With the night train towards the sea

WITH THE NIGHT TRAIN

TOWARDS THE SEA

Is the night train an alternative to reach

European summer holiday destinations?

Dominik Odendahl

Master’s Thesis for the Spatial Planning program, specialization Urban and Regional mobility

Nijmegen School of Management Radboud Universiteit September 2020

I

WITH THE NIGHT TRAIN

TOWARDS THE SEA

Is the night train an alternative to reach

European summer holiday destinations?

Dominik Odendahl

Abstract

This research analyzes the vision of substituting inner-European touristic flights with railway services. This seems to be an interesting way to reduce tourism related emission. The paper should give answers towards questions as “how tourism might be distributed if there was a sufficient night train system and air traffic was less at-tractive” or “whether it is in general possible to substitute those flights by trains”. In order to find out how a distribution would look like if flights were less attractive, trip distribution will be modelled using an adapted gravity model. This model will respect both classical trip distribution and tourism modeling criteria. In a follow-ing step, promisfollow-ing traffic relations will be analyzed in depth. This analysis then includes country- and corridor specific questions. For policy makers, these findings will be translated into recommenda-tions on how to improve European rail services for tourism pur-poses.

Masters’ Thesis

Spatial Planning

Urban & Regional Mobility August 2020

Keywords

Radboud

Universiteit

:

II

SUMMARY

Tourism is currently mostly related to holiday trips made with airplanes. This re-search shows options to promote another way of travelling, making usage of night trains. But currently, the infrastructure differs a lot within and between the regions, making some regions more accessible by rail and others less, even if the flying distance might be the same. After reflecting that tourists might change their travel destination choices because of increased travel times, I decided to model the attractivity of different destinations for different origins.

An appropriate way to do this is an adapted gravity model which models the attractivity between tourist origins and holiday destinations. This requires pa-rameters such as travel time, but also needs to consider how tourists make their choice on holiday destinations.

According to the model, routes from London and Paris to Barcelona, as well as the route from Munich to Split seem to have a high attractivity. After the most attractive routes are identified, the optimal night train, used within the model, is compared to the current state. It shows that the identified routes are currently already partly served by passenger night trains. The comparison be-tween the ideal state and reality helps to identify current disadvantages in routing and governance aspects. Some of those seem to be based on border barriers. Others might be improved by stronger cooperation.

PREFACE

After years of decline in night train traffic and discussions about its’ profitability, recently, night trains seemed to become more popular. This seems to be re-lated to the societal discourses about climate change. After trying to travel from the North to the South of Europe without planes, I wanted to find out whether there are options to make journeys like these more comfortable, easy, and less expensive. My hope is that more competitive alternatives to air travel could help to decrease travel-related emissions.

Night trains seem to be an ideal solution though the current service offer shows that it is not easy to implement. Questions raised what exactly the obstacles were and how the current barriers could be reduced. These considerations lead to the following research.

III

TABLE OF CONTENT

1 Introduction to the research ... 1

1.1 Research problem statement ... 2

1.2 Research aim and research questions ... 3

1.3 Scientific and societal relevance of the proposed research ... 3

2 Literature review and theoretical framework... 4

2.1 Tourism and destination choice ... 4

2.2 Characteristics of international (Night) Rail passenger traffic ... 8

Infrastructure capacity ... 8

Travel Time ... 9

IPRS Communication and distribution channels ... 9

Interoperability ... 10

Infrastructure Charges and Taxation ... 11

Prices ... 12

The Passenger Night Train ... 16

2.3 Business Models ... 24

Governmental Intervention ... 24

Open access Models ... 28

2.4 Critical review of the policy context ... 32

2.5 Brief introduction to the gravity model ... 33

3 Methodology ... 37

3.1 Research strategy ... 37

3.2 Research methods, data collection, and analysis ... 38

3.3 Validity and reliability of the research ... 41

4 Case study choices ... 42

IV

4.2 Origin regions statistics ... 44

4.3 Routing ... 46

Travel Times ... 47

5 Model application ... 50

5.1 General shortcomings ... 52

5.2 Case study ... 52

6 Corridor London – Paris – Barcelona ... 53

6.1 Current Routing ... 53

6.2 Options for Route improvements ... 53

6.3 Governance options ... 59

7 Corridor Munich – Split ... 61

7.1 Current Routing ... 61

7.2 Options for Route improvements ... 61

7.3 Governance options ... 63

8 Discussion ... 63

9 Reflexion and Conclusion ... 65

10 References ... 67

11 Appendix ... 71

11.1 coastal regions with highest amount of accommodations ... 71

11.2 NUTS 2 regions with highest GDP ... 72

11.3 Trip Distribution Model ... 72

11.4 Trip Distribution Model (Rijeka case) ... 74

V

FIGURES

Figure 1: CO2 emissions of flights are steadily increasing again since 2013 (EASA, EEA, & EUROCONTROL, 2019, p. 23) ... 1 Figure 2: Sleeper to Belgrade, own work, photo taken in Villach, Austria in Summer 2012 ... 2 Figure 3: Fare revenue per passenger-kilometer (2012, excluding Luxembourg) (Cartmell, 2016)p 14 ... 13 Figure 4: Interurban fares over 300 kilometers: peak single (November 2015) (Cartmell, 2016, p. 63) ... 14 Figure 5: Interurban fares over 300 kilometers: off-peak return (Cartmell, 2016, p. 63) ... 14 Figure 6: International fares: peak single (November 2015) (Cartmell, 2016, p. 68) ... 15 Figure 7: International fares: off-peak return (November 2015) (Cartmell, 2016, p. 69) ... 16 Figure 8: Long-standing and new air routes (Dobruszkes, 2011, p. 873). ... 19 Figure 9: Number of passengers in the London–Paris/Brussels passenger market (Behrens & Pels, 2012, p. 280) ... 20 Figure 10: Rail and air costs: international trips (lowest observed fare) (November 2015) (Cartmell, 2016, p. 97) ... 21 Figure 11: Rail and air average speeds: international trips (lowest observed fare) (November 2015) (Cartmell, 2016, p. 97) ... 21 Figure 12: Rail and car costs: international trips (November 2015) (Cartmell, 2016, p. 86) ... 23 Figure 13: Classification of social and market services (excluding discounted tariffs) (Beria et al., 2012, p. 117) ... 25 Figure 14: Regulatory strategy for social services. (Beria et al., 2012, p. 118) .. 26

VI

Figure 16: Characteristics of the business models for IPRS (Arx et al., 2018, p. 333)

... 28

Figure 17: Relationship between key influencing factors and business models. (Arx et al., 2018, p. 333) ... 32

Figure 18: Independent variables used in the basic gravity model of Grosche et al. (Grosche, Rothlauf, & Heinzl, 2007, p. 178) ... 34

Figure 19: Fastest routes from each Origin to each Destination (own work, based on OpenStreetMap, n.d.) ... 47

Figure 20: Corridor 1, Routing option 1 (own work, based on OpenStreetMap, n.d.) ... 54

Figure 21: Corridor 1, Routing option 2a (own work, based on OpenStreetMap, n.d.) ... 56

