditions in terms of costs and prices. To avoid situations in which operators encounter entry barriers, such as being refused access to property or lack of space for additional fibre deployments, the first operator, in installing fibre in buildings, must play the role of manager of the network resources installed. Thus, the first operator is obliged to carry out the tasks required to effectively complete the facilities sharing, such as cabling and installation of the referred facilities for other operators.

Facilitating access to the facilities installed in buildings at a reasonable cost is required, thereby guaranteeing that costs do not constitute a barrier to entry for third parties. Finally, because transparency is essential to allow operators to efficiently arrange and generate access requirements, the regulator requested that a number of information fields in an information system be included for that purpose, such as passed buildings, details on the variety of installation performed and technical data on distribution boxes and fibre.

Passive infrastructure

Passive infrastructure accounts for a large part of the cost of building telecommunication networks and represents a very high part of the sunk costs for network installation. For example, civil works (ducts, poles and so on) account for a 68% of the total of the first-year costs of deploying a new fibre network (OECD, 2008). Passive infrastructure is not only expensive, but time-consuming to deploy, constituting a clear entry barrier for infrastructure-based competition. This issue is especially relevant for new entrants, which, unlike incumbent operators, do not own a pre-existing access network inherited from the monopoly era. For this reason, passive infrastructure deployment and sharing should in general be facilitated and encouraged by policy makers and regulators, provided that lowering costs via infrastructure sharing do not raise concerns about a reduction of competition.

The critical policy area in facilitating passive infrastructure is the regulation of rights of way. Costa Rica and Canada are two examples of good practices in rights of way regulation (Box 4.8). Absent any national or regional regulation on rights of way, local administrations (municipalities) are the relevant administrative bodies in charge of authorisation/denial or imposing conditions (such as fees or dates for obtaining permission) for carrying out civil works in a public space.

Box 4.8. Good practices in rights of way regulation

Costa Rica

Costa Rica sets a delimitation of powers among the different public bodies for co-ordinated and expedited approval required for installation or expansion of telecommunication networks in Decree No. 36159-MINAET-S-MEIC-MOPT. Among other issues, it includes directives for municipalities, periods and due dates for responding to requests of rights of way.

Source: Costa Rica (2010), “Normas Estándares y Competencias de las Entidades Públicas para la aprobación coordinada y expedita requerida para la Instalación o Ampliación de Redes de Telecomunicaciones”,


In Canada, the Canadian Radio-television and Telecommunications (CRTC) has wide-ranging authority to settle disputes between operators, and between operators and municipalities or other public authorities on rights of way, and can make recommendations on expropriating property to ensure rights of way through private land. Although operators need to obtain the consent of municipalities to obtain rights of way, the CRTC can intervene. Where an operator cannot reach agreement with a municipality, the CRTC can grant permission and set the conditions to access the rights of way.

Source: OECD (2008), “Public Rights of Way for Fibre Deployment to the Home”,

The lack of harmonised procedures, rights and duties at a national level may negatively affect plans for deployment not only of fixed, but also of mobile networks. Operators in these cases face a complex situation in which conditions, time and costs for deployment differ in each municipality, unpredictably raising costs and adding uncertainty over the amount of time required to before a network can be deployed. National, regional and local regulations of right of ways overlap, leading to disputes among administrations and adding a judicial burden for operators, and involving additional costs and delays.

For these reasons, it is important for national authorities to:

  • Institute co-ordinated national administrative procedures for access to rights of way, ensuring consistency and predictability in the application of these procedures. It is also important to ensure clarification of jurisdiction both for granting rights of way and for settling disputes and arranging co-ordination among the public authorities involved.

  • Develop a reasonable system of compensation for access to and use of municipal public rights of way. Municipal costs incurred in the associated civil works must be covered, but fees for civil works licenses should be limited, to avoid discouraging investment. Benchmarking with similar successful countries can provide a guide on maximum fees. It is also advisable to ensure that operators investing in ducts are required to conform to obligations for remediation and maintenance of ducts, masts and any other passive infrastructure deployed.

  • Set maximum times for obtaining licences (depending on the type of civil work involved), to reduce uncertainty and help operators to schedule their network deployments.

  • Set one-stop shops for rights of way and related administrative procedures. These can significantly reduce the administrative burden on operators during the planning phase of network deployment, and ultimately lead to greater coverage. Time saved in the planning phase can also help operators to bring in revenues faster and begin to compete as soon as possible.

  • Provide advice and guidelines to municipalities on right of way management. Even when there is an existing national law harmonising procedures, local administrations play a key role in applying the regulation. In many cases, municipalities lack the resources and/or knowledge to apply the corresponding procedures. To address this, clear and concise guides may be prepared by the national authorities, including models for forms and other key aspects.

As in any other policy area, regular contacts and seminars with local administrations and operators can help identify aspects for improvements in the regulation of rights of way, as well as to anticipate future needs and bottlenecks.

Encouraging passive infrastructure sharing is also good policy practice that can significantly reduce initial costs for network deployments, facilitating competition and investment in active network deployments. Additional benefits can also be obtained from the reduction in damage to existing infrastructure during excavation work. Some examples are provided below on encouraging passive infrastructure deployment and sharing (Box 4.9).

Box 4.9. Some examples of encouraging passive infrastructure sharing and deployment

Peru: Law on access to infrastructure owned by incumbent telecommunication operators

In 2008, the Legislative Decree No. 1019 or Ley de Acceso a la Infraestructura de los Proveedores Importantes de Servicios Públicos de Telecomunicaciones was published in Peru. Its goal was to make mandatory shared access and use of telecommunications infrastructure from incumbent telecommunication services providers, so that telecommunication licensees could be guaranteed reasonable, non-discriminatory access to telecommunication infrastructure.

The Decree defines telecommunications infrastructure as poles, ducts, conducts, chambers, towers and other elements of the network, as well as rights of way related directly to the provision of public telecommunications service. It also guarantees that the conditions required by the incumbent for the shared access and use of its infrastructure should not be less advantageous than those required of their own subsidiaries or third parties. Two modalities are possible for infrastructure sharing: either agreement of parties (within 60 days of the request for access) or by decision of OSIPTEL (when the 60 days have expired without negotiation).

Source: OSIPTEL (2008), Decreto Legislativo 1019 Ley de acceso a la infraestructura de los proveedores importantes de servicios públicos de telecomunicaciones,

Bahamas: Infrastructure-sharing regulations

The Utilities Regulation and Competition Authority (URCA) in Bahamas passed a set of infrastructure-sharing regulations in September 2015, setting obligations, procedures and directives on price setting for infrastructure sharing among operators. These regulations also include special provisions for construction, use and sharing of communication towers. According to URCA’s regulations, infrastructure providers (operators owning passive infrastructure facilities) must set commercially negotiated access rates based on actual costs and in accordance with the following principles:

  1. Charging should serve to promote the efficient use of assets and sustainable competition and maximise benefits for customers.

  2. Access charges must reflect a reasonable rate of return on capital employed and take into account the investment made by the Infrastructure Provider.

  3. Access charges must only reflect the unbundled components that the Infrastructure Seeker wishes to use. An Infrastructure Provider must unbundle distinct facilities and corresponding charges sufficiently so that the Infrastructure Seeker need only pay for the specific elements required.

  4. Access charges must be transparent; and

  5. Access charges must be impartial, non-discriminatory and must be no less favourable than those the Infrastructure Provider offers its subsidiaries, affiliates partners or any other licensee.

Source: URCA (2015), Infrastructure Sharing Regulations – ECS 04/2015,

Costa Rica: Public consultation on regulation of infrastructure sharing for telecommunication services

The Superintendencia de Telecomunicaciones (SUTEL), the telecommunications regulatory authority in Costa Rica, has performed a public consultation on specific detailed regulation regarding infrastructure sharing (Resolution RJD-181-2015). This regulation is aimed to promote the efficient use of passive infrastructure, encouraging infrastructure sharing under a clear and orderly model, reducing costs for all actors, and increasing competition for services.

Among other issues, this regulation set specific guidelines on space to be left on ducts when installing them, to leave room for future growth, as well as duct sharing among operators. Specific provisions on access by operators to passive infrastructure deployed in public infrastructure (for example roads, bridges or railways) are also included, to help address the need for inclusion on passive infrastructure deployments in public infrastructure projects. Detailed guidelines on setting charges for infrastructure charging are also included.

Source: SUTEL (2015), Consulta pública al reglamento sobre el uso compartido de infraestructura para redes públicas de telecomunicaciones,

United States: Broadband Conduit Deployment Act of 2015

Members of Congress are discussing the Broadband Conduit Deployment Act, with the goal of mandating that federally funded highway construction projects include the installation of pipes to carry fibre-optic cables, assuming that the area in question has a need for broadband in the next 15 years. The Federal Highway Administration estimates that it is ten times more expensive to dig up and then repair an existing road to lay fibre than to dig a channel for it when the road is being fixed or built. According to a study by the US Government Accountability Office, “dig once” policies can save from 25% to 33% in construction costs in urban areas and approximately 16% in rural areas.

Source: United States (2015), Broadband Conduit Deployment Act of 2015,

Some specific good practices aimed to encourage passive infrastructure sharing are:

  • Setting obligations for dominant operators owning ducts, masts and any other passive infrastructure, to share them at regulated prices with alternative operators, even when the passive infrastructure belongs to a parent company (e.g. an electricity utility).

  • “Dig-once” policies may also be applied, encouraging diverse utilities (gas, electricity, telecommunications and water) to adhere to a common shared planning to dig. This can reduce investments for all parties involved, minimise troubles and inconvenience in the public space and better organise deployment and future maintenance.

  • When planning new public infrastructure, such as highways, it is usually worth investing in ducts that could be used by any operator under open-access cost-based conditions to deploy their own networks. This is especially useful when there is a lack of backbone or backhaul infrastructure.

