High Speed Rail - Transport Committee Contents

Further written evidence from HS2 Action Alliance (HSR 153D)


The note is prepared by HS2 Action Alliance.[237] It raises some points on the further submission to the Transport Select Committee (TSC) by Department for Transport (DfT). It is intended to assist the TSC in their questioning of DfT in providing context for some of the assertions and explanations that they give, particularly when they have offered in our view incomplete, misleading or erroneous responses.

This document has been prepared within four days to be available in advance of DfT's Select Committee appearance.


5.  We have received submissions (eg from 51M) that substantial extra capacity could be provided quickly and at relatively low cost by lengthening WCML trains to 12 car sets, and (especially relevant for commuters from Milton Keynes and Northampton) by operating additional trains enabled by a new grade separated flyover junction eg at Ledburn. Do you have any comments on this?

DfT observe that the additional capacity from the 51M proposal is comparatively low. They fail to note that it provides sufficient capacity in total to allow services to have a lower load factor than HS2, and it is able to comfortably accommodate DfT's projected peak demand (that is not addressed by any price changes in the HS2 Ltd modelling).

It is the cumulative or total increase in capacity that matters, and it is this that should be compared with the increase in demand. As demand is calculated from the 2007-8 base, so should capacity.

6.  What assessment have you made of the costs and benefits of running Pendolinos at up to 140mph by further upgrading the existing line?

The reasoning for the exclusion of consideration of 140mph Pendolinos is defective.

ERTMS level 2 is planned to be implemented on WCML by 2030 as part of the UK's commitment to ERTMS implementation.[238] The motivation for this is:

—  Meeting EU inter-operability requirements.

—  Reducing costs.

—  Increasing line capacity.

ERTMS provides in cab signalling which overcomes the current restriction on speed to 125mph that applies to the existing signalling system. So with ERTMS the current Pendolinos could travel at 140mph - which is their design specification.

There are uncertainties about the timing of this programme, but it is worth noting that:

—  ERTMS level 2 is required for HS2 - so any delivery risks apply equally to HS2.

—  Implementation on WCML is scheduled towards the end of the programme (because the programme is based on when the signalling system is expected to be due for renewal and the last WCML renewal is only recently completed) so the WCML implementation is less likely to be delayed than for implementations scheduled earlier.

It is therefore unreasonable not to have considered the benefits of 140mph running (or indeed of somewhat higher top speeds that would be obtainable when rolling stock needs to be renewed). It is also likely that there would be little or no incremental cost to this faster running.

It is worth noting that while ERTMS is not considered in the WCML RUS by Network Rail, Network Rail did consider it in the ECML RUS (published 2009), with its potential benefits to linespeed, capacity (with a headway reduction from three minutes to two minutes) and performance.

In considering the existing network, Atkins did not take into account the effects of implementing ERTMS despite the UK's commitment to implement it. It is clear that it would substantially reduce the assessed benefits of HS2 if it had.

16.  What analysis have you made of business relocations between London, Birmingham and other major cities likely to arise from HS2?

DfT observe that more than half the journeys on HS2 are expected to originate from outside London and the South East, implying that the origin is where the economic benefits would be felt. This is misleading.

—  70% of journeys will be for leisure purposes, it will be the destination that receives the benefit - ie predominantly London.

—  The improved connectivity that DfT describe is primarily with London, which seems to be almost universally recognised in the academic literature as resulting in the benefit accruing to London.

It is worth commenting on the specific evidence that DfT cite for regional benefits:

—  HS1 - the town of Ashford has seen valuable new investment and development since the arrival of high speed services. Analysis carried out by Volterra and Colin Buchanan has estimated that the value of the regeneration benefits of HS1 could be as high as £10 billion.

This study is not based on the observed effects of HS1, but the forecast benefits according to DfT's methodology. The evidence to date is that many of the presumed benefits have not materialised.[239]

—  Lyon, France - high speed rail has helped service sector firms to thrive by providing enhanced access to the Paris market. The Part-Dieu area where the Lyon station is situated has become one of the largest commercial developments in France.

The high speed rail link has enabled firms to relocate their HQs to Paris from Lyon (see evidence submitted to this Inquiry by Prof John Tomaney (section 4.6), and overall the link is assessed as having a negative impact on Lyon.

—  Ciudad Real, Spain - high speed rail has enabled the town to develop into an important regional business centre and its university to expand its student population.

The balance of evidence in Spain seems to show that Madrid has been the principle beneficiary of faster connectivity (see Tomaney's submission to this Inquiry):

—  The Government is considering a significant amount of evidence, submitted as part of the consultation, on the potential positive economic impacts of high speed rail in the North and the Midlands. We understand that much of this analysis has also been submitted as evidence to the Committee by relevant local and regional organisations.

The academically credible evidence does not show positive impacts on the North and Midlands. The KPMG work relied upon to show this is not robust, as shown by work by Laird and Mackie (actually for the Northern Way!) and confirmed by Tomaney. We have summarised the position in Appendix 2 of our observations on HS2 Ltd's further submission of 8 September.


Question 1.  To what extent would demand management on the conventional network delay the need for extra rail capacity?

