UNCORRECTED TRANSCRIPT OF ORAL EVIDENCE
To be published as HC 1258-i

HOUSE OF COMMONS

ORAL EVIDENCE

TAKEN BEFORE THE

Culture, Media and Sport Committee

Spectrum

Tuesday 21 June 2011

John Cresswell and Julian McGougan

Richard Moat, Nicolas Ott, Kevin Russell and Phil Sheppard

Evidence heard in Public Questions 1 - 96

USE OF THE TRANSCRIPT

1.

This is an uncorrected transcript of evidence taken in public and reported to the House. The transcript has been placed on the internet on the authority of the Committee, and copies have been made available by the Vote Office for the use of Members and others.

2.

Any public use of, or reference to, the contents should make clear that neither witnesses nor Members have had the opportunity to correct the record. The transcript is not yet an approved formal record of these proceedings.

3.

Members who receive this for the purpose of correcting questions addressed by them to witnesses are asked to send corrections to the Committee Assistant.

4.

Prospective witnesses may receive this in preparation for any written or oral evidence they may in due course give to the Committee.

Oral Evidence

Taken before the Culture, Media and Sport Committee

on Tuesday 21 June 2011

Members present:

Mr John Whittingdale (Chair)

Dr Thérèse Coffey

Damian Collins

Philip Davies

Paul Farrelly

Mr Adrian Sanders

Jim Sheridan

Mr Tom Watson

________________

Examination of Witnesses

Witnesses: John Cresswell, Chief Executive Officer, Arqiva, and Julian McGougan, Head of Public Policy and Regulatory Affairs, Arqiva, gave evidence.

Q1 Chair: Good morning, everybody. This is the first session of the Committee’s inquiry into spectrum allocation. As our first panel, I welcome John Cresswell, now CEO of Arqiva, and Julian McGougan, who is head of public policy and regulatory affairs. It was suggested to me by a colleague that we should start by asking everybody to turn off their mobile phone. Can I begin by asking you whether you think that the market is still the best mechanism for determining the allocation of spectrum and whether the auction process is the best way of doing that?

John Cresswell: I think there is always a trade-off between the auction and what it is designed to achieve, which I guess is deriving the most value for the taxpayer and balancing that against public policy objectives through the auction. Obviously an auction is a good mechanism in terms of providing competition and openness and transparency, but the key issue, as I say, is what is the objective of the auction? Is it to raise the most money or is it to balance the most money against the public policy objectives? In our submission we made the point that the current auction as it stands does not necessarily achieve that public policy objective of achieving universal coverage to allow the 800MHz spectrum to be utilised for delivering broadband into rural areas.

Q2 Chair: Can you expand on what you said in your submission: that you felt the whole premise may be becoming outdated as a result of sharing of infrastructure and carrier aggregation? To what extent do you suggest that that means we ought to be looking at a different way of determining this?

Julian McGougan: What we have said in our submission is that we thought Ofcom was broadly right in their conclusion that what they wanted to see post-auction is four operators, each with roughly similar spectrum bags, and that would maintain the competitive tension. We thought that was the right thing to do. However, given that the market is moving towards everincreasing infrastructure sharing and there is a strong likelihood of spectrum sharing later on-in fact, the next development, of the 4G technology, almost encourages spectrum sharing-the likelihood is that, if Ofcom secures the outcome it is aiming for, it will have to come back and review that afterwards. What we would not wish to see, surely, is the idea there is a line in the sand drawn under four operators, each with their own exclusive pots of spectrum, that holds forever and a day, which might have a detrimental impact on consumers and citizens.

Q3 Chair: You think that the whole process may need to be revisited once technology has moved on?

Julian McGougan: For now, I think Ofcom’s premise of four more or less equal spectrum portfolios is the right thing to aim for. If they achieve it, it may be that they will have to review it several years down the line to make sure that still holds true.

Q4 Mr Watson: Isn’t that a rather obvious statement-that you have to review things that change in the future? Are you effectively saying that four operators is one too many?

Julian McGougan: No, not in the slightest. In fact, four operators are fighting each other tooth and nail in this country. This is one of the most competitive environments for cellular networks and I am sure the operators following will tell you that they are making less money here than they make in other countries in which they operate. No, I think four operators is exactly the right thing to try and preserve. What I am saying is that what we are seeing in the market is an increasing move to competition for service layer rather than infrastructure layer. The days are moving away from when operators had their own exclusive infrastructure and their own exclusive pots of spectrum and, in fact, the market is much more about the service layer: who has the most attractive tariffs, who has exclusive dibs for six months on the next new iPhone and that sort of thing.

Q5 Mr Watson: You would say that at the moment Ofcom have struck the right balance between regulatory intervention and market mechanisms?

Julian McGougan: Yes. It is very difficult to get this right. There is no onesizefitsall solution for auctions, and when Ofcom decides it is going to award some spectrum there is a whole bunch of decisions they have to take between deciding that and awarding it, one of which revolves around competition-that is, to what extent do they try to address in the auction design any competition concerns they have, as opposed to leave it for competition law after the auction? We think they probably have it about right.

Q6 Mr Watson: Yes. I think we are realising as a Committee that it is staggeringly complex. What I found interesting about your proposal is that it does seem to capture the future a little bit. You might see these as obvious questions, but I think for our deliberations in the future it might be helpful. How many 3G masts would your shared infrastructure need?

Julian McGougan: That would depend what it is you were trying to do with it, I suppose.

Q7 Mr Watson: The proposal that you have given us to get 99% coverage, how many masts would that need?

Julian McGougan: Well, we are aware that one of the operators, which is Three, has said publicly that just using their existing site portfolio, if they added 800MHz to that without any new sites at all, they could achieve that coverage objective of 99% outdoor range to about 97% indoor reception. So if one operator does not need any new sites, though they may need some more antennas, it suggests to us it is perfectly achievable.

John Cresswell: Can I say a couple of words? The challenge is that if you are looking to have broadband into those rural areas, you have the balance between what the cost might be in building out the mobile network in those areas as opposed to getting fibre to the last 5%, 3%, 2%, 1% of homes. The analysis shows that wireless broadband, while not giving the broadband speeds that fibre can give, is a far more cost-effective solution for reaching those homes. While it is a good question-how much is the cost of it in terms of masts and rollout-you need to compare that to the cost of getting fibre to the home in those other areas to have a true comparison.

Q8 Mr Watson: That is good advice. Can I just keep you on masts for the moment, though? Can you tell me how many 3G masts the four providers currently have?

John Cresswell: No, because we don’t service them all, so we would not necessarily have access to the number of masts that they have to an accurate level.

Q9 Mr Watson: It is quite hard for me to understand. If you don’t know how many 3G masts they have, how do you know that your proposal can give 99% coverage?

Julian McGougan: As we said, Three have said publicly that if they put 800MHz spectrum in all the sites they have access to, which I think is about 16,000 from memory but I am sure the Three witnesses coming afterwards will confirm the exact figure for you-they have said that that would provide about 99%.

Q10 Mr Watson: But, essentially, you are saying your proposal is based on Three’s figures.

Julian McGougan: What we are saying is two things. We are saying, firstly, that achieving 99% coverage is absolutely essential if you want to make sure that that last 10% gets served, because fibre will find it increasingly challenging to reach them. 4G is the only way to do it; 800MHz is the only spectrum that is efficient to do it in and no other spectrum like that is going to be awarded for another decade or so. We are also saying that, because at least one of the existing operators could achieve that without adding any more sites, this is not an onerous thing to ask for. It is achievable now.

Q11 Mr Watson: Perhaps we could follow this up in writing. What I am trying to establish, though, is how many 3G masts are now provided by the four operators and how many more would be required to achieve what you say. Three say they can do it with no more masts. I am just trying to work it out because, remember, we are novices; we are politicians, not regulators and lobbyists who live and breathe this. We are trying to understand this on a deeper level.

John Cresswell: We can help with that, with the help of our customers.

Q12 Mr Watson: I think I know the answer to this, but just for our evidence gathering could you explain to me why low-frequency spectrum-it is a simple question, I know-is much better at getting coverage?

