The paper was published in „Horizons of Railway Transport“ (No. 2, Vol. 2, 2011) issued the University of Žilina, Slovakia.
The paper applies the theory of defining relevant market to the railway industry. The first part identifies specific features of the market for rail transport services, the second part deals with issues of the rail industry concentration, the third part analyses the method of defining relevant market for railways respecting the specific features of the market and particularly in relation to application of the SSNIP test and finally the text summarizes the risks of the test’s failure to define the relevant market. A particular attention is focused on geographical and temporal dimensions in defining relevant market with respect to specific features of rail transport services.
At present the railway transport in EU countries is going through far-reaching reforms. Their purpose is to liberalize the market in rail transport services, eliminate national transport monopolies and open up a free access of independent operators to the transport infrastructure. Although the formal principles of undistorted competition were laid down by the Treaty Establishing the European Community (Articles 81 and 82) and since the 1960s the practical policy of economic competition was carried out with great intensity by the European Commission in collaboration with national antimonopoly institutions, the railway transport in Europe practically survived in the form of state monopolies until the end of 1990s. The reasons why for decades efficient competition on the rails did not exist in any European state are certain technological and historical specific features of this transport mode (in more detail see ) as well as particular attitudes of individual states towards railway carriers (in more detail see ). Therefore the present reforms are based on legal rules issued within the framework of the EU Common Transport Policy (mainly the directives of the European Commission, Council and Parliament 2001/12/ES, 2001/13/ES, 2001/14/ES, 2004/49/ES, 2004/50/ES, 2004/51/ES, 2007/58/ES, 2007/59/ES and other regulations included in the so-called First, Second and Third Railway Package; in more detail see ).
Thus a real, even if very specific, market in rail transport services is emerging in Europe – it is no longer separated by national borders but becomes a part of a single European market. Its basic structure is formed by the rules included in the railway packages and so it forms a specific regulatory framework implemented for the whole industry ex ante. Thus the EU member states are required to establish national bodies and agencies governing capacity allocation and responsible for issuing licences, regulatory authorities and internal safety and investigative bodies . It is a case of completed genuine unbundling. The unbundled and regulated market gradually enables efficient competition among new coming independent operators. No matter what regulatory bodies were established pursuant to the above directives or how much power they were, the economic competition among all engaged subjects on the railway transport market is regulated by general legislative rules of economic competition policy, whether at the European level (said articles 81 and 82) or national levels; in the Czech Republic it is regulated mainly by the Act No. 143/2001 Coll. on the Protection of Economic Competition. One of the principal functions of the competition policy is to open up a free competitive space for the new coming operators on the market and supervise, regulate and sanction the conduct of the operators in this space. Thus it is necessary to accurately identify and define market shares of individual rail operators for each intervention of the competition policy, which can be done only in the relevant market. This text aims to theoretically determine options for defining the relevant market in the sector of rail transport services and analyze practical usability of various options with respect to risks of inaccuracy and distortion of such definition.