  • A relevant part of the passive infrastructure deployed by other utilities, such as gas, water or electricity companies, can also be used for telecommunication services. Utility companies performing civil works that are fully or partly financed by public means could be required to meet reasonable requests from telecommunication companies for civil works co-ordination, in order to deploy high-speed broadband networks. This is for example the case in the European Union, where Directive 2014/61/EU, of 15 May 2014, on reducing the costs of deploying high-speed broadband networks, addressed such obligations.

To ensure efficient use and sharing of passive infrastructure, it is critical to ensure that operators have access to accurate information about its availability. This requires development of IT systems showing georeferenced information on this infrastructure, as well as supporting processes for requesting its use, provision and maintenance. When the passive infrastructure to be shared belongs to the dominant operator, the implementation of these systems can be part of the obligations imposed on its access. When passive infrastructure to be shared also includes elements provided by other utilities and/or other infrastructure, the administration should manage the corresponding project to implement and collect data from different organisations. An example of an infrastructure atlas managed and launched in 2012 by the German Regulatory Agency can serve as one example (Box 4.10).

Box 4.10. Infrastructure mapping

In December 2012, a nationwide infrastructure atlas was put into operation by the German Regulatory Agency, the Bundesnetzagentur. The atlas contains spatial data information on existing infrastructure in Germany that can be shared in principle for the construction of broadband networks and to increase the transmission capacity of existing networks. Data on existing passive infrastructure are provided by infrastructure owners from different industries. These include companies in the energy and telecommunications sector, as well as relevant infrastructure in the public sector.

The purpose of the infrastructure atlas is to bring together stakeholders to arrange broadband expansion projects with infrastructure owners. It helps operators access information on the location of relevant infrastructure and obtain contact details for infrastructure owners. On this basis, the user can contact the infrastructure owner and negotiate joint use of the existing infrastructure. The Federal Network Agency has a legal basis for the acquisition of data for the infrastructure atlas, which has hitherto operated on a purely voluntary basis. However, most infrastructure owners have opted to participate voluntarily.

Source: Bundesnetzagentur (2013), “Infrastructure Mapping – The German Infrastructure Atlas”,

Building a complete infrastructure atlas, including georeferenced information and procedural support, is challenging and takes substantial time and resources, not only for the administration, but also for operators and utilities involved in providing data. Some challenging issues in this area include: defining formats for the information to be sent by the different actors who own passive infrastructure, to ensure that the information can be aggregated in a meaningful way; ensuring that the information to be provided by different actors is up-to-date and verified; providing functionalities to allow for co-ordination among actors; restricting access to authorised agents,4 and ensuring the provisions of geographical and technical data that can be useful for operators intending to use the passive infrastructure. A step-by-step process can be followed, starting with simple information and functionalities useful for encouraging infrastructure sharing, incrementally refining the functionality and data that is provided.

Network sharing and co-investment

Network sharing and co-investment strategies are increasingly used by operators to reduce investments and risks associated with network deployment. In the context of mobile services, these strategies have been applied for sharing sites for base stations, including ancillary services such as electricity, masts, antennas or even network elements (Box 4.11). In the context of fixed services, ducts and in-building wiring have also been shared among operators.

Network sharing and co-investment are especially relevant for alternative operators that cannot realistically undertake large-scale deployments in the access network on their own, or when the deployment of new networks or technologies requires substantial investments, which can be shared by several actors. Such agreements can be seen as an opportunity rather than a threat to competition. As noted in Chapter 5, network sharing can be a useful tool to enable governments to achieve the availability of network infrastructure in rural areas or elsewhere given particular circumstances (Box 4.12), and/or to promote an efficient use of infrastructure or investments.

Box 4.11. Mobile network sharing

Mobile network sharing is the generic term used to describe situations where mobile network operators (MNOs) share part of their networks. The term is generally used when larger parts of the network, such as antenna sites and backhaul, are shared, but it can also mean different things. Networks can share many different elements with different competitors, or purchase it from third parties as a service (outsourcing), which can have the same effect as sharing. Other than a spectrum license, which is assigned to a single party by governments, and an operator’s brand identity, there is little in mobile network operations that cannot be shared. In general terms, four forms of network sharing may be noted:

Passive sharing, e.g. sharing of sites, masts and antennas

Site sharing. Finding good locations for antennas can be challenging, particularly in cities where there are few places to erect a high mast or where neighbouring buildings can create shadows on the signal. Planning procedures and site permits can create even more pressure to use the same locations. It is therefore common for antenna sites to be shared between multiple operators, because it may be easier to obtain a permit for an existing site. At the site, the antenna’s base transceiver station (BTS),1 which is at the bottom of the mast and controls the functioning of the antenna, backhaul equipment and other equipment, is still owned by the respective networks. Sites must often be rented from third parties, such as building owners or farmers.

Mast sharing. In the case of mast sharing, the mast, in addition to the site, is shared between the operators. Each network brings its own BTS, backhaul equipment and other equipment. This may require some co-ordination between the MNOs, for example to guarantee the structural integrity of the mast and the location on the mast. As the MNOs use different antennas, and each determines the direction of the antenna, their coverage can also be different. The same companies that offer site sharing as a service to MNOs also offer mast sharing to them.

Active sharing, e.g. radio access network (RAN) sharing

The radio access network consists of the site, mast, antenna, BTS and backhaul. The company managing the site will lease the whole package to an MNO and carry the data to the core network of the MNO. It can use the same radio equipment to broadcast and receive traffic from multiple spectrum license holders. The MNO, however, has less influence over the orientation of the antennas and therefore the coverage of the network. Where and how traffic is broken out to the core networks of the various spectrum license holders is dependent upon the local situation.

Core network sharing

Core mobile network systems are generally not shared between mobile operators. Examples of such systems are the network and switching subsystem (NSS), which carries out switching and mobility management; the High Speed Serial/Home Location Register (HSS/HLR) and systems for data communication (EPC) and for cost optimisation of traffic (UTRAN and GERAN). They are, however often managed or provided as services by third-party service providers, such as Ericsson, NSN, Alcatel-Lucent and Huawei. The ownership and management has therefore been outsourced.2 Cloud- based solutions for core mobile network equipment, where multiple operators can share the same infrastructure, are currently being discussed and developed, but it is not clear how far these systems have made significant inroads into the market.3

Network roaming

Roaming is the term applied when a mobile customer of a network other than that of the owner of the spectrum license makes use of the mobile network. This can be a network from the same country or from another country. In practice, from a technical perspective, there is no fundamental difference between a terminal from the host network or from a visiting network.

1. BTS is also referred to as the radio base station (RBS), node B (in 3G Networks) or, simply, the base station (BS). For discussion of the LTE standard the abbreviation eNB for evolved node B is widely used.

2. The exact construction of the arrangements can resemble business process outsourcing, financial leases, ownership and sub-contracting.

3. SK Telecom, NSN successfully complete Evolved Packet Core virtualisation proof-of-concept, 25 November 2013,

Source: OECD (2014), “Wireless Market Structures and Network Sharing”,

Box 4.12. Network sharing examples

Japan: Tunnel association

In a densely populated country, where infrastructure costs are lower relative to the expected returns, certain types of locations can still benefit from network sharing, including active components. In Japan, for example, tunnels are used to overcome obstacles such as mountainous terrain or terrain associated with urban areas and conurbations. Using tunnels can nevertheless present challenges for MNOs. They may, for example, have limited space for laying cables, and the cost of deploying infrastructure in a tunnel is of course not negligible.

In 1994, the Japan Mobile Communications Infrastructure Association (JMCIA), a public entity, was established to provide a solution for active network sharing in tunnels. Its membership includes all MNOs, major facility vendors and developers. It builds mobile infrastructure shared by those operators in railway, roads and subway tunnels, as well as underground shopping malls. The association provides transmission facilities from base transceiver stations (BTS) to antennas, including the power supply, whereas BTS are separately operated by the MNOs. In the fiscal year ending in March 2014, the association had completed deployments and made mobile broadband services available at 473 points in subway tunnels, covering all underground lines in Tokyo, including 211 points in road tunnels, 82 points in railway tunnels and 765 points in subway stations.

Figure 4.1. Shared facility operated by JMCIA in the subway

Notes: MU = master unit; RU = remote unit. BTS is operated by each MNO.

Source: JMCIA (2016), Japan Mobile Communications Infrastructure Association Website, (accessed on 23 February 2016).

From a policy perspective, it may be noted that the efforts of the association are similar to other successful examples of network sharing, as practised in Sweden. It aims to develop new infrastructure by co-ordination through a joint entity, mainly financed by the operators. Government policy also supports this process. In some cases, the government subsidises the development of shared facilities in less populated areas, although this is proportionally a very small part of the association’s revenue (2.7% in FY2012). The association has also benefited from lower taxation requirements, as it is an authorised public interest entity.

Network sharing in the United Kingdom

The experience of the United Kingdom illustrates how network sharing can evolve and how different forms of sharing can coexist. Site sharing, for example, has been practised in the United Kingdom since the first mobile networks were launched in the 1980s. Coverage was seen as a competitive differentiator, so MNOs were reluctant to make their sites, seen as strategic assets, available to competitors. However, in some cases, sites were shared on a reciprocal basis: that is, each operator provided access to sharing a small number of its sites, in return for access to the same number of their competitors’ sites.

As the third and fourth MNOs rolled out their networks in the United Kingdom, few sites were made available to them for sharing: the incumbents considered their coverage advantage to have too great a strategic value. However, as all of the networks matured, reciprocal arrangements became common across all MNOs. Prior to the award of 3G licenses in the United Kingdom, the regulator secured a voluntary undertaking to provide national roaming to support the entry of a fifth MNO, and in 2003, Hutchison 3G UK (“Three”) launched its 3G service, with its own network in urban areas underpinned by national roaming on Orange’s 2G network. This national roaming arrangement has subsequently been renegotiated, but remains in place today.

