Professional Stuff.

What should the railway cost?

An informal background note by

Roger Ford MCILT AIRSE

Industry & Technology Editor Modern Railways

 

Most procurement decisions are backed by a feel for what something ought to cost. When I joined the railway industry in the 1960s the rule of thumb price for a diesel locomotive like the Class 37 was £1000 per ton. A diesel engine cost around £10 per horsepower.

Today, there are similar benchmarks. A new passenger vehicle costs around £1 million, and SSI interlocking costs around £5 million..

As a reality check we can run historic series of costs corrected for inflation. Figure 1 shows the historic cost of installing overhead electrification equipment, excluding associated civil and signal engineering work.

 

Figure 1

 

But in developing the HLOS and SoFA DfT Rail lacks a coherent series of benchmark costs. The restructuring for privatisation followed by the collapse of Railtrack after the Hatfield derailment and the subsequent Interim Review ( ACR 2003) resulted in an unquantified step change in the cost of infrastructure provision.

There are broadly two reactions to this.

Acceptance

Network Rail takes the view that its income it requirements reflect what the today's railway costs. Incremental savings can be made, but the 31% reduction required by ORR over CP3 represents the major part of post Hatfield economies.

Challenge

We know that during the last economic cycle, passenger ridership growth was similar to that achieved since 1994 (table) . At the peak of that cycle (1990) the railway needed roundly £1 billion in subsidy at 2006-07 prices. In 2006-07, excluding the channel Tunnel, support for the railway in England , Scotland and Wales is roundly £5.5 billion.

 

 

This increase in support is justified on the grounds that the railway was being run down in the 1980s (‘decades of under investment'). In contrast there has been sustained investment and growth for the past decade. However, comparison of physical projects then and now (as opposed to expenditure) suggests that investment has not increased by a similar multiple to support.

I emphasis ‘physical projects' because unit costs of projects have increased and definitions of ‘investment' changed. While the precise multiplier is debatable, there is widespread acknowledgement that the Ford Factor (see Note 1 below) exists.

Thus, I would argue that the railway is not improving and delivering more in proportion to the increase in public support. Compared with the 1980s we are not getting value for money. Which poses the question ‘what should the railway cost' in CP4 and beyond?

Graham Dalton, DfT Rail Group Director Projects, speaking at an engineering conference in June 2005, attempted to answer this question. He estimated that total support for the railway needed to come down to £3-3.5bn to satisfy the Treasury.. While his statement was subsequently disowned, it is unlikely to have been an off-the-cuff figure.

As a part of my on-going analysis of railway costs, I took Mr Dalton's figure and assumed that it would be achieved in the final year of CP4 (2013-14). A further assumption was that £1 billion a year of savings would come from Network Rail and £500 million a year from the train operating companies.

The result of this analysis is shown in Figure 2. You will note that if the rising level of income in the Rail Regulator's settlement for CP2 is extrapolated forward (the red line), by the end of CP4 Network Rail's required income is back to where it would have been had Hatfield not happened and assuming that the Regulator's annual increases in the settlement for CP2 reflected the effect of rising traffic levels

 

Figure 2

 

 

This is, of course, a very simple piece of a analysis, but I hope that Figure 2 may provide food for thought. How much hardware and software is obtained from investment within the implied level of support is a different matter.

 

Roger Ford

February 2007

 

NOTES

 

•  Ford Factor – the ratio of current costs per mile or per unit of infrastructure equipment compared to long run historic costs. When last calculated the value was 3. This chart shows the changing cost per mile for West Coast Main line electrification schemes, all at modern prices. The cost per mile for each stage is shown in red.. Note that the 2004 cost is almost exactly three times the 1997 budget which is in line with long term historic costs .