Influence of the evolution of high-speed railway infrastructure on the success of Italian liberalization

Abstract

The arrival of a new operator on the Italian high-speed railway (HSR) market, its maintanance and its market share made Italian open access experience one of the most successful liberalization models in the HSR sector. Researchers noticed that since the entry of the new operator, expansion of Italian HSR market is mostly due to the presence of this new operator. The aim of this article is to establish whether there were some other characteristics of the Italian HSR market that may explain this success even if it may be in opposition with what theory commonly suggests about competition in HSR markets. This research tries to do a complementary analysis by making a comparison with another successful HSR market that is not already liberalized: the French HSR market. After retracing supply, demand and markets maturity of Rome–Milan and Paris–Marseille lines, results reveal that supply and demand evolutions in both markets were very similar, if not identical after the commissioning of HSR. It suggests that liberalization may not be the only explanation of the significant evolution of Italian HSR market but that the opening of new infrastructures may also lead to positive trends that remain until markets reach a high level of maturity. This maturity understood as the residual capacity of the network and partially determined by infrastructure improvement seems to be another variable that had influenced the success of NTV. Indeed, before liberalization, Italian HSR network was far from saturation that allowed new entrant to capture important market shares.

Keywords

Italian high-speed railway Italian open access competition infrastructure evolution residual capacity of railway network French high-speed railway commissioning of HSR infrastructure

Introduction

With the adoption of Directive 91/440/EC, European Union member states have begun to define their objective of gradually opening their railway markets. With regard to passenger markets, the member states agreed on the third package (2007) to open the market to international services and to allow private operators to practice cabotage, provided that they do not jeopardize the economic equilibrium of local public services. Since the opening of internal passenger railway markets is still limited, the fourth package, adopted in 2016, obliges countries to open their internal markets not later than December 2020 for competing commercial services and not later than December 2023 for public service rail contracts.

Some European member states decided to open their markets well before the date set by the fourth railway package. However, markets that were already open did not follow the same kind of liberalization. As noted by Nash, Nilsson, and Link (2013), the most common type of competition where new operators succeeded in acquiring a large share of the passenger market was when they competed for public service obligation (PSO) routes. Indeed, the tendering process is the most common model of railway liberalization in Europe. In contrast, only a few countries have decided to open their commercial passenger markets with open access; generally only a few entrants gain access to these markets and compete against the incumbents.

After several years of liberalization in these countries, several papers have presented varying results obtained through open access competition, in most cases limited by very few examples of competition for high-speed railway (HSR) services. There is only one experience of HSR service liberalization under open access competition, where a new entrant has succeeded in staying in the market with a non-negligible market share. Indeed, the Italian liberalization of the HSR market can be considered exceptional in the European context, since no other example of effective open access competition for HSR services has been observed.

With that in mind, the literature regarding this experience generally attributes a large proportion of the evolution of the Italian HSR market to the arrival of the new operator, NTV. Furthermore, most of these studies have focused their analysis on the evolution of the Italian HSR market with respect to its opening, judged by default as the main change to be taken into consideration. Indeed, many of these studies did not thoroughly analyse the impact of the construction and subsequent opening of the Italian HSR infrastructure.

To fill this gap in knowledge, the aim of this article is to focus analysis over a longer period that takes into account the differences in the market evolution before and after the commissioning of the new HSR infrastructure. The main purpose of this exercise is to identify the Italian HSR market trend before the arrival of NTV, with the aim of establishing the share of the market evolution due to the commissioning of HSR lines and that of the market evolution that could be attributable to the arrival of the new operator in the market.

Hence, the main hypothesis of this research is that the evolution of the Italian HSR since 2006 is due to two main factors that should be studied jointly: the commissioning of HSR lines and the liberalization of the HSR market.

The assumption underpinning this research is that comparison with another European HSR network will lead to better understanding of the Italian HSR market trend before the arrival of the new operator. In this regard, given that the French HSR network is the most developed European HSR in terms of train-km and passenger-km, as well as one of the oldest HSR European networks, this study considers that a comparison between these two networks may clarify the context and take a more objective view of the evolution of the Italian HSR network with respect to the commissioning of the HSR infrastructure.

This research considers that it is pertinent to compare one market without liberalization (the French HSR market) and another market within a context of liberalization (the Italian market), to identify the impact of commissioning HSR infrastructures in different contexts.

Hence, this article aspires to enrich the literature on the specific effects of commissioning HSR infrastructures and opening HSR markets. However, certain specific aspects of the HSR market should be taken into consideration in this kind of study. Therefore, after a brief review of European open access liberalizations, the second part of this article will focus on certain key entry barriers that might prevent the successful arrival of new entrants from a theoretical point of view. The aim of this is to establish whether or not the Italian HSR market has been subject to such barriers and, if not, to suggest other factors that could contribute to the success of the Italian HSR market.

Analysis will follow in the third part, with a description of the Italian HSR liberalization and the different steps in the evolution of the Italian HSR infrastructure.

The fourth part will be dedicated to the description of the French counterpart HSR market, its institutional evolution and its infrastructure evolution.

The fifth part will focus on a comparison of the evolution of both markets in terms of supply, travel times, market share and, primarily, the maturity of each line, measured with respect to the level of saturation on a line or on a part of it, which will be compared to identify the main drivers behind the evolution of each market.

This fifth part can be considered as the core of the article, explaining its length. As already mentioned, the comparison of the two markets focuses on the magnitude of the infrastructure’s evolution, which can be considered a possible main variable explaining the success of the new entrant in Italy. A central point of the discussion is that the degree of development of the HSR infrastructure is directly correlated with the level of saturation, and hence with the level of residual capacity of the HSR network. The latter is also suspected to be one of the main variables possibly explaining the evolution of both HSR markets.

Finally, the last part will draw several conclusions. It will be noted, among other things, that to our knowledge no research has yet studied the impact of commissioning high-speed lines (HSLs) on the evolution of the Italian market. However, these openings seem to be one of the main explanations for the success of HSR in France and Italy. Thus, one of the conclusions of this article is to signal that research on the impact of competition in the Italian HSR market should not be conducted without taking into consideration the importance of commissioning HSLs.

Rail liberalization and barriers to entry

The following section shows that very few countries have experienced effective rail competition in Europe. In general, the share of new entrants in the markets remains relatively low in most European countries. This may be explained by the characteristics of the rail industry, as the barriers to entry are structurally relatively high.

European experiences

According to Casullo (2016), among the 15 countries whose markets are open by law, only 6 countries have experienced effective competition in the market: Germany, Great Britain, Sweden, Austria, Czech Republic and Italy.

The different experiences of open access in these countries can be categorized into three groups.

