House of Lords - Broadband for all - an alternative vision

Broadband for all - an alternative vision - Communications Committee Contents

Chapter 4: Modifying the Government's strategy

151. Broadband and its infrastructure inevitably represent a highly technical field; and the regulatory arrangements which structure and govern the territory are no less complex. One can be forgiven, therefore, for finding a journey into the detail of policy implementation in this area forbidding. However, it has been our discovery that many of its smaller, more intricate features are responsible for the contours of the wider landscape. As a result, it is important, for the proper scrutiny of the area, not to be deterred by its jargon or detail.

152. It is for this reason that we have taken a substantial amount of evidence on the implementation of the current policy, as well as the regulatory arrangements that surround it. Having considered this, we make a number of observations and recommendations, some of which are inevitably expressed in more technical language. As much as possible, however, we will try to explain the terms used, occasionally by placing background information in boxes, and certainly, we have sought throughout to spell out the important implications of our proposals in the most straightforward terms. Broadly, the observations and recommendations which we now make fall into three groups:

· Those which assume no change in the Government's overall approach;

· Those which call on the Government to change its approach, effectively bringing it closer to our alternative proposal;

· Those which call on the regulator, Ofcom, to change its approach, with the same effect;

We will take these in turn.

Recommendations assuming no change in the Government's overall approach

153. Our focus up until this point has been on a reorientation of Government broadband strategy away from the edges and towards a reorganisation and enlargement of the middle mile of the network, and in particular to the introduction of open access fibre-optic hubs. However, to a large extent, the Government have set the direction for broadband policy to 2015, and clearly much of the work this has set in train is already underway through four strands of activity:

· Three of which are under the auspices of BDUK:

— BDUK management and allocation of £530 million for the 'final third';

— The Mobile Infrastructure Project (MIP);

— Superconnected Cities.

· One further strand of which is a responsibility of Defra:

— Rural Community Broadband Fund (RCBF).

154. Problems with the last of these, the RCBF, are particularly startling. Fortunately, the solutions should be equally straightforward for the Government to develop. By design, the RCBF is intended to subsidise enhanced broadband infrastructure to those areas of the UK considered to be, not only in the final third, but also in the final 10%, which will almost certainly be rural. The reason they need special attention is that communities and properties in these areas are generally acknowledged to present the most severely challenging investment case; even with BDUK subsidy, it is considered to be too weak.

155. Accordingly, the RCBF is only available "if your community is in a rural location identified as being in the 10% hard to reach area covered by your Local Authority's Local Broadband Plan."[84] In other words, it is only available if you are excluded from your Local Authority's Local Broadband Plan.

156. Miles Mandelson, Chairman, Great Asby Broadband, explained that when local authorities are asked whether they can guarantee that a specific community falls outside the boundaries of its roll-out, the answer is:

"'No, we cannot. We have not finished our procurement. We do not know what the boundaries are going to be and the situation keeps changing. But, if you like, you can exclude yourself from the county council roll-out and we will not come anywhere near you.'"[85]

As there is equally no guarantee of inclusion within the RCBF scheme, communities are naturally wary. After all, as Miles Mandelson put to us, if: "their application [to the RCBF also] fails… they are then left with nothing more than the 2Mbps universal service commitment (USC)."[86]

157. As a result, remote communities—particularly those engaged as part of BDUK's pilot scheme—are being forced to take on a considerable amount of work and responsibility, with the entirely likely result that they will ultimately be told they fall outside the boundary of the roll-out.

158. A further failure of the RCBF arises from its specific stipulation that communities in the final 10% fund the construction of their access networks with no guarantee of grant payment from Defra; they will only be able to recover the expenditure once they have provided evidence that they have paid for it—and even then there is no guarantee. According to Defra's RCBF Handbook:

"In simple terms, you will need to spend first—and then claim later."[87]

159. In our view, it is ludicrous to expect all remote communities to be able to provide the levels of funding required to build broadband access networks in their areas. While there may be some which can, inevitably others will not be able to do so. The RCBF process, therefore, will present a number of communities, whom it is intended to assist, with a Catch 22. It must be said that there is a lack of clarity over the origin of this stipulation. The RCBF funds derive from a wider Defra programme called the Rural Development Programme for England, which in turn draws its funds from both the UK Government and from the European Agricultural Fund for Rural Development. Whether the stipulation derives from a Europe-wide policy or is one of the Government's own making, however, does not alter the absurdity of its assumptions, which must be resolved.

160. We urge the Government to provide a more coherent mechanism for the provision of enhanced broadband infrastructure in the final 10% than currently is the case with the Rural Community Broadband Fund. In particular, a new mechanism for distributing funds must meet the criticism that its predecessor was flawed in assuming all communities have the capital required, up front, to invest in their own access network.

161. We also wish to highlight a number of other ways in which the Government could usefully intervene without fundamentally having to change their current approach. The most obvious examples relate to the planning system, street works permissions and the Electronic Communications Code; aspects of all of these, as they stand, can cause delays to the roll-out of broadband infrastructure.