Julian McGougan: Well, quite simply, the lower the frequency, the longer the wavelength. Those of you who listen to Radio 4 long wave will be familiar with that scenario. You can pick it up on the continent very easily, but if you tried to listen to FM Radio 4 you probably can’t. The lower the frequency used, the longer the wave length. The characteristics of long wavelength is that they not only tend to travel very far but they also tend to bend round obstacles and go through walls, which means this spectrum we are talking about, the 800MHz, was specifically chosen for television because it provided the optimal balance between enough bandwidth that you can offer higher-quality TV pictures than we have seen with the old black and white and in colour, but it also carried through walls and hit settop aerials.

Q13 Mr Watson: Would it be too simple a proposal to say that, therefore, based on that technical assessment, four operators having access to the 800MHz spectrum is better for the consumer?

Julian McGougan: Well, it depends what outcome you are trying to achieve. If the key public policy objective is to ensure everybody gains access to broadband, you don’t need four operators to offer that. As long as at least one is doing that, that has ticked that box. From the point of view of competitive tension, one of the things Ofcom is trying to achieve in terms of these minimum spectrum portfolios, as they describe it, is that they have four operators, each of which have some spectrum that is low frequency, below 1GHz. Two operators already do; two don’t.

Q14 Damian Collins: Do you think the Government should set coverage targets or do you think there should be a purely market-based approach? If there was a marketbased approach, what would it look like?

John Cresswell: I am very clear: I think the Government should set a target. The 800MHz spectrum is released through DSO and clearance. It was used to deliver universal coverage for TV. If you want to have a digital Britain or a network nation and broadband for all, this is a once-in-a-generation opportunity to impose coverage obligations on this spectrum to deliver that. Therefore, as we said, 99% outdoor reach coverage will enable virtually every home in the country to be able to access at least 2Mbps, which is enough to watch the BBC iPlayer, for instance. I don’t think the market would provide that, just as television reaching the last 5% would be uneconomic for commercial operators if they were left to choose. I guess that would be the same for a buildout for fibre; it would be uneconomic to do that. Therefore, you have to set down the coverage obligation.

Q15 Damian Collins: What sort of coverage do you think market forces alone would allow for? If there were no targets at all, what do you think we would be looking at?

John Cresswell: Difficult to tell. In 3G, the coverage obligation is about 80%. You can ask the industry colleagues behind me whether they have exceeded that. I do not think the market would necessarily do that because it is uneconomic for them at that level.

Damian Collins: Julian, did you want to come in?

Julian McGougan: I was just going to say the coverage obligation of 3G is an interesting comparator. It was set at 80%, whereas the preceding licences were essentially universal. The Ofcom figures in the latest consultation that has just closed on this very auction, the 800MHz auction, suggest that the 3G reaches 87%-that is, 87% of the population should be able to get at least one 3G signal, although the latest press coverage you may have read suggests that the coverage may not be as perfect as you would like to believe.

The important thing to deduce from that is not only that, left to the market, they will not achieve anything like universality-the commercial imperative runs out long before you hit even 90%-it is also worth drawing the conclusion, if you look at Ofcom’s figures, that those who are not served at all are very unevenly distributed, much like the last 10% who do not have broadband of 2MB; because although 87% is the 3G coverage for the UK, for Northern Ireland it is only 40%. This is one of the reasons why we were suggesting both to Ofcom and in our response to this Committee that, whatever the coverage obligation, it should be set by nation rather than by the UK so the benefits are more widely distributed. The French, in their proposals for 800MHz, have not only set a very challenging coverage target of 99.6%, but they have suggested that it needs to be measured by department.

Q16 Damian Collins: Do you think that setting a high coverage target, be it 95% or 98%, which a number of MPs in Parliament spoke in favour of, or 99% as you said, is a market deterrent? Do you think the companies can deliver that? They may choose, left to their own devices not to, but would their business models allow them to hit that target?

John Cresswell: I suggest that is a question that they can better answer. However, I think it is what I said earlier: there is a balance if you want to maximise the tax take, because the cost of reaching and operating to an extended level is bound to be more expensive than if you didn’t have to do that. In terms of the price that anyone is willing to pay in the auction, there is a slight trade-off for the additional coverage, but I think if you just stand back and look at what BDUK is trying to do and the amount of money it has available to ensure broadband for all, if you don’t have a mixed economy between wireless broadband and fibre, I don’t believe there is enough money in that pot to get fibre to every home. Therefore, the two need to be aligned in some way.

Q17 Damian Collins: Where would you strike the balance? Because it is all money, however you cut it. The Government can set high coverage targets and accept lower bids for it, or it could get as much money as it can and seek to top up the coverage with pockets of money to fund projects, or even maybe local government may decide to support the rollout of broadband to support business development in its areas. From what you have said, it sounds like the best thing to do is set the target and then let the market find a way to hit it.

John Cresswell: I think that is exactly right. Unless you set the coverage target, then the procurement within BDUK, which doesn’t have any coverage obligations in its targets yet either, you would never get there. That is why I say this is the once-in-a-generation opportunity. If you set the target, then you have the opportunity to find funding from wherever. Also, in terms of having the whole thing aligned-all the Government policies aligned-you need to set coverage obligations for the procurement in BDUK; otherwise they could spend the money without reaching the coverage obligations as well.

Q18 Damian Collins: Have you calculated what the additional cost associated with the 99% coverage obligation would be for the industry? A supplementary to that: do you think the Government should ask for bids based on maybe one or two different coverage targets, to assess what the likely cost might be?

John Cresswell: I could not quite hear the first part of the question.

Damian Collins: Sorry. Have you calculated what you think the additional cost associated with the 99% coverage would be for the industry?

John Cresswell: I think it is in the region of £200 million to £230 million to extend the networks in that way. What was the second question?

Q19 Damian Collins: The second was, in terms of the auctioning process, do you think the Government should consider setting, say, two coverage targets and asking people to submit their bids based on, say, a 95% and a 99% coverage?

John Cresswell: I wouldn’t, because I think then you have a level of complexity in the auction and it depends how you are going to decide between them, because then you have a very clear trade-off between coverage and money. I believe the coverage obligation should be there.

Julian McGougan: If I could just make two points on that one, the first is that we shouldn’t look at the costs without also looking at the benefits. There was a substantial OECD-World Bank study in 2009 looking at the benefits to the economy of driving up broadband penetration, which said that for every 10% increase in broadband penetration you get more than one percentage point extra GDP growth. I am sure that figure will dwarf whatever the costs are.

The second point I would make is that it may not be obvious even when the auction is over what the costs of that coverage obligation have been because the six different pots of 800MHz spectrum that the bidders are going to be bidding for are not homogeneous. There are various factors in play affecting how much they will bid for different ones. If you bid for the spectrum at the bottom against prevailing Freeview use, there will be interference from that use into Freeview, which will have to be mitigated. There will be a cost to that.

The release-when all the spectrum is available for UKwide release-varies depending on which bit of spectrum you buy. At the top end you have the risk of continuing unlicensed radio microphone use and the potential interference into the low-power devices immediately at the top, which include personal health care alarms for frail people-the alarm gets triggered if they fall over-which Ofcom is looking at at the moment. All these factors mean that this is not homogeneous spectrum, so even when the auction is over, you may not know what the cost of asking for a 99% coverage obligation was, but you do know what the benefits of plugging that last 10% are going to be.

Q20 Mr Sanders: Given that yours is a company providing media infrastructure, is that one of the reasons why you would be so keen to want 99% coverage?

John Cresswell: If it was only that simple that all the business would come our way, then yes, but we work in a competitive environment. As we explained earlier in terms of the number of 3G masts, we are a provider in the marketplace. There are other providers as well. I think having broadband for all will provide greater business opportunities for us and other businesses as well, but the thing that we have learnt from DSO-digital switchover-which Arqiva is doing and then we are doing the clearance for clearing these channels, is that if you do not plan ahead, then there are, as we know, thousands of sites out there with masts and antennae on them, and the turnaround time is not very quick. The vision and ambition that the Government set out for the longer term, this is an opportunity to seize that.