In 2007, Three and T-Mobile announced plans for “3G RAN” through a joint venture, Mobile Broadband Network Limited (MBNL). In the same year, Vodafone and Orange announced plans to RAN share both 2G and 3G. Although Vodafone and Orange did not complete their agreement, together with the MBNL RAN sharing, it did mark heightened interest among the MNOs for greater network sharing.

There are now two mobile network sharing arrangements in the UK: Cornerstone, covering arrangements between Vodafone and Telefonica-O2 (“O2”), and MBNL, covering arrangements between Everything Everywhere (“EE”) and Hutchison 3G-UK (“Three”).

The Cornerstone arrangement is comprised of two main elements:

  • radio access network (RAN) sharing: sharing of active network components to establish a single RAN

  • passive sharing: consolidation of the two operators’ existing sites and passive infrastructure into a joint venture company.

The active sharing element of Cornerstone was announced in July 2012, and represents an evolution of the two parties’ previous passive sharing arrangements. Under Cornerstone, the United Kingdom is divided into two geographic zones outside London (east and west) and separate treatment of north and south London. Within each territory, one operator is the “host”, owning and operating the single RAN that is used by both companies. The parties co-operate in each territory under the terms of a managed network services agreement. London is treated as a special case to be split only for 4G. The arrangements also establish a joint transmission network, consolidating traffic over a reduced number of sites to achieve economies of scale in backhaul capacity.

Passive assets (existing sites and passive infrastructure) are consolidated into a joint venture company that owns and manages all of the sites of both operators, nationally and in accordance with the single grid of base stations across the United Kingdom, to serve the combined needs of both operators as efficiently as possible. This was essentially the extension of the existing Cornerstone agreement.

MBNL (EE/Three) was initially set up as a joint venture between T-Mobile and Three in October 2007. MBNL was responsible for consolidating two separate 3G RANs of T-Mobile and Three to deliver, and consequently manage, a combined 3G network for both operators. After the merger to form EE, the merged company replaced T-Mobile in the MBNL joint venture and MBNL’s role extended to consolidating Orange’s 3G cells in the combined 3G RAN.

In contrast to the sharing agreement between O2 and Vodafone, Three and EE share nationally passive infrastructure, active 3G base station equipment (“NodeBs”), backhaul transmission and Radio Network Controllers (RNCs). Similarly, however, to Project Beacon, Three and EE maintain separate backbone and Core Network infrastructures, while each operator uses its own spectrum allocation. EE and Three have subsequently updated their arrangements for 4G to share sites, masts and backhaul transmission, but the two companies will not share active 4G equipment. The two operators will maintain their separate Core Networks.

Source: OECD (2014), “Wireless Market Structures and Network Sharing”,

Network sharing 4G network in Colombia

In 2013, the ICT Ministry issued Resolution 449, which established the conditions and requirements for the allocations of the 4G spectrum (AWS and 2.5 GHz) for mobile broadband services. This resolution set some network-sharing obligations for incumbents (i.e. national roaming and passive sharing according to the general regulatory framework) and also encourages network sharing, by allowing the passive and active sharing of network to fulfil coverage obligations, understanding it as a way to promote the efficient use of resources and infrastructure.

In that context, after the spectrum was finally allocated, Tigo and Movistar set a network active sharing agreement for the deployment of the 4G network to provide broadband services in Colombia. The agreement was reviewed by the national competition authority, the Superintendencia de Industria y Comercio (SIC). In its ruling, SIC understood this operation as a collaboration agreement between competitors that have more precompetitive effects than risks for the competitive conditions of the national market or for collusion among the parties.

Because of the network sharing agreement, both Tigo and Movistar launched 4G services in December 2013, only four months after the spectrum allocation, and announced a reduction of at least six months of estimated time for regular network deployment.

Source: MinTIC (2013), Resolution 449 2013,

Although co-investment agreements and network sharing can in general be encouraged, especially when no financial case can be made for the deployment of several competing infrastructures, such agreements should be monitored. In certain cases, the conclusion of these agreements may facilitate behaviour that might lower the level of competition with respect to fully independent partners (e.g. with parallel networks). In general, agreements for sharing passive infrastructure are less prone to raise concerns about competition and leave room for differentiation among operators sharing ducts or base station sites. One issue worth monitoring is the risk of foreclosure by some operators to others, if they are not allowed to enter into sharing/co-investment agreements. Competition authorities and communications authorities should monitor and take action if any sharing/co-investment agreement risks abuse of market power positions or reduces competition.

Market analysis and dominance

Regulatory decisions to require access to essential facilities, price regulation and other associated measures are usually based on a market analysis carried out periodically under a pre-defined methodology set in the regulatory framework and publicly available to facilitate regulatory certainty (Box 4.13). The aim of this market analysis is to identify situations of “dominance” or “significant market power”. Several varying definitions of “dominance” are applied in competition law and ex ante regulation in different countries. All in some way apply the concept of market power enough to act like a monopoly in terms of the ability to raise prices and/or reduce quality, in order to increase profit independent of other firms in the market or consumer pressure. As addressed below, dominance is not simply an issue in relation to market shares, although high market shares usually point to dominance. Other relevant factors focusing on barriers to entry, market performance or economies of scale should also be considered when assessing dominance.

Box 4.13. References for market analysis procedures

OECD Competition Toolkit

OECD’s Competition Assessment Toolkit helps governments to eliminate barriers to competition by providing a method for identifying unnecessary restraints on market activities and developing alternative, less restrictive measures that still achieve government policy objectives. One of the main elements of the Toolkit is a Competition Checklist, which asks a series of simple questions to screen for laws and regulations that have the potential to unnecessarily restrain competition.

Source: OECD (2011a), Competition Assessment Toolkit,

European Union

Market-analysis methodology aimed to identify dominance situations (known as “significant market power”, or SMP in the European regulation) is enshrined in the European regulatory framework and has been extensively applied in all member countries in the European Union. Communications regulatory authorities in the area are required to observe the “Guidelines on market analysis and assessment of significant market power” (European Commission, 2002). The European Commission also defines the set of relevant markets that are required to be analysed every three years by all regulatory authorities in the European Union. It list of relevant markets has been reduced over time, and the recommendation in force, enacted on October 2014 just five relevant markets, all at the wholesale level, emphasising regulation for wholesale access to solve competition problems at the retail level (European Commission, 2014).

Additionally, the Body of European Regulators for Electronic Communications (BEREC, 2014a) has also published a significant number of reports focused on the application of the market analysis procedures. One of them is focused on practical implementation of geographical segmentation in market analysis. This can be very useful in countries where conditions of dominance in different geographical areas (for example, rural and urban areas) vary widely.

Sources: European Commission (2002), Commission guidelines on market analysis and the assessment of significant market power under the Community regulatory framework for electronic communications networks and services – 2002/C 165/03,; European Commission (2014), Commission Recommendation on relevant product and service markets within the electronic communications sector susceptible to ex ante regulation – 2014/710/EU,; BEREC (2014a), BEREC Common Position on Geographic Aspects of Market Analysis (Definition and Remedies),; ERG (2006), Revised ERG Common Position on the Approach to Appropriate Remedies in the ECNS Regulatory Framework,

Market analysis in LAC area

The OECD published a report in 2015 on competition and market studies in Latin America (OECD, 2015) detailing practices on market analysis applied by communications regulators in Chile, Colombia, Costa Rica, Mexico, Panama and Peru.

Source: OECD (2015), Competition and market studies in Latin America 2015: The case of Chile, Colombia, Costa Rica, Mexico, Panama and Peru,

When no dominance is demonstrated, and the market is deemed to be competitive, there is no need to regulate access to essential facilities for alternative operators. If dominance is found, regulatory measures must usually be taken to ensure competition. A careful analysis should be conducted before setting regulatory measures aimed to solve dominance issues. Regulatory obligations should be proportionate, oriented to solve the key aspects leading to dominance, and imposed only on the operators in a dominant position (also known as “asymmetric” regulation, in contrast to “symmetric” regulation, which is applied to all operators in the sector).

As defined by the ITU, essential facilities in telecommunication network markets may include public rights of way, support structures such as poles and conduits, local loops, telephone numbers and frequency spectrum (infoDev, 2000). New entrants typically require access to these facilities for competition to be feasible. Duplication of these facilities may be either technically difficult, or more often, economically inefficient. Control of essential facilities can give an incumbent numerous advantages over new entrants. For example, an incumbent can use its control over essential facilities to increase competitors’ costs or to discriminate them by providing inferior quality to these essential facilities, making the competitors’ services less attractive to end-customers. As a response to this competition issue, the essential facilities doctrine (EFD) predicts that the owner of an essential or bottleneck facility should be mandated to provide access to that facility at a reasonable price.

The market analysis process is typically divided as follows:

  • Market definition. Definition of the set of products to be included in the market, as well as the geographical scope. It is based on substitutability among different products (for example cable and fibre broadband services are typically included in the same market) as well as on homogenous conditions for competition in a geographical area (for example, very different competitive situations may be found in rural and urban areas).

  • Dominance analysis. Based on different aspects such as market shares, control of essential facilities that are difficult to replicate, financial resources, economies of scale and scope, barriers to entry, potential entry, ability to influence prices and so on.

  • Analysis of potential competition problems. These are derived mainly from two situations – vertical leveraging (this applies where a dominant firm seeks to extend its market power from a wholesale market to a vertically related wholesale or retail market, applying tactics such as bundling or cross-subsidies) and horizontal leveraging (this applies where a dominant operator seeks to extend its market power to another market that is not vertically related). Competition problems associated with vertical leveraging are usually related mainly to a refusal to make deals or deny access, discriminatory use or withholding of information, delaying tactics, bundling/tying, quality and price discrimination, and application of excessive or predatory prices for certain services.