2. Specific features of railways
Before discussing the railway transport sector, we meet with a principal problem associated with implemented policy of economic competition: the definition of effective competition. As Bender et al. point out , an ambiguous (or even non-existent) definition of ”effective competition“ makes it impossible to decide whether implemented regulatory measures are successful or not, and decide whether the industry should be regulated ex ante, or to be subjected to only ad hoc antimonopoly supervision as any other market. Formal EU documents work with concepts of „effective“ and “undistorted“ competition (see The Commission of European Communities in 1996 and other above documents concerning railway packages) and national legislation include these concepts as well. For instance, German law implements the EU directives so that the term “effective competition“ is interpreted as actual presence of several mutually competing operators in the market and the term ”undistorted competition“ as a free non-discriminatory access of all operators to the infrastructure and absence of cross-pricing in the incumbent . Enforcement of “undistorted competition“ is secured ex ante by regulation based on the railway packages (unbundling). The definition of “effective competition“, however, is too vague and in practice it is not sufficient for carrying out economic competition policy. Economic theory in no way suggests that effective competition within an industry can take place only if many competing subjects are permanently present in the market. The neoclassical microeconomics virtually assumes the victory of more efficient producers in the market over those less effective, which leads to uneven profit distribution, a growth of market shares in the leading producers and a higher industrial concentration. Kahn in his study  argues that in specific industries (it was the case of telecommunications industry) an undertaking with a very small market share can generate sufficiently strong competitive pressure on a dominant incumbent and so force through ”effective“ competition within the industry; it is particularly the case if the sector is based on existence of sunken costs or if there are barriers to entry within the sector. Moreover, network industries typically generate network economies and economies of scale (these effects are discussed in more detail in ), which characteristically lead to victory of one or more producers in the market, without necessarily decreasing competition intensity . This is also confirmed by the study of Irish service market of mobile operators : although an industry with two dominant producers is strongly concentrated, effective competition takes place in the market thanks to structural characteristics and specific features of the industry. Specific features of railway transport consist mainly in a relatively difficult access of potential competitors to the market: the difficulties are caused by both economic barriers (they incur additional costs to entrants) and non-economic ones (the barriers cannot be overcome by expenditure of costs). A significant factor, which makes railway industry exceptionally specific, is the existence of a network, i.e. specific transport paths and related technological equipment. From the point of view of competition, network industries can be divided into the network as such (”upstream market“) and market of services provided using the network (“downstream market“). Crucial for the decision within the framework of competition policy is the fact whether and under which circumstances there are alternative options for entry to the service market – downstream market . It is the very concept that is a basis for present reforms of railway transport in the EU: actual vertical division of the sector into the transport infrastructure and transport service market (unbundling) is the first step to enforcing modern regulation and introduction of effective competition within the industry. The unbundling alone, however, is not sufficient and must be followed by other measures of economic competition policy, i.e. regulation in unambiguous cases. It is obvious that specific features of railway transport sector give rise to many barriers to entry for potential competitors and to a certain degree create a protected environment for the old, long established nation-wide railway carriers – incumbents. So far, their monopolistic or dominant position has been characteristic and has posed permanent threats to freedom of enterprise within the railway transport sector – it means that it blocks intramodal competition. With regard to gradual liberalization of railway market in the EU countries due to the above reforms (see for instance ), interests of individual real or potential intramodal competitors begin to clash: mutual conflicts of interest appear between newly established operators and incumbents in old national markets or between incumbents on the European market.
The railway industry, however, is not characterized by the barriers to entry only. In Europe it is specific because of overt or hidden support from the part of the state: particularly non-transparent subsidizing of losses incurred by providing services in the public interest in passenger transport, in this connection cross pricing of commercial services, guarantees of loans, dumping price setting for using the infrastructure and the like (in more detail see ). All these things are also reflected in the intensity of intermodal competition, chiefly in passenger transport, when operators in other transport modes (mainly bus/coach carriers, in part also airlines) are exposed to unequal or at least nontransparent
conditions for competition. Naturally, the newly created competitive environment gives rise to more or less intense conflicts between different operators that must be solved by authorized state administration bodies within the scope of economic competition policy. The problem is that railway market has not been fully liberalized and so far it shows (and will probably continue to do so) a number of deformations and specific features – and yet anti-monopoly authorities have to solve these conflicts. Therefore the authorities need a very precisely defined market, on which they could base their regulatory enquiry and measures.
3. The relevant market
Anti-monopoly authorities can interfere, when a dominant subject abuses its position to injure other competing subjects or customers (consumers). In order to solve a case of abuse of economic position in the market it is necessary to find out the position of the investigated subject in the market and mainly to define this market. The dominant position is always linked with the market for a certain product, i.e. associated with the position of an undertaking within a given market segment. The market, where a given subject can be exposed to “effective“ competition, or it can restrict “free” competition, is called relevant market . Although modern economic approach to the competition policy relies on standardized econometric methods using direct cost analysis, these methods cannot be applied to all cases under investigation – above all they cannot be applied to certain specific industries. The reason is a lack or inaccessibility of appropriate data needed for such analysis  and a risk of distorting the results due to the existing specific features within an industry. In these cases it is appropriate to use older, well-established techniques that are based on the very precise definition of the relevant market for a particular product . The relevant market must be defined taking into account the product point of view, i.e. to determine the product market, and further define it spatially, which means to determine the geographic market; in defining the relevant market a temporal dimension applies in addition to product and geographic aspects. A precise definition of relevant market from all three points of view is vital for correct determination of market power in individual producers as well as for correct and adequate application of the competition policy . As
we are going to see next, the application of none of these dimensions is easy and clear-cut within the railway industry.
