ECONOMETRIC BENCHMARKING OF METRO OPERATING COSTS. METHODS AND APPLICATIONS.



ECONOMETRIC BENCHMARKING OF METRO OPERATING COSTS. METHODS AND APPLICATIONS.

Authors

Richard Anderson, RTSC - Imperial College London, Ruben Brage-Ardao, RTSC - Imperial College London, Daniel J. Graham, RTSC - Imperial College London

Description

This paper shows how econometric benchmarking can explain the variation in operating costs across metros. It also identifies the main cost drivers for each of the main constituents of operating costs and how this can be useful for target setting.

Abstract

There is a wide variation in the operating costs of metro operators worldwide. The operating costs per car kilometre or per passenger journey diverge not only in absolute terms but also in the breakdown of the total operating costs, with significant differences between metros in the share of service costs or maintenance costs over the total operating costs.
In order to understand this variation in operating costs, the RTSC has undertaken an econometric research based on the metro database for CoMET and Nova. CoMET and Nova are two metro consortia comprising more than 30 metros with more than 20 years of experience in KPI benchmarking and performance research. Therefore, using this unique database, the RTSC has developed econometric models to explain the main cost drivers of operating costs constituents. These constituents include train service costs, station service costs and maintenance costs for rolling stock, stations and infrastructure.
This econometric methodology quantifies the main cost drivers for each operating costs constituent, identifying potential sources of inefficiencies or savings and assessing the existence of economies of scale. However, the main use of the models is the validation of plans, particularly, to determine whether efficiency or saving targets are sufficiently stretching. The models permit a comparison between the actual costs and the expected costs given the conditions of the metro. Therefore, it is possible to compare the cost efficiency of metros taking into their own peculiar circumstances (e.g. older fleet or higher salaries). Particularly, the analysis revealed important factors that affect metro costs within and outside operator control despite conventional benchmarking can mask these factors. For example, rolling stock maintenance costs were observed to increase less than proportionally to the increase in car kilometre production and there were also found significant economies of scale in the fleet size. Contracting out of station services was observed to reduce costs by around 1% for a 10% increase in outsourced labour. The effects of different unit prices in labour and energy were quantified, substantially improving our understanding of the factors affecting metro operating costs in international benchmarking.

Publisher

Association for European Transport