Discussion Papers

Gernot Müller

Measuring productivity and efficiency in the rail infrastructure sector - Methodology and assessment of productivity change for the German market

No. 318  January 2009


On 26 May 2008, the German Federal Network Agency presented its final report on the introduction of an incentive regulation in the rail sector. A central element of the pricecap formula is the so called X factor, portraying common productivity change and potentials for efficiency improvement. In this context, it is intended to distinguish some categories of companies and service baskets. This necessitates the deployment of applicable concepts for the calculation of general and individual X factors for distinct rail infrastructure managers and market segments. For that purpose, proper variables and data as well as suitable methodologies have to be considered definitely. In addition, empirical studies are to be interpreted, regulatory experience is to be analyzed, and computations of a general X factor for the German rail market are to be executed.

Concerning the determination of productivity change, the use of index numbers and aggregated sector data is recommended. At first stage, it is advised to apply the Tornqvist productivity index to deduce total factor productivity. In the long run, the Malmqvist data envelopment analysis and econometric estimations of production and cost functions can also be an alternative. To measure relative efficiency levels a parametric data envelopment analysis and/or a non-parametric stochastic frontier analysis with translog functions are preferred. The choice between these methods must be based on an assessment of several criteria. Input or cost related approaches are favoured compared to
output, revenue or profit oriented techniques.

These proposals for a qualified methodology are corroborated by the results of previous empirical studies on productivity change and efficiency catch-up in the rail sector. Unfortunately, little relevant statements on the rail infrastructure market are possible. In the face of the poor availability of data, internal comparisons, external benchmarking with other economic sectors and other countries, as well as bottom-up engineering and cost models should be implemented, as it has been shown by the British ORR with regard to
Railtrack and Network Rail.

Most of the former empirical studies for the rail transport sector deduced an annual productivity growth rate of 0 to 3 per cent on average. Using data of the national accounting and of business reports of Deutsche Bahn AG, calculations of this study for the German rail sector lead to results of 0.08 per cent (general X factor of -0.16 per cent), and 3.04 per cent (general X factor of 2.80 per cent) respectively. Due to the assumptions made and certain limitations of data disposability and reliability, the outcomes must be interpreted with caution. Putting aside external benchmarking, estimations on the efficiency level and the company related potential for efficiency improvement in the German rail sector are currently impossible owing to lacking data.

[Full version only in German language available]

Discussion Paper is available for download.

To top  |  Print