DfT is cautious on the benefits of pricing changes. However, given crowding on the WCML long distance services is highly concentrated on a few services (ie after the 19:00 price cliff), this is unimpressive. It hardly reflects Philip Hammond's own statements on this issue post MuNulty.

Unlike commuter railways, long distance railways predominantly carry leisure travellers, who are more responsive to pricing. It seems likely that at least the 19:00 price cliff crowding could be eliminated. As this will represent the greater part of the standing on trains forecast for 2024 by Network Rail (assuming no interventions), the need for additional capacity based on crowding levels would be considerably reduced.

Question 4.  Is it appropriate to focus on the benefits of the Y network given that its case has been assessed in less detail?

DfT fail to mention the problems with 18 trains/hr, which arises with the Y network, and they are not discussed anywhere in the published HS2 Ltd and DfT documentation - despite the consensus of experts being that 18 trains/hr is not achievable.

On this matter a recent analysis has been published by Railway Technical Web Pages, again suggesting that HS2 Ltd's aspirations are unachievable. See http://www.railway-technical.com/Infopaper%203%20High%20Speed%20Line%20Capacity%20v3.pdf

Question 7.  To what extent do you consider that travel time should be considered productive? How realistic is the sensitivity test in Chapter 7 of the Economic Case?

DfT say:

"Furthermore, the evidence provided to the Transport Select Committee by the Guild of Travel Management Companies indicates that business travellers attach a high degree of value to the speed of travel (and that they tend to favour investment in high speed rail)."

The discussion of the importance of journey time savings in the Guild's submission is developed in the context of modal shift in the London/Scotland business travel market. No quantification is given. If this is the best evidence DfT can offer that business journey time savings are highly valuable they have no case!

The Eurobarometer survey of EU countries showed that UK long distance rail travellers are overwhelmingly satisfied (92%) with rail journey times, and reducing journey times in not a priority. A result confirmed by the Passenger Focus surveys. It is the factor with which UK rail passengers are most satisfied.

No one is saying that time savings are without value, simply that DfT's assessment is faulty:

—  reductions in time on board trains cannot be relied upon to improve business productivity; and

—  the relative earnings of rail travellers are overestimated by using very old and inappropriate data.

It is not clear whether DfT is saying that the sensitivity is realistic or not. A powerful reason for contending that it is not is that the realistic alternatives (Rail Package 2, the 51M proposal etc) have less crowding than HS2 - so there is no countervailing benefit from reducing crowding for HS2 to off set travel being useful. Indeed the alternatives are much better at reducing crowding than HS2 (being able to do so earlier and as required) and are less reliant on journey time savings for their economic case.

DfT assert:

"In the absence of better evidence on the very complex choices made by individual travellers, these adjustments indicate that the impact of incorporating a more detailed representation of passenger behaviour would not significantly affect the economic case for High Speed 2."

In the absence of proper studies and peer reviewed estimates, the DfT is not entitled to use discredited estimates of benefits when the business case is reliant upon them. 40% of the £44 billion of benefits (ie £18 billion) for the Y relate to on board journey time savings.

Question 8.  How confident are you in the estimated values of time?

DfT quote Lyons (et al 2011) as saying that reducing the value of time for rail travel in isolation would produce paradoxical results. We do not have a copy of this paper (only the presentation based on it), but on the basis of his previous articles we believe that this is cited misleadingly. Lyons is not suggesting that the reduction in the value is wrong - his analysis shows that it is right! - but he is emphasising it is also necessary to take account of the behavioural impact of reducing rail journey times on modal switching.

The key point that Lyons is making is that time has become more productive - and mobile technology has a clear role in this. HS2 will not be operational now, but at the earliest in 2026: DfT need a basis for being confident that time saving then (and over the 60 year appraisal period that follows) will have the value they are assuming. DfT have no basis to support assessing reductions in on-board journey time as having any productive value at all on this timescale.

DfT say:

"At present, we are not aware of any evidence that provides alternative values that are preferable to our central assumptions regarding the value of time. Only once a credible alternative approach to measuring the value of time-savings for business travellers across all modes has gained sufficient support would we be in a position to substitute the current values."

This is an absurd position to hold.

The numbers are discredited because time on trains has been becoming more productive and past and projected increase in business travellers implies less elite business travellers. Now that this point has been publicly exposed, DfT's position is generally recognised as a wholly unrealistic one to take.

You cannot justify a £30 billion investment on the basis of analysis that you know to be wrong on the basis that you don't know what the right answer is! The DfT is responsible for not having a proper basis for valuing time savings, and they simply have not made a reasonable case that the benefits would exist on anything like the scale they estimate.

The robust thing to do would be to assume that there are no productivity benefits. If the scheme still has a positive case, it doesn't matter that the necessary analysis on the value of time hasn't been done. If it does not then have a positive case, it is not rational to proceed with HS2.

We are not looking at a minor benefit here - it is 40% of all HS2's benefits (as the value of leisure time on trains is also affected).

Question 20.  Is it possible to suggest a likely order of magnitude for these omitted benefits?

It is worth examining the evidence on benefits that DfT suggest

—  The Core Cities Group believes that 400,000 new jobs in the Core Cities and a total 1 million in their wider urban areas will be underpinned by HS2.

The word underpinning gives an entirely false impression of attribution.