Figure 22: Routing Options Corridor 2 ... 62

Figure 23: Rail speed around Ljubljana, Slovenia (OpenStreetMap, n.D.) ... 76

Figure 24: Rail speed around Maribor, Slovenia (OpenStreetMap, n.D.) ... 77

Figure 25: Rail speed around Herzogenrath, Germany (OpenStreetMap, n.D.) ... 78

Figure 26: Rail speed around Willingen (Upland), Germany (OpenStreetMap, n.D.) ... 78

Figure 27: Rail speed around Murau, Austria (OpenStreetMap, n.D.) ... 79

Figure 28: Rail speed around Sauves d'Auvergne, France (OpenStreetMap, n.D.) ... 79

TABLES

Table 2: Origin cells with the biggest cities sorted by the height of GDP (own work, data retrieved from Eurostat, 2020) ... 42

Table 3: Number of bed places in destination regions ... 43

Table 4: Population and age distribution in the origin regions ... 45

Table 5: Share of people younger than 65 and time-sensitivity ... 45

Table 6: Income in Purchasing Power Standard (PPS) ... 46

Table 7: Fastest travel times ... 48

VII

Table 9: Travel time, reduced by nine hours of sleep ... 49 Table 10: Relative travel demand (D/x) ... 50 Table 11: Relative share of travel demand of each relation compared to the total travel volume of the model (rounded) ... 51 Table 12: Route attractivity in the Rijeka case (rounded numbers) ... 62

VIII

ABBREVIATIONS

CO2 – Carbon Dioxide

DB – Deutsche Bahn (German national rail operator) ETCS – European Train Control System

EU – European Union

GDP – Gross Domestic Product GRP – Gross Regional Product

HŽ – Hrvatske željeznice (Croatian national rail operator) ICT – Information and Communications Technology IPRS – International Passenger Rail Service

IPRT – International Passenger Rail Traffic

NUTS – Nomenclature des unités territoriales statistiques (Nomenclature of Terri-torial Units for Statistics)

ÖBB – Österreichische Bundesbahnen (Austrian national rail operator) PPS – Purchasing Power Standard

PSO – Public Service Obligation

SNCF – Société nationale des chemins de fer français (French national rail op-erator)

1

1 INTRODUCTION TO THE RESEARCH

During the last years, the role of traffic emissions for climate change has been discussed within many studies (cf. e.g. Holz-Rau & Scheiner, 2019; Dobruszkes, 2011; EASA, EEA & EUROCONTROL, 2019). As a result, spatial planning tried to reduce emissions of regional and local mobility activities. Many cities are dis-cussing strategies towards emission-free urban mobility followed by investments in e.g. public transport and cycling infrastructure (cf. Eindhoven, Vienna). A big hope also lies in positive side-effects of urbanization processes which seem to cut down commuting distances compared to rural or suburban residential zones as people live closer to their everyday destinations. Also, developments in Information and Communications Technology (ICT) infrastructure promise an increase in teleworking and, thus, a decrease in commuting.

Against expectations, traffic emissions are still growing. According to Holz-Rau et al. (2019, p. 128) this is partly based on increasing travel distances. Especially urban citizens travel more often on long distances for leisure reasons. Many of those journeys are made by flying (Holz-Rau & Scheiner, 2019). As shown in fig-ure 1, the flight-related CO2 emissions raised a lot within the last years and have been expected to rise even more in the future. Due to the current CoV19 de-velopments, the amount of flights might also decrease which also influences the amount of flight-related CO2 emissions. This might be the result of travel restrictions, general fear of passengers, or the economic development.

Figure 1: CO2 emissions of flights are steadily increasing again since 2013 (EASA, EEA, & EUROCONTROL, 2019, p. 23)

For many people, it seems like the decision of going on holiday is always related to flying. Currently, many regional, and international travel industry economies depend on tourists. Raising air taxes or even forbidding inner-European flights

2

would harm local and international economies seriously. To meet both interests of the environmental protection and tourism industry, other means of travel are needed. Possible means to transport people over long distances would be pri-vate cars, coaches, trains, and ferries.

1.1 RESEARCH PROBLEM STATEMENT

The classical summer holiday destinations of middle- and north Europeans are mostly too far away for traveling dur-ing the daytime by trains. For those purposes, night trains seem to be an adequate travel mode to bring tourists from central Europe towards holiday destinations on the coast. Trains can reach high maximum travel speeds, have low operation costs, and low negative externalities such as air pollution compared to other modes of public transport (D’Acierno, Gallo, Montella, & Placido, 2013, p. 76). The substitution of holiday flights by trains seems to be a possibility to find a balance in enabling people to travel and the protection of the environment.

While there seems to be manifold research on high-speed trains, related traffic demand, and how the introduction of those influenced demand on concurring aviation services, the influence of night trains on travel mode choice has not been researched so far.

For vacation destination choice, many models have been developed. But in those models, the airplane played a big role as traffic mode (e.g. Grosche et al, 2007) and it is unlikely that tourists who decide for a destination which is four flight hours away also, if the substitution by rail would take them 24 hours or longer. Accordingly, it remains unclear how the development would look like if airfares raised (which could be results of increasing fuel prices orCO2 taxation) or the fares were restricted by law due to the environmental damage of air traffic. The role of trains, especially night trains, as travel mode towards holiday destinations is also rarely researched so far.

Figure 2: Sleeper to Belgrade, own work, photo taken in Villach, Austria in Summer 2012

3

1.2 RESEARCH AIM AND RESEARCH QUESTIONS

The above-discussed research problems result in two research aims. The first is to discover in which way trains can substitute air travel towards holiday desti-nations. This includes possible differences in the distribution of travelers to holi-day destinations in case of direct (night) trains from the region of origin towards the holiday destination region or in case of islands, the connecting train to a ferry. The second research aim addresses the possibility of introducing these services with help of the current infrastructure, and how potentials (such as ad-ditional stops or through coaches) could be used.

Out of these research aims; the following research questions can be devel-oped:

1) Could a better European (Night-)Train network substitute holiday flights? a. What are factors of modal choice and holiday destination choice? b. How can the attractivity of different routes for holiday passengers

be measured?

2) How can European train routes with high potential for tourism be opti-mized?

a. How could optimizations in terms of routing look like?

b. Which optimization in terms of institutional changes are needed? The research can generally be divided into two blocks which are represented by the two main research questions 1) and 2) the sub-questions 1)a.- b. and 2) a. – b. help to answer the two main research questions. It is expected that an-swering the sub-questions helps to cover the related research aims.

1.3 SCIENTIFIC AND SOCIETAL RELEVANCE OF THE PROPOSED

RESEARCH

The scientific relevance of this research is to answer the question of whether (European) night trains could substitute air travel in tourism. Also, in case it could, how this could be reached, and which steps need to be done. This in-cludes model approaches towards a night-train-based holiday distribution in Europe. This research aims to start a discourse about implementing new railway services, especially night-train services. If those can substitute flights, traffic emissions can be reduced and benefits for the environment can be archived. The main societal value is to propose possible solutions for the travel industry to become more environmentally friendly and help to slower down the climate

4

change. But related to that might also be positive side-effects with societal and economical value. For many European regions which are depending on tour-ism as a big economic factor, additional trains would also mean winning an environmentally friendly connection to economically powerful regions. This could also be part of a branding strategy of tourism destinations as ecologically friendly. Additional international train services can also help to reach European cohesion. Another benefit is the independence of tourism regions from air traf-fic. Even if aviation struggles because of high prices or oil scarcity, the regions will remain accessible for tourists. Another aspect is the possibility to keep mini-mum distances in trains with compartments during pandemics (e.g. CoV19). This enables families or small groups of travelers to minimize the risks of infections compared to flying. Another aspect is the current wish of many people to go to close-by holiday destinations. These chances have been currently used by a private train operator who decided to offer night trains between the alps and the German island “Sylt”, promoting the comparably safe option to travel by night train (Balser, 2020)

2 LITERATURE REVIEW AND THEORETICAL FRAMEWORK

To make assumptions about people using trains to reach their holiday destina-tions, literature researches in the fields of travel behavior, choices of holiday destinations, and travel mode, as well as the basic models such as the gravity model, need to be done. Another research needs to focus on international Passenger railway services (IPRS) and night trains in particular. This also includes possible business models. Also, related policies such as infrastructure planning, taxation, and international coworking need to be examined within this section.

