Transport and the economy

Memorandum from Wharf Weston (TE 30)

1. Summary

1.1. This submission addresses one of the questions that the Select Committee for Transport have asked:

‘Are the current methods for assessing proposed transport schemes satisfactory?’ ’

1.2. The submission argues that the DfT’s approach has serious defects, resulting in recommendations that are wasteful of public funds. The key concerns are:

§ Inappropriate forecasting methodology for long term projections of demand where demand saturation is an increasing risk

§ Inadequate account of uncertainty is taken which requires flexibility for long term projects

§ Lack of a visible framework for generating and assessing project ‘alternatives’ to ensure objectivity and value for money drives the choice of publicly funded projects

§ The unit values attributed to journey time savings by DfT ignore key developments

§ A material monetary cost (property blight) is excluded from DfT’s assessment approach

§ The impact of competition is persistently ignored.

1.3. Some of these criticisms are not new, but the DfT has failed to learn lessons of previous projects, despite studies investigating past failures. The assessment prepared by High Speed 2 Ltd (HS2 Ltd) and DfT for HS2, is a revealing current case study and is used to illustrate the concerns.

2. Introduction

2.1. Appraisal should not be a process justifying investments that are politically driven, ignoring or discarding facts and options that are impediments to reaching the ‘right’ answer. The misallocation of public funds to unneeded and over-expensive projects must be avoided. The arrangements must ensure that only schemes that represent good value for money and are superior to alternatives are selected.

2.2. The DfTs approach is in principle reasonable:

§ It combines limited cost benefit analysis (for impacts apt for expression in terms of money) with other criteria that, while capable of objective assessment, are less readily monetarised eg environmental impacts. The decision maker is left to consider the particular combinations of value for money and other factors in making a choice.

§ The methodology and many of the assumptions or parameters to be used for assessing transportation schemes are published by DfT in webtag (guidance on the DfT web site). This promotes consistency and the use of validated assumptions.

2.3. However there are weaknesses in DfT’s approach which may lead to poor decision making.

2.4. DfT continue to apply a ‘predict and provide’ approach to rail demand, despite its abandonment for roads, and recent effective abandonment for air (with a policy for no new runways for London airports). While not developed in this submission, with increasing environmental concerns and the creation of a portfolio responsibility for ‘non travel’, a reconsideration of this policy for rail seems overdue.

3. Uncertainty and demand projections

3.1. DfT have been criticised for over-estimating demand for major rail projects previously. The Select Committee on Public Accounts criticised persistent over-estimation for CTRL. [1]

3.2. HS2 Ltd’s demand estimates are high compared to other reputable forecasts.

Table 2: Long distance rail growth forecasts

Source

Date

Period

Increase

Annual rate

(DfT [1]

DfT [2]

2007 (July)

2007 (July)

2006-2027

2006-2030

65%

73%

2.4% (1.8% from 2017))

2.3%

Network Rail [1]

2010 (August)

2008-2034

70%

2.1%

Prof J Dargay [1] for ITC

2010 (January)

2005-2030

35%

1.2%

HS2 Ltd (Atkins)

2010 (February)

2008-2033

133% [1]

3.4%

Average (based on rates)

2008-2033

75%

2.3%

ITC is the Independent Transport Commission

3.3. There are strong reasons to believe that all these forecasts will prove overestimates, as they are founded on elasticities with GDP (so that forecast demand increases more quickly than GDP), but this relationship has been breaking down for domestic travel. Clear evidence of saturation limiting recent growth to that of the population has been emerging (see National Travel Survey [2] , Dr D Metz [3] , Bluespace Thinking [4] . The graph shows that since 1995 increases in wealth (real Gross Value Added per capita) have not been matched by any increases in domestic travel.

3.4. DfT correctly observed in their 2007 analysis [5] that the relationship between transport demand and GDP has been changing. But this is not generally recognised in demand forecasts.

Figure 1: Travelling time, journey numbers and distances per person (cf real GVA/capita) [6]

3.5. Just projecting forward recent trends in rail demand ignores the broader picture. There is no long term relationship between GDP and long distance rail travel. From the 1950s to the early 1990s, despite massive increases in GDP, rail usage enjoyed no increase. It is likely that the recent increases are a response to rail improvements, and modal switching in the context of saturating overall demand, not a consequence of economic growth.

