An Taisce’s consultant Matthew Harley has provided one of the best analyses of the proposed €6bn Metro North.
Metro North Oral Hearing. Submission on behalf of An Taisce. December 2009
Contents
Introduction
My name is Matthew Harley. I am a graduate engineer (UCD) and hold a Master’s degree in economic analysis (TCD). I have 30 plus years experience in the Irish Public Service and the OECD.
Metro North is a very impressive piece of engineering design and those involved in that work are to be congratulated. But can we afford it? Is it best value for money?
I am not opposed to the construction of Metro North but I question the social, economic and environmental case for it. I have seen one reference to a Benefit to Cost Ratio of 1.31 for the project. If it produced the benefits claimed and there was no cheaper option, it would be very attractive.
However, I have not yet seen the evidence for that and most of the little I have seen leads me to have some serious doubts about the project’s viability. I am therefore less impressed by the economic case, for what is the most expensive infrastructure project ever in this State.
My submission is by way of a rebuttal of the socio-economic evidence presented by the RPA and will deal exclusively with socio-economic impacts of the proposal. In that context, it will also address the issue of the relevance of Cost Benefit Analysis to the decision as to the sustainability of this development proposal. I contend that evidence from CBA should be admissible and that the Board should have regard to it, and further that, as CBA is a requirement under Department of Finance Guidelines for all public sector capital projects costing over €30 million, the Board should require such an analysis to be presented by the applicant. [1],[2]
Recently, a number of economists have pointed to the need for CBA in this case. In a submission to this hearing Colm McCarthy says urban rail projects such as the Metro “require careful cost-benefit appraisal in order to ascertain whether they are in the public interest.”[3] Dr Edgar Morgenroth of the ESRI, when asked on RTE by Miriam O’Callaghan on 24 March last how he would advise the Cabinet deciding on the Metro, said he would advise them to re-evaluate it: “I would actually check this project out properly. Is it economically viable or not? This is not done or at least not published.” Professor Sean Barrett of TCD told Morning Ireland on 5 March 2009: “I’m afraid that the Department of Finance has been lax over several recent years in not doing CBA, in not publishing the studies. I would have great doubts for instance about the Stephen’s Green central railway station, the interconnector and the Metro to the airport. We don’t even know the costs, but they could be as high as €6 billion.” See also Dr Barrett’s special article for the ESRI in 2006 dealing with the evaluation of Transport 21 projects.[4] The ESRI authors in their report on the ex-ante evaluation of the investment priorities in National Development Plan 2007-2013, said: “as of today there is no comprehensive cost-benefit analysis using a consistent model of all the latest proposals in Transport 21.” Dr Garret Fitzgerald described this ESRI criticism as “a remarkable indictment of an extraordinary absence of public administration oversight in this whole area.” [5]
CBA and indicators of sustainability
The Department of Finance defines CBA as follows: “The general principle of cost-benefit analysis (CBA) is to assess whether or not the social and economic benefits associated with a project are greater than its social and economic costs.” It is important to note that “social” means that benefits and costs are valued from the point of view of society as a whole, and is not a political reference. This usually means including benefits and costs that a private company would normally ignore, such as the costs of pollution or the benefits of job creation, and means using different prices than those prevailing in the market if they do not reflect real values to society.
The main indicators produced by CBA include measures of Benefits and Costs in Net Present Value (NPV) terms. NPV just means putting all the values into one figure in today’s money terms by discounting costs and benefits arising in future years by an annual percentage factor which is much like an annual interest rate. A promise to pay €1,000 in 5 years time is only worth €784 today because that is what I would have to put in a bank today at 5% (real) interest to get €1,000 (in real value) in five years time. The ratio of Benefits to Costs (BCR) in NPV terms is also a useful indicator. The larger it is the better the project looks. Another indicator is the Internal or Economic Rate of Return (IRR or ERR). We consider we are putting into a bank, all our costs, capital and current, as they occur and we look to see what effective rate of interest we are getting for society when we look at all the stream of benefits we get over the lifetime of the project. The Department of Finance requires that we get at least a 4% return per annum in real terms from our investments. (That rate was reduced from 5% to 4% in May 2007). Another way of looking at it is to say, we need to get at least 4% because we believe we could get at least that real return from an alternative investment.
It is important to note that meeting the minimum 4% return requirement is not a sufficient condition to proceed. Clearly, if the condition is not met the project should be scrapped. But before it is approved we must be sure we have chosen the best option. Alternatives have got to be looked at and subjected to the same analysis. In principle, we should choose the option that gives us the highest return on investment.
CBA as a tool of proper planning and sustainable development.
Why is CBA relevant to planning decisions? Here is explanation from a US Government source[6]:
“The most appropriate method in apprising projects from the national point of view is the cost-benefit analysis. The analysis is the most scientific and useful criterion for project assessment. It helps the planning authority in making correct investment decisions to achieve optimum resource allocation by maximizing the difference between the present value of benefits and cost of a project. It involves the enumeration, comparison, and evaluation of benefits and costs. A memorandum provided by the Federal Government for heads of executive departments and establishments defines cost-benefit analysis as “A systematic quantitative method of assessing the desirability of Government projects or policies when it is important to take a long view of future effects and a broad view of possible side-effects”.”
Sustainable development is fundamental to the planning process. The Planning Act 2000 states its purpose as: “to provide, in the interests of the common good, for proper planning and sustainable development”. But, what is sustainable development?
The National Spatial Strategy, page 13, says that: “Sustainable development is development that meets the needs of this generation without compromising the ability of future generations to meet their needs.[7] The concept captures the important ideas that development
- has economic, social and environmental dimensions which together can contribute to a better quality of life
- will only be sustainable if a balance is achieved between these three dimensions,…”
How can we objectively test for that balance? Our then Taoiseach Bertie Ahern said at the “Towards Sustainable Airport Development Conference” on 23 October 2006:
“One issue that has come increasingly to influence our thinking in recent years is now critical to the decision-making processes as regards all forms of infrastructural investments in all sectors of the economy. It’s the issue of sustainability. The sustainability test of major development projects aims to ensure that the impact on the wider economic and social environment is taken into account before projects go ahead at all.”
The only official objective test of sustainability we have is that of Cost Benefit Analysis, as required under Government guidelines for the appraisal of public capital projects and which, for projects of this size (over €30 million), measures the net benefits and the return on investment of proposed projects.[8] These guidelines are reinforced by similar EU guidelines, under which all our NDP projects are supposed to be evaluated.[9] It would seem obvious that if a proposal is not shown to be a plus on balance between economic, social and environmental factors, it cannot be considered sustainable. Further, the only objective test of such a balance is the indicators produced by a full and independent CBA that also assesses feasible alternatives. It is clear that, if sustainable development is critical to proper planning, an objective measure or test of sustainability would be fundamental to taking a planning decision. Apparently not.
Mr Simon Clear in his evidence points out that planning authorities must have regard to the National Development Plan, which envisages Metro North. He does not refer to Chapter 12 of the NDP, which is devoted to Value for Money considerations and specifies that Department of Finance appraisal guidelines apply to all NDP projects. Does NDP Chapter 12 not apply, according to the RPA? It is clear that meeting the requirements of Chapter 12 is a necessary condition for a project to be considered justified by the NDP as Value for Money is an integral part of the NDP, there to ensure that projects are sustainable in economic, social and environmental terms. If a project does not meet these requirements it is not consistent with the NDP. To presume that a project envisaged in the NDP is automatically in accordance with national policy without its having been assessed as required by the same NDP is to argue that waste of resources is consistent with national policy. I am sure the inspector will consider this Metro proposal in the light of the NDP in its totality.
When the Portmarnock Community Association (PCA) tried to argue this at the oral hearings into the new terminal and new parallel runway at Dublin Airport, which is also an NDP project, we were not heeded. Please note that I am not trying to reopen those hearings here!
It was argued by the Board’s inspectors at those hearings that such socio-economic considerations raised by the PCA were not within the remit of An Bord Pleanála and they refused either to require the DAA to produce such analysis or to accept the PCA’s own analysis as evidence. The runway inspector said:
“….I would conclude that such issues including whether the applicant is in accordance or in breach of Government requirements are not germane to the proper planning and sustainable development of the area……..the issue of financial viability is a matter for the appropriate Minister and the applicant. It is not a matter for adjudication by the planning authority or indeed the Board.”
In the case of Terminal 2, the inspector said:
“This also applies to the cost of the lands, or the cost of the airport to the taxpayer. While they are important, assessment of the proposed development in such areas is a matter for other bodies specifically authorised under other legislation, and outside the scope of this appeal.”
It needs to be said that, contrary to what the inspectors said, neither financial viability nor cost to the taxpayer were the issue, but the economic, social and environmental viability that is measured by CBA. CBA is not a measure of financial viability or of cost to the taxpayer but an assessment of the net benefits to society as a whole, e.g., “the common good” or “the national interest”.
It seems odd that the only recognised test of sustainability, based on a solid methodology, and one required by Government guidelines for the appraisal of public capital projects, is not recognised as an essential piece of evidence for a planning process that aims to ensure sustainable development.
On the other hand, The Irish Times of 31 March 2009 quotes An Bord Pleanála’s refusal of planning permission for the Tinerana development beside Lough Derg, where the inspector said: “The net economic benefit to the county of a grant of permission for such uses on an inappropriate site is negligible”. It is not clear how some Board inspectors can decide that net economic benefit is central to the planning decision while others decide it is not relevant.
While the Board’s inspectors and the Board itself have refused to insist on or accept CBA numerical results, it seems that assertions about economic benefits are acceptable. This was evident in both the T2 and R2 cases and is clearly so in this Metro case. There are repeated assertions about “wider economic benefits”, “job creation”, “vital for the national interest”, etc. There is no quantification or proper socio-economic assessment of any of these claims anywhere in the evidence. They are simply unsubstantiated, unbalanced assertions. It seems “waffle” is allowed but objective analysis is not. Dr Massarsch has repeatedly made the point regarding geotechnical impacts that there was a tendency on the part of the RPA not to give numbers, and that for him, statements such as “not significant” are not enough. That injunction should also apply to claims about significant socio-economic benefits. Quantification is needed if the Board is to be able to test the level of significance claimed.
Here are just a couple of examples:
- “Metro North will deliver significant positive benefits within Dublin City Centre and competing centres within the Metro North corridor.” (From Mr. Gilder’s evidence, which mentions “benefits” twenty-three times, but “costs” not once.)
- “There are potentially great benefits for society, including therefore general health and well being benefits, with the development of the proposed scheme, particularly for people living along the route.” (From the NTS)
It would be hard to devise a project that did not produce economic benefits. The issue is not just benefits, but the extent of net benefits, (after costs).
Given the position adopted by ABP on T2 and R2 and the need for some clarity on this issue, it was of some interest to hear John O’Connor, Chairman of ABP, say in an interview on RTE News at One on 6 November 2008:
“In reviewing development plans they [local planning authorities] have to take account of issues such as climate change, the economic situation, the need to get proper return from investment in infrastructure, the need to protect good agricultural land from further sprawl, these sort of issues; and these sort of issues have to be taken on board in a much more serious way than perhaps they were in the past.”
Mr O’Connor made these remarks in the context of the publication of ABP’s 2007 Annual Report.[10] This seemed to be the sort of clarity we wanted: planners should have regard to getting a proper return on investment from investment in infrastructure. We wrote to him asking how his demand could be reconciled with a position taken by the Board in the R2 and T2 cases that the indicators of rate or return on investment, as produced by a CBA, were not relevant to planning.
The text of the reply is below.[11] It says that Mr O’Connor was talking about “a different issue”. In essence, planners must have regard to the need to get a good return from existing public infrastructure when examining applications for new projects. I.e. they should not allow new development that does not maximise the use of existing public infrastructure.
This is an odd position. If it is legitimate to insist on the need to make good economic use of existing public infrastructure, why does it not follow that when public infrastructure projects are proposed, ab initio, a good return on investment is also needed? Why is a proper return on investment in infrastructure only relevant when considering how later developments make use of that infrastructure, and not relevant when such new public infrastructure is being proposed?
Indeed one can ask, if rate of return is not relevant when large-scale public infrastructure is being decided, why is it relevant when later proposals impacting on that infrastructure are being considered? It seems inescapable that if a proper return on investment is relevant in the interests of sustainable development for subsequent projects, it is relevant at all stages.
Was a proper CBA done for Metro North?
We are not sure what was done. Asked if a CBA had been completed for the Metro, Deputy Brendan Smith, speaking for Minister Noel Dempsey, told the Seanad on 4/10/2007, that the Metro North project “must comply with the Minister for Finance’s capital appraisal and public partnership guidelines. This has been done in the case of metro north.” He also said that the details would only be released when the planning and procurement process was completed. Noel Dempsey said on 6 October 2008 about the Metro, “It will get its usual appraisal, value for money, cost benefit analysis and if it meets those, it will go ahead.” Mr Smith said in 2007 it had been done and Minster Dempsey said in 2008 that it will be done. Which is it?
