Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
ENVIRONMENT AND COMMUNICATIONS LEGISLATION COMMITTEE
04/03/2011
National Broadband Network Companies Bill 2011; Telecommunications Legislation Amendment (National Broadband Network Measures—Access Arrangements) Bill 2011

CHAIR —I welcome Professor Walter Green from the Communications Expert Group. Thank you for coming along today to discuss these issues with us. The committee has received your submission as submission No. 7. Do you wish to make any amendments or alterations to your submission?

Prof. Green —No changes, thank you.

CHAIR —Do you wish to make a brief opening statement before we go to questions?

Prof. Green —Yes, please. There were two sheets of paper, one containing slides 1 to 8 and the second one containing a network diagram. The background information to these is that there are developments in the way the NBN is being constructed and built that in fact will lead us backwards in the telecom regulatory environment.

CHAIR —Professor, just before you go any further, I request, if you are seeking to table those documents, that we have the documents.

Prof. Green —I am seeking to table them. I only prepared them as soon as I knew I was coming to the meeting. Copies will be made available.

CHAIR —That is fine.

Prof. Green —The key issues that I wish to raise in my opening statement are why the NBN fees are nearly irrelevant in determining the prices paid by end users and RSPs for services; the potential for as few as three companies, or RSPs to connect to the points of interconnect offered by the NBN, regardless of the number of RSPs and how that will affect the downstream market; and the need for engineering and security audits.

Slides were then shown—

The slides are taken from a talk I gave in March 2009—that is, two years ago—which outline the UK market. These slides were prepared by somebody who works within the UK market and has a very good understanding of what is going on.

What has happened there, through complex negotiations for access to Openreach, has limited the number of people who buy services from Openreach, which is the equivalent of NBN, to 10 in the UK market. Everybody else is limited and that is what we call the UK tier 2 and tier 3 and they are forced to buy from these primary tier 1 RSPs—service providers or retailers as they call them. In effect, those 10 control the prices to everybody else. While sometimes the decreases in price mandated by Openreach are passed on to end users, there are some instances where they are not. In most instances where there is an increase in price that increase is passed on, but where there is a decrease in price there have been instances where it has not been passed on.

The NBN is currently unclear but it is looking like it is following the same path as Openreach, requiring high capital investment and making the process to get access to these points of interconnect very difficult. My expectation is that we will only end up with three or possibly four people who will actually be buying services from the NBN.

If I can refer you to the network slide, you will see at the bottom end there are triangles. They represent the points of interconnect that the NBN will build throughout Australia. While six are shown on the diagram, you will appreciate any number may be involved. It is the black line that goes from the POI up to the rectangles, which are the RSPs. The people who control those fibre links throughout Australia, and bear in mind there is substantial investment required to build those links, are the ones who will ultimately control the pricing to what RSPs and what other people generally refer to in the market. My expectation is that, because of the investment, there will only be three, possibly four. In other words, we will have a similar situation to where we had three mobile carriers, where there was limited competition and prices remained high. It took additional entrants into the market to force a price reduction, and, I might add, the ACCC also had to intervene to get price reductions.

The other thing to note is that if you look at slide 4, which is titled ‘UK 2009’, due to Openreach—the kind of network and some of the prices being forced on but not all of them—they were still able to reduce the average household bill for telephone and internet down to $40 a month. This is nearly half what we pay in Australia. The average call for a national network call was down to less than 2c. It is absolutely imperative that this bottleneck of the link from the NBN POI to the RSPs should not be allowed to be created. Unfortunately, the discrimination provisions that are put in there are exactly the sort of fertile ground for this limited number of tier 1 people, who can provide this link, can grow. As mentioned by the previous people, once you get the group in they are entrenched and you literally cannot move them. The same thing has happened in the US with the group of four tier 1internet services there. Any kind of discrimination is likely to disadvantage people getting access to the NBN in the future.