162. The planning system can cause delays because of the length and cost of consultation processes. As Antony Walker, CEO of the Broadband Stakeholder Group, told us:

"One of the biggest barriers to the use of poles is the planning system. The consultation processes that are required if you want to utilise poles or put up new ones are pretty onerous and off putting to investors."[88]

Similar points were made by Vodafone:

"Many delays during the planning process are caused by tokenistic opposition which is more often than not overturned on appeal. Delays add significant actual costs, as well as the opportunity costs for the local economy."[89]

163. Street works permissions can equally add serious delay. Sean Williams, Group Strategy Director, BT, told us:

"The street work guidance for permit schemes can result in councils designating the putting up of the street cabinet as a major work, and if it does that then it causes all sorts of costs and delays."[90]

164. Finally, delays can also arise from the process for gaining wayleave permissions—a licence granted by the owner and occupier of land giving a third party the right to install, use and maintain its equipment on that land—provided for in the Electronic Communications Code. Vodafone argued that: "The current system of high annual ground rents for the lifetime of the site imposed by the code is one of the greatest disincentives to the roll-out of mobile coverage and should be reviewed."[91]

165. Sean Williams, Group Strategy Director, BT, argued that a better mechanism for dispute resolution with regard to wayleave permissions was required. Wayleaves, he argued, are:

"particularly important for multi-occupancy dwellings, whether they are blocks of flats or retail centres with businesses in them, where it is very often difficult and costly to negotiate wayleaves, and indeed there is no effective mechanism for resolving wayleave disputes. We could very readily do with some support in that area, to get a dispute mechanism, and also to put some obligations on landlords to make this possible, as I think they do in other countries."[92]

166. As part of its wider communications review, the DCMS has asked the Law Commission to conduct an independent review of the Electronic Communications Code. This project began in September 2011 with a consultation paper being published on 28 June 2012. The Commission expects to publish its report in spring 2013 and for any reforms to be implemented through legislation, anticipated in 2015.

167. In their deliberations over the potential reform of the Electronic Communications Code, we encourage the Law Commission to consider the impact of the Code on the roll-out and availability of broadband infrastructure throughout the UK.

168. We urge the Government to consider reform of street works permissions and the current planning system, given their wider impact on the pace and sheer viability of the roll-out of broadband infrastructure throughout the UK.

169. Finally, we note that new build sites can easily have ducting for fibre installed. It is, therefore, our view that for new developments, the provision of fibre connectivity should be as automatic as the provision of mains electricity.

170. We recommend that the Government require that all new building developments be ducted for fibre, with appropriate provision for an internet connection, and that building regulations for this be developed perhaps analogous to those which already require adequate provision, for example, for the delivery of mains electricity and sewage connections.

A change of approach in Government policy

171. Up to this point in the chapter, we have made recommendations which the Government are able to act on now, without rethinking the broader strategy they have set for 2015. However, in exploring its detail and implementation, we have also encountered a number of features which reinforce the case for considering an alternative strategy altogether.

We will pick up three particular points:

· The impact of inflexibility in the current policy targets;

· The wider potential of the Mobile Infrastructure Project (MIP);

· The importance of monitoring infrastructure providers in relatively non-competitive markets

The impact of inflexibility in the current policy targets

172. In its original strategy document, the Government maintained that: "our approach to delivery... [is] technology-neutral."[93] However, it is widely acknowledged that the significance of the original 24Mbps target lies in the fact that it specifically aimed to exclude as an eligible technology any copper between the exchange and the cabinet; as noted in Chapter 2, 24Mbps relates to—what was at the time—the highest headline speed achievable over copper lines from the exchange.

173. In line with the alternative strategy sketched in Chapter 3, we endorse the replacement of copper by fibre deeper into the network. However, there are indications that mandating this shift by means of a speed target has built inflexibility into the implementation of the current policy, with a number of undesirable consequences.

174. One of these was described by Rory Stewart MP, in reference to Northern Fells, an area in his constituency. Having formed plans to bring fibre within reach of its village, the community considered the use of the wireless technology, 'white space', as a way of reaching its outlying properties. The problem that arose was that white space spectrum is technically limited to delivering a maximum of roughly 15Mbps, and as this falls short of the Government's targets, it is ineligible for public money, stranding the outliers with a lower level of service. As Rory Stewart MP put it:

"I think it is about flexibility. I think it is about loosening the reins a little, loosening the criteria and being prepared to look at each case on its merits."

175. Curiously, by mandating a specific speed target, with the intention of pushing fibre—and by extension, enhanced capacity—deeper into the network, the Government have in some cases ensured that communities reliant on more experimental, innovative technological solutions will not get any enhancement at all. In short, because their plans do not meet the Government's aspirations in full, they will be condemned to nothing.

176. We note that the European Commission's draft guidelines on State Aid stipulate that financial support should be provided only to those projects which can deliver a "step change." This is not defined, however, in terms of speed:

"A "step change" can be demonstrated if as the result of the public intervention (1) the selected bidder makes significant new investments in the broadband network and (2) the subsidised infrastructure brings significant new capabilities to the market in terms of broadband service availability and capacity."[94]

177. The refusal to provide financial support for a project, like that in the Northern Fells, on the grounds that its proposal to use a technology (in this case white space spectrum) which would not meet the Government's speed targets, is a further illustration of the way in which such targets are actually counter-productive. We urge the Government to reconsider using speed targets to define the goals of their broadband policy. This would allow them to be more flexible with regard to the technologies used to provide enhanced connectivity, particularly to outlying communities. Loosening the reins a little could very quickly have the effect of bringing enhanced broadband capacity to the final 10%.