Q21 Mr Sanders: I had never come across the phrase "backhaul infrastructure" until this inquiry. We learn something all the time. Do mobile operators have effective enough backhaul infrastructure to cope with a 99% coverage obligation?

Julian McGougan: It is a legitimate concern and we have heard it before. There is no doubt that if you are rolling out into rural areas, the backhaul is an issue that has to be addressed. Now, in some of those areas fibre will reach either because fibre can get there now or because BDUK is looking at how, as part of its procurement, fibre can be driven out as far as possible to the communities. Where fibre does not reach the mast, there is microwave as an alternative, which is a longstanding use of that. You use microwave wireless, which can offer very high bandwidth backhaul up to 100Mbps, which is ample for a small rural site, and it takes that backhaul, takes that traffic, to the nearest fibre point. For example, when we operated the 800MHz 4G trial in Pembrokeshire at the end of last year, we ran it from one of our broadcast sites, which also has virtually all the mobile operators on already for the non-4G. It did not have fibre into it at the time, but we could run fibre into it if we wanted it to. It is a mix of fibre and microwave that will deliver this.

John Cresswell: That is part of the issue in some of the rural sites. They already do have power for backhaul to them because they are used for broadcasting and other benefits. You are right; it is an issue that will need to be addressed with some investment, but it is solvable.

Q22 Dr Coffey: I notice in your submission you mention just 95% could exclude Suffolk, Northern Ireland and Cumbria, so obviously I’m referring to that. I am sure many rural MPs are pushing hard for a higher coverage obligation. I am just trying to understand a little bit more about what you said there. I visited Orford Ness the other day. There is an idea that if they could get fibre out there, they could do a lot of relay on this. If BDUK’s effective role is to create a digital pump in a village, which is what Jeremy Hunt says, what is stopping the mobile operators basically piggybacking on that and creating more masts in a cheap way? Why is it they only ever want to use their own infrastructure, it seems? We are just trying to understand.

John Cresswell: Before we answer the second part of that question, let’s start with the digital pump. I think the digital pump works very effectively with wireless broadband because you still have the pump in the middle; you do not have connectivity to every home and the ranges for wireless broadband would be a very complementary technology to sit alongside that.

In terms of what you were saying earlier in terms of the procurement, I think one of the interesting things about how the procurement is structured is in some counties, Cumbria being an obvious one, it is not the final 5%. It is probably the final 50% that are affected in rural areas, which is why, if you do not have coverage obligations within their procurement, you may find that some areas have very low penetration levels. The second part of the question about the masts I will allow you to have a go at, Julian.

Julian McGougan: I wish I knew the answer. If you want to know why the mobile phone companies prefer their own infrastructure, which they are decreasingly doing, then perhaps it is a question to ask the two operators following us. One point I would make about the digital pump is that, although that was one of the centre points of the Coalition’s broadband strategy that was unveiled at the end of last year, because BDUK is running a whole series of local procurements bubbling upwards, there are going to be 30 to 40 different procuring authorities-some the size of Wales, some the size of perhaps counties-who will decide for themselves what they want broadband to look like for their consumers.

Q23 Dr Coffey: But BDUK is changing that strategy, isn’t it? Or have they not communicated that publicly?

Julian McGougan: It is a slightly fluid situation. Every time we have spoken to BDUK they have made it quite clear to us that they cannot insist on universality. They cannot insist that each procurement delivers 100%. It is for the local procuring authority to decide exactly what it is that they want to buy. Some areas may get digital pumps, which may then have a very large rural dispersed population to serve. Some may not get digital pumps at all.

Q24 Dr Coffey: From what you have just suggested, BDUK may be setting themselves up to fail in delivering the Government’s policy?

Julian McGougan: It is more difficult delivering a Government policy if you are relying on local counties and devolved assemblies to decide what they want to do in their areas. It is one of those localism debates we are familiar with from the NHS and much else, I’m afraid.

John Cresswell: Yes, I think you need to have the two things aligned and certainly as part of procurement. Whether we start at the coverage obligations, certainly standards in the procurement are going to be absolutely essential to ensure that the operators and the network are interoperable.

Q25 Dr Coffey: Stepping back, I appreciate you can’t talk on behalf of mobile operators, but what would be the most cost-effective way to get 99% coverage? Your masts?

John Cresswell: No, not our masts necessarily. I think you need a mixed economy. Principally it will be fibre with wireless broadband, and then the last 1% being able to be served by satellite. That is what happens in most other European countries that are going along the same path. I think you need that mixed economy.

Q26 Dr Coffey: In terms of existing fibre network, is there any reason why that couldn’t be used to, again, be piggybacked on? In your view, are there technical limitations?

John Cresswell: Obviously, fibre has significant advantages in terms of speed. As the Minister said at a dinner I was at, he was aiming for 90% at 10Mbps and 10% at 2Mbps as a starting base. Well, wireless broadband can get you to that. I think essentially it is going to be fibre, but the cost of rolling out fibre to the last 3% or 5% is extremely expensive. Depending on whether you want to deliver high-speed broadband as we would all understand it to everyone, or for the last 5%, 3%, 2%, 1% prepare to, with today’s technology, give them 2Mbps in terms of band speed, then that is just about price-how much you would like to pay for the cost of getting fibre to the home, because it is very expensive in those areas just to dig up the roads to people’s houses.

Dr Coffey: Of course. Thanks.

Q27 Chair: Can I turn to the PMSE sector, with which you have some involvement as a company supplying technical equipment? We have heard considerable concerns have been expressed that the proposals by Ofcom don’t go far enough to safeguard the future of the sector. Do you share those concerns?

Julian McGougan: I think it is unfortunate that they have to suffer this disruption, although we have to bear in mind that television is also being ejected from the same space-there was disruption to Freeview viewers in some places as well-but the PMSE lobby is quite vociferous and they quite rightly say, "Why should we be moved?" I think they lost that argument because of the high-value nature of mobile broadband, but Ofcom, working with Arqiva and the industry, have done their best to put in place something that mitigates that. There is a scheme that pays, I believe, about 55% of the cost of new equipment.

There is a transition process in place where new spectrum was found-channel 38 to be technical-which will provide more spectrum more flexibly than the old channel 69 they are being moved from. Both the old and the new spectrum are available simultaneously and they are available under the same licence you currently have. You don’t need to get a new licence when you buy new equipment. Everything is being done to make that transition as smooth as possible, but yes, it is a bit disruptive and I’m afraid there is not much that we can do about it. There are thousands of radio microphone users to be communicated with and some of them would probably rather just stay where they were. Unfortunately, that is not an option.

Q28 Chair: I think, in terms of their approach to us, to some extent they accept that they are not able to stay where they are now and they are going to have to move. One of their concerns is that where they are being moved to may not be a permanent home and that, in due course, they may come under pressure there as well.

Julian McGougan: Well, I am not sure how permanent any home is, to be honest, in terms of spectrum. After all, the Crown retains the freehold of all spectrum, so to speak, so anybody who has a spectrum licence does not know it is going to be renewed. However, they have been found this new home with the expectation that it will last a considerable amount of time and probably as long as the equipment that they are buying. All the tests have been done to show that there is no interference in that that would affect PMSE use. It should provide more spectrum and more flexible use of PMSE, but, yes, there is a bit of disruption for them, I am afraid.

Q29 Chair: They have also expressed considerable concern about the possible use of white space devices and the fact that this may interfere with PMSE events. Is that a legitimate concern?

Julian McGougan: It is a concern shared by Arqiva, although we are more concerned that it will interfere with Freeview reception than PMSE because that is a wider use than radio microphones. But yes, it is a concern we share and Ofcom share the same concern as well. There is a trial due to start in Cambridge at the end of this month on white space devices to prove the very point that those who are in favour of white space devices have been saying for some time: they can operate safely around Freeview use and around PMSE use. For PMSE, where you don’t know exactly where it is-unlike television, which tends not to move very much in its broadcast-that will require accessing a database, which is to say that every time a white space device wants to power up it will access a database and say, "Tell me what I can use within this location", and it will only then use what the database has told it is free. Thus licensed PMSE use should be protected.

