Assessing the Marginal Infrastructure Wear and Tear Costs for Britain?s Railway Network
P Wheat and A Smith, ITS, University of Leeds, UK
Estimates of the marginal wear and tear cost for the UK network are estimated based on an econometric analysis of regional cross section data from Network Rail.
Given the vertical separation of infrastructure and operations in Britain?s railway industry, it is necessary to price for the use of the infrastructure by train operators be they passenger franchise, freight or passenger open access operators. An elementary result in microeconomics is that pricing at marginal cost will lead to an efficient outcome (in a first best world). In addition, several European Commission Directives (e.g. 2001/14/EC) have been issued which require railway infrastructure to be priced based on marginal (social) cost. As such determination of the constituent parts of marginal social cost of railway infrastructure use is both necessary to price access to the railways in an economically efficient way and to comply with the latest European Commission Directives.
In addition, following the Hatfield accident in October 2000, Britain?s railway network has experienced a cost explosion. A significant part of this explosion was in the maintenance of track and other infrastructure with maintenance expenditure of Railtrack, the then infrastructure manager, 71% greater in 2003/04 than in 1999/00 (or 41% greater compared to 1995/96, the first full year post privatisation). Although the Office of the Rail Regulator set Network Rail efficiency saving targets in the 2003 Access Charges Review, in preparation for the 2008 Periodic Review, analysis of best practice between railways across Europe is of substantial importance in order to continue to bring costs down to efficient levels.
This paper contributes to both areas of railway policy. Estimates of the marginal wear and tear cost for the UK network are estimated based on an econometric analysis of data from Network Rail, the rail infrastructure manager of the UK. We will then benchmark our findings against results from similar studies throughout Europe.
The marginal cost analysis builds up on similar analysis in several other European countries, including Sweden (Johansson and Nilsson, 2004 JTEP), France (Gaudry and Quinet, 2003), Austria (Munduch et al, 2002) and Finland (Tervonen and Idström, 2004). In our data set, the railway network of Britain is divided into 53 Maintenance Delivery Units each containing a mixture of track and traffic types. A log linear variable cost function has been estimated which relates maintenance cost to traffic density, track miles, wage rates and fixed network quality and physical characteristics e.g. maximum line speed. This work is nearing completion and will allow us to derive marginal wear and tear costs of running additional trains/tonnage-km differentiated by existing traffic density and the quality and capability of the track.
We benchmark our results against those from other studies undertaken in Europe. Studies report large differences in marginal cost estimates. These differences are partly due to the impact of exogenous factors such as wage rates and differences in topography and so it is more appropriate to benchmark on usage cost elasticities, which control for these effects. We will thus present analysis of benchmarking on both measures.
Estimates of the marginal wear and tear cost for the UK network are estimated based on an econometric analysis of regional cross section data from Network Rail. We will then benchmark our results with those from studies undertaken throughout Europe both in terms of the elasticity estimates and the final marginal cost estimates.
Association for European Transport