The original study is by Oxford Economics and it suggests that one million (in total) and 400,000 (Core City) new jobs might be created. These results are from their upside forecast for 2020. But this is before HS2 would run a single train (first run in 2026) or have any effect on towns and cities at all (save inhibiting some rail improvements whose benefits would be cut short by HS2). The idea that HS2 underpins these jobs is from the Volterra Arup Study (also done for the Core Cities) and is completely without justification.

—  Analysis by KPMG suggests that HSR could deliver 25,000-42,000 new jobs, contributing £17 billion-£24 billion per annum to the UK economy by 2040, generating £6 billion-£10 billion per annum in tax revenues, or £87 billion-£150 billion NPV to the Exchequer.

These are not estimated benefits from HS2 but from a much more extensive HSR network.

We believe this is the same KPMG study reported in the Greengauge 21 report "High Speed Rail in Britain: Consequences for Employment and Economic Growth" 31 January 2010. KPMG are unsurprisingly using the suspect "GVA" methodology.[240] The lower numbers relate to the full HSR network, the higher numbers relate to what might be obtained with unspecified improvements to local services. Dr Laird & Prof Mackie[241] and Prof Tomaney[242] have made it clear estimates using KPMG's GVA methodology are unsound.

—  Centro found that in the West Midlands HS2, together with local transport improvements, could increase GVA by £1.5 billion, provide 22,000 additional jobs and increase average wages by £300 per worker per annum.

This is again KPMG and is for the HS2 Y network together with unspecified improvements to local rail services.

—  Work undertaken by KPMG for Greengauge21 concluded that a national high speed rail network reaching to Scotland could boost the economy by between £17 billion and £29 billion by 2040, and increase national economic output by 2.1%.

This appears to be exactly the same work described in the second bullet point!

So DfT cite one completely spurious estimation of benefits, and two from pieces of KPMG work. In fairness to KPMG, KPMG do qualify their own results, but these qualifications are not reflected in DfT's use of them. However, DfT do admit that these estimates are not additional to the transport user and other benefits estimated by DfT's own methodology.

In effect DfT have no evidence of additional benefits.

Question 31.  Do the generally favourable ex post assessments of major rail projects (eg The Jubilee Line Extension) suggest that the bottom up BCRs are conservative estimates?

DfT do not mention HS1, and the serious effect that failure to anticipate completion had in overestimating demand. HS2 Ltd's analysis assumes no competition between HSR and the existing network.

DfT do also not mention Stratford International (at which no international trains stop) or Ebbsfleet, both expensive white elephants.

Question 32.  Are the bottom up estimates for the High Speed Rail programme consistent with the top down estimates from other high speed rail examples?

DfT are misleading in implying that the regeneration benefits are in addition to the monetised benefits: regeneration and employment growth are broadly how the productivity benefits are supposed to work their way through the economy - so it would be double counting to monetise both.

The fact that they are not additional is expressly recognised in the DfT discussion paper "Transport, Wider Economic Benefits, and Impacts on GDP", Discussion Paper, July 2005. That they are different views of the same thing is not controversial.


The note is prepared by HS2 Action Alliance.[243] It raises some points on the further submission to the Transport Select Committee (TSC) by HS2 Ltd. It is intended to assist the TSC in their questioning of HS2 Ltd, in providing context for some of the assertions and explanations that they give, particularly when they have offered in our view incomplete, misleading or erroneous responses.

This document has been prepared within four days to be available in advance of HS2 Ltd's Select Committee appearance.


Question 1.  Demand growth and capacity on WCML

HS2 Ltd observe that the 2021 forecast demand levels on WCML have already been reached, and so forecasts are likely to prove underestimates. However this ignores some important facts and is consequently misleading. Crucially the recent actual growth reflects the latest WCML service improvements, but the 2021 forecast excludes them. So it is a case of not comparing like with like:

—  The 2021 demand estimate takes no account of the service improvements from the "Very High Frequency" December 2008 timetable that completed the WCML Route Modernisation. All the "do something" forecasts - including those for HS2 and the rail alternatives - incorporate an uplift in demand for service improvements which include the December 2008 timetable improvements

—  Major service improvements (as with the December 2008 timetable change) normally result in rapid demand growth for a few years which then tail off

—  The three years of RPI +3% pricing will depress rail demand - the HS2 Ltd model predicts a significant reduction in demand for this period (see section 2.2.13 of "Model Development and Baseline Report" April 2011).

—  The period after 2021 has a high background annual growth rate (driven by the assumed economic growth) of about 2.6% per annum - so the model predicted background growth of only 18% to 2021 but 109% to 2043!

HS2 Ltd predict demand on the basis of an already out of date relationships between income and long distance rail demand, and they forecast increases over the excessively long period of 35 years. These are strong reasons for believing that their modelling will give over-estimates not under-estimates.

HS2 Ltd's observation on pricing measures to address crowding is ill considered. That there is crowding by 2043 for the "do minimum" case across the whole day is irrelevant. The "do minimum" case could and would not happen in practice, and, for any realistic alternative, measures to eliminate peak demand would be useful. This is especially true for the artificial demand peak just after 19:00 that results from the "price cliff" of saver tickets becoming valid.

