Home > Consultations > Consultation Documents > Telecoms Review > Telecoms Review  > Section 5

5.1 Section 3 noted that Ofcom has a statutory duty to reduce regulation where possible. We need to see if there are approaches that could lead to withdrawing regulation, where markets can or will become effectively competitive. Where there is a case for continued sector-specific regulation, it must be justified from first principles.

5.2 The central element of the case for sector-specific regulation in telecoms up to now has been the existence of market power. This has led to regulation designed to protect consumers directly and actively to promote competition.

5.3 Not only has the telecoms sector changed beyond recognition over the last twenty years, but it is also likely to change again very substantially in the period to the end of the decade. Among other things, these changes may be caused by technological trends (discussed in Annex J), changes in consumer demand (discussed in Annex I), and changes in the telecoms industry itself (such as mergers between companies). The effect of these changes may be to reduce or remove market power from parts of the market, so creating opportunities to withdraw from regulation.

On the other hand, market power could increase in some parts of the market, or shift to different parts of the value chain, in which case there would be a continued case for sector-specific regulation, but with a change of emphasis and focus.

5.4 This section sets out the key changes, driven by consumer demand and technology change, that Ofcom considers could have a significant impact on market power, and hence on the regulatory requirement. This section looks at possible changes in four critical areas:

• competition in voice services;
• broadband and IP;
• evolution of network design; and
• the evolution of industry structure and strategy.

5.5 This section then considers how regulation can itself affect the transition to more open and competitive markets through the adoption of different regulatory approaches. Finally, it considers an important consequential effect of change on the potential future scope and funding of universal service obligations.

5.6 We would welcome views both on whether we have correctly identified the key trends, and also views on what these trends suggest about the broad approach to regulation that Ofcom should take.

Competition in voice services

5.7 The voice telephony market has been the historical focus of regulatory intervention, and provides the basis for much of today's telecoms regulation. In this section we deal with two possible changes to the voice environment. The first is increased competition for voice services across fixed and mobile platforms, and the second is the challenge to existing voice telephony businesses from VoIP services.

Fixed and mobile convergence in voice telephony

5.8 The market reviews conducted by Oftel in 2003 identified separate markets for fixed and mobile voice and data services. Oftel concluded that fixed and mobile services are insufficiently close substitutes to fall within the same market definition. Nonetheless, customers are already switching some mobile voice calls for fixed voice calls. There is also some switching from voice to non-voice communications, for example with people using text messaging in circumstances where in the past they might have made a voice call.

5.9 To some extent, this switching may represent a shift in behaviour driven by convenience rather than price. The important issue in assessing competition is whether switching between products will increasingly take place in response to changes in relative prices, such that mobile and fixed operations become economic substitutes ⁷ for voice and/or data services. If two products are sufficiently close substitutes for each to constrain the other's pricing then these are likely to form part of the same economic market ⁸.

5.10 Seventy-six per cent of households now have access to both a fixed line and a mobile. In principle, an increase in the price of fixed calling might lead a number of these consumers to substitute a mobile service (voice call or text message) for some of their fixed calls. If such "call substitution" were substantial it could constrain the price of fixed, with a small price rise from the competitive level being rendered unprofitable, and the two would be regarded as part of the same economic market. Currently the costs of mobile calls are significantly higher than those of fixed calls. In conducting its mobile market review, Oftel noted that mobile operators' traffic related costs are expected to remain above those of fixed networks. Oftel considered that consumers might well be prepared to pay some premium for the additional convenience of mobiles. However, it took the view that the cost differential between fixed and mobile exceeds this premium and the availability of mobile services is not a sufficiently strong constraint for them to be regarded as being in the same market as fixed.

5.11 Unit costs are declining both in fixed and mobile technologies. It may be that in the long term, the costs of the two platforms will converge and consumers will readily substitute between them. An important factor could be the release into the market of additional mobile network capacity as 3G networks come on line, reducing unit costs. Even if the unit cost of mobile remains higher than fixed, the difference may be sufficiently modest that consumers are willing to pay the very small premium implied for the greater convenience of mobile.

5.12 The shift in usage patterns could go further than the scenario of consumers who have both fixed and mobile erring on the side of mobile to make individual calls. "latform substitution" would occur if consumers stopped subscribing to one technology altogether because they preferred another; for example, if they did not have a fixed line because they had a mobile.

5.13 There is evidence of some consumers taking mobile services only, with no fixed line: currently 9 per cent of UK households have access to a mobile phone only. However, there is a question of whether such switching behaviour is sufficiently responsive to relative prices to bring the two into the same economic market. Some consumers (such as temporary workers or students) may choose to use a mobile on its own because of transitory living arrangements or convenience factors, and would not have taken a fixed line in any case. Some consumers may keep a fixed line because they prefer to use it for data communications. However, others may be more responsive to relative prices. In assessing whether to switch to mobile only, consumers will take account of the full costs of using each service, i.e. line rentals as well as call charges, or the cost of a bundle including most of the calls that the consumer wishes to make.

5.14 Even if the costs of fixed and mobile operations become sufficiently close to form part of the same economic market, the precise nature and speed of substitution between them will be affected by the business strategies of the various operators. Mobile operators might adopt a strategy of aggressive call pricing in order to grow market share, taking call volumes away from the fixed operators. To stimulate the trend towards platform substitution, and ultimately grow their businesses, mobile operators might offer large bundles of calls at low per-minute prices. Alternatively, an operator on one platform may extend its brand into the other platform, offering combined fixed and mobile services with a single bill and attractive bundles including both fixed and mobile minutes. Service providers without their own network infrastructure might also offer combined fixed and mobile services. Such a strategy could further increase substitution between fixed and mobile calling, although it might perhaps hold back full platform substitution. It should be noted that service providers without their own network infrastructure are dependent on wholesale agreements with existing networks, and hence may not have complete freedom of action in terms of their pricing strategies.

