High Speed Rail - Transport Committee Contents

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



I am writing in connection with the Independent Review that you conducted for the Transport Select Committee. We found it most helpful and were pleased that it covered several of the points that we ourselves have raised concerns about. I thought it might also be useful to identify some other points that were not covered in your review but we feel have a strong bearing on the economic case for HS2. Our purpose is simply to expose the key issues.


As you observe, HS2 is not compared against the strategic alternatives but a "do minimum" case. However, you do not mention that the strategic alternatives developed for DfT are sub-optimal, and other more cost effective solutions exist (some were even mentioned by Atkins but not pursued (for example in Rail Package 1 that concerned longer trains)).

An indication of the unsuitability of the options considered by DfT is that of the load factor. If we use the WCML 2007-8 base and HS2 as a point of reference, these have load factors of 57% and 58%. In contrast, WCML, MML and ECML have load factors in Scenario B of 51%, 28% and 43% respectively. Given the strategic alternatives are composed of a number of separately implementable elements, the strategic alternatives create large quantities of spare capacity. Scenario B also contains some counter-intuitive elements, for example the electrification of MML but also electrifying and upgrading the track between Retford and Sheffield (not currently an ECML service at all).

There are also some important differences in approach in DfT's assessment of HS2 and the strategic alternative that favour HS2:

—  The strategic alternatives are assumed to be implemented in their entirety for a single date, not implemented incrementally as demand requires. Had the strategic alternatives been assessed implemented as required this would greatly improve their economics—in both preventing the build up of crowding and the creation of capacity and cost before it is required. This is not even consistent with how rolling stock for HS2 is assumed to be acquired to match the build up of demand.[221]

—  The key current crowding issue on WCML is on the Milton Keynes and Northampton commuter trains. This could be addressed by the Ledburn Junction work and new fast commuter trains in five to six years, (an element of Rail Package 2 and Scenario B) yet DfT assume that these works would not be done to 2026.

—  Rail Package 2 contains a number of costs for infrastructure works equally required for HS2 but excluded from the HS2 costing, for example the platform and approach works at Manchester (needed for four trains/hour running for RP2 but assumed to occur as part of the Northern Hub works for HS2 that also runs four trains per hour).

—  It might also be noted that the rolling stock and running costs for the strategic alternatives are treated as subject to the same optimism bias as HS2—despite the costs of operating the existing lines and being precisely known and additional Pendolino stock still being in the process of delivery. In contrast HS2 Ltd state that new technology on train control and braking is needed, and the exact scope of works is understood in far less detail.

—  ·  The April 2011 reworking of the economics of Rail Package 2 was even against a different base case from HS2 that caused benefits to be underestimated—despite producing this reworking on the pretext of bringing it onto the same basis as HS2!

The 51 M submission to the Select Committee details a considerably superior approach to alternatives. For example for WCML, the approach is to do the lowest cost alternatives first—ie changing the balance in provision of first and standard class accommodation (practicable without revenue loss due to the low occupancy of first class), lengthening the fleet in stages to 12-car (except for services to Liverpool). 12-car operation was explored and found practicable in the work done by Atkins for the 2010 White Paper, but not used in Rail Package 2 or Scenario B (despite the explicit recognition that it could have been). Because these elements could be implemented in phase with demand growth, they should be commercially viable and have no subsidy requirement. The commuter issue would also be addressed early.

Failing to develop the best alternatives and to compare HS2 against them has resulted in a grossly flattering economic assessment of HS2. In particular, HS2 would not have any crowding benefit (as it would not reduce crowding), so their suggestion that the loss in benefits from time on board trains being productive would not be recouped through reduced crowding.


The demand forecasts for HS2 are much less moderate than DfT present them as being. As we discovered from an FOI, the 1.4% increase per annum is not for long distance (over 50 mile) journeys, but "strategic trips" which includes shorter journeys. The doubling in background demand for WCML to 2043, with an average growth rate of 2.0% per annum is also a misrepresentation, as the 2043 figure is reduced by transferring some of the journeys attributed to Virgin Trains in the 2007-08 base to London Midland. A more representative figure is given by looking at the traffic north of Milton Keynes that gives an increase of 127% and a growth rate of 2.4%. This is the level of long distance rail demand forecast by DfT in 2007—without the recession or the period of RPI +3% fares rises.