  • Imposition of obligations. When dominance is found, one or more obligations are imposed on the dominant operator/s focusing on areas, which create dominance in the market.

Specific good practices in this area should include:

  • Making the market analysis methodology to be applied transparent, as well as the rules for imposing obligations and including them specifically in the regulatory framework. This will allow for regulatory certainty (to encourage investment), as well as to support regulatory obligations if challenged in courts by operators.

  • As technology and competition trends evolve, market analysis for each market should be reviewed periodically (ideally, the period between market reviews should be specified in the regulatory framework) and obligations for operators must be lifted if no dominance is found in the new analysis, or maintained or adapted if needed. In general, market reviews should be carried out about every three years, depending on the pace of evolution in the corresponding markets and the resources available in the communications authority. It is also important to take into consideration that market analysis must be prospective, in the sense that the most relevant concern is the future, and what is likely to evolve while the regulatory measures will be in place.

  • Experts with knowledge of competition law, economics and telecommunication markets are needed to carry out market analysis. Additionally, data collection is crucial (the authority to request nonpublic information from operators is necessary, and procedures must be implemented to keep confidentiality) and data processing and analysis if needed. This implies that, as indicated in Chapter 2, the communications authority must ensure that experts and resources are available to carry out periodic market analysis.

  • Declarations of dominant positions should be based on a set of parameters that take into account the structure of the market. In some countries in the LAC region, a reliance on data from market shares may result in an oversimplified analysis. This can lead to false positives when relatively high market shares are not based on advantages derived from a dominant position, or false negatives that inhibit the regulatory measures needed to encourage competition.

  • Emerging services should not, in general, be subjected to regulatory obligations, and any dominance analysis should be undertaken once the market becomes sufficiently stable.

  • Geographical segmentation of markets can also be used to analyse whether competition exists in a market, if it is deemed that the competitive situation differs significantly in different geographical areas. In some cases, urban areas enjoy a competitive situation, where several operators use their own infrastructure, and none exhibit a dominant position. Rural areas, by contrast, may have a single operator with a clearly dominant position, which may require regulatory remedies. If this is the case, market analysis can be segmented geographically, adapting obligations according to each situation. When such differences exist, relevant geographical data is needed to make the best decisions. As for other regulatory measures, a cost-benefit analysis should be conducted to ensure that the outcomes adequately address the problem identified.

  • Preliminary market analysis must be subject to public consultations, providing an adequate period and procedures for stakeholders’ feedback. This will allow for the correction of errors, resolution of misunderstandings and in general, increasing the robustness of the market analysis. Additionally, many of the issues that may be challenged in the courts can be anticipated and addressed in advance. Once the public consultation is finalised, it is also good practice to include responses to comments in the final market analysis.

With one or several operators in a dominant position, competition problems may arise, hindering or even deterring competition from alternative operators. One potential problem is vertical leveraging, where the dominant operator uses its market power at the wholesale level to leverage its position at the retail level. This is the case of dominant operators with exclusive access to a network where no other infrastructures are available. Horizontal leveraging may also occur. For example, a dominant position in one service (telephony) can be used to extend dominance to other services (broadband access), via bundling or cross-subsidies. Direct entry deterrence may also take place when access to essential facilities for alternative operators is simply refused. Moreover, in general, other problems typical of monopolistic situations such as exploitative behaviour or allocative inefficiencies may also result.

When facing dominance situations, Communication Authorities must impose regulatory obligations aimed to solve the specific competition problem that are consistent with the objectives set out in the regulatory framework. Good practices for regulatory obligations include the following:

  • The set of potential obligations for dominant operators should be listed in the regulatory framework, to provide stakeholders (and dominant and alternative operators in particular) regulatory certainty.

  • When selecting the most appropriate and effective remedies, the communications authority must select the least burdensome and simplest to apply, to avoid imposing high costs that could reduce consumer satisfaction, reduce incentives for investments, and make regulatory monitoring and enforcement more complex and uncertain.

  • A careful analysis should be undertaken of the implications of the obligations for static efficiency (in general related to short-term reduction of prices for final services) and dynamic efficiency (related in general with incentives for investment, innovation and long-term sustainability). Static efficiency in the short term must not compromise medium- and long-term efficiency.

  • Any regulatory measure must be imposed with the aim of developing sustainable, long-term competition with minimal regulation in future. One of the objectives of regulatory authorities is to avoid the need to regulate in the long term, by developing sustainable competition (usually infrastructure-based competition, but not necessarily always if there are enough incentives for infrastructure sharing among actors). This does not mean that complete infrastructure-based competition must be the only focus in the short term. Service-based competition based on use of incumbents’ essential facilities by alternative operators (for example, local loop unbundling) may be a way to facilitate infrastructure-based competition in the long term. Alternative operators may then invest using revenues from service-based competition, to reduce their reliance on competitors’ infrastructure.

  • Encouraging competition at the retail level usually means that regulatory action should be taken at the wholesale level, avoiding retail price regulation that may protect consumers, but in general does not encourage competition. This means that communication authorities must look for competition problems at the retail level and impose regulation, if needed, at the wholesale level. If competition problems are found at the retail level, regulatory measures should be in general aimed to ensure access for alternative operators to wholesale inputs, in order to allow them to replicate incumbent’s retail offers and compete in price and quality at the retail level, using not only wholesale inputs from the incumbents, but also their own infrastructure.

  • The imposition of burdensome obligations on dominant operators or any real pro-competitive regulatory asymmetric measure may be challenged in courts, implying that the regulatory framework must clearly provide adequate powers for the imposition of regulatory obligations.

Regulation at the retail level

Regulatory measures at the retail level are aimed at imposing specific obligations on maximum prices (cap prices), approval of tariffs or restricting any behaviour from the dominant operator to encourage competition. As noted, the main regulatory tools to solve dominance problems are in the scope of wholesale services, but it is worth analysing some of the regulatory tools used in the region, as well as in OECD countries in other regions.

Price-cap regulation is in general applied to avoid high charges for customers when the market is not disciplined by competition. It has been applied (and it is still applied in some countries in the LAC region) for fixed and mobile services. Price-cap regulation is also applied in certain cases in the context of extending broadband services (see Chapter 5) to encourage broadband use by communities with low incomes (see Chapter 6, on affordability, government charges and digital inclusion).

Price approval requirements by the Communications Authority are also applied in a small number of countries in the LAC region. In general, this model is cumbersome for both the operators (who have to wait for regulators’ approval before launching a new service) and for the regulator, which must ensure a prompt response and devote resources to this offers analysis.

One context remains in which price approval requirements are a useful tool, and where they are applied in many OECD countries, is for margin squeeze tests. In national markets where competition from alternative operators is based on the use of wholesale inputs provided by the dominant operator, there is a risk of anti-competitive practices based on margin squeeze for alternative operators. The dominant operator may make offers that cannot be reproduced by alternative operators, using wholesale inputs at regulated prices to reduce the number of competitors in the long term. When there is a risk of margin squeeze for alternative operators, dominant operators are requested to send their planned offers for approval before making them public (typically one or two months before, to allow for analysis by the communications authority). The communications authority performs a “margin squeeze test” to check replicability by alternative operators using regulated wholesale inputs and applying a publicly available methodology. If the offer is not replicable, the dominant operator is allowed to launch the new retail offer only if it reduces the price of the wholesale inputs. Otherwise, the offer cannot be launched.

The application of margin squeeze tests is a good practice to be considered in dynamic markets with alternative operators intensively using wholesale inputs. However, its effective application demands considerable resources, and sometimes it is complex to forbid the application of low price offers by the dominant operator, which, although they may harm competition in the long term, are viewed as attractive by some consumers in the short term.

Margin squeeze tests are particularly helpful in addressing competition problems arising from the bundling of telecommunication services if a dominant position exists, since they can provide a given element for the services of others. Here, alternative operators need to be able to compete by offering their own bundles, using their own infrastructure combined with regulated inputs from the dominant provider. This is different from predatory pricing (selling below cost), which is forbidden by competition law in most countries. In applying margin squeeze tests, it is especially important to adequately select suitable cost-accounting models, as well as to get relevant and updated data on operators’ actual costs.

For example, in 2013, the European Commission adopted Recommendation 2013/466/EU on consistent non-discrimination obligations and costing methodologies, to promote competition and enhance the broadband investment environment. This requires that European Communication Authorities impose obligations to ensure the economic replicability of retail offers of operators with significant market power (SMP), thus preventing them from engaging in margin squeeze practices. Based on the experience of communications authorities, the Board of European Regulators for Electronic Communications Services (BEREC) has developed a guidance document to provide information on how to conduct the economic replicability test and more generally ex ante margin squeeze tests in practice (BEREC, 2014b).

Retail price regulation in the context of mobile services is seldom used. Ensuring that there are enough competitors using their own infrastructure (e.g. typically three to five MNOs in countries with thriving competition) and MVNOs adding competitive pressure is usually enough to allow the market to compete in prices and quality of service. Still, in some countries, on-net/off-net retail price differentiation applied by operators with a high market share may present challenges. This differentiation is applied with the aim of retaining market share, discouraging consumers from switching to a new entrant with low market share, when new entrants face high costs for termination in the network from operators with high market share. As a consequence, on-net/off-net differentiation may harm competition.

On-net/off-net differentiation competition problems may be addressed at the wholesale level by setting cost-oriented low termination prices, making differentiation less profitable and allowing new entrants to set low prices for both on-net and off-net calls. Where the differences between retail prices for on-net and off-net calls by operators with high market share is high and restricts competition from new entrants, regulatory action to forbid this practice can be considered.

