|
|
||||||||||||||||||
|
Comment
on CEPA report: "Cost Benefit Analysis
UPROAR is a subcommittee of the Portmarnock Community Association (PCA) and we are opposed to the unsustainable expansion of Dublin Airport. We have appealed to An Bord Pleanála the decisions by Fingal County Council to grant planning permission for a new parallel runway (R2) and a new terminal (T2) at Dublin Airport. Part of our case has been our insistence that government guidelines for the appraisal of public investment projects have not been followed. These require that a Cost Benefit Analysis (CBA) be carried out for major public projects. In the absence of such an analysis we carried out our own analysis which shows that Dublin Airport's current expansion plan will lead to a national waste of €4.5 billion whereas a second airport for the Dublin Region (one of a number of options) would yield at least the 7.4% rate of return on investment required by the Commission for Aviation Regulation (CAR). Our economic case against R2 and T2 can be read at: www.norunway.com/bp/bp.htm and www.norunway.com/t2a/appt2.htm. We welcome this initiative by the CAR to engage reputable economic policy consultants to apply CBA to the investment plans of Dublin Airport.
A number of documents are relevant to this comment. The following are made available on the CAR website at: http://www.aviationreg.ie/Commission_Paper_2007.HTML On the CAR website these documents are titled as :
The CAR and economic efficiency. As part of its "Public Consultation on Dublin Airport Charges" the CAR has published the above CEPA CBA report. The CAR has always stressed economic efficiency as a principle of its regulation of airport charges and is required by statute to apply it. One of the CAR's stated objectives is: "to facilitate the efficient and economic development and operation of Dublin Airport which meet the requirements of current and prospective users of Dublin Airport." In a press release in 2001(1) the Commission chided Aer Rianta, inter alia, for: "inadequate or non-existent cost-benefit-analysis or business cases undertaken to justify specific CAPEX projects." In its draft determination of airport charges in 2005 the CAR says that: "an assessment as to the required CAPEX programme and its efficiency is…….a central element of the economic regulation of airports."(2) However this analysis could not then be undertaken. CP1/2007 page 6 says: "A revised CIP in September 2005 was submitted too late to permit the CAR the necessary time to analyse the plan against the statutory objective of economic efficiency." It appears that in engaging CEPA, the CAR is endeavouring to undertake that analysis. As this CEPA CBA document says, it goes beyond the time period of the interim review in including the proposed new runway in its analysis. It also states that its primary consideration is linked to the optimal timing of projects. Because a CBA may show a project to be economically non-viable, it would imply in those cases the optimal timing would require never undertaking it. Presumably the model allows for that possibility. The paper's results are presented as a "first-cut", and interested parties are invited to use the published model to determine more credible estimates(3). We take it that the invitation for submissions to the questions asked in CP1/2007, extends to the many issues raised this CEPA paper.
The effort is welcome and offers the opportunity for a better quality debate to take place on the economic merits of Dublin Airport's expansion proposals. The report recognises that DAA's CAPEX plan is subject to Cost Benefit Analysis under the Department of Finance guidelines and that the Department of Transport's own guidelines (possibly the NRA guidelines(4) ) are also relevant. The authors also points to the UK's Green Book as a guide. The report acknowledges that no Cost Benefit Analysis has been done for either Terminal 2 or the new parallel runway as required under various guidelines. As the Dublin Airport's expansion plan is included in the new National Development Programme (NDP 2007-2013) it is now also subject to EU appraisal guidelines and oversight by a Central Monitoring Committee, and a new Central Expenditure Evaluation Unit(5,6). As a consequence, this work by the CAR may be an important input into that process. In our view this CEPA analysis has not followed Cost Benefit Analysis methodology as laid down in these guidelines. However, it does say its results are "tentative" and "must be treated as indicative or first-cut views." Taken in that light, we see it as a significant step forward in our demands to have the DAA's expansion plans properly analysed and we expect that, as it is developed, the model will improve methodologically and in the robustness of some the assumptions made. The intention seems to be to build a model and make it available to "interested parties" to use to test alternative assumptions. It is not clear if that is the final or an interim objective of the exercise. In other words, will CEPA produce a final result or end its work with the provision of a working model based on a "robust but pragmatic methodology" for others, e.g. the CAR, to use?
The only alternative considered to the proposed expansion of Dublin Airport was one in which the proposed expansion does not take place, passengers are delayed and other displaced passengers decide either to use Belfast Airport, sea-ferry travel, or not to travel at all. The alternative of building a second airport in the Greater Dublin Area (GDA) to meet growing demand was not considered. This is a major flaw. The model takes as a baseline the DAA's "do nothing" or constrained scenario and the main benefits of the proposed development are derived from the time saved by the extra passengers neither delayed nor displaced if the project goes ahead in the unconstrained scenario. The main benefits of proceeding with the proposal were measured in terms of time saved over what would be the case if travellers who use a congested Dublin Airport were delayed and travellers who are displaced from Dublin Airport took some combination of the three other travel options. It may be legitimate to attempt to measure benefits by way of time saved, but it is inevitable that a proposal will look good if the alternatives chosen excludes an obvious alternative and leaves only poor ones. The stipulations of the Finance Guidelines about the need for objectivity in the choice of alternatives are apposite here(7). All of this time, and probably more, could be saved if the option to develop a new Greater Dublin Area airport was considered as one of the alternatives. Neither was the option to improve facilities at the other state airports, Cork and Shannon, considered although the Belfast Airport option might have been taken as a proxy for that option. It obviously does not completely represent that option. In effect, the only alternative considered was one that would inevitably show large time delay costs compared to the DAA's proposal. By measuring all delays relative to Dublin Airport the analysis is biased, ab initio, in favour of Dublin Airport. It does not amount to a consideration of alternatives at all. It asks only if the investment proposed could be justified in terms of time saved. It does not ask what the best option is to achieve the fundamental objective of accommodating additional air passengers nationally, and not necessarily at Dublin Airport. To see this, we should ask what would be the decision if the result was unfavourable. We would conclude that we should not undertake the expansion plan at all, as it was not worth it and we should do nothing further. That would be a bad decision if other options were available that would prove viable but had not been assessed. The aim must always be to find the best option. This is facilitated by stating the objective of the exercise in a clear manner that does not in itself bias the analysis in favour of a preferred option. It is of course also possible that no option would prove viable and no expansion ought to be undertaken. Another
alternative that should have been considered is that of extending the
existing northern runway 11/29 at Dublin Airport(8,9).