Also of concern is, under the Freedom of Information Act, one needs to recognise that governance and project management on average only have a 20 per cent effect on the price of delivering a service. If you get the engineering or security issue wrong, you in fact can wipe out the service completely. In other words, you have 100 per cent impact. The difficulty we have is that the NBN was proposing 14 POIs in the first place. This should never have been allowed on management. The first is the fibre network in Australia is just not capable of providing all the connectivity that is needed to connect to 14 POIs. This is simply because if you take the 13 billion subscribers that is announced in the NBN plan of 14 POIs, it is nearly one million customers connected at each point of interconnect. This makes it an obvious terrorist target for maximum disruption but, in the event of a natural diaster, like the floods in Brisbane, you would wipe out the connections to one million people in a disaster situation. So it would be interesting to know where the 14 POIs are and if any of them are in places like the floodplains of Brisbane.

The other point is that if I refer again to the diagram on the network, the NBN is going to be seeking funding for the points of interconnect and the fibre to the customer’s premises. As I indicated, the dark lines from the point of interconnect to there do not exist at the moment, and there will be other people seeking finance to get access to build this network. There will be competition for telecom investment against the NBN. For this reason it is more important that the engineering is fully understood, clearly explained and audited in order to give value for money. Based on my experiences in building equivalent networks in Western Australia, if you use one benchmark—the standards and everything announced by the NBN—we are looking at a price increase of two to four times. If you use the successful business model that was used for a structurally separated carrier in Western Australia, the benchmark price is somewhere between 10 and 20 times the price. In other words, what the NBN is trying to build it for and charge is 10 to 20 times what we were achieving in Western Australia. Thank you very much.

CHAIR —Thanks, Professor Green.

Senator FISHER —Professor Green, thank you. Your concern about volume discounts—do you think that despite the proposed or noted wholesale price of $24 a month that NBN Co. is talking about that NBN Co. can still find ways to charge differing wholesale prices to large and small RSPs?

Prof. Green —Yes, not so much in the access from the POI to the customer but in the costs that are charged to get access to it. There is a whole second tier of tariffs and it is at that point that the gaming and manipulation can occur.

Senator FISHER —Can you expand on that a bit.

Prof. Green —If I can go back to my diagram on the network: where you have the black line that goes from the RSP to the triangle, it is the connection at the tip of the triangle. That is where the NBN has a variety of prices which in fact make it difficult for a smaller RSP to connect but make it easier for a larger one so that in effect will drive down the number of people who can provide the link—in other words, we come down to three or four tier 1 RSPs. The part that does concern me with the debate in Australia is that nobody is looking at the fact that your market is going to be very similar to the UK where you have tier 2 and tier 3 RSPs. Only a few will actually connect to the NBN.

Senator FISHER —In that scenario, for example, for sophisticated applications that need lots of bandwidth, does that mean that the volume discount means that the fees that certain RSPs will have to pay will include an amount for reserving their capacity, if you like?

Prof. Green —That is only one of many ways of doing it. I think it is worth pointing out—

Senator FISHER —Sorry: one way in which NBN Co. might extract more money?

Prof. Green —Correct. From one access seeker compared to another. That is why I am saying there needs to be an audit of the commercial conditions as well as the engineering. The audit of commercial conditions will prove that there is equitable access and that the pricing is not disadvantaging some access seekers against others.

I think it is worth pointing out that in the UK there is not volume discount. They have found that it causes so many problems it is not worth it.

Senator FISHER —They used to have volume discounts, did they?

Prof. Green —They attempted it but it became a nightmare so quickly, so Ofcom, the regulator, stepped in and pretty much forced them to say, ‘You will have the same price throughout.’

Senator FISHER —Over what period of time?

Prof. Green —This was in the very early days nine years ago.

Senator FISHER —So should the government legislate or provide a framework for wholesale pricing—as loath as I am to discuss that scenario?

Prof. Green —I would say you would be creating a can of worms or complexities for gaming that you would never get around. Already there are five generic methods that carriers use to increase prices to customers, and they are all involving ‘nonprice’. As soon as you open that particular one you will never regulate it properly because there are so many combinations that one can do to get around it. This is effectively why the UK said, ‘No, we want a simple one type of contract and one type of commercial condition and one price structure, and that is it.’