Regulation at the wholesale level

In the OECD area, wholesale regulation for essential facilities has been extensively used, and many countries are still applying regulatory measures, following the model of the ladder of investment (Box 4.14). Tools include using local loop unbundling and bitstream access, and even extending it to fibre unbundling. Other countries have lifted wholesale regulation, relying more on infrastructure competition. For those applying such regulations, a wide range of wholesale access products can be provided by dominant operators.

Box 4.14. The ladder of investment model

The ladder of infrastructure investment model assumes that investments are made in a step-by-step manner by new entrants. The model’s premise is that in order to allow new entrants to gradually (incrementally) invest in their own infrastructure, they need a chain of (complementary) access products to acquire a customer base, by offering their own services to end users based on (mandated) wholesale access. In those instances where duplication of access is not considered feasible, promoting service competition is an important goal for the regulator, because service and infrastructure competition are not in opposition. They are linked through the ladder of investment, allowing competitors, through a sequence of regulated access products, to invest incrementally in their own infrastructure. Once they have reached a critical mass, they will deploy their own infrastructure, making them less dependent on the incumbent’s infrastructure. This involves migration from one access product to another (moving to the next rung), as the entrant progresses through several stages of competition in ascending a “ladder” of infrastructure investment (Figure 4.2).

Figure 4.2. The ladder of investment

Source: ERG (2009), Report on Next Generation Access – Economic Analysis and Regulatory Principles,

Many LAC countries show a clear need to encourage competition in the broadband market, so that consumers can benefit from lower prices and more investment. Wholesale regulation has been applied in some countries, but it has not always been effective, due to high wholesale prices, lack of technical detail on products offered, or active intervention by dominant operators to render it ineffective.

According to the ladder of investment model, wholesale access products that are usually regulated include the following steps:

  • Resale is the simplest way to start competition. Reselling is based on commercialisation of complete services provided and operated by the dominant operators. This does not involve any network investment, or any differentiation by new entrants, except in prices dependent on commercial agreements (if prices are not regulated) or prices defined by the regulator. As resale has a limited effect on competition and leaves no room for differentiation, wholesale regulation is not usually applied, except in certain cases forbidding non-reselling clauses set by operators.

  • Bitstream consists of providing broadband access for alternative operators using the access infrastructure of the dominant operator, as well as part of the IP network, delivering the bitstream signal for the alternative operators at central and/or regional points. Depending on the characteristics of the bitstream services considered in the regulation, the alternative operator may have more or less room to differentiate, but any retail service offered must be based on what the dominant operator provides. The use of bitstream wholesale access implies some network deployment by the alternative operators, since they must reach the IP connection points of the dominant operators, as well as connect with the Internet. In general, it is not as effective for competition as other wholesale inputs, such as copper loop and fibre access unbundling in urban areas where access unbundling is economically feasible for new entrants. However, bitstream can be important for competition in countries where it is applied in rural and suburban areas. These areas have low economies of density (the investment for reaching a switch is too high for a small number of customers). In such cases, which typically have only one copper access network and often no cable operators, this is the only way to enable competition. Using bitstream access in these areas does also allow alternative operators to complete their coverage and provide services in a wide area or the whole country.

  • Local loop unbundling (LLU) allows an alternative operator to rent the copper access loop connecting the access switch and the customer’s premises. Local loop unbundling may be full (comprising all services) or shared (where the alternative operator uses part of the frequencies for a set of services and the dominant operator uses the access loop for other services). Contrary to bitstream access, alternative operators must deploy their own network infrastructure for each access switch where local loops are rented, implying more infrastructure deployment and investment, but also more room for differentiated services. Competition based on LLU has been extensively applied in many OECD countries, such as all those in Europe and to a greater or lesser extent in the Asia-Pacific and North American regions. It has been less used or superseded in countries where there is judged to be sufficient infrastructure competition (e.g. Korea where high population densities and apartment living enable multiple players to provide fibre to the basement of these buildings). On the other hand, it has proven very successful in countries such as France and the United Kingdom enabling these countries to lift their penetration rates through the additional competition it enables. LLU use is economically feasible in areas where there are a large number of local loops connected to a switch, as investment is required to reach the switch and equipment to collect traffic must be co-located in the dominant operator’s premises. In the context of broadband services, LLU can be effective in increasing competition in the range of speeds provided by xDSL. This does, however, depend on the length of copper loops, as speed and QoS degrades with distance covered. Further details are shown in Box 4.15.

  • Fibre unbundling is the evolution of LLU applied in the context of next-generation networks based on fibre access. In this case, the dominant operator must provide wholesale access to the fibre access that connects the customer with the access switch. Fibre unbundling is not technically feasible under some specific network architectures, and an alternative wholesale product, Virtual Unbundled Local Access (VULA) is considered by regulators in this case to facilitate access for alternative operators. Although fibre unbundling is regulated in many European countries, it is still in its early stages of implementation.

  • Leased lines, Ethernet services and high speed bitstream services. These wholesale products are designed to increase competition in the business market segment. These customers require high speeds, a high quality of service (including guaranteed speed), as well as connections in several different premises, not all located in urban areas, but also in foreign countries. In some cases, the market for business services is highly concentrated and a single operator accounts for a large market share. This usually entails high prices for business services, which in turn significantly increases costs for business operations, especially those involved in the digital economy, discouraging growth. When dominance is detected and barriers to competition are high, establishing obligations to provide wholesale access for leased lines and high-speed IP services (for example, Ethernet services that are increasingly used as a substitute for leased lines) may be advisable. This can allow alternative operators to complement their own infrastructure (typically based on fibre access intended to provide high speed/high quality IP services for the business segment) in areas where the only available infrastructure is deployed by the dominant operator.

  • Dark fibre wholesale access is typically provided not only by dominant operators, but by other actors. These may include utilities (electricity, water or gas providers), as well as transport providers, for example railway operators using their own passive infrastructure to deploy fibre for their own use. For this reason, dark fibre access is not generally regulated, as many different actors may compete to provide these services.

  • Passive infrastructure access. As noted in the section on rights of way, passive infrastructure, and especially ducts, is the least replicable infrastructure asset owned by dominant operators. Even when rights of way are well managed by public administrations, deploying ducts is expensive and slow. Any operator that has to deploy its own ducts would necessarily take a long time to start competing with dominant operators, and deploying access networks with a national footprint would take decades, as was the case for the PSTN. As a result, facilitating access to the dominant operator’s passive infrastructure is fundamental to enable competition with a lower level of investment and less time for deployment. Passive infrastructure sharing allows also for complete differentiation in terms of services provided, as each operator deploys its own network, increasing room for competition. Passive infrastructure wholesale access obligations are usually structured around a reference offer to be prepared by the dominant operator, under the monitoring of the regulatory authority. This should include prices for the different assets to be shared, procedures for service, maintenance, incident management, as well as address compromises on times and the penalties to be paid if conditions for providing the service are not met. Key factors for success involve guaranteeing that alternative operators know the georeferenced location and capacity of passive infrastructure assets, and ensuring non-discrimination compared to self-supply by the dominant operators and capacity management for existing ducts. This usually implies the implementation of a passive infrastructure database accessible online by alternative operators.

  • Access to submarine cables. The market for submarine trunk connection is in general limited to a few operators. Entry barriers are high for new entrants, and in many cases, infrastructure competition based on competing submarine cables covering similar routes is not sustainable. In cases where no alternative infrastructure exists (as in the case for islands with a single submarine cable, or continental areas with no fixed backbone alternative infrastructure), cable operators are not only in a dominant position, but also act as a monopoly. This can lead to high wholesale prices or even refusal to provide wholesale services, also raising retail prices for broadband. In such cases, regulation is needed to enforce access to alternative operators at regulated prices (usually cost-based), and a reference offer must be prepared by the dominant operator.

Box 4.15. Local loop unbundling models

Full unbundling

Full unbundling (sometimes referred to as access to raw copper) occurs when the copper pairs connecting a subscriber to the main distribution frame (MDF) are leased by a new entrant from the incumbent (Figure 4.3). The new entrant takes total control of the copper pairs and can provide subscribers with all services, including voice. The new entrant can also enhance the copper pairs by adding asymmetric digital subscriber line (ADSL) technology. The incumbent still maintains ownership of the unbundled loop and is responsible for maintaining it.

Figure 4.3. Full unbundling scheme

Line sharing (shared access)

Line sharing allows the incumbent to maintain control of the copper pair and continue providing some services to a subscriber, while allowing an access seeker to lease part of the copper pair spectrum and provide services to the same subscriber (Figure 4.4). Line sharing allows the incumbent to continue to provide telephone service while the competitor provides broadband (xDSL) services on the same copper pair. With line sharing, the competing supplier uses the non-voice frequency of the loop. Consumers can obtain broadband service from the most competitive provider without installing a second line.

Figure 4.4. Line sharing scheme

Sub-loop unbundling

Sub-loop unbundling is a much more far-reaching and complicated regulatory measure than LLU. It makes it possible to gain access to the incumbent’s network on an unbundled basis closer to the customer than at the MDF, that is, at a point between the customer’s location and the incumbent’s site. For example, this arrangement can be used to supply very high bandwidth services that can only be transmitted at a short distance on copper pairs.

Source: OECD (2003), “Developments in Local Loop Unbundling”,

Each of the wholesale products described in the previous subsection cover different needs and have different impacts on competition. In general, they can be combined, provided that prices are set according to product’s characteristics (for example, local loop unbundling should be less expensive than bitstream access, as it involves fewer resources from the dominant operator and more investment by the alternative operator). In many OECD countries and some of those in the LAC region, most or all of them are laid out in national regulation:

  • Bitstream access is used to rapidly spark competition with a low level of investment by alternative operators. Typically, many alternative operators begin by using bitstream access, and evolve toward local loop unbundling, which offers more room for differentiation in areas where local loop unbundling is economically sustainable. Bitstream access is helpful for encouraging competition in rural areas where local loop unbundling may not be profitable for alternative operators.