178. The alternative strategy we have put forward would avert the situation which has arisen under the current policy whereby communities are left stranded with a minimal service because a viable enhancement falls below the—relatively arbitrary—mark set for public funding. We invite the Government to respond to our proposal that bringing open access fibre-optic hubs within the reach of every community would liberate communities and enterprises to evaluate the cost—benefit calculation themselves of the various different technological solutions available in the access network.

The wider potential of the Mobile Infrastructure Project (MIP)

179. The second area of policy implementation which helps make the case for an alternative strategy relates to the Mobile Infrastructure Project (MIP) whose first two "design principles" are as follows:[95]

· "To improve mobile voice coverage to outside of premises for the five to ten per cent of consumers and businesses that live and work in areas of the UK where existing mobile coverage is poor or non-existent;

· To enable the removal of 'complete' voice not-spots without converting them to 'partial' not-spots."

180. The Government do not refer once to data coverage, emphasising instead the role of the MIP in eliminating voice not-spots.[96] This point was reinforced by Robert Sullivan, CEO, BDUK:

"It is important to make that distinction—that rather than being a broadband programme [the MIP] is a voice programme"[97]

181. This strikes us as odd. According to Ofcom's most recent Communications Market Report, "the usage of smartphones as a means of accessing the internet has... risen quickly. Ofcom estimates that in 2011, 32.6 million subscribers accessed the internet via their mobile phones, an increase of nearly 10 million since 2010."[98] They also note the ways in which "new devices shape an explosion in mobile data use" and that "volume of data use increased significantly in 2011." Similarly, a recent report commissioned by 02 showed that among smartphone users, internet access is by some margin the most popular activity.[99]

182. It should be a fundamental 'design principle' of the Mobile Infrastructure Project that where mobile coverage is being widened for the purpose of eliminating voice not-spots, coverage for data is widened and enhanced at the same time.

183. In addition, we note that BDUK is considering implementing the MIP objectives by means of "an investment by government in building Serviced Sites… in an appropriate location to serve one or more not-spots, where site facilities, power, backhaul, connectivity, cabinet and mast space were made ready and available for 'occupation' by all mobile operators."[100]

184. In other words, the MIP, in establishing a number of 'base stations' or 'serviced sites' with fibre backhaul, may bring about something equivalent to the open access fibre-optic hub which we introduced in Chapter 3. In fact, should the MIP be implemented along the lines outlined above, these serviced sites will be procured with clear open access stipulations, allowing any mobile network provider to set up their own access network services from those points.

185. There is a technically strong argument, therefore, given the virtually unbounded capacity of fibre, that it would be a significant enhancement to incorporate into the MIP plan the following stipulation: that access be open not only for mobile network providers, but also for any prospective network provider to build out an access network from that point, should they see a commercial opportunity or otherwise believe there is value in doing so.

186. The Government should consider the potential for serviced sites constructed as part of the MIP to be used as open access fibre-optic hubs more generally, from which independent third parties could extend out their own alternative, local access networks.

187. Incidentally, in considering the harmonies between mobile and broadband policy more widely, we note that in parallel with the investments being made into fixed infrastructure, the forthcoming 4G auction will equally serve, in its own way, to widen access to enhanced capacity broadband. We do not comment other than to welcome this development, in particular the recent announcement by Ofcom that the 4G auctions are set to get underway by the end of 2012 and that a stipulation of these auctions will be that one of the lots of spectrum "will carry an obligation to provide a mobile broadband service for indoor reception to at least 98% of the UK population by the end of 2017 at the latest."[101] Our focus in this report, however, has purposefully been on the fixed infrastructure, and where relevant, the ways in which this provides a necessary foundation for mobile data networks.

The importance of monitoring infrastructure providers in relatively non-competitive markets

188. A final area which reinforces the case for the Government to re-orient their strategy, in the way we advocate, towards the establishment of open access fibre-optic hubs lies in a consideration of the non-competitive nature of much of the UK's broadband infrastructure. From the nine original bidders in the BDUK Framework process, only two remain: Fujitsu and BT. In fact, this may now be reduced to one as recent reports suggest that Fujitsu has withdrawn from the process for the time being, albeit that it is committed to bidding in the future.[102]

189. For some witnesses, this is merely evidence of the inevitable characteristics of investors in broadband infrastructure. Richard Hooper OBE, Chairman, Broadband Stakeholder Group, told us: "this is, whether one likes it or not, an industry of scale."[103]

190. For others, the outcome is evidence of the way in which BDUK's Framework has excluded less established bidders. Gap-funding, the lack of open access to existing infrastructure, and the relatively homogenous nature of the customer base within an area as small as a local authority gives very considerable advantage to bidders with scale. We outline in Box 4 the ways in which the BDUK process may have reduced the ability of a wide range of bidders to compete.