2.1 TOURISM AND DESTINATION CHOICE

To find out how people decide on going to a specific place for their holiday, general research about tourism and destination choice needs to be done. Within the academic literature, the importance of tourism for most regions’ growth and economic prosperity is mentioned. More than 11 % of the EU’s Gross Regional Product (GRP) is based on touristic activities (LaMondia, Snell, & Bhat, 2010, p. 1). LaMondia et al. (2010) also stress the meaning of tourism for infra-structure and urban development (LaMondia et al., 2010, p. 1).

There are several approaches to forecast tourism demand and related modal choice. Researchers acknowledge that destination and traffic mode choice are highly inter-related (LaMondia et al., 2010, p. 2). But so far, studies mostly focused on single destinations or single origins rather than multiple approaches

5

(LaMondia et al., 2010, p. 4) or separate leisure activities (LaMondia et al., 2010, p. 1). To find night train routes with promising demand, these models can not be used, as different origins and destinations can not be expressed within one model.

Decisions about vacation trips combine choices of destination,

accommoda-tion, duraaccommoda-tion, travel party, and transport mode (Grigolon, Kemperman, & Tim-mermans, 2012, p. 1175). Concerning the destination choice for European hol-iday travelers, LaMondia et al. (2010) discovered several factors influencing the decision based on the Eurobarometer survey in 1997. The different factors have been divided into personal, destination, and trip characteristics. “Age, educa-tion, household composieduca-tion, income, and place of residence” are typical per-sonal characteristics (LaMondia et al., 2010, p. 2). Whereas climate, the variety of activities, the coastline, the number and quality of accommodations, the degree of development, the area size of the destination, the Gross National Product, the exchange rates, and the transport, food and accommodation-related costs make part of destination characteristics (LaMondia et al., 2010, p. 2). Examining the Eurobarometer survey, LaMondia et al (2010) found that income does not affect the destination choice so that it is to assume that within the region of choice, cheaper accommodations are chosen. Trip characteris-tics are defined by travel times and distances, costs, and vacation purposes (LaMondia et al., 2010, p. 2). To compare different holiday destinations, they need to be comparable in climate and variety of activities. Also, the number of accommodations need to be kept in mind when comparing holiday desti-nations. It should also be kept in mind to compare regions along the coastline. Traditional standard factors are not enough to model tourism demand as the competitive tourism market made tourists more selective. Researchers now try to gain more insight into traveler preferences and motivations. (European Travel Commission, 2006; In: LaMondia et al., 2010, p. 2). More personal data such as trip motivation and purpose, or prior expectations and experiences can give better insights into personal travel preferences (LaMondia et al., 2010, p. 2). Es-pecially prior tourism experiences seem to be important to analyze holiday travel preferences. Commonly, “more experienced travelers (…) become ex-tremely loyal to certain destinations” (LaMondia et al., 2010, p. 3). As up-to-date in depth analyzes of those factors are not available for the scale, it is im-portant to reflect current touristic demand for travel behavior forecasts.

Within the case study of six European countries in 1997, LaMondia et al. (2010) revealed that travelers prefer to go on holiday within their home countries. Other than that, Spain seemed to attract most European travelers the most (LaMondia et al., 2010, p. 9). Still, variations across different countries could be

6

observed. While Germans and UK citizens are more likely to spend their holidays in another country, Greek and Spanish citizens are more likely to spend their vacations within their own countries (LaMondia et al., 2010, p. 8). It shows that the favorite holiday destination for Germans in 1997 were the countries Spain and Italy while the UK was the least visited country; the French are equally often visiting Germany, Greece, and Spain but travel less often to Italy and the UK for holiday purposes; the Greeks visit all different countries equal often (LaMondia et al., 2010, pp. 8–9). The outcomes of these studies are only helpful to a limited extent as it is outdated and not all European origins and destinations are re-flected. Also, the national level makes it impossible to see regional differences in holiday preferences.

When having a deeper look at individual travel characteristics, it shows that individuals traveling alone or with small children were more likely to choose close by destinations. LaMondia et al. (2010) explain these results with the hy-pothesis that “adults traveling alone would want to get to their destinations quickly to begin their vacation pursuits, while those traveling with young chil-dren may not want to travel for extended periods because of the biological needs of young children and the inherent difficulty in keeping young children occupied when also constrained in physical movement” (LaMondia et al., 2010, p. 9). Due to the increase of air traffic and high-speed train connections, the actuality of this hypothesis, especially for individual traveling adults might be doubted. Greater distances can be overcome in shorter periods. Anyhow, the higher time-sensitivity for adults traveling alone or with young children might still be relevant and the travel time will be included within the model.

The results of the study also show that highly educated travelers do travel to countries which have large cities. The researchers assume that higher edu-cated tourists might be attracted to “the rich culture and heritage associated with large (and typically older) cities” (LaMondia et al., 2010, p. 9). Even though this conclusion seems to be reasonable, I doubt this finding as to the research data were on the national state level and all of the six case study countries (France, Germany, Greece, Italy, Spain, and the United Kingdom) feature both older and larger cities with rich cultural heritage. Larger households tend to travel shorter distances which, according to the authors, might be due to a cost reduction or the easier coordination between the higher number of par-ticipating households (LaMondia et al., 2010, p. 9). If the regions within the model also offer large (and older) cities, all touristic regions will be comparable in this characteristic.

Within the sample, it showed that the choice of holiday destination was not affected by the persons’ income (LaMondia et al., 2010, p. 10). It still might

7

affect the number of holiday trips or the decision whether at all the person goes on holiday. This question was not answered within this study as people who are not going on holiday were not considered at all. Accordingly, it would be help-ful to analyze origin regions with higher Gross Domestic Product (GDP) per cap-ita, as those might generate more holiday trips.