5.15 Finally, there may in future be a substantial blurring of the definition of "fixed" and "mobile" services, with the emergence of new technologies offering limited mobility and enhanced functionality. Wireless LAN "hotspot" solutions using 802.11 technology are currently being deployed by both fixed and mobile network operators, giving high speed connectivity within a small area. Other limited mobility wireless technologies, such as 802.16 ("WiMax") may be deployed in future, giving different combinations of connection speed and coverage. As these technologies are deployed both by fixed and mobile operators, and as devices become capable of connecting to different access networks in fixed, limited mobility and fully mobile environments, the services of the two types of operator could converge, and distinctions between them blur. One implication of this is that substitution of traffic between fixed and mobile could potentially take place in the opposite direction from that described above, away from mobile operators and towards fixed operators.

Why it matters:

5.16 Existing regulation regards BT as having SMP in relation to fixed narrowband exchange lines, certain fixed narrowband calls markets, and call origination. Were a more competitive, cross-platform market for both calls and access services to develop in the medium term, this could affect a future assessment of SMP in these markets.

Question 7: How rapidly and extensively will fixed and mobile networks become substitutes for one another?

 

Effect of VoIP

5.17 Alongside the migration of voice traffic from the traditional, circuit-switched fixed telecoms network to mobile, migration may also take place onto
IP networks.⁹

5.18 Ofcom uses the term Voice over IP (VoIP) to mean a voice call that is delivered over IP networks (either the public internet or a managed private IP network) instead of the traditional Public Switched Telephone Network (PSTN). However, as a retail proposition, it includes a number of different products. One is software which enables two computer users to speak to one another if both are online. Another allows a computer user to call any telephone while online. The most significant and mainstream products allow a conventional phone to be plugged into an adaptor box which converts the signal to IP. These are available both for narrowband and broadband connections.

5.19 VoIP potentially offers substantial benefits to consumers. In principle, it could lead to a step reduction in the core network costs of providing voice telephony. It also offers much greater scope for value-added voice services, such as multi-party conferencing, than traditional analogue circuit-switched delivery.

5.20 Some of the practical barriers to deployment are starting to be addressed, such as VoIP to PSTN interconnection, which allows VoIP users to interconnect seamlessly with existing switched voice telephony users, and problems around access to emergency services. However, interoperability in the VoIP environment remains a significant concern.

5.21 Corporate use of VoIP is now widespread, because complex corporate networks are increasingly used both for voice and data traffic. However, use of VoIP to date by residential and SME consumers has been restricted to use via computers by a small niche of internet users. Many telecoms operators worldwide are now either actively deploying, or at least outlining plans to deploy, retail VoIP services targeting all customer segments. Residential VoIP services may initially be bundled with broadband and, as the number of broadband users grows, this may drag VoIP in its wake.

5.22 However, existing voice telephony providers are likely to face a mixed motivation in deploying VoIP. Existing access operators, having invested in a fixed voice network architecture, might be expected to seek to maximise returns on that architecture rather than incur the costs of a further cycle of investment. This, of course, partly depends where operators are in their investment cycle. An operator contemplating new network investment today might deploy VoIP rather than invest in PSTN infrastructure that is already technically obsolete.

5.23 We need to understand when, and how, operators may make the decision to roll-out large scale VoIP propositions. In principle, VoIP is a lower cost technology, and so could offer operators higher margins on voice calls. However, existing operators might prefer to delay its introduction if it "stranded" existing network equipment, or if it precipitated such falls in prices that their margins ended up being lower. The feasibility of an existing operator delaying entry of VoIP is dependent on whether all other market players are likely to pursue the same strategy. This partly depends on whether all market participants have the same amount to gain from a decision to venture down the VoIP route. It would take just one company to break ranks - and in the case of VoIP, this need not be an existing voice telephony provider - to undermine the efforts of existing players to maintain the status quo.

5.24 Were it deployed on a very large scale, VoIP might display rather different cost characteristics than exist now. Currently, it uses bandwidth which has been put in place for data communications. As it becomes mass-market, it may not be able to take advantage of "costless" access to bandwidth in this way. It will be important for Ofcom to understand the economic characteristics of VoIP in the course of the Review.

Why it matters:

5.25 The implications of VoIP deployment are potentially massive. From a competition perspective, VoIP could erode existing market power in relation to calls and access services. On the other hand, VoIP may not eliminate the need for regulation but merely alter the regulatory challenge - an issue that can only really be addressed by developing a detailed understanding of how a future VoIP market might develop.

Question 8: What impact will Voice over IP have on the telecoms market?

Broadband and IP

5.26 Not only is the market for voice services changing rapidly, but the rise in demand for higher speed data services may change the telecoms market in other, equally fundamental, ways. This section considers future trends in relation to broadband - its growing importance, and the consequences of this for investment and change in the telecoms market.