Martin Cullen writing in the Irish Times on 3 Sept 2006 stated that the economic case for the Metro would have to be made to An Bord Pleanála at an oral hearing. So has a proper CBA been done for the proposal before us? According to the guidelines it has to be done before planning permission is applied for and according to Minister Cullen it has to be made to this hearing. On 9 July 2009 Mr Dempsey told the Dáil: “As soon as the tendering and planning processes have concluded, the cost-benefit appraisal will be carried out. Once it proves positive, which I am certain will be the case, construction will commence.” Note that Minister Dempsey has now stated that the planning process must be finished before the analysis will be done.
At this Oral Hearing Mr Frank Allen of the RPA said:
“RPA has conducted an economic analysis of Metro North, which shows a positive Benefit to Cost Ratio. This analysis has been audited by independent economic advisors to the Joint Committee on Transport. The published report of the advisors to the Oireachtas Committee concluded that the RPA’s assessment of the positive economic case for the Metro tended to underestimate the benefits that will accrue from the project. RPA will update the economic analysis at the conclusion of the tendering process, which will give a reliable estimate of the cost of implementation, but also take account of updated economic and patronage forecasts. It will be on this basis that the Government will make a final decision on the project.”
The only reference in the EIS is to a CBA done by the RPA as part of the route selection. We presume that is the same work Mr Allen referred to. The RPA refused to allow the release of many of the details of its CBA to the Joint Oireachtas Transport Committee on the grounds that it would prejudice RPA’s negotiating position with contractors. RPA would only allow the release of some general results such as the overall Benefit to Cost Ratio. O’Reilly Consultants prepared a report for the Joint Oireachtas Committee in May 2004 with this restricted information. The O’Reilly report cited a Benefit to Cost ratio of 1.31. This is presumably the published report by advisors to the Joint Oireachtas Transport Committee Mr Allen was referring to in his evidence.
The RPA is putting to the Board that its analysis shows a positive Benefit to Cost Ratio but will not provide the details by which that claim can be validated by the Board. Appeals to authority are not valid arguments unless the evidence of that authority is presented. As the RPA admits it does not yet have a “reliable” analysis for the project before us, the RPA does not know if it is viable and cannot claim so here.
Costs
We have limited and confusing information on costs. On 28 May 2003 Mr Padraic White and other senior executives from the RPA gave evidence to the to the Joint Oireachtas Committee on Transport about the Metro.[12] He said that the RPA’s Outline Business Case (OBC) was presented to Government in November 2002. There had been disquiet about a figure of €4.8 billion (or €4.881 billion) that was in the public arena for the City Centre to Dublin Airport Metro option. Mr White explained that this was the total cost in current money terms and not the construction cost of the project. It is an estimate of the “funding requirement”. He further broke it down into a direct capital cost of €1.72 billion, risk and insurance fees of €903 million, a cost escalation component of €811 million and VAT of €458 million. This comes to a total capital cost of €3.892 billion. Interest associated with the PPP arrangement comes to €676 million and financial fees and reserves came to €313 million. That brings the total to the €4.881 billion figure that was in the public domain. It was based on adding up all these figures in current money terms without regard to when they were to be paid. Converted to present value terms Mr White gave a figure of €3.3 billion. Mr White argued that it was the direct capital cost of €1.72 billion that should be used to compare with other systems. What is the appropriate cost figure relevant to CBA? Given the obscure nature of some of the components it is not clear if some of the items are real resource costs. For example, VAT would only be relevant if market price valuation rather than factor cost valuation was used. Interest should not be included.
The RPA prepared a Revised Business Case (RBC), which was presented to Government in June 2003. This presented a lower cost option and may have reduced the funding requirement estimate from €4.881 billion to €4.58 billion. The direct capital costs estimate was reduced from €1.72 to €1.224 billion.
We have examined the O’Reilly Consultants report.[13] It is clear from this examination that the early analysis done by the RPA, as reported by O’Reilly Consultants, is not the proposal before us.
- The analysis reported was for a Metro link from St Stephen’s Green to Dublin Airport, only 10 km long, whereas the proposal before us takes the Metro to Swords and beyond to Belinstown, a distance of 18km. The proposed Metro is therefore almost twice as long.
- Passenger load or patronage assumptions are very different. The proposed Metro is reported to carry 34 (or 40) million passengers per annum when operational. The analysed option was for 18.35 million per annum at the start, about half.
- The report gives some results, which indicate that an extension to Swords would add benefits. That is hardly surprising given that Dublin Airport will only account for up to 20% of the proposed Metro business and most of the population on the Metro corridor lives in Swords. A full analysis of the Stephen’s Green to Swords, or to Belinstown was not done by the RPA.
- The O’Reilly report explores the discrepancies between the RPA’s cost estimate of €4.881 billion in 2003, a Dublin Metro Group estimate of €0.6 billion and Prof. Melis of Madrid’s estimate of €1.05 billion. As Mr White argued, the comparison should rather be with an estimated direct construction cost of €1.72 billion rather than the catchall €4.881 billion figure. The disquiet at the size of the RPA’s figures had led to efforts to reduce costs. O’Reilly Consultants note two sets of reductions that were examined, one amounting to €496 million (fewer stations, plainer stations, no escalators, no substations). A further reduction of €144 million could be made by a switch to mono tube construction from twin-bored tunnels. They say these savings would bring the construction costs down from €1.72 to €1.08 billion. We note that the present proposal reverts to twin bore design for tunnels, so that saving does not apply. Much of that savings must by now have been eroded by the tunnelling options at Dublin Airport and Ballymun, which have now been added. It also looks like the no-frills metro stops have also been dropped.
The proposal analysed by O’Reilly Consultants and for which the very positive result and supposedly underestimated benefits were cited here by the RPA, is a materially different proposal.
Costs and Benefits.
As noted above we do not know what this proposal will cost. Frank McDonald claimed in the Irish Times on 7 August 2007 that he had found a cost figure of €4.58 billion in an FOI document.[14] He argued that the cost must by then have been at least €5 billion. However, it is very likely that this €4.58 billion is a reduced version of the earlier OBC “funding requirement” estimate of €4.881 billion, and is not the appropriate economic cost for evaluation purposes. However, whatever the appropriate figure is, we are looking at a Metro nearly twice as long.
We would estimate (from available information) that the capital costs would be about €3 billion. This is before an allowance for a cost overrun of 45%, the average cost overrun identified by Professor Bern Flyvbjerg from his analysis of 44 urban rail projects. Including the cost overrun of 45%, the probable capital cost comes to €4.4 billion. The NPV of total lifetime costs of the current proposal, including twenty-five years of operation is around €6 billion in 2009 money terms (see below “Sensitivity of CBA results”). Without the cost overrun the NPV of total costs is €4.65 billion. Whatever the total cost in today’s money terms, we have to see benefits of at least the same value to have a Benefit Cost Ratio (BCR) of 1. Where are those benefits? They have to arise predominantly from the value of time saved by commuters. That largely depends on the patronage of the system over its lifetime. We do not have that information but are asked to take it on faith that the NPV of all those benefits exceeds the NPV of costs. That is not good enough, given that we have serious doubts about the validity of the claims made for the patronage of the Metro (see below).
In April I raised the issue of patronage and value of time estimates bilaterally with Mr Rory O’Connor of the RPA. He offered to provide me with the information I needed, as long as it was not confidential. I sent a list of my requirements on 11 May, which was acknowledged. I have received none of the requested data and must therefore rely on deductions made from limited available data.
Outline Business Case (OBC).
An Outline Business Case (OBC) was prepared by Steer Davies Gleave for the RPA and presented to Government in November 2002. A Revised Business Case (RBC) also by Steer Davies Gleave, replaced the Outline Business Case, and was presented to Government in June 2003. This revised report is also frequently referred to as the Outline Business Case, causing some confusion. We understand it is this Revised Business Case that O’Reilly Consultants had access to. While heavily blacked-out versions of the OBC and RBC are available, access is refused to the full versions. There is correspondence between the RPA and O’Reilly Consultants in April 2004, which reveals the conflicts about the release of this information.[15] In the correspondence the RPA confirms the information came from the RBC but would not allow O’Reilly Consultants to publish all of it.
We have examined the earlier OBC document as available on a Department of Transport website.[16] Much of the critical information on costs and benefits in Table 9.1 has been blacked out. Five optional routes were analysed, the longest of which was from Bray to Swords via Blanchardstown and Dublin Airport. Option 3 examined a City Centre to Dublin Airport route and an Option 3a looked at an extension of that route to Swords. The results for Option 3 (stopping at Dublin Airport) show a BCR of 0.49 and a net loss of €1.3 billion at a 5% discount rate. Option 3a to Swords is closest to the option being presented here by the RPA. For that option, the OBC results show a BCR of 0.65 and net loss of €1.1 billion at a 5% discount rate. Losses are lower at the 3.5% discount rate.
These results were presented in the OBC as the “Central Case” analysis. The NDP/CSF Evaluation Unit in the Department of Finance prepared an alternative CBA methodology, which was tested on the OBC. These corrections employed lower values for time saved and assumed a lower growth in the value of time saved in the future. Factor costs rather than market prices were used and a shadow cost of public sector funds of 125%, was applied. Tables 9.6 and 4.1 in Appendix III show the downward effect of these changes. Option 3a (nearest the current proposal) drops to a BCR of 0.48 (from 0.65) and a net loss of €1.7 billion (from -€1.1 billion). Option 3 stopping at Dublin Airport (which later became the Option 3b in the RBC seen by O’Reilly Consultant) falls to a BCR of 0.36 and to a loss of €1.8 billion (from -€1.3 billion).
The OBC concludes: “Options 3, 3a and 4 return a relatively poor economic performance. This is to be expected as they include the most expensive part of the entire Metro network without reaching the range of markets served by Options 1 and 2.” Option 3a was the second worst, only behind the City Centre to Dublin Airport option (Option 3). Under the assumption of demand management (road pricing, etc), and a low discount rate of 3.5% all options do better and Option 3a achieves a BCR of 1.05, but this is without the NDP/CSF corrections. Table 9.7 on page 106 is a hand written addition to the OBC report. The table combines the effects of the demand management arrangements (that tend to enhance Metro options) and the NDP/CSF corrections (that tend to depress them). The combined analysis results are only presented with the 5% discount rate assumption. The overall combined sensitivity results for all options are poor. The best, Option 2, comes in at a BCR of 0.93 while the worst, Option 3, scores a BCR of 0.46. Option 3a, nearest to “our option” comes out at a BCR of 0.62.
The OBC report concludes, “RPA’s view is that, taking the wider benefits into account, the OBC makes a strong case for building a Dublin Metro.” Clearly, they are not referring to the City Centre to Dublin Airport or Swords options. These wider economic benefits include environmental benefits and social inclusion, and city and regional development. None of these wider benefits are quantified. It is hard to see how the RPA reached its conclusion, ever for its best option, the cross-city route from Bray to Swords (Option 2), which had the best Central Case BCR but did not quite breakeven at 0.99. We are now told that Option 3, the Metro from Dublin City Centre to Dublin Airport option, which had the worst result with a combined sensitivity test BCR of 0.46 got to a BCR of 1.31 when revised in the RBC. What has brought about this dramatic improvement?
We then examined the heavily redacted RBC which is also on the Transport 21 website.[17] Table 5.1 gives NPVs of costs and benefits in 2002 prices for a revised Option 3, but again separate figures for costs and benefits are blanked out. The BCR is 1.31, NPV of Net Benefits is €590 million. Results are also given for a 3.5% discount rate. It also reports that using corrections requested by the Department of Finance NDP/CSF Evaluation Unit, the BCR falls to 1.01. No revised results are presented for the OBC Option 3a to Swords. The dramatic improvement in results is explained in the text by a number of factors. One is the reduction in capital costs (figure blanked out). We gather from the O’Reilly report that the direct construction costs might have been reduced from €1.72 billion to €1.224 billion without the reduction for opting for single bore tunnels, as that saving has been abandoned.
Some of the improvement is attributed to an updated transport model. Higher patronage due to the new alignment now proposed is the third explanation (page 5). Page 20 elaborates further on the updated model:
“Work has continued on the RPA’s model since November [2002] and the revised model reflects stronger than expected demand in the forecast year (2016) for public transport and a more congested transport network in the do minimum situation. This means that relative time saving benefits have increased at a greater rate than the patronage. ”
We will argue later that much of the case for the Metro rests on very optimistic patronage projections based on extrapolations from trends during the boom years. It looks as if the RPA’s model was updated to incorporate those trends. In the light of the recent severe downturn we have to question the validity of that model to this EIA. It is particularly important given the dramatic improvement in the BCR figure from 0.49 to 1.31. We remain unconvinced by the dramatic improvement in results, which we are required to take on faith.