This option was rejected early by the DAA on false cost grounds, in
our opinion, which is precisely why it needs to be re-examined in a
more rigorous CBA framework. In rejecting the option the DAA's consultants
ignored the value of perhaps 500 acres of land that would have been
saved if it were chosen over the parallel runway, and wrongly concluded
that the parallel runway was the cheaper option. They incorrectly ignored
the value of their existing land assets and valued only some land that
they would have to acquire. A later FCC consultant introduced a spurious
safety argument that the earlier consultants had not used when they
rejected the proposal on cost grounds. In its application for planning
approval to FCC, the DAA used that false cost argument to reject the
11/29 option, not a safety argument. This 11/29 option needs to be properly
analysed in a CBA context. Land The second major flaw in the model as so far developed and applied, is the failure to take account of the opportunity cost of land. However, it is acknowledged that to do so would have a "huge impact" on the results and it is proposed that it be sensitivity tested. It must rather become a core component of the analysis, with sensitivity testing if necessary around some realistic mean value greatly different from the zero assumed in the present model. It is also stated that the opportunity cost of land could be low because it is already owned by the airport and has no alternative use, (page 7). This is fundamentally wrong, as argued in the attached document: "Valuing Dublin Airport Land." It is contradicted by all relevant appraisal guidelines and by international precedent which are explicit on the need to use market opportunity cost in these cases. If accepted, the implication would be that any assessment, either of a public project using CBA/CEA or of a private project using commercial project assessment tools, would always be biased in favour of the sub-optimal exploitation of existing owned land assets, regardless of their true opportunity cost. By assuming no alternative use, this argument also assumes, wrongly, that existing land zoning is a definitive determinant of opportunity cost. If it were correct to ignore opportunity cost, none of the multitude of private sector redevelopments presently taking place in the case of hotels, pubs, petrol stations, football grounds, etc., on the basis of a realistic assessment of the market value of these sites, would be taking place, It is essential that the model assume a land value based on a full market opportunity cost of about €2 million an acre(10,11). Such a valuation can be subjected to sensitivity testing over a reasonable range around a central value of that order. That is the probable value of land destined for runway use at Dublin Airport(12). Under the standard "do nothing" alternative, which is the basic alternative assumption in all appraisal, that land, ipso facto, would not be needed for a runway and could be put to its best alternative use which would value it at about €2 million per acre. Rezoning is not an issue as the Dublin Airport Masterplan allows for land in the Designated Airport Area to be put to non-aviation-related (commercial) use. Neither has rezoning proved to be a problem for local councils in the past when faced with clearly advantageous proposals to rezone. The UK's Green Book is clear that current zoning does not determine land value(13). This failure to value land properly, biases the results of the CEPA analysis in favour of the Dublin Airport option to the detriment of other options where much cheaper land is available. It is notable that the CEPA text says that even a low opportunity cost of land would "unambiguously increase the cost of providing increased capacity." This is true, but it is even more certain that a realistic market valuation would undermine the case for the DAA's proposal. If the analysis is to be "evidence-based" as the regulator states in his presentation, the overwhelming arguments that exist in favour of a market valuation of land, including existing state-owned land, cannot be excluded from the core of this analysis.
There is a suggestion that job creation has been counted as a benefit but it is not clear what has been assumed, if anything. There are many reasons to question the value of such job accounting exercises as they usually fail to demonstrate that the jobs benefits are either truly additional or additional to what would be created by alternative investment(14).
