Impacts of Unbundling on Competitiveness of Railways

Článek publikovaný v časopise Železničná doprava a logistika 2/2010.


The aim of ongoing reforms of the railway transport in the EU countries is to enhance operational efficiency of railway transport and its competitiveness in comparison with other transport modes; particularly with road transport, in the case of the high-speed rail even with air transport. In order to achieve this aim, the EU Common Transport Policy (CTP) has chosen the instrument of vertical separation, i.e. the separation of railway operation from the railway network ownership and management – so-called unbundling (in more detail see Stevens 2004 or Seidenglanz 2006). However, ownership, administrative, and financial separation of infrastructure from train operations poses a risk of new problems. Where unbundling is implemented under the conditions of inappropriate institutional structure and unclearly defined relationships between individual subjects of railway transport, it can further decrease efficiency of individual train operators, which results in a lower customer satisfaction as well as lower intermodal competitiveness of the railway – hence it can lead to a growth in overall costs of the railway industry. Technically, unbundling means that sunk costs and a considerable part of fixed costs are separated from the budgets of individual train operators and as a result, the increase in overall cost of the industry burdens the public budgets, from which the infrastructure costs or contingent subsidies will be funded. This text aims at identifying the problems which may be caused by unbundling as based on the knowledge of both economic theory as well as practical experience with railway operations. Factors, which may potentially threaten resulting effects of the reform based on unbundling, can be divided into three areas: (i) existence of economies of scale and economies of scope, (ii) existence of economies of traffic density and the effect of intermodal competition and (iii) threats to allocation efficiency by separating decision-making on the infrastructure from the train operating sector.

1. Unbundling from the point of view of the neoclassical economics – economies of scale and economies of scope

Economies of scale are based on the relation between inputs and total volume of operational output performance. Analyses of railway transport based on the Cobb-Dougles production and cost functions (Caves – Christensen – Swanson 1980, Keeler 1974) proved that infrastructure (i.e. land-use and tracks) is a source of a large majority of fixed costs; similar analyses comparing particular transport modes (Gagné 1990, Wetzel – Growitsch 2006, Winston 1985, Xu – Windle – Grimm – Corsi 1994, Ying 1992) established that it is the railway transport, where significant economies of scale are created. However, if infrastructure costs are not taken into account, railway transport operations show economies of scale, but not much higher in comparison with other transport modes. For a vertically integrated railway company the inputs include also the direct costs related to infrastructure (as a substantial component of fixed costs); in the case of unbundling, charges for utilization of the transport route (a part of variable costs) are included in the expenditure of the railway company. Therefore, infrastructure utilization is an important factor of savings for individual operators only if they own the infrastructure or if they pay flat rate utilization charges. If the train operator does not own the infrastructure and pays for its access per unit, the full utilization of infrastructure is not a part of the operator’s economies of scope.

A problem arises if there is natural monopoly of a single railway operator in the market, which means that with the existing or potential traffic flows it is economically more effective if a single operator provides transport services. Thus, competition for capacity of traffic route is ruled out and can be initiated artificially only by means of fixing the access price administratively. If train operators do not pay real costs for using the infrastructure, i.e. if payment for using railway lines do not reflect real costs incurred by traffic, another problem of unbundling emerges: a growing intensity of infrastructure utilization leads to losses – ”diseconomies of scale“, which are borne by the owner and regulator of the transport network.
The existence of another effect – economies of scope – in railway industry has been empirically verified e.g. by H. Wetzel and C. Growitsch (2006). Unlike economies of scale, which are based on decreasing average costs along with increasing the volume of existing production, economies of scope are related to decreasing average costs along with expanding the production with new products. A typical example for railway transport is parallel provision of passenger and freight services by a single operator (Pietrantonio – Pelkmans 2004, Quinet – Vickerman 2004). Economies of scope have their spatial importance: operators can boost efficiency of their service by changing or increasing the number of operated destinations; in this context “economies of spatial scope“ are defined (Jara-Díaz – Cortés – Ponce 2001).