Network Rail (NR) also forecasts demand without consideration of changing peak pricing or indeed any other measures to address demand, but even so they only forecast standing on about 5% of WCML long distance trains by 2024. The greater part of serious crowding must relate to the 19:00 price cliff.

NR say that by the mid 2020s WCML will be sufficiently full that ordinarily additional capacity would be provided. However, despite the suggestion to the contrary, additional measures would be available (including using the spare off peak train path per hour that NR themselves have identified, and is even mentioned in the DfT WCML franchise competition brief). The reality is not that the analysis in WCML RUS supports a new HSR, but that it assumes that a new HSR line is the best solution and will be adopted - so that the normal range of options to enhance services on WCML are simply not considered.

HS2 Ltd do not consider alternatives beyond a new conventional speed railway. NR give no detailed consideration to the opportunities for alternatives to a new HSR. DfT fail to develop the best alternative to a new railway.

Question 2.  Sensitivity tests

HS2 Ltd's explanation fails to recognise why the specified sensitivity tests on the distance term for the income elasticities are important:

—  This distance term was recognised to be problematic in the still in force (but seriously out of date) DfT webtag guidance[244] - so there was a specific requirement to use lower values.

—  If the results of the appraisal are not robust to these lower values there is a serious risk of the project under-performing.

—  This sensitivity is not symmetric with upper and lower alternatives to a central value, as the sensitivity analyses done by HS2 Ltd are - the whole point of this required sensitivity is that a more favourable result is not equally likely.

As it stands the current evidence is clear, ie that the distance term is incorrect (as reflected in PDFH5.0 and the latest Oxera Arup work). Compared to using the PDFH5.0 elasticities the demand predicted for 2043 by HS2 Ltd is 20% more than it should be. Were the growth also capped to 2033, as it was in the 2010 analysis, HS2 Ltd's figures would be 47% too high. Our earlier submission showed the arithmetic on this point.

Question 3.  Evergreen 3 and the "do minimum"

HS2 Ltd are misleading about the use of Evergreen 3 in the "do minimum" cases.

Evergreen 3 was included in the assessment of Strategic Rail Alternatives Packages 3-5 in the White Paper work (March 2010 analysis), and this had the effect of reducing the benefits attributed to these alternatives. But It was excluded from the demand analysis for HS2, the "do minimum", and the Rail Packages. The documentation states:

"Please note that although Evergreen 3 was not included in the Do-Minimum for demand modelling purposes to ensure consistency with HS2 Ltd's assumptions, it is expected to be delivered. The presence of Evergreen 3 has been taken account in the specification for Packages 3 to 5. No benefits for Evergreen 3 were included in the Economic Case."[245]

It is worth noting that Evergreen 3 gives about a 25 minute journey time saving on the Birmingham-London services. If this time saving were included in the benefits of Packages 3-5 it would greatly improve their economics. Deducting this obviously reduces the ascribed benefits.

We would agree that it would be incorrect to attribute improvements that properly belong to Evergreen 3 to the rail packages - but it is equally wrong not to take account of the reduction to HS2's benefits that Evergreen 3 produces.

Evergreen 3 was agreed and therefore fully committed in January 2010. The demand modelling was extensively re-worked for the 2011 consultation materials. All that work was done after Evergreen 3 was a committed project, in contrast procuring Intercity Express Programme trains has been included in all the analyses. The Evergreen 3 London to Birmingham improvements have already been implemented.

HS2 Ltd's observations about the demand model being unstable are illuminating. In essence the projected demand and the capacity developments included in the "do minimum" are incompatible. The "do minimum" simply doesn't have enough capacity for the forecast demand - which supports our contention that the "do minimum" is simply not a realistic comparator. Rather than develop realistic comparators, ad-hoc additions to ECML capacity are made.

We have heard the explanation as to why the reworking of the rail alternatives has a different "do minimum" case before. But it still does not really make sense to us. The problem is that if the "do minimum" that is used for HS2 is stable, how can the rail alternatives not be when they have additional capacity and a reasonably low load factor?

It is important to appreciate that because of how induced demand is treated in HS2 Ltd's assessment, comparing HS2 against the "do minimum" gives a better result for HS2 than using a proper alternative.

Question 4.  Impact of Evergreen 3 on the business case

HS2 Ltd do not correctly explain the position with Evergreen 3.

The key reason why HS2 Ltd think that it would make little difference to include Evergreen 3 is because their demand model ignores pricing differences between different ways of making the rail journey - so it ignores Chiltern Line being much cheaper (by about one third) than Virgin Trains (or HS2). Alison Munro stated this clearly in recent correspondence[246] to HS2AA as her defence of why Evergreen 3 would not affect their results. This ignores how passengers actually make their choices.

Modelled correctly, Chiltern might be forecast to win considerable numbers of passengers from both WCML and HS2. While Chiltern stations are generally less well connected than New Street for onward travel, they are better connected than the HS2 station. If premium prices applied to HS2, the loss of passengers from HS2 would be even greater.

HS2 Ltd's point about Chiltern only going to Birmingham is misleading. Passengers to Birmingham are a material part of those travelling on WCML and HS2. Three of the 10 trains per hour in the off peak specification for HS2 on day one run to Birmingham (four of the 11 trains in peak),[247] and they are the captive sets that run with 1,100 seats per train rather than the 550 seat sets to other destinations. Consequently Birmingham trains they represent nearly half of HS2's phase 1 capacity. They continue to represent half of the 1,100 seat services with the full Y (although they are a smaller proportion of total capacity).