Cost Benefit Analysis is a social accounting framework. I.e. it must include costs and benefits to all affected sectors of society, public and private. In this case there was no accounting for the many social costs associated with the development. As such, this study does not qualify as a CBA even though it says (page 1) that a CBA can help illuminate a core issue such as: "Whether the investment is justified through the broader social and economic impacts it entails." These broad social impacts include the extra costs to near-airport communities of noise, hazard, air pollution, flooding, health damage and damage to children's education, for example. They must also include the impact on the value of private land whose development is to be restricted under the new flight path of R2. This cost has been put by three local valuers at a conservative €0.5 million per acre for some 3,500 acres under the new flightpath. It is likely that most of these costs would be avoided by the development of a new GDA airport. Climate change costs were not estimated either. The climate change cost of the expansion has been estimated by UPROAR at €8.4 billion using the Stern Report(15). It can be argued that climate change costs would be similar for all aviation-based options, but not for those involving greater use of other "greener" forms of transport and less air travel. For example, while similar climate change costs would apply to a new airport, they would not apply if air travelling passenger numbers are reduced under the constrained option. The result would be less air travel and less global warming, estimated to cost at least €17 per Dublin Airport passenger(16). The benefits of the Dublin Airport option, as analysed, are overstated by this omission. Another huge cost not included is the cost of traffic congestion, usually valued as a loss of time for road users, in the same way as reduced terminal congestion costs and other time saved by passengers was taken to be a benefit in this analysis. The road congestion cost should also include the cost of delays caused at Dublin Airport itself whose users will be one of the main victims of increased traffic delays in accessing the airport. If it is legitimate to count as a benefit the time saved by reduced congestion due to a new runway and terminal, it is equally valid to include costs for the extra delay in accessing the airport due to the extra cars arising from this expansion having to use a critically balanced road network(17). UPROAR's consultant found that the runway development would make car journey times to and from the airport unacceptably long and unpredictable due to the congested road network(18). The Metro and Airport Box (roads) proposals will help make a bad situation less bad but will not solve the congestion problems of the expansion. The NRA objected to the IKEA proposal on the grounds that its car load would threaten the €1.1 billion upgrade of the M50. The DAA's expansion proposal would add six times the worst IKEA daily car load to the local road network, including the M50, by 2025. The expansion by 2035 would add twelve times the IKEA daily load(19). Again, most of the associated congestion costs would be avoided if any needed aviation facilities were developed elsewhere, such as a new airport on a greenfield site. The CEPA analysis also adds benefits for an improved quality experience. That should also be done for a new state-of-the-art airport where it would show that such benefits would apply even more to its passengers. According to Annex 7 results are quite sensitive to these quality assumptions. It is notable that these benefits are presented as the opposite of airport congestion, the presumption being that the new facilities will only lead to an improved quality of service. The evidence is tenuously based on low/cost high cost differentials exclusively related to terminal service. If the road traffic impacts of this expansion at Dublin Airport were also considered and the inevitable loss of quality of service arising from the thousands of extra car trips to and from the airport were considered, these quality gains could be totally swamped. If congestion costs were properly assessed it is possible that such quality of access factors would be captured, but using standard time-based methods that is not self-evident and it appears CEPA counted such benefits in addition to the benefits of reduced time delays. Before any such tenuous terminal service quality benefits are allowed, a full assessment of the additional costs of a deterioration in the quality of service due to impeded road access to the airport as a result of this expansion, should also be included.
A key principle of the CAR's mandate is that of economic efficiency and it can be assumed therefore that the economic efficiency of Dublin Airport's investment is to be examined by the CAR in a longer-term context than that delineated by the investments to be made over the 2006-2009 period which are the focus of the revised charges determination. One reason for this is that the investments planned in the 2006-2009 period are only part of a total package of investments. The CAR's review procedure in train is that of its 2005 determination for the four-year period 2006-2009. This has to be done in the light of "the revised capital programme that the DAA released in October 2006", CIP2006. This document specifies an investment of €1.178 billion in the period 2006-2009 involving improvements to T1 and phase 1 of a new T2, all to be completed in 2009. It does not include the proposed new parallel runway (R2) and does not indicate any investment after 2009, such as Phase 2 of T2 due about 2016. The CAR in its CP1/2007 document provides some graphical information for DAA investments after 2009 (Charts 2.1 and 3.1). Chart 2.1 gives yearly figures in bar-chart form which add up to about €1.95 million by 2014. A similar, but not identical chart "Evolution of the capital programme" in the IMR paper, page 3, gives a similar total. These are broadly consistent with the DAA's total of €2 billion and with other sources, but detail is lacking(20,21). It appears that not all the DAA's investment data given to the CAR has been released. DAA's has stated that it intends to invest some €2 billion over the next ten years including the new runway(22,23). It is not clear what the DAA's €2 billion includes in detail. It is unlikely to include anything for Phase 2 of Terminal 2, needed about 2015/16. Further investments are also planned under the Fingal County Council's Dublin Airport Masterplan, which inter alia, foresees a third terminal on a new western campus of Dublin Airport. This is not included in the DAA's €2 billion CAPEX programme, as it is outside the ten year period. On DAA passenger figures, it will be required about 2020 and, with associated infrastructure, will cost something of the order of €500 million in today's money terms(24). (It may also be privately built). It will be the final major component of infrastructure that will allow airport capacity to increase from 35 million around 2020, to a maximum of about 60 million around 2035(25). It would be very helpful to have a clear numerical, chronological and descriptive breakdown of all planned investments over the next fifteen years within which time it is expected all elements of the Dublin Airport Masterplan will be in place. This total investment package, amounting to about €2.5 billion, will take Dublin Airport to its maximum capacity of 55-60 million passengers per year around about the year 2035. This is probably an underestimate given the DAA's track record for its estimates. For example, the figures provided by the DAA to the CAR doubled between September 2005 and October 2006 (see CP1/2007 page 6). The DAA and FCC argue that the "Airport Box" road project and the proposed Metro will help make Dublin Airport's expansion proposal workable. If they are correct in that claim, it follows that most of the Airport Box expenditure of €200 million should be counted as part of the cost of that proposal, as should a substantial share of the proposed Metro currently estmated at between €3.5 and €4 billion. That share can be estimated from the projected share of airport passengers using the Metro. Of course, if such additional transport infrastructure is necessary to make a second airport work, it should also be included in an assessment of that alternative. In our view this package as a whole should be the basis of the CAR's economic analysis. If not, there is a serious danger of project splitting, whereby each element of an investment plan, taken individually, may not be seen in the wider context in which its inherent unsustainability would be more clearly evident. The investments cannot be properly assessed unless analysed as a complete package. It is also a requirement of the EIA Directive that "project splitting" be avoided. See UPROAR's "Legal Submissions" to An Bord Pleanála(26). Another case for a comprehensive approach is to ensure consistency in the benefits claimed. It is essential that the stream of benefits claimed for the investments are causally linked to the planned investments (see below).
