National Broadband Plan: Discussion (Resumed)

We resume discussion of the national broadband plan, NBP. From the University of Limerick I welcome to the meeting Professor Eoin Reeves, head of the department of economics, and Dr. Dónal Palcic, lecturer in economics, to discuss the national broadband plan.

I draw the attention of witnesses to the fact that by virtue of section 17(2)(l) of the Defamation Act 2009, they are protected by absolute privilege in respect of their evidence to the joint committee. However, if they are directed by the Chairman to cease giving evidence on a particular matter and continue to do so, they are entitled thereafter only to qualified privilege in respect of their evidence. They are directed that only evidence connected with the subject matter of these proceedings is to be given and asked to respect the parliamentary practice to the effect that, where possible, they should not criticise or make charges against any person or entity by name or in such a way as to make him, her or it identifiable. Any submission or opening statement provided to the committee will be published on the committee's website after the meeting.

Members are reminded of the long-standing parliamentary practice to the effect that they should not comment on, criticise or make charges against a person outside the House or an official either by name or in such a way as to make him or her identifiable. I remind members to turn off their mobile phones or to switch them to flight mode as they interfere with the sound system. I understand both witnesses will make opening statements. I invite Professor Reeves to make his opening statement.

Professor Eoin Reeves

I am head of the department of economics at the University of Limerick. Along with my departmental colleague, Dr. Dónal Palcic, we specialise in infrastructure-related research into aspects of public procurement, including public private partnerships, PPPs. We thank the committee for the invitation to appear before it this afternoon. A detailed briefing note has been submitted to the committee and I will summarise some of the main points in this opening statement. I wish to emphasise that we have had no direct involvement or advisory role in the NBP procurement. Our access to information has been confined to that which has been placed in the public domain, much of which has been heavily redacted. We wish to state upfront that having observed events and analysed documentation that is publicly available, we have formed the view that the procurement approach adopted has not satisfied the criteria necessary to ensure that the key objective of value for money will be achieved. As the committee is aware, the procurement model being used for the NBP is a gap funding model which is essentially a type of public private partnership.

I will now examine the features of the PPP model and relate them back to the NBP. PPPs can be defined as long-term infrastructure contracts between the State and a private contractor. The main characteristics of a typical infrastructure PPP include: competition for contracts; bundling of the different stages of the project life cycle into one contract; risk transfer to the private sector; and a significant element of private finance. Although the gap funding model possesses those characteristics it differs from other PPPs used in Ireland to date since ownership of the infrastructure will remain with the private contractor at the end of the contractual period.

The principal economic justification for adopting a PPP is that it provides scope for achieving value for money, VFM. However, the achievement of VFM, in turn, depends critically on what may be referred to as key VFM drivers. The main VFM drivers are competition for contracts and risk transfer and I wish to discuss both in the context of the NBP. The importance of competition for contracts cannot be over-emphasised. Competition creates incentives for bidders to reduce costs and delays and to develop proposals of higher quality. The recommendation in favour of the gap funding model was based on the assumption that the market for the contract would be competitive. However, once SIRO and Eir withdrew from the process the principal justification for continuing the procurement no longer applied.

In general, the potential negative consequences of continuing with an uncompetitive procurement are opportunistic behaviour by the bidder during contract negotiations, leading to price increases, and renegotiations after the contract is signed. The international evidence on public private partnerships, PPPs, shows that these problems frequently materialise when competition for contracts is absent. In this context, it is worth considering whether the evidence suggests that the lack of competition for the NBP contract has impacted on the project costs to date. One development that raises suspicions in this respect is the substantial increase in projected costs that will arise due to the preferred bidder rolling out fibre in the 300,000 premises area that Eir carved out of the initial intervention area. Would this have been agreed if there was real competition for the contract?

Optimal risk transfer is one of the main drivers of value for money in PPP projects. The available information indicates that the preferred bidder for the NBP will take on significant risk, particularly in terms of construction, operating and revenue risk. I wish to focus on financial risk. It is typical in privately financed PPPs for financial risk to be assumed by investors that contribute equity and loans to the PPP. Both providers have skin in the game and this incentivises them to ensure that contractual obligations are met. In the case of the NBP, we eventually learned that the level of private equity is €220 million. This is not a trivial sum of money, but it accounts for a very low percentage of total project costs. The concerns expressed by Mr. Robert Watt, Secretary General of the Department of Public Expenditure and Reform, were relevant in this regard when he noted that by 2028 the private bidder will have recouped most of its investment while the State will have invested €2.44 billion at that stage. In addition, the gap funding model for the NBP differs from earlier PPPs in Ireland as it appears that the special purpose vehicle is not borrowing significant amounts from banks or international capital markets. Rather, it appears that it will work on a cashflow basis and rely on payments from the Exchequer as the project progresses. As a result, unlike other privately financed PPPs, this gap funding model approach will not possess the advantages that arise from the due diligence and close monitoring of all aspects of the deal by experienced lenders. Overall, it appears that some of the risk-sharing advantages that typically apply to PPP projects do not apply in this procurement model, which increases concerns that the procurement will not deliver value for money.

I wish to make some points about governance aspects of the NBP. I will preface my comments by stating that the international evidence consistently shows that accountability and transparency are the critical conditions that must be satisfied if good governance is to be achieved. There have been several problems with the governance of the NBP. On the project appraisal stage of the NBP, I wish to highlight that there is very limited information about the cost-benefit analysis, CBA, in the public domain and the available data do not include detailed breakdowns of valuations, assumptions, etc. It is reassuring that the CBA has been scrutinised by the Department of Public Expenditure and Reform, but the fact that the Department concluded that "the CBA is not credible and it is questionable whether it is consistent with the public spending code" is a matter of significant concern. On the choice of procurement model, the principal point I wish to emphasise is that the financial appraisals underpinning the choice of the gap funding model are not publicly available. It is not clear if these were subject to independent scrutiny by the Department of Public Expenditure and Reform or the National Development Finance Agency, for example. Nevertheless, I recommend that these be published in order to improve the accountability of key decision makers.

The governance problems witnessed during the procurement process have been widely discussed. In economic terms, the fundamental weakness has been the decision to continue the process with one bidder. Other issues have arisen regarding the changes in composition of the bidder’s consortium, the drip feed of information about key aspects of the procurement, including the precise level of equity being invested by the preferred bidder, and controversy around interactions between the relevant Minister and investors in the consortium behind the remaining bid. Overall, the governance problems that have arisen are such that the procurement lacks the required level of legitimacy, thereby eroding public confidence in the process. There are legitimate concerns about the overall justification for the project, the chosen procurement model and the potential cost of the project and knock-on effects for other investment priorities. It is our considered view that the current procurement should be terminated and more affordable alternatives explored.

I will hand over to my colleague, Dr. Palcic, whose statement will build on the points I have raised.

Dr. Dónal Palcic

As my colleague noted, the briefing note we submitted to the committee contains more detailed information on the points we wish to raise. I will summarise some of my main concerns.

One of the most important and controversial aspects of the debate surrounding the plan to date relates to the gap funding model that has been adopted. A number of the assumptions made in the KPMG 2015 ownership report that underpinned this decision were questionable at best and appear even more dubious now, given the events that have unfolded since. The original assumptions underpinning the recommendation of the gap funding model were essentially that placing the long-term ownership of the network with the private sector would allow bidders to leverage the use of their existing infrastructure and encourage them to continue to invest in the network to develop and exploit new markets; and that the aforementioned benefits would be reflected in the strategic value that bidders would place on winning the contract, thereby driving down the amount of subsidy required in a competitive tendering process. Strong competition among existing industry players is highlighted by KPMG as one of the key drivers behind its estimation that the gap funding model would achieve the lowest cost to the Exchequer. However, it is difficult to understand how KPMG could assume a strong level of competition when only Eir and SIRO could realistically be considered as having the ability to leverage their existing assets to deliver the NBP at a lower cost to the public purse. Moreover, at the time that KPMG was making its recommendation, Eir had already announced plans to invest in approximately 300,000 premises within the intervention area. The risk of Eir moving forward with its plans and the impact that this would have on the fundamental nature of the plan appear to have been ignored when the decision to proceed with the gap funding model was made in July 2016.

A number of the assumptions in respect of the other ownership options examined by KPMG are highly questionable. For example, it stated that the design, build, finance, operate, maintain, DBFOM, option would involve significant additional subsidy as a result of the Government being likely to seek step-in rights to operate the network in the event of contract termination. The fact that step-in rights will be built into the current contract with Granahan McCourt calls into question why KPMG assumed the cost of such rights would occur only in the DBFOM option.

In order to fully assess all the assumptions made by KPMG in its ownership report, it is imperative that the financial appraisal underpinning its analysis be published in full. Without such transparency, it is difficult to avoid the suspicion that the assumptions made were engineered to deliver a recommendation for the gap funding model only, especially given the belief at the time that it could achieve an off-balance sheet accounting and budgeting treatment.

A critical juncture in the NBP procurement process was the signing in April 2017 of a commitment deal with Eir to remove 300,000 premises from the intervention area. This development had grave consequences relating to competition for the contract and the cost of subsidy. In terms of subsidy, the original estimated budget for the plan was between €500 million and €1 billion. We know from KPMG’s single bidder solution assessment report published in December 2018 that the estimated subsidy for the plan in April 2017 after the removal of Eir’s 300,000 premises was €787 million. What remains unclear is the exact cause of the increase in the cost of subsidy since then. The Department of Communications, Climate Action and Environment and KPMG have recently stated that when Eir and Granahan McCourt submitted detailed solution submissions in September 2017, both bids included significantly higher levels of subsidy than the Department's budget model. We have not been given any detail as to why the bids submitted were far higher than expected. Since Eir withdrew from the competition it appears that the cost of subsidy to the Exchequer has increased further.

KPMG’s 2018 report highlights how the major contributing factors to the increase in subsidy requested by Granahan McCourt are the decision to move from renting Eir’s wholesale network in the 300,000 premises area to the NBP co-deploying its own fibre in that area, and an increase in the blended equity internal rate of return, IRR, reflecting an assumed higher level of risk. Although Analysys Mason touched on this issue when it appeared before the committee last week, there are still serious questions regarding exactly how much additional cost is being incurred as a result of the decision to roll out a parallel fibre network in the 300,000 premises area and why Granahan McCourt is being allowed to do so. The Department needs to clarify if this is being done purely so that Granahan McCourt can connect back to the metropolitan area networks for its backhaul needs and what the comparative cost of using Eir’s regulated wholesale network and backhaul services would have been. Unless the Department can provide convincing answers for why Eir’s fibre cannot be used, it is difficult to see this as anything other than a very costly duplication of infrastructure paid for by the Irish taxpayer.