BOX 4

The non-competitive nature of the BDUK process

There are two particular features of the BDUK process which may have limited the ability of some providers to compete:

· the size of the bodies, particularly in England, acting as procurement bodies (local authorities)

· the gap-funding model adopted by BDUK. The "investment gap" is the public contribution required by a supplier's investment in broadband infrastructure to make a project commercially viable. [104]

In the case of the size of the investment authorities in England, networks within a local authority area may not be of sufficient scale to be sustainable without the assumption that the provider building the network already has scale across the country. If they do, to simplify the point, they can cross-subsidise from densely populated areas of the country with lots of potential end-users, to sparsely populated areas with only a few. As a local authority area may not contain sufficient diversity for this sort of sustainable cross-subsidy itself, an infrastructure provider unable to draw on national scale, and the ability to cross-subsidise over county borders, is likely to be unable to compete. Equally, the selection of the local authority as the focus for investment, rather than fewer bodies representing larger franchise areas, has resulted in a proliferation of procurement processes, requiring more resources to compete in a number of them at once. Inevitably, this has reduced the number of businesses in a position to bid across the country.

Similarly, consider BDUK's preference for gap-funding as the investment model to be used by local authorities. This favours large-scale businesses with a high level of knowledge about the market and likely level of demand. Having this knowledge enables them to make accurate judgements about the value of an investment, and therefore bid with a greater degree of capital, presenting a smaller, more attractive investment gap to the body making the investment. Similarly, it favours incumbents with reliable existing revenues, as these too decrease their risk and mean they can bid with a greater degree of capital, and present the procurement authority with a smaller investment gap.

191. Equally, regardless of the BDUK process, the history of UK communications infrastructure is such that there is now a provider, BT, to whom advantages accrue given its ability to draw on diverse and reliable revenue streams, and an extensive understanding of demand across the UK. These characteristics reduce the level of risk presented by infrastructure investment, and therefore afford those who benefit from them an unparalleled advantage in the competition for infrastructure investment (see paragraph 88).

192. Both views, therefore—those associating the non-competitive nature of infrastructure provision with the characteristics of the UK market and, those associating it with the characteristics of the BDUK process—are right to an extent. In fact, there are other reasons too, relating to the regulatory regime; we discuss these later in this chapter. In our view, however, the foregoing observations are ultimately less important than what one concludes from them. These conclusions should be straightforward:

193. There are clear competitive advantages in the market for infrastructure provision which accrue to those able to draw on economies of scale and scope. If unchallenged, these may result in the dominance of a single provider.

194. As suggested in Chapter 3, the danger that results from the lack of competitive pressure in the construction of the UK's broadband infrastructure lies in the fact that the Government can easily find itself in thrall to the commercial interests of private enterprise, and therefore unable to direct broadband infrastructure in the wider interests of the UK.

195. We urge the Government, therefore, to recognise as a general principle that it will be vital to monitor the dominant, national providers vigilantly and to deploy appropriate incentives to ensure they, and the market in which they operate, behave in the public interest as this will not necessarily follow automatically from competitive pressures alone.

196. In addition, we note the argument of Chapter 3, that despite the presently non-competitive nature of much infrastructure provision in the UK, open access to existing connectivity can enable competition to play a role in extending the reach, connectivity and diversity, and hence also the resilience and performance, of these networks.

PON

197. In the light of the previous discussion, we are concerned about some of the decisions currently being made by the providers bidding for BDUK projects. In particular, where it is rolling out FTTP, BT is doing so using a technology called PON (Passive Optical Network)—see diagram below.

FIGURE 2

Illustration of PON

198. This involves single fibres being laid from the exchange to the cabinet. However, from the cabinet to the premises, these single fibres are not simply extended out, creating a single fibre link all the way from the exchange to the property. Instead, at the cabinet, each fibre (e.g. fibre A in the diagram above) is split into a number of strands, and each premises receives a strand from it (e.g. A1, A2 or A3). As a result, this form of FTTP tends to be called a point-to-multipoint, rather than a more straightforward point-to-point, solution. BT's decision to employ PON for its roll-out of FTTP has a number of consequences about which we harbour concern, notably on the issues of unbundling and future upgrade.

BOX 5

Consequences of PON

At present, physical 'unbundling' of PON is impracticable. Physical unbundling is the mechanism whereby competitor communication providers take control of a single connection from the exchange to the premises even though the physical line (currently in the form of copper) is owned by an incumbent. As well as taking control over the line, the competitor also installs their own 'active' equipment in the exchange and customer premises. As the active equipment is partly responsible for the characteristics of the service (speeds, reliability etc.), this kind of Local Loop Unbundling currently allows BT's ISP competitors such as TalkTalk to differentiate their products from BT—and this in turn forms the basis for real competition.

It is not practicable, however, to unbundle PON because of the way each fibre splits at the cabinet. To do so, given current technology, it would be necessary for an ISP to attract as customers all of the premises which happen to be connected to a single fibre through one of its strands (in the region of 30-100 separate customers). There are technologies in development (e.g. wavelength-division multiplexing) which would enable unbundling of PON and the targeting of individual customers, but these are currently unproven, and require expensive active equipment. As such, until a commercial model emerges to enable a competitor to take over a connection to all premises served from a single fibre, or until new technology emerges, PON cannot be physically unbundled, and it is not possible for BT—as long as it is installing PON technology—to provide a desirable level of open access to its publicly-funded fibre access networks. As TalkTalk put to us in written evidence:

"Under a GPON[105] architecture, all ISPs share the same bandwidth and equipment. This means that BT specify most of the features of the product such as: the maximum speed; the total shared bandwidth; how different traffic is prioritised (important for IPTV and business services); and, the electronics/equipment placed on the both ends of the fibre." [106]

It is on this basis that the European Commission has agreed to Ofcom's proposal that BT must offer an alternative to full, unbundled open access, namely Virtual Unbundled Local Access (VULA). This simulates unbundled access to BT's fibre but is different from full unbundling in one crucial respect: the competitor ISP not only uses BT's physical fibre connection, they must also use BT's active hardware at each end. This makes it impossible to differentiate products to customers, and means that the main basis for competition is brand. It appears to us, therefore, that Government subsidy is being used to fund a new world in which there may actually be less competition than there was over copper—reinforcing BT's market power.