Metro North Patronage
The benefits from Metro North flow predominantly from the increased welfare of users, largely determined by travel time saved. Patronage or demand assumptions are therefore key to determining the benefits of the project and hence its justification if those benefits exceed costs. We have been given very little on Metro North demand/patronage assumptions. We are told in the EIS: “Annual patronage (total journeys) is estimate to be 34 million.” No date given, but is assumed to be at opening. The NTS says 35 million at start. Mr David King in his evidence said: “The demand forecasts indicate that the introduction of Metro North will produce in excess of 40 million trips per annum in 2016.” There has therefore been a substantial increase in the estimate since the EIS.
Mr King also provided evidence of activity in chart form for peak hour at the various stops on the Metro in 2016 and 2040.[18] From this it can be deduced that total peak hour passengers in 2016 in both directions will be about 25,500 assuming all Transport 21 projects are completed. This rises to 80,000 in 2040, “based on all growth assumptions being realised.” If the rise in annual demand is proportionate, this means that annual demand will rise by a factor of over three from 40 million at opening to 125 million in 2040. That is more again than the assumed doubling of population growth in the MNEC, which is highly questionable. Where are all these users to come from?
Note that the projections assume all Transport 21 projects are completed. This cannot be assumed. There must be some doubt about the DART/Suburban Rail interchange at Saint Stephen’s Green, and even more doubt about the Metro West link at Dardistown.[19] We have not been given the information to determine how much they contribute to project Metro North usage. A rough indication based on Mr King’s charts and assuming half the St Stephen’s Green Metro North traffic is due to the DART link, would suggest 10% to 15% of total Metro North traffic relates to DART and Metro West. If these projects do not go ahead, benefits would drop by such a percentage. The forecasts also assume demand management (including cordon charging) will be introduced, encouraging much greater public transport uptake. Its introduction cannot be taken for granted.
We have no evidence to suggest that the economic growth assumptions were reasonable, as we have not been told what they were. Given the drastic recent slowdown and the strong likelihood that the future will be one of lower growth than during the boom years in which this analysis was done, we have grave doubts. Cost Benefit estimates should not be based on what is admitted to be the most favourable scenario with all assumptions about growth and other infrastructure assumed. We do not know what user demand assumptions were built into the RBC that produced the 1.31 BCR result for the short City Centre to Dublin Airport option but expect they were based on a similarly optimistic future. We would have serious doubts if user benefits were projected to grow by a factor of three.
Reverse-engineering of OBC and RBC CBA results.
In order to examine the dramatic improvement in the City Centre to Dublin Airport option, we reverse-engineered the available OBC and RBC results.
Although many figures are blacked out in the OBC, Benefit to Cost Ratios and Benefits less Costs at two discount rates (5% and 3.5%) are given in NPV terms for each option. Using this information it is possible to reverse-engineer the figures to provide some absolute benefit and cost figures.[20] These deductions about total costs and total benefits are accurate (within rounding errors) because they follow directly from the information provided. Further, making some simplifying assumptions about the timing of capital costs, operating costs and user benefits it is possible to make some deductions about them also. These are less certain but must be reasonable accurate given the information provided.
Applying this to Option 3 we get capital costs of €375 million per annum for each of the first five years of construction: a total of €1.92 billion (€1.67 billion when discounted at 5%). Operating costs are about €85 million per year for each of the following 25 years. In a similar fashion we can deduce annual benefits. If these were flat over the period they would be €114 million per annum. If we assume that the value of time saved will grow in real terms at 2% per annum and that population will double in the MNEC in 30 years, as the RPA seems to have assumed, we get benefits of €55 million in year 6 and growing at 6.7% per annum to €260 million by year 30. For this option the total costs are €2.56 billion discounting at 5%, benefits are €1.26 billion and there is a net loss of €1.3 billion. The BCR (at a 5% discount rate) is 0.49, as it was reported in the OBC. The IRR (not reported in the OBC) is -1.8%, well below the Department of Finance then minimum rate of return of 5% and the current minimum rate of 4%.
We did the same for the revised Option 3 in the RBC Table 5.1. We get capital costs of €333 million per annum for each of the five years of construction: a total of €1.67 billion (€1.44 billion when discounted at 5%). Operating costs drop to €42 million per year from €85 million for each of the following 25 years. Flat annual benefits jump to €226 million per annum, compared to €74 million. If assumed to rise with the value of time and doubling of patronage, we get benefits of €110 million in year 6 rising to €520 million by year 30. Total costs are €1.91 billion discounting at 5%. Total benefits are €2.5 billion and the net gain is €600 million. The BCR (at a 5% discount rate) is 1.31 as reported in the RBC. The IRR is 7.3%. These are dramatic changes by any account.
Table 1: Improvement in City Centre to Dublin Airport Option from OBC to RBC.
| OBC Option 3 | RBC Option 3b | |
| CAPEX per annum yrs 1-5 | €375 m | €333 m |
| NPV CAPEX (5%) | €1.67 b | €1.46 b |
| Operating Costs p.a. | €85 m | €40 m |
| Benefits p.a. | €114 m | €226 m |
| NPV Benefits (5%) | €1.26 b | €2.5 b |
| NPV Costs (5%) | €2.56 b | €1.91 b |
| NPV Net Benefit | -€1.3 b | €0.6 b |
| BCR Central Case (5%) | 0.49 | 1.31 |
| IRR Central Case | -1.8% | 7.3% |
| BCR NDP/CSF Case (5%) | 0.36 | 1.01 |
The overall drop in capital costs of €210 million is less than the €500 million reported by O’Reilly Consultants, but may be due to other adjustments factors we do not know. What we have termed operating costs have halved. These are recurring annual costs and would be dominated by operation costs but may include other elements such as capital replacement. It is still hard to imagine how operating costs could drop so dramatically when it is claimed that patronage is actually higher than previously expected. Operating costs should be dominated by wages and energy costs. What could have altered so dramatically, between the two evaluations? In addition, benefits have doubled. That also seems to be unlikely. Even updating of a model to reflect what were then more recent trends should not lead to such dramatic changes in forecasted patronage and higher savings on reduced congestion. As said above, we expect that this model would now need to be totally revised in the light of more recent trends which are much more likely to show sharp declines in future patronage from what has been assumed here. Mr Allen says an update of the economic analysis is necessary to “give a reliable estimate of the cost of implementation, but also take account of updated economic and patronage forecasts”. We would rather insist on it, as the analysis on which his claims are based is not reliable. A doubling of benefits and a halving of costs is not credible.
Was it partly the presumption of continued spectacular growth in passenger numbers through Dublin Airport? Airport passenger numbers determine about 20% of user benefits. As explained below, forecasts of passenger numbers have been shown to be very badly wrong.
It is unfortunate that we have to speculate about these important matters because the RPA refuses to provide the information, which would allow us to see if the assumptions are reasonable. In the absence of this information and given the implausible arguments that lead to such dramatic improvement in what was the worst option previously analysed, we believe the Board should err on the side of caution and either demand this information or reject the application.
It is important to note that when the NDP/CSF corrections were applied, the result was just barely breakeven (BCR of 1.01). It follows that any downward revision in the light of more recent trends in economic growth, etc., will push the results back into negative territory. Why has the RPA only referred to the results before these corrections? Surely the RPA must apply the corrections required by the NDP/CSF Evaluation Unit. This is the body responsible for methodology and standards of appraisal in the Department of Finance. The result based on their prescriptions is the definite one.
Extension to Swords
O’Reilly consultants report some results for an extension to Swords (but not Belinstown). It appears to be an analysis of the extent of the extra benefits of extending to Swords. It is said that overall benefits will increase by 59%. This is not a BCR ratio for the option but refers only to extra benefits. An estimate of cost of the extension has been deleted from the O’Reilly Consultants text. We are told that while the St. Stephen’s Green to Dublin Airport option 3b has been examined, an overall examination of a St.Stephen’s Green to Swords (or beyond) option has not been done. For some reason, the analysis of option 3a in the OBC was not updated. This extension to Swords analysis does not appear in the heavily redacted RBC available online. There is little mention of the Swords option beyond recommending that the option to extend to Swords be kept open in the tenders so that a decision can be made against the actual price. We do not know where the partial Swords analysis comes from. It may have been some additional work shown to O’Reilly Consultants but not included in the RBC.
It is inevitable that an extension to Swords will add benefits. That would happen if any extra passengers resulted. The extent of the increase in benefits seems very large but as only 20% of the Metro patronage will arise from Dublin Airport, much of it must involve Swords and beyond, so that considerable extra passengers and hence benefits can be expected from the extension to Swords. But is it worth the extra cost? As we do not have absolute costs and benefits or a BCR ratio for the extension we cannot say the extension is worthwhile when compared to costs. As we cannot build an extension to Swords without building the part to Dublin Airport, a good case at the margin would not be enough to justify the whole. In the OBC, Option 3a (City Centre to Swords) was better that Option 3 (City Centre to Dublin Airport), but it was still inadequate overall. Extra benefits exceeded extra costs so it improved the result but not by enough to breakeven overall. We would have to accept the dramatically improved BCR of 1.31 for the City Centre to Dublin Airport and that the extension benefits exceed its costs, in order to accept that thesis. We therefore do not have a solid basis for the claim that this proposal has a positive Benefit to Cost performance. In any case the results reported for the extension to Swords do not cover the extension to Belinstown.
Summary of RPA CBA claims
The RPA claims their own CBA analysis shows a positive BCR for the Metro option proposed. They claim this was verified in a published report by independent economic advisors to the JOTC who also claim the benefits are understated. They thereby endorse those claims. At the same time, the RPA also admit that the analysis needs to be redone to give reliable estimates and they refuse to release all details of their analysis.
O’Reilly Consultants were the advisors to the JOTC who evaluated the RPA work. When we examined their report, we find they reported CBA results on a City Centre to Dublin Airport option and not on the much longer route to Swords and beyond to Belinstown, proposed here. They report a BCR of 1.31 and that they believed it was underestimated.
O’Reilly Consultants source for the CBA information was the RPA’s Revised Business Case (RBC) for Metro North. This is confirmed in correspondence with the RPA and O’Reilly Consultants. Both the older version of this work, the Outline Business Case (OBC), and the Revised Business Case (RBC), both heavily redacted, are on the Transport 21 Website.
We examined these reports. In the OBC the CBA results for all route options were negative. They were best for a cross-city route (from Bray) but poorer for City Centre to Dublin Airport or Swords routes. The report did not examine an option to go beyond Swords. The results were worst for the option from the City Centre to Dublin Airport (Option 3).
The RPA favoured the cross-city Bray to Dublin Airport option and pointed to the weaknesses of the City Centre to Dublin Airport or Swords options (low patronage). These evaluations were redone in the RBC. We only have the heavily redacted version of this report and the limited information in the O’Reilly Consultants report. Option 3, Dublin City Centre to Dublin Airport in the OBC, became the Option 3b O’Reilly Consultants examined, and for which they report a BCR of 1.31. That was the option with the lowest BCR of 0.49 in the earlier OBC. How did it get from 0.49 to 1.31?
When the analysis was redone following the corrections required by the NDP/CSF Evaluation Unit in the Department of Finance the BCR falls from 1.31 to 1.01, i.e., just breakeven. The NDP/CSF Unit is the arbiter of standards and methodology so this result should be the definitive one. However, this result was not mentioned by O’Reilly Consultants, or by Mr Allen to this hearing. If the changes that brought about the initial dramatic improvement from 0.49 to 1.31 cannot now be justified, the result with the NDP/CSF corrections will not show breakeven.
We require a full explanation for the miracle that transforms the worst option in the OBC to the splendid result reported by O’Reilly Consultants. We found that estimated annual benefits had to be doubled and annual costs halved to produce a BCR of 1.31 from the earlier value of 0.49. We require access to all relevant reports, including the RBC and the report by Goodbody Economic Consultants for the Department of Finance that reviewed the costs of the project proposed. This Option 3 (OBC) or 3b (RBC) is not the City Centre to Belinstown option proposed by the RPA to this hearing, yet the finding of a BCR of 1.31 for that option, as reported by O’Reilly Consultants, is the basis of the RPA claim by extrapolation, for a positive and supposedly underestimated BCR for their proposal. There must be serious doubts about the validity of the BCR of 1.31 result and about its extrapolation to a Metro nearly twice as long.
Non-disclosure.
Mr Rory O’Connor referred in his evidence to the non-disclosure terms of the PPP guidelines. [21] What the guidelines actually say is: “Current policy is that the final PSB, or any elements thereof, is not made public [original emphasis] on the basis that revealing the amount that the State is willing to pay for a service may give tenderers an opportunity to increase their asking price above what they might otherwise seek.”