General outcomes of the case study from 1997 concerning the modal choice are that most people (64,9%) chose their personal vehicle to reach their desti-nation. 19,5% used surface public transport such as ferries, coaches, and trains; 15,6% of the trips were made by air transport (LaMondia et al., 2010, p. 7). The modal choice preferences are different for different destinations, though. Whereas the distribution of the three different modes was equal, Spain was preferably traveled to by airplane. France and Germany were mostly visited using a personal vehicle. The same counts for the United Kingdom and Italy but with a weaker difference compared to other modes (LaMondia et al., 2010, p. 8). Also, the travel group size is correlated to the modal choice. People

trav-eling with others “have a strong preference to travel by a personal vehicle,

suggesting that the travel to the vacation destination itself is viewed as part of the overall vacation experience when traveling with others” (LaMondia et al., 2010, p. 9). Even though the income does not influence the holiday destination according to the researchers, lower-income groups prefer surface modes com-pared to travelers with higher incomes who prefer to travel by airplane (La-Mondia et al., 2010, p. 10). These findings differ from those of Grigolon et al. (2011), who discovered that students let themselves influence by the destina-tions low-cost operators offer as “they are mostly adventurous and quite flexible tourists, even if they look for variety when seeking for new places they have never visited before” (Grigolon et al., 2012, p. 1182). In general, it needs to be stated that the aviation market developed a lot within the last 2 decades. Es-pecially the presence and growth of low-cost carriers such as Ryanair, Easyjet, and others probably contributed to a radical change within the modal choice. It is assumed that especially price-sensitive individuals can now use the aircraft to reach their holiday destinations. This seems to be the case for students who are “a major market segment for low-fare airlines” (Grigolon et al., 2012, p. 1174). Another point might be the induction of more air travel demand as the offers of low-cost carriers start at very low prices and are said to lure espe-cially students, who are mostly price-sensitive but less time-sensitive (Castillo-Manzano, López-Valpuesta, & González-Laxe, 2011, p. 1093). These findings seem to be in favor for the usage of night trains which offer several comfort classes, enabling people to travel on their own, together or in low-cost seating wagons.

8

2.2 CHARACTERISTICS OF INTERNATIONAL (NIGHT) RAIL

PAS-SENGER TRAFFIC

As this research focuses on trains as transport mode towards tourist destinations, different particularities of International (Night) Rail passenger Traffic need to be examined. International Rail Passenger Traffic (IPRT) combines the already challenging characteristics of Rail Passenger Traffic with the challenges of bor-ders and related different institutional, technical, and financial systems. This comes along with several barriers for the involved actors. Compared to the total Rail passenger traffic, in 2016 only 6% was international (European Com-mission, 2019, p. 3). It also shows that domestic railway services have a higher demand than international services (Arx, Thao, Wegelin, Maarfield, & Frölicher, 2018, pp. 329–330).

The low share of international rail passenger services can be partly explained by the fact that “the provision of long-distance rail services is mainly deter-mined by the government (not by market mechanisms)” (Finger & Rosa, 2012 In: Arx et al., 2018, p. 327). Examples for current and past governmental deci-sions influencing the state of international rail passenger services such as infra-structure, charges, and taxation are explained within the following sub-chap-ters.

INFRASTRUCTURE CAPACITY

The availability of railway infrastructure is a basic need to provide railway ser-vice. IPRT asks for adequate infrastructure and service facilities in all countries (Independent Regulators' Group - Rail, 2015, p. 4). Especially the railway net-works in north-west Europe are used intensively (European Commission, 2019, p. 7). Congested tracks are mostly found within the UK, Germany, Italy, and Romania (European Commission, 2019, p. 8). This shows that infrastructure on some relations is already limiting additional railway passenger services in Eu-rope (EuEu-ropean Commission, 2019, p. 7).

Infrastructure limitations can also prevent higher effectiveness (European Com-mission, 2019, p. 8). Resulting travel time extensions can also negatively affect the profitability and competitiveness of International Railway Passenger Ser-vices (IPRS). As cross-border railway infrastructure is often less developed, and missing cross-border high-speed lines occur, the competition to other travel modes is further affected (Independent Regulators' Group - Rail, 2015, p. 3). In case a decision between different railway services needs to be made, EU states mostly prioritize PSO. Also, international passenger and freight services benefit from a higher prioritization (European Commission, 2019, p. 8).

9

TRAVEL TIME

A big factor for the competitiveness of railway passenger traffic is the travel time (Arx et al., 2018, p. 331). This is in line with the findings of Maurer, Burgess, Hilferink, Kroes, & Whiteing (2010) who describe a lack of profitability for most border crossing railway passenger lines except for most high-speed rail services (Maurer, Burgess, Hilferink, Kroes, & Whiteing, 2010, p. 10). This might be partly explained by the fact that high-speed infrastructure is mostly built for promising corridors, but also by the time competitiveness with other modes on the same relation.

Travel time extensions on international railway relations can occur due to dif-ferences in infrastructure or formal requirements such as border controls, change of locomotives or staff, and poorly coordinated train connections (cf. (Arx et al., 2018, p. 331). Also “fears of unreliability on the part of the railway companies are factors contributing to” longer travel times (Maurer et al., 2010, p. 9). Arx et al. (2018) name the example of the Swiss national operator SBB who separated connections and introduced longer waiting times because of fre-quent delays of international passenger rail services. But both, too short and too long waiting times are unattractive for passengers (Arx et al., 2018, p. 331).

IPRS COMMUNICATION AND DISTRIBUTION CHANNELS

For people who start planning to travel, related communication and distribu-tion channels are the main source of informadistribu-tion. Those channels help to de-cide whether or not to make a trip, where to go to, and which mode or oper-ator to use.

Arx et al. (2018) discovered that channels for IPRS mostly remain invisible to po-tential customers and comparing prices remains difficult. While airline fares and schedules can be found on various platforms, international railway information is mostly not present and IPRT fares are therefore perceived as pricier than air-line fares (Arx et al., 2018, p. 331). Another limitation of the distribution channels is the, compared to airfares, limited pre-sale time (Arx et al., 2018, p. 331). Mi-chael Cramer, former chairperson of the European Parliament’s Committee on Transport and Tourism mentions that connecting tickets often need to be bought separately as national railways considered their price and schedule in-formation as “business secrets“ (Bellinghausen, 2019, p. 81).

Own tryouts showed interesting results. Using the inter-modal channel

https://www.fromatob.com/, both coach and airfares can be displayed for nearly every relation connecting major European cities. Railway fares for inter-national relations remain the exception (e.g. Düsseldorf-Venice). When looking

10

at national railway operators’ websites, fares for the relation Düsseldorf-Barce-lona can be neither found on the German nor Spanish national railway sale channel, but on the French one. Of course, some international train tickets can be bought in the booking offices of the national incumbents. But whilst flight tickets can be compared and bought through several online distribution chan-nels, the option of going to booking offices appears outdated. These examples show that a common distribution platform for international passenger rail ser-vices is not available.

INTEROPERABILITY

An IPRS specific requirement is the interoperability. Within Europe, several track widths, energy, safety, and signaling systems hinder cross-border train services. But also different laws and safety requirements as well as language barriers hin-der interoperability. "(A)s soon as a train crosses a borhin-der, figuratively 'every-thing' changes” (Arx et al., 2018, p. 332).

The lack of harmonization of technical and operational rules for accessing in-frastructure is seen as the biggest barrier for additional IPRS services (Arx et al., 2018, p. 327). This comes with many technical differences which ask for expen-sive solutions (Maurer et al., 2010, p. 9).