The first group includes countries where new operators have focused their interest on, or have found opportunities to enter into, the market via niche markets. Niche markets are most commonly regional or specific railway lines. Different factors may explain new entrants’ decision to focus on this type of market, but the most common is that the incumbent has decided not to develop these markets/lines. New entrants have seen this situation as a means of entering the market.

Germany and Great Britain belong to this first group of countries that have faced very limited entry of new operators in the long-distance market. In Germany, new operators have focused their services on markets where the incumbent had very few incentives to invest its energy, as could be seen with the market for night trains (just one operator for night journeys between Germany and Sweden: the Berlin Night Express). The only operator that provides a significant long-distance service (Hamburg–Cologne–Frankfurt) is the company HKX. Nevertheless, the frequency provided by this new entrant is limited (between one and three trains per day in 2016). Indeed, the market share of new entrants in Germany has not exceeded 1%. This situation contrasts with what has happened in regional services, where the market share of alternative operators reached nearly 30% of train-km and 18% of passenger-km after almost 20 years of competition in the market (Link, 2016).

The case of open access competition in Great Britain is not very different from the German situation. Indeed, although there are eight open access operators nowadays,¹ their market share does not amount to more than 1% of the whole British market (CMA, 2016). Moreover, since the 1990s, the British railway market has been organized into more or less 20 franchises through which train operators regularly compete for the market.

The second group of open access experiences can be characterized by markets where the entry of new operators took place on the main railway corridors of each country with a non-negligible number of daily frequencies between origins and destinations. This group includes the Czech Republic, Austria and Sweden. However, Sweden has new operators that managed to enter through niche markets as well as on the main Stockholm–Gothenburg line, and the presence of private operators is significant on regional markets under PSO contracts (Alexandersson & Hulten, 2008).

Sweden was one of the first European countries to open its market to open access competition. The Swedish government opened these services to competition in 1990 for public service rail contracts and in 2006, 2009 and 2011 for competing commercial services. However, for more than a decade, competition in the Swedish industry has taken the form of competition for the market rather than competition in the market. Several new operators have entered the Swedish market through the mechanism of tendering, which concerns PSO services. As observed by Nilsson, Pyddoke, Hulten, and Alexadersson (2013), for a long time, those who tried to compete in the long-distance market with the incumbent (SJ) had negligible market shares. Nonetheless, since 2015, a new operator has decided to compete in the market against the incumbent for the Stockholm–Gothenburg route. On this route, the Chinese company MTR Express reached a significant level with seven daily services, which is considerably more than the other operators in this market (e.g. Bla Taget and Snälltaget). According to Vigren (2017), the incumbent reacted to this new service by reducing prices by 13%.

The case of open access competition in Austria and the Czech Republic is relatively similar. In both countries, a new entrant decided to compete on their main routes. In Austria, a private company (WESTBahn) took almost 20% of the market share for the Vienna–Salzburg route 2 years after its arrival in 2011. In the Czech Republic, two new entrants arrived on the Prague–Ostrava route: RegioJet since 2008, and Leo Express since 2012. The market share of new operators for these services reached almost 40% according to Tomeš, Kvizda, Nigrin, and Seidenglanz (2014) and Tomeš, Kvizda, Jandova, and Rederer (2016) or 50% according to CMA (2016). Moreover, in both cases, the arrival of new entrants has had a great impact on prices on the market: after the entry of these new operators, ticket fares decreased by 33% to 50% in both markets.

The third group of successful open access competitive experiences is represented by just one experience in which the new operator has achieved tenure and a large share of the HSR market. This experience is attributable to the liberalization of the Italian market and the arrival of a new operator, NTV. The situation is characterized by very intensive and direct competition with the incumbent, with frequent services between origins and destinations.

In brief, three main conclusions can be drawn from these experiences. Firstly, effective competition is more likely to happen in the case of competition for the market (tendering for PSO) than competition in the market. Indeed, this could be the reason for on-track competition in niche markets. Secondly, when open access competition is effective, as in the Austrian and Czech cases, competition in the rail market concerns a relatively small number of operators (one or two new entrants versus the incumbent), and mostly on main rail corridors. This means that the rail market is far from being a perfect competition model in which atomistic operators compete in the market. Thirdly, it appears that apart from the Italian case, which will be explored later, there are no cases of effective open access competition in the high-speed rail market.

Barriers to entry

The preceding section suggests that competition is relatively poor for long-distance passenger services in the European rail market, since the number of new entrants is relatively low. However, during the 1980s, economic theory took a new view of monopolistic structures, traditionally believed to yield inefficient and non-optimal outcomes. In particular, the theory of contestable markets suggests that potential competition leads to more efficient outcomes than was previously thought. The presence of an incumbent company which holds a large share of the market is not necessarily sub-optimal if the market is contestable. Stated differently, the absence of new entrants in the rail market is not necessarily problematic if there is freedom of entry.

Observation of the rail market nevertheless suggests that barriers to entry are quite high for a new entrant. Indeed, barriers to entry for newcomers in the railway market are a critical issue, particularly for HSR services. Firstly, it appears that the cost structure of the railway industry – and, more precisely, the HSR market – mainly depends on the size of the market (economies of density) or on how the operator can adapt its costs with respect to the different services it may offer in the market (economies of scope). In both cases, a new entrant in the HSR market has a big disadvantage compared to the incumbents because of its lesser knowledge of the market. Indeed, new operators cannot enjoy the economies of scale that an incumbent might derive from its previous dominance as a monopolist. As Beria and Grimaldi (2017) suggested for the German experience, on-track competition may be very limited ‘because of the market power of DB and by its extreme level of integration with local services, making point-to-point open access operators less attractive’ (scope and density economies).

Moreover, other factors related to demand characteristics explain that it is hard for a new entrant to compete with the incumbent in the rail industry. In particular, the academic literature points out that a passenger is more likely to choose the operator with the higher number of frequencies (Mohring, 1972). This effect, called the Mohring effect, suggests that the incumbent has a competitive advantage in comparison with new entrants.

In addition, the cost that a newcomer must bear to enter an HSR market implies very intensive investments in rolling stock, since HSR rolling stock is more expensive than traditional trains, and in staff training, principally in a historically monopolistic market. As all these costs are sunk costs, entry in the market is not free. These characteristics considerably reduce the probability of having several railway undertakings in open access competition. The contrary can be observed in the case of liberalization for the market. Indeed, new operators looking to compete for a market by participating in different tendering processes are not confronted by this kind of investment because rolling stock is generally ensured by transport authorities. However, researchers have highlighted certain conditions under which open access competition may take place.

From a theoretical point of view, it is therefore relatively difficult for a new entrant to provide a sustainable service in open access competition. One of the main strategies that a newcomer might follow to break into an HSR market could be based on differentiating the service, by focusing on niche markets. This strategy may be equated with a monopolistic competition, as developed by economists following the seminal research of Chamberlin (1933).

