Review of the wholesale local access market

Summary

Summary

Purpose of this document

1.1 This document is a consultation on our analysis of the state of competition in the Wholesale Local Access ("WLA") market, and the measures that we are proposing to address concerns over the degree of competition in that market.

1.2 These proposals matter for consumers because ultimately they will affect the price, choice and availability of critically important retail services, such as current broadband and traditional voice services. However, the WLA market directly concerns services provided between different communications providers ("CPs"). Our proposals also matter because they are intended to promote competition and investment in new super-fast' broadband networks, in the important early stages of development for such networks.

1.3 The WLA market concerns fixed telecommunications infrastructure - the physical connection between a consumer's premises and the local telephone exchange. This connection is needed to support fixed line services such as voice calls and broadband internet access. The cost of this connection therefore affects the prices that consumers pay for those services. Also, if this connection fails then consumers' services will fail. The WLA market is therefore critical to all fixed line services.

1.4 In reviewing this market we are seeking to establish whether there is sufficient competition in the supply of fixed telecommunications connections. Our ultimate goal in this market review is to protect consumers' interests by using regulation to promote competition and choice in the delivery of fixed line telecommunications services. This will help to ensure that consumers do not have to pay excessive prices for those services, and that they benefit from innovation and investment.

The context for our proposals

1.5 As well as looking at current competitive conditions, this market review looks ahead to likely developments over the next few years. This period will coincide with the early stages of the UK's evolution to new super-fast broadband services. These services will be based on network upgrades, to provide connections over optical fibre some or all of the way to the consumer. Such next generation access ("NGA") networks fall within the scope of the WLA market.

1.6 Super-fast broadband will support higher network speeds than those currently available to most UK consumers. Whilst broadband based on traditional copper networks has a maximum download speed of up to 24 Mbit/s, super-fast broadband will offer speeds in excess of this.

1.7 Currently, evidence suggests that, where deployed, NGA networks are being used for video-based applications, including broadcast-quality TV. NGA networks are also likely to be important in providing broadband access to multi-PC homes. Nonetheless, considerable uncertainty remains regarding the range of services that will be provided over super-fast broadband. Experience from overseas deployments shows that there is experimentation in the types of services being offered. We consider that it is important to take a regulatory approach that is flexible enough to allow for experimentation in innovative new products and services, whilst at the same time preventing consumer detriment as a result of firms exploiting market power.

1.8 BT and Virgin Media are both now offering retail services that are based on NGA developments. BT's current plan for deploying its NGA network runs to the end of 2012, by which time it aims to cover around 40 per cent of UK premises. Under BT's plans, three-quarters of those premises would be supplied using by fibre-to-the-cabinet ("FTTC") technology and the rest by fibre-to-the-premises ("FTTP") technology.

1.9 In January 2010, BT's retail business announced the prices it will charge for its BT Infinity' super-fast broadband service as it upgrades its network. Virgin Media has upgraded its cable network, which is capable of serving around half of UK households (although its UK market share is 16 per cent^(-1-)). It has been offering super-fast broadband services since the end of 2008. Other CPs have indicated that they are interested in deploying NGA, either through private investment or based on public funding to allow deployment in areas where commercial deployments may not be as attractive.

1.10 Despite these NGA developments, it is important to recognise that in the next few years, most of the supply in the WLA market is expected to be based on current generation access ("CGA") copper-based networks. We understand that BT does not intend to cease using its CGA network in the foreseeable future. Whilst NGA networks will be used to supply super-fast broadband, CGA networks will continue to support telephony and lower-speed broadband services. Therefore, it is important to continue to deliver effective regulation of CGA networks at the same time as supporting a smooth transition to NGA networks.

The market review process

1.11 We periodically review various markets, according to both European and domestic legal requirements, including the Communications Act 2003 ("the Act"). The market review process is divided into three parts. First, we define the scope of the market that we are assessing (both the products in the market and its geographic scope). Then we assess whether any CPs have a position of significant market power ("SMP"), which in simple terms means the power to influence markets to a significant degree in a way that could harm consumers^(-2-). Then, if any CPs have SMP, we assess which regulatory remedies might need to be imposed to address that SMP.

1.12 As well as this WLA market review, we are currently also reviewing the related Wholesale Broadband Access ("WBA") market, and have also published consultation proposals on that market today ("the WBA consultation document")^(-3-). Our WLA proposals should be viewed in conjunction with the WBA proposals in order to understand the overall proposed impact on consumers.