Senator FISHER —Thanks, Professor Green.

CHAIR —There are a few points here, Professor. You indicated that there was a problem with 14 points of interconnect, is that correct?

Prof. Green —Correct.

CHAIR —You might not be aware of this. Since December there have been 121 points of interconnect.

Prof. Green —I am aware that 121 had been proposed. The statement I am making is the fact that the issue of 14 points of interconnect ever got as far as it did indicates significant concerns in the way security and engineering issues are being managed at the moment. This is why I am recommending that not only the management aspect but the commercial and engineering aspects need to be audited as well. You do not want to create a single point of failure in a natural disaster or create a terrorist target—because you take out one of those and a million people lose services.

CHAIR —But, given there are going to be 121 points of interconnect, that minimises some of your concerns?

Prof. Green —It substantially minimises concerns. But the real issue is that the 121 is all that should have been offered, not 14, and the fact that they were even prepared to go as far as they wanted for 14 indicates some serious concerns. I believe it is that kind of discussion and that kind of pushing those issues that led, and it was reported in the Australian Financial Review, to some of the leading financial banks now saying, ‘Work on the principle that the NBN won’t happen, because they won’t get the financing.’ In fact, the audit process will help the NBN get the funding. It is important for Australia—and for WA, and I say WA because we need it more than anyone—to move to a fibre-to-the-premise environment.

CHAIR —Do you support the principle and what we are trying to achieve with that?

Prof. Green —I support the proposal. As for what you are trying to achieve, I am saying at the moment there are concerns that by not doing things correctly it could either significantly slow it down or make it not sufficiently competitive or cost effective for the bulk of the population to get access to it. I support the NBN in principle but there are details that I think need to be looked at carefully. Unfortunately, they are parts of the legislation which I believe are weakening the NBN’s case.

CHAIR —And one of them was 14 points of interconnect?

Prof. Green —Correct. To me there was inadequate engineering and security review in even letting that proposal come out to the public.

CHAIR —But you are now comfortable with 121 points of interconnect?

Prof. Green —I am actually saying the figure should be north of 200—121 is better but there is substantial investment required within Australia to support 121. If you look at today’s technology you see you are better off with 200. Let us put it this way, 121—

CHAIR —I suppose 1,121 would be even better.

Prof. Green —is far better than the 14, yes.

CHAIR —But you say you want 200 and I am saying I suppose 1,100 would be even better.

Prof. Green —Sorry, 1,100?

CHAIR —Yes.

Prof. Green —No. The optimum is somewhere between 200 and 300.

CHAIR —So we are getting close to the optimum.

Prof. Green —Yes.

CHAIR —That is good. I am glad that is not such an issue for you anymore. Can you explain how the NBN costs will be 20 times more?

Prof. Green —Sorry?

CHAIR —I think you finished with your last point being about costings in Western Australia and you said that the NBN—this is what I wrote down—would be 10 to 20 times more expensive.

Prof. Green —Yes, if you use the benchmark of what is provided within greenfields properties and you take the cost of providing fibre to the premises—and WA actually has more houses that are equipped with fibre—what the NBN are proposing, if you take their standards and implement them, is somewhere between two to four times the price of what providing fibre in Western Australia or Perth is costing at the moment. There is a business model that was used successfully in WA. If you use that as your reference point for costs, then the NBN is costing something like 10 to 20 times. The business model of the NBN needs to be looked at carefully. There are various ways you could do it. The opposition has suggested the Productivity Commission. There is no real audit or check of whether you are actually doing it the right way to deliver costs when we already have experience both overseas and in WA where we have done it cheaper.

CHAIR —But you cannot really compare the rollout of a national network across the country with the demographics of Australia to the rollout of a network in Perth as a city, can you, really?

Prof. Green —In regard to the connection with from the point of interconnect to a customer’s premises, the size of Australia is irrelevant, because that is something that is local. The only thing that varies is the link between them.

CHAIR —The point of interconnect might be local but the infrastructure has to be rolled out nationally.