In this regard, Preston, Wardman, and Whelan (1999) suggested that the way for new entrants to succeed in HSR markets is to adopt a low-cost model and marginal infrastructure pricing, as in the air transport market: this is the only way for a private operator to achieve financial sustainability. Ivaldi and Vibes’ (2005) conclusions take the same line: if alternative operators want to obtain a considerable HSR market share and not be present in reduced niche markets, they should follow a low-cost model. They must considerably reduce their costs, or at least create enough cost difference with respect to the incumbent, to allow them to reduce their prices sufficiently and thus create effective competition with the incumbent (Johnson & Nash, 2012).²

With respect to the Italian case, theory suggests that the new operator had to face all these types of barrier to entry. However, NTV has succeeded in obtaining a large market share. Consequently, the pertinence of this article is to provide better understanding of the reason for this success.

The Italian case

Experiences in different European railway markets that have opened their long-distance passenger services suggest that the effectiveness of open access competition is relatively limited with respect to the market share of new operators. Since 2012, the arrival of a new railway undertaking in the Italian HSR market has shown an inflection point regarding this type of competition in Europe. The much publicized success of the new Italian HSR operator makes the Italian experience the only example of open access competition for HSR in the world.

Italian HSR liberalization

Institutional changes and public railway policies are important factors that can help in understanding the big picture of this market’s evolution. Although our research does not study this question in a detailed way, it tries to present the principal institutional changes and the principal public policies in railway transport that may have an impact on the evolution of the railway market.

In Italy, the rail industry has evolved over the last two decades in accordance with the evolution of the European legal framework, creating a single European railway area. Between 1960 and 2010, the governance of the railway industry changed radically in Europe and in Italy. In particular, a major change in the legal context of European railways occurred at the beginning of the 1990s with Directive 91/440/EC. The Italian government implemented this directive in Italian legislation. Hence, the Italian government decided to reorganize the governance of the railway industry, firstly through the creation of the new Ferrovie dello Stato SpA as a joint-stock company in 1992 and then by separating the two main divisions of the company, infrastructure and transport service, in 1998. This separation was supposed to be the first step in preparing the market for the liberalization of railway activities.

Nonetheless, although Italy was one of the first countries to open its railway market, the government took on the task of regulating the market from 2004 to 2013. This created a paradoxical situation intended to ensure real competition in the market. It was only in September 2013 that an independent regulatory body, the Autorità di Regolazione dei Trasporti (ART), was founded. Since then, the ART has played a very active role in the railway market.

Moreover, in Italy, the passenger transport service has been open to competition since 2002 for regional railway transport, but also for long-distance, high-speed trains (HSTs) (Desmaris, Croccolo, & Patuelli, 2016). Regarding local services, regional authorities are free to choose, through a tendering process, the operator that will supply the regional railway services. Consequently, some regions decided to open their markets through tenders but no newcomers won a tender against the incumbent. Bitzan (2003) and Preston (2009, 2014) suggested that the small size of Italian regions, which does not involve a large volume of passengers, could be the main reason why the incumbent always won the tenders. The small size of regional demand can also be explained by the liberalization of the coach market.

Concerning the HSR market, the end of 2006 was marked by the creation of a new private railway company, NTV, which aimed to compete with the incumbent in the long-distance market. Nevertheless, NTV only began its operations in April 2012 for the Milan–Naples route (with a stop at Rome), after several years of waiting for permission to operate on the Italian railway network. According to numerous papers, the arrival of NTV radically changed the face of the Italian market. Many studies have sought to analyse the impact of this liberalization on the HSR market and suggested that it had direct impacts on prices, frequencies, number of passengers carried, quality and variety of services and so on (Bergantino & Capozza, 2015; Beria, 2015; Beria, Redondi, & Malighetti, 2016; Finger, 2014; Patuelli, 2015).

For instance, Cascetta and Coppola (2014) estimated that a reduction of about 30 to 34% in the average price between Milan and Rome was due to the arrival of the new entrant, NTV. The authors also suggested that the decrease in prices was due mainly to a new pricing scheme implemented by NTV and previously adopted by air market operators: yield management. Concerning frequencies, Cascetta and Coppola (2014, 2015) and Bergantino, Capozza, and Capurso (2015) estimated that frequencies increased by almost 80% after liberalization. Desmaris (2016) concluded that all these characteristics contributed towards increasing the consumer surplus, creating a virtual spiral and enhancing railway demand and supply.

However, to our knowledge, none of these studies analysed recent trends in the Italian rail market, considering the evolution of the larger market of long-distance mobility. In our opinion, the high-speed market should be analysed within this broader view. More precisely, intermodal competition also plays an important role in the evolution of the rail market. It is essential to consider the quality of service of each mode in terms of travel time, for instance, to understand the evolution of rail demand. In other words, it is important to understand the ‘big picture’ of the Italian long-distance transport market framework before and after liberalization. This means that we should not only focus on the HSR market but also observe what happened to the infrastructure and analyse the air, coach and car transport markets (Albalate, Bel, & Fageda, 2015; Beria, Debernardi, & Ferrara, 2017; Capozza, 2016). Our analysis will show that the high-speed market in Italy was also characterized by the construction of a new HSL between Rome and Milan between 2007 and 2009. This evolution may provide a hint at new leads in understanding the evolution of supply and demand regardless of the opening of the market.

Infrastructure

Institutional changes are not the only factor that can explain the evolution of the railway market. Another key element capable of explaining the growth of this market is its competitiveness in comparison to the other modes. More particularly, travel time is a crucial factor for the rail service in its competition with air transport. In other words, the competitiveness of rail services is strongly influenced by the presence or not of a dedicated HSL (Prussi & Lonza, 2018).

Indeed, the origin of the success of Italian rail operators can be traced back to 1960, with the Italian government’s decision to build a direct railway path between Rome and Florence. This link was the first European HSR project, and it took the name la Direttissima. The project was approved in 1968, construction began in 1970 and the first 138 km (of 254 km) section of the line was inaugurated in 1977. This first section of line allowed trains to reach a maximum speed of 250 km/h. The construction of the remaining two other parts of the line was achieved in 1985 and 1991.

At the end of the construction of la Direttissima, the fastest train travelling between Rome and Florence was supposed to take 1 h and 20 min. Based on the information found on the public timetables of Trenitalia, this new HSR corridor had a direct impact on the travel time between Rome and Florence. Travel time fell from 2 h 42 min in 1976 to 1 h 37 min in 1992, representing an improvement of 40%. However, these trains never reached the ex ante objective of a travel time of 1 h 20 min. At the same time, it can also be observed that travel time between Rome and Milan was reduced from 5 h 55 min to 4 h 25 min, which is an improvement of 25%.