Question 7.  Problems at Euston

HS2 Ltd's analysis of the number of people travelling through Euston on underground trains fails to focus on people trying to get on the trains at Euston, to which HS2 adds considerably with just phase 1 - 5,500 extra passengers in the am peak. The problem is that the trains will be full and so the additional passengers cannot board them

Phase 2 clearly makes the situation much worse. It is unsatisfactory to say that they haven't done the work for the Y, as the Y is the basis of their business case. It seems that this (like the 18 trains/hr) is an area that is likely to be resolvable only with considerable additional cost (Crossrail 2?).

HS2 Ltd's position on disruption is interesting. They are claiming that current services could be accommodated with four fewer platforms, despite DfT questioning whether 51M's proposal to increase peak services by just three trains per hour could be done without three additional platforms! See their supplementary submission. DfT and HS2 Ltd seem not to have consulted on this!

It is hard to square HS2 Ltd's supplementary submission with their earlier submission, and the warning given to bidders for the WCML franchise by DfT.

Question 8.  Capacity north of Litchfield

DfT have clarified[248] that the Manchester station and approach works that are part of Rail Package 2 (RP2) are not required to operate four Euston trains per hour, despite the wording of the documentation that says they are. These works are part of the cost of RP2 but not HS2, which also requires four trains per hour to Manchester in phase 1.

DfT now claim that the works are only required to save 1 minute on the time of Manchester services for the cost of £0.4 billion. This is surprising as the time saving is not mentioned as a reason for these works anywhere in the documentation, and it represents poor value for money as it only benefits the journeys for Manchester.

Question 9.  18 trains/hr

The assessment of line capacity in terms of trains per hour is at odds with all expert opinion. Furthermore, HS2 Ltd have not provided sufficient details of the line design to make it possible to properly check the analysis.

HS2 Ltd has previously declined to provide even this information, despite being repeatedly requested to do so. It is the subject of a formal FOI complaint. Indeed it may be that this material has been generated only after the consultation materials were published. Andrew McNaughton has said that he was doing the calculations in July 2011, and that they would be peer reviewed. We have been promised but not had sight of any product.

Due to the lack of detail, and the inconsistency between this material and the position HS2 Ltd held as reported in the Greengauge 21 workshops last summer ("High Speed Two Interfaces", Greengauge 21, July 2010), there must be serious doubts about whether the 18 trains/hr is really deliverable.

If HS2 Ltd really believe that they can demonstrate that the claimed 18 trains/hr is feasible then they should be prepared to make their analysis available for scrutiny. There persistent refusal implies only one conclusion.

The design of HS2 is unfavourable to high capacity, with a diverging junction for Heathrow, and six of the 18 southbound trains per hour originating from the classic network (often after travelling long distances).

HS2 Ltd say they assume grade separation and acceleration and deceleration tracks at junctions, which implies extensive four-tracking in connection to Heathrow, although there has previously been no mention of this.

While HS2 Ltd say that they take account of the variability of arrivals from the existing network, there is no explanation of why they reach such a different conclusion from Network Rail and the summer 2010 workshops.

The idea that 18 trains per hour represents a relaxation from 20-21 trains per hour and is robust seems difficult to square with expert evidence. Before applying other constraints, Systra assessed the maximum line capacity at 16.6 tph for a maximum speed of 350km/hr.[249]

Greengauge 21's report[250] of the workshops last summer makes it clear that the view of attendees was that 18 trains per hour was not possible. Participant included HS2 Ltd, DfT, Atkins, and Systra.

Alison Munro has offered as an explanation[251] of HS2 Ltd's reported position in the summer 2010 workshops: planning for a second north/south HSR was in the context of demand exceeding 18 trains/hr and that it was not questioning the deliverability of 18 trains per hour. However, this explanation is plainly inconsistent with the second extract shown in Appendix 1. Appendix 1 contains several extracts from the workshop report that show clearly that 15 trains/hr was accepted as a practical limit. Please note that this is after the publication of the 2010 White Paper materials in which HS2 Ltd expressed the view that 18 trains/hr would be achievable.

The suggestion that the inability to achieve 18 trains/hr could be addressed by joining and splitting at the Birmingham Interchange is impractical. Joining and splitting is currently practiced, but it is normally avoided because of the performance risk that it introduces. For HS2 this would be very grave as there would be no spare capacity for recovery.

The latest publication in this area can be found on Railway Technical Web Pages, that argues that in reality more modest capacity should be expected. See http://www.railway-technical.com/Infopaper%203%20High%20Speed%20Line%20Capacity%20v3.pdf

Question 10.  Four tracking HS2

The answers that HS2 Ltd give are at odds to other ones that they have given in the supplementary submission, and demonstrate how vulnerable the case for HS2 is to a more reasonable demand forecast.

The acceleration and deceleration lanes suggested in answer to Question 9 imply four tracking to some extent, and very extensively if speed and capacity is not to be materially affected.