4. The product market
It is crucial to determine groups of products (goods or services) that make up one market, i.e. which are mutual substitutes. The key question, to which the competition policy seeks an answer, is: what group of products can satisfy a certain need of their consumers, whether a given product can be replaced by a substitute if its price is increased and if the consumers have a real possibility to switch to a substitute in case of shortage. Modern methods of identifying the product market rely on the approach based on identical consumer response [1, 19, 20]. These methods assess mutual substitutability according to the response of consumers to the change of price of the product. The consumers´ response is widely measured by the SSNIP test (Small but Significant Non-transitory Increase in Prices), which is based on quantifying a change in profit of a hypothetical monopolist as a response of consumers to a price rise 5-10% of a given product [19, 20]. If the demand for a particular product has a low elasticity and a price rise results in a relatively small outflow of consumers from the market and producer´s profit rises, then the given product constitutes a specific relevant market and does not have real substitutes. If the demand for a considered product is more elastic so that a price rise makes consumers to replace the product with its substitute, and the producer´s profit decreases, such relevant market cannot be defined for the given product, but it is necessary to consider other products (substitutes) and define the relevant market more broadly.
One of the main problems posed by defining the product market for the railway industry is non-homogeneity of transport services. The principal unsubstitutable service is transport of persons or goods in space; logically the product market can be easily defined separately for passenger and freight transport. Frequent but not automatic is a further classification of passenger transport into long-distance intercity transport and regional or suburban transport and classification of freight transport into bulk transport of substrates, carload transport and piece shipment. This classification is also quite broad and is not suitable for many examined cases ; moreover, it is often combined with the definition on the basis of the network, which means it does not consider the intermodal competition.
When defining the relevant market, it is also necessary to take into account the possibility of demand-side and supply-side substitution. Identification of potential demand-side substitution answers the question, whether there are other alternatives for a given product at a given moment, i.e. available existing substitutes. Identification of potential supply-side substitution answers the question whether it is viable for another producer to enter the market of a given product and create sufficient substitute supply – in response to a price rise by the incumbent. As for railways, the definition of the relevant market at certain destination can thus include the existing operators from the demand-side point of view, while the definition from the supply-side point of view can be enhanced by potential competitive supply of carriers operating at other destinations and chiefly in other transport modes. Such definition of the relevant market requires taking into account the geographical and temporal aspects.
5. Geographic market
Definition of the geographic market is inseparable from the product market, since it answers the question, in what area the market for a concerned product and its substitutes really exists. As in the case of the product market, the SSNIP test is currently applied. However, due to the nature of the thing, it is necessary to use descriptive methods as well. The common method is to measure exports and imports of a concerned product from/to a region and their share in the product total consumption in a particular region (shipment test); another definition takes into account the distance in which a producer can find feasible possibilities of selling with regard to transport costs (such as a catchment area of some raw material extraction), local preferences and habits of consumers (e.g. preference of certain foods, focus on fancy goods etc.) or regional technical parameters and national standards (such as use of particular types of transport modes and their homologation, composition and origin of foods etc).
In delineating the geographic market for transport services the definition according to individual destinations automatically emerges. Such a definition, however, is not self-explanatory and can result in a distorted definition of the relevant market. The problem must be seen from two viewpoints: (i) whether the relevant market is delineated for determining the measure of transport market concentration in a particular territory (in this way it is identified by the EC in assessing national implementation of the railway packages ), or (ii) whether it should serve to determine the shares of individual operators in certain destination (e.g. for proceedings to deal with the abuse of dominant position). In individual investigated cases it is also necessary to study and take into account specific features of each region from the point of view of real substitution of transport modes  – definition of relevant markets for certain transport services using geographic definition can differ by (not) including the intermodal competition.