The title of CEPA paper is "Cost Benefit Analysis of Terminal 2 and Runway 2 at Dublin Airport" and their analysis only includes €951 million of CAPEX. The €951 million is made up of €757 million for T2, €150 million for R2 and €44 million for a new ATC tower (Annex 6). This is €227 million less that the €1.178 billion to be spent in the 2006-2009 period that is the subject of the current review of charges. It does not include Pier D and the extension to Terminal 1. It is €1 billion less than the €2 billion the DAA stated that it will invest over the next 10 years. And it is €1.5 billion less than the probable total Masterplan package of €2.5 billion over the next 15 years. How sensitive are the results of this analysis to this understatement of CAPEX? This is an unacceptable basis for the assessment of the economic efficiency of the DAA's CAPEX programme. The analysis cannot be confined artificially to just two (or 3) elements of a much larger CAPEX plan. The CAR's brief was to analysis the revised CIP of October 2006. As pointed out above, that amounts to €1.2 billion by 2009 and nearly €2 billion by 2014. For consistency, we believe that, at least, the full Airport Masterplan package of planned investments amounting to some €2.5 billion should have been analysed by reference to passenger numbers planned to reach a maximum of 60 million by 2035(27). A share of other necessary transport infrastructure (roads and Metro) should be added. Investments and their assumed consequences must be cognate in order for benefits to be reliably assessed and correctly attributed to the originating investment. If not, the benefits claimed may be exaggerated. While the CAPEX of less than €1 billion included in the CEPA model is much too low, the benefits credited to it are consequently much too high. CEPA has confined itself to an analysis of an investment of T2 and R2 only. In our opinion it is impossible to separate that investment from the total investment package of which it is only a part, much less than half (38%). First of all, it is impossible to envisage these selected investments functioning without much of the other related investments, and secondly it is impossible to separate out the stream of passenger-related benefits that would accrue uniquely to those selected investments, from the benefits that may accrue to all passengers. For consistency one has to be able to demonstrate that the benefits claimed were those due only to those selected and essentially inseparable investments and, as the benefits of the investments are determined by associated passengers, the stream of passenger benefits claimed must correspond causally, no more no less, to the investments made. The DAA projected passengers per annum to 28.4 million by 2025 if the airport is constrained by lack of investment, and to 38.4 million by 2025 if unconstrained. Although the projections were initially developed in the context of the new runway EIS, the unconstrained outcome with its 10 million extra passengers was not due uniquely to the provision of a new runway. The provision of the runway is presumed to be critical, but other investments would obviously be needed. As the R2 EIS says: "This cannot happen if the growth of Dublin Airport is constrained by a shortage of runway, terminal or other on-site capacity or by off-site surface access congestion."(28) It is clear the unconstrained projection involved much more than the runway investment. It follows that the benefits to extra passengers and improvements in all passengers' welfare in the unconstrained scenario must be attributed to the full package of investments in the preceding period. At the very least, it would appear the entirety of the €2 billion CAPEX to be undertaken over the next ten year period is involved. It will be fully in place ten years before 2025 when these extra 10 million passengers are expected. Otherwise, it would have to be assumed that about half of the €2 billion CAPEX will have no affect on passenger numbers or their welfare. More properly, it should be the full €2.5 billion to be spent by 2020 as envisaged by the Masterplan. It will also be in place well before 2025. However, it seems the CEPA CBA model counted all the cost-saving benefits of all the passengers who are not delayed or displaced if the project goes ahead and attributed those benefits to an investment of less than €1 billion(29,30). If some of the full €2.5 billion CAPEX is to be excluded from the analysis, then some of those passenger benefits must also be excluded. Alternatively, some clear explanation is needed as to why some €1.5 billion of investment will not increase passenger numbers or contribute to an increase in their welfare, as measured in this exercise. If such an explanation can be given, it would strongly suggest that the €1.5 billion investment excluded will have no benefits and should not therefore be undertaken at all. To put this another way, in the R2 EIS the unconstrained scenario envisaged 38.4 million passengers per year by 2025 (centreline forecast), 10 million more than in the constrained scenario. If that projection did not depend on the full €2.5 billion to be spent by about 2020, we have to presume that numbers will be even higher than so far projected by the DAA, unless that extra €1.5 billion investment is to have no effect on passenger numbers at all. But, as the DAA persists with its position that passenger numbers will be about 38/39 million per annum by 2025, and the Masterplan itself agrees with that, we have to take it that all of the €2.5 billion CAPEX of the Masterplan is inseparably linked to those 38/39 million passengers by 2025(31). CEPA based its passenger benefits on that unconstrained figure, it therefore should have included all of the associated CAPEX as a cost. We do not know exactly how the model treated passenger growth over the full period of analysis and the extent to which future benefits may have been overstated by reference to the clearly understated investments included. The model was to have been made available for inspection but has not been. For example, it is not clear for how many years into the future the analysis is continued. It is necessary to make a distinction between the analysis period and the investment period. The investment period may be well defined in time with a clear end point. The analysis period, on the other hand, should extend for the lifetime of the assets being invested, and if, as in the CEPA analysis, those assets are assumed to be maintained indefinitely, the analysis period should also be indefinite. There is a reference to a "perpetual project" (page 3) which is consistent with the assumption made that capital will be replaced at a rate of 2.5% per annum (page 8). Whatever the period of analysis, the benefits claimed must correspond