European Union currently tries to make interoperability a main priority. But standardization of administration, safety standards, and infrastructure proceed slowly. In fact, “(m)any different national technical and safety rules overlap and/or conflict with EU legislation” (European Commission, 2010; OECD, 2013; Perennes, 2017; In: Arx et al., 2018, p. 327). As a result, the flexibility of Railway services is restricted. A train traveling on one route can often not be used on another route (Independent Regulators' Group - Rail, 2015, pp. 3–4)

European Union also called for a common standard in signaling systems. But the standardization towards the European Train Control System (ETCS) takes a long time (Arx et al., 2018, p. 328; Feuerstein, Busacker & Xu, 2018, p. 307). And due to the little significance of international railway services, technical stand-ardization often does not seem to be worth the costs (Maurer et al., 2010, p. 9; Independent Regulators’ Group – Rail, 2015, p. 4). Michael Cramer even as-sumes that national operators block the installation of ETCS to hinder competi-tion (Bellinghausen, 2019, p. 81).

To offer IPRS, trains need to be equipped with technological systems that ena-ble services in multiple countries (Independent Regulators' Group - Rail, 2015, pp. 3–4). These are followingly called multi-system trains. Compared to the past, when mostly standardized carriages where used and locomotives were

11

changed at the borders, nowadays customized multiple-units are used thor-ough Europe. Specifications of those multiple units for multiple countries usually come with additional costs (Arx et al., 2018, pp. 327-328; Maarfield, 2018, p. 33). Big difficulties with adopting rolling stock come with the different development strategies of the different countries’ infrastructure. Contrariwise strategies might make it difficult to adapt rolling stock for use in several countries (Arx et al., 2018, p. 332). It is to add that the approval of multi-system trains is more com-plex than comparable trains for national use only. Even the European Train Control System (ETCS) comes with national specifications that are sometimes incompatible with other countries’ systems or which integration comes with high prices (Arx et al., 2018, p. 328).

Another problem is the difficult process of vehicle authorization and safety cer-tification (Arx et al., 2018, p. 332). The Fourth Railway Package includes the shift of those tasks to the ERA which is regarded as a possible facilitator for IPRS (In-dependent Regulators’ Group – Rail, 2015, p. 5; Feuerstein et al., 2018, p. 307). Another option is the change of locomotives at borders, this comes with addi-tional staffing requirements and does not go along with the trend of using mul-tiple-unit trains instead of locomotive-hauled trains (Arx et al., 2018, p. 334). But not only the differences in signaling, timetabling, security, and regulatory compliance are hindering IPRS. Also, the personnel of the trains need to be trained for both countries. While the operation of other modes only requires the knowledge of one language (e.g. English for the aviation), language require-ments “differ from one country to another” (Independent Regulators' Group - Rail, 2015, p. 4).

All these factors demonstrate how complex and cost-intensive international passenger rail services are. But these costs are not the only challenge IPRS cur-rently faces.

INFRASTRUCTURE CHARGES AND TAXATION

Infrastructure charges are, next to costs for energy, staff, rolling stock, and com-munication and distribution channels part of the fixed costs of passenger rail-way enterprises. On the other side, the infrastructure access charging schemes enable infrastructure managers to recuperate costs for building and maintain-ing the infrastructure. Charges can also have a steermaintain-ing function by givmaintain-ing dis-counts to less damaging trains or for the usage of less frequented routes or time-slots (Schroten, Scholten, van Wijngaarden, & van Essen, 2019). In general, these charges are perceived as one of the main barriers to passenger railway

12

competition because of their height and structure (Independent Regulators' Group - Rail, 2015, p. 4).

For IPRS, especially the immense price of high-speed and night train infrastruc-ture charges are seen as a constraint (Arx et al., 2018, pp. 331–332). In 2016, UK high-speed rail infrastructure access charges were as high as 19€ per train kilo-meter while the majority of EU countries only charged 3€ per train kilokilo-meter for conventional long-distance passenger trains (European Commission, 2019, p. 6). Additionally, changes in those charges are a high risk for IPRS operators (Arx et al., 2018, p. 332). This could be the case if operators decide to invest in multi-system trains and infrastructure charges in one of the countries increases significantly. According to Maarfield (2018), there are cases in which charge international services higher than domestic services. He, thus, pleads for EU or national regulators’ intervention against IPRS discrimination (Maarfield, 2018, p. 33).

In the case of night train infrastructure access charges, a Steer Davies Gleave study for the European Union argues that “night trains might appear more via-ble if infrastructure charges were reduced to the costs directly incurred as a result of operating them” which they calculated being less 2€ per train-kilome-ter (Steer Davies Gleave supported by TRASPOL - Politecnico di Milano, 2017, p. 104).

It is important to not only focus on the amount of infrastructure access charges. In fact, those need to be examined together with taxation schemes. Also, tax-ation contributes to a large part of production costs (Arx et al., 2018, p. 332). The exemption of international air travel and other differences in fee-charging “creates distortions in intermodal competition, usually forcing rail transport into a weaker position. This might change in the future since the EU is working on establishing a level playing field for different modes of transport” (European Commission, 2018; European Parliament, 2017 In: Arx et al. 2018, p. 332).

According to the findings, it is necessary to introduce fair and reliable taxation and access charge schemes to enable IPRS. Those should ideally also reflect charges to competing travel modes (cf. Arx et al., 2018; Independent Regula-tors’ Group – Rail, 2015; Maarfield, 2018).

PRICES

Also, the differences in prices can describe differences for passengers in the different countries well. EU railway fares ranged from less than one cent up to 48 cents per passenger in 2012. The average was about 11 cents per passen-ger-kilometer (cf. Figure 3). For longer interurban relations, the prices differed

13

from ca. 2 cents per passenger-kilometer in the Czech Republic to 30 cents per kilometer in the United Kingdom in 2015. Also, the difference of yield manage-ment can be discovered as in some countries, the prices changed a lot de-pending on the date of purchase or the distinction of on- and off-peak prices (cf. Figure 4 & Figure 5).

14

Figure 4: Interurban fares over 300 kilometers: peak single (November 2015) (Cartmell, 2016, p. 63)

15

When comparing international with national fares, it becomes clear that more (8 instead of 5) relations were charging different passenger-kilometer fares de-pending on the date of purchase. It can also be seen that fares for international relations are mostly higher than for intranational relations (cf. Figure 4 - Figure 7).

16

Figure 7: International fares: off-peak return (November 2015) (Cartmell, 2016, p. 69)

The higher discounts for booking in advance can be explained by the likeli-hood of those being operated by non-incumbent operators. Their prices are less commonly “regulated by competent authorities”, which leads to higher flexibility to use yield management strategies (Cartmell, 2016, p. 69).

THE PASSENGER NIGHT TRAIN

Within this work, Steer Davies Gleaves’ definition is used: “A passenger night

train is any train consisting partly or wholly of rolling stock dedicated to, or re-configured for, overnight travel”. (Steer Davies Gleave supported by TRASPOL

- Politecnico di Milano, 2017, p. 17).

POLLUTION

This work bases on the assumption that using night trains is a less polluting way to travel than using other modes, especially airplanes. When regarding pollu-tion, noise and air pollution seem to be relevant factors to look at. Steer Davies Gleave (2017) concluded that noise pollution is of course an existing factor but seems to be accepted by the majority “without complaint” (Steer Davies Gleave supported by TRASPOL - Politecnico di Milano, 2017, p. 99).