The answer provided in explanation for not four tracking is illuminating, they quote their 2009 work that says that they should not rely on additional demand materialising after 2033:

"The level of demand we have assumed - with growth continuing to 2033 and then levelling off - does not appear to support such a substantial increase in capacity. That is not to say that the demand may not materialise at some point after the next 25-30 years. However it does suggest that HS2 would need initially to bear a substantial degree of additional cost and environmental impact on the basis that the demand may materialise. We do not believe this to be a credible position."

According to HS2 Ltd's own 2011 analysis, Phase 1 of HS2 is not viable unless growth continues past 2033. Consequently HS2 Ltd provide a clear, concise and reasonable justification of why HS2 should be abandoned. The subsequent shift in position appears to have no more basis than it is necessary in order to continue to claim there is an economic case for HS2.

Question 15.  Premium fares and fares competition

There are some simple and certain effects of premium pricing that HS2 Ltd have declined to mention. They are:

—  Fewer passengers will use HS2.

—  More passengers will use WCML and Chiltern Services.

—  Leisure passengers are more price sensitive, and will be more likely to choose the slower non-premium services.

—  The less affluent will be more price sensitive and hence not chose HS2, making HS2 even more a railway for the affluent.

—  There will be less scope for reducing WCML and Chiltern intercity services because there will be more passengers, so costs will be higher and freed-up capacity less.

—  Unless classic services have lower prices than currently modelled, forecast total passenger numbers will reduce.

—  If classic and HS2 services compete, surplus capacity may result in driving down prices and increasing the subsidy resulting from HS2.

There is already differential pricing, for example with Chiltern charging less than Virgin Trains. No one realistically expects that all the train services will charge the same. With time on board trains becoming more useful, and classic services having better connections to other train and public transport services, the economics of HS2 under price competition will be considerably worse.


10.  How was the level of the demand cap determined? What evidence is there to support it being set at the level selected?

HS2 Ltd suggest that there is no evidence of when demand for long distance rail may saturate and an increase of 61% per person (that gives the doubling when population growth is taken into account) is reasonable on the basis that those with higher incomes do more travelling.

This is entirely spurious:

—  Long distance domestic travel by all modes is also predominantly made by those in higher income households.

—  But people have not travelled more per person as incomes have risen. Since 1995 there has been no increase per person in long distance travel (for all modes), with an unchanging average of seven trips per annum over 100 miles and 20 for over 50 miles.

—  The cross-sectional relationship between income and the amount of travel undertaken is not mirrored by a correlation between income increasing through time and more travel - there is no such increase in travel.

Over this period rail has gained a larger share of long distance travel. HS2 Ltd's model (the rail part of which is generally used in the rail industry) assumes that long distance rail travel increases in a fixed relationship with changes in income. The question is how long can this be reasonably projected into the future. As the income elasticities are above one for long distance journeys, this cannot continue indefinitely as it leads to a higher and higher proportion of income being spent in such rail travel.

The reason for capping the period of increases is that fixed elasticity demand models assume that the key relationships remain permanently unchanged. But the further into the future the more unlikely it is that the assumed relationship would actually apply. This is because technical and social changes will alter the way demand develops.

There are good reasons to be cautious about relationships continuing in the longer term. In particular domestic travel (total and long distance) has shown signs of saturating in the UK, and the key relationship between economic growth and domestic travel has been breaking down. It is consequently unreasonable to assume that past relationships will remain unaltered until background growth doubles demand - however it takes.

The issue is how long one can have reasonable confidence that past relationships will hold:

—  Eddington (2006) suggested about 10 years was the limit of confidence.

—  DfT suggested 18 years (to 2026).

—  HS2 Ltd last year suggested 25 years - and offered reasoning why further increases should not be relied upon (see Question 10 above).

HS2 Ltd have shifted from using the rail model to forecast demand, to using the model to calculate how an independently specified level of increase would take to be reached. HS2 Ltd's explanation of why they project demand increases for 35 years to achieve the background doubling in demand previously forecast to 2033 is entirely inadequate. Not only is projecting background demand until it is has doubled arbitrary, as Oxera pointed out, but it is unreasonable. There is nothing inevitable about long distance rail travel doubling. Looking back, rail demand has only increased over the last 15 to 16 years. Looking 35 years into the past (to 1975) instead of the future, for the first 20 years there was no growth in rail passenger numbers at all - not even to reflect population growth.

It is worth noting that if there were more than a background doubling in demand (and a tripling in demand with HS2) HS2 would be incapable of accommodating the demand on HS2 trains running onto the existing infrastructure. If demand is less than a background doubling, the economics of HS2 are worse. So perhaps the doubling is not so much arbitrary as optimal for HS2's assessment.

Only projecting increases for 25 years would mean that the economic case for HS2 does not work. Rather than accepting that HS2 should be abandoned, HS2 Ltd and DfT cling on to background demand doubling without any reasonable basis.

11.  Have other scenarios of higher or lower fare increases been tested?

HS2 Ltd's approach is based on assuming that a doubling in background demand is in some way inevitable. This is inappropriate as discussed above (in Oxera question 10).

17.  How have the cost savings on the conventional network been estimated?

DfT state that they increase the cost saving from reducing classic services by a 41% optimism bias. As a result HS2 Ltd have substantially overestimated the savings (by £1.6 billion as shown below).