After the opening of la Direttissima, the Italian government announced the construction of new high-speed sections between Florence and Bologna (79 km) and Bologna and Milan (214 km) in 1996 and 1998, respectively. The construction of these links radically changed the trend of passenger transportation between these cities and, more generally, the travel time between Rome and Milan.

Indeed, after the opening of the Florence–Bologna section in 2009, and the lines between Milan and Bologna in 2007 and 2008, the two main Italian cities, Rome and Milan, were finally connected by an HSL. This change had many consequences for the long-distance transport market in Italy.

As Beria and Grimaldi (2016) noted, the opening of these two new high-speed tracks has significantly reduced travel time for rail transport. The opening of these lines may lead to an increase in the train modal share between Rome and Milan. The trend of modal share in supply and demand between Rome and Milan must be studied from dates before the commissioning of these HSLs along the entire path to better understand the present market shares.

The French counterpart

The French market is an interesting counterpart to the Italian case. As with the Italian HSR market between Rome and Milan, high-speed infrastructure is highly developed. However, unlike the evolution of institutional changes and public policies undertaken in Italy, France has not yet liberalized the HSR market, and much less the regional and long-distance markets (Crozet, 2014). It is therefore interesting to compare the evolution of the rail market in both countries to observe whether they follow significantly different paths.

Rail liberalization in France

As stated above, from the outset, the decisions taken by the Italian government were in harmony with the evolution of European railway policies regarding railway infrastructure and the European legal framework for the railway market. While Italy opened its entire railway passenger market (HSR included) to private operators, the French government decided not to open domestic markets until it became mandatory due to European regulation. Indeed, the international passenger market opened in 2009. The French regulatory body for railway activities (ARAF, later Arafer) was created in December 2009.

Although the French government decided to wait before opening the domestic rail market, changes took place with the introduction of other kinds of service offered by the incumbent. To face the opening of the HSR market, the French incumbent developed a new HSR service after 2004 with the introduction of low-cost services on several routes. The first low-cost service, iDTGV, was opened on the route between Paris and Marseille. One of the characteristics of this new service was that reservations could only be made online. Also, among other particularities, luggage restrictions allowed the operator to reduce its costs.

This service is currently generalized under another commercial name: Ouigo. The principles of Ouigo are the same as iDTGV. Indeed, Ouigo is the translation of a low-cost service for the rail market. Tickets are still sold online, the on-board staff is reduced, luggage is limited and passengers have to pay if they want to carry more luggage. Another particularity that permits this new company to stay in the market is the significant difference in the use of rolling stock per day. Indeed, Ouigo trains operate for 13 h per day on average, while regular TGV trains operate for 7 h per day. This service is expected to represent 20–25% of the demand in 2020.

Infrastructure

Although there are certain differences concerning the liberalization of the rail markets in Italy and France, both countries are characterized by the existence a relatively highly large high-speed network. Eleven years after the Italian government took the decision to build an HSL connecting the two main Italian cities, the French government agreed to evaluate the possibility of building an HSR linking its two main urban areas.

Whereas the first European HSL was projected by the Italian government with la Direttissima, the first line fully opened to commercial service in Europe was the French line LN1 connecting Paris to Lyon, the two largest conurbations in France, in 1981. LN1, or LGV Sud-Est, covers 427 km of HSR between two stations, one in the centre of Paris (Gare de Lyon) and the other in the centre of Lyon (Lyon Part Dieu). This new line is 87 km shorter than the regular line and trains can reach a cruising speed of 270 km/h. The project was declared a public utility project in 1976, and the construction of the line began at the end of the same year. The opening of the first 302 km of the line took place in September 1981, while the second portion was commissioned 2 years later. As expected, travel times changed radically after the opening of each section of the line (Bonnafous, 1987). After September 1981, the travel time between Lyon and Paris fell by 56 min and has continued to fall since the opening of the second section in 1983, passing from 2 h 50 min to exactly 2 h.

After the successful opening of LN1 in France, the French government decided to continue the construction of its HSR network. Other lines were opened, like the LGV Atlantique in 1990 and the LGV Nord in 1993, and the extension of LN1 from Lyon to Valence in 1992–1994, which took the name LGV Rhône-Alpes, or LN4.

The success of high-speed rail encouraged the French government to extend LN4 from Valence to Marseille to reduce travel time between Paris and Marseille. The decision to study the possibility of extending the HSR network from Valence to Marseille was taken in 1992, and the government declared the construction of this connection a public utility project in 1994. The construction of the LGV Méditerranée, or LN5, began in 1996 and it opened in June 2001, with two additional new stations in Avignon and Aix-en-Provence. With the opening of LN5, the journey between Paris and Marseille was reduced to nearly 3 h by the fastest train.

Comparison

The rail markets in France and Italy did not follow the same trajectories. Italy liberalized its market earlier, but both countries created new HSLs to reduce the travel time between their largest cities. In what follows, we propose to compare the evolution of the rail market in these two countries to find similarities or differences for several variables that characterize the market.³

Travel time

As noted before, in both Italy and France, travel times changed significantly with the improvement of rail infrastructure. For the Italian case, to precisely determine travel time, our research conducted a re-enactment of frequencies based on the commercial timetables of the Italian incumbent Trenitalia found in the archives of the Italian railway company Fondazione FS Italiane. We performed the same analysis for travel time (see the next section). The periods observed in our study were between 2006 and 2010 and then 2016 and 2018. No detailed information on timetables between 2011 and 2015 was found. However, it was possible to reconstitute the timetables for 2016, 2017 and 2018, leading to a data set with two different periods, used to determine the impact of two different kinds of change on the composition of the railway market.

This reconstitution represents one of the main novelties of this research. Indeed, it includes detailed information on all the services using HSL before the opening of the last HST between Florence and Milan. This data set also allows us to observe the evolution of the frequencies and travel times of trains with different numbers of stops on their itineraries, considered an important characteristic to take into account in view to seeing the ‘big picture’ of the Italian HSR market.

Figure 1 shows a continuous decrease in travel time between Rome and Milan. Since 2016, travel time for services with more than three stops has been less than for non-stop services in 2009 (3 h 25 min vs 3 h 59 min). This could be the main reason for the considerable increase in supply of this kind of service offered by Trenitalia. In addition, the difference in travel time between non-stop services and one-stop services is very small (2 h 55 min vs 2 h 59 min).

Figure 1.