A precise definition of the geographic market can be very significant when there is a high share of international transport at a particular destination; it is very common in freight transport, e.g. for Estonia international transport accounts for 93% of output (in the year 2005, in tonne-km), in Latvia 88% and in Slovakia 86,5%; passenger transport has significantly lower shares: the highest share of international passenger transport was in Luxembourg 24%, Austria 17% and Belgium 15% outputs in passenger-km . In delineating the relevant market within a destination it is necessary to take into account not only the current transport flow but also potentially competitive routes between prevailing point of origin and the destination of international transit transport flow.
Fig. 1 Diagram of transport routes for definition of the geographic relevant market A-B
Fig. 1 illustrates the example when a geographic relevant market must be defined for a case solved at a destination between points A and B. At first we can consider the relevant market limited by rail services on the route between the two points. It is possible, however, that an important competitive potential may have a combination of routes A-Z and Z-B; thus the relevant market could be identified at destinations A-B and A-Z-B together. The next step could examine the origin and destination of the prevailing transport flow: if a considerable part of transport between points A and B was formed by transport from X to Y, the relevant market could be identified by including alternative destinations connecting X and Y. This can be significant with respect to international transport. If the combination of routes A-Z and Z-B was not identified as a substitute in the previous step, it is still possible that a competitive alternative to the transport X-A-B-Y will be the transport X-AZ-Y. Then all the combinations must be assessed in view of intermodal competition and substitutes must be identified within road, air or water transport. Even if the combination A-Z-B with a bend at Z does not represent a demand substitute for line A-B, it does not necessarily mean that the route via Z can be excluded from considerations: we are still to explore the market from temporal dimension.
6. Temporal dimension
In delineating the product and geographic relevant market, temporal dimension must be considered as well. For example in retail business, this time dimension expresses the same opportunity for buying and selling over a given period, which can be specified administratively (shopping hours) or naturally (seasonal selling of agricultural produce). Similar restrictions can arise in transport because of a limited capacity of terminals (both passenger and freight) and roads at peak hours, etc. Temporal dimension is especially important for assessing potential possibilities of entry of new producers onto the market, i.e. for assessing potential supply substitution. Increasing of a price of a product in the relevant market is a strong motivation for other producers, who have so far produced other, either similar or relatively different products, to enter the market of a given product; but it takes a time. If the delay of other producers entering the market is not in excess of 6 to 12 months, then the existence of significant barriers to entry is not expected . Determining the relevant market for transport services one must consider other potential carriers operating at other destinations as well and their chances to begin operating at the destination in question. The possibility to enter a destination is given by several factors: the speed of obtaining the licence, if required; existence of subsidies for the destination and the possibility to apply for them; period of contracts; existence of economies of scale and economies of scope; transport flow density; existence of network economies; integration into regional integrated transport systems; level of necessary sunk costs; consumer habits
and additional services.
European air transport often defines as relevant markets individual destinations with existing carrier/carriers, without taking into account other operators that would be able and willing to enter onto the given destination under the condition of growing price; regardless of the fact that the EU air transport is currently more open to free competition than the railway transport. The reasons for it are existing barriers to entry of other operators: large European airports are congested and obtaining landing slots is a lengthy process and sometimes impossible over the decisive time period . The situation is similar at many railway destinations: certain stations or track segments are afflicted by congestions in particular at peak hours and the entry of another intramodal competitor is impossible because of insufficient capacity of infrastructure.
A specific case of definition of the relevant market from temporal viewpoint is the timetable of passenger services: peak-hours suburban and intercity transport cannot generally be substituted by transport at off-peak-hours, since real substitution of commuting at fixed daily or weekly times cannot be provided. In contrast, special transport products based on off-peak discounts will be freely substitutable during daily and weekly timetables.