I wish to conclude on some issues relating to governance of the NBP. The success or otherwise of the plan will critically depend on the appropriate implementation and governance of the final contract. We have been told that an in-house unit will be established within the Department to monitor the contract and will require a budget of up to €10 million per annum which will cover up to ten permanent civil servants and support by specialist external advisers. However, given the sheer complexity and scale of the contract, it is worth asking how realistic the proposed budget is in terms of ensuring the appropriate level of governance and oversight that will be required.

Outside of the cost implications of establishing a largescale governance and regulatory structure for the plan within the Department, the decision to so do also had a significant impact on the competitive tension for the contract. For example, one of the stated reasons for Eir’s withdrawal from the competition was the fact that it would have to provide access to its network at prices below the regulated level set by ComReg in the intervention area. As a regulated entity, Eir must ensure equivalence all over the country and would therefore be forced to offer access to its network across the country at the same lower intervention area price if it was awarded the contract. Another factor affecting the incentives for Eir to bid for the contract was the restriction that winning the contract would impose on its wholesale market activities. If it won the contract, its wholesale division, Open Eir, would not be allowed to sell wholesale products in the intervention area, whereas other operators would be allowed to offer wholesale products across the country, including the intervention area.

While the logic behind such clear lines of separation is understandable, the fact that Eir is already regulated by ComReg across the country but the Department has proposed setting up a separate regulatory structure for the intervention area with conflicting requirements for access prices, for example, made it very difficult for Eir to continue with its bid. This again calls into question how KPMG and the Department could have assumed that there would be strong competition for the contract from bidders such as Eir that would drive down the cost of subsidy to the Government. Our briefing note raises a number of other concerns that we have with the procurement process to date. Professor Reeves and I thank the committee for the opportunity to present our observations and we would be happy to take any questions that members may have.

I thank the witnesses for their presentations. They both have significant experience with public private partnerships at an economic and academic level and we value their input to our analysis. Will they tell me about their experience in the telecoms market as opposed to other areas?

Professor Eoin Reeves

If the question is whether we are specialists in the telecoms area, I would not consider that to be a fair description. With regard to our experience, we have written a book about the privatisation experienced in Ireland which was largely centred on the story of Eircom. That was published in 2011. We subsequently published peer-reviewed papers on telecoms policy arising from the privatisation of Eircom which prompted much of our interest in this area. We have a degree of experience in the telecoms sector specifically but it is confined to that.

Does the same apply to Dr. Palcic?

Dr. Dónal Palcic

We do not have telecoms engineering backgrounds but we have extensive experience of looking at the impact of privatisation on telecoms industries across Europe, including regulatory impacts. We have published different peer-reviewed papers and, as Professor Reeves mentioned, a book on privatisation in Ireland. That largely related to Eircom and the telecoms sector in Ireland. From that point of view, with regard to policy and ownership-level regulation, I would say that we have a good bit of experience.

I thank Professor Reeves and Dr. Palcic for attending. The documents and executive summaries they have provided set out a very clear picture. There is very little further to question them on. They have made it clear, using their academic position and experience in procurement, that while the process that was undertaken may have started out as a sound endeavour and there was a certain context to the gap funding model, it effectively evaporated as soon as the other potential bidders pulled out. We have been concerned about that for some time in the committee. From a policy perspective, it is clear that the tender process and consultative dialogue process should have halted at that point and we would not be in the mess we are in if it had halted at that stage. It seems that if one follows the process and the position of the Department and Government, they just did not want this to fail. It could not fail because it was too important even when all the fundamentals that underpinned the particular process had changed. The top line issue of getting a broadband service to consumers in the intervention area was such that they had to keep going, which leaves us in a difficult position.

Professor Reeves spoke about opportunistic behaviour and the potential for such behaviour when there is no competition in the process. That is certainly something that we identified here and the witnesses identified it with regard to the parallel building of the network. Based on their examination of the information that has been published, do the witnesses have any idea what the additional cost of building the network out over the existing network, rather than splicing into it, would have been?

With regard to state aid rules, I acknowledge that the witnesses may not have expertise in this area. Are they concerned that the additional taxpayer-funded network being rolled out in the area with 300,000 premises may not comply with state aid rules? Do they believe that a contractual agreement can be put in place that will prevent the final bidder from using its network in that area?

I will leave it at that because the witnesses have given an exceptionally comprehensive and useful presentation.

Dr. Dónal Palcic

On the additional cost of rolling out their own fibre, we are dealing with heavily redacted documents and that is all that we can effectively analyse. We have no idea what the additional cost will be given that it is always redacted but it is clear that it is much higher than the Department's budget model, which assumed that it would be renting Eir's wholesale fibre product. If one reads through Analysys Mason's technical assessment report and KPMG's single bidder solution assessment report from 2018, they highlight many other areas where there is an increased cost as a result of a redesigned network and redesigned solution being put forward by Granahan McCourt. There is obviously a substantial additional cost and the extra fibre that will have to be laid will increase the timeframe. That is partly why we have gone from a five year to a seven year roll-out. There are other issues relating to the technology that has been chosen. It is using next generation technology for optical equipment, which is different from what the Department had assumed would be the case, whereby the current technology would be refreshed in five to seven years as the industry evolves.

Will Dr. Palcic address that issue? One of the reasons given for using the gap funding model was that the benefit to the State would be that the ultimate winner would continue to invest in the technology. Dr. Palcic is saying that because there is no competition, Granahan McCourt is getting the latest generation of technology right now, funded by the taxpayer, rather than being funded from within the business.

Dr. Dónal Palcic

Government subsidies can only cover a certain amount of capital expenditure. It seems that to get the up-front capital expenditure, Granahan McCourt is putting in more advanced technology now, which means that the refresh will happen later in the 2020s. Considering that such low take-up is projected in the early years, gold-plating the technology that it is putting in right now when it is assuming very low take-up would, according to Analysys Mason, represent a 25% additional cost compared with putting in the current technology when it is up and running with take-up increasing in the future. It could then do the typical refresh that any other Internet service provider would do in the mid-2020s and fund that from user revenues rather than from subsidised capital expenditure. If there was more competitive tension in the process, the company may have been forced to go with the lower cost capital expenditure to try to win the bid.

Professor Eoin Reeves

I suggest that the fact that there has not been competition and we have reached the last leg of the process has significantly increased the scope for the preferred bidder to bargain to its own advantage due to the lack of competitive tension. This is exactly what our fear was when we first learned that the process was down to one bidder. That is exactly how it is playing out.

The Department's defence has been that it knew the costs anyway because there was a comparison available based on a submission made by Eir at an earlier stage in the process. The witnesses are clear that the competitive tension would have arisen in the latter stages of the process and usually as part of the final bid price, but that did not happen.

Professor Eoin Reeves

Exactly. One must also take into account that while the Department appears to have a good handle on the costs, the other side of the equation is the revenues. It was in the mix between the cost and the revenues that the competitive tension would play out. That is where one would expect to get the advantages that underpin the gap funding model. Unfortunately, that did not happen.

If there was greater competitive tension closer to the final point, the level of investment by the company concerned would be greater because that could be scored against.

Professor Eoin Reeves

Exactly.

Dr. Dónal Palcic

Even Analysys Mason pointed out that Granahan McCourt was being far more conservative than the Department in estimating future demand. That allowed it to lower its projected revenues. It has been increasing its projected cost by making questionable decisions that have increased the cost beyond what it maybe should be.

It has increased its projected costs by taking quite questionable decisions that have raised the costs perhaps beyond what they should be. It means it needs a higher subsidy and there is less strategic value. All that has forced up the subsidy, which happened after Eir pulled out.

Professor Reeves stated he was concentrating on the financial risk but I turn to the complex area of public private partnerships, PPPs, because he is accustomed to examining them. When one speaks to Department officials and others, they say we do not understand that the developer in this case is taking on significant revenue risk. I do not know whether our guests have assessed that. It is difficult because we are all working with blacked-out sections of documents. Have they examined it in any way to assess where the risk resides?

Professor Eoin Reeves

No, we have not examined it in detail, for the reason the Deputy outlined, namely, that we simply do not have the information. It is fair to say we have been told, and provided with a specimen contract to the effect that if the revenues are higher than projected, there will be a clawback. In that sense, the developer takes on risk but it also takes on the downside risk. If the revenues do not accrue, the developer will be penalised within the contract on that basis, which is important.

Nevertheless, we need to take into account a few general aspects of risk, of which there is a great deal. At the initial stages, I assume that the developer ran risk workshops to identify at least ten or 20 risks, worked out the probabilities of them occurring and placed values on them. We mentioned construction operating revenue but there are other risks. An important factor in risk transfer is that it is central to the value-for-money argument in favour of PPP. What is in the contract is not necessarily what materialises when the contract is up and running and managed. An interesting lesson emerged from the schools PPPs, where it was identified that in the first three years of the schools' running, the management of the contracts were such that they were not really up to scratch and risk transfer was not imposed in reality, that is, with penalties imposed for poor performance and so on.

That brings us to the question of how the contracts will be governed. In the case of schools' PPPs, the National Development Finance Agency, NDFA, took over the management of the contracts. I am curious about where the NDFA has been in respect of broadband. It is a specialist PPP procurement agency. The PPPs that have caused the most problems in the country are those that were not procured by the NDFA but rather by individual Departments, such as that of Department of Education and Skills in the case of the schools projects. The Department of Communications, Climate Action and Environment procured the incinerator project and, now, the national broadband plan. The question is where the NDFA is. I acknowledge that it played an advisory role but in the overall governance plans for what happened in respect of appraisal and so on, it is a question that needs to be asked and should certainly be addressed.

I recall attending a meeting of the Committee on Budgetary Oversight on the issue of PPPs. I cannot remember whether representatives from the Department, the Irish Fiscal Advisory Council or others made the point that one of the downsides of PPPs is that much higher costs for money must be paid when compared with when the State funds a project. That is on the basis that there would be efforts in capital markets to issue a bond or raise lending, which, as Professor Reeves noted, would be a more standard PPP model. In this instance, it is probably even yet more expensive because a return on equity will be the payment rather than a return on capital. It means we will cover the capital costs. We do not know the figure for the return on investment but it is likely to be even higher than what it would be if it was a more standard PPP project.