3.6. Inadequate account is taken of market maturity and saturation. This, if uncorrected, will lead to a major misallocation of funds when applied to assessing long term projects, as expenditures will be made to accommodate demands that will not materialise.

3.7. At minimum there is risk that the high levels of growth (267%) predicted by HS2 Ltd will not occur. This has important consequences for the choice of options. HS2 is inflexible – a new railway is built to Birmingham or it is not: it is all or nothing. In contrast the rail and road alternatives for improvement existing infrastructure are incremental, have short lead times and can be implemented against the emerging trend in demand. The alternatives are robust to shortfalls in demand, HS2 is not. Any consideration of risk (as commended by the Treasury Green Book and DfT’s 2007 White Paper) favours adoption of the alternatives. A methodology is needed to make this happen.

3.8. To illustrate, if we consider each of the estimates in Table 2 equally likely, the expected NBR of HS2 becomes 1.35 [7] . In contrast the NBR of the alternative schemes might increase, as only the easier and cheaper options would need to be implemented.

3.9. The 2007 White Paper recognised the problems of long term forecasts and the need for flexible solutions:

Forecasts have been wrong before, and any strategy that tried to build a rigid investment programme based on fixed long-term forecasts would inevitably be wrong again.

To overcome this challenge, the guiding principles in this strategy are:

To invest where there are challenges now, in ways that offer the flexibility to cope with an uncertain future; and

To put in hand the right preparatory work so that, as the future becomes clearer, the necessary investments can be made at the right time.’ [8]

3.10. Unfortunately DfT’s assessment methodology does not reflect this advice.

4. Alternatives and their assessment

4.1. DfT’s processes neither generate an appropriate range of alternatives, nor are they properly employed in option selection. The effect of this deficiency is that DfT recommend projects that are inferior to entirely feasible alternatives. This is a persistent problem. The Foster Report [9] , observes inadequate assessment of alternatives to the Intercity Express Programme

‘I am not convinced that all the credible alternatives to Intercity Express Programme (IEP) have been identified, worked up and assessed on an equal footing with it……..

The team’s preliminary analysis suggests that these alternatives could achieve better value for money than IEP, realising a greater proportion of the currently desired IEP benefits with reduced expenditure over the coming 15 to 20 years, and especially during the next decade.’

4.2. Similarly for HS2 there are weaknesses in the option development and the selection process:

§ The ‘do minimum’ case is inadequate– it cannot meet the projected demand and so when compared with HS2 causes key benefits (on crowding) to be overestimated

§ The HS2 Ltd alternative is only for a new railway (not an improved existing one) and the DfT developed alternatives are not required to be used in the selection process

§ DfT alternatives do not include a least cost option that meets the projected demand

§ The assessment rejects alternatives that have better Net Benefit Ratios than HS2 (on the basis they do not provide surplus capacity – yet they do meet the demand required)

§ There is no assessment involving incremental benefits over incremental costs of HS2 compared to the best alternative

The HS2 Ltd ‘do minimum’ reference case

4.3. For HS2, the assessment is done against a ‘do minimum’ reference case for rail and road network development, which essentially only incorporates committed future developments or those very likely to happen within the next 10 years. This means it will fail to meet the projected demand growth.

4.4. Appraising HS2 against this case will overestimate key benefits, eg crowding. This is because, with predicted passenger trips into London greatly outnumbering those originated from London [10] , an all day load factor of 81% [11] is unachievable. Like commuter trains, those running against the flow will be sparsely occupied – but unlike commuter trains, passenger numbers cannot be expected to continue to grow if this necessitates increasing levels of standing.

DfT developed alternatives

4.5. DfT had some credible alternatives developed by Atkins, but they were not seriously considered as alternatives to HS2, or used as a basis against which to assess HS2.

4.6. Importantly, no attempt was made to develop the least cost means of meeting this demand. Generally this is achieved by running longer trains. This option was developed as ‘Rail Package 1’, but was dismissed as it could not accommodate all the projected demand without infrastructure amendments that were claimed to be impractical. But it was also deliberately not incorporated in other packages of changes [12] , although it would have made them both less expensive and less disruptive means of meeting demand. Indeed it was not even costed.