to the investment made. That does not appear to have been the case here. This says no more than that while the investment may have well defined costs in a fixed time period, the benefits and other costs which result may continue for a considerable time, and need to be properly accounted for in a sound analysis. Passenger forecasts DAA's passenger forecasts have been hard to pin down and DAA sources have given quite different figures even in recent months. What appears to be the latest DAA passenger forecast was made available late on the CAR website. The paper by Coveney and Butler, updates earlier work by the same authors and is dated April 2006 with textual updates in August 2006. It updates earlier forecasts in the light of the recent jump in growth, especially in 2006. However, it only goes as far as 2015 and does not cover much of the period of the CEPA analysis which depends critically on differential passenger forecasts, constrained and unconstrained, at least until 2025. The authors do not appear to have updated that analysis and one has to suppose that CEPA is depending on the original work done for the runway EIS in 2004(32). It hardly needs to be said that this limitation raises serious questions about the reliability of an analysis based on such out-of-date material. We presume that the constrained/unconstrained analysis will be updated to help this CEPA work advance to a more complete stage. The graphical analysis is presented with the underlying assumption that there will be a break-even point or range which might be moved somewhat either way in time, under sensitivity testing. It is stated that the "primary consideration of this report is linked to the optimal timing of the project". This assertion excludes the possibility that the project was basically non-viable and that no such point would ever be reached. That is very likely to be the case if the missed costs as listed above were included. No Cost Benefit ratios or IRRs were produced(33). The time scheduling analysis is of interest but it seems to be exaggerating the precision of this model. It is very doubtful if the sensitivity of this model to such obviously missed assumptions would allow any really meaningful conclusions to be drawn from this scheduling analysis. The results appear to be very sensitive to assumptions. When an additional €180 million (19%) was added to CAPEX for works associated with T2, the "break-even" date (under a high benefit assumption) was shifted from 2013 to 2018. Under the low benefit assumption it gets pushed out to about 2023. CEPA also admits that it assumed no CAPEX cost overrun (Annex 6) when the UK Treasury recommends a 40% factor. They rightly recommend that this scenario be tested. It would obviously make the results much worse. Annex 8 provides some "Traditional CBA Results". This Annex is hard to interpret (lots of negative NPVs, "n/a"s and "2015+"s) but it suggests that the case for the T2 and R2 investments is precariously balanced, even with the many cost omissions and overstated benefits. It is clear that including the costs we believe have been excluded, would totally swamp the results and postpone the "break-even" date indefinitely. For example Table 3.1 has the understated CAPEX of €951 million and no opportunity cost of land. Taking that land factor alone and valuing the 840 acres of airport land at the going rate of €2 million an acre would add another €1.7 billion to the cost. That alone would probably produce a negative rate of return before adding other costs such as: the understatement of CAPEX of some €1.5 billion, the loss of value of flightpath land, the costs to communities and congestion costs. The CSF (Department of Finance) also recommends a 50% penalty for the use of public funds. That would have to be applied at least to the state-owned land to be consumed by the runway (€1.7 billion). This time scheduling approach can be used to present poor results in a more palatable form and avoid coming to a blunt conclusion on the DAA's proposal itself. It would be very helpful to have a clear statement of the CBA results for the timing envisaged by the DAA, with all the components of the proposed expansion included This would be as shown in Chart 3.1 in CP1/2007 for investments up to 2015. What are the NPVs, Benefit/Cost ratio and IRR for the investment plan as actually proposed by the DAA? All these tentative analysis results say is that the heavily understated costs of construction and operating costs would be covered by the time saved (passenger and airport) if this project goes ahead rather later than planned(34). These gains in time would be also achieved and possibly more so, by an alternative such as a second GDA airport, but with much lower economic, social and environmental costs. Indeed UPROAR has identified (but not quantified) spin-off benefits for a well-sited new airport that would not arise in the case of a congested Dublin Airport. A CBA analysis must also be done of the option to build a second airport. It would have a construction cost similar to the DAA Masterplan investment of €2.5 billion, not surprising as it includes all the main elements of a new airport, without the land. Unlike the Dublin Airport option, however, it would benefit from a low opportunity cost of land and the much lower social and environmental costs. It would very likely prove to be very positive in comparison with the Dublin Airport option, as found by UPROAR. Time benefits
of a new airport, would likely be similar for the average traveller
as between Dublin Airport and a new airport in the Dublin region. If
these timesaving benefits were likely to be similar, it would suggest
that a Cost Effectiveness Analysis (CEA) may be the more appropriate
than attempting to measure benefits which are very uncertain and difficult
to quantify reliably(35). UPROAR's analysis was essentially
a CEA as discussed in "Note on Runway Cost Benefit Analysis"(36)
As that note argues, the benefits (consumer surplus) arising from both
options for the Greater Dublin Region (expansion of Dublin Airport or
a new GDA airport) would be similar but those benefits are achieved
in the second airport case without the huge land subsidy (taxpayer transfer)
implicit in the passenger charges at Dublin Airport nor the further
economic distortions generated by that subsidy and manifest in congestion
externalities. It follows that UPROAR's finding of a loss of €4.5
billion for the Dublin Airport expansion option is the minimum cost
of the Dublin Airport option relative to the second airport option that
we estimate will yield the regulator's required annual return of 7.4%
at an unsubsidised passenger charge of €6-€7. This is less
than the €7.50 minimum currently being sought by the DAA to fund
only its €1.2 billion 2006-2009 investment plan(37).