Unbundling really means the possibility of a change in “structure“ – but only as long as the structure is suitable and the operator has a possibility to modify it. A problem occurs if the change in infrastructure utilization is costly, blocked administratively or unfeasible in the short term, or if utilization charges of infrastructure are discriminating, or service of various destinations is tied to various subsidies and obligations.

Empirical studies of railways in North America and Japan (Pietrantonio – Pelkmans 2004) as well as in Europe (Wetzel – Growitsch 2006) have arrived at the conclusion that separating network ownership from operations does not lead to higher efficiency; on the contrary negative effects prevail. Obviously, advantages of separating infrastructure from operations outweigh the economies of scope in the event of high density traffic, where advantages resulting from unbundling can be fully manifested: higher operational efficiency due to competition among individual operators and simultaneously due to economies of scale (Pietrantonio – Pelkmans 2004). This concept has been used in the reform according to CTP: efficiency of individual operators can be enhanced if they extend their service to other destinations throughout the EU. This is incorporated in the concept of service liberalization on the EU single market. The question is, however, what impacts this effect will have within various regions or national markets.

In order that railway operators may attain economies of spatial scope, the railway network must be adaptable to their business plans. In vertical integration, infrastructure is a part of fixed costs of the operator who bears additional costs incurred by its modification; the operator himself decides on the change of destinations, network and investment on the basis of subjective economic calculations. In unbundling the operator does not pay the additional costs of network modification (or at least not in full), but the possibility of changing destination exists only within administrative restrictions imposed by the system of train path allocation by the network regulator. It can be easy and economically effective to change operated destinations, if the system of regulation is flexible, the network has a sufficient capacity, covers all destinations which can be taken into account for transport operation and if the price for using the route to new destinations is non-discriminatory. While the road networks meet these characteristics without bigger problems, the railway network is much further from this ideal.

2. Unbundling from the point of view of practice – economies of transport density and intermodal competition

Economies of density are based on the relation between inputs and outputs with a constant size of the network, which entails that unlike in the case of economies of scope, economies of density are generated only due to intensive utilization of traffic routes (Keeler 1974). The possibility to benefit from economies of transport density is considered a dominant operation efficiency factor (Jara-Díaz – Cortés – Ponce 2001) and is one of the principal factors which should be considered when deciding on unbundling (Stelling – Jensen 2005). The higher the volume of transport an operator concentrates on a traffic route, the higher economies of density are achieved – this is obviously valid only under the conditions of vertical integration. In vertical separation it is the owner and infrastructure regulator who receive revenues from high density.

At the same time the transport density is the factor which gives railways a competitive advantage over other transport modes (Fisher – Bitzan – Tolliver 2001, Pietrantonio – Pelkmans 2004). However, it requires economically and technically optimum allocation of investment in infrastructure, which can be interfered with by various factors. The danger of a network owner’s failure can be twofold: (i) over-designing of the network in certain segments, i.e. investment into tracks in the directions where the limit traffic density, which could ensure return on investment, is not and will never be reached. This threat is relevant primarily to the European continent in connection with constructing corridors and high-speed lines (one example for all – the Channel Tunnel). Attempts of the regulator at price adjustment will not be effective – a high price will discourage operators from using the line while a low price does not ensure return on investment. (ii) Equally dangerous is under-designing of a certain segment, where its insufficient capacity does not enable to exploit the potential of economies of traffic density in the neighbouring network. In this way a bottleneck will be created, which will burden the traffic efficiency by congestions and will decrease the ability of the railway to compete with other transport modes. It is also evident that in such a case a considerable loss of economies of scope is incurred in individual operators (Nash – Preston 1993, Preston 2001), although there are practically no empirical studies that would exactly calculate the loss (Pittman 2005).