  • LLU is one of the most useful regulatory tools for encouraging competition in urban areas. Where there are dominant operators, it can offer fixed broadband access from alternative operators without their own access infrastructure. Although LLU is difficult to implement, it can help cultivate competition and allow alternative operators to develop networks. This can increase competitive pressure in markets where the dominant operator enjoys a large market share and prices are high, as in Mexico. The Instituto Federal de Telecomunicaciones (IFT), the Mexican regulator, introduced this obligation, together with resale and bitstream access, in 2015 (Box 4.16).

  • Passive infrastructure access and especially duct access encourages infrastructure-based competition so that the competing operators can deploy their own active network infrastructure. This encourages complete differentiation and innovative services based on different network technologies. Duct access is useful not only to boost competition for fixed broadband access, but also to encourage competition in mobile broadband. The increased penetration and use of mobile broadband has increased the traffic handled by base stations. When mobile operators can access existing ducts, they rent them out to connect base stations to their core networks deploying their own fibres. Additionally, mandating duct sharing for the dominant operator may reduce the need for civil works and digging, reducing any inconvenience to the public at large. In general, imposing obligations for duct access encourages infrastructure-based competition, reducing the need for investment while preserving incentives for innovation.

  • Wholesale access obligations for submarine cables are generally needed if no other alternatives exist for operators to cover the route. This can help avoid high prices and/or refusal of access that would reduce competition and increase prices at the retail level. However, while wholesale access obligations are important when no alternatives exist, encouraging (and in some cases funding) alternative infrastructure (other submarine cables, or when possible, fixed backbones) can let competition develop in the long term.

  • Leased lines and high-quality IP wholesale products are intended to help encourage competition for business services. They should be considered when a dominant position for these services has been established. Products, configuration and competition dynamics for the business services markets are usually very different from those for mass markets, and demand specific analysis. In any case, high prices and/or low quality for business products influence productivity and competitiveness.

  • Price regulation for reselling is seldom used, as its influence on competition is very limited. Dark fibre is usually provided by entities such as utilities, which are not subject to telecommunications regulation.

Box 4.16. Local loop unbundling in Mexico

In December 2015, Mexico’s national regulatory authority for telecommunications, Instituto Federal de Telecomunicaciones (IFT) approved a reference offer on local loop unbundling for the access network of Telmex. The dominant telecommunication operator in Mexico, Telmex has a market share of more than 50%. This regulatory measure was one of the obligations imposed on Telmex as a result of its preponderant position in the services sector, as determined by the IFT in 2014.

The primary purpose of this regulation is to allow competitors to access Telmex’s network on a non-discriminatory basis, in order to replicate the retail services it provides. By lowering entry costs for competitors, this measure is expected to promote innovation, new services and to lower prices for consumers. The local loop unbundling reference offer includes wholesale line rental (voice and Internet) and bitstream access services (at a local, regional and national level) as well as full and shared local loop unbundling.

The IFT also determined the rates that allow competitors to supply competitive services, while allowing Telmex to cover its incurred costs. Rates are based on cost models using the “retail-minus” methodology for resale and bitstream access, and the long-run incremental cost (LRIC) methodology for local loop unbundling and co-location. IFT authorised discounts on the retail prices of Telmex as follows: 28.9% for voice line resale, 24.5% for internet resale and 54.9% for bitstream access. Fees authorised by IFT for the local loop unbundling services are 89.1% lower than the services retail prices of Telmex for the full unbundling services, and 98% lower in the case of shared-access services. Local loop unbundling in Mexico enforces Telmex to make available information about its infrastructure to competitors, especially related to the local loop. Resale and bitstream services are available at a national level, while local loop unbundling services are feasible at Telmex exchanges with at least 5 000 lines in the 24 largest cities of Mexico.

Local loop unbundling in Mexico is technically feasible only for the legacy network of Telmex. Given that the operator’s fibre-optic network is based on Gigabit-capable passive optical network (GPON) technology, physical unbundling is not feasible. Thus, its access is available only through bitstream access and is limited to Telmex profiles offered to its end users. The reference offer for local loop unbundling (OREDA) is applicable until 31 December 2016, and Telmex is required to submit a new reference offer to IFT by 30 June 2016 at the latest.

Source: IFT (2015), OREDA: Oferta de Referencia para-la Desagregación del Bucle Local,

In general, bitstream, local loop and fibre unbundling are usually used to encourage competition for fixed broadband access when there is not enough infrastructure-based competition (for example, the competitive pressure exerted by cable operators) and when competition is not expected to develop given the absence of these obligations in the medium term. In certain cases, mobile broadband can act as a complete substitute for fixed broadband. Competition from mobile providers should also be taken into account in designing regulatory measures to impose wholesale access for essential facilities for competition in the broadband market. At present, in most situations, especially in urban areas with short loops, the speed and quality of service for fixed networks cannot in any case be wholly replicated by mobile infrastructure.

Wholesale regulation is key in encouraging competition in markets where one operator accounts for most of the market share. It is also critical where there is insufficient infrastructure competition and new entrants need access to essential facilities that are not replicable in the short term, as is true of most LAC countries. However, wholesale access is not regulated in many cases, and in others, while at least some of the wholesale products are regulated, the obligations are not effective. This is especially relevant in the context of fixed networks, where access network deployment demands time and resources. The key wholesale access obligations are focused on facilitating access to bitstream services, local loop unbundling, leased lines and other high-quality wholesale products aimed at the business segment, passive infrastructure access and trunk segments based on submarine cables. Regulating wholesale access implies setting prices as well as technical conditions and procedures for access.

Wholesale prices for dominant operators should be set under a cost-based assumption. This means that the regulatory authority must perform a careful analysis of the costs involved, so that the wholesale inputs can be regulated. Regulatory cost-accounting is a complex discipline involving specialised knowledge and extensive data. Additionally, setting wholesale prices too low discourages investments by the dominant operator subject to the obligation, while setting regulated prices too high renders the obligation ineffective, and discourages alternative operators from using these wholesale inputs. The key good practices for setting prices include the following:

  • Experts on regulatory accounting must be involved in price setting for wholesale access. Knowledge and experience in applying different cost models is needed.

  • Extensive information on the accounting for the services involved is needed and must be provided by the dominant operator on a regular basis. This information must be audited by a third party, and the regulator must also review the key aspects of the information provided by the dominant operator.

  • Depending on the information available and the context and purpose for the wholesale regulation, different cost models can be applied when performing cost-accounting exercises (Box 4.17). Top-down modelling attempts to measure costs starting from the firm’s actual costs, as set out in its accounts. This method does not involve detailed network modelling. Instead, a top-down model separates the firm’s assets and costs into service groups, and then adds the extra costs associated with interconnection, to arrive at an estimate of LRIC. The bottom-up approach develops the cost model on the basis of the expected demand in terms of subscribers and traffic and sets the network design and estimates the related costs on the basis of a network engineering model. There are advantages and disadvantages of both models (Table 4.1). Selection on the cost model to be applied should take into consideration data available, resources available, as well as the impact on competition and investment for both access seekers and access providers.

  • Regulated prices must be reviewed periodically to adapt to the evolution of costs, as well as effects of the wholesale regulation in the evolution of the market. When there is a significant difference between wholesale prices applied at the moment of enacting the regulation and costs obtained from the cost accounting exercise, it can make sense to define a glide path to facilitate a smooth non-disruptive adaptation of wholesale prices from the existing situation to the objective situation. Glide paths prices are defined setting descending prices along time. This is, for example, the model applied for reduction of termination rates by a number of member states in the European Union.

  • The information on how regulated prices are set must be published (with the exception of confidential data from the dominant operator) and be subject to public consultation. This helps improve the methodology, correct errors and anticipate potential conflicts among operators or judicial challenges by any of the stakeholders involved.

  • Well-defined technical specifications and procedures covering all the areas are as relevant as prices and very important to ensure prompt and effective access. When technical specifications are not well defined, there is room for the dominant operator to restrict access and make it ineffective. Related good practices in this area are:

    • Impose an obligation on the dominant operator to publish a reference offer where all technical, procedural and price issues are covered. This reference offer should be aligned with the obligations required, address all the issues defined by the regulator and be approved by the regulator before its publication. A deadline must be set for the publication of the reference offer.

    • Given the complexity of the contents of a reference offer, it is good practice to have regular meetings with both the dominant and the alternative operators while drafting the reference offer. This can help ensure that all aspects are well covered and that the reference offer can be used effectively by alternative operators.

    • The definition of processes and deadlines for each step is key in ensuring that issues are solved in time, do not delay provision of essential facilities and ensure effective operation (including incident management and maintenance). Such processes should also include providing information regularly to the regulator and alternative operators about key performance indicators, as well as penalties for not meeting objectives. To make these processes effective, specific IT systems may be needed, so that key processes, such as provisioning or incidence management, can be automated as far as possible.

    • Non-discrimination is an important issue that needs to be addressed and monitored by the regulator. In principle, alternative operators must face the same conditions when accessing the dominant operator’s wholesale products as the dominant operator’s retail business faces. Terms for access, prices and times for provisioning and incidence management should be as similar as possible.

    • Ensuring that reference offers are effective for competition is a process that usually takes many years. Experience from OECD countries shows that reference offers are incrementally improved based on experience of their application, as well as on conflict resolution between the dominant operator and alternative operators using the reference offer. Fortunately, extensive experience of preparing and tuning reference offers in OECD countries is available. That being said, context conditions differ from country to country, and adaptation to the local context is always needed.