Travel time in minutes between Rome Termini and Milano-Centrale with Trenitalia. Source: Own elaboration based on Trenitalia timetables published on https://www.archiviofondazionefs.it/it/

As mentioned above, the change in travel time depends directly on the evolution of the infrastructure. Nonetheless, it may also depend on the rolling stock used by operators or the signalling system along the paths. Since 2010, the infrastructure has not changed but Trenitalia’s rolling stock has changed (Frecciarossa 1000). This is probably the main reason for the small change in travel time perceived between 2010 and 2016. Since the last improvements to the infrastructure in 2010, it can be expected that no reduction in travel time will occur unless certain improvements are made to the rolling stock or signalling.

The same analysis was undertaken for the French case using the open data of SNCF. As Figure 2 shows, reductions in travel time can be observed after the opening of each track section connecting Paris and Marseille, as with the Rome–Milan HSL. As stated above, the opening of the two main sections of LN1 reduced the journey time between Paris and Marseille from 6 h 40 min to 5 h 30 min and then to 4 h 40 min (−30%). Then, the reduction of travel time between Paris and Marseille after the opening of LN4 was about 20 min, passing from 4 h 40 min to 4 h 20 min, representing a 7% reduction of the previous travel time.

Figure 2.

Travel time in minutes between Paris and Marseille. Source: Own elaboration based on SNCF data published on https://data.sncf.com/explore/?sort=modified

Therefore, before the construction of LN5, travel time between Paris and Marseille was reduced from 6 h 40 min to 4 h 20 min, which represents a decrease of 35% due to the opening of two other HSLs. Once LN5 opened in 2001, the shorter journey between the two cities was reduced from 4 h 20 min to exactly 3 h, that is, a significant reduction of 31%. In total, since the opening of the first track section of LN1 in 1981 until the opening of LN5 in 2001, travel time between Paris and Marseille fell by 55%.

Figure 3 summarizes the findings on the evolution of travel times in France and Italy. Figure 3 shows similarities for the Rome–Milan and Paris–Marseille routes. It is very interesting to see that even if we look at different periods and two different countries, the evolution of journey times and the impact on journey time of the opening of both lines are very close. This similarity is also perceived in the number of steps undertaken between the beginning and the end of the construction of these HSLs.

Figure 3.

Travel time in minutes for Paris–Marseille and Rome–Milan. Source: Own elaboration.

Supply

A similar analysis can be performed for the supply of train operators – that is, the number of trains proposed to consumers. Regarding Italy, Figure 4 shows that there were no direct trains between Florence and Milan before the whole of the new line was opened in 2009. After 2009, non-stop services were opened to passengers, with 19 services in total (10 services from Rome to Milan and 9 services in the opposite direction). One year later, the number of non-stop services increased from 19 to 34 per day (17 in both directions). The number of non-stop services provided by Trenitalia between 2016 and 2018 does not seem to be significantly different from the offering in 2010. Indeed, they even appear to be less numerous in 2016 than in 2010: non-stop services increased from 28 (14 + 14) in 2016 to 33 (17 + 16) trains per day in 2018. These numbers are corroborated by the findings in other papers: according to Bergantino et al. (2015), the incumbent offered 36 and 38 services in 2012 and 2013, respectively. This suggests a decreasing number of non-stop services for Trenitalia between 2013 and 2016.

Figure 4.

Evolution of the number of services of Trenitalia HST per day by number of stops (only non-stop and one stop in this figure) between Rome and Milan. Source: Own elaboration based on Trenitalia timetables published on https://www.archiviofondazionefs.it/it/. HST: high-speed train.

Concerning the number of services with one stop (generally in Bologna or Florence), it can be observed that the number of journeys was not very significant. Indeed, there were no more than four services per day between 2006 and 2009, and these were stopped in 2009. In contrast, in the second period, it can be observed that one-stop services offered by Trenitalia numbered between 14 and 19 per day.

Services with three stops (generally two in Florence and one in Bologna) were always offered in significant number by Trenitalia. Indeed, the total number of this kind of service increased from an average of 27 in 2006–2009 to 37 in 2010 as a result of the opening of the two last tracks connecting Florence to Milan. In the second period, the average number of services fell to around 29 services per day. Finally, services with two stops were never offered by Trenitalia on this track.

The trend of total services provided between Rome and Milan shows an inflection point in 2009–2010 with the opening of the totality of the two last sections of the new line between Florence and Milan. Indeed, total HST services increased from 29 per day to 50 (+72%) in 2009 and 73 (+152%) in 2010.

This increase was mostly due to the evolution of the number of non-stop services that were introduced in 2009 and then in 2010 (19 and 37, respectively). In addition, it can be observed that one-stop services and services with more than three stops also increased significantly. Indeed, almost 50% of the overall evolution of services between 2006 and 2018 was due to the development of these two other services that were not sufficiently developed.

In brief, when the evolution of the whole service is observed, it can be noted that until 2010 its trend was significantly positive. Nonetheless, the market had a single operator, which reveals that before the entry of a new operator into the market, the increase of supply had already begun, placing the infrastructural changes at the core of the evolution. It can therefore be expected that this trend, accompanied by the poor railway market share on the intermodal market that this research will address, will continue.

The evolution of French supply between Paris and Marseille faced different strategies at different times (Figure 5). As stated above, the Paris–Marseille track benefited from the opening of LN1 and LN4. Concerning LN1, the total number of trains between both cities evolved considerably after the opening of the second part of the Paris–Lyon track. Indeed, it increased from 16 in 1982 to 24 one year later, reaching 30 in 1993 – a change of 87.5%. This increase seems to be the result of the incumbent’s strategy of expanding the number of services with two stops from two services per day to eight services per day. The other types of service that were better developed between 1983 and 1993 were those with three stops (from four per day to ten per day) and those with more than three stops (eight per day), which were not offered in 1983.⁴

Figure 5.

Evolution of the number of services of SNCF HST per day by number of stops (only non-stop and one stop in this figure) between Paris and Marseille. Source: Own elaboration based on SNCF timetables. HST: high-speed train.

Despite its positive evolution, the French HSR was not exempt from the generalized decline of the French rail sector and the success of the air transport model in the 1990s (Delaplace & Dobruszkes, 2015). In this context, the evolution of HSR frequencies after the opening of LN4 was different compared to the opening of LN1. Supply was readjusted by the incumbent, which resulted in a decrease of some of the different services offered. From the opening of LN4 until 2001, services with three stops and more than three stops had been stopped. In contrast, services with one stop were highly developed, as were non-stops trains, which had not been proposed before. The incumbent proposed 14 trains per day with one stop and four non-stop trains in 2000, versus four and zero trains in 1993. Finally, this trend was confirmed in 2001 and after the opening of LN5: the incumbent continued to develop non-stop services and two-stop services, finally reaching a total of 36 daily trains between 2001 and 2004. Only three kinds of service were preserved on the Paris–Marseille line: non-stop, one-stop and two-stop services.