Professor Eoin Reeves

It is quite hard to work that out. Given that such a small amount of equity - €220 million, or less than 1% of the overall project costs - will be provided, the rest of it will be financed by the Exchequer. In that regard, it will not suffer from the cost disadvantage that would previously have applied in the case of PPPs, where they were entirely privately financed, usually at a equity-to-debt ratio of 1:9. Those PPPs probably had a higher cost.

As I outlined, the concern is that the degree of equity is so low that it will pull the rug from under the mechanisms that should drive the advantages of a PPP model, such as a gap funding model. Developers just do not have enough skin in the game.

One of the variations is that the cost of the debt is not so great an issue. In a sense, the ultimate risk for the State is that if the problem goes wrong, it will carry the large capital-cost risk. If there was at least a private financing element, the banks or their bondholders would carry some of the risk cost, in the event of it terminating.

Professor Eoin Reeves

I agree. When Carillion went bust in January 2018, the banks and the private sector took the risk. At the time there was much panic because the schools were not opening, which showed that the ultimate risk taker is always the service user. When Carillion went into bankruptcy, the Government and the NDFA were managing the project and were in a position to stand back and let it be resolved within the special purpose vehicle. Eventually, a cost was paid by the service user but that will not apply in this case.

Professor Reeves stated it was his "considered view that the current procurement should be terminated and more affordable alternatives explored." In many ways, that is similar to what the Department of Public Expenditure and Reform advocates. Does Professor Reeves have any suggestions as to how more affordable alternatives might work? What might be the timescales, and what would be the downsides?

Professor Eoin Reeves

It is difficult to say, given the lack of information we have and where our expertise lies. I have read what the Department of Public Expenditure and Reform has proposed as a "plan Z". To a large degree, I can understand where it is coming from. Two words in the proposal that struck me are "budget led". In other words, it will be affordable. The Department has suggested that the large escalation in the costs of the project mean there will be an opportunity cost. Other projects within the national development plan will have to be forgone. That should not happen under the more modest concerns with the plan the Department has put forward.

On whether my colleague or I have specific plans or proposals, I cannot come up with something better than the Department. Dr. Palcic might like to contribute.

Dr. Dónal Palcic

Short of there being a retendering and a more public auction, there is not really any other alternative and going back to the market and so on would come at a time cost. The Department of Public Expenditure and Reform's proposal seems to be the more realistic and would allow for the time to retender a more State-led approach.

On the revenue risk, as I mentioned, in the report by Analysys Mason, it appears that Granahan McCourt is being very conservative in what it expects the demand to be, especially in the early years. It will, of course, have to bear some of the risk in respect of how many people it connects as it rolls out broadband, meets its various deployment targets and receives its subsidy payments. As well as a higher internal rate of return, it is building into its cost model many different margins in respect of labour, the capital expenditure and the various aspects that subcontractors will carry out on its behalf. The Analysys Mason and KPMG reports, however, state such margins will impose significant additional costs. Granahan McCourt is covering itself as much as possible by taking on the risk.

I apologise for asking a personal question but it is a small country. Is Professor Reeves related to Mr. Tom Reeves, the former Commissioner for Energy Regulation? They have a striking resemblance.

Professor Eoin Reeves

I can confirm we are not related.

I welcome our guests and thank them for taking the time to attend. My concerns relate to the cost to the taxpayer; the risk to the public finances; the total absence of competition, given that there is now only one bidder; the fact that the owner will ultimately be the investor; the subsidy and the 340,000 households and premises to be connected by Eir; and the governance.

What is set out in the plan for National Broadband Ireland is a structure whereby one person will be representing the investor who puts in 95% of the cash upfront, which is the taxpayer, and there will be eight representing the outfit that puts in 5%.

In light of his comments that he has extensive experience of PPPs - this project is a PPP of sorts - has Dr. Palcic ever seen a PPP such as this and was it a success? Dr. Palcic also referred to the NDFA and stated that it would normally be involved in major projects. Ministers have indicated time and again in the Dáil and in these rooms that this is the biggest tendering process ever undertaken by the State. It is an absolutely massive project and the scale has increased in the context of costs. Is there any reason the Government would not hand this over to the NDFA? According to replies I received to parliamentary questions, up to 80 people have been working on this in the Department for a number of years. I suspect that there are bale loads of information, such as tender, legal and procurement documents, and that the entire process is bogged down in a quagmire. Can Dr. Palcic think of any reason the project would not go to the NDFA?

Dr. Dónal Palcic

I will take the Deputy's first question about other projects that use the gap-funding approach. KPMG's ownership report looked at other countries and the only other gap-funding model that was used as a case study was the Broadband Delivery UK programme. One of the central criticisms in the KPMG report relates to the lack of competitive tension in that programme in the early stages. It was just down to one bidder every time in terms of lots, and BT won it. The report effectively states that one has to ensure competition in order for the gap-funding model to work. Just last night, Indecon's review of the national broadband scheme, which was carried out in November 2017, was published on the Department's website. The national broadband scheme involved 3, which won the contract to roll out mobile broadband 3G technology to rural areas that otherwise could not get broadband. That scheme used the gap-funding model but on a much smaller scale. The amount involved was approximately €200 million, of which the State provided €80 million through capital expenditure and it received €36 million back from the European Regional Development Fund.

Indecon's recommendations on the gap-funding model are instructive in this regard. Indecon indicates that the calculation of strategic value is central to all of this in terms of the subsidy that is required. One would hope to have the revenues and costs estimated somewhat accurately, and then there is the strategic value that the bidders will attach to the fact that, because of the gap-funding model, the network remains in private ownership at the end of the process. Indecon has stated that it is extremely difficult for Government to come up with a figure when estimating this strategic value and that the only way to ensure that the figure is more accurate is to have strong, competitive tension in the whole process. This review has only been published now. It took a bit of pushing; I submitted a freedom of information request for it last March and there was a delay in the report coming out. Part of this is probably due to that recommendation. Indecon zeroed in on the strategic value and pretty much stated that coming up with the figure in that regard was dubious. The KPMG ownership report assumes the strategic value in its financial appraisal, which has not been published. That drives the recommendation of the gap-funding model. This is the major difference between it and all the other ownership options. KPMG has assumed similar costs and revenues for the public ownership model but states that the difference, along with some of the start-up costs, is the strategic value and the estimate on that. The gap funding model has already been used in the national broadband scheme. The recommendation is that it is all about competition and especially in a telecoms sector where there is a lot of uncertainty regarding market developments, potential competition and technological developments. When all of that complexity is in the mix, one wants a very carefully designed mechanism, which is the recommendation coming from Indecon.

I just have a few questions.

Professor Reeves would like to come in on that point.

Professor Eoin Reeves

If it suits. I want to address the Deputy's other question on the governance and the NDFA.

Yes, absolutely.

I will let Professor Reeves in on that, and then Deputy Stanley can ask his further questions.

Professor Eoin Reeves

In the context of getting the best value from the specialist PPP procurement agency, which is the NDFA, that time has passed unfortunately. I would have recommended that the NDFA be involved in the appraisal, especially the financial appraisal, but instead KPMG was contracted to do it. I would also have suggested that the NDFA should have been involved to a large degree in the procurement of the project. Again, it is too late for that now.

In general, it is a good idea for the governance of infrastructure projects to see separation between the agencies that take care of the different stages of procurement. It does not always happen, even internationally. We are witnessing the fact that the Department does not want to let go of this project. It has invested so much time and so many resources - internally and in hiring outside consultants - and the project has run on for years. There is a huge reluctance, institutionally and at a particular level, to change tack. Unfortunately, this appears to be where matters stand. If I had a blank page and I was to suggest how to improve governance going forward, a good way of doing it would be to separate the roles of the agency that sponsors the project, which is the Department, with the agency that conducts the appraisal, which is the cost-benefit analysis and the financial analyses such as value-for-money assessments. That role would need to be separated again between the agency that procures the project and a separate agency to manage it thereafter. This would avoid the kind of rut the project has gotten into now. This is just a suggestion.

I mentioned the resources that have been expended on this project and I expect it has been a lot of resources. PricewaterhouseCoopers, PwC, made reference to a figure at the committee's meeting last week. A lot of resources were spent on the cost-benefit analysis but, unfortunately, we do not have the details in the public domain in order that we might interrogate that analysis. I find it very difficult to understand how the bottom-line figures have changed so much. It is good that the Department of Public Expenditure and Reform has scrutinised this and held the consultants to account on the assumptions that underpinned some of the cost-benefit analyses. After the Department of Public Expenditure and Reform had consulted, the benefits were reduced in the region of €1 billion, but suddenly the costs went down due to an error that had been recurring throughout the whole process but had not been spotted. When I learned of this, I found it very difficult to understand. It calls into question how, generally speaking, we govern different stages of big procurement projects like this and who we rely on to do so. Not relying on State agencies such as the NDFA with its built up expertise is hard to fathom.

Deputy Stanley will ask all of his questions.

I was just following the pattern of the others. In his submission, Professor Reeves stated that the current procurement process should be terminated and a more affordable alternative should be explored. Would the ESB or a model based on ESB Networks be a viable alternative? I ask this in the spirit of trying to be helpful and to move the matter further along. Given Dr. Palcic's experience of studying PPPs and his broader experience in procurement processes, and considering that we have been told that a lot of the heavy lifting has been done in the context of detailed analysis and so on - there have been 80 people working hard on this so there is a lot of information, experience and learning already built up - generally speaking, do our guests believe that we could complete a competitive tender and procurement process within one year or so? I have a related query that may tie the three pieces together.

The witnesses are familiar with the negotiated procedure without prior public notice that can be used. Could that be used to select the ESB or a similar body to deliver the NBP without a competitive tender?

Professor Eoin Reeves

On the Deputy's final point regarding an appointment without a competitive tender, that would involve serious state aid issues. My understanding is that if the ESB were to be brought in, that would have to be under competitive conditions. On the feasibility of using the ESB, I understand from the deliberations of the committee that there are issues regarding the viability of using the ESB's infrastructure. I will not get into the details, which have previously been discussed. Ideally, the State should lever the network infrastructure of the companies it owns, such as the ESB, Ervia and so on. I wish it were possible to turn back the clock and that the Government had run with the original option to build a national broadband company as stated in the NewERA plan published in the run-up to the 2011 general election. However, that ship has sailed.