The model situation in Fig. 1 illustrates that the relevant market definition can encompass potential competition, which can result from a hypothetical price rise. Even if we eliminated the combination of routes A-Z-B from the relevant market, carriers operating on routes A-Z-B would be able to adapt to changed conditions over reasonable time period and to create competitive supply on the direct route A-B. The same applies to the transport X-Y and, of course, inclusion of intermodal competition.
7. Failure of the SSNIP test – cellophane fallacy
A practical application of the SSNIP test is rather exceptional in railway industry  and relevant markets are usually delineated mainly on the network basis. As shown above, this solution is not quite appropriate and poses risks of significant distortions in the dominance power of individual operators in the market. Therefore the SSNIP test has a potential application in network industries as well , but as an auxiliary method with certain corrections.
Application of the SSNIP test to defining the relevant market fails if there is a powerful, obviously dominant entity in the considered market. For this purpose the methodology of the SSNIP test cannot be based on comparing existing market price of a product, since the rationale of the test implies that it would lead to the definition of a very broad relevant market and therefore to underestimating the dominance power of the investigated subject. In this case the test must be based on hypothetical competitive price  and the relevant market precisely defined using qualitative criteria of the product concerned and its potential substitutes. Then it may become apparent that the substitutes identified by the SSNIP test are not real substitutes at all and that the test only indicated that consumers stopped using the product considered. This problem is known as “cellophane fallacy“ both in theory and practice (in more detail see ).
The principle of the test failure in the case of ”cellophane fallacy“ is that a drop in consumption of the considered product is mistaken for a natural consumption switch to other substitutes. The problem is to identify the product that can still be considered as substitute and what is assessed as consumption constraints. In the railway industry this problem concerns both products, geographic and temporal definitions of the relevant market. Due to the fact that demand for transport services is derived demand, the broadest definition of the relevant market would answer the question, whether transport of passengers or goods as such could be substituted by consumption of other service. Although such delineation does not make sense at first sight, it is logical in certain cases. In freight transport it practically corresponds to a reversed logic of defining the geographic relevant market and the question can be formulated like this: following a rise in fares will consumers have a possibility to substitute transport e.g. in purchasing some raw materials by purchasing the raw material at the point of consumption (probably at a price higher than the sum of the raw material price bought at the original point plus the fares, but it will be still lower than the price of the raw material at the point of consumption without transport costs increased by 5-10%)? In a similar way we can think about passenger transport: if a consumer has an option to substitute e.g. daily commuting to work by taking a job in their domicile following a rise in fares (so that the income in the original job diminished by more expensive commuting will be lower overall than the income in the new job, which is itself lower than the income in the original job together with the original fare)? Although such definition does not lack economic logic, definition of relevant markets in transport industry will primarily define markets in a narrower sense, while meeting the prerequisite of non-substitutability of transport service. And yet the relevant market definition will have to comprise geographic and temporal specific features.
8. Issue of intermodal competition
What is crucial in the respect of intermodal competition is the price of transportation in comparison with the quality of the service offered. In passenger transport, speed, quality of the interior, feeling of safety, additional services are considered important features; that all causes continuing price differences between individual modes. In freight transport, speed (there is a great difference between transport of various products: the extremes being individual packages on the one hand and substrates on the other) delivery reliability and the possibility of precise timing (just-in-time logistics). When the ratio of the perceived service quality and its price are changed, willingness to switch to other modes will change too. However, the possibility of substitution between transport modes may incur switching costs, which the consumer must expend as lump payment ; it may pose a problem chiefly in the freight transport market definition. Thus to correctly define the relevant market it will be essential to assess physical and technical properties as a complementary method, which means to earmark the products, which could be perceived as substitutes under certain circumstances and to test the definition of the relevant market including these substitutes. It is also possible to use historical experience and find out, whether any intermodal shift occurred under similar circumstances in the past and by doing so to determine a potential substitute for testing the relevant market .