"Optimism Bias is the demonstrated systematic tendency for appraisers to be overly optimistic about key parameters."[252]

Applying optimism bias to increase an estimated saving acts in the opposite direction from the intended manner. Optimism bias is intended to prevent underestimates of costs - not inflate estimates of savings!

To illustrate how HS2 Ltd's approach must be wrong - if services were stopped entirely the cost saving would not be 100% of the cost but 141. So as opposed to the cost being eliminated, a surplus of 41% would be generated - which is clearly nonsense!

Netting off operating costs savings on the classic network from the operating cost estimates for HS2 - before applying optimism bias to the HS2 costs - is plainly wrong. If the classic network cost savings and the estimated operating costs of HS2 (before optimism bias) were identical, HS2 Ltd's approach would say there was no net cost. But the point is that the operating costs of the classic network are known and understood, while the operating costs for a new HSR are substantially unknown and therefore likely to be subject to optimism bias - ie be systematically estimated at too low a level. So the net increase in cost should be the optimism bias appropriate to the HS2 costs. Basically HS2 Ltd's approach treats all operating costs as equally subject to optimism bias - even those of operating existing systems that have been operating for years.

HS2 Ltd estimate the cost savings on the classic network with the full Y to be £5.4 billion. This suggests that the savings have been over-estimated by £1.6 billion - which represents almost a 10% increase in subsidy.

28.  Are there expected to be significant distributional effects between socio-economic groups as a result of the construction of the HS2 line?

HS2 Ltd's answer is less than complete.

The obvious impact is that a substantial subsidy is provided, for whom the principle beneficiaries with be HS2's passengers. With existing pricing arrangements, the majority of passengers (business and leisure) will be from considerably above average income households (47% of all long distance rail passengers are from the top income 20% of households).

Clearly, if there is premium pricing, passengers will be even more concentrated amongst the affluent.

30.  What is the relative size of the economic impacts on cities expected to be served by the high speed network? What proportion of these economic impacts is abstracted from other regions not served by the high speed network?

HS2 Ltd suggest that benefits accrue where trips originate and that the great majority originate outside London and the South East. This is incorrect on the facts.

70% of passengers will be travelling for leisure, and for them the economic benefits accrue at their destination (where they shop, visit the theatre and attractions, etc) - which is predominantly London.

For business travellers the situation is more complex, but reducing the barriers to competition through faster transport is likely to benefit London, which is economically more efficient than cities in the North and Midlands.

While cities with a HSR station may benefit, the evidence seems to be that this is at the expense of their hinterland and bypassed towns. The TSC has received extensive evidence on this topic. We summarise this at Appendix 2, which also explains why some of the evidence the Committee has received adopts a different position.

33.  How would substantial long-term oil price rises or falls have an impact on demand for rail? Would the impact be greater than those in the tested fuel duty scenarios?

HS2 Ltd's "sensitivity" is not a proper high oil price scenario. In a high oil price scenario high efficiency and electric cars would achieve market penetration much more quickly, counteracting the effect of higher prices.

As cars have an average age of seven years, technical improvement can achieve rapid market penetration.










Supporters of High Speed Rail (HSR) claim HS2 will create substantial numbers of new jobs in the Midlands and North. These claims are groundless and the facts quite clear. This note explains why.


HSR on routes to London will reduce regional employment and increase jobs in London:

—  DfT say more than seven out of 10 of the 30,000 jobs caused by HS2 around stations will be in London ie not the Midlands or the North, for example Old Oak Common, in London, will generate 20,000 of the jobs. But most of the jobs will not be genuinely new jobs but ones associated with shopping malls that have simply moved from other areas in the Midlands and North. So it's not a net increase in jobs. HS2 Ltd also concluded this, after taking expert advice.

—  DfT assume trips to London grow at twice the rate of those from London. Given most trips are for leisure, (70%), more people and more money will go to London and so will the jobs that support this. This outcome is what might be expected when a high speed link connects to a dominant city (and happened for example with Madrid).

—  HS2 impacts on the service sector, in which London is dominant. So work is more likely to move to London, not away from it, as faster journey times reduces the barrier to more efficient London businesses competing directly with less efficient regional ones.

So who says this? Just opponents to HS2? No. It is actually the leading academics who had been invited to give evidence to the Transport Select Committee (TSC) on the relationship between transport and the economy:

—  Prof Mackie (ITS, Leeds) says "For various reasons HS2 is rather unlikely to make much difference to the north south divide. A spatial analysis would probably show London to be the main benefiting region".

—  Prof. Tomaney (CURDS, Newcastle) who did a full literature review, says "The impacts of high speed rail investment on local and regional developments are ambiguous at best and negative at worst"…"In countries with dominant capital cities net benefits tend to accrue to these".

—  Prof Overman (LSE, London) said to the same Committee in October 2010 Claims about the "transformational nature of transport investments for particular areas should be generally discounted in assessing these benefits because they have no convincing evidence base to support them".

—  Even the work by HS2 Ltd itself, assisted by Prof Vickermann (Kent) recognised that some regions might lose out to London as a result of faster connections.