Another unfavourable, consequence of the use of PON technology relates to its 'upgrade path.' PON is not—technically speaking—an interim step towards full point-to-point fibre; in fact, it is something of an evolutionary cul-de-sac. If a future Government were to have as a policy goal the achievement of full point-to-point fibre, much equipment and fibre, once PON is already installed, will simply have to be replaced, conceivably with further public outlay.

199. In the light of Box 5, we believe it is incumbent on policy to ensure that there is a clear upgrade path from any intermediate steps which may be taken, to the roll-out of point-to-point FTTP. This will avoid major investments being made, particularly with public money, into technologies which represent an evolutionary cul-de-sac, and cannot be upgraded to point-to-point FTTP without being completely replaced, possibly with more public money.

200. We recommend that the Government's approach be explicit in its insistence that the technologies and infrastructures in which companies using public funds decide to invest be ones which offer a clear 'upgrade path' to point-to-point FTTP.

201. Incidentally, there is an approach using PON with fewer harmful effects on competition than is suggested above. If a network is operated as a PON but the splitter is moved up to the level of the local exchange, and alternative operators can connect at that point, this would realise most of the capital expenditure savings of PON, but would also enable passive unbundling, and as a result, would equally enable real competition between ISPs.[107]

202. Where infrastructure providers using public money decide to invest in Passive Optical Networks, we recommend that the awarding of public money should be contingent on the installation of the splitter at the level of the local exchange rather than the cabinet, as this would enable passive unbundling, and thereby real competition between ISPs.

Regulatory changes

203. The final area in which we propose changes to the current regime in order to make it possible to realise our vision is regulation. The success of the alternative strategy we outlined in Chapter 3 would depend partly on a number of specific shifts in the overall regulatory model for the broadband industry. One shift in particular would be crucial: the establishment of a set of open industry-led standards for both the physical network, and the administrative interface between infrastructure and services providers.

204. In order to understand the significance of this measure, envisage the alternative broadband infrastructure we depicted in Chapter 3, providing newly established open access fibre-optic hubs within reach of every community. Now consider a prospective infrastructure provider, weighing up the decision to build and own a new access network, connecting the hub with the local premises around it. This could be a community interest group, an SME or a large-scale infrastructure provider. Beyond the investment case itself, they will be aware of a number of hypothetical risks and pitfalls, including the possibility that once the network is built, no service provider will be willing to provide a service to the premises on the network, effectively rendering it stranded and redundant.

205. Unfortunately, this has happened. Digital Region Limited, a regional FTTC network in South Yorkshire, provides the textbook case, as Antony Walker, CEO, Broadband Stakeholder Group, told us, "where a network was built and then the operator really struggled, in fact failed, to get any big national ISPs to provide service on its network." [108]

206. There are a number of reasons why an ISP might find it uneconomic to provide a service over such a network. These include:

· The administrative systems of the infrastructure provider and the interface through which the ISP interacts with them are a serious hurdle. Most seriously, an ISP will not be able to countenance the administrative cost of dealing with a proliferation of different such interfaces and procedures. Equally, administrative systems have an impact on the procedures for matters such as switching customers and handling complaints. For example, when an issue is raised with the ISP about a fault whose source is in reality something physical (e.g. a broken line), there needs to be an easy mechanism in place for alerting the infrastructure provider and arranging a repair visit.

· The technical standards of the products which the ISP can offer over the infrastructure are partly determined by the wholesale product offered by the infrastructure provider. This creates two additional issues. First, the ISP will want these technical standards to be harmonious with those of the wholesale product it purchases elsewhere in the country, in order to avoid a situation in which its retail product, and by extension its brand, are tarnished through an inconsistency in their quality. Second, the ISP will also be concerned that these standards remain under continuous improvement, and again, as such, in harmony with the wholesale product it purchases elsewhere from other infrastructure providers.

207. The effect of these concerns is manifest in the current broadband market, where ISPs face overwhelming disincentives to provide services over fragmented and isolated access networks. If unresolved, this will continue to be the case, and would be a particularly acute issue under the alternative strategy we propose in which a greater degree of competition at the access network level would be a desired outcome.

208. One significant source of these issues lies in the absence of a universally adopted and open set of standards with definitions for both the technical specifications of the network and its wholesale products, and for the operation, administration and maintenance (OAM) systems which create the interface between infrastructure and service providers. Universal adoption of such a set of standards would have an important effect: it would make it easier—in fact actually make it possible—for ISPs to provide services over networks other than just those belonging to the dominant, national providers.