Mutual competition of particular transport modes is marginal in passenger service; the principal division line does not separate the modes but the mass passenger transport (of any mode) and the individual transport (car transport). In freight service, absolutely critical is the type of transported freight: the railway operator is in a completely different position when transporting raw materials, substrates, etc. than in the transport of consumer goods. On the other hand, railway cannot have a competitive advantage over road transport when transporting low-weight heterogeneous products with a high added value between different destinations over short and medium distances, where the emphasis is on speed and accuracy of delivery (as is the case in most European countries). Where freight transport in based on a regular transportation of bulk of a homogeneous product (typical of raw materials) or transports freight over a long distance between the same destinations (mines, ports, industrial sites), competitive advantage of the railways is highly probable (e.g.. transcontinental railways in Asia or North America). This type of transport and its high density enable the development of intra-modal competition, even under the conditions of vertical integration, i.e. a connection of individual destinations by two or more railway companies owning transport routes (typical of the USA, Canada or some Latin American countries). From the point of view of economies of traffic density the railway can efficiently operate in certain segments of transport network regardless of its vertical integration or separation. The reverse is true about the segments where the railway transport is not efficient due to a low density where unbundling cannot trigger any turnaround.

3. Unbundling from the point of view of management – separation of decision-making on infrastructure from operators

As mentioned in the first part, unbundling leads to separation of a substantial part of fixed costs related to the infrastructure from operators – the subjects, whose rational decision-making on costs of infrastructure should be critical for competitiveness of the industry. Decision-making on optimum allocation of resources to innovations and development investment is more efficient if it is made simultaneously and in mutual interaction between operations and infrastructure within one integrated company. It has been deemed for theoretically proven in economics at least since publishing the book by Ronald Coase in 1937.

Technically speaking, the railway transport considerably differs from other network industries such as electricity industry and other transport modes. It is not predictable that a transfer of identical institutions and mechanisms from these industries to railway undertaking will yield the same results. Economic theory itself and specifically regulatory economics cannot properly explain and estimate potential impacts of unbundling in railway industry, without taking technical specifications of railways into consideration (Ehrmann 2003).

It may seem correct and logical to separate railway’s ownership and operations of transport routes from providing transportation services, i.e. from train operations on the lines. Technically it means to split the rail track and wheels. The problem hidden behind it is technical but its consequences are economic. Optimization of maintenance costs and meeting technical standards of wheels and rails are completely different in a vertically integrated company and in vertically separated operations and infrastructure (Pittman 2005). Then it is logical and justifiable that operators try to shift a part of costs to the infrastructure manager, while a part of costs is displaced outside the system of transport in the form of negative externalities. Changes in destinations and changes in transport plans of individual operators in unbundling can be burdened by additional costs incurred by enforced optimization of connecting of rail superstructure and travel of trains (Ehrmann 2003).

A far more important problem is associated with traffic route technology. Most railway tracks in Europe were built in the second half of the 19th century and so they were often designed with the main emphasis on national, regional, or even only local short-term economic interests and after some time the interests became more often political and military-strategic (Heymann 2006, Kvizda 2006, Kvizda – Pospíšil – Seidenglanz – Tomeš 2007). Today, large parts of this network have neither economic nor social raison d’être (such as in former industrial agglomerations related to extraction of raw materials, mainly coal). In contrast, bottlenecks are created in some sites with absolutely insufficient capacity (big metropolitan regions, backbone traffic flows, junctions, and the like) where various types of trains arrive (fast long-distance passenger trains, suburban commuter trains, long-distance freight expresses, special-purpose freight trains, etc).

Unbundling in railway transport could result in deformation of investment plans of both the owner and infrastructure operator and individual train operators, who invest in a certain type and quantity of rolling stock on the basis of condition and development of infrastructure. There is a real danger that in vertically separated industry decision-making on investments will not be coordinated and if so, then only by means of administrative tools at the level of government bodies – with all drawbacks of central planning. It could give rise to conflicts between development and investment on individual tracks and their future utilization (in the short and medium term) and when setting charges for infrastructure utilization on particular tracks with regard to their parameters and demanded utilization (in the long term).

The danger of inefficient or totally misplaced investment allocations to the development of railway tracks is further intensified in Europe by the Common Transport Policy in the EU linked with anticipated subsidizing of passenger transport. In such a system, the link between supply of the railway route and demand of rail operators for it is completely lost because supply is not based on solid foundations of market relations. Any administrative deviation in the complex system of regulated and subsidized transport market can cause marked changes in the behaviour and decision-making of rail carriers. At the same time the owner and the regulator will have only a small and misleading feedback about the demand, because it will be only derived demand – real demand is that for transport services from customers of transport companies.