Q30 Chair: Ofcom have suggested if there is interference then they will turn it off within an hour. If you are Bon Jovi about to play the Milton Keynes Bowl, an hour is rather a long time while you wait for Ofcom to get round to clearing your frequencies.

Julian McGougan: Well, this is one of the things that are being debated hotly within the industry and Ofcom. It is a live issue.

Q31 Chair: Does Arqiva have a view on it?

Julian McGougan: Well, we would not want Bon Jovi to be switched off either.

Chair: Good, nor would I. I think that is all we have. Thank you.

John Cresswell: Thanks very much.

Examination of Witnesses

Witnesses: Richard Moat, Deputy Chief Executive Officer, Everything Everywhere, Nicolas Ott, Vice President of Strategy, Planning and Regulatory, Everything Everywhere, Kevin Russell, Chief Executive Officer, Three, and Phil Sheppard, Director of Network Strategy, Three, gave evidence.

Q32 Chair: Good morning, gentlemen. I welcome for the second part of this morning’s session Richard Moat, the Deputy CEO of Everything Everywhere, and Nicolas Ott, the Vice President of Strategy, Planning and Regulatory Affairs, and from Three, Kevin Russell, the CEO, and Phil Sheppard, Director of Network Strategy.

Q33 Mr Watson: First, I thank the executives for coming, taking time out of your busy schedule. Richard, could I just ask you why couldn’t your Chief Executive make it this morning?

Richard Moat: This meeting was arranged at relatively short notice and I was available but he wasn’t. I am the person within the organisation who is most directly responsible for this area.

Mr Watson: What does he have on this morning?

Richard Moat: I don’t know exactly but-

Q34 Mr Watson: Okay. He is a busy guy. My first question, perhaps for Kevin and Richard, is: is it fair to say that complaints about the effects of liberalising the 900MHz licences are special pleading?

Richard Moat: I think there is an imbalance in the UK with respect to the allocation of sub-1GHz spectrum, which started when that was awarded entirely to Vodafone and O2 back in the early 1980s. It has been perpetuated by the fact that that spectrum was liberalised without charge at the beginning of this year, which means that it can now be used for 3G, and if provisions are not put in place to allow Three and Everything Everywhere to get a reasonable portion of the 800 spectrum, that problem will perpetuate itself into the 4G environment in the future.

Kevin Russell: There is a well documented and well understood significant competitive advantage to low-frequency spectrum. It transmits about three times as far as highfrequency spectrum. That is an advantage that O2 and Vodafone have enjoyed in the marketplace for a long time. It has now been carried over directly into 3G with the liberalisation. It is a distortion that has been largely addressed in other European countries, so in every other European country that Three operates in, there has been an allocation or a reauction of 900MHz spectrum. There is no other country that we operate in, but UK has not addressed that distortion.

Q35 Mr Watson: I am sure I speak for most Committee members when I say we are still getting used to the technical terms about types of spectrum, but it does strike me that there are certain portions of the market where you are advantaged and disadvantaged depending on your position. For example, I read in the Financial Times this morning that for the first time Everything Everywhere will be taking advantage of Ofcom’s new spectrum trading laws and that you will be selling 25% of your 1,800MHz spectrum. Is that right?

Richard Moat: Yes, that is right. We had to relinquish 25% of our 1,800 holding as part of the remedies with the EU and with the UK competition authorities, which allowed the merger to go forwards in the first quarter of last year. Now that the spectrum trading regulations have been put in place, we can effect that.

Q36 Mr Watson: That is valued at £450 million?

Richard Moat: I don’t know what the value of it is at the moment. I think that can only be determined when the rules for the spectrum auction are finalised, because that will have a massive impact on the value of any block of spectrum, 1,800 included.

Q37 Mr Watson: Would you say that the FT report is more or less accurate-that £450 million is probably the figure that you could realise for that?

Richard Moat: I would not want to speculate on that publicly. It is highly commercially sensitive.

Q38 Mr Watson: But surely in your own forecasting you have a figure in your own head. You might not want to share it with a parliamentary Committee, but your shareholders presumably-

Richard Moat: I don’t have a figure in my head at all. As I have said, there is such a wide range of outcomes, depending upon what rules are put in place specifically with respect to the large amount of spectrum that is for sale in the spectrum auction, which would determine the value of a single block of 15MHz that could be sold separately from that auction.

Q39 Mr Watson: Let me just get this straight. You are telling me that nowhere in your company there is a single person who has estimated what the value of this spectrum might be?

Richard Moat: Do you want to comment on that, Nicolas?

Nicolas Ott: Yes, I can comment on that. One key driver of being in a difficult situation for assessing the value of that block of 1,800 is that Ofcom has not yet formally stated the spectrum fee they will want to charge, whoever is going to buy the spectrum once we have sold it. The current proposal of Ofcom seems to attach to that block of 1,800 a very high spectrum fee; therefore whoever buys it will have to pay, endlessly, a very high fee yearly back to Ofcom. That is one of the fundamental points of discussion with Ofcom, which is to tell them, "Until the time you finalise your decision about that yearly spectrum fee, no one is capable of assessing the value of that spectrum."

Q40 Mr Watson: To repeat my question, is there anybody in your company that has estimated the value, or are you saying that, as Deputy Chief Executive, all you get is reports saying, "We have this very valuable piece of spectrum but we don’t know what it is worth."?

Richard Moat: I am saying there is a very wide range of possible outcomes and I wouldn’t want to speculate on what that is.

Q41 Mr Watson: Well, could you just let me know what you think the range might be?

Richard Moat: Let’s look at the prices that are being quoted for 800 and 2.6 in the context of the current spectrum auction rules. If you are looking at 800 reserved spectrum, then it is £200 million per 5MHz.

Q42 Mr Watson: So the FT might not be far off, then?

Richard Moat: But 1,800 has much worse propagation characteristics and, therefore, can provide less coverage than 800, so it is not going to be worth the same amount. It is going to be less than that.

Nicolas Ott: If you take the 2.6 reserve price-

Q43 Mr Watson: I have the answer on that. Could I just move it on then? You received this in 1993, is that right? Can I ask what you paid for it?

Richard Moat: It was in 1991, I think.

Mr Watson: 1991, was it?

Richard Moat: Yes.

Mr Watson: Can I ask what you paid for it?

Nicolas Ott: We are paying yearly the spectrum fee to Ofcom.

Q44 Mr Watson: No, when it was allocated to you, what did you pay for it?

Nicolas Ott: It was paid against a yearly fee. In the 900MHz only 1,800MHz have been allocated to the operators against a yearly fee that is paid by the operators to Ofcom.

Q45 Mr Watson: My understanding is you were basically given this chunk of spectrum. Do I have that wrong?

Richard Moat: Yes. The original companies that were in existence at that time were given that chunk of spectrum.

Q46 Mr Watson: Yes, so the component parts of Everything Everywhere were given a chunk of spectrum by the Government and did not pay anything for it. That would not be a misrepresentation?

Richard Moat: Yes, the same is true of Vodafone and O2-

Mr Watson: I understand that. They are with us next week.

Nicolas Ott: But we pay £33 million per year.

Q47 Mr Watson: Let me just say I am going to go off the £450 million figure that the FT have speculated on this morning. Can I ask what you pay in licence fees for that chunk of spectrum?

Nicolas Ott: For the 800MHz?

Mr Watson: Yes.

Nicolas Ott: We paid £33 million per year.

Q48 Mr Watson: £33 million. If I go off £450 million-I did do the maths before I came in-that means that, if I go from 1993, you have paid £104 million since 1993 in licence fees. Is that right?

Nicolas Ott: Well-

Mr Watson: Thirteen times eight.

Nicolas Ott: Thirteen times eight would be probably there.

Q49 Mr Watson: Yes, okay. If the FT are right, you are about to get a third of a billion pounds dividend out of this. Is that right? With the FT figure of £450 million, you are going to get £350 million surplus for a piece of spectrum that you were given for free?