4.7. To omit the least cost approach only makes sense if there is no interest in developing the best value for money option. A 65% increase in demand could be accommodated by longer trains [13] with no further infrastructure works beyond that scheduled for 2012 and part of the ’do minimum’ reference case (except perhaps some limited work at Euston). Currently WCML Pendolinos have four of their nine cars as first class. With the decline in first class travel, it may be possible to reduce this to increase overall capacity without loss in revenue. Consequently, over half the projected demand (a 133% increase without HS2) could be satisfied with no disruption to services at all. Typically this means of satisfying additional demand is commercially justified, and would not require DfT subsidy.

4.8. Rail Package 2 (RP2) [14] allows projected demand to be met, and while HS2 Ltd predict that it would generate less benefit than HS2, it has much lower costs, and so has a better Net Benefit Ratio (NBR). But DfT do not use RP2 as the comparison basis for HS2, to judge whether the additional benefits of HS2 justify the additional costs of HS2. This is not because it had been rejected on the basis of other criteria in the assessment framework, as it has better sustainability.

4.9. RP2 is rejected primarily because it does not provide as much additional capacity as HS2 [15] – despite the fact that on HS2 Ltd’s demand estimates there is not actually any need for this additional capacity. The creation of un-necessary capacity hints at the misallocation of resources rather than compelling justification. Other ways in which HS2 is held superior (journey time, reliability) are already incorporated in the assessment of NBR for which RP2 is better than HS2.

4.10. In a climate of public spending cuts, to reject a more cost effective solution on the basis it does not provide as much excess capacity is profligate. As RP2 satisfied the forecast demand, it seems common sense to use this rather than an unrealistic ‘do minimum’ case.

4.11. Had RP2 been used as the base against which to assess HS2, this would have shown that HS2 could not be justified in terms of its NBR. The table below summarises the costs and benefits for HS2 and the best value for money alternatives. If we assume that all the benefits of RP2 and half those of Road Package 2 are benefits that HS2 would otherwise deliver, we can see that HS2 has a net benefit ratio of 1.88 (instead of 2.7).

Table 1: HS2 and alternatives: net benefit ratios

Against ‘do minimum’ case (source: HS2 Ltd)

Incremental case

HS2

Rail Package 2

Road Package 2

HS2

Present value of benefits

32.3

7.35

5.14

22.38

Present value of net cost

11.9

2.03

1.40

11.9

Net Benefit Ratio

2.7

3.63

3.66

1.88

4.12. In reality some combination of parts of Rail Packages 1 and 2 would be necessary to accommodate the demand growth projected by HS2 Ltd before HS2 is completed.

4.13. DfT’s process also fails to do justice to alternatives, assuming a common start date for HS2 and the alternatives of when HS2 could be complete. This fails to credit the alternatives with a benefit stream that could predate that of HS2. This means that the NBRs for the alternatives will be significantly understated compared to an assessment that allows phased early implementation of alternatives against emerging demand.

4.14. Under DfT’s methodology there is no requirement to adopt the best value for money solution, nor, when selecting a different option, is there a requirement to justify this in terms of the other criteria in the assessment framework (but not incorporated in the cost benefit analysis).

5. Value of time savings

5.1. Time savings are the principle benefit used to justify government subsidy in transport investments. However, the webtag values fail to take account of some travelling time being useful and are based on out-of-date data.

Usefulness of some travelling time

5.2. Webtag values for time savings derive from a resource cost for business travel (the cost to the employer of the time) and a willingness to pay basis for commuting and leisure travel.

5.3. Time spent travelling is not split between useful and non-useful time. For both business and other travel, any saving of time spent on trains is credited at the full value of the entire time saving. This is inappropriate for long distance rail journeys. It ignores that some of the time within rail journeys is useful and productive. Mobile phones, laptops and mobile broadband can make the train as productive as the office, and similarly useful for commuters and leisure travellers.

5.4. However, DfT’s webtag guidance disregards the usefulness of this time – despite it being generally recognised [16] that shortening long distance journeys may have little or no value for saving productive time, including in work sponsored by DfT:

‘Rail Business travellers in the UK are now using travel time highly efficiently. Marginal reductions in travel time (10, 15, 20 minutes) are not guaranteed to lead to much extra productive time at work, whether in the 'usual workplace' or elsewhere.’ [17]

5.5. For business travel, if saving travelling time does not reduce unproductive time, the time saving has no value to business. For a project starting in 15 years time, it is not credible that technology will not permit business travellers to be fully productive.

5.6. For leisure travellers and commuters, willingness to pay is the basis for valuing time savings, which relates to the degree of disutility of travelling time. Different values are identified for different travel activities – ie waiting for transport is costed at 2.5 times actually travelling, walking to access public transport is valued at twice the value of travelling [18] , but there is no value specific to useful time on trains.