CEPA Congestion Paper The CEPA paper on congestion charging confines itself to congestion within the airport itself and excludes consideration of congestion on access roads or rail, or parking. The basic idea is to vary charges over time to smooth demand for services between peak and off-peak times. The paper makes the following important comment on page 1.
UPROAR has pointed out the passenger charges at Dublin Airport are grossly understated by the failure to charge for the use of the land at Dublin Airport at its real opportunity cost value. It is, in effect, a heavily publicly subsidised airport. One consequence of this is that the demand for its services are not a true reflection of the real cost of providing those services and is therefore artificially enhanced. The consequences are manifest in congestion within and without Dublin Airport. The planned expansion of the airport is predicated on a projected future demand which is driven by these subsidised charges and is therefore an illegitimate basis for a determination of the airport's real investment needs. Among the distortions created by this huge subsidy is the elimination of real competition for Dublin Airport, either from other state airports or any private interests that might consider building extra airport facilities. It is obviously in the airlines' interest to fail to reveal their willingness to pay for those subsidised services. Indeed, they continue to demand ever lower charges in their own commercial interest. Ryanair has been very critical of the DAA but when asked if it would not build a competing airport did not give a straight answer(39). The reason is obvious; Ryanair (or anyone else) has no incentive to do so, as long as the subsidisation of Dublin Airport continues. If that subsidy were removed, it is very likely that the case for a competing airport in the GDA would be made. The benefits of such an outcome do not need to be repeated here. On page 4 the paper makes the odd claim that the new parallel runway will support an increase in potential passenger from 20 million to 30 million. That is an error. The CEO of the Dublin Airport has said that the airport's maximum capacity is 60 million passengers(40). The runway, scheduled for completion about 2012, is the key foundation on which the airport's expansion to its potential maximum of 60 mppa, will depend. Catering for this full passenger throughput by about 2035 will require the construction of Terminal 3 in about 2020. As no further runways are planned, that capacity increase is critically dependent on the proposed parallel runway. The analysis of congestion charges applied to the existing runway structure at Dublin Airport finds "limited support for additional runway construction." This is on the grounds "that the net present value of congestion charging revenues from the existing infrastructure is less than the net present value of constructing the new runway, which is approximately €146 million". "Limited support" is a carefully chosen phrase. The analysis actually finds against additional runway construction. The authors go on to say that, as the congestion charge revenue is likely to grow, the case for the new runway gets better over time. That is very doubtful as the real commercial cost of the runway proposal, including the value of Dublin Airport's own land is about €2 billion, it is clear that this analysis shows that such a runway investment can never be justified. Needless to say, the case for the runway is even worse on these grounds if we were to take account of the full net social cost of some €3 billion. It would be of interest to have this analysis repeated with these costs included rather than the wholly inadequate construction cost of €146 million, that itself is probably underestimated(41). The analysts also says that if congestion charging were introduced, "it would postpone the date at which any new runway were needed and provide a much stronger basis for an economic appraisal of new runway capacity." The first point is consistent with our view that the airport is heavily subsidised and if that subsidy were removed and charges were to rise, demand would fall and undermine the need for a new runway. The scope for congestion charging (beyond that justified by peak/off-peak differentiation) is simply a reflection of that hidden subsidy. The second point is a welcome restatement of the need for proper economic appraisal but it is not clear why the imposition of congestion charging would improve a case that is not only self-evident now, but actually a requirement under current Department of Finance and NDP guidelines. Perhaps it is repeating the sentiment expressed on page 1 that airlines' reluctance to reveal their willingness to pay for airport services makes "expansion decisions problematic" and is making the same point that we have just made, that the existence of the scope for congestion charging is indicative of economic distortions which need to be thoroughly investigated as part of a comprehensive analysis. We agree. Our primary concern is with the economic appraisal of the DAA's CAPEX plan. However can make a few observations on the questions raised in CP1/2007 about passenger forecasting. The CAR wonders if the DAA should bear the consequences if its forecasts prove optimistic (CP1/2007 page 4). In our opinion the DAA has an incentive to underestimate passenger numbers on the grounds that the charge determination is on a per passenger basis. We have noted that in its passenger projections recent growth trends drop abruptly at the present, and continue at a lower rate thereafter. This has been seen in a couple of recent instances. It can be seen in the DAA's 2006 Demand Forecast Report, Table 3.1a. Passenger growth in 2006 was 10.6% but it drops sharply to 6.1% in 2007 and continues downward thereafter. These figures are supposedly determined by a model and the authors point out that the data will be smoothed by that process (page 23). This sudden drop in growth in 2007 does not look very smooth. A similar pattern was evident in a chart shown by DAA CEO Declan Collier at the Towards Sustainable Airport Conference in October 2006. The figures were not the same as those provided in the above DAA document but show similar discontinuities and interrupted trends. The primary drivers of traffic growth (Table 1.1) are Economic Growth and Yield (undefined). One wonders what is expected to happen to economic growth per head in Ireland and all those other countries included in the model between 2006 and 2007 to produce such a dramatic