5.27 Ofcom has reviewed projections for broadband penetration in the UK, which range from 40 per cent to 50 per cent of households by 2008. This is comparable with projections for the US and the majority of European states, but well behind the economies of the Asia-Pacific rim. Whatever the exact number, by the end of the decade, broadband is likely to have grown from a niche product to a mass-market product and will be critical to the prospects of the sector. This is borne out by the importance attached to broadband growth strategies by telecoms companies themselves.

5.28 Some commentators have argued that broadband take-up could be even more rapid than this. Rather than a steady, predictable rate of increase, they suggest a "tipping point" at some stage in the evolution of the broadband market. At this point, broadband would move very rapidly from a niche product to a mass-market product, and then to a "must-have" product similar to voice telephony.

5.29 One important question is the extent to which this broadband growth occurs at the expense of the existing narrowband data market. The narrowband data market has evolved from the first dial-up services using local, national and premium rate telephone numbers through to the development of unmetered internet services. All of these continue to be provided over the traditional circuit-switched network. This market is likely to decline in importance as broadband services largely substitute for narrowband, and therefore the principal competitive and regulatory issues in future are likely to arise in relation to broadband. The transition from narrowband data to broadband data may be less fundamentally disruptive than the potential transition of voice services to the IP environment. However, as we discuss below, the combination of the narrowband to broadband transition and the migration of voice services could, in combination, have important implications for the existing circuit-switched network architecture.

5.30 This sub-section considers the impact on market power, and the telecoms regulatory requirement, of four aspects of the growth of broadband. These are

• the scope for competitive delivery of broadband services over different platforms;
• the scope for competition in IP networks;
• the deployment of new "broaderband" services; and
• broadband content.

Question 9: How rapidly and extensively will broadband be taken up in the UK, and what are the regulatory implications of such growth?

Scope for competitive delivery of broadband services over different platforms

5.31 When assessing the scope for future competition in relation to broadband services, Ofcom must take a view on the extent to which inter-platform competition could emerge. Some larger business customers in metropolitan areas have a choice of alternative providers of fibre access. For smaller businesses and residential consumers, today's broadband services are largely being delivered through a combination of incumbent telephony networks upgraded to deliver DSL products, and cable TV networks upgraded to deliver "cable modem" services.

5.32 However, wireless networks, satellite, "powerline" (which delivers telecoms services over electricity cables) and passive optical networks are all possible sources
of further access-level competition. Local loop unbundling (LLU), while not "pure" platform competition, involves an element of infrastructure competition in access. In the course of the Review, it will be important for Ofcom to understand the scope for new competition from these sources. For example, many analysts do not include mobile networks when assessing the scope for competition in broadband. However, mobile networks are also deploying broadband architectures. They may be able to
deliver broadband services which would, in terms of functionality, match some of those currently available on wired networks.

5.33 There has been substantial investment in broadband mobile access technologies; in particular, 3G. Wireless LAN "hotspot" services are also being rolled out in many public places. However, in recent years there has been little inclination on the part of operators or the capital markets to fund large-scale new fixed access networks in the UK. But this may well change over time. Japan, Korea, the US and France are all attracting capital into access infrastructure, either in the form of new primary infrastructure (typically fibre-based) or Local Loop Unbundling (LLU), albeit in very different market environments to the UK's. Understanding whether such market developments could also happen in the UK (recognising that regulatory choices partly affect the answer to this question) is an important issue for the Review.

5.34 It is also important to understand how extensive the geographic reach of such new networks might be. There are likely to be different levels of competitive intensity in different parts of the country. Regulatory policy might therefore need to balance the requirements of urban areas where there may be scope for further network development, and rural areas where there is less likely to be any source of competition to BT's access network.

Why it matters:

5.35 If there is scope for further additional access network construction which competes with DSL and cable, this might suggest a declining case for regulation of broadband access over time. On the other hand, if there is little scope for roll-out of additional broadband networks in the medium term, this would suggest that broadband access needs to be an area where regulation is particularly focused. This is important because, as discussed in Section 4, there may be a trade-off between regulation to promote further access infrastructure build, and regulation designed to promote access to existing infrastructure

Question 10: What scope is there for new, competing broadband platforms to be rolled out, and which technologies are most likely to be used?

 

Scope for competition in IP networks

5.36 Broadband is a significant driver of the migration by operators towards a so-called "all IP" network. Such an "all IP" environment would involve the progressive replacement of existing legacy networks over a period of time. Potential advantages for the operator include moving to a lower overall cost base, principally through the ability to "de-layer" the network and strip out intermediate network protocols such as ATM and SDH, and also through benefiting from the cost savings associated with a single network architecture.¹⁰ What were previously said to be disadvantages with this approach - most noticeably, concerns about the security and quality of service of an IP architecture - are now being addressed. For example, multi protocol label switching (MPLS) is already allowing very robust virtual private network (VPN) products to be rolled out to corporate consumers.

5.37 At present, networks interconnect to one another at different levels of the network architecture. BT, for example, offers a range of wholesale broadband offerings. These include IP Stream, which bundles the access part of the BT network with BT's IP network; and Data Stream, which allows other operators with their own fixed networks to take traffic from the BT access network and divert it, via an ATM interconnection, onto their own networks for onward delivery. As part of the Broadband Overview, Ofcom has been looking at options for increasing competition in the backhaul part of the market.

5.38 If networks were to deploy "all-IP" architectures, an arrangement such as Data Stream could potentially be rendered obsolete, and new interconnection products might be required. This might shift the focus of regulation toward the viability and sustainability of competition in relation to the IP layer itself.