If tenderers are in a position to increase their asking price, as this text assumes, they are not competitive bidders. If there is a number of them, that suggests a cartel, requiring immediate investigation by the Competition Authority. Mr Allen told us this is a good time to build the Metro because of its economic stimulus effects and the “likelihood of competitive implementation costs.” One wonders why the RPA thinks construction firms are likely to be more competitive, yet they will also seek to up their “asking price”. It seems Minister Lenihan believes bidding will be competitive when he said presenting his revised 2009 Estimates to the Dáil on 23rd April:”…the current competitive environment for tender prices will result in better value-for-money for State spending.”[22] Another implication of the RPA’s contrary position is that the Government estimate is too high, otherwise why would bidders be tempted to increase their asking price?
Note that the PPP guidelines in fact say that it is the Public Sector Benchmark (PSB) that should not be disclosed (before the contracts are finalised). The PSB is an estimate of that part of the project on which the private sector is being asked to bid. It is built upon the costs used for Capital Appraisal (CBA) but is not the same. Even if there was some validity to the fear that bidders may be influenced by the release of detailed cost figures, there is no need to suppress all cost figures. The cost figures needed to validate the analysis are aggregate annual capital outlays and operating costs. Such annual aggregates are hardly likely to be of much use to bidders who might be trying to guess how high the state might go for a particular piece of work. In our opinion that is a useless strategy anyway, if there are enough competitive and independent bidders in the field. This policy of non-disclosure has been strongly criticised by Transparency International.[23] This kind of secrecy is the very opposite of the transparency we need.
The decision to withhold this information would appear to be in breach of Directive (2003/4/EC) on public access to information on the environment (AIE), which defines environmental information to include “cost-benefit and other economic analyses.”
It appears, that in spite of what Brendan Smyth told the Seanad, a full Cost Benefit Analysis, as required by Finance Guidelines, has not yet been produced for this project as it is now defined (at least not to the Board). It may be produced in time to influence the Government’s decision but cannot influence the Board’s decision, as it will only be made available after that decision. If such a study is the definitive test of sustainability, it is perverse that the Board cannot have that information in time to inform its decision. Given the decision to redo the early analysis to produce “reliable” estimates it is also clear that the original analysis no longer applies. That would also accord with our opinion, as it was based on a much smaller project, and a rosy economic and patronage outlook.
If the applicants refuse to provide the information which supports the socio-economic claims they insist on making, in terms of economic benefits etc, those claims have to be set at nought. The Board should find that the information on socio-economic impacts of this Metro proposal before the Board, both in the EIS and in submissions to this oral hearing, is not sufficient to allow for a proper and fully informed assessment. In my opinion the Board should request this information by way of additional information or reject the application. The Board might agree with us that the limited economic information available actually indicates that the proposal is not viable in social, economic and environmental terms.
Jobs
I would like to address a number of specific socio-economic issues raised in the EIS and the evidence of the RPA.
In these difficult times it is disappointing to see old jobs scam back in fashion (evidence of Mr Gilder). To quote Ronald Utt[24]:
“With the collapse of most centrally planned economies, use of I/O analysis is now largely confined to economic consultants hired to justify costly and underutilized building projects such as a convention center or football stadium because they will “create” jobs. In fact, such projects never create anything approaching the benefits projected through the misuse of these models, but there always seem to be local boosters, businessmen, and politicians willing to exaggerate the potential benefits.”
“Furthermore, arguments for a costly highway bill on the basis of potential job creation fail to recognize that creating jobs is not the same thing as creating value. The expenditure of any sum of money on nearly anything will contribute to a job, but whether that job leads to the creation of products and services of broad public value is another question. Hurricanes, tornadoes, and forest fires create large numbers of jobs, but they also destroy value in the process, an outcome not materially different from much of today’s federal spending on costly and underutilized light rail systems.”
Bent Flyvbjerg makes the following comment:
“Job creation and other local economic benefits invariably used by proponents to justify the billions spent on such projects often do not materialize or are so weak that they can’t be measured outside of the temporary jobs generated by construction itself, which may be substantial, but the benefits of which end the day the last construction worker leaves the site.”[25]
It may be a surprise to some, but jobs are not economic benefits; they are economic costs. If you are in business you know that is so because you have to pay wages. If you propose a business plan to a bank you cannot say: “It’s OK, I’m going to create jobs!” The bank will still want to see that your proposal is viable after you pay the wages at the going rate. In a time of unemployment, wages do not easily adjust downwards for many reasons, wage agreements, minimum wages, etc. A Registered Employment Agreement (REA) applies to the construction industry in this country. Therefore, for a private business, wages may only drop a little in a recession and that could make a business proposal a little more attractive, but maybe not by much.
However, at a time of significant unemployment, the cost to society of employing people who would otherwise be unemployed is lower than the wage rate. The real cost to society is lower even if wages do not fall. Therefore when we are assessing a project from society’s point of view (as in CBA) we can legitimately charge less for the labour we will employ. The concept is that of the opportunity cost of labour. That just means what society would have gained if that labour was used for its best alternative, and therefore it means what we lose by diverting the labour to our project. That cost is not likely to be zero. It is often taken to be close to the rate of unemployment benefit. This means that a project, which might not have passed the test of economic viability for a bank, might do so because the labour costs are now really less, when evaluated from society’s point of view. In that sense job creation by the proposed project is a benefit – but it is only in the sense that the cost to society of engaging that labour is less than it would be in better times: a lower cost is, in effect, equivalent to a benefit. One immediate consequence of this reasoning is that the value of a job is not infinite; it is determined by the difference between its opportunity cost and the value of what it can produce.
To return to Prof Utt, it is sad to see again the endless repetition of the “jobs, jobs, jobs mantra” as if all projects that would create employment are justified. That logic leads to justifying digging holes and filling them in again. That is not good public policy. For the project to be in the national interest, the benefits that flow from the project, allowing for the lower opportunity (real) cost of labour, still have to exceed the total costs.
By way of illustration: Mr Gilder says that the construction of Metro North will generate some 1,550 jobs in Full Time Equivalent (FTE) terms – before applying a multiplier of 1.5. Let us assume that half of these people would otherwise be employed given that many of them would be highly skilled and many brought in from abroad for the project. Assume the jobs going to the unemployed will be paid at the average annual industrial wage of about €33,000 and that the opportunity cost of this labour is only about €10,000 (about the rate of unemployment benefit). That means the “benefit” of employing these people is about €23,000 each per year or €18 million per year in total. Considering this to apply over the five years of construction, and without discounting, we get about €90 million. Assuming the best and unlikely case that the opportunity cost of labour is zero, we get figure of €25.6 million per year or a total of €128 million. While far from negligible sums, they have to be compared to total costs, in the range of €4 billion to €6 billion. The essential point is that general assertions about the benefits of job creation are worthless. These claims have to be quantified with clearly stated assumptions and put into the equation with all the other costs and benefits before a decision is made. One must also consider if alternatives might not have had greater labour creation effects. To justify this investment on the grounds that it will generate €90 million (or a maximum of €128 million) in job creation benefits would be absurd, but that is the proposition being put to us by the RPA. Minister Dempsey does not take that view. He said on the opening of this hearing: “We’re not building this just to create jobs.” [26]
As usual, these job creation arguments ignore alternatives. Dr Edgar Morgenrath pointed out in the Irish Times on 24th April last that according to the CIF every €100 million spent on construction projects will create 1,000 jobs and that the net cost of a construction job is €40,000, after taking account of tax receipts from income tax and non-payment of social welfare. However, Dr Morgenrath adds, “This turns out to be almost four times larger than the cost per IDA Ireland created job.” In other words, the same spending could create many more jobs if used in a different way.
We should point out that we do not know if any explicit value for jobs created has been or will be included in any CBA, for example by reducing the labour costs component of the capital costs. Even if it is not, unquantified and populist arguments about job creation are often made to justify proposals that fail objective analysis. We need to disregard those claims.
Multipliers.
The multiplier effect is widely used to inflate benefits, but often with questionable justification. The notion is that, as newly employed people spend their earnings they generate more employment. Thus, directly created jobs lead to indirectly created jobs. This is the well-known multiplier effect. It is often a multiplier fallacy. It is unfortunately common for Economic Impact Assessment (EIA) studies to claim these multiplier effects to the sole benefit of the project under examination. Extra spending may have multiplier effects, but as alternative projects would also involve spending, a priori, any extra economic activity does not distinguish one option from another. Use of such multipliers will favour the most expensive option, as it will have the higher multiplier effects. As a first approximation, it is safer to assume that, because all alternatives would generate additional economic activity, it should not be counted in analysing the project under examination.
Indeed, if it is raised from taxes, leaving it in the pockets of taxpayers to be spent by them would also have multiplier effects. Advocates have to be able to argue that their proposal will have real economic effects greater than would otherwise be the case for the claim to have any validity. Most analysts avoid multiplier claims because it is fraught with danger of misleading. Professor Utt’s boosters have no such problem.
The most transparent method of putting a value on job creation is to reduce relevant labour costs to reflect the real opportunity cost of that labour. This avoids a usually arbitrary and undiscerning application of a multiplier.
Contrary to the application of a spurious multiplier “bonus” the Department of Finance, in 1999, recommended a surcharge of 50% on Exchequer spending because it has disincentive effects.[27] The CSF document says: “Price of Public Funds: In order to take account of the distortionary effects of taxation, a shadow price of public funds of 150% should be applied to Exchequer cash flows (taxes, grants and subsidies) to make them commensurate with private cash flows.”
That surcharge (or negative multiplier) was reduced in recent years of exchequer surplus, and the surcharge required by the NDP/CSF unit for the OBC report in 2002 was down to 125%. I would have thought that in the light of recent increases in the cost of sovereign borrowing by this country, there is likely to be a case to increase the penalty again. The implication is that public funding of this proposal should attract such a surcharge. That public funding may come in the form of annual payments to PPP partners to compensate them for their capital costs, at least. Experience shows that metro schemes tend to generate sufficient fare revenue to cover only operating costs and depreciation.[28] The public purse is left to pay the capital costs. Such subsidies may well be justified if the total benefits to society exceed total costs. As a first approximation, subsidies are a transfer within society and do not effect the Cost Benefit result. However, if a surcharge is applicable, that is not so, and the surcharge on any public expenditure needs to be added to the costs of the project.
That being said multipliers at least have the benefit of being specific. A multiplier rate is proposed and at least may be challenged. The same does not apply to the “wider economic benefits” argument. These nebulous untestable benefits are usually appealed to as a desperate last resort when the objective evaluation fails to justify the preferred proposal. The OBC case made such a claim to wider benefits when its analysis, especially after the NDP/CSF correction, failed to achieve positive benefits. We do not know if this claim was used in the RBC to help get the BCR of 1.31 for the City Centre to Dublin Airport option. If so, it would have required some quantification of that claim. These wider benefits are not usually quantified, because they are so vague, and quantification might lead to a challenge. They are in any case essentially useless, because, as in the multiplier case, they also apply to the alternatives. Further, if there may be wider economic benefits, there may also be wider economic costs. Where are these addressed? Decision-makers should not be misled by such poor arguments.
Mr Gilder claims that 90% of the multiplier-derived benefits will stay in the Greater Dublin Area.[29] I am sure he is aware of the well-identified problem of leakage of expansionary public spending in a small open economy such as Ireland. Much of borrowed injection would be spent on imports, foreign holidays, etc., and will not have the domestic simulative effects claimed. He uses a UK multiplier, which may be more applicable in that larger economy. Even at EU level, it appears we are reluctant to rely too much on Keynesian expansionary policies in the present crisis, because of the likelihood of such leakage. The Department of Finance is explicit that “induced effects should not be included in the analysis unless it is demonstrated that they have been calculated on the basis of a rigorous methodology using data for the Irish economy.”[30] U.K multipliers are not appropriate.
Other benefits in perspective: reduction in car emissions.
Another benefit presented to justify the project is the reduced climate changing car emissions with the modal shift to Metro when operative. The NTS says: “It is estimated that the proposed scheme will remove in the region of up to 5,000 peak hour car trips from the road network.” It is not clear how many cars per day are meant. Minister Martin Cullen said the Metro would remove 41,000 cars per day, although that claim has been disputed.[31],[32] We will assume that higher figure. The shadow price of carbon dioxide being used by the U.K Government in its assessments is about £25; say €30 per tonne. (It is assumed to increase in real terms over time, but we will ignore that here). Assuming the average car trip saved by the Metro is 10km this means a saving of 1.8 kg saved per trip. If 41,000 trips are saved per day, that is 74 tonnes of carbon dioxide per day, or a benefit of €2,200 per day. That is about €800,000 per year, which over 30 years is about €12 million (discounting at 5% per annum). Therefore the total benefits from the reduction in car emissions of CO2, if we take Minister Cullen’s estimate, are about €12 million over the project’s lifetime. A small result, and negligible compared to a cost of €4-6 billion. Again, presenting us with wordy statements about benefits with no quantification in money terms is a worthless exercise. Modal shift from car to public transport has other benefits including reduced congestion (measured in terms if time saved), reduced accidents, and reduced air pollution. The RPA did not confirm the claim of 41,000 fewer cars per day. They provided percentage changes in travel times, average speeds, etc., but such indicators do not to quantify the possible benefits. I understand the traffic models used by the RPA have the capacity to estimate the time saved by reduced congestion. Such figures would be very useful in allowing claims for these benefits to be objectively assessed as to their significance in absolute and relative terms, thereby avoiding general claims about their significance.