fall in passenger growth. If it is not determined by economic growth what is the explanation for such a sharp drop? Has the model output been adjusted in any way? Some explanation is needed. As pointed out above, in our opinion much of this passenger growth is determined by the subsidised operation of Dublin Airport. It seems that passenger charges are not included explicitly in the forecasting model but (we assume) are part of air fares, that are projected to decline slightly on most routes in the short-term. It seems odd that passenger charges are not separately modelled in view of the importance DAA places on those charges and in having them increased to pay for its expansion plans. A 24% minimum increase is being sought at present. If charges were included as an exogenous variable we could test the effects of higher passenger charges on passenger demand, and in particular the effects of removing the subsidy. The subsidy is about €18.5 per passenger with 20 million passengers per annum. As passenger numbers would probably fall with the removal of such a large subsidy, it could be expected that the actual charge per passenger would have to increase even further. Shocking as that might seem, it is nothing more that a consequence of the correction of the current unsustainable operation of Dublin Airport and its expansion plans. Matthew
Harley
(1) See: http://www.aviationreg.ie/images/ContentBuilder/press270801.pdf
(3) We have been unable to locate the published model and requests to the CAR for access to it have not been successful. (4) NRA guidelines: "Guidelines for the Appraisal for Cost Benefit Analysis.", National Roads Authority, St Martin's House, Waterloo Road, Dublin 4, June 2005. (5) "Guide to cost-benefit analysis of investment projects", (Structural Fund-ERDF, Cohesion Fund and ISPA), Evaluation Unit, DG Regional Policy, European Commission, 2002. See: http://ec.europa.eu/regional_policy/sources/docgener/guides/cost/guide02_en.pdf (6) See UPROAR Press Release: http://www.norunway.com/Archive/Media%20Coverage/UROAR%20PR24jan07.htm (7) The 2005 Finance Guidelines say: "Objectives should be expressed in a way which will facilitate consideration and analysis of alternative ways of achieving them. They should not be so expressed as to point to only one solution." The 1994 guidelines say: "Objectivity is important in considering options. There is a danger that the selection of options may be manipulated in order to make a case for a course of action which is already favoured. For example, options for which there is a very weak case may be put forward in order to make a poor option look good." (8) See "Rejecting the Option to Improve Runway 11/29 - the Value of Land Ignored." Chapter 4 of "The Economics of the Proposed Runway", submission to An Bord Pleanála at www.norunway.com/bp/bp.htm. (9)
See "DAA Consideration of Alternatives" at www.norunway.com/alts.doc
which deals with the DAA's failure to properly assess alternatives to
their runway proposal both within Dublin Airport and elsewhere. (11) The Cargobridge land, a parcel of private land completely surrounded by Dublin Airport land, was rezoned unanimously by the County Council in 1993 from agricultural to industrial use. The Irish Independent on 20 September 2006, referring to the Cargobridge land (under investigation at the Mahon Tribunal) says in that in 2000, 8 acres of the Cargobridge land was sold for almost €20 million or €2.5 million per acre seven years ago. This is the best indicator of the value of land at Dublin Airport for its next best use. (12) URPOAR estimates that some 840 acres of land will be absorbed by the runway. This is greater than the 645 acres (261 hectares) mentioned in the R2 EIS but that figure does not include some airport land, particularly at the western end of the new runway, that will be useless for any other purpose if the runway is built and it must therefore be counted as a cost. (13) The UK Government's Green Book on project appraisal advises that the valuation of a site should be based on the most valuable possible use, rather than the highest value that could be obtained for its current use. In the case of land currently subject to planning restrictions, the prospects for a lifting of such restrictions should be taken into account in making the valuation. "In all cases, the prospect for obtaining a higher planning consent should be considered by the appraiser and his professional property advisor." Annex 3, "Land and Buildings" in "Appraisal and Evaluation in Central Government: The Green Book", HM Treasury, London, 1997. See: http://greenbook.treasury.gov.uk/ (14) See: "The Economics of the Proposed Runway", Chapter 3, "Economic Case as Put by the DAA" at www.norunway.com/t2a/bp.htm. (15) See: http://www.norunway.com/Archive/Media%20Coverage/UROAR%20PR12Feb07.htm (16) UPROAR did not include this €8.4 billion cost in its CBA. (17) See UPROAR Press Release on the NRA's selective objection to IKEA: http://www.norunway.com/Archive/Media%20Coverage/UROAR%20PR%2022may06.htm (18) See: http://www.norunway.com/t2a/Road%20Traffic%20Report.pdf (19) Half of the load is due to the extra 10 million passengers in the unconstrained scenario, the other half is due to the assumed 10 million extra passengers arising even in the constrained scenario. See: http://www.norunway.com/Archive/Media%20Coverage/UROAR%20PR%2022may06.htm (20)
At a conference on Sustainable Airport Development in October 3005,
Declan Collier DAA CEO said: "…we are proposing to spend €1.2
billion over the next four to five years with a further €800 (21) The new NDP 2007-2013 has €1.9 billion for all three state airports but it is not broken down. (22) See: http://www.dublinairportauthority.com/media-centre/press-releases/022007.html. (23) The DAA's assumed construction cost of the runway included in the its €2 billion figure is not known. The DAA put it at €141 million to the regulator in 2005. It is probably nearer €200 million. (24) However, Minister Martin Cullen told the Dáil on 7th June, 2006 that Terminal 3 would not be needed until about 2015. That is too early by 5 years, if CEO Declan Collier is to be believed. See: http://debates.oireachtas.ie/Xml/29/DAL20060607A.PDF (25)
There is a multiplicity of estimates of future passenger throughput
from various sources. That is understandable given the complexities
and uncertainties involved. However, it would be very useful if the
DAA were to provide, perhaps online and regularly updated, its latest
opinion on future passenger numbers and the underlying assumptions.