5.39 IP networks are frequently stated to exhibit large returns to scale, principally in terms of customer numbers and overall traffic volumes. Ofcom would like to explore this issue further, given its implications for the scope for future competition in the market. In particular, we would like to understand whether these scale economies are greater, lesser or different in nature from the scale economies which have traditionally been associated with the PSTN.

Why it matters:

5.40 Competition at the backhaul and core network level was an important component of the regulatory approach adopted by Oftel. If the introduction of IP changes the economics of these networks, it is likely to change the current assumptions underlying the regulatory structure in this market. We need to understand the scope for competition in this part of the market in the future, and what form that competition could take. We can then determine what regulatory regime (if any) is required to ensure effective competition.

Question 11: When are operators likely to move towards "all IP" architectures, if at all?

Question 12: What are the implications of "all IP" networks for the way networks interconnect with one another, and for the scope of competition?

Deployment of new "broaderband" services

5.41 The current range of higher-bandwidth offerings available to UK consumers includes entry-level products aimed at residential consumers and SMEs offering download speeds below 512kbit/s; mid-range products offering 512kbit/s to 1 mbit/s; 2 mbit/s ADSL and SDSL services aimed at SMEs; and leased lines of 2 mbit/s and upwards aimed at larger businesses, which offer bandwidth plus no "contention", guaranteeing a set amount of capacity for the customer.

5.42 In some other countries, a new generation of much higher speed services is becoming available, both for business and residential consumers. An issue for the Review is to what extent demand for these "broaderband" services will develop here, and how the market will serve these needs.

5.43 Two possible stimuli for demand are the need for symmetrical bandwidth services which cater for peer-to-peer activities and transactions such as file sharing; and demand for download speeds which are sufficient to cope with full-motion video and video gaming. The technological capabilities for such services already exist, and they are now being deployed in Korea and Japan.

5.44 Further network investment would almost certainly be required to support these services, even by operators who already provide broadband. The cable network infrastructure may already be capable of supporting such download capabilities in theory, but we understand it would be necessary to make further investments to support a large-scale roll-out of higher bandwidth services, not least in relation to core network transmission capacity and server architecture. BT would need to make very significant network investments to support speeds of, for example, 5 to 10 mbit/s. It may need to deploy fibre much closer to the customer, potentially all the way to some customers' locations. Other broadband transmission infrastructures, including fixed wireless access technologies such as WiMax (802.16), might also provide a means for meeting this demand for higher speeds.

5.45 As well as demand from residential consumers for "next-generation broadband", there is also potential demand from businesses, particularly SMEs. At present, DSL-variant products and cable-modem based products occupy a niche, serving businesses who do not require or cannot justify the expense of leased line services. Whether there will be continued demand depends on the relative pricing of these products and leased lines, and whether businesses themselves require ever-increasing bandwidth.

Why it matters:

5.46 In principle, decisions to invest in "broaderband" networks, even if the scale of investment required is considerable, do not in themselves present any new issues or challenges for regulation. But it is important that regulation facilitates network investment where it is efficient and desirable. Investment may not always display these characteristics - as Section 4 sets out, there may be a trade-off between dynamic and other forms of efficiency.

5.47 Therefore, in practice it is important for Ofcom to understand if major new network build is required in the near term. A linked issue is where, realistically, such investment may come from. Whether there is scope for such investment from BT, from the cable companies, from mobile operators and/or from new entrants and new technologies could also have important consequences for regulation.

Question 13: Is there likely to be widespread demand for services that require "broaderband" networks to be rolled out and, if so, how will such infrastructure be supplied?

Broadband content

5.48 One constraint on the uptake of broadband is the need to develop viable content models. This is particularly relevant to the question of whether a business case for "broaderband" networks will emerge.

5.49 Broadband-enabled "entertainment-type" services may need to compete with traditional linear broadcasting technologies (such as satellite DTH), and off-line DVDs and video games, for a share of consumers' expenditure. In the medium term a "converged" market for video entertainment services could perhaps emerge, encompassing both these more traditional broadcast platforms and "telco-type" platforms, at least for some categories of entertainment service. Clearly, there are strong players already operating within this space, in some cases with exclusive rights to certain premium content. Broadband-based services would need to find sufficient differentiation in order to secure the kind of margins which would justify the high upfront investments that would need to be made. For example, time-shifted TV is one such possible market opportunity.

5.50 A practical constraint on the development of broadband as a content delivery platform is the desire of major rights holders to restrict illegal file-sharing activity, and their corresponding reluctance to set up legitimate online content distribution channels. Our discussions with stakeholders suggest that two things are likely to be needed to give some rights holders sufficient comfort to release content for broadband distribution. The first is effective and agreed standards for digital rights management (DRM).11 The second is a legal framework and enforcement mechanisms that can be used effectively against intellectual property infringement.

Why it matters:

5.51 Without compelling broadband content, there may be a ceiling on the demand for broadband from residential consumers. Access to entertainment content is also likely to be the stimulus for any residential demand for "broaderband" connectivity in future. In this way, access to broadband content is likely to affect the speed of many of the other changes discussed in this section. Therefore, Ofcom is particularly keen to hear the views of stakeholders on the likely evolution of content business models, and the possible constraints on them.

Question 14: How rapidly are broadband content businesses likely to emerge, and what factors will affect their viability?

 

The evolution of network design

5.52 Alongside changes in the way that consumers wish to use telecoms, technological progress means that telecoms companies may increasingly change the way that services are delivered. In this section we discuss three trends in telecoms network architecture that may have implications for the sources of market power in the telecoms sector, and therefore for regulatory policy. They are:

• the growth of functionality and intelligence at the edge of networks;
• the control over standards and interfaces; and
• the "legacy network" problem.