The "background" growth does not include the effects of improvements to services—including the improvement from the WCML December 2009 timetable—which are on top of the 2.4% growth per annum. The "do minimum" growth is for exogenous factors and fares only. The demand forecasts for HS2 and RP2 (etc) include uplifts for service improvements over the 2007-08 service specification.

You note that DfT still employ crucial PDFH4.1 values in their guidance and HS2 assessment—and not version 5.0. Readers might not be aware that DfT continue to use income elasticities on demand of 4.1 that have a distance term—unlike with version 5. We understand that the work that Oxera yourselves have recently done confirms that there should not be a distance term. The removal of this distance term has a major impact on the demand forecast.

You note that the demand forecasts are the area most thoroughly subjected to sensitivity tests. I am not sure that you are aware that the sensitivity test required by DfT's own guidance on income elasticities was not performed. Again we had this confirmed in an FOI from HS2 Ltd and are still waiting for a reply from Phil Graham as to whether permission was given not to follow the guidance. Had it been it would have showed that the economic case is not robust.

We agree that the doubling in demand does not seem to have a proper basis. However, we feel that using PDFH to make forecasts over a 35 year period with income elasticities as high as 2.8 should attract specific questioning.

An important but unobvious omission in DfT's case is that of the Evergreen 3 improvements to the Chiltern Line. These reduce London—Birmingham journey times to near Pendolino levels this autumn. This will be bound to win traffic from WCML, thus reducing demand and capacity requirements on WCML, and increasing competition for HS2. It is omitted from the "do minimum" case, and the assessments of HS2 and RP2.


In your discussion of premium fares you did not note that HS2 Ltd's view[222] is that:

—  demand for HS2 would be price elastic for the induced demand (some 35% of passengers) so that increases in ticket price would reduce overall revenues;

—  there would be limited scope for higher prices for the leisure market (70% of passengers); and

—  The availability of non-premium competition affects the ability to generate additional revenue through premium fares.

Interestingly, HS2 Ltd are more positive about using premium fares to price off excess demand, which they recognise here cannot be accommodated on some of the services running onto the classic network but fail to in their economic assessment.

It would seem that competition would increase costs (as passengers will chose cheaper classic services, so that more such services will be run) and may well reduce revenues as over-supply reduces fares, worsening the overall economic performance.


You note that the impact on reliability of the strategic alternatives is unclear. While we agree, the reliability of the strategic alternatives to the London—West Midlands phase was discussed in the 2010 documentation, that concluded:

"Even with higher levels of train frequency, the packages may enhance train performance at a network level.....These locations [where grade separation of extra track is built] may more than compensate for other area where there will be an enhanced train frequency but no infrastructure enhancement" [223]

There are grounds for suggesting that the claimed reliability benefits of HS2 are questionable.

DfT claim HS2 will deliver £5.7 billion of reliability benefit for the full "Y" network. However such reliability benefits are predicated on HS2 being largely isolated from the classic network to avoid importing delays. Such isolation will not apply either to phase 1 or phase 2 of HS2. Given the very intensive time tabling planned for HS2 (18 trains/hour), this suggests the claimed benefits may be doubtful.


A major uncertainty that DfT have not covered is the deliverability of the "Y" network service pattern. This requires 18 trains/hour in the peak—without allowance for trains to Heathrow and directly onto HSI. This is recognised by HS2 Ltd not to be deliverable with current technology, and HS2 Ltd have given no evidence to support it becoming deliverable in the future—despite an FOI request for the basis of their expectation.

This is clearly—if not a show stopper—at minimum a major uncertainty about the deliverability of benefits (reducing the services to 14/15 trains per hour would do serious damage to the economics)—or on the costs as building an additional line or four-tracking the stem of the "Y" would have major cost implications.

DfT have not even considered issues surrounding their proposals being capable of supporting 18 train per hour in the sensitivity analysis.

It may be that HS2 must be regarded as incapable of supporting services to Leeds, and that it needs to be appraised on the case for London West Midlands and Manchester by itself.

If you would like me to clarify any of these points, or others made in our submission, I would be pleased to in correspondence or in person.

June 2011

221   "A Summary of Changes to the HS2 Economic Case" April 2011, Section 2.1.1, page 6 Back

222   "HS2 Demand Model Analysis", February 2010, Section A2.4.5, page 155 Back

223   "High Speed 2 Stategic Alternatives Study: Rail Interventions Report", March 2010, Appendix B, Section 1.1.1, page 16 Back

previous page contents next page

© Parliamentary copyright 2011
Prepared 8 November 2011