Nicolas Ott: We also have invested on that spectrum. We have invested in network; we have invested in capex.

Q50 Mr Watson: Sure. Would I just be accurate in that? I do not want you to defend that position just yet.

Richard Moat: No, it is £33 million per annum for 20 years.

Q51 Mr Watson: You are paying £33 million per annum, yes, but you are only giving a quarter of this spectrum away.

Richard Moat: Are you talking about only the quarter that is being given away?

Q52 Mr Watson: 25% of 33 times 13, 25% of 33, broadly £8 million times 13-when I did it on the calculator this morning it made £104 million, but that was-

Richard Moat: Times 20. It was given in 1991.

Q53 Mr Watson: 1991? So it is even more than that? What would that be? That would be £8 million times 20, that would be-

Richard Moat: £160 million.

Q54 Mr Watson: £160 million. You have paid £160 million and you will be making £290 million surplus?

Nicolas Ott: But if you look at the NPV of that spectrum you have to consider that we have been investing on our network and that investment was-

Q55 Mr Watson: Can you just answer the question? It is broadly a third of £1 billion, yes?

Richard Moat: On that hypothesis, yes, but there is 70MHz of 2.6 spectrum and 30MHz of 800 for sale.

Q56 Mr Watson: Let me just ask you, do you think it is right that the Government should give away a public asset to a private company on which they then realise a third of £1 billion pounds surplus?

Richard Moat: These licences were awarded between 20 and 30 years ago, so it is difficult to comment on what public policy was then and what this decision means now.

Q57 Mr Watson: No, I do understand that, but do you think that your customers, who for the next three years will be facing reducing living standards, would think it was fair for the Government to allow you a surplus of a third of a billion pounds to be made on spectrum that is owned by them?

Richard Moat: Well, it is currently owned by Everything Everywhere.

Q58 Mr Watson: That is because the Government gave it you.

Richard Moat: Yes.

Q59 Mr Watson: Do you not think they should get a share of some of that surplus? Do you understand the point I am making?

Nicolas Ott: We understand the point you are making. That decision has been taken by the public authorities when they accepted the merger. So it is not-

Q60 Mr Watson: Is it the right decision?

Richard Moat: It is not our decision to sell it in the first place.

Q61 Mr Watson: But do you think it is right that you should get a benefit from that spectrum that was given you for free? Just say yes or no.

Richard Moat: The spectrum has a value and the fact that it was awarded for free 20 years ago I don’t think is relevant to today’s circumstances when we have invested £58 billion in this country since the start of the business. That represents 7% of FDI over the last 10 years.

Q62 Mr Watson: Some of your shareholders are the German and French Governments, is that right?

Richard Moat: Yes.

Nicolas Ott: Well, nominal.

Q63 Mr Watson: They will be getting a share?

Nicolas Ott: It is not the French and the German Government. The French and German Governments have a shareholding of DT and FT, but they are not the prime shareholder.

Q64 Mr Watson: I see. The German and French Governments have a share in Deutsche Telecom and France Telecom that own you?

Nicolas Ott: Yes.

Q65 Mr Watson: So essentially they have a share?

Nicolas Ott: But they do not control the companies.

Q66 Mr Watson: They are going to get a share of a public asset that was given freely to your company and the British taxpayer is not going to see any of that?

Nicolas Ott: The key issue for the company, I think, is not to take that disposal and just to cash it. The point is that we are here in the sense that we need to invest in a new series of network. We need to invest in 4G, we need to invest in rural coverage and we need to invest in the licences that Ofcom will be organising very soon now, hopefully. The priority for Everything Everywhere is to be in a position of investing in the UK more than anything else, and that is what matters. These proceeds, should we indeed keep them, will be indeed reinvested into the UK market by acquiring more spectrum, by rolling out 4G network and by trying to do as well as we can in the rural areas. That is what Everything Everywhere is about and that is the vision and the mission of the company. That is why we have named the company Everything Everywhere. We want to offer the best coverage possible in all the country. That is, if you want, the industry or project of the company and I think that is what matters quite significantly for the country, hopefully. They going to benefit from one of the four players in a position that allows sustainable competition, in a position that allows them to invest, and in a position that will allow them to do as we are doing in Cornwall now, offering very deep rural coverage, in other regions of the country. It is definitely where we are focusing more time and effort than on anything else, which is how can we play a fair role in that new world of mobile internet.

Q67 Damian Collins: The profit you might make on the sale that we have just been discussing, are you saying that will be ring-fenced for investment back in the UK network?

Richard Moat: I think that it is not necessarily ring-fenced but, practically speaking, we need more spectrum. We are being forced to sell 25% of our 1,800 spectrum holding as part of the remedies to allow the joint venture to move forwards. That being so, we need to acquire more spectrum to replace it and more on top of that because we have 27 million customers in the UK, 30 million if you include the mobile virtual network operators that we host, which is by far the largest number of customers, the largest customer base, of any of the four operators in the UK. If you look at our spectrum holdings on a per customer basis, we have one of the lowest spectrum holdings because of the large number of customers that we have and we have to maintain that spectrum per customer. We have to reinvest not only in replacement spectrum but in more spectrum on top of that to cater for the huge growth in data that is going to occur over the next 10 years.

Q68 Damian Collins: Okay, so that was no, but given the figures we heard at the previous session, what might strike some people from what Mr Watson has been asking about and the numbers he was talking about is that the profit you are making on the spectrum sale would potentially pay for the infrastructure costs of extending coverage to 98% or 99% of the country. People might question whether the money would be better spent doing that than being returned to your shareholders.

Nicolas Ott: If you go back to effectively the values that were quoted and if you look at the reserve price that Ofcom is currently proposing for the 800MHz and for the 2.6GHz, quite clearly you can see that, with these proceeds, we can pay for only a fraction of the spectrum that we need, as Richard said, for replacing what we sell and for acquiring the spectrum we need to cope with the mobile data increase. It does mean that, only on the pure spectrum scope, it is quite clear that the net outcome will be a cash outflow from the company. That is obvious. We would almost like to be in the opposite situation where we would think the spectrum will not be very expensive, so it would be easier.

Now, when you look at the reserve price proposed by Ofcom and the auction outcome in the other western European countries and you try to compare things likeforlike, quite clearly the netnet for Everything Everywhere will be a very significant cash outflow for the spectrum itself. Then comes, effectively, the rollout of the 800 network that we do not have, so we need to build it from scratch because we do not even have a 900 network. We need to create something from scratch. Same thing for the 2.6. If, hopefully, we can acquire also some 2.6GHz-as we say, it is a further capex investment-at the end of the day the netnet is going to be in the years to come definitely a very significant cash outflow from the company either to the Treasury for the spectrum, or to the companies we are using in the UK for rolling out and maintaining our networks.

Q69 Damian Collins: I accept that, but you have a model where you know you are making cash investment in the networks and you know you are going to recover that from your customers over a long period of time and that is protected for you. Where we have these sorts of pots of extra cash that maybe you had not budgeted for, to me it suggests there is enough money in the system here to fund some of these extra things we might want in terms of certainly expanding coverage.

Richard Moat: We have not said that we are opposed to high coverage obligations. If you look at the coverage that we have on our existing 2G network, it is 99%. On our 3G network, despite the fact that the original coverage obligation was only 80%, recently going to be increased to 90%, our coverage once again is 99% of the population. Despite what was said previously, market forces push you in that direction. It is very difficult to get to 100%, particularly with the spectrum we have and its propagation characteristics, how much coverage it can provide, but we firmly believe that we have a role to play in terms of reaching the maximum possible amount of population in the UK. The trial that Nicolas referred to in LTE at 800, which we are running in Cornwall in association with BT, is a big push pilot to make sure that that will work in the future.

Q70 Dr Coffey: Can I just clarify in terms of the sale of the 25% spectrum that was forced upon you-that was income you may have expected in the past but the European Commission said otherwise-will it be based on market forces, what other people are prepared to pay, or is there an element of Ofcom introducing almost a kind of a set price for you?

Richard Moat: I think it will be heavily influenced by the spectrum rules that are eventually agreed, but it will be an open market transaction.