1.13 The WBA market concerns the wholesale broadband services which are used by CPs to provide retail broadband services to business and residential consumers. The WBA market therefore sits between the WLA market and the retail broadband market. Regulation in the WBA market takes into account how much infrastructure competition there is in the WLA market (including as a result of regulation of the WLA market). Effective and sustainable infrastructure competition tends to give rise to the greatest benefits in terms of the mix of lower prices and faster innovation. Where there is effective competition in the WLA market, further regulation of the WBA market is unnecessary. This is why we are reviewing these markets at the same time.

Summary of proposals

Market definition

1.14 We propose that the WLA services within this market are those based on copper loops, cable networks and optical fibre, at a fixed location. We propose that the market excludes WLA services based on mobile, fixed wireless and satellite technologies. We also propose that WLA services for business and residential use are in the same market.

1.15 We propose that there are two separate geographic WLA markets:

• The UK excluding the Hull area; and
  
• The Hull area.

Market power assessment

1.16 Our proposed SMP findings in the WLA market are as follows:

• BT has SMP in the UK excluding the Hull area; and
  
• KCOM has SMP in the Hull area.

1.17 One of the key reasons why we are proposing that BT has SMP is its high market share, which is 84 per cent^(-4-) . As the WLA market covers both CGA and NGA networks, Virgin Media's earlier NGA deployment does not in itself radically change its market share. Market shares are based on take-up, not deployment (although the two are obviously related).

1.18 As we consider that there is SMP in both of these markets, we are proposing to introduce regulatory requirements on BT and KCOM to address the identified competition problems. Under the legal framework governing our proposals, it is only possible to impose obligations on those CPs that have SMP. We therefore are not proposing regulatory obligations on other CPs in this market including Virgin Media.

Remedies for market power

1.19 To address BT's SMP, we are proposing a number of complementary regulatory obligations (SMP remedies). We propose that BT should have to provide other CPs ("OCPs") with access to its network in the following ways:

• Local Loop Unbundling ("LLU"): a current remedy, this allows OCPs to physically take over (or share) BT's existing copper lines between the local telephone exchanges and the customer premises;
  
• Virtual Unbundled Local Access ("VULA"): this would have to be provided by BT wherever it has deployed its NGA network^(-5-) . The intention is that VULA would provide access to the NGA network in a way that is similar to how LLU provides access on the CGA network. However, rather than providing a physical line, VULA would provide a virtual connection that gives OCPs a dedicated link to their customers and substantial control; and
  
• Physical infrastructure access ("PIA"): like VULA, this is a proposed new remedy. This remedy would allow OCPs to deploy fibre in the access network using BT's ducts and poles - either to support deployment of FTTP technology, or to support deployment of FTTC technology (by enabling a backhaul' connection between street cabinets and the OCP's network). BT would be required to produce a draft reference offer ("RO") for duct access within three months, with a view to launching a product within eight months.

1.20 In addition we propose that BT should continue to provide sub-loop unbundling ("SLU"). This is a current remedy (but currently only used in very limited situations) which allows OCPs to physically take over (or share) the part of BT's existing copper lines between a street cabinet and the customer premises. This remedy would allow OCPs to deploy FTTC technology where they consider this to be economic.

1.21 Based on the proposed specific access products, OCPs will be able to use BT's network infrastructure to develop their own services to offer to consumers, thereby lowering barriers to entry and investment. VULA would also support competition in (downstream) voice markets by providing BT and OCPs with an equivalent input for developing those voice products. We would expect BT's downstream businesses to use VULA as an input when providing voice services over fibre.

1.22 At this point, we consider VULA to be the primary focus of NGA competition, to supplement the continuing effective LLU remedy over at least the next four years. Our economic analysis suggests that VULA is very likely to be the most cost-effective NGA remedy to support competition. However, we think that both SLU and access to BT's ducts and poles could also play a part in supporting competition, as well as investment in NGA. Partly, this is because VULA will only be available where BT deploys its NGA network.

1.23 We propose that prices for LLU, SLU, and PIA should be related to the cost of providing them^(-6-). However, we are not proposing to set regulated prices for the product(s) that BT provides under its VULA obligation. This would give BT the flexibility to price its VULA services according to emerging information on the demand for, and supply costs of, NGA services.