5.7. While leisure and commuting time on trains cannot be used as flexibly as time at home or some other free time, it may be used in an increasingly broad range of activities that must be substantially eroding previous disutility [19] .

5.8. HS2 Ltd estimate that time savings are worth more than £13bn [20] (£23bn [21] including reliability) and this is mainly the value to business. It seems unlikely that the savings are worth more than a small fraction of this (perhaps a quarter). A £10bn reduction alone would reduce the NBR from 2.7 to 1.87. The road and rail alternatives would not be similarly affected: time in cars is likely to remain much less useful, and the rail package benefit is mainly crowding reduction (worth £5bn [22] for HS2, and trains are less crowded under RP2 than HS2) and from increasing service frequency (reducing waiting time that is not useful time).

5.9. The cost of business time will increase with employee costs, which are expected to rise due to continual improvements in productivity. However, if the relevant travel time is productive, increases in the unit value have no effect. For commuting and leisure time, the money value of time can be expected to increase, because the utility of money falls with increasing income. But the disutility of time during long rail journeys is or will become low, so that while the value of time savings generally can be expected to increase in money value, these savings will be of modest value.

5.10. Serious crowding may prevent travelling time from being useful. This is a further reason for increasing capacity incrementally in line with emerging demand and in advance of HS2’s start data.

Out of date data

5.11. The value of business travel time on trains has not been updated to reflect recent trends. Business travel by train is increasing (despite a fall in total business travel), so that while rail business travellers’ average salary was previously much higher than the all transport average, the additional business travellers will have reduced the rail average.

5.12. By using 2002 information [23] , DfT are substantially overestimating business rail travellers’ salary levels. If the numbers of business travellers increase by a further 360% [24] to 2033 (as HS2 Ltd forecast), while population only grows by 16.6% [25] , using 2002 data will give a substantial overestimate of value.

5.13. This illustrates how important it is that webtag parameters are evidence based and properly reflect known trends.

6. Ignores a material monetary cost

6.1. DfT assessments deliberately leave out a social cost that can be monetarised – the reduction in value of properties near line of route [26] . This cost is not included anywhere in DfT’s assessment framework. The costs of compensation are included, but these are relatively small. Blight is not due to a redistributive effect (with losses offset by gains elsewhere), but a result of degrading the local environment.

6.2. Property blight data is not readily available because currently the individual and not public purse meets the costs of the blight that transport infrastructure projects generate. It is, however, not difficult in principle to estimate the costs of property blight, as relatively simple techniques can establish divergences in property value trends for a locality. An unsuccessful attempt was made in the 1990s, as reported by the Interdepartmental Work Group on Blight [27] However, modern positional software has been extensively used to map dwellings and other buildings, making the identification of the large numbers of properties needed to provide reliable analysis easier. A study to quantify the effects of blight is being planned for HS2 [28] .

6.3. Just as social benefits (eg time savings) are included in the cost benefit assessment, the social cost of blight should also be included – irrespective of who pays it.

7. Competition

7.1. DfT assessments are persistently marred by a failure to adequately consider competition. The Channel Tunnel failed to take account the response of ferry companies. The CTRL failed to addressed the impact of low-price air travel. The high speed Kent commuter services have failed to recognise that passengers would prefer the residual ‘classic’ services with lower fares and overall shorter journey times to commuters’ places of work.

7.2. HS2 Ltd assume-away the competition between HS2 and conventional services:

‘ HS2’s approach has effectively assumed a regulatory framework that allows the joint (social) optimisation of both high speed and classic rail services.’ [29]

7.3. Unless competition is suppressed, conventional long distance services will compete with HS2. Competition will reduce both the number of passenger on HS2 and overall revenues, which may force HS2 to run with an operating subsidy. The natural use for surplus long distance rail capacity is not to provide more local services (that are already only provided at their current level due to subsidy) but to undercut the high speed services on price (as Chiltern Railways does to the WCML services for London-Birmingham).

7.4. If HS2 fails to deliver additional fares income because of competition, this alone reduces the NBR to 1.20. The DfT assessment process should involve proper consideration of competition.