Q71 Dr Coffey: Then any profits you gain from that will, I assume, be included in your accounts and your corporate tax base for paying corporates?

Richard Moat: Yes, but to reiterate what Nicolas said, there is almost bound to be a net cash outflow in terms of replacing and adding to that spectrum, before we even start talking about the cost of investment in the infrastructure equipment.

Q72 Dr Coffey: But there is no reason why, given the structure of how the spectrum is currently held, the money would not be included as part of your financial accounts within the UK?

Richard Moat: No.

Dr Coffey: I have finished my questioning on that.

Q73 Chair: Can I just clarify? Both of you have essentially suggested that you are disadvantaged by the fact that the original two operators have 800MHz and you don’t and the lower the bandwidth the better the coverage, yet in the case of Everything Everywhere you are now the biggest operator in the UK and you have more base stations than everybody else. Given that you have all those existing base stations, surely the spectrum you already have access to is perfectly as advantageous as the lower bandwidth?

Richard Moat: We want to reduce the overall number of base stations that we have, because that makes economic sense. We had two completely separate networks and we only need one moving forwards into the future. Also, obviously, if we reduce the number of base stations, then that is good for the environment as well. There would be a positive environmental impact from doing that.

Chair: It is altruistic reasons that cause you to-

Richard Moat: Well, partly. Partly it is sound commercial reasons and partly, yes, it is environmental reasons, but it is a combination of the two.

Q74 Chair: With the spectrum you have at the moment, while you nevertheless do have to maintain all the base stations, you can offer a perfectly viable service?

Richard Moat: Ofcom have previously said themselves that if you are using 800MHz spectrum, which we are, you need, roughly speaking, three times the number of base stations than you do if you are using 900 or 800 as we are now contemplating in the context of this auction. That being so, your cost base is going to be much higher, so you are at a significant commercial disadvantage if you have 900 spectrum rather than 1,800.

Nicolas Ott: If I may, just a quick point. Ofcom in their consultation state very clearly that they consider that roughly 10% of the calls-whether it is voice, or it is mobile internet-can be served only by 900 or 800 because these calls are being made by people or received by people who are what they call "deep indoor". Here, because of the lower physics that the representative of Arqiva was explaining before, the 900 and the 800, because they go further deeper into a building, are the only spectrum capable of making that call or receiving that call. That is why, effectively, the ownership or the access to 900 or 800 is absolutely critical for servicing that part of the deep indoor coverage. Then, going back to the rural coverage and going back to BDUK’s ambition of delivering 2Mbps in mobile data, the work we are currently doing with BT in Cornwall evidences very clearly, on a scan of the industry standard, that if you don’t access 2x10 minimum of sub-1GHz there is no way you can provide that coverage in deeper rural areas. That is also why if we don’t have it, despite the number of sites if you want, we couldn’t do it at all.

The last point: on the surface we seem to have more spectrum but when you divide that spectrum by the number of customers we have, including VMOs such as Virgin Mobile, then you have roughly the same level of megahertz per customer. So, yes, we have a bit more spectrum but, yes, we also have a few more customers. All in all there is a balance without over-engineering.

Q75 Chair: Can I put a similar question to Three? You don’t even have 1,800MHz, you have even higher, and yet, despite the fact that you are deprived of the lower bandwidth, which we are told is the most effective, you have 50% of all mobile data traffic, so it doesn’t seem to have disadvantaged you too much either.

Kevin Russell: No, and I think that is fair for the last eight years. All four operators in the marketplace have operated in the same frequency, the same spectrum, which is 2100. So for the last eight years we have all competed based on exactly the same spectrum. The 900 being liberalised only happened in January. The impact of that in the marketplace, in terms of how indoor coverage will fundamentally improve for O2 and Vodafone, will take effect over the next 18 to 24 months; that is the impact going forward.

The other point I think it is important to clarify is that when we talk coverage, it is important for us to be very articulate and very clear around indoor coverage and outdoor coverage. In the last session, that talk of 80% for the previous licences, that is population coverage, outdoor coverage. That could represent only 50% indoor coverage. If I look at where we, Three, are today, our population coverage, our outdoor coverage, could be perceived to be 97% to 99% in that range. Our indoor coverage, however, is only 79% and the challenge with the high-frequency spectrum is, while you might cover the population, you struggle to penetrate the buildings as well as the low-frequency spectrum can. For ourselves and Everything Everywhere, it is critical we acquire more frequency spectrum to be able to take indoor coverage up to levels in the high 90s to be able to compete on a sustainable basis going forward. That is the key point around low-frequency spectrum.

Q76 Chair: What is the indoor coverage of your competitors?

Kevin Russell: Based on pre-900, Vodafone and O2, in my strong view, have fundamentally under-invested in the network and they have, I think, probably close to 8,000 sites; Everything Everywhere and ourselves have 12,500-plus sites. Our indoor coverage would be 79%. I would expect their indoor coverage to be somewhere probably in the 60s-that would be my guess, but that is an estimate. I don’t know whether you have a different view?

Richard Moat: You are talking about 3G there.

Kevin Russell: Talking 3G. This is mobile broadband.

Q77 Chair: Yes, I understand. So you see the competitive disadvantage of not having access to the bandwidth as emerging over the next couple of years, not current today?

Kevin Russell: It is emerging today but it will take traction over the next 24 months. O2 very rapidly redeployed their low-frequency spectrum in London and they are rolling out very rapidly across the country. Vodafone has not redeployed yet. It will emerge over the next, I would say, 18 months.

Q78 Dr Coffey: If your fears about O2 and Vodafone cornering the market at the next auction are realised, what could it mean for each of your operations? Perhaps Three could go first.

Kevin Russell: For us it is simple. We have approximately 10% of the marketplace today. We have invested in a nationwide infrastructure and we have ambitions and necessities to double the size of our business to get ourselves up towards a 20% market share. We must do that with leadership in mobile broadband and data. The fundamental requirement of customers is to have a strong network. You are not going to move from your existing operator to somewhere like Three, even though you might get a better value proposition, unless the network is as good or better than your competitor, so our ability to take customers from O2 and Vodafone and Everything Everywhere to build market share has a fundamental premise that we are able to have at least a comparable network in terms of inbuilding coverage and overall capacity and speed to support customers. In terms, we will not be able to grow as rapidly as we need to unless we can match our competitors on spectrum.

Q79 Dr Coffey: Would it have any impact for Everything Everywhere?

Richard Moat: Yes, I agree. I think exactly the same. There would be an impact on revenues because customers would naturally migrate to what they consider to be the best network and, in this context, if two players have a significant advantage over, in particular, as Kevin was saying, in-building coverage over the long term then that is going to have a massive impact, particularly for the use of mobile broadband. We want to remedy that by getting part of this spectrum ourselves to be able to use it and to have a more balanced playing field. As I said earlier, in addition to that kind of top-line impact there is also a massive impact on costs, because if you are trying to achieve the same end results using higher frequency spectrums-2.6 or whatever it might be-then that is going to be significantly more expensive. It is a kind of double whammy: it is a hit on revenues and a potential hit on costs as well.

Q80 Dr Coffey: Technically, in the rollout of 4G broadband, is there anything significantly different between 800MHz versus 900MHz, those two neighbouring parts of the spectrum?

Phil Sheppard: In essence, not one that counts a great deal. They both have near identical characteristics in terms of in-building coverage and coverage. There are differences in standardisation and timing. The 900 spectrum is usable now for what is called HSPA and HSPA-plus, which is basically a 3.9G service, if you like. That is usable now and that is what O2 are doing. The speeds that that technology provides are going to be quite similar, with small differences, to the other technology, which is LTE, which is a technology likely to be rolled out in 800. There are subtle differences in what can be deployed technically and what devices but, to all intents and purposes, they are very, very similar.

Q81 Dr Coffey: If we were to recommend-I am not saying we are-that O2 and Vodafone should not even be allowed to have any 800MHz, would that unduly affect their customers?