On the intervention area, a decision was taken to allow Eir to connect 300,000 easy-to-reach households. It effectively encircled many villages, towns and areas of significant residential development in rural areas. This was of significant concern to me at the time. Some 340,000 homes and premises are now being serviced in those areas and that appears to have derailed the whole procurement process. Is that the reason for the other two bidders pulling out? Is it the reason for the cost of the subsidy to fill that gap increasing from slightly less than €800 million to €3 billion? The gap is €2.95 billion.

Dr. Dónal Palcic

On whether it is the sole factor, SIRO hinted that it is one of the reasons it walked away. As they were probably the most commercially viable premises, would have been connected in any fibre roll-out and would have been the initial revenues on which one would have relied, it definitely had an impact on the cost of subsidy. It removed a big chunk of potential revenue and that was revenue which would have been more quickly gathered in terms of the deployment.

Was it necessary to allow that to happen? A person with a lot of knowledge on these matters stated at a meeting in the past fortnight that it was comparable to saying one was going to build a motorway, but that a section of it would have to be privatised, which would not make sense.

Professor Eoin Reeves

Under state aid rules, the Government had no choice. The state aid rules and report for this plan indicate that if the ESB or any other commercial operator has plans to invest in the intervention area within the next two years, it will not qualify for state aid. The Government cannot provide any other State-subsidised investment in that area. When Eir progressed its plans, there was no choice but to sign a commitment deal. In summer 2015, before a contract notice was issued, Eir stated that when it was finished its urban investment roll-out programme, it would move on to the 300,000 premises. It seems to me from speaking to various people that there may have been a calculation by the Department that that may have been a bluff or some way of Eir trying to insert itself into that position. The Department called Eir's bluff by going ahead with the full tender. Once Eir followed through on its plans and talked to the Directorate-General for Competition or the European Commission, there was simply no choice as a commitment deal must be signed if it is to invest as a private operator in the intervention area.

I welcome the witnesses and thank them for the briefing document they provided. In the document and their presentations, they put considerable focus on ownership of the network, which is reasonable. In their extensive submission they refer to schools and road projects and use them as a kind of benchmark. In their analysis, did they apply the same assumptions to the fibre broadband network as they might to a school or hospital? Breaking this down, what lifespan does fibre wire have? How does that compare to the lifespan of a school or hospital? In that regard, is it a valid comparison? Where is the value of the comparison? What is the value of fibre when it sits on rented Eir poles? These are very pertinent, real and practical questions to which the taxpayers would like answers. What value will the fibre have after 25 years, including 25 Irish winters? What would the depreciated value of this fibre and communications infrastructure be after 25 years? In other words, what end product will we have after 25 years? I am not referring to the KPMG analysis of its other values, but, rather, to the physical value of the infrastructure. To what extent will reinvestment in the network be necessary after 25 years, keeping in mind the lifespan of the infrastructure? What level of reinvestment will be necessary for the following ten years? That is a key question. To continue the analogy, although some maintenance work might be required in a small school or hospital after 25 years, much more significant replacement of fibre and telecommunications infrastructure could be required. The comparison does not fully stand up in that regard.

The Minister stated that he estimates the turnover of National Broadband Ireland will reach a peak of €150 million. That is one tenth of the current turnover of Eir. KPMG suggested that its peak turnover may be less than €150 million. Is this a valuable asset for the State to own after 25 years? It is a great popular and populist concept to get out there and people identify with this kind of thing and believe we may miss out on something of great benefit to the State in 25 years. I would like a little quantification of that in the context of the figures put forward by the Minister.

The Senator asked a lot of questions. Have the witnesses taken note of them all?

I will repeat them if that is necessary. With the permission of the Chair, I will finish asking my questions. If the Chair would prefer, I will ask them one by one. I was assuming that the witnesses have a high level of capacity.

Professor Eoin Reeves

No pressure.

Have they considered the success or otherwise of the gap funded model used for broadband in the UK and Scotland? Have they looked at how that model worked there? That would be a valid comparative study. In fact, it would be the most valid comparison. A previous witness told the committee that in the UK, BT initially assessed a 20% take-up in the early years of the service, but the actual take-up was 40%.

This resulted in the Government being able to claw back a total of €700 million. Is a series of clawbacks not a good protection for the State? Does it not outweigh anything we might talk about in the context of a reversion of the asset at that point? Would it not be far more significant for us to be getting that clawback? My belief is intuitive and instinctive, and is based on my own professional background, anecdotal evidence and observation. The demand for the service will increase exponentially. I did a bit of history in the past. Initially there was almost a fear of the electric light and there was a limited take-up. Subsequently, the level of take-up skyrocketed and only total flat-earthers and people living on the margins of society as a result of various social vulnerabilities were left without electricity at the end of the process. I expect a similar process here, although, when saying that, I am gazing into a crystal ball. Based on the UK experience, could we not have optimism in that regard and in respect of the clawback? If our guests want me to repeat any of those questions, I will do so but I think they have them. For each of my questions there is a taxpayer who wants to know the answer.

Professor Eoin Reeves

I will do my best to answer some of the Senator's earlier questions. I am not sure I will answer every single one of them. In our briefing note, we discussed other PPPs like schools, roads and so on. The real reason for bringing them into our discussion was that in those cases, as the Senator knows, ownership is handed over to the State at the end of the contractual period.

With respect, this is the nub of my question. As a layperson, I humbly suggest that it is not a valid comparison by virtue of the longevity of a school or hospital and the nature of their capacity to go on-----

Professor Eoin Reeves

I understand-----

-----and the product over time.

Professor Eoin Reeves

What I have tried to do is explain why we brought them into the statement.

Professor Eoin Reeves

That was simply because the justification for the gap-funding model was that if the ownership stayed with the private sector, that would help drive the subsidy down. When they went through the different ownership options, there were initially five options on the table but the detailed analysis was of just two. One was the DBFOM model and the other was the gap-funding model. The DBFOM model is that which is used for schools and roads and the key difference between the two models is ownership of the asset at the end. The reason-----

With respect, will Professor Reeves come to what that would be?

Professor Eoin Reeves

The reason the gap-funding model was recommended in the KPMG ownership report was that it would increase the value to the bidder, which, in turn, would help drive down the subsidy and the cost. That was the reason the model was chosen. That is why we brought it in, just for the purposes of comparison.

On the question of ownership of network infrastructure, I take on board a number of the points the Senator made. However, I am always mindful of the experience with the Eircom privatisation and the decision made back then to divest full State ownership of vital telecoms infrastructure. We would not be here today if the decision in that regard had not been made. Governments should be extremely careful about giving up vital network infrastructure, be it roads, telecoms or electricity infrastructure. For that reason alone, I would be loath to recommend to Government that it opt for something like the gap-funding model, not only because the justification for it is not very strong, particularly in view of the points that have been made, but also because giving up vital network infrastructure is not necessarily in the strategic interests of any country. That is how I would respond to those two points.

Professor Reeves might come to the subset questions as to the value of the fibre at the end of the process in net terms.

Dr. Dónal Palcic

On the figures they are sending in terms of the value of the fibre, we do not know. They have obviously come up with a residual value. The figure for what the value of this network will be has been redacted. If the residual value of the network is higher in the future, then there will be a clawback to the State. We do not know what is that estimated residual value.

The Senator and Deputy Eamon Ryan asked a series of questions on the lifespan of fibre. I do not think people yet know what will be the full lifespan of fibre. ComReg applied a 20-year lifespan in the context of the depreciation rates. Some industry guys state that it could be 25 or 40 years. At some stage, the fibre will have to be replaced. The optical equipment, the active technology, has to be refreshed-----

On balance, though, would it not be circa 25 to 35 years? There are very few suggesting it would have any value afterwards. Is there any technical suggestion that it would have any value after that?

Dr. Dónal Palcic

There will be value but they have to keep refreshing the optical technology. That develops rapidly enough in terms of the active equipment they use to transmit light down the fibre. That has to be refreshed every seven to ten years at a minimum, although one could sweat it a bit if one wanted. That must be built in. All of these things can be built into a contract in any event.

Will Dr. Palcic just explain that again to me? The bit about the light going into the cable and refreshing every seven to nine years.

Dr. Dónal Palcic

I am not a telecoms engineer so I probably will not explain it very well but I do know that the optical equipment they use to light the fibre-----

To send refracting-----

Dr. Dónal Palcic

-----needs to be refreshed as technology develops. The current technology can handle maybe up to-----

What will be the net value to the State after 25 or 30 years?

Dr. Dónal Palcic

It is very difficult to tell because we do not know how the active equipment, the optical technology, will develop over the next 25 years.

I know, but-----

Dr. Dónal Palcic

Maybe the fibre-----

Does it not become obsolete if we get new technology? What I am trying to ask is if there is any real value.

Dr. Dónal Palcic

The speed of light is fairly fast. Obviously, something may come in the future that surpasses it-----

What about the effect of winter weather on the fibre?

Dr. Dónal Palcic

Again, in the context of current estimates, they still think it could have a lifespan of at least 25 years but even the manufacturers-----

We are back to the 25 years. What I am trying to nail is the populist argument that the State is foregoing some big asset and that it is recklessly letting of something that is valuable. The truth is that after the span and for the subsequent ten years to reach the 35 years, there is nothing in it for the State other than hopefully a clawback from a better take-up than we anticipate.

If I could come in on the same point, there is a slight variation on that.

Perhaps Dr. Palcic can take my question and then we will take the variation.

Dr. Dónal Palcic

Again, there has to be some kind of residual value here. It will not be a worthless asset in 25 or 35 years.

No, but there is nothing to suggest it will be a very valuable asset either.

Dr. Dónal Palcic

It could well be. How can anybody, especially in this sort of-----

It is speculative. I could win the prize bonds, too. It is speculative, basically.

Just to clarify, would our guests accept that the real asset value in this entire endeavour is unfettered access to 542,000 premises and the fact that the company concerned will in essence be paid to build up a pool of customers who will pay, at current rates, €30 a month? Rather than just the cost of the piece of cable or whatever, the value is actually the monopoly access which is being supported to build that pool of customers, which will really drive the future valuation.

The Department's expectation is that in approximately ten years 80% of the 542,000 premises will have signed up to the network. Therefore, that is the value.