Modal shift from buses to Metro North.
In evidence by David King, presented by Mr Rory O’Connor on 29 April 2009, it is revealed that according to the RPA’s indicative forecasts, the Metro will add 45.6 million trips to the Metro network. However, there would be a corresponding decrease in bus trips of about 16.3 million due to bus users transferring to Metro North from parallel bus routes. That is 36% of the expected Metro trip impact. This is a staggering admission that the Metro will have a huge detrimental effect on environmentally positive bus usage. The impact is appeased by the suggestion that “bus services would be rearranged”. We have to presume therefore that bus services will be expanded on routes where they are not currently justified. It is not obvious how this will offset in any way the negative environmental impacts of the shift from currently operating bus routes.
Rosy future?
There is a serious lack of information on which to assess the validity of claims made about the prospects for the proposal. There are a lot of assertions but few facts. To revert to the CBA problem, if the analysis had been done and published we would have, on a year by year basis, the assumptions about patronage on which the estimates of most of the benefits of the project would be based. We have a number of statements about capacity and patronage in some years, but we do not have the basis for these patronage figures, passenger origins and destinations, the assumptions about economic growth on which they are based, etc. Without this detail we cannot fully test the validity of the claims made for this proposal. There is good reason to believe that the future will be very different from that assumed. We are not convinced by the claim of Mr Gilder, that it will be fine anyway and indeed that the recession is good for this project. In his evidence Mr Gilder said: “The importance of progressing Metro North and the benefits it will deliver are, if anything, magnified by the downturn and are reflected in my evidence.”
Dublin Airport
Dublin Airport is expected to account for up to 20% of Metro usage, but the growth outlook is very different from what is was a short time ago. The NTS says: “Dramatic growth is forecast for Dublin Airport. Currently carrying over 20 million passengers each year, this is expected to grow to over 30 million by 2015 – 2016.”
Until recently, the DAA was predicting continued strong passenger growth, e.g. Declan Collier gave a figure of 29 million passengers per annum (mppa) by 2015, similar to the NTS forecast. With the recent sharp drop in passengers we estimate that 29/30 million will not now arise until about 2026. Passenger numbers will decline by 15% in 2009 and a further 5% in 2010. By 2015 numbers may only be back up to 21 million. If the Metro opens in 2015 there will be 9 mppa (nearly one third) fewer through Dublin Airport than expected, according to the NTS. Airport employees will be important users of the Metro. With 9 million fewer passengers, there will be 7,600 fewer employees.[33] Using modal shift to Metro figures provided at the Terminal 2 oral hearing, this fall in passengers and employees will mean about 10,000 fewer Metro users per day or 3.5 million fewer per annum. While passenger growth may recover somewhat in a few years time, it will not compensate for the lower benefits implied by the downturn. Therefore estimated benefits over the lifetime of this project are lower than previously assumed and BCR estimates based on them cannot stand.
Adding further to the uncertainty about Dublin Airport’s future, two proposals for new international airports in the Midlands are being considered at the pre-application consultations stage by An Bord Pleanála: one by Midlands Airport Development Ltd, for an airport near Tubber, Co. Offaly and another by Tundra Holdings for a Leinster International Airport near Portarlington, Co. Offaly. Either of these developments, which will draw passengers from Dublin Airport, must have significant implications for a future Metro North serving Dublin Airport.
The board is not optimistic about future passenger growth, and has already refused permission for a development on the grounds that future passenger numbers do not justify it. In the case of a 2714 space car park near Dublin Airport, the Board, among its reasons and considerations for refusing permission in November 2009 said[34]:
“..the emerging and likely future pattern of passenger traffic at the airport in the short to medium term, which is one of significantly slower growth if any, and the consequently reduced need for additional parking facilities,”
Further, the Board made the passenger issue central:
“In deciding not to accept the Inspector’s recommendation to grant permission, the Board had regard to the factors set out above and, in particular, to the recent indications of declining air passenger numbers at Dublin Airport and the likely pattern of such traffic in the short to medium future.”
If additional car-parking cannot be justified because of a future shortage of passengers, how can the Board believe a €4-5 billion Metro to serve those disappearing passengers is justified?
MNEC
The strip of land 1km either side of the Metro line has been designated as the Metro North Economic Corridor (MNEC). A report was prepared for Fingal County Council by Indecon in 2008 on the prospects for the MNEC. [35] We believe that many of the optimistic assumptions about the future for the Metro contained in the EIS and in the RPA submissions come from that report, or a common source. For example, as far as Dublin Airport is concerned, the report, not even one year old, says:
“The figure below highlights the recent trends in and projections for passengers using Dublin Airport out to 2020. This shows the recent very rapid growth recorded in passenger numbers and the prospects for further strong growth, underlying the importance of this critical infrastructure, both for the MNEC and the country as a whole.”
The chart on page 10 of the report shows about 27 mppa for 2009 when it is clear there will be only about 20 mppa: a 25% error for a one year projection. The same chart shows between 43 and 58 million passengers by 2020. In information provided recently to the Aviation Regulator, the DAA now expects about 31 mppa by 2020. We believe, on very recent trends, there will only be about 25 million by 2020. The Indecon passenger estimate is based on blind extrapolation of earlier trends and looks to be about 100% wrong. We believe that little credence can be given to the optimistic claims made for the Metro based on this type of analysis. This sort of wishful thinking may explain much the dramatic improvement to 1.31 in the BCR for the City Centre to Dublin Airport route in the RBC and as reported by O’Reilly Consultants.
The population of the MNEC is put by Indecon at 59,000 in 2006 growing to 128,100 in 2025/30; a more than doubling. The Non-Technical Summary says that the “proposed scheme will permit Fingal County Council to plan for an increase in the MNEC’s population of almost 59,000 (2006 data) to 128,100 (+117%) by the period 2025/2030.” That is not quite a projection to 2025/30 but is remarkably similar, in fact identical, to the figure which Indecon projects for 2025-30. This analysis looks like the same dubious extrapolation that was applied to the airport passengers. In the Indecon report, it is a target rather than a projection. From the NTS we conclude that this doubling of population depends on the provision of the proposed Metro and is not an exogenous growth in population that would require such a service.
The density of population in the catchment area of the Metro is critical to its viability. Nearly all the benefits from Metro systems accrue to the users of the system and those users are largely determined by the catchment population.
Alain Bertaud states: [36]
“The metro network in Barcelona [20] is 99 kilometers long while 60% of the population lives at less than 600 meters from a metro station. Atlanta’s metro network is 74 km long – not so different from Barcelona – but only 4% of the population lives within 800 meters from a metro station. We should not be surprised if in Atlanta only 4.5% of trips are made by transit vs. 30% in metropolitan Barcelona, where the high density allows also an impressive 8% of all trips to be walking trips.
“The comparison between Atlanta and Barcelona shows in an anecdotal form why density is important in transit operation. Empirical evidence confirms that there is indeed a density threshold below which it becomes impossible to provide transit service.
“The literature review conducted by Holtzclaw [16] on transit and density suggests that there might be a density threshold around 30 people per hectare (p/ha) for intermediary bus service, 35 p/ha for light rail and 50 p/ha for metro (see Table 1).”
But patronage forecasts are frequently exaggerated. Fouracre and Mounder have this to say from experience with Metro systems in developing countries:[37]
“ There is a major risk for the financial viability of the project. Forecasts of patronage are the basis on which a metro is economically and financially justified. Over-optimistic predictions are the result of many assumptions that cannot always be realised. Of the newly built metros in nine cities for which data were available, only one achieved its expected ridership levels, while three achieved only half, and five had patronage that was between 50% and 90% lower.”
Bent Flyvbjerg, finds that rail projects are particularly poor at forecasting demand. In a study of 44 urban rail projects in North America, Europe and developing countries he found that passenger ridership was 40% lower than forecast. [38]
Mr. David King showed us a table comparing the capacity of Metro North to other cities. It is quite striking that the city population density shown for Metro North was the lowest at 1,300 per km2 (13 p/ha), except for Manchester. Madrid, Vancouver, London, Seville, Paris, Budapest and Munich were from 3,500 to 5,200 per km2. Nevertheless the capacity of 20,000 persons per direction per hour is given for Metro North. Only Madrid, London Jubilee Line and Paris Line 14 are higher. This proposal is seriously out of line with international comparators.
If, when the Metro opens, the MNEC population is about 60,000, of which the working population is about 46,000 (and falling), a capacity of 20,000 per hour each way means that almost the entire working population of the MNEC needs to spend the day going up and down on the Metro to fully use its capacity. That will not help economic recovery.
The population density of the MNEC is about 1,500 per km2 or 15 per hectare. If a population of 128,100 is reached by 2025/30 that will be a density of 32 per ha. That is below the threshold for light rail and well below the metro threshold of 50 per ha given above. The population of the Dublin City Council area is 44 per ha, for the Dublin area as a whole it is about 30 per ha. It appears that the MNEC is a relatively unpopulated area of Dublin, at present, but would still be less densely populated than Dublin City if the Metro were provided and MNEC population did grow to 128,000. This is rather obvious when we look at the MNEC map. Apart from Swords, there are no populated areas from Santry to Belinstown. It has to be asked if the proposed Metro is in the right place.
Of course the MNEC contains Dublin Airport, whose passengers are not counted in population density. However, the Airport will account for less than 20% of Metro usage and passenger numbers will be severely below expectations.[39] Recall that this higher population density is aspirational: it does not yet exist and may not arise, even if the Metro is built. This introduces a different order of risk beyond advocating a Metro system to meet the needs of an existing population whose density meets international norms for metro construction. Was such a risk element incorporated in the CBA Mr. Allen claims was done? It is of the nature of a “build it and they will come” proposal. What is the BCR if they do not come? If the Metro has to generate the population density to justify itself, there is no particular reason why it should be located in this low-density area, rather than another similar area where it might also generate its own sustaining population.
Expectations of a doubling of population in the corridor have to be very weak. The Fingal area, especially the Swords area, was the fastest growing area of the country during the boom years.[40] The population of the MNEC grew by 33% between 2002 and 2006. Sixty-four percent of the MNEC population is in Swords and that is projected by Indecon to grow to a 78% share. Immigration accounted for a significant proportion of that growth.[41] We are seeing a reversal of that flow and are unlikely to see previous rates of immigration for a very long time, if ever. Net immigration was about 70,000 in 2006 and 2007. It was reported in September 2009 that the adult population declined by 5,000 in the first quarter of 2009 and renewed net outward migration in the 12-month period to Q1 of 2009, meaning further reduced demand for housing.[42] The CSO confirmed on 22 September 2009 that net outward migration, at 7,800, had resumed for the first time since 1995 and that the adult population (over 19) had actually declined by 17,100 over the year to April 2009.[43] An exceptionally high number of births in the year to April 2009 led to a net population growth of 37,200, but this will not add to the adult population for some time and we can expect a sharp fall in the adult commuting population for many years, and an even greater fall relative to previous expectations. Further, we must factor in the probably higher unemployment levels for years to come, which will further reduce the expected adult commuting population.
The Swords area is key to the Metro’s viability. According to Minister Noel Dempsey: “The greatest passenger demand will come from the Fingal, North Swords area beyond the airport which will be directly served by the metro.”[44] The volume of passengers determines most of the benefits: the fewer the passengers, the fewer the benefits.
Earlier growth in Swords was based on a property bubble that has almost destroyed this economy. Fixing this problem will not mean a return to previous growth trends after a few difficult years. The RPA’s thesis rests on a return to the glory days that will lead to a doubling of current population in the MNEC. Here is Minister Dempsey in February 2009, explaining why we cannot afford not to proceed with the Metro North project: “Fingal is the fastest growing county in Ireland. Its population is set to double over the next 20 years and most of that growth will happen in the residential areas along the Metro North corridor.” His population expectations are highly dubious, based as they are on even more housing development in an area around Lissenhall/Belinstown, which is still largely rural.
