EHS0037/EAC/14-15

HS2 Action Alliance—Written evidence

HS2 Action Alliance is a not for profit limited company formed with the objective of improving Government decision making on High Speed 2 (HS2).  We are supported by over 100 affiliated groups and over 15,000 registered supporters.  HS2AA adopt an evidenced-based approach to challenging the business case, the environmental case and compensation arrangements.

Summary

i)          There is no robust economic case for HS2:

• There is no commercial case for HS2, even on the Government’s own evaluation.
• The benefits are greatly inflated, particularly:
• the value of the time savings are exaggerated, as there is no relevant evidence that businesses would pay the full cost of their employees time to reduce the time that they are actually working.  Government ignore the fact that time on-board is more useful than it was
• the proportion of current long distance rail business travellers are a third higher than in all previous business cases, despite reliable evidence that previous estimates were correct
• relief of crowding is a product of using an unrealistic comparison basis for HS2, and could not exist in reality
• benefits are assessed as being constant over a time period during so long they may have largely ceased to exist, as they are not robust to likely technological change.
• The costs exclude major items ie cost of capital, cost of new infrastructure needed to connect HS2, uncompensated blight, and it underestimates others - rebuilding Euston, operating costs.

ii)        The benefit cost ratio, properly assessed, is less than 0.5 (“poor” value for money), meaning the subsidy is more than twice the size of HS2’s economic and social benefits. 

Neither the standard case, nor the sensitivities tested cover realistic outcomes, let alone explore the downside risk.

iii)      HS2 is unnecessary:

• The relationship between domestic transport and economic growth no longer exists, and building new transport infrastructure cannot be relied upon to deliver economic benefits
• Travel demand is unlikely to materialise on the scale predicted by Government because:
• domestic travel in general (by all modes) is saturated
• the particular reasons that have driven long distance rail growth will not be sustained
• long distance rail demand has already flat-lined
• If the predicted demand occurred it could be met more cheaply by improving existing railways
• Creating capacity that is not required, as HS2 would, is a waste of resources
• The WCML railway has considerable spare capacity – more so than other comparable routes
• The level of demand required for HS2 is unlikely to arise, but because HS2 is an ‘all or nothing’ project, the expected losses are very much larger than the Government predict.

iv)      There is no reasonable basis to assume that HS2 will redress regional disparities, or cause regeneration in the regions:

• Academic opinion is that such connections benefit the capital city
• Jobs created near stations are likely to be simply redistribution from the hinterland
• The Government’s KPMG study has fundamental flaws that make its results ‘meaningless’
• High speed rail itself has very small agglomeration benefits

v)        The October 2013 business case contains a number of departures from previous business cases.  Some are predictable and reasonable, eg using later version of the forecasting model; others are unexpected and lack justification (eg level of business use).  But without these ‘unexpected’ changes, the benefit cost ratio would be well below I – even before correcting other defects, such as optimistic forecasts in growth or omitted costs (of financing, Euston, HS2 connections etc).

vi)      Similarly the Strategic Case that HS2 would cause regeneration of the Midlands and North is made by employing a methodology that had already been discredited.

Q1              Is there an economic case for HS2?

1          The Government accept that there is no commercial case for HS2, and that it requires a subsidy which, on their own optimistic figures, is £16.7bn for Phase 1 and £31.5bn for the full Y. 

2          It is assumed HS2 makes an operating profit, but this is highly dependent on optimistic demand (and revenue) forecasts and realising savings (£8.3bn) from the classic network.

3          The Government seek to make an economic case for HS2 using social cost benefit analysis, where economic and welfare benefits which would not form part of a commercial case are taken into account.  Their latest Strategic Case (Oct 2013) identifies benefits £57.7bn for the full Y, costs of £40.5bn and a benefit cost ratio (BCR) of 1.8 which increases when wider economic impacts (WEI) are included to 2.3. Phase 1 has a BCR of 1.4 (1.7 with WEI).

4          Unfortunately the cost/benefit analysis fails to be as claimed an “objective, rigorous and accurate assessment of the costs and benefits of proceeding with HS2”.  It overvalues the benefits (particularly of reduced journey times) assumes unrealistic growth and business use, excludes major costs such as interest and additional infrastructure needed to achieve local connectivity and hence the time savings claimed. A more realistic BCR is well below 0.5, making HS2 very “poor” value for money.

5          We focus on our concerns on the benefits, costs and the BCR under Q1, and demand under Q2.

Benefits

Journey time savings

6          There is no proper basis for the current unit values used for time savings.   This is crucial as time savings account for 79% (£46bn) of the benefits claimed for HS2 and are greatly dependant on savings claimed for business. They concern not just the on-board time (worth £31bn) but time spent waiting, at interchanges etc.

7          Business travellers. Previously the values of time used contained two major flaws: that salaries were unrealistically high (at £70k/a) and the wrong assumption that people wasted their time on trains ie did not work. The salary level has been corrected (with values one third lower), but the second flaw has been disguised.

8          DfT has been forced to admit that time spent on trains is productive, indeed their own studies evidenced it.  One of the reasons that Government themselves promoted train travel is the usefulness of time on trains!  Mobile technology has greatly increased the utility of on-board time, and there are sound reasons for assuming that time will be fully useful by 2026 when HS2 starts. 

9          However, rather than remove this benefit, as businesses would get no more productivity from their staff by getting there 30 minutes quicker, DfT simply relabelled the unit value a ‘willingness to pay’ figure.  This is despite there being no pertinent survey evidence to support the switch or the values adopted[1].   Such relabeling is irrational – it assumes that businesses would pay the same rate to reduce already productive time as unproductive time.  It is inconceivable that businesses are indifferent to whether their staff are working or not!

10      This approach is also inconsistent with the trend to less office based work, with an increasing proportion of ‘briefcase’ employees working from home.  Why would there be a material difference between home and on a train that is worth the entire cost of the time on the train to the employer?

11      While adopting a “willingness to pay” approach is acceptable, it requires evidence to support using the values that are still calculated as if people did not work on trains. Instead they refer to overseas studies that predated the increased utility of on-board time, and generally employ the DfT’s previous approach of the cost of time to businesses.  A more robust approach would be to accept that time saved is 100% productive, and that time savings no longer have a productive value.  We assume, conservatively, that it is 75% productive. This wipes £20bn off the benefits (for the Y).

