RIEB Seminar

10:30am-12:00pm

Topics

Assessing the Efficiency of Railway Maintenance with Heterogeneous Maintenance Units: The Case of France

Speaker

Andrew SMITH (Institute for Transport Studies, University of Leeds)

Abstruct

Since the unbundling of Europe's railway, there has been a growing interest in the literature on studying the upstream infrastructure manager, as a regulated natural monopoly within a wider rail system. The first objective of this paper is to investigate the possibilities of internal benchmarking to derive efficiency measures of railway maintenance in a context of heterogeneous units and see if there is scope to reduce network costs. To do so, we use an original dataset consisting of the 29 maintenance decision units in France over four years (2013-2016) with a rich set of variables to characterise the cost, the outputs and the network's characteristics and condition. Our dataset is important for two reasons: firstly, previous studies of rail efficiency (and indeed in regulation of other network industries) have been beset by the criticism that heterogeneity between decision making units (DMUs) is not adequately controlled for prior to estimating efficiency; secondly, previous applications have typically seen relatively small sample sizes. To measure efficiency we apply appropriate econometric techniques to estimate a cost function; and derive efficiency estimates using a range of methods, including alternative variants within the family of Stochastic Frontier Analysis (SFA).
The second objective is to increase our understanding of railway maintenance economies of scale and density, taking account of heterogeneity of service types and network complexity. There are a number of dedicated maintenance units reflecting the particularities of the French network. France indeed has the third largest high speed (HS) network in the world and each line has a designated maintenance unit for a total of 4 high speed maintenance units. Moreover, the Paris region which accounts for 10% of the network in length but 74% of the passenger journeys was split into six maintenance centres. Therefore, an important dimension of the research is to study if the cost structure of these different line types (economies of scale and density) differs, and how the results compare to the wider literature on integrated railways.