However, according to Mr Frank Allen future population is not the justification for the Metro at all. In his evidence he said (paragraph) 5.2: “However, the case for Metro North does not depend on that future growth [in the Fingal area] but is in large part a response to development that has already taken place.”
The population of Swords is currently 38,000 (working population 29,000, and falling). There is some logic to Minister Dempsey’s claim, however optimistic, that the Metro case hinges on serving a population that is “set to double over the next 20 years” due to housing development North of Swords. There is no sense to a case based largely on current population along the MNEC in Fingal (essentially Swords). How can a working population of less than 29,000 justify a Metro that can move 10,000 passengers an hour each way when it opens, and eventually 20,000 per hour each way?
Economic prospects.
There is a wide consensus among economists that our recovery must be based on export growth led by improved competitiveness but that it will lag EU recovery, which in turn will lag US recovery. The IMF has pointed out that the combined synchronous and financial nature of the worldwide recession means that it will be severe, protracted and recovery will be slow and weak: “Recessions that are associated with both financial crises and global downturns have been unusually severe and long-lasting.” [45]
Nobel Prize-winning economist at Princeton, Paul Krugman said in Dublin also on 5 June last that Ireland’s economy will experience “long slow grinding deflation”. There will be no rapid bounce back after the sharp fall, which was due to over-borrowing and a very relaxed attitude to debt. He said it would be at least five years before Ireland’s economic growth matches that of the rest of the euro zone and Ireland could be facing a “lost decade” of economic stagnation.” In Dublin on 4 June last Professor James Galbraith said that “the forces [credit booms and inflows of private capital] that gave us these periods of prosperity cannot be recreated anytime soon and indeed even if they could be it wouldn’t be a good idea to attempt to do so.”
While we cannot be sure of the path of recovery we know that future conditions will be very different. There will be very little reckless lending, or over-leveraging and we can expect much more modest levels of FDI. Our recovery will not be based on a new building boom. We are likely to see some years of sharp decline in economic growth, followed by some years of stabilization before a much more modest path of growth of perhaps 2% per annum resumes. Some would say that was the optimistic scenario, as economic collapse and even sovereign default are not ruled out. If the IMF is right, it will take a long but uncertain time to return to the disposable income levels of recent years. We will only get back to the 2007 peak GDP around the year 2020. We will not return to the 2006 madness of building 90,000 housing units a year that was exemplified by the Fingal/Swords/MNEC area and is required to meet the Minister’s expectations.[46] Only 10,000 units will be built in 2010. Fingal County Council’s and Minister Dempsey’s ambitions for massive housing development north of Swords, on which he says this proposal depends, is a continuation of the property bubble illusion and is wholly unrealistic and even dangerous. Professor Krugman accused the Government of “cheering on the bubble.” It appears some have learned nothing, and certainly not the humility John Hurley hopes for.
Simon Clear in his evidence pointed to population projections published by the DoEHLG in January 2009, which show little change over earlier projections made in 2007. Although apparently recent, one has to doubt that these projections include the more recent and projected downturn in economic activity. It would be useful to have the detail of the growth assumptions made in those projections but the report, contained in a “pre-draft issues paper” for a review of National Spatial Strategy, is not readily available. The weakness of supposedly recent evidence put out by the DoEHLG is shown, for example, by a report on the construction industry for DoEHLG by DKM Economic Consultants, dated September 2008, which foresaw a decline in GNP of only 1.5% in 2009.[47] It further adds “the prospects for 2010 are positive, with a return to growth of 3.5%.” We now know these projections are wholly unrealistic as we are facing a decline of 14% (now maybe 12%) between 2007 and 2010. The Central Bank also says that only 18,000 housing units will be built in 2009 compared to DKM’s forecast of 25,000. Were the DoEHLG population projects based on similar, apparently recent, but seriously outdated assumptions?
This is not to belittle the problems of forecasting in the present volatile environment, but pretending there are no problems will not make them go away. The RBC assumed a 4% growth per annum over the period 2005 to 2010 and 3% per annum thereafter. Using the latest IMF and OECD figures and assuming recovery starts in 2011 (the IMF expects 1% growth in 2011), we see that the RBC assumed a cumulative 25% growth from 2005 to 2011 when we will see a cumulative decline of 3%. That is quite a gap (28%) and one we will not get back, except in an unlikely future where growth exceeds what was previously predicted. The opposite is much more likely. If growth only recovers to 2% per annum after 2014 as the IMF expects, rather than the RBC’s 3%, by 2035 real GDP will have increased by 58% over 2005, only about one third the RBC’s predicted increase of 155%. This demonstrates how the recent sharp downturn and lower future growth prospects can radically alter the predicted future and hence assumed future benefits that are closely related to economic growth assumptions, such as patronage and the value of time saved.
Part of the projected population growth along the MNEC is attributed to the Metro itself. It is probable that people will be attracted to live near a Metro line, but to assume that the tragic bursting of the property bubble, with all our other economic ills, will have no lasting effect on the viability of this Metro is not credible.
Mr. Gilder deals with this when he says in his evidence: “Even if development is slower than previously expected in the next year or two I believe that the long term forecasts which underpin the strategic transport programme, including Metro North, are robust.” Note the euphemistic “next year or two” when we are facing many years of slower development than previous expected. He further argues that the case is better for the Metro because of the recession. He said: “The importance of progressing Metro North and the benefits it will deliver are, if anything, magnified by the downturn and are reflected in my evidence.” In effect, the argument is that we needed the Metro to cope with exponential growth but now that this growth has disappeared we need the Metro to help bring it back. That is denial in the face of adversity, and demonstrates the persistence of the poor reasoning that got us into the mess we are in. These long-term forecasts are patently not robust. If an argument is based on projections that are no longer tenable, the argument falls. Sophistry will not put Humpty Dumpty together again.
Sensitivity of CBA results.
To test the sensitivity of the CBA results we can construct a simple CBA model from the bits of information we have. We assume capital investment is at €600 million per annum (in 2009 terms) in each of the first five years (hence our earlier €3 billion capital cost estimate). This starts from the estimate of €333 per annum from the RBC and allows for inflation in construction costs, tunneling at Ballymun and Dublin Airport, and the 8 km extension to Swords and Belinstown. [48] Patronage is assumed to grow from 40 million in year six (the first year of operation) to 125 million in year 30 (the twenty-fifth year) and that total benefits are proportional to patronage. We assume unit benefit (essentially the value of time saved) grows at 2% per annum in real terms. Operating expenses are taken as €1.50 per trip, on the assumption that fares will just cover operating expenses, as is the widespread experience. Using the official discount rate of 4%, a starting unit benefit per trip is determined by solving for breakeven, BCR = 1.0. That value is almost €5.22 per trip.[49] This also incorporates a shadow price of public funds of 125% as recommended by the NDP/CSF Unit, which is applied to the capital cost to be repaid eventually out of public funds. We get an NPV of costs and benefits of €4.65 billion. This exercise is essentially a calibration to a baseline against which to test alternative assumptions. However, it is probably close to reality, given that with the NDP/CSF corrections, the claimed BCR of 1.31 for the RBC option dropped to breakeven.
Table 2: Sensitivity tests of CBA results.
| Baseline growth 5% pa | Cost overrun 45% | Ridership over by 50% | Combined tests | |
| Patronage year 6 | 40 m | 40 m | 27 m | 27 m |
| Patronage year 30 | 125 m | 125 m | 84 m | 84 m |
| CAPEX yrs 1-5 | €600 m | €870 m | €600 m | €870 m |
| BCR (4%) | 1.0 | 0.76 | 0.74 | 0.55 |
| IRR | 4% | 1.6% | 1.5% | -0.7% |
Professor Bent Flyvbjerg, (cited by Colm McCarthy in his submission) deals extensively with the problems of delusion and deception in large infrastructure projects. [50] He notes that:
“Unfortunately, cost overruns, delays, and exaggerated benefits are the norm rather than the exception…”
He adds that where project promoters believe their venture is in the public interest and will benefit society and posterity,
“They feel that they should do anything possible to make the project happen, including cooking forecasts of costs and benefits.”
“Underestimating the costs and overestimating the benefits of a given project results in an artificially high benefit-cost ratio, which in turn leads to two problems. First, the project may be started despite the fact that it is not economically viable. Second, a project may be started instead of another project that would have yielded higher returns had the actual costs and benefits of both projects been known.”
While we would not suggest deliberate deception here, there may be considerable delusion and political pressure. Professor Flyvbjerg found that on average for 44 urban rail projects cost overruns were 45%. He also found that ridership was overestimated by 50%. He recommends the use of “reference class forecasting”, under which experience with similar projects elsewhere is factored into the analysis. To apply this to our baseline we first add 45% to capital costs. It is important to note that even if the State manages to switch some of the risk of overrun to contractors, the cost to society will include that overrun. The BCR drops to 0.736 while the IRR drops to 1.6%. When we cut patronage back to 27 million in year 6 and to 84 million in year 30 (without the cost overrun), the BCR drops to 0.74 from 1.0 and the IRR from 4% to 1.5%. When we combine both these tests, we find the BCR drops to 0.55 and the IRR goes to a negative 0.7%. This future is not unrealistic given international experience on the extent of the two common assessment errors of underestimated costs and overestimated benefits. The gravity of the recent economic downturn, the much reduced prospects for economic and population growth in the GDA and the MNEC given the permanent demise of the property bubble, the fall in Dublin Airport passenger throughput, and the real prospect that Metro West or other connecting transport infrastructure will not be built. Metro North is not a viable project in this scenario. We do not claim to be able to predict the future, but we claim this future is more probably than the rosy future behind the RPA assumptions, which do not appear to have taken account of international experience, as Professor Flyvbjerg recommends and to have factored in the dramatically different future we now face.
Alternatives
It is notable that in the EIS and in RPA presentations, alternatives seem limited to alternative routes for the Metro, essentially between Parnell Square and Dublin Airport. This is not a sufficient consideration of alternatives, which must include alternatives to the Metro, not just alternative routes within the Metro option.
The obvious such alternative is buses. They are much cheaper but are said to be limited in their passenger carrying capacity. Over and above what can be delivered by a cheaper bus system, is there a case for a Metro system? In a perfectly competitive system based on social pricing, we would have a combination of modes of transport such that at the margin, marginal social costs equaled marginal social benefits for all services. It is not therefore a choice between a bus system and a Metro but the optimal mix of the two. The issue is whether or not the Metro proposal can be justified in co-existence with a viable bus-based system that is free to reach capacity. If thousands of people could be moved more cheaply by a bus system, the more expensive metro system would have to be sustainable without restricting a competing bus service.
Mr Rory O’Connor in his evidence said (paragraph 3.8):
“A bus alternative was not assessed as a bus corridor cannot provide the capacity needed or the journey time to attract large amounts of car users to public transport to meet the objectives in terms of modal shift. The maximum capacity of a bus corridor is around 2,000 passengers per hour per direction.”
However, according to A Platform for Change, buses can transport 3,500 people per hour each way, possibly up to 5,000, with difficulty.[51] That is double Mr O’Connor’s figures but still only a quarter of the maximum capacity of the Metro (by 2040). Is there a better bus option?
O’Reilly Consultants who examined the RPA’s analysis also looked at a Bus Rapid Transit (BRT) system used in Curitiba, Brazil.[52] The system uses extra long, five-door buses boarded by passengers via “tube” stations after pre-paying for tickets. Each bus carries up to 300 passengers. With loading taking only 20 seconds, buses can run every 30 seconds moving up to 36,000 passengers per hour, each way, when running in a dedicated bus lane (busway). It is claimed a longer articulated bus can move up to 48,600 passengers every hour, each way.
[SEE MOVIE]
This 36,000 – 48,600 capacity per hour each way is double the maximum design capacity of the proposed Metro of 20,000 each way, or 40,000 per hour. [53] That would be twenty times the maximum capacity Mr O’Connor allows for buses. The possible maximum demand of 20,000 each way for the Metro is “based on all growth assumptions being realised” by 2040[54]. We believe these Metro forecasts are overly optimistic and that we are therefore very unlikely to need that capacity before 2040. The BRT option may cope with the real demand. Dublin is not Curitiba and the claims for Curitiba and BRT systems have been challenged.[55] We cannot say what maximum capacity such a system could deliver in a Dublin environment but it deserves serious study. If BRT system could move even 15,000 per hour per direction, that would put off the need for Metro for quite some time. Metro North would eventually have to justify itself in competition with such a BRT in place.
Consider an alternative scenario to the Metro North proposal. To build a Metro we need a tunnel and some type of mass people-carrier. Metro North is mainly to serve Swords and, to a lesser extent, Dublin Airport (20%) from Dublin City Centre. But we have a tunnel already: the Dublin Port Tunnel, which already serves Dublin Airport and Swords via the M1, and is underused. Information is scarce on the maximum capacity of the tunnel and projected future usage. It appears to have a capacity of 70,000 vehicles per day, of which 55,000 are cars.[56],[57] Current usage seems to be about 16,000 vehicles per day. There is good reason to believe that will continue to be significantly underused, especially as the alternative port at Bremore near Balbriggan, and possibly another at Arklow develop.