Growth of functionality and intelligence at the edge of networks

5.53 Annex J describes the trend towards greater and greater performance capabilities being built into the apparatus and devices connected to (or "at the edge of") networks. The increasing demand from customers for options which allow personalisation and customisation, and the development of peer-to-peer file sharing, appear to be strongly correlated with this. Over time, the relative intelligence and functionality of consumer devices connected to the network may continue to increase, particularly while a bandwidth "bottleneck" continues to exist between consumers' devices and the core network.

5.54 This could, in turn, make some elements of the consumer experience more dependent on the device and less dependent on functionality built into the network itself. On the other hand, some kinds of customer data (such as location data) may become more important in offering a rich customer experience. Therefore, control over these kinds of data may become an increasing source of competitive advantage.

5.55 While the deployment of intelligence at the edge of networks may be a strong trend, it may not take place evenly across all networks or service segments. The fixed voice, mobile voice, and broadband business models of different operators currently support widely divergent degrees of intelligence and autonomy at the edge of their networks. It will be important for Ofcom to understand whether this is a short-term divergence or indicative of a longer-term trend.

Why it matters:

5.56 If ever-greater service differentiation occurs in the device, rather than in the network, increased creativity and service innovation may not rely on control over the network. This could allow some of the regulation relating to control over networks to be removed. On the other hand, service delivery could increasingly rely on ownership of data such as customer location data, customers' personal data, software, intellectual property and so on, and this could pose challenges for regulation in new areas.

Control over standards and interfaces

5.57 The trend towards delivery of both fixed and mobile, and voice and data services over IP, has been discussed above. IP is an open standard, over which many different applications (including voice) can run. In principle, the increased use of IP means that much of the scope for innovation and competitive differentiation does not rely on control over the network. Any service provider can design a new service, and provided it runs over IP, they don't need control over the network in order to introduce it.

5.58 In contrast, proprietary standards may be an increasing source of competitive advantage. For example, if network operators deploy proprietary standards over their network, and own the licence to those standards, this may be used as a means of restricting or exploiting activities by service providers. Alternatively, if computer operating systems expand and become network operating systems, developers of those operating systems could gain control over the whole service activation environment.

5.59 In parallel, network operators may themselves come into conflict with equipment manufacturers over control of standards. As Annex J notes, telecoms standards have increasingly migrated towards being specific to, or controlled by, vendors rather than network operators. This may lead equipment vendors to have increasing market power in telecoms markets in the future.

Why it matters:

5.60 To date, much of the focus of telecoms regulation has been about the price that competing operators should pay to access each others' networks. In future, the standards that are deployed over networks, and the terms on which proprietary standards may be used by others, may take on a greater importance and pose new issues for regulation

Question 15: How will future network evolution, such as growth of intelligence at the edge of networks, and the increased importance of control over technical standards and interfaces, affect the requirements of telecoms regulation?

The "legacy network" problem

5.61 The rise in broadband relative to narrowband, and in mobile relative to fixed voice, and of IP networks, could all have an important knock-on effect. There could come a point where the existing circuit-switched architecture starts to become uneconomic for operators to maintain.

Why it matters:

5.62 During the transition, a group of customers may remain reliant on this "legacy" circuit-switched architecture. The fixed costs of this network would remain constant, but would have to be recovered from this (declining) group of customers. If the rate of this traffic migration is not correctly anticipated by the regulator when setting the rate of depreciation of network equipment for the purposes of determining interconnection prices, this could cause windfall gains or losses to consumers or to telecoms network operators.

5.63 The transition is also unlikely to take place evenly across customer groups or geographies. Customers remaining on the old network are likely to be clustered in particular locations and demographic groups. As a result, this transition is likely to raise significant universal service and consumer protection regulatory issues.

Question 16: Will it become uneconomic for operators to maintain the existing circuit-switched architecture at some point and, if so, when? What regulatory issues will this transition to IP networks raise?

The evolution of industry structure and strategy

5.64 Along with changes driven by technology and consumer demand, there may also be a realignment of the telecoms industry itself during the period under review. This may result in changes in market concentration or market power.

5.65 The financial performance of the telecoms sector is discussed in Annex H. It describes how there has been a fundamental re-evaluation of the value of businesses in the telecoms, media and technology sectors. Many network operators are either going through a financial restructuring process or are expected to do so. Many industry commentators have predicted for some time, and continue to predict, realignment in the sector, although opinions differ as to when it will take place.

5.66 This document does not address the underlying reasons for this realignment, but it is very important that Ofcom understands the prospects for the sector and the likelihood of changes in market structure. Were some kind of realignment to take place, it could be through:

• horizontal consolidation, (for instance, mergers of existing mobile networks, or alternative fixed telecoms networks);
• vertical mergers (for instance, service providers merging with network operators, or network operators with content companies);
• alliances which, although not mergers, are designed
  to reap some of the benefits that would flow from horizontal or vertical integration;
• market entry through brand extension, such as mobile telecoms companies extending into the fixed market and vice versa, or service providers and retailers from entirely different markets entering the telecoms sector.

5.67 Given that, in the short run, the majority of network operators' costs are sunk, the principal benefits of horizontal consolidation may be a merging of customer bases and a reduction in the number of competitors. Equally, there are trends towards vertical alliances. For instance, there is a tendency for network operators to partner with systems integrators who are bidding for the managed services contracts of large corporate customers, and tentative alliances are being formed between networks and content providers in the broadband environment.

