Looking Beyond Separation - a Comparative Analysis of British, German and Swedish Railways from a New Institutional Perspective

Looking Beyond Separation - a Comparative Analysis of British, German and Swedish Railways from a New Institutional Perspective


R Merkert, C Nash, A Smith, ITS, University of Leeds, UK


This paper discusses the effects of different institutional and contractual arrangements on the interactions between train operators and infrastructure manager in the three most liberalised rail systems in Europe.


Although the EC imposes many regulations on the rail systems of its member states it leaves some space for different approaches of rail organisation, which include much more than different degrees of vertical separation. This paper aims to reveal key differences and similarities between the institutional and contractual set ups in the British, German and Swedish rail systems and the key interactions at the train operation/infrastructure interface. For a snapshot of the year 2007 our intention is to show what effects these differences have on the attributes (e.g. uncertainty) of key interactions and what really happens in practice at the interface. Finally we aim to answer which approach of rail organisation produces the best alignment with transaction attributes without losing the benefit of facilitating competition. This paper is not only highly relevant for EU policy but also aims to contribute to the international debate on rail organisation.

For this cross-national comparative study, more than 60 in-depth semi-structured interviews with train operating companies, their parent companies, infrastructure managers, regulators and industry associations have been conducted. Key policy documents and contracts have also been reviewed. From a methodological perspective, this paper applies Williamsons transaction cost economics to the train operation/infrastructure interface in seven pre-specified transaction/interaction areas (e.g. timetabling/slot allocation). The interviews focus on how the interaction in each area works, including its frequency and intensity, as well as the uncertainty and complexity that are associated with it. We also address disputes as well as formal and informal dispute resolution procedures. Unlike earlier studies regarding transaction cost economics in railways this paper is not only interested in investment decisions in respect of rolling stock but is also trying to reveal the whole range of investment and coordination aspects associated with interactions between train operators and infrastructure managers. Although the focus is on track infrastructure, other assets such as stations, depots, rolling stock and human capital are also included in the analysis.

Our results show many commonalities (e.g. safety standards or timetabling issues) between the systems mainly because of the implementation of EU directives and the nature of the railways. Coordination and contractual aspects are seen as more crucial (the major source of disputes is around track possession for engineering work) than the traditional investment hold up or lock in issues and the importance of good relationships and trust is highlighted in all systems.

Nevertheless the three systems are still very different, not only because of the increased heterogeneity and complexity of today?s EU railways. Institutionally the key difference is that in contrast to Sweden and Britain, most German rail transport and infrastructure activities are governed by the DB holding structure. This has advantages for the ?DB family? (e.g. joint functions and IT-systems, escalation of disputes) but creates uncertainty for other operators on the German network and disputes that often involve courts. In Sweden courts get rarely involved and the UK?s industry dispute committees prevent court involvement. Contractually, the key contracts are track access agreements, but they are very different in each country (duration, content, format, framework contracts etc.) with consequences on their application and frequency of their amendments. One of the key differences is the inclusion of performance regimes into these contracts. Typical for the very contractual British system is also the leasing and sub-leasing of stations/depots which creates many additional contractual relationships at the infrastructure train operation interface. For many areas there are, besides national and regional distinctions, differences with regard to the type of train operation. The paper discusses all differences and their effects on the interaction attributes in detail.

The results indicate, in conclusion, that the fully separated Swedish model aligns best to the transaction attributes although the relevant contracts within the Swedish system are still not entirely aligned to each other. The German model has clear advantages for the ?DB family?, and the British one is the most advanced one in terms of economics and completeness of contracts. The model of vertical separation is seen as workable and best to provide non-discriminatory access to the infrastructure. For the effectiveness of implementation of EC policy it is, however, important to address also the effects of competition and contractual arrangements for each type of operation.


Association for European Transport