1.24 In addition to requiring the specific products above, we are also proposing a set of general access remedies on BT, all of which are imposed on them currently in this market. These include a requirement to provide network access, an obligation to not discriminate unduly when providing services, various transparency measures (including on its quality of service), and a requirement to keep separate accounts for different services (to support effective regulation). For VULA, we propose to apply a strict interpretation of the no undue discrimination obligation, which would mirror the definition of Equivalence of Inputs ("EoI") in the BT Undertakings.

1.25 For KCOM, we are proposing to maintain the general access remedies that are currently imposed on it in this market^(-7-). These include a general requirement to provide network access and a no undue discrimination obligation. We are also proposing to add a general requirement on them, to meet reasonable requests for new network access. We are proposing this to encourage OCPs to enter the WLA market in Hull. At this time, we are proposing - in line with the current position - that KCOM should not be required to provide any specific access products, such as access to its duct and pole infrastructure.

Reasoning for our key proposals on remedies

1.26 In proposing the combination of remedies on BT outlined above, our primary goal is to promote effective and sustainable competition across the range of different situations that will exist in this market in the next few years. At the same time, we are mindful of the beneficial impact on competition and consumers of promoting investment in NGA services at this important time for their development.

1.27 We have commissioned independent research to assess the costs and practicalities of achieving NGA competition in BT's network. The evidence suggests that where BT deploys an NGA network, the most cost-effective and straightforward way to support competition is to allow OCPs to access that new network, rather than to invest in their own parallel NGA infrastructure. This is why we are proposing the VULA obligation on BT. The information available to us at present also suggests that, as long as BT delivers a service with the requisite characteristics, VULA would offer the best prospects for reproducing the kind of competition in NGA (at least in the short- to -medium term) that LLU has provided in CGA.

1.28 However, we think that there is potential for other WLA access remedies to contribute to delivering competition and NGA investment, because:

• SLU can be used (in combination with duct access) to deploy an FTTC network; and
  
• PIA can be used to deploy an FTTP network and/or for FTTC backhaul.

1.29 There are a number of reasons why we consider these two additional remedies to be appropriate:

• BT's NGA deployment plans currently only cover 40 per cent of UK premises. Also, the sequence of those deployments is not currently certain. Therefore, these other remedies could enable OCPs to deploy NGA infrastructure before BT in some areas, thereby competing with BT's existing CGA network^(-8-) and providing NGA services earlier to some areas;
  
• Whilst the cost of competing based on SLU or PIA appears at present to be high, compared to using VULA, technological developments could change the relative costs over time;
  
• The relative benefits of providing NGA services in different ways are currently unclear, because the demand for NGA services is still at an early stage. The type of services and level of demand might in due course give more support to SLU and/or to PIA;
  
• The prospects for using SLU and PIA will differ between areas, so both of these options should be available. The reasons for this include: the variability in the capacity available in BT's network; small street cabinet sizes that make it uneconomic to serve a limited number of customers using SLU; and the varying locations of OCPs' existing networks. We consider that BT's own plan for a mix of NGA deployment types supports our proposals.

1.30 As well as providing OCPs with the opportunity to compete with BT's and Virgin Media's NGA deployments, these WLA remedies will give OCPs the opportunity to provide NGA services in areas where there are no current BT plans to deploy NGA services. We consider that having a choice of remedies available to suit different locations is in the best interests of consumers across the UK, and may help to limit the prospects of a digital divide in the delivery of NGA services.

1.31 Our proposals to support different means of NGA deployment are separate from current central government proposals on universal broadband provision and public subsidy of NGA developments. However, we consider our proposals to be complementary to those initiatives, as they should lower the barriers to investing in NGA infrastructure. Additional public subsidy to encourage NGA deployments could be used in conjunction with the additional network access products that we are proposing. If such investments occur in locations that would not otherwise have attracted commercial interest, at least not on the same timeframe, the proposed access obligations on BT would also contribute to meeting universal broadband targets. However, as with CGA, even with effective SMP remedies in place, there are likely to be geographic limits to how far CPs will want to invest in NGA infrastructure.