209. The exact nature of these standards is important. Unfortunately, specifying them, even in outline, can only be done sensibly in light of a brief tour through another area of dry, regulatory detail. This is contained in Box 6.

BOX 6

Background to the introduction of open standards

At present, there is no set of standards for the broadband industry which meets all of the following criteria:

(1) is open, and industry-led

(2) is universally adopted

(3) contains a technical specification for:

(a) the physical network itself;

(b) the wholesale products it should provide;

(c) and stipulations regarding the operation, administration and maintenance (OAM) systems interface between infrastructure providers and ISPs.

The significance of a set of standards which succeeds in meeting these criteria is partly set out above in paragraphs 204-208, and is elaborated below.

There is, however, an existing regulatory tool with some of these characteristics. As such, unless its inadequacies are made clear, there may be a temptation for some to suggest that, to an extent, the criteria above have been met. We will briefly, therefore, describe this tool.

Its name is Virtual Unbundled Local Access (VULA). VULA is a requirement placed on Openreach by Ofcom. Ofcom would normally require full physical unbundling of Openreach's local access network, in line with European Commission guidelines. As mentioned in Box 5, the benefit of full physical unbundling—and the reason for the European Commission to require it—is that it enables any ISP to take control of a single, independent connection from the exchange to a premises. As the connection is physically unbundled, the ISP can physically attach their own active equipment to each end, and that, in turn, is the basis on which they are able to tune the service they provide over it. Without full physical unbundling, on the other hand, ISPs cannot physically install their own equipment, and lose, therefore, their ability to tune and differentiate their service from their competitors.

While the European Commission, and by extension Ofcom, normally require full physical unbundling, there are cases where this is accepted as impracticable.[109] One has already been mentioned: PON (see Box 5), which is the FTTP network technology in which Openreach is investing; another arises in Openreach's FTTC network. While in this latter case, full physical unbundling is now technically achievable, in order to achieve it, an ISP would face the prohibitive cost of installing their own complex, active equipment (specifically, a Digital Subscriber Line Access Multiplexer, or DSLAM) in the cabinet; a cabinet serves less than 300 subscribers—there is no commercial justification for each ISP to install this equipment in the cabinet even if there were space available. As a result, in these two cases (where Openreach is investing in PON, and in its FTTC access network), full physical unbundling has been accepted to be impracticable, and Ofcom requires Openreach to offer Virtual Unbundled Local Access (VULA) instead. To reiterate, VULA refers to a requirement placed by Ofcom on Openreach; it is a regulatory tool.

Openreach, in implementing this requirement, has developed a specific wholesale product called Generic Ethernet Access (GEA). In summary, GEA is a wholesale product which ISPs can buy from Openreach; GEA is an implementation of the VULA requirement placed on Openreach by Ofcom because of the impracticability of Openreach offering full physical unbundling over its PON and FTTC local access networks.

Returning to the standards set out at the top of this Box, which would help pull down a major barrier to the ability of ISPs to provide services over networks other than those belonging to the major, large scale providers, it should now be understood that GEA, which implements VULA, does not meet all of the criteria we mention for these standards.

The most important criterion which GEA fails to meet—quite fundamentally—is the first (that the standards be open and industry-led) which, by extension, means that it cannot meet the second (be universally adopted). GEA is a closed, proprietary specification.

210. While it is outside the scope of this report to define precisely what an open set of standards should specify, a set of industry-defined standards does already exist which, in our view, provide a very good model. These are the Active Line Access standards (ALA) fostered by Ofcom and defined by the UK's Network Interoperability Consultative Committee (NICC), a technical forum for the UK communications sector that develops interoperability standards for public communications networks and services in the UK. They are designed to provide a standard interface marrying the needs of Communications Providers and Infrastructure providers in the UK, and meet all of the requirements mentioned in Box 6, barring universal adoption: Openreach, of course, applies its own proprietary specification, Generic Ethernet Access (GEA); see Box 6 for more detail.

211. As is made clear in Box 6, Active Line Access distinguishes itself from Generic Ethernet Access on a number of fronts. As Chi Onwurah, MP for Newcastle upon Tyne Central, former Head of International Technology Strategy, Ofcom, told us:

"The most obvious difference is that GEA is a set of proprietary specifications defined by BT Openreach … while ALA is a set of openly standardised specifications defined by industry consensus through NICC."[110]

212. This distinction is equally as obvious as it is important. As Chris Gallon, Ethernet Working Group Chair at the NICC told us:

"This difference in open standard rather than BT product is important because a small operator cannot implement GEA but they can implement ALA."[111]

213. In other words, the universal adoption of an open set of standards like Active Line Access is absolutely crucial in enabling local access network providers without national scale to compete with large-scale national networks like that of Openreach which envelope many parts of the country, and at present apply their own exclusive, proprietary specification, inaccessible to others.

214. While there are a number of detailed technical distinctions between GEA and ALA, and differences in how they meet the letter and the spirit of Ofcom's VULA regulations, such detail is outside the scope of this inquiry. We do not necessarily suggest that Active Line Access provides a perfect set of standards—indeed any such standards will need the flexibility to adapt to commercial and technological changes. However, ALA seems to provide a good initial standard, and a good model for the open standardisation process which is manifestly required

215. It still remains to conclude exactly what Ofcom should do. In fact, there are a number of mechanisms at Ofcom's disposal to bring about the universal adoption of open, industry-standards like ALA. We will briefly mention three of these: General Conditions of Entitlement, SMP obligations and gentler encouragement and facilitation of the industry.