Market share

Regarding the HSR market share and its evolution, our study bases its analysis on data from the estimations of FS Italiane used in a presentation that took place in March 2018.⁵ This presentation shows that the HST market increased from 2007 to 2012 (Figure 6). Indeed, the market share of HSTs between Rome and Milan evolved from 36% in 2007 to 55% in 2011 and to 66% in 2016. This evolution was due in particular to a fall in the number passengers transported by air between these two cities (Figure 7). Indeed, in 2007, 3,160,422 passengers travelled by plane to commute between Rome and Milan. In 2011, the number of passengers decreased to 2,197,020 (−30.5%) and continued to decrease in 2015 to 1,683,736 passengers (−23.4%).

Figure 6.

Market share evolution between Rome and Milan. Source: Estimates by FS Italiane on origin–destination: presentation from March 2018.

Figure 7.

Evolution of number of passengers travelling by plane between Rome and Milan. Source: Own elaboration based on Ente Nazionale per l’Aviazione Civile (ENAC) passenger data.

Figures 6 and 7 show the trend in market share before and after the entry of NTV. It can be seen that the increase in rail modal share was already very significant before the entry of the new operator, and it continued to grow after the liberalization (at a slower rate). These findings suggest that liberalization may contribute to the increase of market share, but growth in the rail market was taking place before the entry of NTV. The decrease in the air transport share corroborates this interpretation, showing a modal shift from air to rail since the opening of the new HSL between Rome and Milan. The closer the travel time is to 3 h, the larger the market share won by rail.

Concerning the French market, the reconstitution of the evolution of the HSR market share between Paris and Marseille was very difficult to do due to the lack of public and historic data for rail and air transport.⁶ However, a study conducted by Steer Davies Gleave for the European Commission in 2006 and other research studies like Fourniau (1989), Guyard and Chapulut 2004 and Leheis (2009) suggest that the rail market split between Paris and Marseille was around 45% in 2000 and 68% in 2004 – that is, before and after the opening of LN5.

This evolution is in line with a study by Widmer (2002), in which he noted that this market share had reached 66% in 2002 as shown in Figure 8. It is noteworthy that this same study estimated a rail market of 51% in 1984, which is in total accordance with the decline of the train before the opening of the HSR system and the growth of air transport.

Figure 8.

Market share (air–rail) evolution between Paris and Marseille. Source: Own elaboration based on multiple sources presented in the article.

Zembri (2010) noted an HSR market share of 83% in 2007, mostly induced by EasyJet’s decision to leave the market between Paris and Marseille. Indeed, EasyJet and Air France were the only airline companies that resisted the first years of competition with HST on this line. However, after 4 years of trying to consolidate its position in the market, EasyJet decided to strengthen its presence on other routes where HSR average travel time was longer than 3 h 30 min (Paris–Toulouse, Paris–Biarritz and Paris–Nice).

The impact of opening the HSR network is therefore notable: the market share evolved from 45% in 2000 to 83% in 2007, which represents an increase of 84% in 7 years. Furthermore, the opening of LN5 had a direct impact on the number of airline passengers.

As Figure 9 shows, the number of air passengers between Paris and Marseille has decreased considerably since the commissioning of the HSR. One year before the opening, almost 3 million people had used air transport to commute between the two cities. A year later, that number had decreased by 500,000 passengers (−16.7%). Four years later, the fall in the number of air passengers between the two cities continued, decreasing by −32.7% (2.05 million passengers); six years later, this number continued to drop until only 1.8 million passengers (−40%) were being transported by air. Finally, it seems that the number of air passengers travelling between Paris and Marseille levelled off 7 years after LN5 had been commissioned (between 1.5 and 1.6 million passengers).

Figure 9.

Evolution of number of passengers travelling between Paris and Marseille by air. Source: Own elaboration based on DGAC data.

We synthesize these findings in Figure 10, which compares market shares at five different moments: 1 year before the opening of the HSR lines (t – 1), and 1, 3, 5 and 6 years after the opening. For the Paris–Marseille route, t − 1 is 2000; for the Rome–Milan route, t − 1 is 2007. Comparing the evolution of market share in both cases, we can observe similar trends in the rail market share.

Figure 10.

Evolution of HST market share (air–rail) with respect to the year of the opening of HSR. Source: Own elaboration. HST: high-speed train; HSR: high-speed railway.

More precisely, this figure shows that 1 year before the opening of the HSR for Rome–Milan, the HSR market share was 41% versus 45% for Paris–Marseille. Nonetheless, 1 year after the opening of HSR lines, the market share grew, and 3 years after the opening of the HSR line, the HSR market share had reached 61% in the Italian case and 68% in the French case. Five years after the completion of the HSR network, the market shares for each line were very close: 69% for the Paris–Marseille line and 67% for the Rome–Milan line.

Finally, 6 years after opening, HSR market shares were, according to the data collected, 83% for the French line and 69% for the Italian line. An explanation for this difference may be the desertion by EasyJet, one of the two airline companies in the French market, which may have led air passengers to use HST services. However, Ryanair and EasyJet also decided to leave the Rome–Milan market, but at different moments. Indeed, Ryanair decided to leave the market in April 2014, while EasyJet decided to stop its services in October 2017, thus 8 years after the opening of the HSR service but just 2 years after the arrival of the company on this track. The effect of this departure is not included in Figure 10 though it may help to explain the difference between market trends.

We can also observe the evolution of air transport demand for both routes. Figure 11 shows similarities in the trends in both markets. The evolution of air passengers was studied for the period between 4 years before and 4 years after the commissioning of the HSR network. It is clear that the evolution is very similar, despite the fact that it took place at different times: t − 1 is 1997 for the Paris–Marseille line market and 2005 for Rome–Milan. In addition, it can be seen that the number of air passengers was very similar in the two markets before (a difference of fewer than 300,000 passengers at t − 1) and after the opening of the HSR lines.

Figure 11.

Evolution of air transport demand with respect to the year of opening of HSR t. Source: Own elaboration. HSR: high-speed railway.

Saturation

The findings seem to suggest that the two markets faced similar trends several years before and after the opening of new HSR lines, although these new infrastructures opened at different times and in different countries. However, nowadays the two markets are different.

On the one hand, a new operator entered the Italian market in 2012 and researchers suggest that the evolution of the last few years was due to this liberalization. On the other hand, the French market will remain closed until 2021, but this does not mean that the market has not continued to evolve, reaching a high level of maturity.

One of the indicators of the maturity of HSR markets is the residual capacity of the network. To analyse this point, it is important to underline another principal similarity in the two HSLs concerning infrastructure. Each line has to support high levels of traffic, but although most of the traffic is on the Rome–Milan and Paris–Lyon/Marseille routes, a considerable share of the traffic connects other cities and uses a portion of the same path. The use of a part of the path to get to and from other origins and destinations signifies that the availability of the whole route is limited.