We have repeatedly come across with the issue of revenues without pointing to one of the biggest problems of unbundling: pricing the access of operators to train paths. It is difficult if it is not a case of setting a price on the basis of market supply and demand – which can never happen in vertical separation – and it always poses a risk of system failure. It is nearly impossible to determine an adequate price (equilibrium price in the economic sense) for using the infrastructure in the environment where this price is derived from investing in infrastructure, while the investment does not reflect real demand by railway operators. And vice versa, when deciding about investment, price as a decisive informative indicator in unbundling fails and cannot be used for decisions about adequate maintenance costs and investment in different parts of infrastructure (Pittman 2005).


From the state antimonopoly policy point of view, unbundling is one of the potential tools for creating a competitive environment within an industry with a natural monopoly based on the existence of network. An application of a theoretical model to practice has its pitfalls. A number of case studies of unbundling reach an ambiguous conclusion: a formal separation of infrastructure from producers yields good results in certain industries, certain countries and certain periods of time, while this need not be true about other locations and other periods of time (Ehrmann 2003, Hirschhausen 2002, Newbery 1999, Pittman 2003, Pittman 2005, Vickers 1995). Unfortunately, the quantitative analysis of profit and loss from unbundling on the railway is complicated by unavailability of data or by their inadequate structure – it mainly concerns the railway systems, which had long been monopolized, as in most countries of continental Europe. Empirical studies, which tried to make the quantification of a sample including North American or British railways, where the data are available, estimated 20 – 40% loss of technical efficiency in case of vertical separation (Ivaldi – McCulough 2004).

If we summarize the results of unbundling analysis, we find that the state is in a singular position in its role as a central regulator and planner. The railway reform based on unbundling means that the state has to designate a strong, independent owner and infrastructure regulator – if it is not the state itself, which is so typical for European conditions. British experiences with the private owner and infrastructure regulator (Railtrack) are forbidding. The state probably will always be the owner, manager and regulator of a vertically separated transport network. The state will decide about directions as well as about technical and economic parameters of the network. It is the state that will determine major costs of train operators through the regulated price for access to infrastructure. Since this describes the situation in Europe which has existed for decades, unbundling means only a small progress. Therefore it cannot be overestimated and the method conceived in this way cannot be associated with high expectations of enhancing competitiveness of the railway transport. In present-day world economy these limits that are hard to overcome.


