Licensing is authorization to build a network and/or to offer services of different kinds over a network. The arguments behind licensing usually relate to the need to regulate the activities of operators and service providers for the public good, such as quality of service and customer care, protection against price gouging and unfair or anti-competitive practices. Licensing arrangements are a way to ration scarce resources, but in cases where resources are not scarce other mechanisms may be used, such as authorizations, class licences that cover a variety of services to different devices, or simple registration.* For example, in the 1990s Hong Kong issued paging licences on-demand as long as there was spectrum available.
An Adaptable Licensing Framework Needed
In a pre-digital, pre-IP and pre-NGN era the costs of building and operating a network were very high giving rise to claims, not always justified, that a telecoms network was close to being a ‘natural monopoly’. A natural monopoly occurs when the unit or average costs of output fall as output rises across the entire market until all demand is satisfied. Under these circumstances, no new entrant could be more efficient at serving even a select portion of the market. In reality, there are almost always segments of the market where a specialized new entrant can serve more efficiently and in a more innovative way. In a world of digital telecommunications and IP networks new entrants can choose to come into the market using NGN systems with significantly lower operating costs. If there are any natural monopoly elements left they are likely to be found in long-distance traffic networks, but even here if non-telecom entities such as electricity utilities and rail networks are licensed to lease their long-haul cable or microwave capacity to new entrants, competition is possible. Licensing and authorization therefore need to take into account the changing economic realities that arise from new technology paradigms.
An important first step policy-makers and regulators can take towards a competitive telecommunications market is to make the licensing regime responsive to the emergence of new technologies that make new entry commercially feasible. Responsiveness to change requires a flexible licensing regime. There have been several different approaches to reforming the licensing process. One approach is to separate telecom activities into a tiered stack with infrastructure at the bottom, service delivery in the middle and applications and content as the top layers of the stack and issue different classes of license for each layer. For example, the Malaysian Communications and Multimedia Commission (MCMC) issues licences according to the following categories:
In the case of neighbouring Singapore, facilities-based and services-based licences are also issued, but the third layer consists of individual licences and class licences. In Africa, Kenya has a similar approach (see Toolkit Broadband in Kenya Case Study) but with twelve licence categories embracing ancillary sectors such as contractors, vendors and even business process outsourcers under a Unified Licensing regime. One of the problems with this segmented approach is that the lines of delineation between networks, services, apps and content begin to crumble. For example, cloud computing service providers may offer Infrastructure-as-a-Platform (IaaP) with more and more content accessible through apps; the licensing framework ceases to reflect the way in which facilities and services are offered.
An alternative approach to meet the needs of NGNs is the issuing of a multi-service general licence which enumerates the specific types of services the operator can provide. The most flexible form of general licence is a service and technology-neutral unified licence which permits entry to any field of service the operator wishes to invest in. This was an approach pioneered by India. The most radical approach is to replace licences altogether with registrations, but this may require codes of practice to ensure good behaviour by operators and service providers.
In some jurisdictions, licenses have to be approved by the legislature. While this process may ensure a thorough vetting of the application, it can also mean lengthy delays and the involvement of policy-makers who are not specialists in the field. It is also a process that lends itself to lobbying. A better option is to remove the specifics of licensing conditions as applied to each operator from the legislative process and place them in licences issued by the telecoms regulator, even if final approval or endorsement lies with a higher authority. Legislation can then focus upon the overall principles of licensing in the sense of creating a template which the regulator can use.
Transparency and Investment
What is fundamentally important is that the terms and conditions of any licence are absolutely transparent and, ideally, available for all to see on the regulator’s website. Both users and competitors should have the right to know what are the terms and conditions of service. Users need to know for the protection of consumer rights, and competitors need to know so every operator has equal access to commercially non-sensitive information. Where information is commercially sensitive, for example in the cost accounting data upon which a Reference Offer of Interconnection (ROI) is agreed by the incumbent operator, then the regulator needs to have a copy so that they can determine whether there is discrimination in case a dispute emerges between operators.