The Paris–Marseille path has four main rail nodes between Paris and Marseille, while Rome–Milan has just one. These nodes represent inflows and outflows on the line, which slow speed and may limit the capacity of the network.

Today, the residual capacity of the ‘common’ portion of the Paris–Marseille path is very limited, especially during rush hours, indicating that it is close to saturation (Figure 12).⁷ From Paris to Lyon, the HSL is saturated, especially in the afternoon (departure from Paris from 16 h to 19 h), but the number of trains running during off-peak hours is considerably lower.

Figure 12.

Actual overload level of Paris-Lyon HSR corridor in 2018 (normal day from Paris to Lyon). HSR: high-speed railway.

Where the Rome–Milan line is concerned, Figure 13 shows that traffic is flatter. Indeed, traffic during peak hours is not very different from traffic during off-peak hours. Today, the maximum number of trains running on the ‘common’ portion is not higher than 10 trains per hour, and the total number of trains running daily is on average 124 in one direction and 245 in both directions.⁸ The difference between the Italian and French corridors is not very significant in terms of the number of daily trains. Nevertheless, the fact that traffic is flatter in a day in the Italian corridor than in the French one means that Italian lines still have the potential to achieve a 20% improvement in supplementary traffic during rush hours.

Figure 13.

Actual overload level of Rome–Bologna HSR corridor in 2018 (normal day from Rome to Bologna). Source: Own elaboration based on SNCF, Trenitalia and NTV operators timetables. HSR: high-speed railway.

Moreover, capacity on the Italian HSL will soon be higher due to the commissioning of ERTMS by the end of 2018. In France, this new signalling system is expected on the Paris–Lyon HSL but not before 2025–2030.

However, 3 years after the opening of LN5 (2004) and 1 year before the opening by the incumbent of the IDTGV service (which could be considered a new operator in the French market despite the fact that it belonged to the incumbent), the traffic on the HSL between Paris and Lyon was almost 243 trains per day, thus almost 121 one way.⁹

The Italian context is radically different. The Italian HSL is not close to saturation at the moment, which means that this was also the case before the entry of NTV. This big difference suggests that the Italian and French HSR markets do not have the same level of maturity.

Three years after the opening of LN5, the residual capacity of the line was as limited as today, while 3 years after the opening of the Rome–Milan line, there was significant residual capacity. Given the trend in each market, the residual capacity on the Italian line appears to be one of the main reasons for the success of the new Italian operator.

Conclusions

The liberalization of the European rail market in 2021 is leading researchers to conduct in-depth studies seeking to explain the impacts of this liberalization on markets that have already been opened and to estimate the possible impacts on markets that will soon be liberalized regardless of the model of competition – whether for the market or in the market.

With regard to markets already opened under the open access model, research has shown that, in most cases, liberalization was very limited, and it was perhaps impossible for a private operator to stay in a market with a decent market share. Italian liberalization is the only case where a new arrival has been able to win a considerable share of the HSR market and to remain in it for several years. Several studies have suggested that this success was due to the liberalization of the market itself: the arrival of the new operator and its business model drove an increase of the HSR market share.

In this article, we proposed another interpretation to understand the evolution of the Italian HSR market, by comparing it with the evolution of a French line not yet open to competition. The aim of this comparison was to know whether or not the rising trend of the Italian HSR is due to the arrival of a new competitor in the market.

After reviewing and comparing the evolution of both markets, this study stressed that their evolution in terms of supply, demand and market share is very close. Indeed, these variables can be explained by factors other than liberalization, since the trend in the French market is completely independent of the opening of the HSR market. Indeed, infrastructure seems to play a central role in explaining this evolution.

To our knowledge, the role of infrastructure in the HSR Italian market trend has not yet been studied. Indeed, most studies that have tried to understand the success of the HSR market have focused on the evolution of this market since 2008. As a result, research has not taken into consideration the impact on short-term travel time after and even before the opening of new HSLs. This was one of the main reasons for comparing the evolution of the Italian HSR infrastructure and its correlation with market trends with the French situation.

The results are in some cases very similar. The evolution of supply and of the number of passengers, the market share and even the number of air passengers and its evolution are almost identical. Even when taking into account the 8 years’ difference between the opening of both HSLs, the rail and air market trends were very close before and after the commissioning of new HSR infrastructures. Airlines seemed to anticipate the competition from HST and reacted by removing most of their services 1 year before the HSLs were commissioned.

The number of HSR services seems to follow the same evolution in both cases, with a continuous increase during the first 4 years and a stabilization after 5 years. The decreases in air services are almost identical in both cases, falling drastically during the first 2 years and then continuing to drop more gradually for 3 or 4 years afterwards. Finally, the market shares also evolved very similarly: the level of saturation and the market split seemed to reach stability.

Last but not least, the saturation of HSR markets on the Italian and French lines were very different 5 years after the opening of the HSR lines. Indeed, the French HSL was close to saturation shortly after the opening of LN5. In contrast, the Italian HSR had considerable capacity available after the opening of the HSL (and still has), which facilitated the entry of a new operator.

Overall, the results provided in this article suggest that infrastructure lies at the core of the success of the Italian and French HSR markets. It appears that HSR markets reach maturity 4 to 6 years after the commissioning of an HSL. This latter assertion is made on the basis of the level of saturation of each line and implies that where the French case and its future liberalization are concerned, regarding the Paris–Marseille line, the market cannot continue to evolve as rapidly as in Italy and that the arrival of a new operator may prove very difficult.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Carlos Augusto Olarte Bacares

Notes

References

Albalate

Bel

Fageda

(2015). Competition and cooperation between high-speed rail and air transportation services in Europe. Journal of Transport Geography, 42, 166–174.

Alexandersson

Hultén

(2008). The Swedish railway deregulation path. Review of Network Economics, De Gruyter, 7, 1–19.

Bergantino

A. S.

Capozza

(2015). One price for all? Price discrimination and market captivity: evidence from the Italian city-pair markets. Transportation Research A, 75, 231–244.

Bergantino

A. S.

Capozza

Capurso

(2015). The impact of open access on intra- and inter-modal rail competition. A national level analysis in Italy. Transport Policy, 39, 77–86.

Beria

. (2015). High-speed rail in Italy: Is it a story of success? Paper presented at the International Workshop on Advances in Research on Surface Transportation, University of Barcelona, 24 November 2015.