12      The other time savings (for walking, waiting, station access, interchanges) that make up £15bn of the £46bn benefits are also over-valued.  Previously such matters were not considered by DfT to be the concern of the business but of the individual (for which the lower commuter rate was appropriate).  DfT now use the new business unit cost, despite no evidence that businesses actually do value these matters greatly more than the employees actually affected.  We believe this change is unjustified.

13      Leisure and commuter travellers: Time savings for this group have always been assessed on the ‘willingness to pay’ basis (ie as a welfare rather than productivity gain) which is what they are prepared to pay as indicated by surveys of what people say or reveal in their behaviour.  But the evidence used pre-dates the arrival of mobile technology so the values (£6.04/hr and £6.81 for leisure & commuting respectively) cannot reflect the improved utility of on-board time.  The more useful the time, the less passengers will value a reduction to that time. There is no evidence to quantify by how much the values are now too high: HS2AA have assumed by one quarter.

Current % of business travel

14      The estimated current proportion of long distance rail business travel is not credible.  The latest business case increased the proportion of business users by a third – from 28% to about 38%. This alone accounted for a massive a 52% rise (£14.6bn) in claimed business benefits[2] since the previous economic case.  The new increased level is inconsistent with the evidence.

15      The latest approach remaps ticket type data to journey purpose (using assumed conversion factors), rather than using the extensive readily available survey evidence.

This is despite:

• The fit of the new estimates to the survey evidence is very poor
• The survey evidence is consistent with the previous estimates of business travel
• The survey is large scale, frequent and the basis for commercial rewards, making it highly reliable.

16      Passenger Focus conduct the National Passenger Survey (NPS), which is a 6-monthly large scale survey that gives statistically meaningful results for individual train operators’ customer satisfaction scores.  Amongst other factors it gives the reason for travel, and shows that Virgin Trains (the operator of the fast trains on the WCML) have about 30% of passengers travelling on business, and all long distance services about 26%.  The increase in 2010 coincides with completion of major WCML improvements.  This result conforms to the pre-October 2013 HS2 business case assumption (of 28%).  There is simply no question that this survey is not statistically reliable.

17      The NPS survey also gives route examples eg London to Manchester has 31% business users (Mon. to Sun). This contrasts with 63% from the remapped exercise that covers Mon. to Fri. only.

18      It is consequently hard to justify any increase in the current level of business use.  It is also not supported by the general trends in business travel, (see Q2 below).  We assume it stays at 28%.

19      We note that the one third increase in business users entirely compensates for the one third reduction in their salary rate that DfT were forced to make when we and others exposed it.  Clearly without the increase in business users, the BCR would be much lower.

Crowding

20      Crowding benefits (worth £7.5bn) are ascribed to HS2 on the basis that crowding will be less on HS2 than the “do minimum” alternative (which is based on currently committed developments only).  The business case assumes that these crowding benefits continue to apply through to 2092.  The ‘do minimum’ is unrealistic as it assumes that demand grows entirely unconstrained by the absence of capacity to carry it.  In reality improvements occur constantly.

21      As an example, the Stafford area works received approval from the Secretary of State earlier this year.  The works will deliver more capacity (for long distance, local and freight trains) and journey time reductions.  Re-signalling across the existing network will occur (several times) before 2093, permitting greater top speeds, shorter journey times and increasing capacity.

22      The crowding benefits are entirely an artefact derived from the use of an unrealistic ‘do minimum’ comparator, rather than a realistic and credible alternative.

23      The crowding benefits from Phase 1 alone are £3.9bn[3], which is more than the cost of alternative ways of providing the capacity that would address the overcrowding[4].

Reliability

24      The economic case assumes that HS2 will deliver improved reliability (worth £5.5bn).  This is despite the fact there will be 18 high speed trains an hour each way (that no one else in the world delivers, even France only has 12), and HS2 is not a ‘closed’ high speed rail system ie it must interact with the classic network as many services are partly on the classic network.  It is hard to see how such improvements are deliverable.

Costs

25      The case for HS2 fails to include some major items:

• The cost of capital – funding for HS2 is currently treated as ‘free’ money
• The cost of connecting HS2 to existing infrastructure – these are serious omissions
• Incorrect cost escalation – a mismatch in the treatment of costs and benefits
• Social dis-benefits eg uncompensated property blight – their omission is illogical

26      Cost of capital omitted: DfT ignores its own guidance[5], assessing HS2 without any inclusion of the cost of capital.  This is a massive item, a real 2.5% charge on £50bn (neither infrastructure nor trains are ‘free’) is £1.25bn each year, and cannot be dismissed by claiming that HS2 would be funded directly out of taxation, or of selling the completed railway into the private sector.  This has an NPV (2011) of £17.5bn, in itself adds more than 50% to the subsidy required (the net cost to government) and reducing the BCR to 1.2 for the full Y (or 1.5 with WEI) without any other correction.

27      Even if paid for out of tax, the money still has an opportunity cost from the borrowing that is required or repayment of debt that is prevented.  If sold to the private sector, it is inescapable that the owner will require payments to give a return on the investment – creating a clear cost that must be met from operating revenues and subsidy.

Connecting HS2

28      Euston dispersal: In a recent parliamentary debate on HS2, Frank Dobson reports[6] that HS2 Ltd now accept that Crossrail 2 (a £20bn cost) is needed to resolve an acute passenger dispersal problem. While some passengers will leave HS2 at Old Oak Common many continue into Euston.  DfT/HS2 Ltd estimate that the number of passengers using the Underground in the three hour morning peak will more than double by 2041 (from 24,682 to 56,420), and the Victoria Line is already at peak capacity today.  Crossrail 2 is a large omitted cost from the 2013 Strategic Case for HS2.

29      Euston rebuild: Frank Dobson reports in the same 9 September debate on HS2 that the cost of the Euston rebuild is now estimated by HS2 Ltd at £7bn rather than the £2bn in the Economic Case.

30      Out of town stations: HS2 is planned to have out of city centre stations at Old Oak Common, Birmingham airport, Toton, Sheffield (Meadow Hall).  To provide effective public transport additional investments are needed at all these locations except Old Oak Common (where although TfL plan no Crossrail station, the cost is within the HS2 Case).  Without these other investments, costed at £2bn by Taxpayers Alliance[7] HS2 will lack sufficient connectivity to offer their claimed journey savings.