    • Benchmarking with other countries, and maintaining a network of contacts among regulators in different countries, is a valuable tool for refining reference offers, as well as for gaining useful experience. The Board of European Regulators for Electronic Communications Services (BEREC) is a forum for sharing experiences of regulatory action, and allows for an exchange of views and experiences. The LAC area could also benefit by building network of regulatory experts to exchange information.

Box 4.17. Key concepts applied when regulating cost-oriented prices

Broadly, the key concepts in the regulator’s access pricing toolkit are:

  • Cost-oriented prices. As required by the World Trade Organization (WTO) Reference Paper, these can be developed from bottom-up or top-down cost models or from benchmarking rates in similar countries that have used cost models.

  • Cost models. Bottom-up costing for long-run incremental costs (LRIC) where a firm sets prices to cover only the incremental costs of the product (i.e. the product’s LRIC), and sales of that product make no contribution to the firm’s common costs. These cover many variations, but it is sufficient to consider LRIC to understand the issues and principles.

  • Regulatory accounting is top-down costing associated with fully distributed costs (FDC) where all costs, including joint and common costs, are fully allocated to all the operator’s services/products according to a specified distribution/allocation key. The costs of a given service/product are composed of direct volume-sensitive costs, direct fixed costs and a share of joint and common costs.

  • Benchmarking. This compares access prices across a peer group of countries to determine what price would be reasonable.

  • ECPR, the efficient component pricing rule, is closely related to “retail minus avoided retail costs”. ECPR is cost-based because it includes opportunity cost.

  • BAK. Bill and Keep has been used for mobile termination in countries with “receiving party pays” (as in the United States) and seems related to “peering” in Internet traffic exchange.

  • GB is volume-based charging. This is a possible alternative access pricing to address changes in the industry that BAK cannot address.

Related technical concepts:

  • DAC (depreciated actual cost) – based on historic cost accounting (HCA). Some regulators also require current cost accounting (CCA), in which assets are revalued at replacement cost; which may then require further adjustment to “mean equivalent assets”.

  • DORC (depreciated optimised replacement cost) – takes accumulated depreciation from ORC (optimised replacement cost) calculated for TSLRIC (total service long run incremental cost).

  • SAC (stand-alone-cost) – the sum of the incremental cost of the product, plus all the costs which are common between that product and other products. The stand-alone cost is therefore higher than long-run incremental cost (LRIC).

  • WACC (weighted average cost of capital) – derived from the capital asset pricing model and used to set the return to capital to be applied when setting capital costs.

Detailed information on the different cost models and additional references can be found in other toolkits, as the ITU/InfoDev ICT Regulation Toolkit or the World Bank toolkit “Broadband Strategies”.

Source: ITU (2016), ICT Regulation Toolkit,

Table 4.1. Comparison of bottom-up and top-down cost models




Can model the costs that an efficient entrant would face.

Flexible: can change assumptions readily.

Transparent: much of the information used is publicly available.

Incorporates actual costs.

Useful for testing results from bottom-up model.

May be faster and less costly to implement, but this depends on how well categories in the final accounts match the data required.


May optimise too much or omit costs. If this happens, the operator will be undercompensated and will reduce investment in the network.

Modelling of operating expenditure is usually based on simple margins instead of real-world costs.

Data needed for the model may not exist.

The modelling process can be time-consuming and expensive.

Includes actual costs, which are likely to incorporate inefficiencies.

Less transparent: confidentiality issues mean that other stakeholders may not have access to the information used.

The parties may dispute the cost allocation rules used to allocate shared and common costs among specific services.

Data may not exist in the required form.

Source: ITU (2016), ICT Regulation Toolkit,

Functional and structural separation of operators

Functional separation, sometimes referred to as operational separation, aims to achieve non-discriminatory conduct of an operator with significant market power in providing access products and in downstream competition. It requires a dominant operator to separate, but not sell, its network infrastructure (or wholesale services) from its retail services division. The key feature of functional separation models is that the network provider is required to operate at arm’s length from downstream service operators, providing competitors and the incumbent’s own retail operations with non-discriminatory, equivalent services.

Structural separation goes further than functional separation. It requires the separation of a vertically integrated firm, not only operationally but also in terms of ownership, into a company owning the local access network, providing wholesale access (the network operator); and the rest of the company that provides retail services. The separation of ownership is intended to eliminate the incumbent’s incentives to discriminate.

Functional and structural separation are complex to implement and can substantially affect the structure of the dominant operator. For this reason, both should be considered as a last-resort regulatory measure used only when other regulatory measures cannot solve the problem of competition problem in a given market. A careful impact analysis should be conducted to demonstrate that the benefits to be obtained are greater than costs and that wholesale access regulation is considered insufficient. Although functional and structural separation in the telecommunications sector have not been applied in the majority of OECD countries, some experiences in the OECD area, such as Australia, Italy, New Zealand, Sweden, the United Kingdom and the United States (OECD, 2011b), can be used as a reference. The recommendation of the OECD council concerning structural separation in regulated industries can also be used as a general reference for issues to consider (OECD, 2001).


Significant advances have been made in competition in the LAC region in the last few years, but in general, regulatory measures are needed to encourage entry and investment by new operators, as well as to encourage price competition and innovation.

Good practices to encourage competition in both fixed and mobile broadband markets include:

  • Improved authorisation and licensing processes for new entrants, which are still burdensome in many countries. Licensing processes should be faster and simpler, and fees for registration should be as low as possible, aimed to cover associated costs, following when possible a registration-based authorisation process not linked or specific to any service.

  • Foreign investment restrictions in the broadband access market, where they exist, should be lifted to maximise funding of networks and service deployment needed in the area, as well as to encourage competition. With a few exceptions, most of the LAC countries do not apply restrictions to foreign investment.

  • To ensure a level playing field, state-owned operators should be regulated and operate under rules and regulations as close as possible to those that apply to private operators. The administrative body responsible for overseeing the ownership of state-owned operators should be separated from policy-making administrative bodies.

  • Interconnection prices for fixed and mobile voice are still high in a number of LAC countries, leading to on-net/off-net differentiation, raising prices for voice services and limiting competition by small operators and new entrants. As broadband services are usually marketed with voice services, this also effects broadband competition. Ensuring that interconnection prices are public and cost-oriented and in line with prices in other regions where competition is thriving is good practice. At the same time, ensuring rapid and efficient number portability procedures in both fixed and mobile services is critical to facilitate switching by consumers and to develop competition. Countries that have still not implemented number portability should prioritise this issue.

  • The prices for Internet access are in general very high in the region, resulting in high retail prices for broadband access. Regulators should encourage IXP deployment, development of local data centres, encourage backbone deployment (see Chapter 8 on regional integration), ensure Internet openness (see Chapter 7) and monitor prices for Internet interconnection.

  • Regulatory authorities should as a general rule refrain from regulating emerging services too early, except where justified, such as for consumer protection.

  • Customer retention practices must be monitored. If periods for customer retention are too long, regulatory measures can limit the lock-in period and set conditions for customer retention. This can help improve competition.

  • Network sharing and co-investment should in general be encouraged by regulators. Administrative barriers for co-investment should be lifted, as this can significantly reduce investment needed to provide services. Regulators should monitor network sharing and co-investment agreements, setting conditions when needed to prevent undesirable negative impacts on competition.

  • Rights of way are an important area where national administrations and regulatory authorities can take measures to reduce barriers to entry. National harmonising administrative procedures, setting rules for quick and reasonable fees, one-stop-shopping procedures, as well as guidelines for municipalities are all important to ensure that operators do not face high administrative barriers in deploying networks.

In the context of mobile broadband, which can play an increasingly key role in the region for network expansion and competition for broadband services, competition can also be encouraged by:

  • auctioning more spectrum for broadband mobile access, encouraging new entry in the market (see Chapter 3 on spectrum management)

  • reviewing the regulatory framework to ensure that MVNOs have no regulatory barriers to entering the market, and if needed, imposing regulatory obligations on MNOs to facilitate wholesale access for MVNOs

  • imposing obligations for national roaming access when needed, to facilitate competition from new entrants, while they are deploying their network and for a time limited period

  • SIM lock-in practices should be monitored, and regulation may be applied to ensure SIM unlocking for customers under reasonable circumstances, e.g. when the terminal has been completely paid for.

In the context of fixed broadband, competition can also be encouraged by ensuring that in-house building wiring is not a barrier for competition. Another expedient is regulating existing in-house building wiring sharing, as is regulation to ensure adequate passive infrastructure in new buildings configured to accommodate several competing in-building wires.

Dominance is a key factor preventing competition, since dominant operators are in a position to set high prices and stall or prevent entry by new actors. Analysis of dominance in relevant markets must be performed regularly, since it provides the evidence base for subjecting dominant operators to regulatory measures aimed to facilitate entry of new market actors. Good practices in this area include:

  • Performing dominance market analysis on a regular basis (every two to four years) using recent data collected from operators, and applying sound methodologies that address not only market share, but other structural parameters in the market. A preliminary market analysis must be subject to public consultation.

  • Operators must have clear powers, set out in the regulatory framework, to impose regulatory measures derived from market analysis. These regulatory measures should be justified, be adequate to address the competition problems, and should be the least burdensome possible.

In general, the regulatory measures applied should be focused on the wholesale market and aimed at facilitating access of alternative operators to essential facilities. Bitstream, local loop unbundling, fibre unbundling, duct access, trunk services supported by submarine cables and wholesale access to leased lines are the typical wholesale access measures to consider.

Setting cost-based maximum prices for wholesale access and ensuring that an adequate and effective reference offer is prepared by the dominant operators are key issues for ensuring success and cultivating competition. The extensive experience in OECD countries can be used to address these issues.