[1] BANNOCK, G. – BAXTER, R. E. – DAVIS, E. (2003) The Penguin Dictionary of Economics. London: Penguin
[2] CAVES, D. W. – CHRISTENSEN, L. R. – SWANSON, J. A. (1980) Productivity in U.S. Railroads, 1951-1974. Bell Journal of Economics, N° 11/1, s. 166-181
[3] COASE, R. (1937) The Nature of the Firm. Economica 1937/4, s. 386-405
[4] EHRMANN, T. (2003) What can regulatory economics learn from franchise systems? Some additional arguments for vertical integration of railway companies. paper on The
First Conference on Railroad Industry Structure, Competition, and Investment, Toulouse
[5] FISCHER, P. A. – BITZAN, J. – TOLLIVER, D. (2001) Analysis of Economies of Size and Density for Short Line Railroads. Fargo: North Dakota State University
[6] GAGNÉ, R. (1990) On the Relevant Elasticity Estimates for Cost Structure Analysis of the Trucking Industry. The Review of Economics Statistics, Vol. 72, s. 160-164
[7] HEYMANN, E. (2006) Competition in European railway market. Morning has broken. Reports on European Integration – EU Monitor 39, Deutsche Bank Research
[8] HIRSCHHAUSEN, C. VON (2002) Modernizing Infrastructure in Transformation Economies: Paving the way to European Enlargement. Cheltenham: Edward Elgar
[9] IVALDI, M. – MCCULOUGH, G. J. (2004) Subadditivity Tests for Network Separation with an Application to U.S. Railroads. Toulouse: CICT,
[10] JARA-DÍAZ, S. R. – CORTÉS, C. – PONCE, F. (2001) Number of Points Served and Economies of Spatial Scope in Transport Cost Functions. Journal of Transport Economics and Policy, Vol. 35, N° 2, s. 327-342
[11] KEELER, T. E. (1974) Railroads Costs, Returns to Scale, and Excess Capacity. Review of Economics and Statistics, N° LVI, s. 201-208
[12] KVIZDA, M. (2006) Ekonomické dějiny železniční sítě České republiky – mýty, omyly a iluze v hospodářské politice a path dependence železných drah. Brno: Masarykova univerzita
[13] KVIZDA, M. – POSPÍŠIL, T. – SEIDENGLANZ, D. – TOMEŠ, Z. (2007) Železniční doprava – institucionální postavení, hospodářská politika a ekonomická teorie. Brno: Masarykova univerzita
[14] KVIZDA, M. (2007) Vertikální integrace versus separace v železniční dopravě – cui bono? In: Rozvoj systémů osobní transport z hlediska respektování požadavků uživatele. Pardubice: Univerzita Pardubice, s. 103-109
[15] NASH, C. A. – PRESTON, J. (1993) The Policy debate in Great Britain. In European Conference of Ministers of Transport: Privatization of railways. Paris: ECMT, s. 85-119
[16] NEWBERY, D. M. (1999) Privatization, Restructuring, and Regulation of Network Utilities. Cambridge: MIT Press
[17] PIETRANTONIO, L. DI – PELKMANS, J. (2004) The Economics of EU Railway Reform. Bruges European Economic Policy Briefings, Bruges: College of Europe
[18] PITTMAN, R. W. (2003) Vertical restructuring (or Not) of the Infrastructure sectors of Transition Economies. Journal of Industry Competition and Trade 2003/3, s. 5-26
[19] PITTMAN, R. (2005) Structural Separation to Create Competition? The Case of Freight Railways. Review of Network Economics, vol. 4, N° 3, s. 181-196
[20] PRESTON, J. (2001) An Overview of Public Transport Reforms in Great Britain and Forecasts for the Future. International Journal of Transport Economics, vol. 26, N° 1, s. 23-42
[21] QUINET, E. – VICKERMAN, R. (2004) Principles of Transport Economics. Cheltenham: Edward Elgar Publishing
[22] SEIDENGLANZ, D. (2006) Železnice v Evropě a evropská dopravní politika. Brno: Masarykova univerzita
[23] STELLING, P. – JENSEN, A. (2005) Train Operator’s Economies of Scale and Business Strategies. Third Conference on Railroad Industry Structure, Competition and Investment: Stockholm School of Economics,…
[24] STEVENS, H. (2004) Transport Policy in the European Union. New York: Palgrave Mcmillan
[25] VICKERS, J. (1995) Competition and Regulation in Vertically Related Markets. Review of Economic Studies, vol. 62, N° 1, s. 1-17
[26] WETZEL, H. – GROWITSCH, C. (2006) Economies of Scope in European Railways: An Efficiency Analysis. IWH-Discussion Paper 5, Halle: Intitut für Wirtschaftsforschung Halle WILLIAMSON, O. E. (1975) Markets and Hierarchies: Analysis and Antitrust Implications. New York: Free Press
[27] WINSTON, C. (1985) Conceptual Developments in the Economics of Transportation: an Interpretative Survay. Jouranl of Economic Literature N° 23/1985, s. 57-94
[28] WORLD BANK (2002) World Development Report 2002: Building Institutions for Markets. Washington D.C.: World Bank
[29] XU, K. – WINDLE, R. – GRIMM, C. – CORSI, T. (1994) Re-evaluating Returns to Scale in Transport. Journal of Transport Economics and Policy, Vol. 28, s. 275-286
[30] YING, J. (1992) On Calculating Cost Elasticities. The Logistics and Transportation Review, Vol. 28, s. 231-235