31      Strategic imperatives: HS2 for the purpose of its costing has been de-scoped.  The costs of the connections to Heathrow and to HS1 are no longer included in the costing.  Both these links worsen the benefit cost ratio of HS2 but were previously described as strategic imperatives.  Connecting to Heathrow was even part of the original HS2 remit.

Incorrect cost escalation:

32      The costs of factors such as labour (eg train divers salaries) are escalated only to 2036, while the benefits including that based on assumed real earnings (eg value of time savings) continue to escalate for the entire evaluation period (through to 2092). So not only do benefits continue unabated to 2092 (itself questionable), but they are increased in value for the entire period while costs stop growing far sooner.

33      This is clearly inconsistent and underestimates the costs compared to the benefits.  Escalating the costs in the same manner as the benefits adds an additional £6.6bn to the operating costs and hence the subsidy required for the full Y:

• £2.3bn extra in wages/salaries
• £2.8bn extra in maintenance costs (based on 1%/a increase)
• £1.4bn extra in electricity costs (assuming price increases in real terms at 1%/a)

34      Diesel costs, which are used to compute a saving from reduced classic services, are assumed by HS2 Ltd to escalate throughout the evaluation period!

Social dis-benefits forgotten:

35      The social benefits of HS2 (eg transport user benefits) are included in the assessment but the social dis-benefits of the loss in property values near to the route are not included – with the exception of those few properties that HS2 Ltd expects to acquire under its compensation arrangements (eg under Safeguarding, the Voluntary Purchase and Need to Sell schemes).  HS2 Ltd assess the cost of land purchase and their compensation schemes at £2.5bn for the Y.  But this omits the cost of blight that remains uncompensated and must be borne personally by both individuals and businesses. 

36      HS2AA estimate excluded uncompensated loss (on blight out to 1km) at just under £10bn (the Y).

Benefit cost ratio (BCR)

37      There are also concerns with the BCR itself, particularly given the uncertainties:

• The “standard case” results are unsound – inappropriate use of value for money test; its ‘do minimum’ assumption, the period of evaluation and assessment of demand/ use of the cap
• Inadequate sensitivity testing of the results

38      Value for money test: The value for money test that is indicated by the BCR is exaggerated, as the categorisations of ‘poor’, ‘high” etc values should apply to the BCR without Wider Economic Impacts (WEI). Government instead use the figures including WEI – presumably it is only these that are large enough to count as “high” ie exceed the threshold of 2 (£2 benefit for £1 spent). 

39      The “do-minimum” assumption: the case compares HS2 against the current position improved only by committed rail enhancements ie a “do minimum” package out to 2092 (the benefits period). This is a serious flaw as it is wholly unrealistic:

• It freezes enhancements at the level currently committed (in 2013) despite the fact that they will and do keep happening
• It assumes like-for-like renewals, implying the perpetual renewal of obsolete technology.  In reality renewals is always an opportunity for improvement eg ERTMS signalling which is applied to HS2 will be implemented across the intercity network long before 2093 (most likely before the second phase of HS2 is complete) will increase capacity and reduce journey times
• It means that the assumed demand growth is far too high for the capacity that is provided under the “do minimum” ie the “do minimum” is artificially constrained and the demand is projected without reference to supply limitations hence the spurious crowding benefit attributed to HS2. In reality capacity constantly evolves, eg  to accommodate demographic change from a growing London, through Crossrail, Thameslink, longer platforms etc

40      Best practice in cost benefit analysis requires assessment of options against the best alternative, not the status quo (ie the ‘do minimum’ is) which would not be credible or feasible to 2092.

41      Curtailing benefits much sooner: benefits are evaluated out to 2093 (60 years from Phase 2 first being operational).  This is an unacceptably long period, particularly because of the inherent future uncertainties eg due to technical change, discussed under Q2. It is beyond even the period for which the Office of Budgetary Responsibility produce figures for economic growth!  A shorter period eg to 2050 should have been used. This gives a BCR of ‘poor’ value (at 0.49)[8] as we wrote to labour about.

42      Sensitivities tested: Not only is the ‘standard case’ unjustifiably favourable to HS2, but the sensitivities focus on the up-side potential.  Given the uncertainties, the costs and history of inadequate previous assessments eg for HS1[9] this is unsatisfactory:

• Lower rail growth: DfT appears to test lower rates of growth but in reality they do so by just extending the year when the demand cap is achieved.  In combination with assessing benefits over such a long period the reduction is cosmetic. They do test the scenario in which growth is capped at 2027[10], in which the BCR with WEI is reduced to just below 1.5 for the full Y (from 2.3).  But demand is still increased 40% - which is nearly three times the population growth to 2036.   Instead they should have assumed it grows in line with observed general domestic transport growth, with rail travel growing at only the rate of population increase (which itself may be optimistic as developments in other modes, eg road, may win back market share).  This gives a BCR of ‘poor’ value for money, ie below 1 (see Table 1).
• Journey time savings: these dominate the benefits and are highly questionable.  The DfT should have considered the BCR if the on-board time savings to business have no value, as this time is productive. Instead they consider values of time that are actually even higher[11].
• Lower business demand: the business traveller percentage should have been assessed using the Passenger Focus derived survey evidence – not those from the ticket type re-mapped in a new way.  The survey results were not even used in sensitivities.  The ticket type basis resulted in an increase in benefits by 1/3rd for Phase 1 and 25% for the fully Y[12].  It is extraordinary that the position without this change is not given in the range of BCR results, particularly given its very questionable basis.
• Classic rail savings not achieved: the costs of not making the £8.3bn savings in classic services should have been tested.  This might be because of delays to completing the re-modelling of Euston (making mass transfer to HS2 from WCML unlikely or materially delayed), or, price competition causing demand for the classic services to continue.  Such an outcome would also prevent creating the re-use of freed up capacity.

A more realistic BCR

43      A realistic assessment of HS2 would have the ‘standard case’ with a benefit cost ratio below 0.5, which is very poor value for money (requiring a subsidy at least twice the benefits). Table 1 shows the appropriate changes from the Government’s ‘standard case’ and the resultant benefit cost ratio, using more realistic assumptions.