5.68 Any proposed integration would be subject to the normal approval processes under competition law. However, Ofcom would like to understand the strength of the drivers for integration and alliances in the industry, whether vertical or horizontal.

5.69 In principle, aggressive organic growth or brand extension represent alternative strategies which do not involve merger or partnership. If current operators are sub-scale, or lack the efficiencies which flow from vertical integration, entry into related markets through brand extension may address this by making greater use of existing assets.

Why it matters:

5.70 If a trend towards mergers and alliances outweighs any trend towards market entry through brand extension, greater concentration in the market would be the likely result. This could raise issues for the scope of future regulation. In particular, there might be a case for increasing the flexibility and effectiveness of regulatory devices available to regulate markets where there is more than one player with some degree of market power.

Question 17: Are consolidation, alliances, market entry or other forms of market evolution likely? What will their implications be for telecoms regulation?

The transition and development of the market

5.71 So far, this sub-section has discussed how change in four areas might cause shifts in the nature of market power in telecoms, and therefore have implications for regulation. However, some effects go the other way. Regulation could itself affect the rate of transition towards these outcomes. This is a very important topic; for example, many commentators argue that the rate of deployment of broadband will be significantly affected by the form of regulation adopted.

5.72 In deciding whether to make changes (for example, by entering a new market or investing in new infrastructure), telecoms companies consider the risk and expected returns from doing so. These can be affected by regulation, and this section considers the impact it may have in two areas in particular:

• the impact of regulation on investment; and
• the impact of consumer protection measures, and the impact of regulation on consumer behaviour.

Regulation and investment

5.73 Investment is critical to development in most industries, and particularly in telecoms. Without substantial capital outlays, major improvements to the network, such as digitalisation in the 1980s and the creation of new services such as broadband, would not take place. Nonetheless, encouraging investment should not itself be an aim of regulation; instead it is important that regulation facilitates the economically efficient level of investment. It is therefore important for Ofcom to understand not only the scope for new technologies and services but also the drivers of the investments necessary to convert them from technical possibilities to practical offerings.

5.74 Investment depends on risk and return. Most investment projects involve up-front costs, such as installing physical infrastructure or designing and marketing a new product, which are incurred before significant revenues are earned. These revenues are risky; the investor does not know for certain at the start of the project exactly what revenues will accrue in the future. Risk is therefore an inherent part of undertaking investment. Firms will invest only if, on balance given the range of possible outcomes, they expect to make a commensurate return.

5.75 Similar principles apply to providers of finance, as well as the investing companies themselves. Financiers will not provide funds unless the expected return compensates them adequately for the risk involved.

5.76 In a regulated sector such as telecoms, regulation interacts with the investment process in critical ways. Once capital investment has been made, it may seem desirable to the regulator to push prices down towards the avoidable cost of production. In doing so, the return on investment may be undermined. If investors anticipate this behaviour then they will no longer be willing to invest, even if the project were an intrinsically desirable one.

5.77 In its extreme form, the problem is one of a "regulatory hold-up": knowing that the investment cost is sunk and cannot be recovered, the regulator could, in theory, deliberately expropriate this by reducing prices. However, even in the absence of deliberate intention, regulation might undermine investment through a number of mechanisms.

5.78 First, the mere fear of subsequent expropriation might be sufficient to deter investment. If investors do not trust the regulator, or if the regulator cannot signal its intentions in a way that is credible and reliable, investment may not take place - even though, in fact, the regulator would have taken due account of the capital outlays involved and included an appropriate return in determining regulation.

5.79 Second, assessing the appropriate rate of return to include in regulated prices is not straightforward. There are likely to be debates about the precise cost of capital, even if the calculation methodology is agreed. Moreover, for investments where the initial probability of failure is high, a return to success that merely covers the cost of capital is unlikely to generate an adequate return overall, as viewed from the start of the project.¹² This issue is particularly important in sectors where risks are high and innovative investments are highly desirable; it is less relevant for well-established businesses, or for companies undertaking a range of projects, some of which may succeed and some fail.

5.80 Regulatory systems in different countries use various mechanisms to reduce this problem. In the US, regulated utilities traditionally received a legally guaranteed rate of return on their qualifying investments, and had legal redress to courts to enforce this. On the other hand, the US system leaves open considerable scope for litigation around the details of the regulation and this can introduce unpredictability. The UK has pursued a more flexible strategy, instead relying on the length of the regulatory reviews and the informal "regulatory contract" between regulated firms and the regulator to provide predictability. If the regulator were to behave opportunistically at one point in time, companies are likely to respond to this by cutting back on future investment. Given that companies make a sequence of investments over time, with the need for investment rarely diminishing, it is in the interests of the regulator to respect past investments in order to ensure that investment takes place in the future. In this way, and assisted by stated policy aims, regulators in the UK have been able to create an environment in which investment has taken place.

5.81 Competitors and new entrants, as well as incumbents, are affected by regulatory decisions. Competitors may not be formally regulated but the price that may be charged by a regulated firm generally creates a ceiling on the return that a competing firm may achieve. Where a competitor uses part of the incumbent's network, the competitor's return is partially dependent on the price that it must pay for this access. When both types of competitor are present - those that purchase access to the incumbent's infrastructure and those that compete directly with it - the effects of regulatory decisions become correspondingly complex.