It is good that we have nailed the concept that it has some intrinsic value. Assuming the high level of take-up outlined by Deputy Dooley, do the witnesses accept that the clawback will achieve the taxpayers' requirements, as happened in the UK? When Deputy Dooley was out of the room I made that the point that in the UK whereas BT anticipated 20% take-up of the service, it was 40%. In his absence, I was drawing the analogy of rural electrification. I think we are only at the margins in terms of this service in that there will be huge take-up of it.

Do the witnesses accept that the manner in which the contract is structured is such that the greater the loading of investment by the company in a network that it will end up owning after the 20 years, the greater will be the reduction in the clawback to the State? The clawbacks are meant to ensure that the contractor continues to invest in the network that it will ultimately own. While this clawback sounds good, in essence, it was never included to benefit the State. Rather, it is a methodology of ensuring that the contractor does not cream off the profits while the contract is ongoing.

Yes, but the infrastructure will be continually renewed.

Dr. Dónal Palcic

As I understand it, the point being made by Deputy Dooley is that in the rundown to the end of the contract the incentive of the operator to reinvest in the network, knowing it is going to walk away, will be fairly minimal. The operator will not be trying to increase the residual value so that the State can then take a big chunk of that back. I would have thought that the operator would be trying to sweat the asset as much as possible towards the end of the contract.

The operator might also want to have something to sell.

Dr. Dónal Palcic

It will have made all its money back. From my reading of the Department of Public Expenditure and Reform report, the operator will have made its money back in seven or eight years, such that everything made after that is additional to it. On the asset, one of the concerns highlighted in our briefing note is the reliance on Eir's infrastructure in the intervention area. Eir has a major infrastructure programme that it is going to have to implement-----

What is the alternative?

Dr. Dónal Palcic

It has to make ready all the poles and the ducts. This is part of the reason KPMG has now set a seven year roll-out timeframe. The question is what happens if there is any delay? In other words, where does the risk lie and what is Eir's incentive to deliver on that? If it does not deliver or if there is a delay, will buildco, NBPco or the subcontractor be on the hook? Who takes on that risk in terms of not being able to deliver?

We can all pose questions such as what happens if the Ice Age returns or there is a tornado or an earthquake this evening. I am interested in knowing the alternative to using the Eir infrastructure.

Dr. Dónal Palcic

It is good to reuse existing infrastructure. The contractor can use either the ESB's infrastructure or Eir's infrastructure or build its own infrastructure. Obviously, if there is existing infrastructure, it should be utilised.

It should be used and that is what the contractor is doing.

Dr. Dónal Palcic

Down the line it might utilise some of the ESB's infrastructure. Much of this comes down to regulation and the Commission for Regulation of Utilities, CRU. It would have to make a determination on the valuation of the masts to energy versus telecoms.

Dr. Palcic's criticisms of the project are criticisms that anyone could make. I am not suggesting he does not have an expertise in this area - far from it - but his criticisms are ones which the man in the bar or anyone could make. For example, the man in the bar could say this will not work, we should retain ownership of the network because it will be great asset, take-up will be low, the company will exploit the clawback and so on but one could say that about any project. One could similarly say in regard to the construction of a school or hospital that the building will be badly built and so on.

I want to allow Deputy Stanley in at this point.

Senator O'Reilly's comparison with rural electrification is not valid because only one company was supplying electricity, namely, the Electricity Supply Board, ESB. In this case, the difficulty is the risk in terms of take-up because almost three quarters of the population have been already serviced. I was in a townland recently where there are 56 houses, 40 of which have been serviced in the past month or two by a private operator via a radio signal.

I am fascinated with the discussion about the value after 25 years. The Government, the Department and those defending this process have advised us that there are built-in obligations for the technology and the fibre to be kept up to scratch. In terms of comparisons, the motorways are to return to State ownership after 25 years but this asset will not come back to the State. If the motorway is not resurfaced every two years and maintained in terms of drainage there will not be much of a motorway to get back. The difference is that the taxpayer's buy-in was less. I am not happy with some of the motorway contracts, particularly the contract for the N7, which I think is a bit dodgy, not in the sense that anybody set out to do anything wrong but in that its terms were not good for the taxpayer. I will give another example. Various private companies have flipped Eir in the 20 years since it was first privatised. It was sold recently to a French investor, who paid billions of euro for it despite the fact that half of the poles are falling over in the ditches. To say that a telecommunications network or a broadband network would not be valuable at the end of 25 years is ridiculous. Taking what the Government has told us regarding the compulsory mechanisms built into the contract to ensure that the investor and operator will maintain, upgrade and renew cables and so on over the 25 years, this network will be very valuable. When one adds to that the point made by Deputy Dooley that it has the customer base cornered in the 542,000 intervention area this will be a very valuable asset after 25 years. That is my contention.

What is Dr. Palcic's assessment of the clawback? Does he think it has enough teeth and will it work?

Dr. Dónal Palcic

We do not know. There are so many residual clawbacks built in I cannot recall if this one is 60% or 100% or 50%. There is a clawback of anything over the agreed residual value on the contract.

Would that be effective?

Dr. Dónal Palcic

Obviously, it provides for a return. I reiterate the question I posed earlier. If an operator knows that it is going to walk away, would it really want to ensure that the residual value was any higher?

It would hardly want to walk away from a good product. Nobody is going to want to do that with an asset of this value. If the operator does want to walk away, it cannot flip the asset under the contract. Assuming that after a set number of years it could flip it, the company buying it would have a commercial imperative and an imperative to its bank to keep it going. One could make the argument that the people who own the Shelbourne Hotel could walk away some day and run it down in the interim. Surely, they would want the hotel to be in a state which allows them to move it on. The people who buy it will have bankers who make sure they maintain it. I would have thought the same would apply here.

Professor Eoin Reeves

What was at stake at the beginning was a choice between different ownership models, one of which was the design, build, operate finance model used for road projects.

That is the exact same as the gap funding model in so far as the incentives will be built into the contract to make sure that the asset is maintained and at the end of the contractual period it would be handed back in a state that was fit for purpose. The main reason for using the gap model was for a different set of considerations, which was it would drive down the subsidy because it was expected the winning bidder would exploit the advantages of owning the infrastructure in the long term.

Which is not available now.

Professor Eoin Reeves

The models are very similar when it comes to the question of maintenance and the condition of the asset when it is handed back.

Senator O'Reilly has finished his questioning. All his questions have been answered. I call Deputy Kate O'Connell.

To follow on from those questions, there are essentially two schools of thought, namely, if one has an asset up to 25 years that one neglects, one sweats the asset, gets the value out of it and leaves bad cable, or one invests in it and sells something worth value. What the witnesses are saying, or perhaps I am misinterpreting them, is that of those two schools of thought, the first one of sweating the living life out of the asset, getting as much money back from it and then having a defunct cable is the more likely outcome. Is that what they are saying is the more likely outcome?

Dr. Dónal Palcic

Absolutely.

If it is, how can Dr. Palcic say that?

Dr. Dónal Palcic

I am not saying it would be the most likely outcome-----

Of the two outcomes I have given.

Dr. Dónal Palcic

Yes. If as an operator one knew a big refresh would be needed involving new active equipment and perhaps replacing fibre, perhaps that fibre could last 40 years, and that would be after the 35 years are up and perhaps they would be okay on that, but if one knew when one's potential exit date is, I would have thought that one might try to avoid some of those aspects, especially-----

That is Dr. Palcic's personal opinion.

Dr. Dónal Palcic

-----knowing that this gets clawed back.

This is important because it is a subject of major public debate. From the conversations we have had and the extensive questioning, there are two options, one of which is a valuable asset fit for purpose and updated. One of the witnesses mentioned that the technology being put in now is more advanced than day one. Let us say they do all the investment and everything works out great. We will not find faster than light transmission, and we take on board. Fibre optic is the way forward. It just depends on the thickness and the insulation of the cable. Of the two options we discussed today, having a defunct asset or a fabulous gold standard asset, Dr. Palcic is telling me his expert opinion is that we will have a defunct asset at the end of the process.

Dr. Dónal Palcic

I am not saying it will be a defunct asset.

What is Dr. Palcic saying?

Dr. Dónal Palcic

I am saying that-----

I do not find it funny by the way.

Dr. Dónal Palcic

Yes, but I am not saying that. I am saying there is a residual value that will be agreed in this contract. There will be a clawback for anything of value. above all that.

I understand that.

Dr. Dónal Palcic

If one looks at how the gap funding model works, and it came out in the report on the national broadband scheme yesterday, where there is economic value in an asset at the end of the contract, when it is going back to private ownership, the Indecon report indicates that should then lower the amount of subsidy. Accordingly, the higher that economic value is-----

Of the residual asset.

Dr. Dónal Palcic

Yes, at the end of whatever is the term of one's contract. The higher that economic value of that asset is, the lower the State subsidy should be because one is giving ownership over a very valuable asset to the-----

The lower it would be as one goes through the process.

Dr. Dónal Palcic

Yes, but at the end of that contract, if it is a very valuable asset that one has-----

However, we do not know its value. That is my point. I am trying to sum it up. I gave Dr. Palcic two options, one of which was having a defunct asset of a gold standard modern asset. Forget the gap funding model and all of that. He is an expert who was brought in here to inform us. Where in his opinion does he see it ending up? Will it be a defunct asset or-----

Dr. Dónal Palcic

I would never say "defunct".

Okay, worn out. "Sweated" was the word Dr. Palcic used. He said "sweated the asset".

Dr. Dónal Palcic

Everything depends on the future take-up and technological and market developments.

What Dr. Palcic has said is really speculative.

Dr. Dónal Palcic

I would say that in terms of coming up with a residual value, everybody is speculating here.

We must be careful here. With respect, this is a point of order. We have asked Professor Reeves and Dr. Palcic to come in to talk about a procurement process and give us their views on public private partnerships, PPPs, and how they might be better structured.

With respect, this is core to it.

Deputy Dooley-----

There is a line of questioning here that quite frankly would not be tolerated in general terms. I am a little concerned that we are-----

No. In fairness, Deputies and Senators are trying to get clarification and when witnesses come before our committee I think these are valid questions.

They are core questions.

This is a very important issue. We have come in here to try to analyse and this is a robust discussion. I am sure the witnesses do not have any objection or is there any objection to this?

Professor Eoin Reeves

No. I am fine with that.

We need to have robust and challenging questions. We will not know the answer to some of them.

Taxpayers want to know.

I am speaking in generalities. We do not expect the witnesses to have definitive answers but they are here to give their professional opinion where they can and, if they cannot, that is absolutely fine. We all accept that.