44      The table does not take all the factors we have discussed above into account.  This is partly because the combined effects are difficult to assess.  However, it is indicative of the scale of the concerns HS2AA have for the economic case:

• On the benefits we show individually the impact of three large items – business value of time, business use and crowding.  We did not include reliability or capping the benefits at 2050.
• On the costs we cover including the capital charges and the cost escalation correction.  The table omits many items: HS2 connections; uncompensated property blight, Crossrail 2; Euston rebuild; un-achievability of classic service saving; and environmental impacts).
• For demand we assume population growth (15% over the period[13]) - discussed under Q 2.

45      In summary, there is no economic case for HS2 and it is hard to make sense of key elements in the remodelling of benefits for the October 2013 Strategic Case, unless the purpose was to ensure a favourable result.  The evidence on the best available data shows that HS2 is in fact very poor value for money. 

Q2              Should the Strategic Case for HS2 published in October 2013 by the Department for Transport and analysis from HS2 Ltd have taken account of any other factors in making an economic case for the project? Is the expected range of the benefit cost ratio persuasive?

46      The Government in the Strategic Case argue that there are strong economic and demand based reasons for having HS2.  This is based on a number of false assertions:

• That there is a continuing relationship between economic growth and growth in travel
• That the trend in transport growth is faster than that used for HS2’s demand forecasts
• That stopping the assumed growth in 2036 is cautious and that the economics of HS2 will be better as growth will continue
• That high speed rail gives rise to regeneration (including in East Kent for HS1)
• That the environmental impacts are modest, and positive for carbon
• That there are no satisfactory alternatives to HS2
• That the costs are under effective control

47      So there are a range of factors that the Strategic Case should have taken account of, and did not.

Economic growth and transport

48      There is no longer a relationship between economic growth and domestic transport – crucial to the HS2 case – in either the UK and other developed economies[14]. 

49      More than 30 eminent transport professors (including advisors to HS2 Ltd) wrote an open letter to the Secretary for Transport in the Financial Times[15] pointing out that the evidence base for the link between improved transport infrastructure and economic growth and more jobs can no longer be relied upon. They emphasised in their letter that as Eddington had said in 2006, the UK is already well connected.

50      David Metz[16] has demonstrated that the link between economic growth and domestic transport is broken.  The graph below shows that while there was a clear relationship until about 1997, since then economic growth is not mirrored by a corresponding increase in domestic travel (the top line that shows distance travelled). Incomes are rising but we are not travelling more.  This remains true with the subsequent recovery in the economy since 2010.

51      If a cross-sectional analysis of travel is made, there remains a clear relationship between income and travel, with the affluent travelling more than those with lower incomes, eg MPs travel more than factory workers.  This reflects the different roles within society having different characteristic travel demands.  But because the overall demand for domestic travel is saturated, total travel per person has not grown as the UK has got richer.  Metz suggests it is not income that constrains us each travelling more but the time individuals wish to spend making visits. 

52      DfT offer a graph of rail demand and GDP[17] as evidence of a relationship between rail growth and GDP.  There is such a relationship as central London employment is a key driver for commuting rail demand in London, which is an important component of total rail demand.  But as discussed below, business travel is reducing per capita, and hence is not a driver of economic growth. 

Demand

53      HS2 are at pains to argue that their demand estimates are conservative.  This is transparently incorrect, when their estimates are set in the context of recent trends in domestic travel.

54      A study of more than 200 infrastructure projects[18] showed that there is a tendency to overestimate passenger forecasts by more than 100% on rail projects. HS2 does not look to be an exception. The demand modelling exaggerates demand for HS2 because:

• The projected background growth is unlikely to happen
• The uplift in demand assumed from introducing HS2 is too large
• No allowance is made for new technology that will remove rail’s current competitive advantage (from mobiles etc making on-board time useful) , which may reduce rail demand

Growth in travel

55      Background long distance rail travel growth, ie without HS2, is unlikely to happen because:

• Domestic travel (all modes) has been a saturated market for a long time
• Long distance rail demand has now flat-lined
• The demand model used to predict large future rail growth is defective, and produces an estimate of growth 5 times greater than population growth. 

56      Market saturation: The overall domestic travel market is saturated.  UK domestic travel per person has not been growing for nearly 20 years (this is not just a UK phenomenon, but what happens in affluent societies).  Total long distance trips (over 100m) per person has been at about 7 per year.

57      Any further growth in long distance rail will therefore be at the expense of other modes.  But due to changes in technology, it is more likely that long distance rail will lose market share – particularly to automated cars.

58      Business travel generally is declining: Domestic business travel is also not growing in absolute terms (in kilometres per annum).  It is not growing measured with either GDP or population.  Since 1996 GDP has grown by 40%, population by 10% (and 5% since 2005) but business travel has fallen by 10% (and 16% since 2005). So a link to GDP growth no longer exists.

59      Long distance rail demand has flat-lined: evidence shows long distance rail travel has plateaued.  The strong growth in long distance rail demand following privatisation in 1995 has stalled, with no growth for 2.75 years for passenger km and 2.25 years for passenger journeys, despite the economy rallying.  Modern Railways highlighted this in August 2014 article “The growth delusion”.

60      Rail growth has not always occurred.  Rail travel was stagnant from the Second World War until the mid 1990s when particular factors applied eg a large rail subsidy increase.

61      Defective rail demand model: The rail modelling for HS2 that predicts future rail growth (at 2.2%/a for next 25 years or 79% increase by 2036) is based on a version of the Passenger Demand Forecasting Handbook.  It is defective and results in too optimistic a forecast because