216. First, under the European Communications Framework (as enacted through UK legislation), Ofcom has the power to impose conditions on all providers of communications networks and services (known as General Conditions of Entitlement). Such conditions can include an obligation to comply with relevant national or international standards, such as General Condition 2 on Standardisation and Specified Interfaces, which facilitates interconnection between providers in order to promote competition. Ofcom can impose such obligations to fulfil its duties to further the interests of citizens and consumers, where Ofcom deems such conditions to be objectively justifiable, not unduly discriminatory, proportionate and transparent in relation to what they are intended to achieve. Were Ofcom to judge it appropriate to do so, Ofcom could impose an obligation on all providers of wholesale local access services to comply with ALA or an analogous set of open, industry-led standards.

217. Second, Ofcom can impose specific conditions on providers, where they are found to have significant market power (SMP) in a relevant market. Although using this mechanism inevitably means any such obligations would not apply universally, we note that this may not actually obstruct universal compliance with them. While this mechanism may not provide Ofcom with the means to impose the adoption of these standards on those without SMP, recall that it is those without a dominant market position who are likely to find it entirely in their interest to adopt such standards anyway, as they help to level the playing field. As such, it is likely only to be necessary for Ofcom to use coercive measures to impose their adoption on those like Openreach, who do have SMP in relevant markets. Using SMP remedies, Ofcom can impose an obligation to adopt these standards where it counts.

218. Finally, Ofcom can use its 'soft power' to convene telecoms stakeholders, facilitate dialogue and encourage the take up of open, industry-led standards more gently.

219. We recommend that Ofcom draw on one of the mechanisms at its disposal to encourage, if not require, the universal adoption of standards like Active Line Access, if not ALA itself, which are open, and industry-led, and contain a technical specification for the physical network itself, the wholesale products it should provide, and stipulations regarding the operation, administration and maintenance (OAM) systems interface between infrastructure providers and ISPs. The universal adoption of such standards would do much to level the playing field between alternative infrastructure providers and would help to stimulate competition at the access network level.

220. The institution of the standards raised in the foregoing discussion would have a significant, mitigating effect on the disincentives which confront ISPs considering providing a service over dispersed and isolated networks. We recognise, however, that even once these standards are in place, ISPs would face a number of other challenges, equally likely to discourage them from providing services over such networks. Chief among these is the commercial viability of offering services over networks with a small number of potential customers. This is an important issue in its own right, but is not a regulatory problem, and does not have a regulatory solution. Accordingly, we discuss it further in Chapter 5, where we describe the realisation of our alternative vision in greater depth.

221. An area in which a change in regulatory approach could have a significant impact on the ability of different infrastructure providers to compete—and which would form an important part of the regulatory backdrop of our alternative proposal—relates to the provision of access to passive infrastructure such as ducts and poles. At present, Ofcom requires that Openreach provide open access to its ducts and poles at a regulated price. This manifests itself in a product Openreach offers to its competitor infrastructure providers, called Physical Infrastructure Access (PIA). It provides prospective infrastructure providers with access, not to any fibre, but to the passive infrastructure through which new fibre can be run. By using these existing passive elements of infrastructure, other providers can lay their own access networks from an exchange, for example, to the community or end-user; once they have done so, they would arrange for a connection to backhaul from the exchange and be able to offer wholesale and retail products over the new access infrastructure which they then own.

222. At face value PIA would appear to provide a powerful stimulus to competition in the access network; it is commonly recognised that 80% of the cost of laying access networks lies in the civil works,[112] which to a significant extent can be avoided altogether by using existing passive infrastructure.

223. PIA, however, comes with a number of important restrictions. In origin, PIA was designed by Ofcom as a remedy to Openreach's SMP in the fixed line, local access market. By design, therefore, it has been constrained so as not to have any impact on other adjacent markets as Ofcom defines and segments them. Accordingly, networks built using PIA are not currently permitted to provide middle mile backhaul, leased line services to businesses or backhaul for mobile and wireless masts; they are restricted to supporting fixed line services in the local access market.[113]

224. The effect of these restrictions, however, appears, in the end, to be a significant limitation on the potential of PIA to provide stimulus to competition in the local access network. In fact, in its recent Business Connectivity Market Review, Ofcom acknowledges the following argument which had been put to it by a number of stakeholders in the foregoing consultation:

"the current usage restrictions of PIA could place bidders of public funds available from Broadband Delivery UK (BDUK) to support investment in superfast broadband infrastructure in less populated parts of the UK at a disadvantage relative to BT who benefits from economies of scope"[114]

225. By 'economies of scope', Ofcom means that an investor laying a fibre network is, in the end, in the market of optical connectivity, and will naturally seek to aggregate demand from a range of the segmented markets that fall under that broad umbrella; these will include domestic, mobile and business markets. BT Group Plc, as a parent company of both Openreach and BT's other divisions, has 'scope' across and is able to draw revenues from all of these markets. Accordingly, the risk of investing in local access networks is less intimidating to BT Group Plc than it is to an infrastructure provider without the scope across these revenue streams. Lifting the restrictions on PIA could help provide other infrastructure providers with that breadth and thereby, in our view, play a significant role in introducing more competition at the level of the access network.