Air pollution can be a cause of direct emissions (direct energy consumption of trains) and embedded emissions (manufacturing emissions) (Steer Davies

17

Gleave supported by TRASPOL - Politecnico di Milano, 2017, p. 98). Direct emis-sions depend a lot on the energy source used. Common ways of energy supply are diesel with direct local air polluting effects, and electric energy which pol-luting effect varies according to the source (e.g. coal, gas, nuclear, renewa-ble) (Steer Davies Gleave supported by TRASPOL - Politecnico di Milano, 2017, p. 98). Another important factor for direct emissions is the loading factor of the trains. To offer more comfort for passengers, night trains mostly offer way more space per passenger than day trains which comes with a higher energy de-mand per passenger (Steer Davies Gleave supported by TRASPOL - Politecnico di Milano, 2017, p. 98). It thus is to conclude that day passenger trains are more energy effective and less air-polluting than passenger night trains. Steer Davies Glee (2017) mentions that around 40 % of night train passengers in Sweden claimed to use day trains if there was no night train service. Still, this is unlikely to be the case for international tourism purposes as ways are further and cur-rently, rail seems to play a small role on that corridor (cf. chapter 2.2).

The embedded emissions consist of the production emissions of rolling stock and infrastructure. For the Rolling Stock production emissions, the loading factor effect, as described above, can be mentioned. This is the case as the same length of night train vehicles transport way fewer passengers than passenger day trains would. This is even enforced due to a lower daily or only seasonal utilization. Anyhow, this factor seems to be made up with because of a longer vehicle life of night train compared to day train rolling stock (Steer Davies Gleave supported by TRASPOL - Politecnico di Milano, 2017, p. 98).

COSTS

Next to the costs other passenger train operators face as well (cf. chapter 2.2), night trains come with higher maintenance costs such as change of linen, cleaning, and food supply (Arx et al., 2018, p. 332); (Tourist Austria International, 2020, p. 21). The long lifetime of the rolling stock also comes with additional investments to update it to current needs and expectations. Using night trains only during nights also increases unproductivity (Arx et al., 2018, p. 332).

Different labor costs have a high influence on competitiveness. This is also men-tioned to be the case for the Austrian national incumbent ÖBB and the Ger-man national incumbent DB (Arx et al., 2018, p. 332; Tourist Austria International, 2020, p. 21). ÖBB even sourced out the passenger services in order to safe on staff costs (Tourist Austria International, 2020, p. 21).

High costs ask for high prices which can be up to 200€ for last-minute booking in single compartments (Steer Davies Gleave supported by TRASPOL - Politec-nico di Milano, 2017, p. 104). Rides from Berlin to Zürich are announced to start

18

at 90€ in shared 2-bed compartments and 140€ in single compartments (Schle-siger, 2019, p. 64).

COMPETITIVE MODES

IPRS competes with many other travel modes. Most namely these are private (or shared) cars, airlines, and coaches. Competition performance varies be-tween different relations depending on local peculiarities such as infrastructure quality and accessibilities (Arx et al., 2018, p. 331). Bad performance of railway in comparison to other modes in International Passenger services can also hap-pen due to the costs (and thus ticket prices), travel time, or other, more quali-tative, reasons (cf. Feuerstein et al., 2018; Bellinghausen, 2019; IRG, 2015; and Arx et al., 2018).

Especially high infrastructure access charges are named when it comes to cost barriers for intermodal competition (Feuerstein, Busacker, & Xu, 2018, p. 307). Traffic Researcher Andreas Knie blames the low prices of low-cost airlines who don’t pay Kerosene taxes nor VAT for international relations (Bellinghausen, 2019, p.81 quotes Knie; see also IRG, 2015). Next to the charges, also problems concerning the rolling stock are named. These include, next to the price, the long certification process and the low-developed second hand market for roll-ing stock. Also, the inflexibility compared to other modes is mentioned (Inde-pendent Regulators' Group - Rail, 2015, p. 4).

Travel time is seen as an important factor that negatively influences modal split for the railway on relations with travel times longer than four hours. But it shows that passengers traveling for leisure purposes are willing to accept longer travel times (Arx et al., 2018, p. 331).

Of course, there also cases where the rail is favorable for international passen-ger transport. As mentioned above, this counts for travel times under four hours, but also on corridors where only a few alternatives are available (e.g. Switzer-land-Northern Italy) (Arx et al., 2018, p. 333). Trains also benefit from high energy efficiency, high volumes, and the possibility for train-to-train transfers on long-distant relations (Independent Regulators' Group - Rail, 2015, p. 4).

AIR TRAFFIC

One of the concurring modes, especially on longer distances is air passenger traffic. Airplanes are often unbeatable in travel time on long routes which are the focus of this thesis.

19

While high costs reduced the attractivity of air travel for masses in the past, air traffic liberalization, mainly since 1993, led “to a quantitative and geographical development in supply and its democratization, in particular on account of the boom in low-cost airlines” (Dobruszkes, 2011, p. 871). Figure 8 impressively shows the expansion of European Air traffic since the liberalization until 2010. This comes with a massive increase in air travel within the modal split.

Figure 8: Long-standing and new air routes (Dobruszkes, 2011, p. 873).

In Europe and Asia, also High-Speed Rail connections play an important role on medium and longer relations (Behrens & Pels, 2012, p. 278). Even on rela-tions, where several airlines are competing, High-Speed Rail can often find suf-ficient demand (Behrens & Pels, 2012, p. 287). Within the competition between the Eurostar and air travel, Behrens & Pels (2012) identified “frequency, total travel time, and distance to the UK port as the main determinants of travelers’ behavior”, with leisure travelers reacting variously on fares (Behrens & Pels, 2012, p. 286). Dobruszkes (2011) agrees on the importance of frequency, travel time, geographical structures, and fares and adds the importance to be aware of the total travel chain including waiting and access time (Dobruszkes, 2011, p. 878). The case of the Market between London and Paris indicates that travel

20

time has significant effects on the share within the modal split. After each im-provement within the high-speed network (that means with each travel time reduction), the number of railway passengers increased significantly. At the same time, the share of flights dropped slightly or fell (cf. Figure 9). These find-ings are in line with the research of Dobruszkes (2011). But it is also to mention, that the total passenger volume for the Eurostar is way above the reduction in air passengers.

Figure 9: Number of passengers in the London–Paris/Brussels passenger market (Behrens & Pels, 2012, p. 280)

While passengers on trains benefit of more legroom and unlimited luggage, air traffic offers luggage services (Arx et al., 2018, p. 331; Dobruszkes, 2011, p. 873). Airlines also offer different seating types and different forms of fares and flexibil-ity as well as loyalty programs. These strategies are now partly copied by some rail operators (Cartmell, 2016, p. 102). It shows that the distribution of air fares and railway fares is equal, showing cheaper fares for each mode on specific routes (cf. Figure 10). When comparing the average speed of train and air-plane on international routes, it turns out that traveling by air-plane is not always faster than traveling by train. Indeed, there are many routes with similar aver-age speeds, and trains are even faster on some routes (cf. Figure 11).

21

Figure 10: Rail and air costs: international trips (lowest observed fare) (November 2015) (Cartmell, 2016, p. 97)

Figure 11: Rail and air average speeds: international trips (lowest observed fare) (November 2015) (Cartmell, 2016, p. 97)

22

A special form of air-travel concurrence is those of low-cost airlines, as incum-bent rail operators have difficulties to lower fares (Dobruszkes, 2011, p. 871). Rail operators as Kurt Bauer, CEO of the ÖBB long-distant department (Interviewed in (Bellinghausen, 2019, p. 84)), pleads for equal treatment of all traffic modes to realistically compete with low-cost airlines. Dobruskes (2011) summarizes that “Low-cost carriers exploit their staff and are often able to obtain significant re-ductions in airport charges or subsidies whose legality is disputable” (Barbot, 2006; Dobruszkes, 2008; Hunter, 2006; Marty, 2004; quoted from Dobruszkes, 2011, p. 879). This is a contrast to the high infrastructure charges for railway op-erations (Dobruszkes, 2011, p. 879).