I do not mean to be difficult; I am just trying to get to the answers. There has been much toing and froing. Is there is any issue, please correct me.

Professor Eoin Reeves

The Deputy has given a binary choice between one or the other. I do not believe the point was made that there is any guarantee the investor will sweat the asset and run. We did not make that statement. That was not in our opening statement or in our detailed brief. The truth is that we do not know. Nobody knows what will happen in terms of changing technology and the value of the asset at the end of the day. No one ever said that we did know. It is up to the people who are involved in the contract to do their best to try to factor this into the agreement. The problem is that what is in the agreement will be set in stone to a degree. One of the downsides of using a PPP type model is that one loses a bit of flexibility when there is a lock-in to a 25-year contract. That is another aspect we did not mention.

In terms of where this ends up-----

Allow Professor Reeves to finish.

Professor Eoin Reeves

There will be contractual provisions to ensure that the asset is maintained within the period of the contract and that is as far as anybody is looking. They are looking at a 20 to 25-year period. On the question of whether it will be worth anything in 25 years, Deputy O'Connell does not know, I do not know and the Department does not know. We simply cannot answer that question. We have not suggested and nobody else that I am aware of has suggested there is an incentive for the operator to sweat the asset or to run it down. They might have an incentive at some point to exploit the value of the asset and sell it. That happens in PPP contracts.

They deliver an incentive----

Allow Professor Reeves to finish.

Professor Eoin Reeves

That does not necessarily mean there will be issues in terms of the service quality for the users. What it becomes then is a question about who is extracting value and where does the value go? The problem with the Eircom privatisation - and I have referred to it already and exactly what we are talking about here was relevant is that case - was that the true value of the network infrastructure we privatised in 1999 was only realised when it was flipped not once, not twice but three or maybe four times. The question then is, and this is what the Deputy may have detected from some of our points-----

That is where I am going now.

Professor Eoin Reeves

It is a distributional question. The question is who extracted the value. In that case the value was extracted by private equity groups abroad. As I said earlier in that context, we would not otherwise be here today discussing the broadband plan and how we can address the market failure that arose as a result of the botched privatisation in 1999.

That is where I was going next. Professor Reeves compared it to Eircom and explained it very well to us. In the case of Eircom, and Professor Reeves has done a lot of work on it, there were poles and cables attached to them and the ground around them, as such, but that is gone now and we cannot reverse that. I know it is regrettable as we could do with them to attach these cables but we are where we are today. This is a vital infrastructure just like electricity but it is not really comparable to Eircom because what we are dealing with here is the infrastructure as a cable but with Eircom it was the pole and the cable. With Eircom, it was an entire pathway through the country where in this instance this is not comparable.

Professor Eoin Reeves

The Deputy has just described it. That comparison only goes so far.

Yes, it only half of the story.

Professor Eoin Reeves

Literally, that is it exactly.

There is the top half and the bottom half.

Professor Eoin Reeves

It only goes so far - it is poles, cables, ducts.

It is not really a comparison then.

Professor Eoin Reeves

It is comparable to a point but the Deputy has almost answered her own question.

Yes. I was just making sure that I had not misinterpreted anything. Professor Reeves mentioned in his opening statement the duplication of the fibre, but it is not really duplication or perhaps I have it wrong.

Eir will be providing the poles with the cable on it already. Is the proposal to lease it from Eir as an entity rather than just the cable?

Dr. Dónal Palcic

It is overlaying its own fibre over Eir's fibre.

Is its fibre on top of Eir's fibre?

Dr. Dónal Palcic

Yes.

Okay. Will there be any cutting into the fibre? Will it be able to bridge both ends? Will Eir have any control over the overall loop?

Dr. Dónal Palcic

Eir will have its own separate network in the-----

Dr. Dónal Palcic

It will be connecting to its own national network. NBP Core, NBI will be overlaying its fibre in that 300,000 premises area. I think Deputy Dooley asked earlier whether we would have any state aid concerns around that in that it is meant to be passing through the exclusion zone to get to the intervention area and whether there would there be any concerns that it could then try to connect premises within the exclusion zone. Obviously that is technically not allowed in the exclusion zone but the fact the Department had assumed that it would rent Eir's fibre - it signed a commitment deal with Eir to roll out fibre in the 300,000 homes-premises area. I think the Department's budget model then assumed that the winning bidder would effectively rent the fibre used, Eir's wholesale active product, but now it is not. There is a much higher cost to what it is doing.

Initially, it was renting a pole with a cable on it from Eir, now Dr. Palcic is saying is that the pole and the cable is there to service Eir's own people and the company proposal is to layer other cable on top of it.

Dr. Dónal Palcic

Yes.

Is Dr. Palcic suggested that it should have used the existing Eir's network?

Dr. Dónal Palcic

The Department's own budget model assumed that would be the case. Effectively what we want to know, and should be told, is what were the comparative cost of not deploying a parallel fibre network in a fairly extensive area. There is a big timeframe to do that and there is a big cost to do that. That is a lot of capital expenditure. What was the comparative cost of renting or using Eir's active product?

On that point, let us say we have done that and let us say that Eir is sold again, how would that have benefitted the Irish taxpayer if we had gone with that model, which was not the overlaying of cable - in other words, if Eir had control over the connectivity of those 300,000 households and was then sold? How would that play out for the people, if an integral part of the network could be pulled out? I think the option that the company is going with is better in terms of-----

Professor Eoin Reeves

Insofar as I understand the Deputy's question, I do not think it would really matter, because Eir would be sold, the company would have a different name plate, but it would still own the network and still be using it and selling the same products and the prices of those products would be regulated by ComReg.

There is nothing to say that if Eir had been sold and with the position of the 300,000 homes being serviced, that the company that bought it would have had the same priorities. It could have been just a small part of its overall business. Do the witnesses understand the point I am making? The witnesses are assuming that this theoretical company would have had the same model, but if the new company did not have that vision, that potentially would have taken 300,000 homes out of the priority for broadband.

Professor Eoin Reeves

I am confused, as I am not sure what the Deputy is talking about when she talks about Eir taking the 300,000 houses out and whether she is talking about now or in a few years' time.

What if Eir provided to the 300,000 homes with its own cable, and it had nothing to do with NBP people and that it was just a part of the solution? It could reuse the cable it already had and could be part of the whole solution to broadband. What if Eir, which is a private company with control over those 300,0000 homes, had been sold and the new company decided to specialise in electric cars? I am raising questions about the idea that we could have used a chunk of a private company's network. Would that not be high risk for the State?

Dr. Dónal Palcic

It is a regulated entity in terms of its wholesale services. Anyone else who comes in and takes ownership of it would be regulated in the same way. All we are questioning is-----

Somebody purchasing this company would be similarly regulated.

Will Senator O'Reilly allow Dr. Palcic to finish?

Dr. Dónal Palcic

All we are questioning is the comparative cost, that is, the cost the Department assumed of renting Eir's fibre in the 300,000 premises area as compared to the additional cost of laying fibre. We have been told it is a significant additional cost to the winning bidder rolling out its own fibre, overlaying Eir's existing fibre in the 300,000 premises area. That is just a point of comparison that we are looking for and to see what justifies it. Why would Eir's product not be dependable? I cannot remember which company was here at the time - perhaps it was Analysys Mason - but it said something about a view that Eir's product would not be dependable. That company did not go into significant detail, but what are the precise reasons that Eir's fibre would not be good enough and that Granahan McCourt need to overlay that fibre with its own fibre in the 300,000 premises exclusion area?

Professor Eoin Reeves

This is the big question that needs to be answered. The big issue for all of us is to try to understand why the subsidy goes from €787 million to that figure, plus another €1 billion. It appears, and Analysys Mason mentioned this last week when it appeared before of the committee and the penny dropped with a lot of us, that this is what is going on. One of the reasons that the cost is going up is that Granahan McCourt has decided to overlay its own cable on the Eir network within the 300,000 premises exclusion area. This goes back to the Deputy's original question. That is what Granahan McCourt is actually doing.

Professor Eoin Reeves

We, as independent commentators, have our eyebrows raised and wonder why this is happening. Does it make sense to duplicate the existing network? We do not know.

I totally get that.

Professor Eoin Reeves

We do not know. When Analysys Mason was before the committee last week, it could have been probed more deeply to try to understand why it does not have confidence in Eir's products. Again, we do not know the answer, but what we do know is that it seems to have been a major driver of the huge increase in the subsidy that the taxpayer will pay.

The point that I make is that obviously duplicating an overlaying of cable would have had a cost impact but the whole point is to roll out broadband to all of Ireland. If one did not overlay fibre on the Eir poles, one was at risk of leaving 300,000 families that were in the intervention area.

Professor Eoin Reeves

That does not appear to have been the assumption of the Department at any stage, but I would imagine, and the point I made in my opening statement is that this would have happened once we got down to one bidder. For me this is a suspicion, because we do not have full facts. When one is down to one bidder, then the bidder is in a very strong position and can say that it will not go with the original plan, which is to use the poles and ducts, and that it will overlay its own cable. There is an advantage to doing that, because the subsidy is based on the capital expenditure. The company gets a bigger subsidy and builds a much higher quality network for itself. We do not know what will happen down the line, but Granahan McCourt will be able to exploit that value for a longer period. This is the problem with having a procurement with no competition.

Senator O'Reilly and Deputies Stanley and Dooley want to come in. I wanted to keep the thread of thought going and allow Deputy O'Connell finish her questions.

I will come back in later. Deputy O'Connell asked the question.

I have two related questions. Eir was sold for a significant sum, and the figure was €4 billion was mentioned at one stage. Perhaps the witnesses will be able clarify that. The Eir network had very little investment in it over the 20 year period, or in the nearly two decades since privatisation. Very little cable was replaced. The cable has been exposed to the severe weather in the winters since then. Eir has what the witnesses would call distribution and what I would call connectivity and a connection and an infrastructure value despite the fact that it needs a lot of upgrading and repair. Will the witnesses remind us of the price achieved when Eircom was sold?

When Eir exited the competition, I believe it had what it needed at that stage. It got 300,000 households and premises that were easy to reach and it then said, thank you very much.

The real issue is the overlay. The whole thing is bananas, that one was going overlaying fibre already there, which was going to happen. I said it would act like Dick Turpin on the highway and say, "You won't pass us unless you pay". To combat that, one must go around it or overlay another system on its one.