Unfortunately not all licences are given out on a transparent basis. Where corruption or cronyism is involved licences are seen as rewards and favours, with the inevitable consequence being that users and the economy-as-a-whole are denied the full benefits of competition. This will make the economy less appealing to investors, especially to investors in carrier services who see the market as being rigged, but also to investors in sectors that are heavily reliant upon broadband communications networking such as financial institutions, trading companies, business process outsourcers, call centres, and other companies that are part of the global supply chain of multinationals. A World Bank study of investment in telecoms in the Asia Pacific region in 2004 found that regulatory uncertainty was the No.1 deterrent to investors in telecoms, while the availability of high quality and affordable telecommunications services is regularly cited as among the top three factors that attract overseas investment, along with the rule of law and good transportation systems. A good example from North Africa of transparent regulation was Morocco’s creation of an independent Agence Nationale de Réglementation des Télécommunications (ANRT) in 1998.*
Ex-Ante, Ex-Post and Incentive Regulation
In many cases the conditions of the licence as they apply to different operators will depend upon whether the operator in question has been designated as dominant in any given market, such as in voice termination or in leased circuits. Dominance is usually measured by market share of users or by the share of total revenues. A widely used measure, for example is the Hirschman-Herfindahl index (HHI) which is constructed by the sum of the squares of the market shares of each operator, so in an evenly competitive market of 4 players each with 25% of the market HHI = 252 + 252 + 252 + 252 = 2,500. A number larger than this is indicative of a more concentrated market, but this is a rather static measure that in itself does not explain market behaviour.* A more refined measure is significant market power or simply SMP. This estimates how far a single operator can increase profitability by changing the price it charges for its services. In a highly competitive market, it will lose customers when it hikes its prices, and if it lowers its prices competitors will match it. Traditionally, where an operator has been designated as dominant, some regulatory restrictions apply; for example, an incumbent operator could be required to submit proposals of tariff changes to the regulator for approval. This is regulation before the event or ex ante regulation. However, as competitors establish a firm foothold in the market there has been a change towards regulation after the event or ex post regulation. This is appropriate when it cannot be assumed that the dominant operator is abusing its market position. It is often referred to as ‘light-handed’ or ‘light-touch’ regulation. Just as the threat of competition may be sufficient to deter the dominant operator from acting in an anti-competitive manner, so the threat of regulation may achieve the same result. If it does not, then regulation is called for.
A half-way house between ex-ante and ex-post regulation arises when competition has been introduced but is not yet well established. A good practice widely used in the US and in the UK in the 1980s came to be known as incentive regulation or economic regulation. The regulator would develop a formula to govern when and by how much the incumbent could change tariffs in a way that rewarded the operator for becoming more efficient. One such formula used in the UK was the price cap which allowed British Telecom (BT) to raise its prices by RPI-X, the retail price index minus X% where X is determined by the regulator. So if the rate of retail price inflation was 8% and X was set at 5%, the operator had an incentive to increase efficiency by over 3% to make more profit. The formula referenced a basket of services and within the basket there were individual sub-caps to allow BT to rebalance prices between services. X was adjusted every 3 to 5 years. This example is not to suggest that a price-cap formula is the best solution for all circumstances; it is simply a good solution if it works under local circumstances. The key point is that this was an inventive way to apply solid economic principles to a regulatory problem, combining longer-term regulatory flexibility with short term financial clarity for the investors. It is just one of several options.
Maintaining a Competitive Market
Competitive markets can become less competitive when companies exit the market and when mergers and acquisitions (M&A) reduce the number of network and service suppliers. It is therefore important that the regulator is given powers by the legislature to evaluate potential M&As. This means imposing a requirement on the companies concerned or on the acquiring company to notify the regulator and seek guidance. In some cases this function may be carried out by a separate competition or monopolies commission. A good practice is for the regulator to give advanced advice to the parties concerned to smooth the process. This avoids undue delays, but identifies early on any substantial issues that may need addressing. The regulator can either aim to avoid a company becoming dominant as a result of an M&A by imposing conditions such as the disposal of assets, or can require certain behavioural guarantees such as equal access to network facilities and directory databases in exchange for ex post regulatory oversight. These steps may require changes to the terms and conditions of the licence before the M&A is given the green light.
3.2.1 Technology and Service Neutrality
Section 3.2 emphasized the need for the licensing and authorization framework to keep pace with changes in the technological landscape because new technologies and standards can give rise to new market opportunities. The traditional problem is that licences issued many years previously may specify the technology or the standard to be used. This was often true by default in the case of wireless cellular telephony. If the CDMA standard was used then an allocation within the 800 MHz band was assigned by licence, and if the standard was GSM an allocation within the 900 MHz band was assigned, and so on. With 3G and 4G standards now available innovation in frequency usage enables mobile network operators (MNOs) to use a range of different spectrum bands including 900MHz, 1.8GHz, 2.1GHz, 2.5/2.6GHz, etc.