The concurrence of Low-cost airlines for passenger night trains has become visible within the lower charged seat segment by the former German operator (DB Autozug), and with the current Austrian operator (Rogl & Schmidt, 2003, p. 78). But it is important to state that low-cost airlines’ passengers are mainly not substituting rail travel. Instead, LCA offer way more connections than night trains (Steer Davies Gleave supported by TRASPOL - Politecnico di Milano, 2017, p. 18) and thus help to increase European mutual understanding (Bellinghau-sen, 2019, p. 78).

This all shows that air traffic is a very strong competitor for international night trains towards holiday destination. This is because of lower costs and a higher flexibility.

COACH

Another public transport mode competing with (international) passenger rail-way is the coach. Compared to train fares, coach tickets are usually less ex-pensive. The railway can mostly offer faster connections and is thus competitive in fields of travel time (Arx et al., 2018, p. 331). Coaches also compete with night trains, offering overnight services which prices mostly lower than “the cheapest seats on night trains” (Steer Davies Gleave supported by TRASPOL - Politecnico di Milano, 2017, pp. 18–19). Coaches are thus seen as a strong competitor for international night trains. They are a convenient option especially for travelers who are price-sensitive and comfort-unsensitive. Still, overnight coaches come with less comfort than sleeper trains. This means that sleeping might be difficult for many travelers so that for many travelers, night train rides might appear faster than overnight coach rides.

23

Arx et al. (2018) see private cars as the main competitor for IPRS as they make a high share within modal split on international trips (Arx et al., 2018, p. 331). When having a look at travel costs, Cartmel (2016) found out that in November 2015, most international passenger railway connections were more expensive than traveling by car. This is especially the case for the relations between France and the United Kingdom (cf. Figure 12Figure 12). Another factor making cars more attractive is the flexibility in time and location. There is also no need to change the train, and baggage does not need to be transferred (Arx et al., 2018, p. 331; Dobruszkes, 2011, p. 873). It is also to state that cars offer more privacy than public transport. Also, car drivers are not affected by yield man-aging strategies, only by traffic density. A negative aspect of driving is the fact that the travel time can not be used for other occupations such as reading or sleeping. This might especially affect longer journeys.

Reflecting all the competitive modes, it is to state that rail traffic needs to in-crease travel comfort and dein-crease prices for being an attractive travel alter-native for tourists. It is to assume that business models affect those factors.

24

2.3 BUSINESS MODELS

To offer International passenger night train services there are several possible business models. Those can be roughly divided into those affected by govern-mental interventions and those which are running on open access market. Cartmel (2016) calls those types ‘competition “for the market” and competition “in the market”’ (Cartmell, 2016, p. 104). The first describing public service ten-dering and the second describing open access competition.

As international night trains cross at least two states, also mixtures are possible granting e.g. financial support for late night or early morning services in only one state.

GOVERNMENTAL INTERVENTION

Railway services are more and more liberalized. But according to Finger & Rosa (2012), liberalization does not mean “disengagement of public powers from the sector.” They pledge for “Systems to govern interfaces” as well as to identify services that need to be subsidized (Finger & Rosa, 2012, p. 136).

PUBLIC SERVICE OBLIGATION (PSO)

To provide socially desirable services that are not profitable direct subsidies or cross-subsidization are needed (Beria, Quinet, Rus, & Schulz, 2012, p. 117)7; (Fin-ger & Rosa, 2012, p. 35). This requires the willingness of society to pay for those and the services need to be efficient (Beria et al., 2012, p. 117).

Public Service Obligation is defined as “a requirement defined or determined by a competent Authority in order to ensure public passenger transport services in the general interest that an operator, if it were considering its own commer-cial interests, would not assume or would not assume to the same extent or under the same conditions without reward” by EU legislation (Reg. 1370/2007, art. 2(e)); quoted from Finger & Rosa, 2012, p. 35).

The compensation of railway services by PSO is common (Beria et al., 2012, p. 118). This can be executed by exclusive access rights or direct subsidies (Fin-ger & Rosa, 2012, p. 35). Those compensations are seen as “significant source of revenue for railway undertakings in a majority of Member States.” (European Commission, 2019, p. 11)

It is to mention that the duration of PSO contracts for train services should not be longer than 15 years and that those contracts need to be awarded by the competent authority or group of authorities. “Contracts for carrying out such

25

PSOs must be awarded based on clear, consistent and fair rules, and be open to providers from throughout the EU” (Regulation 1370/2007 “on public passen-ger transport services by rail and by road” quoted from Finpassen-ger & Rosa, 2012, p. 35).

But as in international markets several authorities are involved to grant PSO, the organization becomes more complex (Maurer et al., 2010, p. 9). Only a few international services are currently carried out under PSO (European Commis-sion, 2019, p. 11). Some basic barriers can be seen when the question is posed whether or not services are seen as socially desirable. While answers for regional services are mostly consistent for most of the countries, long-distance services can not always be agreed on. While Germany and Spain see long-distance services as a market service, Italy and France agree on the classification of some long-distance services, partly also night train services, as social desirable services (cf. Figure 13; Beria et al., 2012, p. 117).

Figure 13: Classification of social and market services (excluding discounted tariffs) (Beria et al., 2012, p. 117)

Long-distance services are usually paid by the central government (Beria et al., 2012, p. 117). Lately, the responsibilities of the railway sector shifted more and

26

more to independent regional authorities (Finger & Rosa, 2012, p. 136). Re-gional services under PSO subsidies showed “an important increase in the num-ber of passenger-km as well to the renewal of rolling stock” (Finger & Rosa, 2012, p. 41). Another result is an increasing competition between railway oper-ators, while ‘open access only’ did not always show these results (Finger & Rosa, 2012, p. 137). Finger & Rosa (2012) also stress the ability of those public bodies to overcome administrative boundaries for seamless passenger services (Finger & Rosa, 2012, p. 136). This could serve as an example for international passen-ger railway services.

Figure 14: Regulatory strategy for social services. (Beria et al., 2012, p. 118)

COMPETITION FOR THE MARKET

Competitive tendering is the more common and, often, the only possible form of competition (Johnson & Nash, 2012, p. 15). Cartmell (2016) identifies five dif-ferent contract models for competitive tendering processes (cf.

27

Figure 15: Intramodal competition for the market (Cartmell, 2016, p. 104)

- Management contracts pay operators either fixed fees or reward perfor-mance-related outputs. In this contract type, the operator does not bear financial risks. This is because the tendering mostly only consists of the management and not the cost and quality of services (Cartmell, 2016, pp. 104–105).