• The utility of travelling time is ignored. It takes no account of the recent improvement in the utility of on-board train travelling time in explaining the growth of long distance rail.  This will have caused the overall cost (including journey time) of rail travel to fall.  There was a major review of the rail forecasting model by Oxera/Arup (in 2009) which failed even to consider this as a variable to explain the increase in rail’s market share[19].
• No allowance for technical change: It has a blinkered view of technological advances and inter-modal competition.  Changes in the utility of travelling time would explain the countercyclical growth in rail long distance travel.  Over the relevant timescale rail’s advantage in utility of travelling time is likely to erode rather than increase. Future technology (Google or self-driving cars, mobile technology on aircraft, etc), will eat into rail’s current competitive advantage in providing an environment suited to work or leisure activities compared to other transport modes.  Ignoring this over an 80-year period seems certain to result in over-estimation of rail travel and over-optimistic conclusions about HS2.
• GDP and rail growth assumptions are also not robust: It assumes that increases in real income result in somewhat larger increases in long distance rail travel.  But this relationship is incorrect. Long distance rail has grown during the recession, but stagnated when the economy came out of recession and subsequently. 
• It is a short term forecasting model: It is unsuited to long term forecasting, because it has fixed elasticities.  Sir Roy Eddington recommended such models be used for 10 year forecasts – not 25yrs. 25yrs prior to the current (2011) base year was 1986 (as the previous long term graph shows) – long before mobiles and broadband. The model is fitted to the circumstances of the past rather than the determinants of future transport requirements.
• Price competition between routes is ignored.  Different routes are assumed to compete on the basis of journey time and crowding only and not price.  This is unrealistic. It will cause a misallocation of demand to HS2 from other rail routes. Plainly passengers also take price into account in making choices (eg Chiltern line to Birmingham is one third cheaper than Virgin but takes longer). It has also been assumed, quite improbably, that HS2 would not be a premium priced railway – when all other high speed services in the UK and overseas are.  This exaggerates forecast demand for HS2, as Public Accounts Committee forcefully pointed out.

62      As a consequence demand might be expected to grow at most at the rate of population – ie at 15%, compared to the five times greater figure assumed by DfT in the economic case (79%increase to 2036). HS2AA assume a 15% (14.5%) increase in their revised quantification (this equates to a 0.5%/a increase in long distance rail growth).

Uplift from introducing HS2

63      The uplift in rail travellers when HS2 becomes available is also unlikely to be as large as forecast.  This is because the reductions in journey time are now less valuable than modelled.  HS2 Ltd currently assumes 26% of its passengers are new travellers, this will be an over estimate, as it is based on the reduced journey time being massively valuable to potential travellers (and un-eroded by the inevitable improvements from technical advance and numerous minor improvements).

Capacity

64      The Government’s case ignores the real position on existing and available capacity.

65      It is argued by the Government, for example by Patrick McLoughlin and Baroness Kramer[20], that additional long distance capacity is urgently needed on the WCML, and this underpins the need for HS2.  But the WCML currently has considerable spare capacity.

66      First, the railway carriages are half empty. Data released to the High Court (as part of 2012 HS2 Judicial Review challenge) showed that the average evening peak load factor (the ratio of passengers to seats) for InterCity services from Euston was only 52% (or 229 passengers/ train).

67      Second, this loading data is before 35 out of the 56 Pendolino trains used on the route were lengthened from 9 to 11 cars, completed by 2013 giving an additional 150 standard class seats in each train. This reduces the 52% evening peak loading to about 43%.

68      Third, it is the commuter trains and not intercity services (that HS2 is designed for), that have the chronic overcrowding on WCML, as any commuter knows, confirmed by standing passenger data, and the analysis reported in the Strategic Case. Euston[21] itself is the second least crowded London station (after HS1 into St Pancreas!).

69      Government attempted to shift the issue to say HS2 was also to meet freight and commuting needs by freeing up space on the WCML.  But there is no anticipated rail freight capacity shortage that HS2 addresses (the freight issue is north of Rugby), and maintaining services to intermediary stations to London leaves little/no spare capacity on the fast lines for more commuter services. 

70      Fourth, there is proven potential to increase capacity on the existing WCML route without HS2:

• Reducing the number of first class carriages from 4 to 3 per train and running longer trains (increasing to 12 car everywhere but Liverpool).  Together this creates 693 seats a train – three times the average evening peak loading (229) – more than the growth DfT forecast
• Relieving three pinch points on the WCML would allow more trains to run (extra commuter trains to Milton Keynes and Northampton and 15 extra intercity services in each direction as well as increasing freight capacity by grade separation).

71      This solution was developed and promoted by 51m (a cross party alliance of local authorities opposed to HS2). The 51m ‘optimised alternative’ was costed by Network Rail at £2bn and Atkins for DfT gave it a net benefit ratio of 5.17 (compared to HS2 at under 2 at that time).  Disruption to create the extra capacity is minimal and similar to ongoing current works.  At the very least the HS2 cost benefit analysis should have used the 51m as part of the best alternative instead of using the “do minimum” as the base comparator.  The 51m solution is risk free, incremental and can accommodate more than forecast long distance demand. 

72      Interestingly although the Government insist that the WCML is full, Network Rail are now accepting the use of more train paths.  Network Rail have agreed with Alliance Rail that there is capacity to operate an additional 12 InterCity trains a day on the WCML (six in each direction between Euston and Blackpool and six between Euston and Leeds via Manchester Victoria and Huddersfield)[22]  Network Rail have also agreed the timetable for September 2014 to include extra fast commuting in peak that were previously thought to have required the Ledburn Junction grade separation.

73      Fifth, the Strategic Case compares the capacity against that which would be available with HS2.  This is misleading: there is no virtue in creating unused capacity (it is a waste of resources).  DfT and Network Rail have conceded that the 51m approach creates more than sufficient capacity to meet their forecast demand.

74      Given there is no difficulty in meeting the demand that the Government forecast with modest improvement to the existing network, and that renewals will as a matter of course increase capacity and reduce journey times, betting £50bn on building a new railway that depends on high demand even to deliver poor value for money is perverse.

Environmental implications

75      HS2 is opposed as needlessly environmentally damaging by a range of environmental pressure groups including the Green Party – some of which support high speed rail in principle (eg CPRE).

76      Landscape impacts: The environmental implications of HS2 are not reduced to monetary values in the assessment, although a woefully inadequate attempt is made to assess the landscape impacts.  These are valued at a £1bn dis-benefit but on a methodologically unsound basis that underestimates its impact eg it places the lowest value for some of the land in the only Area of Outstanding Natural Beauty it cuts through, and truncates the period of assessment without allowance for restoration.

77      Landscape impacts and the loss of ancient woodland are major considerations, because as proposed HS2 does not follow existing major transport corridors, it permanently spoils the character of areas that are relatively tranquil.

78      Health impacts: HS2 also damages people’s health and wellbeing in the areas through which HS2 will pass. This was not covered by the Strategic Case and the Health Impact Statement was not even consulted on.