Promising changes in spectrum use are arising from the so-called digital dividend as frequencies used by analogue radio and TV broadcasts in the VHF and UHF bands are freed up with the shift to digital radio and terrestrial TV or DTT. The 700MHz band especially is seen as having excellent propagation characteristics that could mean much wider area coverage by cellular and other wireless technologies for the same or less investment. This will pose a challenge and an opportunity for policy-makers and regulators. The challenge will be how to allocate these frequencies between the claims of competing services. The opportunity will be to assign these frequencies to new services and to meet the needs of populations in more remote and rural areas. This will be discussed further in section 3.3. The point at this stage is that although regulation through licensing may be needed to ration a scarce resource such as radio spectrum, if the licensing process itself is to become ‘future-proof’ it needs to take a step back from specifying exactly which technologies and standards are to be used. For example, allowing 3G services to be offered over 2G assigned frequencies—a process known as spectrum refarming—is a step in this direction. UMTS900 networks have already been deployed by AIS in Thailand, by Digitel in Venezuela and by DNA in Finland.
No amount of licensing or regulation can predict which technologies and standards will be successful in the market. Some will come and go quite quickly. Paging for example, is not used today but was a popular and conveniently cheap method of text communication in the 1990s, before it was pushed aside by lower cellphone prices. Some standards will enter the market but fail to gain widespread adoption. Local Multipoint Distribution Service microwave or LMDS was a case in point, and the WiMax (Worldwide Interoperability for Microwave Access) may be another. Licensing policies need to be adaptable to accommodate these market experiments because no amount of regulatory foresight can predict the outcomes and, as in the case of WiMax, although it may not be adopted in its mobile version it can nevertheless play a role as a substitute for digital subscriber line (DSL) in some markets. Also a new entrant choosing the standard is more likely to develop a sustainable business plan if the licensing conditions permit it to migrate to an alternative technology (DSL) or wireless standard (4G) as it discovers the market need. This is what is meant by technology neutral regulation: it leaves the choice up to the investor to test the market.
Apps and Services
So far neutrality has been discussed in terms of technology and standards, and in terms of medium, fixed or wireless. Ultimately neutrality is about different services requiring different media and technologies. For example, most standalone MNOs are required to lease backhaul from fixed-line carriers, but they would be in a much stronger competitive position if they could invest singly or jointly (through facilities sharing) in their own lines. If needs be, their license conditions could restrict the usage of those lines to their own business so they do not compete for the retail business of the fixed line carriers; however, even that solution may be considered second-best to outright competition. Certainly unified or converged carriers offering both fixed and mobile services (triple play, including Internet) have a cost advantage in this regard. An alternative solution is to licence independent broadband wholesalers which encourages more cost-effective competition at the retail level.
A new dimension to technology and service-neutral regulation has arisen with the spread of web-based applications that can be accessed through a range of fixed line and wireless network devices, such as smartphones. As many of these apps are OTT and therefore potentially by-pass the tariffs of the licensed carriers’ networks, carriers are tempted to restrict the bandwidth made available to their download, either by throttling bandwidth, outright blocking or by levying additional charges on their users. This is the net neutrality issue which is also referenced in module 3.7. It is a point of argument whether regulation that tries to preserve equal access to services for consumers is straying too far into the commercial pricing decisions of operators. In Hong Kong, for example, the regulator took the view that while operators could introduce tiered pricing schemes for consumers to choose from, they could not discriminate against the supply of different apps or content. In other words, once a consumer has established and paid for their chosen level of demand in terms of total broadband capacity they can use per month at the basic fee or in terms of bit rate speeds, they are entitled to access any apps and content up to their chosen limits.
3.2.2 New Authorization Options and Their Implications for Broadband
Regulation of the media and regulation of telecommunications networks and services have traditionally been separate domains of government. The spread of broadband, the convergence of technologies and the search for business synergies between publishing, broadcasting, IPTV and the delivery of content over the Internet has prompted several countries to converge their telecoms and broadcast regulation agencies, although it should be noted that this does not imply that the regulations have fully converged. For example, the FCC in the US was created in 1934 and manages the regulation of both telecoms and broadcast in separate departments. A useful ‘Environment Scan’ conducted by the Canadian Radio-television and Telecommunications Commission (CRTC) in 2011 cited the Australian Communications and Media Authority (ACMA) as finding that:
Even in converged legislative frameworks that adopt an industry-agnostic approach to carriage regulation, at this point in the evolution of converged regulatory models, when it comes to content, sector-specific media regulatory measures still generally apply.*
However, these changes are bound to impact upon the way in which new authorizations will be made. Modules 3.2 and 3.2.1 have highlighted the important development of technology-neutral and service-neutral licensing regimes and the challenge of keeping the licensing structures relevant to developments in the marketplace. The innovation of multiservice and general licensing was noted, but even these approaches can have their red tape. The multiservice approach, for example, may require different licences for different categories of service and cross-ownership rules may restrict the licences available. In an ideal world, all services would be open to all comers, but in reality markets often support only two or three profitable ventures,* raising fears of dominance or, possibly, of collusion. For this reason, easing the licensing requirements for service providers and content distribution networks (CDNs) may prove an effective countervailing power within the market.