[1] Select Committee on Public Accounts, Thirty eighth Report, 27 March 2006

[1] ‘Delivering a Sustainable Railway, Summary of Key Research and Analysis, July 2007 ’ , DfT . S lide TPF9a, page 27

[2] ‘Delivering a Sustainable Railway’ , Cm 7176, Dft, July 2007, paragraph 6.6, page 60

[1] ‘Planning Ahead 2010’, page 6, section 2.10, Network Rail, ATOC and Rail Freight Operators Association, August 2010

[1] ‘The prospects for longer distance coach, rail, air and car travel in Britain ’ J M Dargay, January 2010

[1] ‘Command Paper 7827’, March 2010, section 5.38 page 91 , growth without HS2 uplift

[2] National Travel Survey, 2009, Tables NTS0403 and NTS0404

[3] ‘Saturation of demand for daily travel’, David Metz, May 2010 . published in’ Transport Reviews’

[4] ‘A review of high speed rail – HS2 proposals’, Bluespace Thinking Ltd, April 2010, section 4

[5] DfT op cit, slide EST1, page 3

[6] After Dr Metz based on NTS 2008 Table 2.1 with GDP trend added

[7] Reducing all benefits in the proportion to the reduction in forecast passengers ( assuming HS2 uplift equals background growth in demand, as in HS2 Ltd analysis)

[8] ‘Delivering a Sustainable Railway’, Cm 7176, Dft, July 2007, page 9

[9] ‘A Review of the Intercity Express Programme’, Sir Andrew Foster, June 2010 , page 22

[10] Webtag elasticities for rail travel on GVA for journeys to London are 2 to 3 times those of journeys from London (Table 11.1, Tag 3.15.4), meaning that growth will be predominantly to London

[11] High Speed 2 Strategic Alternatives Study Strategic Outline Business Case, Table 3.7, page 38

[12] ‘High Speed 2 Strategic Alternatives Study: Rail Interventions Report’., page 34

[13] Lengthening 31 sets to 11-car from 9-car and having 4 new 11-car sets are planned for operation in 2012. This increases capacity by 32% (source Virgin Rail 31 July 2008). Lengthening the rest of the fleet to 11-car creates a further increase in capacity of 14%, and extending all to 12-car a further 19%, totalling 65%.

[14] described in High Speed 2 “Strategic Alternatives” Study: Rail Interventions Report, and the economic assessment is in ‘ High Speed 2 Strateg ic Alternatives Study: - Strategic outline business case’

[15] Command Paper 7827, Section 2.20 to 2.22

[16] See for example ‘Travel Time Use in the Information Age: Report’, Centre for Transport & Society, UWE, Bristol , and Centre for Mobilities Research, Lancaster University , October 2007 , or ‘The use of travel time by rail passengers in Great Britain ’, Glenn Lyons, Juliet Jain and David Holley, January 2007

[17] ‘The Productive Use of Rail Travel Time and Value of Travel Time Saving for Travellers in the course of Work’ The Mott MacDonald IWT Consortium, 2008

[18] Webtag 3.5.6D, paragraphs 1.2.19 and 1.2.20

[19] See ‘Values of Travel Time Savings in the UK ’, P J Mackie, M Wardman et al , January 2002 , page 50

[20] ‘High Speed 2 Demand Model Analysis’ section 10.4.3

[21] From HS2 Ltd spreadsheet HS2_Day1cWiderImpacts_MidRange.xls, sheet SummarybyArea , obtained under FOI10-039 by Dr J Savin, 4 June 2010

[22] ‘High Speed 2 Demand Model Analysis’ section 10.4.3

[23] Tag Unit 3.5.6, Dft, March 2010

[24] 24% of Virgin Rail’s customers are travelling on business (source ‘National Passenger Survey Wave 2,1 Autumn 2009, Virgin Trains’ Passenger Focus, page 21), which is forecast by HS2 Ltd to increase to 30% with an overall demand that increases by 267%

[25] ‘ National Population projections 2008 base’ ONS, 21 October 2009 , Table 1

[26] HS2 Ltd gives this explanation of the assessment that they had made in a meeting on 17 August 2010

[27] ‘Interdepartmental Working Group on Blight : Final Report’, December 1997

[28] by HS2 Action Alliance working with Councils, eg Buckinghamshire County Council

[29] ‘Outline for Technical Annex’, 091123- ACP technical note, HS2 Ltd, page 19 ‘Remaining Issues’, This document records issues raised by the Analytical Challenge Panel

[29]

[29]

[29] September 2010