Buses are very efficient people carriers, so that we could provide a good service at very low capital cost putting a string of buses through the Port Tunnel. With the spare capacity in the tunnel there would be little loss of benefits to other users of the tunnel.
In economic terms the Dublin Port Tunnel is a sunk cost. Its historic cost (€751 million) is not relevant to any decision now about its future use. The extra cost of tunnelling for this bus option is therefore zero. Many will baulk at that assertion but it makes sense now to exploit the Port Tunnel to maximise the return on that old investment. (I believe Mr John O’Connor would agree with maximising the use of existing public infrastructure.)
Some expenditure would be needed to optimise access to the airport and Swords but it would be miniscule compared to that needed for the Metro. We would need a busway down the North Quays from O’Connell Bridge to the Port Tunnel mouth at some capital cost. According to Mr Ian Byrnes’s evidence at the hearing, a North Wall QBC has already been proposed. As the Luas is planned to run down to the Point (now open) we could also run it on to the tunnel mouth at a small extra cost. The buses could run in the dedicated Luas corridor. As the Luas red line extension to the Point is already decided and presumably already cost-benefit justified, that capital cost can also be ignored for this purpose, adding only the extension to the tunnel mouth. In effect, for the extra capital cost of buses and some QBC kilometres we could effectively provide a Bus Rapid Transit (BRT) service for thousands of commuters.
The principle determinant of benefits is the volume of passengers. Of course, without the Metro stops through the city we would miss some benefits. However, the main purpose of the Metro is to serve Dublin Airport and Swords from the city centre and that is where most of its economic justification must lie. That is also the opinion of Minister Dempsey. The charts presented by Mr David King of peak hour demand in 2040, imply that about two-thirds of demand comes from City Centre, Dublin Airport and Swords stops. If the Cost Benefit Analysis was done of a bus service using the Dublin Port Tunnel we expect it would give a very positive return as the capital costs would be very low and many of the benefits now being used to justify a €4-6 billion cost on the Metro would still be gained, depending on the relative capacities. It is unlikely that the relatively few users of intermediate stops could justify the massive extra investment needed to build a Metro. The bus system could be easily and cheaply extended to use the M50 to serve Ballymun, and further. There could be opportunities to add patronage, for example, by linking to the new Docklands main line station at Spencer Dock serving the Maynooth line, which will also be served by the Luas C1 extension to the Point. There is also the prospect of providing the 40,000 employees and 22,000 residents of the Docklands area, including the Point Village with a 15-minute link with Dublin Airport. The area already has a daytime population greater than that of Swords, and has growth prospects at least as good.
On the cost side we have the extension of the Luas from the Point to the tunnel mouth, a bus connection down the North Quays to the Port Tunnel, and special bus lanes M1 to/from airport and M1 to/from Swords. A bus lane from O’Connell Bridge to the Tunnel mouth would be necessary even with the Luas extension, as the Luas alone can only deliver about 6,000 per hour. The busway could run down the North Quays or possibly share the same Luas corridor through the Docklands. A bus lane from St Stephen’s Green, down Merrion Street to the Quays might also be attractive. The Parnell Square proposed Metro stop is expected to generate considerable demand. The preferred route for the proposed Luas line BX/D is northward on O’Connell Street to Parnell Street and looping back down Marlborough Street to Eden Quay. Surely this line could feed those potential Parnell Square passengers to a QBC running East along Eden Quay.
The Luas C1 extension of 1.7km is to cost €54 million.[58] Let us be very generous and allow €50 million for a further extension of less than 1km from the Point to the Tunnel. Buses cost about £150,000 each, say €200,000. At a guess to provide a near continuous bus service throughout the day we might need about 30 buses. Let us be generous and say 50 buses; that is a €10 million outlay for buses. Bus lanes cost about €750,000 per km – say €1 million per km. How many kilometres do we need along North Quay, M1 to/from Airport and M1 to/from Swords? Let us be generous and say 30 km – so we need €30 million for bus lanes. That is a total capital cost of €90 million; say €100 million.
There would be operating costs (wages, fuel, etc.,) and we assume they would be higher than the operating costs per passenger of a Metro. We use our simple calibrated CBA model and assume capital costs of €50 million per year for two years, a shadow price of public funds of 125%, operating costs 50% higher per passenger than Metro, and that we achieve half (rather than a possible two-thirds) of estimated Metro patronage. For this BRT option we get: NPV of benefits of €2.5 billion, an NPV of costs of €1.2 billion, Net Benefits of €1.3 billion, at the currently required discount rate of 4%, a BCR of 2.1 and an IRR of 45% (see Table 3 below). That would be a very good investment and far superior to the breakeven baseline for the Metro option. Even if that option were better than breakeven, this BRT option would be proportionately superior, and always a much better option.
Exits from the Port Tunnel to the M1 and exists to Dublin Airport and Swords from the M1 would take some engineering, but that would seem to be a small problem for the designers of this Metro. If further engineering costs of €100 million were required to solve these problems, the CBA indicators would fall somewhat: BCR 1.9, IRR 26%, still a very viable project.
Table 3: CBA of BRT options
| Baseline Metro North | BRT baseline | BRT high CAPEX | BRT pessimistic | |
| CAPEX p.a. | €600 m 5 yrs | €50 m 2 yrs | €100 m 2 yrs | €100 m 2 yrs |
| Patronage first | 40 m | 20 m | 20 m | 14 m |
| Patronage last | 125 m | 58 m | 58 m | 39 m |
| NPV Benefits(4%) | €4.4 b | €2.5 b | €2.5 b | €1.7 b |
| NPV Costs (4%) | €4.4 b | €1.2 b | €1.3 b | €0.9 b |
| NPV Net Benefits | €0.0 b | €1.3 b | €1.2 b | €0.8 b |
| BCR (4%) | 1.0 | 2.1 | 1.9 | 1.9 |
| IRR | 4% | 45% | 26% | 25% |
Taking the worst case assumption that was applied to the Metro North baseline and assuming a cost overrun of 45% and a patronage reduced to allow for a 50% overestimate, we see that the BRT options still scores very well with a BCR of 1.9 and a IRR of 25%. Even though capital cost is higher in that scenario, costs overall are down because of lower operating costs due to lower patronage. This very positive performance is largely determined by the low initial capital cost compared to Metro North.
Using the Port Tunnel and the M1 we have virtually the ideal dedicated busway infrastructure (for most of the route) for such a bus system. It would also avoid the years of disruption to businesses in the City Centre and along the route, and associated costs. To assess that loss we need an estimate of net loss in national value added, allowing for some displacement of business activity. This seems to have been ignored by the RPA. In October 2006 Robert Watt, then of Indecon, told the Dublin economists’ conference: “We need to count the very considerable disruption costs during construction of projects like Luas.” In his submission for the Dublin City Business Association, Mr Jerome Casey provides a “Moderate” estimate of disruption costs of €227 million per annum for 3.5 years of construction (Table 3). [59] If we add this to the costs of the baseline Metro North scenario, the BCR drops from 1.0 to 0.87 and the IRR drops from 4% to 2.7%. If Mr Casey was only half right when we allow for displacement, the BCR drops to 0.93 and the IRR to 3.3%. That drop in performance alone would put the viability of project into question.
The BRT system would also deliver the benefits years earlier, adding to the Net Benefits and IRR. I would also expect the BRT system to have the same (pro rata) decongestion and emissions-lowering benefits by reducing car use to a similar extent as the Metro. There would be no need for the potentially wasteful extension of the system beyond Swords to Belinstown. If and when population density does develop there, the service could be extended to provide the service needed.
Would the Port Tunnel route be much longer? From O’Connell Bridge to the mouth of the Port Tunnel, along the North Quays, is 3km. From the tunnel mouth to the Malahide Roundabout at Swords is 13km. Therefore O’Connell Bridge to Swords, via the Port Tunnel is 16km. O’Connell Bridge to Swords, via the proposed Metro, is 14km. Therefore the bus route is 2km longer; hardly a critical difference. The Port Tunnel/M1 route is only, on average, about 1.5 km (one mile) east of the essentially parallel Metro route. With no stops between the Port Tunnel Liffey entrance and Dublin Airport, the BRT journey time from the City Centre to Dublin Airport and Swords would be similar to the Metro, which is expected to take 25 minutes from Swords to O’Connell Street.[60] The current route 142 bus service using the Port Tunnel takes only five minutes end to end of the tunnel.[61] That would suggest the BRT option might even better the Metro time. In his evidence Mr Ian Byrne told us that taxis to Dublin Airport use the Port Tunnel and M1. They therefore find it an attractive and efficient route option.
O’Reilly Consultants dismissed the Curitiba BRT option, largely on the grounds that it would take up road space and would be slower than the Metro, but they did not consider its using the Dublin Port Tunnel, which would largely negate both those concerns. Had they considered costs and benefits they might not have dismissed it so casually. The BRT system may have a capital cost greater than our higher estimate of €200 million. However, that could be further multiplied by a factor of six before the BCR drops to 1 and the IRR gets down to the Finance minimum rate of return of 4%. The BRT option is only a fraction of the capital cost of the Metro option, because busways are much cheaper and because the Port Tunnel already exists. Therefore, because it could generate half of the patronage and hence about half the benefits of the Metro, the BRT alternative appears to be a much more viable and sustainable option in economic, social and environmental terms than the proposed Metro North.
Our purpose is not to do a rigorous assessment here, but to provide a brief sketch of an option that looks very attractive, but has not been considered or assessed by the RPA, because the RPA believes buses can only carry 2,000 per hour per direction. That is odd given that the BRT system is said to have a capacity ten to twenty times that figure, and has been the focus of much attention among transport specialists for its innovation. Our point is to illustrate that an assessment of alternatives, such as a fast, high capacity bus service from the centre of Dublin using a dedicated lane/busway through the Port Tunnel and onto the M1 to Dublin Airport and Swords, could deliver half the proposed Metro benefits, at very little capital cost. Further, there are serious doubts about the need for the carrying capacity designed into the proposed Metro, given the more modest future for passenger growth at Dublin Airport and for population growth in the Greater Dublin Area and MNEC including Swords. A capacity of 20,000 each way is unlikely to be needed in the next 30 years.
The BRT system also offers the option of scalability. The system could provide up to and possibly beyond the same maximum capacity as has been set for the Metro. Yet, if such capacity is not needed, the BRT service can be run at a lower level in response to actual demand and is easily expanded later, if necessary, for negligible extra capital cost. Costs are dominated by operating costs rather than initial capital costs.
With the Metro option, it is all or nothing. We have to incur the massive capital costs, including disruption during construction, even if the numbers never justify it. If such a low-cost bus scheme proved feasible, and it could be delivered in a year or two, it is then up to Metro advocates to demonstrate that their proposal is still justified with this bus system in place. If the BRT system is really capable of delivering the passenger capacity claimed (about 40,000 per hour each way), that would be a difficult challenge. In the present economic and Exchequer climate it would be advisable to consider putting such a BRT system in place at very low cost, and very quickly, while deferring the higher cost Metro option until economic conditions can clearly justify it, if ever.
Further, it would be deplorable to forbid the use of the Port Tunnel to such a BRT service in order to make the Metro option look better on paper. The Metro must stand in its own right with other competitive mass transit options in place, such as the BRT option. By making best use of existing public infrastructure (the Port Tunnel and the M1) thereby maximising the return on that public infrastructure investment, this bus proposal is fully in line with Mr John O’Connor’s injunction to planners.
This case is a good example of the need to ensure we are choosing the best option. Even if the Metro North option was to show positive net benefits and an adequate IRR (4%), it would be foolish to proceed with it, if another option, delivering half the same benefits, but at massively lower cost, and a much higher rate of return was available. This underlines why it is important for the Board to allow CBA evidence, as it will naturally force the debate towards the best option rather than a merely adequate proposal that barely meets the planning requirements.
It would be a shameful failure of governance and potentially huge loss of national welfare to precede with the Metro proposal without having subjected it, and alternatives such a BRT system, to a full Cost Benefit Analysis. I urge the Inspector and the Board to insist on the quantitative evidence that this proposal is a sustainable development and to refuse permission in the absence of that evidence and in the light of the arguments we have made which raise serious doubts about the socio-economic case made so far.