- A contract type that includes more risks for the operators is the Gross cost contract. This is a wide-spread contract form within urban and suburban systems. Operators compete for the contract in quality and price. Within gross cost contracts, operators do not bear revenue related risks but they do bear cost-related risks (Verband Deutscher Verkehrsunternehmen e. V. [VDV], n.d.a); Cartmell, 2016, p. 105). Transport authorities are usually in charge of setting and changing fares, as well as offering competing services. Fares can, but do not necessarily have to be collected by the operator (Cartmell, 2016, p. 105).

- Within net cost contracts, operators predict fare revenues and calculate the subsidies it needs to run the services. Operators take the risks of lower fare revenues but win in case the revenues are higher than expected (Verband Deutscher Verkehrsunternehmen e. V. [VDV], n.d.b). The op-portunities for operators to change prices vary, often they are fixed by the traffic authority. Still, operators can often influence revenues by at-tracting more passengers with qualitative improvements such as better information or marketing (Cartmell, 2016, p. 105).

- Another, more flexible contract type is the “flexible franchise” model which included the payment of PSO based on revenue uplifts in the Netherlands. This ensures that operators have incentives to raise the de-mand and also offer services above the minimum standards which have been neglected beforehand (Cartmell, 2016, pp. 105–106).

- Out of those contract types, “concessions” grand the highest freedom to railway operators. Concessions come with exclusive rights to operate trains on a certain route. Sometimes, they also include the exclusive right to build infrastructure. This contract type does not restrict services nor

28

fares (Cartmell, 2016, p. 106). According to Cartmell (2016), this model is not fully suitable when socially desired services are not cost-effective (Cartmell, 2016, p. 106).

It shows that all types of contract types have specific advantages and disad-vantages which make it possible to react to specific market particularities.

OPEN ACCESS MODELS

Even more operational freedom is granted within the open-access model or with Cartmell’s (2016) words, the “competition within the market”. In theory, every rail operator can offer passenger services and compete with incumbent operators. Competitors can decide on timetables, routing, fares, quality, etc. and react to passenger demand independently. Of course, the practice is far more complex.

There are different business rationales for international passenger rail services common examples for point to point connections are the western European Highspeed connections offered by Eurostar and Thalys. Another often-used ra-tionale is the extension of domestic routes across borders to increase the de-mand. This strengthens the demand for the domestic network (Arx et al., 2018, p. 332).

BUSINESS MODELS WITHIN OPEN ACCESS COMPETITION

Arx et al. (2018) summarized three different business models for operators to offer international passenger railway services in open access competition. Those are the classical cooperation, joint ventures, and the intermodal com-petition (Arx et al., 2018, p. 333) (cf. Figure 16).

Figure 16: Characteristics of the business models for IPRS (Arx et al., 2018, p. 333)

29

In countries where incumbent operators are active, many IPRS are carried out with the help of a cooperative model (Independent Regulators' Group - Rail, 2015, p. 3). Within the classical cooperation, two or more incumbent operators are partners in cooperation and are legally acting as independent operators which’s responsibility changes at the national borders (Arx et al., 2018, p. 333). That means that each operator mainly needs and keeps the information for their own market, and they care for their own network’s technical, legal, and organizational particularities. Management decisions need to be made in con-sensus and information is only partly shared (Arx et al., 2018, p. 333). As the op-erators need to agree but responsibilities are split, the transaction costs tend to be higher than in joint ventures but can remain low, when “incumbents retain full decision-making powers” (Arx et al., 2018, p. 333). For customers, classical cooperation usually comes with the benefit of a high integration in the national markets. That means that connecting trains can be included within one ticket (cf. column one, Figure 16). But it can also occur that standards differ on one journey (Arx et al., 2018, p. 333). Within a survey between operators, IRG-Rail (2015) found “a clear preference for the cooperation model”. Reasons are ac-cess to foreign resources, less administrative and technical involvement, and the full exploitation of the foreign market (Independent Regulators' Group - Rail, 2015, p. 3). The cooperation between the German DB and the Swiss SBB is an example of this business model (Arx et al., 2018, p. 333).

JOINT VENTURE

A business model with increased cooperation is the joint venture model which is seen as a suitable approach to serve international point-to-point relations or relations with an insignificant domestic market (Arx et al., 2018, p. 333). In joint ventures, two or more incumbent operators found a common company, with an own legal status, to run international passenger railway. Profits and losses are shared between the owning companies (Arx et al., 2018, p. 333). Different revenues for the owning companies can still lead to different priorities or deci-sions in how to develop the cross-border services (Arx et al., 2018, p. 334). Due to the own legal status, the (subsidiary) managerial freedom in joint ven-tures is higher (Arx et al., 2018, p. 333), and the transaction costs to run services are lower than in classical cooperation. Transaction costs are also depending on the mutual trust of the companies (Arx et al., 2018, p. 333). Information on the different national markets is shared and as incumbents are involved (Arx et al., 2018, p. 333), differences between the national networks take a low to me-dium important role. For passengers, the integration into national transport sys-tems can be the same level, or slightly lower than with a cooperation model (Arx et al., 2018, p. 333) (cf. column two, Figure 16). In comparison with services

30

based on cooperative models, joint-venture services are usually perceived as a consistent product. An example of joint-venture services is the French (SNCF) and Swiss (SBB) owned company TGV Lyria (Arx et al., 2018, p. 333).

INTRAMODAL COMPETITION

According to Feuerstein, Busacker, & Xu (2018), the emerging open access competition possibilities are playing “a role in the revitali(z)ation process of senger rail” (Feuerstein et al., 2018, p. 307). Competition on long-distance pas-senger rail can lead to higher efficiency, less overall transport emissions, mod-ernization, and a higher focus on customers (Feuerstein et al., 2018, p. 304). Cartmel (2016) made clear, that intramodal competition is not only limited to direct competition between two commercial operators on the same route. PSO operators can compete against each other or a commercial operator. Typical competition cases are alternative routes to reach the same destination, and overlapping services on the same route sections (Cartmell, 2016, p. 109). New competitors mostly offer lower ticket prices than incumbents (Cartmell, 2016, p. 115), and therefore consumers benefit from lower fares in general. Also, services seem to be improved by competition effects (Johnson & Nash, 2012, p. 21).

Next to the benefits, there are some concerns about open access competition, namely “additional strain upon limited capacity” (Cartmell, 2016, p. 112; also mentioned by Feuerstein et al., 2018, p. 304), fewer services by competitors than by incumbents (Cartmell, 2016, p. 115), and a reduced profitability for the incumbent as well as it is hard for competitors to run profitable services (John-son & Nash, 2012, p. 21; Feuerstein et al., 2018, p. 304). It is also important to note that the passenger rail system becomes more complex with additional operators (Feuerstein et al., 2018, p. 304). Another risk in open access competi-tion is that the incumbent uses resources from other markets in case those are not restricted by service contracts (Cartmell, 2016, p. 112). Competition can lead to different scenarios (Johnson & Nash, 2012, p. 16).

INTRAMODAL ‘ON-TRACK’ COMPETITION

Since 2010, the IPRS market has been opened up to competition within the Third Railway Package (Directive 2007/58/EC) (Independent Regulators’ Group – Rail. 2015, p. 2; Finger & Rosa, 2012, p. 21; Arx et al., 2018, p. 327). Within the business model intramodal 'on-track' competition, IPRS could be theoretically run by any railway operator if the provision of IPRS is legally allowed, barriers are low, the market has enough potential for two operators, and cooperation with a local operator is impossible or not attractive (Arx et al., 2018, p. 333).