79      Comparison with HS1: HS2 would be far more damaging that HS1, which is frequently used as exemplifying the expected effects.  Noise increases greatly with increasing speed (it massively increases in volume and becomes more episodic (with shorter onset and attenuation making it more noticeable and disruptive), and harder to screen (as with speed it is increasingly aerodynamic noise and generated by the body or above the train (by the pantograph) rather than by the wheel rail interface).  HS2 does not follow existing transport corridors, while HS1 extensively follows the M2, and needs to be straighter because of its 400km/h design speed.

80      Slower speeds: HS2 has devastating environmental implications largely because to achieve the very high speeds it cannot be routed to follow existing transport corridors.  Neither the Strategic Case in 2013 nor the first economic assessment in 2010 examined alternative routes based on slower speeds eg at 300k/h (the maximum HS1 speed).

81      The Government say that dropping the speed to 300k/h adds 4.5minutes to the journey time[23] for Phase 1.  Given the journey time savings are substantially over valued (and this leads to a distortion in decision making, as remarked by Public Accounts Committee[24]), the economic case for HS2 should now be tested against a range of less environmentally damaging alternatives.

82      Carbon emissions: HS2 claim that HS2 reduces carbon emissions over its life – 120 years.  This is because the HS2 carbon assessment is based on a number of inappropriate assumptions:

• That HS2 will cause a transfer of freight from road to rail, but no case is made for this and there is no evidence that HS2 would relieve any capacity constraint on carrying freight by rail
• The carbon impact of additional electricity consumption from HS2 is assumed to be the average carbon impact of electricity generation, not the marginal emissions of additional peak consumption which would actually occur.  This means that the generation to supply HS2 would add to carbon emissions until the generation system is almost entirely de-carbonised
• The boundaries of the ‘system’ for assessing carbon emissions are drawn to exclude:
• The energy requirements of some of its construction, and materials for construction
• The potential re-use of any freed-up airport runway space for long haul flights.
• Assumes operational carbon savings for a period of 120 years from starting operation is absurd.  These benefits depend on an absence of change until well into the next century.

83      Using reasonable assumptions HS2 substantially adds to carbon emissions.  Speed isn’t green.

Opportunity cost of HS2

84      It is surprising that a Strategic Case for HS2 should not adequately review alternative investments, or how it fits as part of a coherent UK transport policy. 

85      DfT admitted to the Public Accounts Committee (in 2012, in the context of lessons to be learnt from HS1) they had not considered any range of alternative costs and benefits to HS2, eg on local services or broadband. 

86      The Strategic Case in 2013 now says that alternatives like broadband have been examined.  But the stance taken is (that they will happen too) is not an evaluation of them as alternatives.  The current plans for broadband involve very limited expenditure (under £2bn).  Investment in rural broadband seems a more credible way of creating employment opportunities in the regions (by giving them access to global markets) than giving a few cities faster access to London.

87      Curiously the alternative of using price to constrain demand, up to the point where there is a commercial case for additional capacity, is not considered.  Travel is energy intensive, and is in decline for business purposes.  Travel by car, which is the dominant mode of travel for most of the population, is constrained by taxation that exceeds the direct cost of its provision.  It is also constrained by the failure to address pinch points that have greatly better BCRs than rail investments[25].  The environmental justification for this is disappearing with electric cars which, because they store electricity, will be de-carbonised far earlier than electric trains.

88      The discussion of rail alternatives to HS2 in the Strategic Case is also misleading:

• The alternatives generated by the Government are not realistic.  They produce far too much capacity for the demand that even the Government forecast.  For example there is simply no need to electrify the connection between Doncaster and Sheffield for Phase 2.  Creating capacity in excess of demand is wasting resources that could be put to better use, and its inclusion in the alternatives serves no purpose but to worsen their economic case.
• While it accepts the 51m alternative is realistic for Phase 1 (and far cheaper) it says it produces less capacity than HS2.  It should instead have compared it with the forecast demand.  Network Rail in their review of the 51m alternative did not dispute it could meet all the required demand!
• An advantage (besides costing very little) of the 51m solution and other improvements to the existing railway is that they are incremental and so could be implemented to the degree that emergent demand is required. In contrast HS2 is essentially an all or nothing scheme highly dependent on the high levels of long distance rail demand that are unlikely to arise. This advantage is completely ignored.

89      Clearly there are alternate plans that could be focused on the north eg the £15bn “One North Transport Plan”. These should have been covered in any comprehensive evaluation.

Costs under control

90      The Strategic Case argues that costs are under effective governance. 

91      The Major Projects Authority (MPA) has made independent reviews of the project, its governance and the prospects of it being successfully delivered.  They have classified HS2 as ‘amber red’ meaning that its success was assessed as ‘in doubt’.

92      The November 2011 MPA report raised the project’s alert status from ‘amber’ to ‘amber-red’.  This status has been maintained in subsequent assessments including that in September 2013.  The assessors said they had low ‘delivery confidence’ in HS2, according to sources familiar with the report[26], and identified four core weaknesses in it:

• The timetable is risky, with too little time allowed for the passage of the hybrid bill
• The management and governance arrangements are inadequate, with a confusion in roles between DfT and HS2 Ltd
• The management methods were inadequate, without the use of even a ‘critical path’ method
• The skills, capability and resources deployed were insufficient.

93      The Government vetoed the release of the report, but this decision is subject to judicial review.

94      While the original report that was suppressed is now quite old, HS2 has had the same rating by the Major Projects Authority repeatedly since, and, as Andrew Gilligan reports these issues are echoed in DfT’s recent internal audit report.

95      It is plain that the Government are concerned to conceal an on-going lack of control of the HS2 project.  Since the MPA first raised the alert status to amber-red, HS2’s costs have already risen significantly – from £32bn to £43bn for the infrastructure (in 2011 prices).

Subsidy justification

96      It might be expected that the Strategic Case explain why HS2 should merit a £30bn+ subsidy.

97      Long distance rail is used by a tiny and minority (0.18% of all trips are long distance rail) as the next graph shows and rail users are relatively affluent (47% are in the 20% income households).  It is also clear that on the evidence, HS2 will not generate material economic activity, rather it will be a drain on the economy.  This makes it a strange priority for subsidy.

98      While a local bus services subsidy is justified by beneficial effects on the labour market (bringing economically inactive people into employment), a similar case cannot be made for long distance rail.

99      Interestingly, 60% of MPs are said to be long distance rail users, which may give them a skewed appreciation of the importance of long distance railways. 