Prof. Green —Yes, but we are only dealing with the fibre that goes from the point of interconnect to it. Whether it is rolled out in one city or small villages and so forth is actually irrelevant. In fact, I have cases with two shires in WA where the cost of providing fibre to the premises for a small village community of 100 houses is cheaper than in many metro areas. The issue is the connection from that community back to a point of interconnect or wherever you can aggregate those services. It is the backhaul that really controls the price, not the last mile.

CHAIR —Yes, but doing in this nationally, if we are going to provide equality of access across the country and a standard price, surely there has to be some give and take in some of the high-end areas in terms of the costs to subsidise some of the other activities that are taking place?

Prof. Green —We have had to work across a variety of territory from rock, sand—all the variables. The volume of the infrastructure of the NBN is, in our opinion, excessive. You can do the same thing and yet achieve all the goals. It comes back to whether the engineering design is an optimum design. The view is that there are more economic and effective ways where you can build a national network at a lower cost.

CHAIR —That is your view?

Prof. Green —That is the view and that is my experience in actually building these things.

CHAIR —What is the biggest network you have built?

Prof. Green —It was 3,000, but I have actually been involved with three large estates and developments.

CHAIR —That is fibre to 3,000 properties?

Prof. Green —Correct.

CHAIR —This is a much larger undertaking than that, isn’t it?

Prof. Green —Once you go to that size, you just duplicate that across the country.

CHAIR —Are you sure it is that simple?

Prof. Green —I am pretty sure from the engineering it is simply a project management issue of getting the right materials at the right time with the right government approvals.

CHAIR —What is ‘pretty sure’?

Prof. Green —The biggest problem is actually the connection, as I say, from the estate back to the centre of Perth, which is that thick black line from the point of interconnect, which in WA is called the head end, which is equivalent to the RSPs. That is where the costs are controlled and affected in delivering services to the end customer.

CHAIR —I am just trying to get my mind around your experience on this and it is doing 3,000 households. I do not want to diminish that experience at all, but this is not 3,000 households. The NBN is of huge magnitude compared to that. Didn’t the ACCC recommend that POIs be located where there is competitive backhaul? Wouldn’t that lead to lower backhaul prices?

Prof. Green —This comes back to the point that there is not enough backhaul in Australia. We effectively have Telstra and that is it. We have very few locations, mainly the capital cities and a few others, where there is an alternative fibre provider. In order to get three or four people who can provide that competitive backhaul, that is where additional extra investment is required. I come back to the logistics and the design. Once you have proved the design for 3,000 or even 2,000 properties, rolling it out across the different suburbs and the different localities within Australia is fairly repetitive. You are literally down to asking ‘What is the money?’ and organising the supply chain. That is the only real difference. You have to get the engineering to make it cost-effective right on the first one and that is where I am saying I think there are better, more cost-effective engineering designs that could be used.

CHAIR —Your submission basically says that, if we can do it 20 times cheaper, we can build the National Broadband Network for $1.8 billion.

Prof. Green —That is using a benchmark with an alternative provider. In fact, looking at where I believe the NBN should have been done, there has been a substantial increase in costs because the successful WA model relied on the telecommunications being put in at the same time as pavements, powerlines and sewerage. The price difference there was, say, $1,000 per property. Because we were doing it in conjunction with another provider we were doing it for $70. That is why the business model was acceptable. Bear in mind that 80 per cent of the NBN is in fact putting the pit and pipe and the tubes in the ground.

CHAIR —Professor Green, I say to you that I am sure there will be ministers, there will be bureaucrats and there will be expects beating a path to your door if you can build this for $1.8 billion.

Prof. Green —I am not saying I can do it for $1.8 billion. I am saying there are substantially cheaper ways of doing it.

CHAIR —But that is the effect of your submission here.

Prof. Green —I am saying there is a substantially cheaper way.

CHAIR —That is a changed submission. That is okay. The ACCC rule specifies that the POI location is defined as places where there are three backhaul providers. Are you aware of that?

Prof. Green —Yes, and that is why I am saying there will be three principal connectors to the network who will provide the connections to the other RSPs—in other words, the tier 2 RSPs.