226. Ofcom keeps these considerations under review and states straightforwardly that:

"We remain open to evidence that shows that investment in next-generation access networks could be unlocked if PIA could be used for leased lines services, to help us formulate our policy in relation to such investment."[115]

227. However, at present, Ofcom is not minded to do so for a number of reasons. These include real concerns that:

· "introducing PIA as a remedy into the business connectivity market could undermine the remedies that we have already imposed in that market."

· VULA already provides "an additional option to support competition and investment in NGA networks"

· And finally, "in response to the point about BT not being subject to any restrictions in the use of its ducts and poles, we would note that the purpose of remedies is to appropriately address the identified SMP in the relevant market." [116]

228. In our view, these arguments are understandable, but not persuasive. Under the Communications Act 2003, which implements the EU telecoms package, Ofcom is generally bound to duties both to promote competition and to promote investment, and in particular in so doing, to secure "the availability throughout the UK of a wide range of electronic communications services."[117]

229. On occasion, Ofcom's duties are likely to conflict, and such an issue seems to arise in the case of PIA. Ofcom's competition-based regulatory mechanism has led it to a logically sophisticated segmentation and sub-division of the broadband market, while its duties to promote investment would, on the evidence, appear to suggest that a more open field is what is required.

230. Confronted by this, we fully appreciate Ofcom's cautious stance. It has built up a coherent framework for promoting wholesale competition and serving consumers' needs through its remedies in the various sub-divisions of the broadband market; it is perhaps reluctant, understandably, to undermine these.

231. Such a cautious stance may, we believe, be usefully complemented by a more creative approach alongside it. Its regulatory edifice is not to be preserved for its own sake, and where a conflict emerges between its existing remedies to promote competition and those it might draw on to promote investment, it may be that a root-and-branch reconsideration of the relationship between the two and of the entire framework is required. This is not an easy job, but it is Ofcom's responsibility and it could lead to a clearer, more coherent vision of the total market. As an OECD report concluded: "segmentation of the market may be excessive, resulting in an overly complex regulatory environment."[118]

232. We understand Ofcom's cautious stance with regard to the removal of restrictions on Physical Infrastructure Access. However, we urge Ofcom to give the benefits of doing so full consideration.

233. In our view, the benefits of opening up the restrictions on PIA are likely to be significant, particularly were policy to be re-oriented towards the establishment of open access fibre-optic hubs, as we advocate. Removing the restrictions on PIA may, of course, have knock-on effects for the effectiveness and coherence of other aspects of the overall regulatory edifice. We, therefore, recommend that Ofcom evaluates alternative approaches to the regulation of the broadband market as a whole, in line with EU guidelines.

234. In the light of this discussion, we note that since May 2011, Ofcom has had the power by virtue of Article 12 of the Revised EU Framework Directive, transposed into UK law,[119] to impose regulatory obligations with regard to infrastructure sharing in the broadband market without reference to SMP.

235. While Ofcom is cautious about putting these powers into practice, as they are new and untested and any use would have to be justified by a legally robust proportionality test, imposing regulatory obligations on the basis of Article 12 may have a number of benefits, including but not limited to:

· Allowing Ofcom to impose regulatory obligations on the basis of a broader set of considerations than merely SMP, e.g. public interest and its duty to promote investment to secure the availability of a wide variety of electronic communications services throughout the UK;

· Adapting its regulatory obligations to marry with Government policy, where this is proportionate;

· Removing the imperative to maintain the distinction between fixed and mobile access markets, which is increasingly illusory.

236. In light of the potential benefits, we recommend Ofcom actively considers the possible implications of putting its Revised EU Framework Directive Article 12 powers to use, by undertaking an Impact Assessment of doing so, including an open public consultation. Of course, some of Ofcom's existing remedies in the broadband markets may be rendered ineffective or incoherent by implementing these powers. In consulting on their use, therefore, Ofcom should make positive proposals for how these issues would be overcome.

84 RCBF webpages: http://rdpenetwork.defra.gov.uk/funding-sources/rural-community-broadband-fund Back

85 Q440 Back

86 Miles Mandelson Back

87 Defra, Rural Community Broadband Fund (RCBF): Applicants' Handbook - Expression of Interest Round Two, 10 May 2012 to 6 July 2012. Available online:
http://rdpenetwork.defra.gov.uk/assets/files/New%20Offer/DEF-RDPE-RDPE-Bk-R2_WEB.pdf Back

88 Q 583 Back

89 Vodafone Back

90 Q 509 Back

91 Vodafone Back

92 Q 511 Back

93 DCMS and BIS, Britain's Superfast Broadband Future, December 2010. Available online: http://www.culture.gov.uk/images/publications/10-1320-britains-superfast-broadband-future.pdf Back

94 European Commission, EU Guidelines for the application of state aid rules in relation to the rapid deployment of broadband networks. Available online:
http://ec.europa.eu/competition/consultations/2012_broadband_guidelines/en.pdf Back

95 DCMS, Mobile Infrastructure Project: Industry Stakeholder Engagement, January 2012. Available online: http://www.culture.gov.uk/images/publications/Mobile_Infrastructure_engagement_Rel_1_0.pdf Back