Beria

Grimaldi

. (2016). An ex post cost benefit analysis of Italian high speed train, five years later (Working paper). Italian Society of Transport Economics (SIET). Retrieved from http://www.sietitalia.org/wpsiet/Beria%20Grimaldi%20-%202016.pdf

Beria

Grimaldi

(2017). Reality and opportunities for on-track competition in HSR. In Albalate

Bel

(Eds.), Evaluating high-speed rail. Interdisciplinary perspectives. Vol. 7 (pp. 140–159) Abingdon, UK: Routledge.

Beria

Debernardi

Ferrara

(2017). Measuring the long-distance accessibility of Italian cities. Journal of Transport Geography, 62, 66–79.

Beria

Redondi

Malighetti

(2016). The effect of open access competition on average rail prices. The case of Milan-Ancona. Journal of Rail Transport Planning & Management, 6, 271–283.

10.

Bitzan

J. D.

(2003). Railroad costs and competition: The implications of introducing competition to railroad networks. Journal of Transport, Economics and Policy, 37, 201–225.

11.

Bonnafous

(1987). The regional impact of the TGV. Transportation, 14, 127–137.

12.

Capozza

(2016). The effect of rail travel time on airline fares: First evidence from the Italian passenger market. Economics of Transportation, 6, 18–24.

13.

Cascetta

Coppola

(2014). Competition on fast track: An analysis of the first competitive market for HSR services. Procedia—Social and Behavioral Science, 111, 176–185.

14.

Cascetta

Coppola

(2015). New high-speed rail lines and market competition: Short-term effects on services and demand in Italy. Transportation Research Record: Journal of the Transportation Research Board, 247, 8–15.

15.

Casullo

(2016). The efficiency impact of open access competition in rail markets: The case of domestic passenger services in Europe (ITF Discussion Paper 2016-07). Paris: International Transport Forum-OECD.

16.

Chamberlin

(1933). Theory of monopolistic competition. Cambridge, MA: Harvard University Press.

17.

Competition and Markets Authority (CMA). (2016). Competition in passenger rail services in Great Britain: A policy document. Competition and Markets Authority. Retrieved form https://assets.publishing.service.gov.uk/media/56ddc41aed915d037600000d/Competition_in_passenger_rail_services_in_Great_Britain.pdf

18.

Crozet

(2014). High speed rail performance in France: From appraisal methodologies to ex-post evaluations. In ITF/OECD (Eds.), The economics of investment in high speed rail. Roundtable Report 155 (pp. 73–105). Paris: OECD.

19.

Delaplace

Dobruszkes

(2015). From low-cost airlines to low-cost high-speed rail? The French case. Transport Policy, 38, 73–85.

20.

Desmaris

(2016). High Speed Rail Competition in Italy. A Major Railway Reform with a ‘Win-Win Game’? (OECD-ITF, Discussion paper No. 2016-11). Paper presented to the Working Group: Public Transport Organization, p. 24.

21.

Desmaris

Croccolo

Patuelli

(2016). The high speed rail competition in Italy: How are the regulatory design and practices concerned? 14th World Conference on Transport Research (WCTR) Shanghai, Shanghai, 10–15 July 2016.

22.

Finger

(2014). Governance of competition and performance in European railways: An analysis of five cases. Utilities Policy, 31, 278–288.

23.

Fourniau

J.-M.

(1989). La genèse du TGV Sud-Est. Innovation et adaptation à la concurrence. Culture Technique, 19, 85–94.

24.

Guyard

Chapulut

J.-N

. (2004). Multimodalité Avion – TGV. Direction générale de l’aviation civile et Direction des transports terrestres. Ministère de l’équipement, des transports, de l’aménagement du territoire, du tourisme et de la mer.

25.

Ivaldi

Vibes

(2005). Intermodal and intramodal competition in passenger rail transport (IDEI Working Paper no. 345). Toulouse, France: IDEI.

26.

Johnson

Nash

(2012). Competition and the provision of rail passenger services: A simulation exercise. Journal of Rail Transport Planning & Management, 2, 14–22.

27.

Leheis

. (2009). TGV Méditerranée – France – Profil du projet. Retrieved from halshs-00550187.

28.

Link

. (2016). Liberalisation of passenger rail services: Case study Germany. Centre on Regulation in Europe (CERRE). Retrieved from https://www.cerre.eu/sites/cerre/files/161206_CERRE_PassRailComp_CaseStudy_Germany.pdf

29.

Mohring

(1972). Optimization and scale economies in Urban bus transportation. American Economic Review, 62, 591–604.

30.

Nash

Nilsson

J.-E.

Link

(2013). Comparing three models for introduction of competition into railways. Journal of Transport Economics and Policy, 47, 191–206.

31.

Nilsson

J. E.

Pyddoke

Hulten

Alexadersson

(2013). The liberalization of railway passenger transport in Sweden. Journal of Transport Economics and Policy, 47, 307–312.

32.

Patuelli

(2015). High-speed rail: is competition in the market sustainable? An Italian case. International Research Society for Public Management Conference 2015, University of Birmingham, Birmingham, 30 March–1 April 2015.

33.

Preston

(2009). The case for high speed rail: A review of recent evidence. London, UK: RAC Foundation.

34.

Preston

(2014). The economics of investment in high speed rail: Summary and conclusions (OECD-ITF Discussion Paper 2013-30). Paris: International Transport Forum.

35.

Preston

Wardman

Whelan

(1999). An analysis of the potential for on-track competition in the British passenger rail industry. Journal of Transport Economics and Policy, 33, 77–94.

36.

Prussi

Lonza

. (2018). Passenger aviation and high speed rail: A comparison of emissions profiles on selected European routes. Journal of Advanced Transportation, 2018, Article ID 6205714, 10. doi: https://doi.org/10.1155/2018/6205714.

37.

Steer Davies Gleave. (2006). Air and rail competition and complementarity. Final report for DG Energy and Transport. Commission for the European Communities. Retrieved from https://ec.europa.eu/transport/modes/air/studies/internal_market_en

38.

Tomeš

Kvizda

Jandova

Rederer

(2016). Open access passenger rail competition in the Czech Republic. Transport Policy, 47, 203–211.

39.

Tomeš

Kvizda

Nigrin

Seidenglanz

(2014). Competition in the railway passenger market in the Czech Republic. Research in Transportation Economics, 48, 270–276.

40.

Vigren

. (2017). Swedish Passenger Railway: Entry in an open-access market and its effects on prices. Economics of Transportation, 11-12, 49–59.

41.

Widmer

J.-P.

(2002). High speed rail: partner or competitor? Paper supporting the presentation at the Airports Council International Conference Air Links, Salzburg, Austria, 5–7 March 2002.

42.

Zembri

(2010). New objectives of the French high-speed rail system within the framework of a highly centralized network: A substitute for the domestic air transport market? ERSA Congress, NECTAR Special Session on HSR, Jonköping (Suède), p. 17.