CHAIR —Coming back to this cost, I am sure you do not want the front page of the Australian saying that you can build this for $1.8 billion. You are now saying it is for substantially less. Have you discussed these issues with NBN?

Prof. Green —I have had meetings but, as I said, they have chosen to do it that way and they have disagreed with me on what I have said.

CHAIR —They have disagreed with you?

Prof. Green —Yes, they wish to put in too much infrastructure and I believe there are alternative engineering solutions where they can achieve their objectives using less infrastructure at a lower price.

CHAIR —Why did they disagree? Can you outline the technical reasons why they disagreed with you?

Prof. Green —They did not give any reasons. They just said, ‘No, this is what we have decided we are going to do. That is it.’

CHAIR —But you said they disagreed with you. That is not disagreement; that is ignoring.

Prof. Green —Ignoring or disagreeing—they said that that was what they believed was necessary and had to be done. They are in charge and that is what they have to do.

CHAIR —Any further questions for the professor?

Senator TROETH —I have some. Professor, you have mentioned in part 3 of your submission:

The provisions of the Legislation with regard to the determination of when the “NBN” is complete, is vague, and subjective.

Would you like to elaborate on that? I would also be interested in your view of Optus saying that there should be a limit of 15 per cent ownership of NBN Co. by any one retail service provider.

Prof. Green —The first point is—sorry, I just need to check my papers.

Senator TROETH —I am referring to page 4, part 3, conditions for sale of NBN Co.

Prof. Green —Coming to the sale of it, I was a chief engineer in Rhodesia which was managing the changeover of the analogue telephone network to the digital telephone network. I have had to deal with a large-scale national network. A network is always evolving and changing. There are parts being removed and there are parts undergoing upgrading at any one time. I have tried to think through how to define the NBN as being complete. Using the simple metric ‘I have connected 13 million subscribers’ is irrelevant because in 10 years time we may need 14 or 15—in which case it is not complete; it is still incomplete. Or it may only get to 11 million and that may be because of demographics or various conditions, so you will never get to the 13 million.

It is very difficult to define any kind of metric which will say, ‘This network is complete.’ That is why I say that at the point in time that any government in the future should, subject to the provisions in the act, be able to say, ‘Okay, yes, it is acceptable to sell the NBN.’ There may be opportunities to get better value for money for the taxpayer by selling it halfway through construction rather than trying to wait for some nearly arbitrary decision as to when it is ready. There is no clear metric for any regulator or minister to work to as to whether it is complete. There are so many variables and it is even more complex on a 10-year project. I was only dealing with a five-year project but the changes within the first two years were quite phenomenal and I am expecting the same thing to happen here.

My second point is that I disagree with any RSP having any funding, capital or equity within the NBN because the British Telecom and Openreach still have the allegation that BT actually receives a subsidy from Openreach because there is a joint holding or the financial control is from a single entity. So you have this back path. To eliminate that particular one there is no need for an RSP to invest in it because they will automatically be seeking some kind of discount beneficiary—you name it—and that undermines the whole basis on which the NBN was formed, to get equitable access.

Senator TROETH —Do you consider that it should remain as a government owned entity?

Prof. Green —I am saying that it is acceptable for it to go to a private entity but there should be no shareholding allowed by an RSP. In fact Art Price from AXA Communications has clearly stated the only reason he succeeded was that he had no interference from his downstream purchasers.

Senator TROETH —It should be another entity, not necessarily government, but one formed strictly with the view of running the NBN as a wholesale service provider?

Prof. Green —I believe it should be wholesale only. My other concern is that there are also difficulties in determining who can get access. This is why I am saying you would probably need to redefine the business activities of the NBN from the point of interconnect to the customers’ premises. That is all they are allowed to do. It gets rid of all sorts of problems. To me the current definition in the legislation is too vague and I believe will undermine the viability and effectiveness of the NBN in delivering its desired outcomes.

CHAIR —Professor Green have you any further qualifications as an economist?

Prof. Green —No, I do not but I have had to work with pricing and seeking funding on enough cases to put the business case together.

CHAIR —The business case for a 3,000-house—

Prof. Green —This was for quite a wide range of projects.