1.32 It should be acknowledged that there are challenges in developing competition based on SLU or PIA. However, we nevertheless believe that these remedies could have a significant role to play. Therefore, for the PIA obligation, we are proposing a firm timeline for BT to develop and implement usable products and processes. Our aim is to reduce the barriers to entry, and to equip OCPs with the relevant information on their choices. It will then be for the market to determine the best way of competing, which may differ between areas and between CPs.

1.33 With regard to our proposals to allow BT flexibility in the pricing of its VULA products(s), we consider that this gives BT the opportunity to promote the efficient deployment and use of its new NGA network (and thereby recover the costs of developing it). We consider that this approach will promote investment in NGA. We also consider that whilst BT still has its CGA network, and is providing services to OCPs using that network, the (regulated) prices that OCPs pay for CGA services will help to constrain the prices that BT can charge for the (upgraded) NGA services.

1.34 With regard to the Hull area, to date OCPs have chosen not to compete in the WLA market (or even in the retail market). We therefore consider that it would be disproportionate to require KCOM to invest in developing specific access products, as it seems unlikely that they would be taken up. We consider that the general access remedies on KCOM are sufficient as they will allow any OCPs to request access if they wish to enter the Hull market. However, we are proposing to introduce a new requirement on KCOM as a means to encourage entry into the Hull market. Under this requirement, KCOM would have to create a Statement of Requirements ("SOR") process, which should clarify the process for requesting new access products.

1.35 We are aware that consumers in the Hull area have a very limited choice of providers, because of the lack of entry into the market by providers other than KCOM. Potentially, this lack of competition could result in consumers in Hull paying higher prices and receiving less attractive service propositions. This might suggest a need to consider closer regulation on KCOM's prices and services at the retail level. However, we have examined the retail offers available to consumers in the Hull area. This shows that whilst consumers in Hull may not have access to the best offers available in some other parts of the UK (where providers such as Sky and TalkTalk have deployed their own networks), they do have access to products that are comparable in terms of price and specification to those available to many consumers in the rest of the UK. Therefore, whilst we will continue to monitor the position in Hull, we do not propose to carry out a further review of the retail market at this time^(-9-).

Consultation and next steps

1.36 We invite comments from interested parties on the proposals in this document. The consultation period runs for 10 weeks, to 1 June 2010. We aim to publish our conclusions during Autumn 2010.

1.37 There are also a number of important practical issues to consider if we decide, following the consultation process, to implement the proposed new SMP remedies on BT. Firstly, there would be a need to ensure that BT's VULA-based product(s) is fully compliant with the characteristics that we specify for this product.

There are currently some differences between VULA and the Generic Ethernet Access ("GEA") products that BT (Openreach) has been developing to provide a wholesale service to CPs based on its FTTC and FTTP NGA deployments:

• For the FTTC GEA products, BT has so far proposed to make them available only in combination with other products/service. However, we consider that a VULA service would need to be made available on a stand-alone basis; and
  
• For the FTTP-based GEA products, BT has chosen to embed a voice analogue telephone adaptor ("ATA") into the network termination equipment ("NTE"), which currently is a necessary part of the GEA product. We recognise that there are good economic and commercial reasons for this approach. However, to meet the VULA requirements, we consider that the arrangements known as open ATA' (which include control for interconnecting CPs) will be an essential requirement if the voice ATA is embedded into the GEA/VULA product.

1.38 Further work would also be needed on the details of the PIA products that BT should offer. This would require the involvement of OCPs as well as BT, and would need to happen before detailed work on the pricing of these products could be done. Our proposed obligation requires BT to produce an initial RO for duct access, describing the service to be made available, within three months of the obligation being introduced. Significant OCP involvement would be needed at that time, and ideally before, to take BT's proposal forward.

1.39 We will be considering these practical issues further during the consultation period, and we would welcome discussions with CPs during that time to inform our decisions on any relevant obligations.

Footnotes:

  1.- As a proportion of active lines

  2.- Formally, SMP is defined as a position of economic strength affording an undertaking the power to behave to an appreciable extent independently of competitors customers and ultimately consumers

  3.- See http://www.ofcom.org.uk/consult/wba/

  4.- Of active access lines in the UK

  5.- This would apply both to FTTC and FTTP deployments

  6.- Based on the long-run incremental cost of provision, including an appropriate element of BTs common costs

  7.- in the Hull Area only

  8.- We are proposing that CGA and NGA services are in the same WLA market

  9.- See the WBA consultation document for retail price comparison for KCOM services

The full document is available below