The real issue Deputy O'Connell missed is that the taxpayer must pay a subsidy for each pole, each year and for each metre of cable that the fibre runs through. That is the real issue because we are going with the model of using one private company to use the network of another private company and Joe and Mary Soap will subsidise the whole thing, which completely ties into the cost escalation of €0.8 billion up to what it is now. The other factor in driving up the cost is the fact that we are now in a situation where €1 billion of €3 billion of taxpayers' money is going to be handed over and it has to be handed to a separate private company by the investor to allow it to use its infrastructure. Is that not the real issue plus the overlay in the Eir intervention and controlled area? Is that not the reason for driving up the cost from an estimated €0.5 billion to €1 billion to a bill of €3 billion for the taxpayer? We must pay for each pole and each metre of ducting must be subsidised. It is not as if Eir is saying to us that we can use it and it is all right to overlay it. The taxpayer must pay for this each year with the crazy scheme that has been designed.

Dr. Dónal Palcic

I cannot remember the price that Eir was sold at. For the most recent sale, the sum of €3.5 billion is in my head for some reason.

Yes. It was a very valuable asset but some cables were 30 years old.

Dr. Dónal Palcic

In fairness, Eir has a substantial wholesale operation business, a retail business and a mobile element. Eir is involved in a lot of other things beyond just a wholesale broadband operator. Eir has invested in recent years in fibre roll-out mainly in urban areas and now in the 300,000 premises. Of course there are parts of rural Ireland where poles are falling down, as the Deputy mentioned, and things like that so there has not been massive investment. Eir is involved in so many other things that the value is a very different situation. Eir took 300,000 premises out of the area, which obviously had an impact in terms of competitive tension, the cost of subsidy, etc. A lot of that subsidy is going to renting the actual poles and ducts.

At a cost of €1 billion.

Dr. Dónal Palcic

I think the additional cost that we are seeing now would have always been there. The additional cost that we are seeing now, once we are down to one bidder, is the overlaying that the Deputy mentioned and that is increasing subsidy even further.

The reason I raised the point is that if the ESB network, a parallel network, was used, and which already carries fibre in some areas, the taxpayer would not be paying rent of €20 per pole on which to hang cables.

Dr. Dónal Palcic

I imagine it would depend on what the CRU would determine the price should be.

What if the ESB or a model attached to the ESB did that?

Dr. Dónal Palcic

Again, there might be an issue around how the ESB does cost separation of its accounts between its energy division and a telecoms division, and what the price the that CRU or ComReg, in conjunction with it, would say should be charged for access to that project. It might have to be brought under regulation.

It would be easier to deal with than a French investor.

There are a lot of unknowns and the witnesses are trying to look at what will happen in 20 or 25 years' time. In terms of the best protection from the perspective of the State, do the witnesses accept that full State ownership is the best way to ensure that the State is protected into the future considering that there is a relatively low level of private capital being invested versus the very significant amounts of money that the State has invested? We are all of the view that burden and risk sharing would ultimately be carried by a third party - the private sector. The State lost the long-term potential. On the basis of the medium term, the risk was elsewhere and, therefore, lower subsidy which made sense. In a nutshell, that has been turned on its head. We have capital worth €175 million coming in and then there is some working capital that will be paid back together with dividends, interest and all of that good stuff. Before we really know the success or otherwise of this venture, based on Mr. Watt's communication, the cost will already have been recouped by the bidder. The risk remains with the State throughout and in return there is no potential upside. Is that not the case?

Professor Eoin Reeves

I will try to answer the question as best as I can. Ideally, the answer should have been given as accurately as possible in the ownership report that was published at the end of 2015. I read the report then and I read it again in preparation for today and I cannot get away from my conclusion that the State-owned model was dismissed too easily. The arguments were not very convincing. As a corollary, a huge degree of faith was put in the PPP approach which, as it turns out, is always more complicated than often anticipated at the beginning. There is a constituency that believes in a more private sector led approach but basic economic theory will tell one that in such complex projects, where there are huge degrees of uncertainty around issues like technology, it is almost a common sense thing. Economics is a common sense subject sometimes but it would suggest that the solution will be closer to the State-owned model and away from the market-based model because, in circumstances like that, markets do not work well when there is huge uncertainty, great complexity and so on. This is exactly what we see being played out. I think the State would be better protected with a State-owned model and that choice should have been made at the beginning.

There is really no financial benefit for going the other way because the State is carrying such a huge chunk of the cost.

On the 300,000 premises, do the witnesses accept that if, as has been suggested, that the notion of overlaying the fibre was to protect the future demands of the 300,000 premises that are already in the intervention area? If any consideration was given to that as a reason for the additional cost it would breach the state aid rules because there is already an incumbent. Therefore, any further provision of taxpayers' money to future-proof the 300,000 premises would be a breach of State aid rules.

Dr. Dónal Palcic

It is fairly clear that it is not allowed in any way to connect to a premises within the exclusion zone. Basically, it is on its way through and nothing else. It is passing through in order to reach the intervention area.

I wish to address the future potential or the future value. The fact that it would have a cable and have an arrangement to go through the area opens up opportunities once one is out of contract in 25 or 30 years' time and, therefore, creates a potential competitive environment against Eir or whoever else is there.

Dr. Dónal Palcic

Yes. Once it is out of contract that could be the case. It is very hard to see. It is the complexity that is key, as Professor Reeves mentioned earlier. In a more State controlled or State-led approach, one is better able to deal with that complexity, having to negotiate contracts for all of these things and arrange PPPs. Even for roads, which is easier infrastructure in that it will be a motorway and there will be a certain value at the end of it, that is still a complex contract and there are a lot of complexities around it in calculating value for money, the PSB, etc. In this type of contract, in a fast moving sector, with technology and competition changing, there is so much extra complexity on top of what we would normally see with a PPP. Earlier people asked what would the economic or residual value. Maybe I would be too glib if I said it would down, and I am not saying it would be. A lot of the value would come down to the competitive process.

The NBS review suggested that the subsidy to be paid would be a mixture of the costs, revenues, strategic value and final economic value. Bidders will have to come up with figures for what they believe that final value will be, what the strategic value of their ownership of the asset will be, what the revenues will be and what the costs will be. The subsidy to the Government will need to be worked out with the complex interplay of these factors in mind. When there is just one bidder, though, and it is not under competitive pressure to be more realistic about future demands and, therefore, revenues, attaching a higher strategic value or placing a higher economic value on the asset so that the subsidy will be reduced, all of that tension is gone. When dealing with this type of gap funding or PPP model, complexity and uncertainty make it extremely difficult.

This will be my concluding question or comment. Given the number of unknowns and variables, including the vagueness, which is reasonable in some respects, around the asset's value after 25 years, the likely uptake, and whether the asset will be stress tested or worn out by the company, is it not too emphatic for Professor Reeves to say that "the procurement approach adopted has not satisfied the criteria necessary to ensure that the key objective of value for money"? How did he reach that conclusion, given the level of unknowns? It is theoretically possible that Professor Reeves could be correct, but there is no evidence to suggest he is. Given that he does not know the detail of many elements, were the comments he put into the public domain not too emphatic?

Professor Eoin Reeves

I will respectfully make the point that I was not emphatic enough. The Senator's question relates to the paragraph on value for money. I have outlined clearly what the drivers of value for money are - competition and risk transfer. If we want to boil all of this down, they are the two elements.

But there could not be competition when the other bidders pulled out of the process.

Professor Eoin Reeves

There is no competition. The point we are trying to make is that the gap funding model will not achieve value for money, as there is no competition. There are also concerns about the level of risk being transferred to the private sector. These are the two drivers of value for money in a model like this one.

Regarding the unknowns, complexities and so on, we make that point in our briefing note. It is Infrastructure 101 that, in such circumstances, the safest option, and probably the most economically advantageous in all respects, is to go for a State-led model. Within that model,-----

Have the witnesses costed that?

Professor Eoin Reeves

Some people find it difficult to agree with the fact that a State-owned enterprise would be far more nimble and flexible than a long-term contract, which is much more rigid. The past 20 years have shown that one of the downsides of the PPP model is that the costs of such contracts tend to escalate because, in circumstances where flexibility is required, the private sector does not provide it and instead sticks with the contract if doing so is to its advantage.

Is that it, Senator?

I will ask a brief question before we abandon this. A great deal of the information has come out. Let us say that the witnesses' hypothesis or proposition is that the State-led model would be better. Have they evaluated that model in terms of the cost inputs required to reach the sector of the community that we are trying to reach, the longevity of maintenance and so on? Have the witnesses undertaken a cost-benefit analysis of it?

Dr. Dónal Palcic

We have not done that, but KPMG and the Department have. Consider KPMG's ownership report. When it examined the public ownership option, it assumed that that option would have more or less the same costs and revenues as all the other options. The key differences were some higher start-up costs, given that this would be a new State enterprise, and the strategic value that KPMG assumed would arise under the gap funding model. That is what KPMG said was the difference. It has access to all of the costs and figures whereas we do not because everything has been redacted.

What did KPMG say would be the difference in start-up costs?

Dr. Dónal Palcic

The differentiator that KPMG highlighted was the initial start-up cost and, critically, the strategic value that it assumed would arise under the gap funding model. In terms of elements that distinguish the DBFOM, or PPP, model from the gap funding model, though, the only thing that KPMG referred to was step-in rights, which are included in the gap funding contract that we have. Without seeing what KPMG's financial appraisal included and the figures used in that, we cannot say for certain what the situation is. Based on what KPMG has written in its document, though, one must question many of the assumptions that it made.

Professor Eoin Reeves

Another assumption that it made was that, under the State-owned model, the absence of competition would be a disadvantage. As it turned out, however, there is no competition in the gap funding model anyway, so how can one argue at this point that the State-owned model would not have the same advantages as proposed for the PPP model?

At the end of the day, it is all speculative.

That is one of the Senator's wiser contributions.

Ignoring the analysis for a moment, full State ownership would require land CPOs with Eir. Essentially, the process would be starting again. How would the witnesses do this? There are poles and cables currently. What does a State-led model entail? Does it involve owning the ground, the poles and the cables?

Dr. Dónal Palcic

The simplest version would be to do exactly what Granahan McCourt is doing, except it would be State owned and utilise existing infrastructure.