Phil Sheppard: There are two issues here, basically. One is that the coverage will be the same with the 800s; there would be no difference there. The technology deployed would be slightly different, certainly in the next few years, so they would have maybe more benefits initially because they have the HSPA technology. Over the longer term, they have to clear 2G customers from that band and we believe that is fully possible to do. The number of 2G devices is rapidly diminishing in the market, so that will be possible. We don’t believe that will have a negative impact if that were to happen. We think it is feasible for them to provide very competitive services on that, but that is not what we are asking to happen in a sense. We are asking for a minimum requirement such that there can be competitors in the market.

Q82 Dr Coffey: Could I cover the pricing and value? I live in a part of Suffolk where I have to go to one corner of the bedroom to get an indoor phone signal.

Mr Sanders: Too much detail.

Dr Coffey: I am just trying to give an indication; if anybody tries to call me on the mobile phone, getting through happens by chance because I have to be in a particular place. What I am trying to get at, frankly, is: who are your more valuable customers? Is it not going to be the people at the 2.1GHz, the 2.6GHz anyway, as opposed to the wider ranging 800MHz? What is more valuable to you in terms of data services and usage and applications and similar? I would have thought it would be the higher end of the spectrum.

Kevin Russell: I will have first kick because I start with a degree of frustration. In the previous session I wanted to jump in on population coverage. We want and need, in my view, as a company to roll out 97%, 98%, up to 99% indoor coverage-not population coverage, indoor coverage-because UK consumers and businesses increasingly expect comprehensive coverage wherever they go. It is not just about covering Suffolk because people live in Suffolk; it is covering people who travel to Suffolk. I fundamentally believe that the level of coverage that is there in 2G today, which is up at 99%, should in future be matched or surpassed in 3G in terms of indoor coverage, not just outdoor coverage. It is not a case of someone in London being more valuable than someone in Suffolk. You have to, as a basic, cover the whole of the country up to a level where it becomes uneconomic and I believe that level is more in the last 1% or 2%, not a number higher than that, if that makes sense.

Dr Coffey: Yes, it does. Thank you.

Q83 Paul Farrelly: I want very briefly to try and put this in some international context. Could you tell us how competitive the UK market is at the moment compared with France or Germany, or even potentially Hong Kong, and how good the UK has been at maintaining a competitive market as against other countries?

Richard Moat: If I can go first, I think that the UK is one of the most competitive markets in the EU. If you look at call prices, they are among some of the lowest of anywhere in the EU. Finland was often regarded as being the cheapest place to make a mobile phone call at one time, but I think now the UK has overtaken it. That is obviously very positive for consumers.

At the same time, I think that regulation has been quite strong here and that the latest proposals from Ofcom with respect to termination rates will result in very low rates, half of one penny at the end of the glide path, if that is implemented. The competition associated with the regulation means that the margins that are achievable in the UK are among the lowest, if not the lowest, in the EU 27. So there has been quite a significant impact from all of those factors in terms of the way the UK market looks and obviously that was one of the driving forces behind the consolidation from five operators to four, which took place when Everything Everywhere was created.

I think that, looking forwards, Ofcom has taken the right approach to try to create four viable operators who can compete effectively in the future and that the platform is being laid for that. We believe that, with some of these tweaks that we are hoping can be made, we can lay the right scenario for coping with this huge data growth that is coming over the next 10 years. The adoption of mobile data and of data generally in the UK is among the highest in the world, so we have a major role to play in coping with that.

Q84 Paul Farrelly: The supplementary I wanted to ask is this. If Ofcom doesn’t get this right and doesn’t give you what you want in terms of your level playing field, there a risk of there being a big blot on the UK’s successful record, but is the consensus that the UK has a pretty successful record?

Kevin Russell: I think you have to go down a level on the number of competitors in the marketplace. You talk about four competitors in the marketplace. The competitive tension, in my view, comes from the fact that you have three established players but one new player who is subscale and having to grow, so the pressure on Three to acquire customers from other operators drives us to expand our network, drives us to put aggressive pricing into the marketplace, drives us to innovate. When you look across Europe and the markets that have led on mobile broadband penetration and led on mobile broadband pricing, they tend to be the markets where there is a greenfield 3G operator-a smaller, newer entrant who has to force the pace in the marketplaces. It is too simplistic just to say it is a matter of having four operators or five or three. Someone has to have the need to grow. That is one point.

The second point is that I think the competitive impact of the merger of TMobile and Orange has yet to be felt in this marketplace. There is a fundamental consolidation going on where five retail brands will go to four. That has not happened yet. When it happens I do believe there will be a degree of competitive tension that comes out in the marketplace. I believe there is a fragility to the competitive dynamic of the UK marketplace that we almost shouldn’t take for granted. This decision on spectrum fundamentally sets up the competitive structure for the UK mobile market for the next 10 years, and I think that competitiveness is going to ease already because of the merger. The danger here is that five goes to four goes to three in a very short period of time.

Today, I think we are okay, but I don’t think we are incredibly aggressive. I push back on the benchmark that says it is the most aggressive marketplace in the world-go to India, go to Hong Kong; there are many markets that are far more aggressive than the UK competitively-and I think there is a fragility to it.

Q85 Mr Watson: I think this might have been covered in your earlier answer, Kevin, but we heard from Arqiva that they are proposing that 99% coverage should be based on one of the auction licences. Do you both think this is technically feasible and financially feasible?

Kevin Russell: Working through coverage forecasts is not a difficult thing to do. We know exactly where we have 12,600 sites today, we understand the propagation of the transmission capabilities of 800 and other frequencies, and we know where people live, so I know that with the site infrastructure that we have today, which is shared with Everything Everywhere, as well from our calculations, which are accurate, that if we had 2x10MHz at 800 we would be able to provide, as soon as that frequency was deployed, coverage with speeds of over 2Mb to 97% of the UK population. That 97% is indoor population coverage. Outdoor would be comfortably over 99%. To go beyond that, to go from 97% to 98%, we estimate probably would incrementally cost about £100 million in terms of sites. To go to 99%, we think it would incrementally cost about £270 million. That is our internal calculations.

Q86 Mr Watson: Presumably that is because you have to get to the Outer Hebrides or something like that?

Kevin Russell: You are not going to the Outer Hebrides at that point. You are going to a further point. But there is an incremental cost on sites. So that 97% number is based on the site footprint that is there today. It is not about going out to acquire additional sites.

Q87 Mr Watson: Richard, would you concur with that?

Richard Moat: Yes, I concur with that. Obviously the calculations are pretty identical because we share the same network, and that would be the backbone of the network that would be created using this extra spectrum. I do think, especially from what we have learnt from the broadband trial that we have been running with BT in Cornwall, that when you get up to the marginal coverage beyond 99% there has to be a combination of fibre, satellite and mobile to make sure that those people get that coverage.

Mr Watson: That is from 99% to 100%, or that is from 95% to-

Richard Moat: Yes, 99%. Well, 1% or 2%, it depends where you are, at the top end.

Kevin Russell: This is very, very doable. This is not a drama, the coverage obligation, in my view.

Q88 Mr Watson: Can I ask why you think this rather commonsense proposal has not been enacted before now, or is that a stupid question to ask you guys?

Kevin Russell: Because you have differences in site footprints. Everything Everywhere and we have very clear views that, as data takes off, you are going to have to have a lot of sites to support it. You are going to have to have a lot of sites for capacity and speed. We have proactively over the last three years built or expanded our site footprints from 7,000 or so sites all the way up to 13,000. We have done that from 2008 through to the present day. We have prepared ourselves for it. I strongly believe that O2 and Vodafone have not done so. Their site footprints are significantly less and, therefore, the coverage numbers look more daunting for them and would take longer.

Q89 Mr Watson: Just for the benefit of us novices, when I use the word "3G masts" does that mean "sites" to you?

Kevin Russell: It does.

Mr Watson: Just so that I don’t end up getting confused.