CHAIR —But the biggest project was 3,000—

Prof. Green —I saw the funding in detail for 3,000, yes. I have also seen Telstra’s funding and various other people’s as well.

CHAIR —It seems to me you are coming at this almost from a test tube approach—3,000 houses—compared to what is happening. You know that moving from the test tube to real life does not always equate. The point I am making to you is that you are in the test tube area, with 3,000; NBN is in real life, building a national network, and to compare the two is just not easy, is it?

Prof. Green —I am prepared to say that, adopting different engineering designs and a different approach to the way you go about it, yes, there are significant savings and that is why I said about the benchmark that, depending on if you take the one with the different approach, the savings—and I accept there are a whole lot of things that affect this and that is why I gave a price range of 10 to one—can be reduced quite substantially. By exactly how much is going to be dependent on the cooperation with shires, local government and in fact state planning.

CHAIR —Have you done any modelling of the savings?

Prof. Green —The only effective way I could do it was to take what is the cost that we have done at not just one site but three sites, what is the cost of doing it there, now take the experience that we had where we had to do a brownfield development in Perth and look at those costs because that is what the NBN is currently using as its approach—in other words, take the ratio of those two; that is where you get the significant savings.

CHAIR —So in my terms you take the test tube results and apply them to the real live NBN?

Prof. Green —In some instances you would be able to do what was done successfully; in others you have what I call a raw brownfield where you have to dig up the pavements and everything else, and it is that ratio, how much you can get cooperation with the state planning organisations, local government and so forth, where you just have to go and dig everything up and start again, which is what the NBN is committed to. That is why I cannot say yes, I will do it for $1.8 billion, because it assumes 100 per cent with the wind blowing behind you. It will probably be more than that.

CHAIR —I am not being critical of this—you are entitled to have a view on all these issues and I thank you for bringing your views here—but you are now into the area of liquidation, where you say that, if NBN requires a business that has got retail operations, then the divestiture of those retail operations should be undertaken by a liquidator. Why a liquidator?

Prof. Green —There are two situations that occur. A carrier might be up for sale that has point two customer premises access assets, which are what the customer will be buying it for. If that carrier has gone into liquidation there will be a liquidator in place and the instruction to the liquidator in that case is transfer the assets required by the NBN to the NBN, and the rest the liquidator needs to sell off. The other case is where by private negotiation a carrier wishes to sell its assets to the NBN, then, rather than the NBN take over all ownership and do the divestiture itself, the people who have the experts in managing the company, effectively breaking up the company, are liquidators and I am saying keep it out of NBN’s hands, let the liquidator do the transfer of assets and the selling off because in fact liquidators do sell off the assets of companies that are being broken up and sold.

CHAIR —I think there would be business people listening to this around Australia who would be horrified to think that liquidators would be the ones that would be allocated the job of selling a going operation’s retail assets.

Prof. Green —I understand the difficulties within the liquidation industry, but it is a case where letting the NBN get into the retail space will, I believe, undermine the effectiveness in the delivery of services.

CHAIR —But that is in the legislation. Senator Troeth?

Senator TROETH —Surely, Chair, an economist has an equally valid view on some of these issues. I would have thought that for the chair to criticise Professor Green’s qualifications on this is somewhat lacking in the professional conduct of a chair. Surely it is the chance for a wider view to be taken of this, not just telecommunications specialists but an economist and other people who are equally well placed?

CHAIR —I am happy to hear your economist come and tell me this. I am happy to hear businesspeople tell me why liquidators should be allocated to a going enterprise which is simply selling off a retail business. I think I am entitled to ask the professor, given that he is putting that position to this committee, as to how that operates. I am entitled to pursue that in an appropriate way, and I think I have. That is fine. It is not the professor’s area of expertise; he concedes that. He is entitled to these points of view—

Senator TROETH —Good. I am glad you acknowledge that.

CHAIR —and I am entitled to pursue what the implications are of those points of view. Professor, thanks very much for coming along and assisting us. It has been very helpful.

Prof. Green —Thank you for the opportunity to present.

[11.51 am]