Range of benefit cost ratio (BCR) outcomes are not persuasive

100  Under Q1 we expressed concerns about the BCR.  The Strategic Case even created longer term growth scenarios eg out to 2049.  These cannot be credible outcomes on which to base any decision. They rely upon a relationship between economic growth and travel that is spurious.

101  The case for HS2 uses an unrealistic benefits timescale – another 78 years to 2092, and 120 years for the carbon assessment. Over this period, the changes in technology, social patterns and human behaviour are unforeseeable.  It is unsound to set the benefits so derived against entirely foreseeable costs.  Looking backwards 78 years illustrates this: in 1936:

• There was almost no air travel
• Few people had private cars
• No digital devices existed including any form of computer
• Few homes had even a telephone
• The world’s first regular public broadcast of television was started by the BBC that year
• The NHS had yet to be created
• The 2nd World War had yet to be fought
• Nuclear weapons had yet to be invented
• China was in no sense an economic power

Q3              What are the likely economic benefits of HS2 to the Midlands, to the North of England and to Scotland? Do they also depend on complementary action by governments, local authorities and Local Enterprise Partnerships, for example measures to attract investment and skilled workers?

102  HS2 is being sold on a completely false prospectus.  An HS2 Ltd spokesman[27] recently claimed:

“Analysis by KPMG has shown that HS2 could boost the economy by up to £15bn per year when operational. HS2 will rebalance the economy, skill a generation and support new homes and jobs”.

103  Not a single element in this claim is supportable.  What is more disturbing is that the Government must know this.

Likely economic benefits

104  The claimed benefits, of more jobs and wider benefits, are not justified because:

• The weight of academic evidence finds find little/no supporting empirical evidence to say the regions will benefit, and inequalities reduce. London is likely to benefit most
• While localities in the North and West Midlands close to new stations may benefit, it can be at the expense of their hinterland ie not a net regional benefit
• Reports for Government, such as that by KPMG have been shown to be based on false assumptions and are not robust statistically
• Even regeneration-related benefits at stations are not guaranteed. IEA research into the impacts of HS1 on East Kent do not reveal regeneration benefits
• To gain even these limited benefits requires complementary investments beyond that in the HS2 cost envelope.

105  Simple competition theory sees distance as a barrier to competition, with improvements in transport reducing this barrier.  London completes with regional cities for the provision of a range of professional services, and while London is the most efficient location at service provision, distance affords regional cities some protection from London’s greater efficiency.  So without other factors, reducing the cost (time or money) of transport ought to benefit London at the cost of regional cities.

106  The empirical evidence suggests that this ‘theoretical’ effect actually occurs with high speed rail. Prof Tomaney’s report on the evidence confirms this:

‘The report examines evidence of the experience of five countries where HSR has been introduced to assess its impact on their economic geography. Taking this evidence in the round it is very difficult to substantiate the argument that high speed rail is likely to have a positive impact on regional inequalities. Cities which are the location of HSR stations may gain some benefits, but distribution of net benefits needs careful analysis. Some of the benefits accruing to regional cities may be at the expense of neighbouring places, while in countries with dominant capital cities net benefits tend to accrue to these.’ Transport Select Committee written evidence HSR14, 2011

107  Work done by Prof Tomaney[28] shows that experience from actual high speed railways fails to show the effect claimed by Government, and that there is a tendency to increase centralisation.

108  David Higgins in his HS2+ report highlighted the concentration of top companies in London and the opportunity to decentralise. He failed to note that in France (with its more extensive high speed rail network) 82% of CAC 40 companies are based in Paris. None in cities served by the TGV.

109  Agglomeration benefits: The work done by Graham and Milo[29] shows that, on the empirical evidence, in practice the agglomeration effects of high speed rail are very small.  This work was commissioned for, and published with, the 2010 case for HS2 by HS2 Ltd. However, its conclusion was misquoted in the 2010 White Paper to give the impression that such agglomeration benefits might be material.  The work had concluded that they would be worth £10million at most.

110  Back-office services: It has been argued that HS2 would make the location of ‘back-office’ activities in the service centre to the Midlands and North more attractive, as this would allow the utilisation of less expensive labour than in London, but with ready senior managerial access from London.  However, it is unclear what scope there is for this, beyond current practice.  ‘Back-office’ provision is relatively lowly skilled, and UK locations compete against low cost providers (eg India), with the internet and telecoms increasingly providing business integration (rather than personal travel).

111  There is no good evidence that HS2 will have any beneficial effects to the areas it connects, except possibly London (but further concentration of economic activity in London has costs).  There is likely to be a concentration of development near stations, but at a cost to places in their hinterland.

KPMG

112  The KPMG analysis, commissioned by Government and published in 2013, shows something different.  But there are serious problems with it technically and with its assumptions: 

• The method of statistical assignment of effects is defective (leading to major overestimation)
• The assumption that transport linkages are the only constraint on business location is transparently false (skill, in particular, is an issue for Midland and Northern cities)
• There is a complete failure to recognise the absence of a concentration of appropriate skills in regions, and the consequential gravitational pull of London for those with the required skills and aptitude for professional services.
• The use of a future date for money values is also misleading as it inflates them compared to the other numbers standardly used for HS2 – which are on a current or 2011 basis.

113  The problems with KPMG’s approach were well appreciated before Government commissioned the study.  Mackie and Laird[30] conducted a detailed critique on behalf of the Northern Way in 2010.  KPMG’s approach is essentially unchanged.  KPMG’s study for Government was savaged in expert evidence from Profs Graham and Overman at the Treasury Committee meeting (5 Nov. 2013).  Prof Overman suggested that KPMG had overestimated the benefits by a factor of between six and eight.

114  Expert opinion is that the benefits estimated by KPMG are not additional to those captured in a conventional transport appraisal with WEI[31].  The key additional information that the conventional analysis does not capture is the regional distribution of benefits.  Unfortunately, the failure to take into account skills and land availability mean that the KPMG analysis is unhelpful for even this.

IEA research

115  Dr Richard Wellings[32] (IEA) has published an analysis of the effects of HS1 on East Kent, which demonstrates that the claimed benefits[33] to East Kent have not materialised – rather the reverse – with below average economic growth accompanying the new services.  He also cites Birmingham and Doncaster as locations with good transport connectivity (including by rail) but poor economic performance, which provides a real world demonstration that KPMG’s assumption about transport connectivity being the key constraint on regional economic development is false.