Existing prices in the markets belonging to specific (predominantly network) industries, which are regulated ex ante, may not be competitive prices. It may be caused by direct price regulation, regulation of access to the infrastructure, price regulation for the access and a number of other restrictions or on the contrary supports  – for example a system of subsidies in passenger railway transport. A correct application of the SSNIP test assumes that the hypothetic monopolist is not influenced in their conduct and price setting by any administrative regulatory intervention, and at the same time that the price level of products beyond control of the hypothetical monopolist is stabilized at the competitive level by market forces. Bender  believes that in such a case the situation will develop in contrast to what is described by the cellophane fallacy: existing price in the market will be lower than the competitive one, the applied SSNIP test will indicate an increase in incomes along with the increase in price, which will lead to defining of too narrow relevant market (without including other transport modes) and to defining strongly dominant position of the subject under investigation, without taking into account intermodal competition. Bender  also warns about another distortion, which is characteristic for network industries: existence of economies of scale, network economies, economies of scope and sunk costs result in difficulties in estimating how high is the competitive price in the market and whether and how it actually differs from the existing market price. Again, this is very likely situation within the railway industry.
The danger of distortion in defining the relevant market in the railway industry can be caused by overestimating the importance of a network for product demand substitution. The relevant market should not be a priori limited by network technological parameters, since they need not be critical for finding substitutes: e.g. the existence of fixed phone network is not a limiting factor for substitution of cellular phone services provided by mobile operators . A time dimension is not negligible because of technological progress: for instance thanks to developing high-speed wireless communications the fixed telephone
network is not important for defining the relevant market for internet connection. Identically, defining the relevant market in transport services cannot be automatically limited by the existence of network of railway tracks.
If the existence of a dominant subject in the market is proved beyond doubt, Motta  suggests that the SSNIP test is corrected in several ways. First of all it is necessary to verify price elasticity of demand, econometrically compare its relevance while taking into account the current price change of potential substitutes, revenues, aggregate price level etc. A good guidance may be the cross-price elasticity of demand: its relatively high value will point to substitutes and such products can be included in one relevant market; if the cross-price elasticity is low, it is not a case of substitutes and the products will not be a part of one relevant market. In railway transport it means that intermodal competition must be included in defining the relevant market. A similar concept, based on the correlation analysis of price development for a market substitute, was applied for the first time by Stigler and Sherwin : the higher the correlation of the development of a product’s market prices, the more likely it is that it is a case of close substitutes . However, the test can be distorted due to the movement of macroeconomic aggregates (revenue, inflation etc.) and also due to tariff regulation, access prices and the like. Another possibility is to use direct comparison of average price level of individual products: if it varies greatly, products will not be part of the same relevant market, since they will probably be demanded by different consumer groups.
Regulatory measures should principally protect the competition within an industry as such, not individual competitors ad hoc. As illustrated above, railway transport industry naturally tends to the victory of a couple of dominant operators and is typically strongly concentrated even in the conditions of unbundling. The market of railway transport services displays a number of specific features, which considerably complicate the definition of the relevant market. The biggest problem is the application of the SSNIP test – this currently common method used for defining relevant markets fails as far as railway is concerned and its application in this industry is rather limited. In practice the SSNIP test analysis of the industry must be enhanced by using additional tests based on a detailed definition of the product market as for demand substitution and also supply substitution, with a special focus on temporal definition of the market and in particular a very precise definition of the geographic market. Geographic definitions will require the analysis of viability of intermodal competition, technical limitations, track capacity, and in particular the size of real and potential transport density.