The mainstream view is clearly expressed by the Economist (3 September 2011):

"In most developed economies high-speed railways fail to bridge regional divides and sometimes exacerbate them. Better connections strengthen the advantages of a rich city at the network's hub: firms in wealthy regions can reach a bigger area, harming the prospects of poorer places."


The evidence from HS2 Ltd is that the wider economic benefits of HS2 would be small:

—  The productivity benefit from shorter journey times is the key benefit, but it's already in the business case (and is greatly overstated now that DfT admit time-on-board is not wasted).

—  The Wider Economic Impacts of better connectivity are relatively small, £4-£6 billion NPV (in 2009 prices), and are mainly driven by the use of freed-up capacity, but which will need a new further subsidy to realise.

—  HS2 Ltd asked Imperial College (Graham and Melo) if faster connectivity had any further direct benefits - they said "very little" (max £8-£10 million/a) - but their conclusion was not only left out of the White Paper last year, but not even referred to in the consultation materials.


There are reasons that explain why northern cities and businessmen think they will gain:

—  Cities served by an HS2 station think that their local economies will gain, but while the city may indeed benefit it is only at the expense of its hinterland. And only a few cities are proposed to have a station (London, Birmingham, Leeds and Manchester). The rest will not benefit.

—  Cities and businesses are not being asked to choose between having HS2 and having other transport investment. There are major benefits related to incremental non HS2 transport improvements which would be available far earlier than if the money is wasted on HS2.

—  Regional development and spatial economics are complex areas. It is not surprising people take at face value what Government tell them.


There are studies that attribute much larger numbers of jobs to HS2, than DfT say. We need to get the facts agreed before we get into debate about potential benefits.

The recent Volterra Arup study (for the Core Cities) claims that HSR would support the creation of one million extra jobs. This is simply untrue on their own evidence:

—  The million extra jobs are from a study by Oxford Economics in which the causes of the extra jobs are entirely unrelated to HSR. Clearly it is true that if there are one million extra jobs (in Local Enterprise Partnerships (LEPs), then extra journeys will need to be made to get them to work

—  This estimate of extra jobs is for 2020 - before HS2 has any effect at all - except to actually inhibit transport development because some improvements have to wait for HS2! So the 1 million extra jobs would have to happen despite HS2 - not because of it!

—  If HS2 is built it won't even help with journeys to work directly - it only helps through the creation of spare capacity on the existing railways. And HS2 only provides freed up capacity on the routes it serves - for 2026 this is only London to Birmingham! It does nothing for any other Core City or LEPs.

Extra capacity where it is needed on the existing network could be created well before HS2 and actually support the extra 1 million jobs that will be created (but not by HS2) in LEPs.

There are also earlier studies done for Greengauge 21, Centro and Northern Way studies (by KPMG) that attribute big employment increases to HSR. But these studies:

—  Are not on a reputable basis for forecasting extra jobs - rather than using the DfT approach they are based on an unsound "GVA" methodology.

—  Work commissioned by the Northern Way itself shows the problems with the GVA approach and how it overestimates the effect on jobs.


So far from supporting the government's claims that HS2 will redress the North/South divide, the evidence (affirmed also by Oxera the independent analysts appointed by the Transport select Committee) actually suggests that HS2 would re-enforce London's dominance.

To genuinely benefit the North/Midlands what is needed are transport investments that improve the efficiency of their labour markets - not ones that expose them to greater competition from London, as HS2 does.

8 September 2011

237   This paper was prepared by Bruce Weston, on behalf of HS2 Action Alliance, which is a not for profit organisation that is campaigning on an evidence based approach against HS2, and has 73 affiliated groups (see www.hs2actionalliance.org). Back

238   "ERTMS National Implementation Plan" DfT, September 2007 Back

239   See Andrew Giligan's articles in the Telgraph on unemployment and Ashford house prices in connection with HS1 Back

240   See Appendix 2 of our observations on HS2 Ltd's further submission to this Inquiry. Back

241   "Review of methodologies to assess transport's impacts on the size of the economy", James Laird and Peter Mackie (ITS), September 2010 Back

242   "The Local and Regional Impacts of High Speed Rail in the UK: A Review of the Evidence", John Tomaney, Pedro Marques and Penny Marshall, April 2011 Back

243   This paper was prepared by Bruce Weston, on behalf of HS2 Action Alliance, which is a not for profit organisation that is campaigning on an evidence based approach against HS2, and has 73 affiliated groups (see www.hs2actionalliance.org). Back

244   Tag Unit 3.15.4, section 6.1.1 Back

245   "High Speed 2 Strategic Alternatives Study - Strategic Outline Case", March 2010, page 10 Back

246   Letter to Bruce Weston, 29 June 2011 Back

247   'Economic Case for HS2' Feb 2011, figure A1 page 59 Back

248   Letter from Alison Munro to Bruce Weston, 28 July 2011 Back

249   "Fast forward a high speed rail strategy for the Britain" Final Report Appendix B, October 2009, Greengauge 21, Figure 2 Back

250   "High Speed Two Interfaces", Greengauge 21, July 2010 Back

251   Letter to Bruce Weston, 28 July 2011 Back

252   Webtag Unit 3.5.9 section 3.5.1 Back

previous page contents next page

© Parliamentary copyright 2011
Prepared 8 November 2011