5.82 It is important for Ofcom to understand the sensitivity of capital investment decisions in the telecoms sector, by various parties, to regulatory interventions. It has been argued by some commentators, most recently during the passage of the Communications Act, that previous regulatory approaches have had a "chilling" effect on investment. On the other hand, the evidence would suggest that the UK has not performed appreciably worse than other countries in attracting investment as a result of its approach to regulation - in fact the available evidence suggests rather the opposite; this is reviewed in Annex H. Whatever the answer, the question going forwards is how Ofcom can adopt the optimal approach with regard to future investment.

5.83 One possible conclusion from this analysis might be that the regulator should, in certain circumstances, tolerate higher prices in order to ensure that the appropriate investment incentives are maintained. However, this is a step which no regulator charged with protecting consumers' interests can take lightly.

5.84 It would also be desirable to consider ways in which the regulator may achieve transparency and predictability in its approach to investment, in order to reassure investors that it will not act opportunistically or randomly to the detriment of investment. Some countries have experimented with mechanisms such as "forbearance" from applying regulation (pioneered in Canada, and more recently a feature of the Federal Communications Commission (FCC)'s policy approach in the US), and the setting of clear margins and differentials between products on a forward-looking basis (as used in France). The UK has sought to implement the EU framework in a way which provides as much clarity as possible, but has stopped short of both the approaches identified above. We would like to understand the value that investors place on regulatory certainty, recognising that the granting of certainty comes at a price in terms of loss of flexibility to respond to changing market circumstances.

Why it matters:

5.85 Ofcom recognises the importance of investment by both incumbents and their competitors, where this is efficient and productive. Healthy levels of investment are fundamental to the process of creating competitive markets and, by extension, to satisfying consumers' needs. We therefore need to be aware of the ways in which regulation might inhibit investment that would otherwise be desirable - without, of course, encouraging investment that is inefficient. In particular, Ofcom would like to consider what approaches might be adopted towards creating a consistent and secure regulatory environment in which investment may
take place.

Question 18: What impact do different regulatory approaches have on investment decisions in telecoms, and what regulatory approaches does this imply that Ofcom should adopt?

Consumer protection and behaviour

5.86 Regulation may also affect the risk and return that operators perceive in other ways. In particular, consumer protection regulation, and the impact that regulation has on consumer behaviour, may have an impact. In both cases, there is a trade-off between the beneficial intent of regulation and the possible damage from over-regulation.

5.87 Most competitive markets attract a certain number of unscrupulous suppliers. Telecoms is no exception, and measures are sometimes needed to protect its consumers. Often, self-regulatory or co-regulatory solutions can be most effective in achieving this, provided they include effective enforcement powers. However, over-application of this kind of regulation can slow the progress of emerging products or market trends. Occasionally, gentle application of such regulation might facilitate the emergence of new products, by preventing the unfettered market behaviour of disreputable suppliers from causing such a loss of public confidence that the development of the market in question is retarded or damaged. For example, regulation of premium rate calls was introduced as a response to consumers running up large phone bills unintentionally.

5.88 VoIP is an example of this balance. Requiring VoIP to conform to quality, or other thresholds of traditional telephony, could mean that many of its benefits do not flow through to consumers. On the other hand, regulation may be desirable to prevent VoIP products being launched which are of insufficient quality and not labelled as such, or which attract negative publicity because, for instance, customers do not realise that certain elements of the standard voice telephony service are not available.

5.89 The impact that regulation has on consumer behaviour may also affect the risks and returns that operators perceive. For example, the EU regulatory regime provides national regulators with powers which allow them to remove or limit the impact of certain barriers to consumers switching from one supplier to another. Again, it is important that Ofcom should get the balance right between acting to remove such barriers and interfering unduly in the normal operation of the market.

5.90 For markets to function effectively, consumers must be able to exercise choice and "vote with their feet". In the past, the Office of Fair Trading and Oftel have successfully removed certain barriers, such as lengthy mobile contracts and lack of number portability. On the other hand, over-zealous application of regulation to encourage switching could mean that operators fear that they would not recoup any investments they make in establishing customers with their products (for example, through their spend on marketing). This could discourage operators from introducing new services, to the detriment of consumers.

5.91 A further issue is whether all consumers enjoy adequate access to information, and are able to exercise choice, in the same way. It may be that some consumers naturally find it easier to exercise effective choice than others. Our initial consumer research (discussed in Annex I) suggests that some consumers resent a lack of comparability and clarity in pricing for telecoms services. Research conducted by the National Audit Office¹³ supported this conclusion and also found a lack of awareness of alternative call providers among residential consumers. The House of Commons Committee of Public Accounts¹⁴ has also suggested that the range and complexity of tariff packages and the lack of clarity of bills are confusing to consumers, and prevent them from making meaningful comparisons.

5.92 Ofcom recognises the importance of this issue and that it merits further study. While in theory it may not be necessary for all consumers to be able to exercise effective choice in order for the benefits of competition (for example, lower prices) to flow through to all, it might be a desirable objective for regulation to promote a wider understanding of the choices on offer. As part of the Review, it will be useful for Ofcom to understand the extent to which certain pricing and packaging options might work against this. A critical issue for Ofcom is the extent to which regulation should take an active role in signposting, providing, or ensuring that the market provides clear information to enable consumers to make informed choices.