Would the witnesses contract-----

Dr. Dónal Palcic

What would make the most sense - people have been discussing this for well over a decade and it has been recommended since I do not know when - would be to bundle together our existing State telecoms assets, including those of Bord Gáis, Irish Rail and so on, and build out from that or utilise existing infrastructure. Granahan McCourt would be utilising the existing infrastructure wherever possible anyway, be that the ESB's or Eir's, and only building new poles or ducts where absolutely necessary because none exist. The key difference is that a State-owned enterprise would be managing, delivering and deploying that process.

Instead of Granahan McCourt doing it, an agency would be set up and incur these costs, which would probably not be too much. That agency would approach Eir in one area, Virgin in another and so on and use their cables.

Dr. Dónal Palcic

It could be Eir, the ESB or any operator or infrastructure owner.

The person at the head is the issue. Does Dr. Palcic believe the State would be more efficient or is that just another assumption? Would it be less efficient?

Dr. Dónal Palcic

The State would be better able to adapt to changing market and technological conditions. When one is locked into detailed and rigid contracts with a 35-year lifespan, there is no flexibility unless that is built into the contract. A State-owned enterprise can, if something gets taken out of or added back into the intervention area, adapt easily and be as flexible as possible.

Professor Reeves mentioned two drivers of value for money. They were competition and something else.

Professor Eoin Reeves

Risk transfer, which is essentially optimal risk sharing. That would involve transferring more risks to the private sector than would occur under a traditional procurement process or a State-owned enterprise. Optimal risk sharing is a very strong advantage of PPP if it is done right, in that one cannot transfer too much or too little. It is a stick-and-carrot situation. It is an incentive mechanism.

The private sector would have skin in the game.

Professor Eoin Reeves

Those are the two main drivers. In their absence, the case for a PPP approach is gone. It is not-----

The opening statement raised a question about the cost-benefit methodology. There can be qualitative and quantitative assessments. Would the cost-benefit analysis that the witnesses would have preferred provide a figure for the future value of remote working? To my mind, that cannot be quantified now. No figure can be put on the person outside Skibbereen being able to sell his or her pottery to the man in America. No figure can be put on the person in Portlaoise who is commuting to Dublin being able to stay three days per week at home. In the witnesses' raw analysis of the sums, have they included a variable, for want of a better word?

Professor Eoin Reeves

Not at all. We have not done a raw analysis at all. PwC was paid hundreds of thousands of euro to do this.

Professor Eoin Reeves

Is that okay? We do not have the information.

I suppose I refer to Professor Reeves' raw analysis of the analysis.

Professor Eoin Reeves

Exactly. In turn, we do not have access to any of the data because that is the way much of this happens. This is one of the big problems with infrastructural policy. By and large, it is problematic because these details tend to be hidden. Why are they hidden? It is a good question. My suspicion is because people do not want to be accountable for decisions that they make. I am glad I made that point.

Going back to the Deputy's point, how does one put numbers on that? This is the art of cost-benefit analysis. It is most inexact because there is much educated guess work. My knowledge of the analysis that was undertaken by PwC was that it was heavily invested in by PwC. PwC would have relied on many previous studies, for instance, in order to come up with a number for something like the value that one could attach to remote working. There is an established body of work around how to put numbers on time and cost savings and I would have a degree of faith in the rigour of the analysis up to that point.

The important point is, because it is imprecise, the key then would be to have it scrutinised by an independent agency of some sort. That is why I mention the fact that the Department of Public Expenditure and Reform, DPER, scrutinised the cost benefit analysis, CBA. As a result of that, many adjustments were made. They identified, for instance, how sensitive the ultimate result of the CBA was to certain assumptions. For me, it highlighted the fact that one can never get this right. One can never get it nailed-on perfectly.

Professor Eoin Reeves

That was demonstrated in this case insofar as the details that are out there allow us to draw a conclusion. However, the key is that it was scrutinised and we are down now to a benefit-cost ratio of 1.3 to 1, but DPER has problems with that.

To summarise it for myself, there are levers here in terms of the qualitative impact of something like this but there are variables taken in by the people who do this analysis to put a value on whatever, for instance, staying in Portlaoise three days a week.

Professor Eoin Reeves

Sure. This is a big field in economics and there are specialists.

No one can be definitely right.

Professor Eoin Reeves

True. One will always be able to shoot holes through the analysis but at the end of the day, I would argue - I teach this stuff - it is better than no analysis at all. When one is talking about investing approximately €3 billion of taxpayers' money, it has to be legitimately supported by a cost-benefit analysis. That happened in this case, but the value of the CBA depends largely on the scrutiny.

It also depends on what lens one is coming at it with. If one is coming at it as an economist looking only at figures one would have one view but if one is coming at it, I would argue, as a Minister trying to deliver high-speed cable across Ireland, one's lens is totally different, wishing no disrespect to the academic lens. Legislators and Ministers have a different job to do. They must balance the sum that the man has done in the office on the cost-benefit analysis with the fact that we want the woman outside Skibbereen to be able to sell her pottery. We are back to the academics of it. All of these variables, first of all, are bespoke to Ireland and one cannot really compare it to get an international perspective. I have looked at it. One cannot compare it to any other jurisdiction and it is bespoke. The variables are arbitrary by which I mean they are not set in stone. It is merely what lens one is looking at it through. Looking at the plan, the lens it is being looked at through is the delivery of broadband to the people of Ireland, not whether it is worth our while to have the woman outside Skibbereen selling the pottery. That is the difference.

Professor Eoin Reeves

We could be here all night discussing it.

Professor Eoin Reeves

I was aware of the fact that when we got into CBA, it could go on for a long time. However, to put it simply, the CBA is not about - Deputy O'Connell did not say this - financial costs.

I understand that.

Professor Eoin Reeves

It is actually social costs and social benefits. The CBA that was used in this case was complete and looked at all those costs and benefits. The tricky part is the benefits. KPMG, as I understood it, looked after the costs - we talked about that - but PwC was commissioned to look specifically at the benefits. We are talking about social benefits here. Identifying them is one big challenge, and one has to make choices. The tricky part is monetising or putting the value to them.

Professor Eoin Reeves

There are established methodologies.

Internationally?

Professor Eoin Reeves

Yes. That is why I say, because it is so inexact, one will only get the full value from it if it is interrogated by independent observers.

Because it is Ireland and it is bespoke, different weighting could be attached to a different one.

Professor Eoin Reeves

Absolutely.

Professor Eoin Reeves

There are a huge number of assumptions-----

A huge number of variables here.

Professor Eoin Reeves

-----across every benefit that is identified and evaluations.

Finally, and I am sorry I am keeping the Chair this late, according to the briefing we got, Eir should not have been required to establish a new company. I found that because I sit on the Committee of Public Accounts and I have been through this previously on a number of occasions. My understanding was Eir was a larger company, Open Eir, and it had all sorts of parts selling other bits and bobs. Would there not have been an issue there if that company, with business to its name other than the roll-out of broadband, was selling phones or whatever? Would the State subsidy, potentially, in a private company situation, be pulled off to pay a pension, etc.?

Dr. Dónal Palcic

The point we were making on that paragraph was not that Eir should not have had to do that.

Dr. Dónal Palcic

The point was that Eir would have to create a wholesale division separate from Open Eir in the intervention area and then, if it is competing in the wholesale market, anyone using its services would have to go to two different companies - its wholesale one outside the intervention area and the one inside the intervention area - and Open Eir would not be able to compete in the intervention area whereas other operators would. What we were basically saying there was, in terms of Eir's incentive to want to win this contract, the complexity that type structure would create for Eir would probably diminish its desire to win that contract whereas they are assuming there will be strong competition from the likes of Eir or SIRO with existing infrastructure that they can leverage, etc. Our point was not that this should not have happened. The clear line of separation is perfectly understandable but it definitely would have affected Eir's incentive, particularly because the Department is setting up a separate governance and regulatory structure for that area. They are setting a different price. ComReg is already regulating Eir outside of that area across the country. The conflict between what the Department might want in this area, what ComReg requires of Eir outside of that area and-----

Is Dr. Palcic saying that within Eir, it could be delivering two different things?

Dr. Dónal Palcic

Probably, we would have two different companies. One of the biggest reasons Eir stated it withdrew - another was the uncertainty around the regulation - was that Eir's access price in the intervention area would have been lower than ComReg's regulated price that it must charge and, through regulatory equivalence, Eir would have to offer that price to everybody.

It was making it unpleasant for Eir.

Dr. Dónal Palcic

It was the conflict in what the Department is doing. That is a big structure that they will have to put in place. They have budgeted €10 million per annum. It is a very big structure that they will have to put in place. If one reads the KPMG document, practically every second line reads "if the contract is implemented appropriately ... if there is appropriate governance on monitoring ... if there is this or if there is that...". It is constantly saying it. It states a number of times that the success of this plan or otherwise will all come down to the monitoring of this contract. That will require a big, well-resourced multi-disciplinary and thorough team within the Department.

They have talked about a longer-term vision of a State agency to do all of this. That is a big duplication of regulatory functions in the telecommunications sector that a highly experienced ComReg is already doing. I would wonder what the cost of all of that will be purely in order to deliver this type of plan and the conflict that there might be, and maybe the overlap, with some of ComReg's existing functions.

I am obsessed with the overlaying on the Eir poles.

Will the Deputy wrap it up?

Somewhere along the line, they made reference to next-generation technology. Is what is going on the overlaying, the next layer of cable, fancier than the Eir cable that is under it? Is it more modern?

Dr. Dónal Palcic

I cannot be sure. I think Eir is using the current gigabit passive optical network, GPON, technology which in the future will probably be refreshed to the next generation.

There is probably a value in putting in the next generation technology. If we had incorporated the not-so-new generation cable in the system, it could have needed an intervention.

Dr. Dónal Palcic

The Department assumed any bidder would try to meet its minimum 13 Mbps download requirement and would do so by using current technology. The main issue is around the refresh rate - for example, if one puts in, say, the current optical equipment and it has to be replaced in seven years. It is much further in the future when there are revenues. Much of that could be funded by commercial revenues. However, there is a much higher cost of putting in the more advanced technology right now. Analysys Mason said that is a much higher cost and a big margin in terms of what it was assuming. If take-up rates are low in the next several years and it will have to be refreshed in ten years time, it is big upfront capital expenditure cost subsided by the State versus the future refresh costs which could be coming out of revenues.

I thank both of the witnesses for attending the committee. It was a robust but worthwhile engagement.

The joint committee adjourned at 7.22 p.m. until 5 p.m. on Tuesday, 25 June 2019.