Endnotes
[1] See: http://www.finance.gov.ie/documents/publicaIRRtions/other/capappguide05.pdf
[2] “Value for money is also not simply an item to be ticked in project appraisal. Rather it is the outcome of a carefully considered appraisal system and culture that takes into account, as objectively as possible, the overall benefits and costs of a given project and seek to make sure that budget estimates are met….Currently, Department of Finance Guidelines require that projects with value in excess of €50 million are subject to full scale cost benefit appraisal. I believe it is appropriate to reduce the project value level to €30 million. This approach will include identification and carefully quantified analysis of all the relevant project costs and benefits, including indirect costs as well as the identification of any risks of cost escalation. The principle of cost-benefit analysis is that a project is only desirable if the benefits exceed the corresponding costs.” See: http://www.finance.gov.ie/viewdoc.asp?DocID=3561
[3] Reported in The Sunday Independent 39 March 2009.
[4]Evaluating Transport 21 at: http://www.esri.ie/UserFiles/publications/20061220145614/QEC2006Win_SA_Barrett%20.pdf
[5] Dr Garret Fitzgerald, “State attempting to do too much, too quickly.” Irish Times, 4 November 2006.
[6] http://www.manipuronline.com/Features/May2002/shadowpricing14_1.htm
[7] This definition comes from the 1987 report of the UN established Brundtland Commission, “Our Common Future.”
[8] See: http://www.finance.gov.ie/documents/publications/other/capappguide05.pdf
[9] “Evaluation is a requirement of the EU Regulations governing the Structural Funds, which are co-financing elements of the National Development Plan (NDP) through the Community Support Framework (CSF). Because the Irish Government is concerned to ensure the best possible return to the resources committed to the Plan, the evaluation arrangements apply to all expenditure under the Plan, regardless of funding source. Ireland has built its NDP evaluation arrangement around the evaluation framework set out in EU Council Regulation No. 1260/99.” See: http://www.ndp.ie/viewtxt.asp?fn=%2Fdocuments%2Foverview%2Fevaluation.htm
[10] http://www.pleanala.ie/news/pressrelease07.pdf
[11] Reply too PCA letter to John O’Connor.
An Bord Pleanála
11 February 2009.
Dear Sir,
The Board has asked me to refer to your letter dated 7th November, 2008.
As a general rule the Board does not comment on the planning issues in particular appeals once they are decided. It is important to avoid reopening issues, to ensure that the planning authority can properly carry out its function such as enforcing the terms and conditions that the Board has included in its order.
In relation to your central point about economic viability, the records in the appeal file show that this issue was before the Board for consideration. The Board only arrived at its decision after it had given full and careful consideration to all of the issues. There has been no change in the Board’s position as set out in the Order that issued in the case.
Regarding the Chairperson’s remarks that you have cited, these are about a different issue. The Chairperson was referring to the need for development to be planned in a manner that makes the maximum use of the large scale investment in infrastructure for example development should be concentrated on public transport corridors so that a good economic return can accrue from such investment. He was also referring to the need to avoid planning decisions that would devalue investment in infrastructure for example, by allowing access onto main roads from private developments, where such access would adversely affect the flow of traffic on the road.
Yours sincerely
Chris Clarke,
Secretary.
[12] See Joint Oireachtas Committee on Transport: http://www.oireachtas.ie/documents/committees29thdail/jct/jt280503.rtf
[13] See: O’Reilly Consultants report: http://www.oireachtas.ie/documents/committees29thdail/jct/metro-report/Report.doc
[14] http://www.irishtimes.com/newspaper/ireland/2007/0807/1186424805489.html
[15] See letter to Myles O ‘Reilly from Rory O’Connor of RPA, 23 April 2004.
http://www.oireachtas.ie/documents/committees29thdail/jct/metro-report/correpsondence.pdf
“All of the information listed in your letter was supplied in confidence and on your explicit undertaking to keep the information confidential. It should not be included in any report which is made public. Any such report should also not include any commentary on the information given in confidence,which could reveal the nature of the information.
The information consists of extracts from the Outline Business Case (OBC) for the metro project and supplementary reports. The OBC and supplemental reports were prepared for the RPA for the confidential consideration and use of the Government.”
[16] http://www.transport21.ie/Publications/upload/File/FOI/Dublin_Metro_Project_Outline_Business_Case.pdf
[17] Dublin Metro project Revised proposal, RPA, no date. See: http://www.transport21.ie/Publications/upload/File/FOI/Dublin_Metro_Project_Revised_Proposal_June_2003.pdf
[18] There is a large error in the Swords 2016 southbound data, where passengers boarding are less than half what the need to be to balance the rest of the chart.
[19] An Article in the Sunday Tribune of 24 May 2009 quotes a government report, which says Metro West and the Luas line to Lucan are to be delayed indefinitely.
[20] If the BCR and NPV of Net Benefits (NB) are given (at 5% discount rate) it follows that the NPV of costs is NB/(BCR-1) hence also the NPV of benefits. Similar deductions can be made for the 3.5% discount rate. Given that we then have separate estimates of the NPVof costs and benefits at two discount rates we can break costs into capital and non-capital items, making some assumptions about their structure, e.g. that capital will be spent in equal tranches in the first five years. If we assume a regular pattern for annual benefits we can estimate those as well.
[21] See Guidelines for the Provision of Infrastructure and Capital Investments through Public Private Partnerships: at : http://www.environ.ie/en/Publications/DevelopmentandHousing/PPP/FileDownLoad,17003,en.doc
[22] http://www.finance.gov.ie/viewdoc.asp?DocID=5746
[23] http://www.transparency.org/content/download/41661/663522
[24] From: Highways and Jobs: “The Uneven Record of Federal Spending and Job Creation” by Prof. Ronald Utt
[25] Flyvberg, Bent, Design by Deception: The Politics of Megaproject Approval, Harvard Design Magazine, Spring/Summer 2005.
[26] http://www.herald.ie/national-news/city-news/cutbacks-will-not-derail-euro4bn-metro-north-plans-1695532.html
[27] See: CSF Evaluation Unit, Proposed Working Rules for Cost-Benefit Analysis , Department of Finance, June, 1999.
http://www.csfinfo.com/documents/publications/evaluation/Workingrules-cost-benefit-analyis.doc
[28] http://www.transport-links.org/transport_links/filearea/publications/1_629_PA3551_1999.pdf
[29] Ian Gilder’s evidence page 10: “Again, using English Partnerships research, I estimate there will be a low level of leakage of benefit outside of the Greater Dublin Area of no more than 10%.”
[30] Proposed Working Rules for Cost-Benefit Analysis, CSF Evaluation Unit June 1999, page 12.
[31] http://www.transport21.ie/MEDIA/Press_Releases/Cullen_launches_Metro_North_public_consultation.html
[32] See: Costs and Benefits of a proposed Metro North, Research Report for Senator Pascal Donoghue, Oireachtas Library and Research Service, Sept. 2007.
[33] In the Irish Independent, “In Business” supplement, Summer 2007 (14 June) Declan Collier, CEO of the Dublin Airport Authority is reported as saying: “Economic studies indicate that airports generally support between 750 and 1,000 jobs per million passengers that travel through the airport”. He went on to imply that an extra 12 million passenger per annum would need 10,000 extra employees. That is 850 employees per million passengers.
[34] http://www.pleanala.ie/casenum/233974.htm, case number PL 06F.233974, FCC File number: F08A/1288
[35] Economic Development Strategy for the Metro North Economic Corridor (MNEC), May 2008. http://www.indecon.ie/download/pdf/fingal_coco_mnec_v1.pdf
[36] Bertaud, Alain, Clearing the air in Atlanta: Transit and smart growth or conventional economics? Jan 11 2003, Revised May 29,2003, See:
http://alain-bertaud.com/AB_Files/AB_Clearing_the_Air_in_Atlanta_JUE1.doc
[37] See: Phil Fouracre and David Maunder, Experiences from Metro Schemes in Developing Countries, http://publications.ksu.edu.sa/Conferences/Business%20Briefing%20World%20Urban%20Economic%20Development%20in%202000/FOURACRE.PDF
[38] Flybjerg, Bent, Cost Overruns and Demand Shortfalls in Urban Rail and Other Infrastructure, Transportation Planning and Technology, February 2007 See: http://flyvbjerg.plan.aau.dk/Publications2007/URBANRAIL61PRINT.pdf
[39] Minister Dempsey was adamant: “I’ve also heard it said that Metro North is just a rail line to the airport and not worth doing. That is totally wrong. The airport will only account for a relatively small proportion of passengers on this line. In fact less than 20% of all trips on Metro North will be to or from the airport.” See Press Release: http://www.transport.ie/viewitem.asp?id=11303&lang=ENG&loc=2347
[40] FCC Press Release of 30/3/2007: “Fingal is officially Ireland’s fastest growing county and the area of Dublin undergoing the greatest expansion.” See: http://www.fdb.ie/PR_Census06_final_300307.doc
[41] Minister Brain Lenihan said the proportion of foreign nationals is “probably more likely to be 13 to 15 per cent”. See: Irish Times, September 17, 2007. This is a national figure but it is likely to be an understatement for the Fingal area, which showed the highest population growth between 2002 and 2006.
[42] Jim O’Leary, Business Agenda, Irish Times, 18 September 2009.
[43] Migration and Population Estimates, CSO, September 2009.
See: http://www.cso.ie/releasespublications/documents/population/current/popmig.pdf
[44] See Press Release: http://www.transport.ie/viewitem.asp?id=11303&lang=ENG&loc=2347
[45] World Economic Outlook, Crisis and Recovery, International Monetary Fund, April 2009. See: http://www.imf.org/external/pubs/ft/weo/2009/01/pdf/text.pdf
[46] The Central Bank in its latest Quarterly Bulletin April 2009, says: “House completions declined by about one third last year to under 52,000 units and an even sharper decline of almost two-thirds to about 18,000 units is projected for 2009. Housing starts, which lead completions by about a year, point to a further one-third decline to about 12,000 units next year.” See: http://www.centralbank.ie/frame_main.asp?pg=nws_article.asp%3Fid%3D441&nv=nws_nav.asp
[47] Review of the Construction Industry 2007 and Outlook 2008–2010, DKM Economic Consultants September 2008. http://www.environ.ie/en/Publications/StatisticsandRegularPublications/ConstructionIndustryStatistics/FileDownLoad,18630,en.pdf
[48] Construction costs indices suggests growth of 19% from 2002 to end of 2008. See: http://www.akc.ie/index.php?pageID=283 Costs have been falling recently and tender prices are said to be back to 2000 levels, indicating very competitive conditions and some biding below cost. See: Construction industry Indicators, DKM, March 2009,
http://www.environ.ie/en/Publications/StatisticsandRegularPublications/ConstructionIndustryStatistics/FileDownLoad,19736,en.pdf This suggests that contractors are cutting their margins and absorbing cost increases. It is possible that in this environment, contracted costs may be more in line with the evolution in tender prices than construction costs. In such a case our allowance for cost inflation may be too high.
[49] The claim by Senator Paschal Donohoe that the amortised capital costs of the Metro would be about €22 per ride is not correct. First of all, the €5 billion cost referred to is almost certainly not the capital cost but an early estimate of the total funding requirement in current terms. The capital cost will be somewhere around €2.5 billion. Using the 4% discount rate required by the Department of Finance the annual charge over a thirty-year lifetime would be €148 million. If we start with 40 million passengers a year that would be €3.6 per trip, not €22. If passengers rise to 120 million per annum the capital charge would be €1.20 per ride. See Senator Donohoe’s report. “Costs and Benefits of a proposed Metro North”, Oireachtas Library & Research Service, Enq. No. 2007-1962, September 2007.
[50] Flyvbjerg, Bent; Garbuio, Massimo; Lovallo Dan, Delusion and Deception in Large Infrastructure Projects: Two Models for Explaining and Prevent Executive Disaster, California Management Review, Winter 2009.
[51] A Platform for Change, page 27. See: http://www.dto.ie/platform1.pdf
[52] See: http://www.hindu.com/mag/2009/03/15/stories/2009031550130400.htm, and: http://www.fta.dot.gov/assistance/technology/research_4391.html
See also the movie here: http://www.streetfilms.org/archives/curitibas-brt/
[53] This article by Brendan Keenan notes that “it had been shown in other cities that dedicated busways could carry five times as many passengers as the Luas 6,000 per hour,..” http://www.independent.ie/business/irish/luas-system-too-costly-and-limited-73493.html
[54] See chart presented by Mr David King to oral hearing on 1 April 2009.
[55] http://www.publictransit.us/ptlibrary/specialreports/sr1.curitibaBRT.pdf
[56] http://www.village.ie/Election_07/National_Issues/Election_scandals:_the_Dublin_Port_Tunnel/
[57] http://archives.tcm.ie/businesspost/2007/09/02/story26274.asp
[58] See: http://www.railusers.ie/transport21/costs.php
[59] See: “Do No Harm”, submission by Jerome Casey for the DCBA, Dublin City Business Association.
[60] Evidence of Mr Rory O’Connor, paragraph 2.4.6.
admin
james
planbetter