 BENDER, M. – GÖTZ, G. – PAKULA, B. Effective Competition: Its Importance
and Relevance for Network Industries. Intereconomics, 2011/1, s. 4 – 10
 BRENAN, T. J. (2008) Applying ‘Merger Guidelines‘ market definition to
(de)regulatory policy: Pros and cons. Telecomunications Policy, 32/6, s. 388-398
 BOONE, J. – VAN OURS, J. –VAN DER WIEL, H. (2007) How (not) to Measure
Competition. CPB Discussion Papers, No. 91, CPB Netherlands Bureau for Economic
 BOUTTES, J. P. – LEBAN, R. (1995) Competition and regulation in Europe´s
Network Industries from Theoretical Approach to Sectoral Application. Utilities Policy,
5/2, s. 127-146
 DALJORD, O – SORGARD, L. – THOMASSEN, O. (2007) Market Definition
with Shock Analysis. discussion paper SAM 36, Norwegian School of Economics and
 GRIFFITH, A. – WALL, S. (2004) Applied Economics. Harlow: Pearson
 HAUSMAN, J. A. – SIDAK, J. G. (2007) Evaluating Market Power Using
Competitive Benchmark Pricing Instead of the Herfindahl-Hirschman Index. Antitrust Law
Journal, 74/2, s. 387-407
 IBM (2011) Rail Liberalisation Index 2011 Market Opening: Rail Markets of the
Member States of the European Union, Switzerland and Norway in comparison. Brussels:
IBM Global Business Services
 KAHN, A. E. (2006) Telecommunications: The Transition from Regulation to
Antitrust. Journal on Telecommunications and High Technology Law, 5/2006, s. 159-188
 KOMISE EVROPSKÝCH SPOLEČENSTVÍ (1996) A strategy for revitalizing the
Community’s railways. COM(96) 421 (Bílá kniha), Brusel
 KOMISE EVROPSKÝCH SPOLEČENSTVÍ (1997) Notice on the Definition of
the Relevant Market for the Purposes of Community Competition Law. OJ C372, 9.12.97),
 KOMISE EVROPSKÝCH SPOLEČENSTVÍ (2007) Sděleni Komise Radě a
Evropskému parlamentu o vývoji na železničním trhu. KOM(2007) 609
 KVIZDA, M: (2006) Faktory efektivnosti železniční dopravy – Jak rozhodovat o
dopravní politice? Národohospodářský obzor, VI/4, s. 37-49
 KVIZDA, M. (2008) Unbundling a konkurence na železnici. In: Kvizda, M. –
Tomeš, Z. (eds.) Konkurenceschopnost a konkurence v železniční dopravě – ekonomické
a regionální aspekty regulace konkurenčního prostředí. Brno: Tribun, s. 7-20
 KVIZDA, M. (2009) The State and Railways – A Historical Comparison of
Economic Policy in Relation to Railways. In Issues in Economic Performance – Business,
Regional and Transport Issues. Berlin: LIT Verlag, s. 115-137
 KVIZDA, M. (2010) Impacts of Unbundling on Competitiveness of Railways.
Železničná doprava a logistika, 2010/2, s. 66-72
 KVIZDA, M. (2011) Vymezování relevantního trhu v odvětví železniční dopravy.
In Regulace konkurenčního prostředí na železnici – teorie v centru a praxe v regionech.
Brno: Masarykova univerzita MuniPress
 MASSEY, P. (2000) Market Definition and Market Power in Competition
Analysis: Some Practical Issues. The Economic and Social Review, 31/4, s. 309-328
 MOTTA, M. (2004) Competition Policy. Theory and Practice. Cambridge:
Cambridge University Press
 OFFICE OF FAIR TRADING (2004) Market Definition. London: Office of Fair
 PELTRÁM, A. (2010) Hospodářská soutěž na železnici v pojetí Evropské unie. In:
Kvizda, M. – Tomeš, Z. (eds.) Konkurence na evropských železnicích – ekonomické,
právní a regionální faktory. Brno: Masarykova univerzita, s. 20-31
 SLANÝ, A. a kol. (2003)Makroekonomická analýza a hospodářská politika,
 STIGLER, G. J. – SHERWIN, R. A. (1985) The Extent of the Market. Journal of
Law and Economics, 28, s. 555-585
 Zákon č. 143/2001 Sb. o ochraně hospodářské soutěže ve znění pozdějších
 TOMEŠ, Z. – POSPÍŠIL, T. (2006) Ekonomické aspekty železniční dopravy.
Brno: Masarykova univerzita
Full text can be found here: Relevant_market_2012