Richard Moat: One thing I wanted to point out as well is it is not just a question of providing coverage to 99%-plus of the population. It is also the throughput that you provide. One thing to emphasise in terms of the minimum amount of 800 spectrum holding we would like to have, which is 10MHz, is that if you only have 5MHz the technology basically says that you can only deliver 1.3Mb. If you have 10MHz then you can do more than double that because it becomes more and more efficient the more spectrum you have. It is a question of the throughput that you want to deliver to as many people as possible, as well as simply the coverage itself.

Q90 Mr Watson: Would that be the reason why every iPhone user in SW1 can never get any phone calls or signals, just anecdotally?

Richard Moat: No, I think that is a more general capacity-related issue.

Q91 Mr Watson: Okay. On a more technical point, do you think Ofcom’s proposals to limit future licence fees on the outcome of the next auction will help remedy the distortion in the market of the operators who don’t have 900MHz licences?

Kevin Russell: I apologise, Mr Watson. Could you state the question again?

Mr Watson: Yes. Will Ofcom’s proposals to link the future licence fees on the outcome of the next auction help remedy the distortion in the market for operators who do not have the 900MHz licences?

Nicolas Ott: Our view is that the 900 has been liberalised since January, and since January the people benefiting from the 900, or at least one of them, are communicating on the benefits they are getting from it. The question back to Ofcom is, as they already have the benefit, why don’t they pay the spectrum fee already of leveraging that benefit? Why should we wait one or two years before Ofcom adjusts the spectrum fee on the 900 band?

Mr Watson: Sorry, we are having trouble understanding. Could you repeat those last few sentences again?

Nicolas Ott: I understand your question being, "Is Ofcom’s proposal for the spectrum fee going to contribute in rebalancing the imbalance versus the 900?"

Mr Watson: Yes.

Nicolas Ott: My answer to that was the 900 has been liberalised early this year and there has been already a refund into 3G for O2; for Vodafone we do not know. Our point is what we have said in our answer: that from the day an operator can refund his spectrum into 3G they should immediately pay the increased spectrum fee because they get the benefit of it. Our question to Ofcom was, "Why don’t you do it faster?"

Kevin Russell: I am taking two parts to that question. One is if someone has low-frequency spectrum and others do not, recalculating the licence fee to charge that 900 holder more would not cure the distortion. That distortion will be a fundamental, ongoing shortfall in coverage that will go to your brand and your ability to build your business. All operators need to access low-frequency spectrum.

There is a second point in there, which is that a fair price should be paid for both those allocations of spectrum-the 900 spectrum and I believe the 1,800 spectrum. There are three trunks to spectrum: one is auction spectrum in 2000, another one is refund spectrum in January of this year, and then there is another auction coming up. This bit in the middle was allocated administratively and we have to be very careful, as you alluded to in your first question, that a fair price is paid for it. Currently, and this is what Nicolas was alluding to, the 900 has been used and the 1,800 is available to be used for 3G and mobile broadband without an incremental change on the licence fees. That, to Three, does not seem fair and appropriate.

Q92 Mr Watson: How do you feel about that?

Richard Moat: I feel exactly the same. I think that there have been some well publicised statements by some of our competitors about the fact that if some 1GHz spectrum were to be reserved for either Three or for Everything Everywhere, this would be classified as "state aid", but our view would be that the ability to use for 3G services 900 that is available now-bear in mind whatever spectrum we purchase in this auction will not be available until 2013 or beyond, so we are talking about a significant advantage over a number of years-is not a fair approach either.

Q93 Chair: What do you say to the argument that your competitors have used, that although they are now able to use the 900MHz for 3G services, it is already being used for all their voice telephony, but they can’t just clear out thousands of customers in order to vacate it for that purpose?

Nicolas Ott: It is difficult to comment on the way they are managing their spectrum and their customer relationships but, from what was available in the press, I think they were explicit, at least one of them was, that they perceived a significant advantage in being in a position to offer their customers 900 3G coverage. Back to what was explained before by the representative of Arqiva, it is a well known fact that you have immediately a better indoor coverage and a better rural coverage, so yes, they are benefiting from it. That is why it also matters that the auction proposed by Ofcom takes place reasonably fast, because if it is further delayed then the availability of the 900 is postponed even more and that will put us in a very difficult situation. It matters to progress fast now. Going back to what we were saying before, that is also why we do believe it is critical to be in a position of benefiting from 2x10 minimum of 800 because then, hopefully, when it becomes available, we will be back on track to be in a position to compete.

Kevin Russell: My view on the very strongly raised concerns of O2 and Vodafone is influenced by hearing large volumes of comments over the last five to seven years to the same effect, that once 900 is liberalised it would be a long time before you could redeploy it; it could be years; there is an incredible level of difficulty. To wake up a couple of months after it was liberalised and find that it is now operational in London leaves me slightly sceptical that it is as difficult as previously claimed.

Q94 Dr Coffey: I think I heard you say earlier, Kevin, that you have estimated it would cost £270 million, and you didn’t demur from that, Richard; so it might be similar. What is to stop the operators sharing infrastructure for that last extra bit of coverage and split the £270 million between you?

Kevin Russell: Nothing, practically. When we look at that final piece of coverage, it is eminently sensible for there to be a level of collaboration across the industry. I think the UK market has matured dramatically over the last three years and I think we will take some credit for that, EE and ourselves, in infrastructure sharing. There were groundbreaking deals done in 2007, followed through by O2 and Vodafone in 20082009, and there are now two networks that share a lot of infrastructure. The ability to have, for example in Northern Ireland, one network would be more economic and I don’t think it would be giving much away competitively. There is scope for discussions to collaborate on the more marginal areas, which is why I still think that strong coverage obligations should always be put on these licences, because the industry will then work through solutions to do it economically.

Dr Coffey: Do you have a view, Richard?

Richard Moat: I agree with that absolutely. I think that we should seek every possible means of collaborating to reduce the costs and the impact on the environment, as I said earlier.

Q95 Dr Coffey: Forgive me, I wasn’t due to ask this question but it is the one I put to Arqiva about Jeremy Hunt’s idea of a digital pump. We are hearing different things about whether the BDUK money will deliver that. Is there anything stopping you basically piggybacking on the new rural network that this money is supposed to provide for?

Richard Moat: No, I don’t think so, is the basic answer to that question.

Nicolas Ott: The digital pump in the village has the strong advantage of being in a village where there is probably a BT cabinet already, so there is probably also already some BT fibre. I would suspect, from that point in time, that mobile sites become reasonably easy; provided we find the right landlord, that becomes reasonably easy. I suspect that you already have, in fact, most of the mobile sites, so it is something that can be leveraged. I suspect it is de facto already leveraged because the BT cabinet is already there, because in this area we buy all the backhaul, if you want, mostly from BT, if you want to simplify a little bit.

I guess the key issue in the coverage when you go for 95% or 96% or 97%-more than a village and, therefore, the pump of the village-is you are dealing with much more remote places, such as the road between the two villages and so on. It becomes more of a trick to have the right site where there is no BT cabinet, because we are too far away in the countryside. That is where, as I was explaining, microwave can play a role. In a nutshell, the digital pump might help, but I suspect it might help only where we are already present in one shape or another. The issue is where there is no BT cabinet at all; that is where the challenge is.

Q96 Dr Coffey: Funnily enough, I thought the whole point of the extra money was to go where BT wasn’t already. I thought that was the whole idea.

Nicolas Ott: The point is even with the subsidies of BDUK, BT will not be in a position to go everywhere, because it is just too far away from their main network. It is back to the laws of physics. Some regions are better off than others-for example, in Cornwall where we tried with them, it is reasonably okay because the population is reasonably dense; there is always a piece of network somewhere. If you go in Cumbria, though, BT has a far lower density of networks, so it is much more difficult, even with BDUK, to pull their lines much further. In Cumbria, quite clearly, if you want to cover all of the region, the cellular aspect of digital Britain would be far more important. For example, I know where BT can go but I suspect what the micro-cellular network will do will be quite a fair chunk of coverage. It is back to what was said before by the Arqiva representative-that all the regions are not equal between themselves-but the issue is also what BT has or does not have in these regions. That is why it needs to be tailored on a by-region basis.

Chair: I think that is all we have. Thank you very much.

Prepared 24th June 2011