Why it matters:

5.93 Ofcom is responsible for consumer protection regulation in telecoms as well as regulation designed to influence consumer behaviour (for example,through making switching easier or through ensuring that certain market information is available). It is very important that Ofcom strikes the right balance between protecting and delivering benefits to consumers, and over-regulating markets. Over-regulation could affect the risks and returns that telecoms companies perceive for introducing new products or other activities, potentially to the detriment of consumers.

Question 19: What is the right role for consumer policy? What impact do different approaches have on telecoms companies' perceptions of risk and return?

Question 20: What role should Ofcom take in signposting, providing, or ensuring that the market provides clear information to consumers, enabling them to make effective choices?

Emerging pressures on existing universal service arrangements

5.94 Finally, it is worth considering whether the trends we have identified also have significant implications for the final limb of current regulatory policy: the maintenance of universal service arrangements.

5.95 When BT was a nationalised monopoly, it funded unprofitable services that the Government deemed desirable, through cross-subsidisation from profitable services. Since then, BT and Kingston Communications have continued to be required to provide a set of services to disadvantaged consumers and to consumers in remote locations. A number of other providers also have certain Universal Service Obligations (USOs). The current regulations are described in Annex G.

5.96 Oftel carried out a review of universal service in 1999-2000 which concluded that the costs to BT of being a universal service provider were broadly neutral. However, at some point in the future, the growth in competition could complicate the funding of USOs. Hitherto, elements of USO such as the provision by BT of low-cost access for disadvantaged consumers (the residential Light User Scheme) have been possible through a subsidy from other profitable parts of BT's business. Competition tends to grow fastest in high margin services and hence can be expected to focus on these sources of potential funding for USO arrangements, thus eroding margins. This could potentially call into question the current costs and benefits of USO provision and require a re-evaluation of the funding options. The introduction of new sources of competition, such as the Wholesale Line Rental product, are likely to put downward pressure on call prices in the near term and developments such as VoIP may accelerate this process in the longer term.

5.97 In addition, there may be pressure to change the scope of USOs in the future. Annex G describes how USOs have been applied to services when there are significant network externalities associated with bringing more subscribers onto a network, and/or because inability to access a service by a particular group would result in significant exclusion of that group from society.

5.98 At privatisation in 1984, the proportion of households with a fixed line was just under 80 per cent. The current situation is very different, and the environment is likely to change yet again. Today, less than 1 per cent of households have access to neither a fixed nor mobile phone. The growth in fixed penetration and the maturity of the mobile sector could have a number of implications for the scope of USO obligations relating to voice telephony. For example, it might become increasingly anachronistic to oblige fixed operators to supply a fixed connection at a uniform price if the property is within mobile network coverage and mobile could provide connectivity instead.

5.99 Broadband uptake stood at around three million at the end of November 2003 (about 12 per cent of households), so a long way short of the 80 per cent of households that had a telephone line in 1984 when the USO was brought in for fixed telephony. Although people who do not subscribe to broadband could hardly be considered to be excluded from mainstream society, this assessment may change in the future.

Low-cost access to broadband is likely to continue to be unavailable to consumers who live outside cable TV areas, a long way away from a BT exchange, or in an extremely small BT exchange area. In future, it may become increasingly important to design mechanisms, possibly using wireless or satellite technology, to ensure that these types of consumer have access to broadband at similar prices to consumers living in less remote areas.

Why it matters:

5.100 Ofcom recognises that there is an important "citizenship" dimension to telecoms regulation, much of which is currently articulated through USO provisions. Ensuring that USO remains properly focused on the right outcomes, and that the method of delivery and of funding for USO is appropriate, is therefore extremely important.

Question 21: How many universal service arrangements need to evolve in response to changes in the telecoms market?

Footnotes

7. Two products being (demand) substitutes for one another has a precise meaning in economics. In response to an increase in the price of one product, consumers will switch to purchasing the other one, so that the demand for the latter rises as the price of the former goes up.

8. Substitution between products underpins the market definition approach used by Oftel for the market reviews, which is itself based on competition law. The European Commission’s Notice on the Definition of the Relevant Market for the Purposes of Community Competition Law defines markets by examining substitutability between products. Suppose that the price of a product were permanently raised above the competitive level by a small margin (in the range 5-10 per cent). If substitution to other products were sufficiently high to render the price rise unprofitable because of the resulting loss of sales, the closest substitutes should be included as part of the relevant market. The market is the smallest set of products such that a profitable price rise can be implemented for the group. Although demand substitution is generally the most significant factor in this assessment, substitution may also occur on the supply side, i.e. when a supplier of another product (that is not necessarily a demand substitute) switches its production to the good in question.

9. These trends are not mutually exclusive – mobile networks may also use IP.

10. These technology trends are discussed in more detail in Annex J.

11. Ofcom intends to consult on DRM later in the year, as part of its project to advance broadband development.

12. This point can be illustrated by considering investment to create a new pharmaceutical product. Suppose that, at the point when research is commenced, it is known that it may either succeed in creating a drug that treats a certain medical condition, or fail completely, each with probability one-half. If in the event of success the price that is permitted for the drug merely covered the cost of capital, the expected return on the project as viewed from the start will be half the amount needed to generate an adequate return.

13. Source: Helping consumers benefit from competition in the telecommunications market, National Audit Office, July 2003. The research found that 77 per cent of residential consumers could not name any indirect access call providers. This research is summarised in Annex I.

14. Source: Helping consumers benefit from competition in telecommunications, Eleventh Report of Session 2003-04, House of Commons Committee of Public Accounts.

Back to top