Complementary investments

116  Because high speed rail has a tendency to centralisation, the issue with other initiatives to encourage investment and attract skilled workers to regions is therefore not one of complementary action.  Other measures may be successful, but one might expect a new high speed rail connection to make the task harder rather than easier.

117  However, as discussed above, complementary investments are required for HS2 to be useful in transport terms alone.  The lack of connectivity for out of city-centre stations necessitate additional investment, which is not inside HS2’s economic case.  Even Euston requires investment to address passenger dispersal problem (Crossrail 2) which is not within HS2’s business case – although the time savings which depend upon them are included.  An alternative that might address the Euston dispersal issue is extending Crossrail to replace some commuting services into Euston (as is now being considered), but this will have costs which are not in the HS2 assessment.

118  HS2 is an unusually poor infrastructure expenditure for generating UK jobs, eg compared to road building.  The UK has no indigenous high speed rail industry, and so will import the technology (and most probably the high valued added components) from foreign manufactures.

Q4              Might some parts of the UK suffer economic disadvantage from HS2?

119  As discussed above, HS2 is likely to contribute to a trend of centralisation of cultural and services activities in London, at the cost of Birmingham, Manchester, and Leeds.  The reason for this is that the time cost for visiting London is reduced, thus improving access to London’s facilities from cities on the route.  This puts regional facilities in closer competition to London.  As London has a much greater provision than regional cities, this may adversely affect the viability of regional activities. 

120  In contrast to the results from KPMG’s analysis of the effect on HS2 on regional economic activity, which they see as favouring those connected by HS2 with detriment to those cities which are not, eg Wales and South West, the increased competition with London from HS2 is likely to have the reverse effect.  The economies of cities connected to London might lose service sector activity to London, while the economies of cities unconnected are unaffected.  The assumption that drives KPMG’s wrong conclusions is that it is only transport connectivity that prevents businesses locating in regional cities – rather than (in particular) a shortage of the necessary skills.

Q5              Is London likely to be a main economic beneficiary of HS2?

121  Even KPMG identify London as the largest beneficiary.  It is plausible that London would be the only beneficiary, and itself not be greatly benefitted.  The basis for this is discussed at Q3 above.

Q6              How might the expected benefits of HS2 to the national economy be realised?

122  The claimed benefits are said to be in terms of jobs and regeneration around stations eg Old Oak Common and regions served by the services.  However a net benefit should not be expected.  This is because it will require a subsidy that needs to be financed, yet delivers economic and social benefits that are worth less than the subsidy.

123  Other factors for expecting a net negative effect are:

• The subsidy is likely to be much larger than the Government claim. An on-going subsidy for services is likely to be necessary:
• If the infrastructure is sold off once complete there will be capital charges without fares income or economic growth to pay for it
• Long distance rail ridership and that of HS2 in particular is likely to be much less than forecast, so that fares income would not cover operating costs
• There would be a tendency to strengthen London’s economic and social dominance, exacerbating London’s transport requirements, requiring Crossrail 2 etc. – weakening the economies in the regions and underutilising the less congested infrastructure there
• The UK lacks an indigenous high speed rail design or manufacturing capability, meaning that the high value added activities will be imported
• Property values will increase in the vicinity of the new stations, but properties along the line of route will lose value, due to the adverse local environmental impacts of very high speed trains and the lack of any counterbalancing improvement in connectivity. The positive economic effects are captured in the WEI, but the costs, eg from environmental damage are not.

Q7              How should HS2 be operated? As a franchise in competition with West and East Coast Main Lines?

124  It is not obvious why HS2 should be operated on a different model from the body of the UK railway, unless the Government has different policy objectives. 

125  The core issue is the degree of Government control over the management of the infrastructure and the provision of train services.  In broad terms the options are competition with a degree of regulation, or state management.  The choice will dependent on the extent of price fixing and control of service specification intended to be applied.  Price fixing to prevent alternative rail routes competing on price, as is implied by the assumption that high speed rail and classic rail be priced the same, goes beyond the current level of control, because Virgin Trains, London Midland and Chiltern Railways currently compete on price with their London-Birmingham services.

126  Allowing competition is inconsistent with the current HS2 business case, as:

• Competition and overcapacity will drive price competition and in turn larger subsidies (as prices will relate to avoidable costs not total costs)
• Price competition allows slower services to offer a discount from the price level of fast routes to win business.  This is what London Midland and Chiltern Railways do with their London-Birmingham services today – and win business from Virgin Rail as a result.  HS2 would not have as large a market share with price competition, and cuts to classic services would be difficult to justify when the trains concerned achieve reasonable occupancy levels, so the assumed savings on the classic railway would not be deliverable in practice.

Q8              Should travellers expect to pay higher fares on HS2 than on other lines?

127  Faster routes and high speed routes have to date commanded a price premium.  This is the experience internationally, and also on the high speed Kent commuter services that run on HS1. The Government’s assumption that HS2 would be priced the same as the classic network is unlikely to be the case when HS2 operates (as the Public Accounts Committee emphasised[34]) unless the Government regulate fares to make it so.

128  HS2 will create surplus long distance rail capacity, with a number of undesirable consequences.  With competition the surplus capacity will force down fares, increasing the subsidy to long distance travel, ensures low occupancy of HS2 and inability to cut classic services as planned (and hence preventing planned savings from being realised).

129  It is irrational to increase the subsidy for long distance rail travel, as it is used by the relatively affluent and in encouraging travel it wastes energy.

Q9              Does the prospect of HS3 affect the economic case for HS2?

130  HS2 might be expected to reduce any regional benefits from HS3.

131  Any regeneration benefits to the north of HS3 would be lost if HS2 is built first.  This was Ultraspeed’s conclusion, and one reason why people have proposed building HS2 from north to south – rather than the commercially less costly south to north.

132  Ultraspeed was a proposed Maglev (a very high-speed sort of railway) network, for which supporting analysis by CURDS[35] showed benefits to the North (primarily for Manchester) from connecting the North up first.  However the study showed that this regional benefit was lost to London, if the North-South connection were built first.  This confirms the conclusion that a connection with London sucks economic activity South.

133  The CURDS work for Ultraspeed preceded the analysis by Graham and Milo, which suggests that the regeneration benefits would in any event be modest.

September 2014