From these varied sources it seems that, with its expansion plans allowed,
the DAA expects passenger numbers per annum will reach about 35 million
by 2020 and 40 million by 2025. For some reason, these projections imply
a rate of growth of about only one million passengers a year which is
much lower than recent growth of about 1.5 million a year. Average assumed
growth is only 3.2% between 2006 and 2025 whereas growth between 2000
and 2006 was actually 7.5% per annum. Assuming only a growth rate of
4.2% p.a., modest by recent trends, we would get to 60 million by 2035.
If the recent trend of 7.5% p.a. continued, 60 mppa would be reached
by 2020! See also "Passenger Forecasts" in text. (27)
UPROAR's CBA assumed a total infrastructural investment of €1.5
billion with passengers reaching a maximum of 55 million by 2040. As
we now know that €1.5 billion is underestimated by about €1
billion, but was the figure available at the time of the analysis. (29)
The CEPA CBA paper page 6 says: "DAA, as part of the planning application
for R2, provided Fingal County Council with a constrained forecast assuming
neither R2 and T2 occur. The annual impact of constraining the airport
has been taken from this report and grafted on to a more recent unconstrained
traffic forecast." Although we are not sure what the more recent
forecast is, we take it that CEPA has used the DAA's figures which assume
about 10 million passengers displaced (30)
The CEPA CBA paper says on page 5 that in using the consumer surplus
approach to measure the cost of displacement it used "the quantity
displaced as per DAA's constrained forecast." We take this to mean
that, for example, the full 10 million passengers displaced by 2025
was taken. (33) In Annex 6 it appears that the cost of remunerating the CAPEX costs have been included as an additional cost. While it is correct to include the cost of maintaining the assets, it is double counting to include the cost of capital. That cost should emerge as an output of the analysis in the form of an IRR. The issue is then how the IRR compares to the guide rate (7.5% or 5%). The project would be judged on that basis as to whether it can remunerate its CAPEX adequately in social terms. It may be however, that the analysts have assumed that the cost of capital should be the regulator's 7.4% and determined the "break-even" date as the key dependent variable. This is a valid alternative to the estimation of an IRR. It is similar to what UPROAR did in its analysis of the second airport. We assumed the 7.4% as the required rate and determined the passenger charge necessary to achieve that rate of return. We found it was about the same as the €6-€7 currently applying at Dublin Airport, but without any land subsidy. (34) CEPA uses a percentage (7% to 10%) of CAPEX to represent operating costs. Perhaps CEPA should use a per passenger operating cost. That would be more transparent and consistent with the method used by the CAR in its charge determination. It would also mean that total operating costs would grow with passenger numbers, as one would expect, and not fluctuate with CAPEX. (35) Consider only the uncertainties facing the aviation industry due to the possible imposition of environmental taxes on aviation fuel, the increasing volatility of fuel prices and the probable upward pressure on those prices. How reliable are any estimates of benefits that depend critically on long-term future passenger numbers? If the same uncertainties apply to both options, the next best criterion for decision making is that of least cost. (36) See: http://www.norunway.com/t2a/CBA%20Note.htm (37) See: http://www.dublinairportauthority.com/media-centre/press-releases/142006.html (38) The EU Cost Benefit Analysis Guidelines are explicit in specifying that historical cost is an inappropriate (page 32): "Many projects in the public sector use capital assets and land, which may be state owned or purchased from the general Government budget. Capital assets, including land, buildings, machinery and natural resources should be valued at their opportunity cost and not at their historical or official accounting value. This has to be done whenever there are alternative options in the use of an asset, and even if it is already owned by the public sector." If historical cost should not be used in project cost-benefit analysis, it should not be used in a charge determination which purports to follow principles of economic efficiency. SEE: "Guide to cost-benefit analysis of investment projects", (Structural Fund-ERDF, Cohesion Fund and ISPA), Evaluation Unit, DG Regional Policy, European Commission, 2002. See: http://ec.europa.eu/regional_policy/sources/docgener/guides/cost/guide02_en.pdf (39) At the Towards Sustainable Airport Development Conference in October 2005, a PCA representative put this issue to Michael O'Leary. The exchange was as follows:
(40) Declan Collier, CEO of DAA gave a maximum capacity figure of 60 million passengers per annum to Pat Kenny on RTE Radio 1, on 21 September 2006. Ciaran Scanlon, DAA programme manager, gave a capacity figure of 55 million to Fingal Independent, 30 August, 2006. (41) Chart 3.1 in CP1/2007 shows a new runway investment of about €170 million. Even discounting half of this at 7.4% we get €164 million as CEPA did to get their €146 million from a two year spend of €150 million.. |
||||||||||||||||||