ANATEL (2012), Plano Geral de Metas de Competição (PMGC) – Resolução No. 600 de 8 de novembro de 2012, Agência Nacional de Telecomunicações,

BEREC (2014a), BEREC Common Position on Geographic Aspects of Market Analysis (Definition and Remedies), Body of European Regulators for Electronic Communications, Vol. 14, No. 73, Riga, Latvia,

BEREC (2014b), BEREC Guidance on the regulatory accounting approach to the economic replicability test (i.e. ex-ante/sector specific margin squeeze tests), Body of European Regulators for Electronic Communications, Vol. 14, No. 190, Riga, Latvia,

BEREC (2011), BEREC Report on Open Access, Body of European Regulators for Electronic Communications, Vol. 11, No. 5, Riga, Latvia,

Bundesnetzagentur (2013), “Infrastructure Mapping – The German Infrastructure Atlas”, TRIS Workshop Infrastructure Mapping – Approaches, Benefits and Challenges, FTTH Conference 10th Edition, London,

Costa Rica (2010), “Normas Estándares y Competencias de las Entidades Públicas para la aprobación coordinada y expedita requerida para la Instalación o Ampliación de Redes de Telecomunicaciones”, Decree No. 36159,

CRC (2013), Resolución 4112 de 2013, Comisión de Regulación de Telecomunicaciones, Bogotá,

ERG (2009), Report on Next Generation Access – Economic Analysis and Regulatory Principles, Vol 9, No. 17, European Regulators Group, Riga, Latvia,

ERG (2006), Revised ERG Common Position on the Approach to Appropriate Remedies in the ECNS Regulatory Framework, European Regulators Group, Riga, Latvia, (accessed on 23 February 2016).

European Commission (2014), Commission Recommendation on Relevant Product and Service Markets within the Electronic Communications Sector Susceptible to Ex Ante Regulation – 2014/710/EU, European Commission, Brussels,

European Commission (2002), Commission Guidelines on Market Analysis and the Assessment of Significant Market Power under the Community Regulatory Framework for Electronic Communications Networks and Services – 2002/C 165/03, European Commission, Brussels,

IFT (2015), OREDA: Oferta de Referencia para-la Desagregación del Bucle Local, Instituto Federal de Telecomunicaciones, Mexico City, (accessed on 23 February 2016).

infoDev (2000), “Module 5: Competition Policy”, Telecommunications Regulation Handbook, The World Bank, Washington, DC,

ITU (2016), The ICT Regulation Toolkit, International Telecommunication Union, Geneva, (accessed on 23 February 2016).

JMCIA (2016), Japan Mobile Communications Infrastructure Association Website, (accessed on 23 February 2016).

Mexico (2014), Ley Federal de Telecomunicaciones, Secretaría de Gobernación, Mexico City,

Minetur (2011), Real Decreto 346/2011, Ministerio de Industria, Turismo y Comercio, Madrid,

Minetur (2009), Reglamento sobre mercados de comunicaciones electrónicas, acceso a las redes y numeración, Ministerio de Industria, Turismo y Comercio, Madrid,

MinTIC (2013), Resolución 449 de 2013, Ministerio de Tecnologías de la Información y las Comunicaciones, Bogotá, (accessed on 23 February 2016).

OECD (2015), Competition and Market Studies in Latin America 2015: The Case of Chile, Colombia, Costa Rica, Mexico, Panama and Peru, OECD Publishing, Paris,

OECD (2014), “Wireless Market Structures and Network Sharing”, OECD Digital Economy Papers, No. 243, OECD Publishing, Paris,

OECD (2011a), Competition Assessment Toolkit, OECD Publishing, Paris,

OECD (2011b), Next Generation Access Networks and Market Structure, OECD Publishing, Paris,

OECD (2008), “Public Rights of Way for Fibre Deployment to the Home”, OECD Digital Economy Papers, No. 143, OECD Publishing, Paris,

OECD (2003), “Developments in Local Loop Unbundling”, OECD Digital Economy Papers, No. 74, OECD Publishing, Paris,

OECD (2001). Recommendation of the Council concerning Structural Separation in Regulated Industries, OECD Publishing, Paris,

Ofcom (2015), The National Telephone Numbering Plan, Office of Communications, London,

OSIPTEL (2008), Decreto Legislativo 1019 Ley de acceso a la infraestructura de los proveedores importantes de servicios públicos de telecomunicaciones, Organismo Supervisor de Inversión Privada en Telecomunicaciones, Lima,

SCT (2016), Red compartida, Secretaria de Comunicaciones y Transportes, Mexico City, (accessed on 23 February 2016).

SUBTEL (2014a), Licitación 700 MHz: SUBTEL define frecuencias para Entel, Movistar y Claro, Subsecretaría de Telecomunicaciones, Santiago, Chile,

SUBTEL (2014b), Bases del concurso público para otorgar concesiones de servicio público de transmisión de datos en las bandas de frecuencias 713-748 y768-803MHz, Subsecretaría de Telecomunicaciones, Santiago, Chile,

SUTEL (2015), Consulta pública al reglamento sobre el uso compartido de infraestructura para redes públicas de telecomunicaciones, Superintendencia de Telecomunicaciones, San Jose, Costa Rica,

United States (2015), Broadband Conduit Deployment Act of 2015,

URCA (2015), Infrastructure sharing regulations, Utilities Regulation and Competition Commission, Nassau, Bahamas,

Vivo (2013), “Oferta Pública de Referencia de Roaming Nacional”, Vivo Website,

Further reading

AHCIET (2013), Operadores Móviles Virtuales en América Latina, Asociación Iberoamericana de Centros de Investigación y Empresas de Telecomunicaciones, Montevideo, Uruguay,

APEC (2011), Survey Report on Infrastructure Sharing and Broadband Development in APEC Region, Asia-Pacific Economic Co-operation, Singapore,

BEREC (2010), “BEREC Common Statement on Next Generation Networks Future Charging Mechanisms/Long Term Termination Issues”, BoR, Vol. 10, No. 24, Body of the European Regulators for Electronic Communications, Riga, Latvia,

Chung, H. (2010), “Developments in Cable Broadband Networks”, OECD Digital Economy Papers, No. 170, OECD Publishing, Paris,

CRC (2013), “Strengthening competition policy in Latin American countries: The application of competition law in the telecommunications sector”, Centro Regional de Competencia para América Latina,

GSMA (2014), License Renewal in Latin America, GSM Association, London,

OECD (2015), OECD Digital Economy Outlook 2015, OECD Publishing, Paris,

OECD (2014), “International Traffic Termination”, OECD Digital Economy Papers, No. 238, OECD Publishing, Paris,

OECD (2014), “The Development of Fixed Broadband Networks”, OECD Digital Economy Papers, No. 239, OECD Publishing, Paris,

OECD (2013), “Broadband Networks and Open Access”, OECD Digital Economy Papers, No. 218, OECD Publishing, Paris,

OECD (2012), “Developments in Mobile Termination”, OECD Digital Economy Papers, No. 193, OECD Publishing, Paris,

OECD (2011), Next Generation Access Networks and Market Structure, OECD Publishing, Paris,

OECD (2011), Recommendation of the Council Concerning Structural Separation in Regulated Industries, OECD Publishing, Paris,

OECD (2010), Geographically Segmented Regulation for Telecommunications, OECD Publishing, Paris,

OECD (2009), Network externality premiums and International Telecommunication traffic exchange, OECD Publishing, Paris,

OECD (2003), “The Benefits and Costs of Structural Separation of the Local Loop”, OECD Digital Economy Papers, No. 76, OECD Publishing, Paris,

ANNEX 4.A1. Number portability implementation in the region

Number portability in the LAC region


Portability for mobile operators

Maximum time to port the number (working days)

Portability fixed operators

Maximum time to port the number (working days)








Currently, there is only one mobile service provider, thus no effective portability


5 (Only on the same island)2







No, implementation by 2016


No, implementation by 2016









3 to 5


3 to 5





Depending on technical feasibility






Costa Rica





Dominican Republic










El Salvador


























1 to 2

































Trinidad and Tobago






No, but there is a law project


No, but there is a law project







Note: x = not applicable; .. = data not available.

1. According to information provided by the regulator, mobile number portability will take effect when the new mobile operator enters the market in 2016.

2. While the rules only require porting within the same island, CBL (one of the fixed providers) ports between islands within this timeframe (on its network and if a number is ported from the BTC network to the CBL network).


← 1. In 2012, the Peruvian government published Law No. 29904, (Ley de Promoción de Banda Ancha y Construcción de la Red Dorsal Nacional de Fibra Óptica). This states that telecommunications companies have the right to access and use the infrastructure of electricity and hydrocarbon companies. Access cannot be denied unless it threatens the continuity of these services. For details, see:

← 2. See, for example, Ley 8660 de Fortalecimiento y Modernización de las Entidades Públicas del Sector Telecomunicaciones, enacted by the Costa Rica regulator, SUTEL, aimed to set a more flexible framework for the Instituto Costariccense de Electricidad (ICE), the public telecommunications operator (

← 3. It may not always be necessary to regulate interconnection prices in an asymmetrical situation if compensation charges are reciprocal. If that is the case, and one party tries to charge more than its cost, it faces the problem that traffic flows may reverse and it will end up being the net payer. This happened in the United States in the 1990s. Some incumbents set their reciprocal compensation rates too high, and instead of receiving money, found themselves forced to pay it out as new entrants promptly set up modem banks to receive dial-up Internet access calls from incumbent customers. That being said, cost may differ among the incumbent and new entrants, and a reciprocal-rate model may not be fair. Moreover, high reciprocal rates may harm competition from alternative providers in the context of fixed and mobile telephony.

← 4. Unrestricted public access to all information about passive infrastructure may not be advisable, as it may be considered sensitive information for security reasons. As a result, controlling and granting limited access to specific persons or companies should they need to use the information may be advisable.