Much of the content today is likely to come from overseas, through broadband, through the Internet and through CDNs. Policy-makers and regulators are therefore faced with additional decisions to make: should they confine licensing and regulation to domestic-based service providers and permit open access to external services? Or, should they attempt to impose domestic conditions upon external services? If so, how? For example, in some jurisdictions, overseas service providers are required to register an official presence in the country. Vietnam has proposed this, but only as a point of official contact and not, it seems, as a means to carry liability in the case of a dispute.
The arguments for regulation usually relate to the need to screen out unacceptable apps and content. This is clearly a decision that has to be taken at the national level, but a good guideline is that if there are to be regulations and restrictions they should meet two conditions. Firstly, what is regulated and restricted should be a fair reflection of what is legal and illegal in local law, otherwise there is a danger of arbitrariness, lack of policy transparency and even of due process. Secondly, regulations and restrictions should be proportional to the threats or dangers involved. For example, content that glorifies or promotes hatred and violence is far more harmful than content that makes parody and criticism. An additional important element to consider is the intention of the content. For example, content that makes a damaging untrue statement about a person may be deliberate libel or it may be an unintended error. Proportionality in the application of law should be able to draw the distinction and apply that in the remedy, in this case an apology and/or a takedown rather than a prosecution.
3.2.3 Disputes Resolution Procedures
Moving from an incumbent monopoly towards an open entry and free market in telecommunications is a process that from time to time inevitably generates disputes between the regulator and the operators, between different operators and between operators and customers.*
Disputes with regulators are most frequent when licensing conditions are being imposed such as steps to curb anti-competitive behaviour, the imposition of price controls, and the granting of permissions to market new services. Disputes between operators are more likely to centre around competition issues, such as discrimination in the cost of interconnection or in providing access to unbundled network elements (UNEs) – see Module 3.6 – and claims of misrepresentation arising during combative marketing campaigns. Billing, quality of service and waiting lists are the types of issues most likely to arise between customers and operators.
Dispute resolution procedures usually take one of two forms: either official government channels such as the regulator, a statutory arbitration court, an appeal to a minister or to the Supreme Court, or through unofficial channels – also referred to as alternative disputes resolution (ADR) – involving a voluntary process and an arms-length arbitration panel. Consumer councils often play an important role representing user interests, but they do not always enjoy an official status. Sometimes, as in Thailand, the regulator’s office helps to create a consumer protection agency.
Where disputes arise over a regulator’s decision it is best practice to establish some kind of telecoms tribunal, in Hong Kong called the Telecoms Appeal Board, to ensure transparency. But the appeal will be confined to examining whether the regulator followed due process and not a challenge to the regulator’s statutory powers to make a judgement. Like many other jurisdictions, Hong Kong has also established an ADR mechanism called the Customer Complaint Settlement Scheme (CCSS) similar to Australia, New Zealand and the United Kingdom to help resolve disputes between operators and users.* In 2013, Hong Kong’s six mobile virtual network operators (MVNOs) agreed to join the scheme which already covered fixed and mobile network operators. The importance of ADRs is they can provide fast and also less costly solutions to relatively small disputes without taking away the legal rights of the parties concerned.
Other approaches to dispute resolution include judicial or semi-judicial tribunals that adopt court-like procedures. For example, India has placed the procedure in the hands of an independent Telecom Dispute Settlement and Appellate Tribunal (TDSAT) presided over by a retired high court judge. In his assessment of the dispute resolution in India and influences upon it from around the world, R.U.S.Prasad notes the importance of Malaysia’s approach, although in the latter case the tribunal’s chairman, also a retired high court judge, is subject to appointment and possible dismissal by the minister.* In the US, where the role of litigation is more accepted as part of commercial life, the decisions of the Federal Communications Commission (FCC) are generally regarded as final. However, law judges are involved during hearings leading up the final decision of the Commissioner, and FCC rulings are frequently subject to Court of Appeal challenges.
The legal and commercial histories of countries differ widely, so different approaches to dispute resolution apply, but if there is one underlying principle that should be common to all is transparency.