EUROPEAN COMMISSION

Brussels, 8.12.2016

SWD(2016) 427 final

COMMISSION STAFF WORKING DOCUMENT

Accompanying the document

REPORT FROM THE COMMISSION TO THE EUROPEAN PARLIAMENT AND THE COUNCIL

Fifth report on monitoring developments of the rail market

{COM(2016) 780 final}

List of abbreviations

List of Countries

Table of Contents

List of abbreviations    

List of Countries    

Table of Contents    

Figures    6

Tables    8

Boxes    8

Introduction    9

1.    The state of the Union railway network    11

1.1.    Description    11

1.2.    Missing links and cross border vision    15

1.3.    Infrastructure management    16

1.4.    Infrastructure density    17

2.    The evolution of the internal market in rail services    18

2.1.    Traffic volumes    18

2.2.    Passenger market    19

2.3.    Freight market    22

2.3.1.    Evolution of volumes    22

2.3.1.    Future of rail freight    26

2.3.2.    Rail noise    27

2.3.3.    Rail freight in multimodal context    28

3.    The evolution of the internal market in services to be supplied to railway undertakings    30

3.1.    Passenger rail stations    31

3.2.    Other service facilities    34

4.    The evolution of framework conditions in the rail sector    38

4.1.    Infrastructure charging    38

4.2.    Capacity allocation and congestion    42

4.2.1.    Network utilisation rates    42

4.2.2.    Congested sections    43

4.2.3.    Managing capacity shortage    44

4.3.    Infrastructure expenditure and funding    45

4.3.1.    Infrastructure expenditure    45

4.3.2.    Funding and financing    49

4.4.    Developments as regards prices of passenger services    52

4.4.1.    Overall evolution of fares and tickets    52

4.4.2.    Fares and tickets in different market segments    55

4.4.3.    Intermodal competition    61

4.5.    The quality of passenger rail services    69

4.5.1.    Punctuality and reliability     69

4.5.2.    Safety    73

4.6.    Rail transport services covered by public service contracts    76

4.6.1.    PSO scope    76

4.6.2.    PSO compensation    78

4.7.    Licensing    80

4.8.    Degree of market opening and utilisation of access rights    81

4.8.1.    Legal liberalisation and actual entry in the rail market    82

4.8.2.    Freight market opening    83

4.8.3.    Passenger market opening    84

4.9.    Development of employment and social conditions    86

4.9.1.    Employment in rail    86

4.9.2.    Socio-demographic structure of the rail labour market    88

4.10.    Harmonisation    90

4.10.1.    EU legislation    90

4.10.2.    Transposition    91

4.10.3.    Infringements    92

4.11.    Digitalisation    94

5.    Funding and efficiency of the railways    96

5.1.    Funding of railways    96

5.2.    Performance in terms of main inputs and outputs    97

5.3.    The challenge of performance management    101

6.    Conclusions    104

Figures

Figure 1 –Length of national rail networks (2014) and relative change since 2009 (length of lines, thousand km)

Figure 2 –The proportion of electrified networks (2014) and relative change since 2009 (%)

Figure 3 –Length of dedicated high speed lines (km, 2015)

Figure 4 –Long term evolution of high speed lines in Europe (km)

Figure 5 –Institutional setting in the Member States (end 2015)

Figure 6 –Density of railway network in terms of surface area and population (2014)

Figure 7 –Evolution of traffic volumes since 2005 and average annual change of train-km since 2009 (%)

Figure 8 –Evolution of rail passenger traffic volumes

Figure 9 –Passenger land transport modal split (%)

Figure 10 –Passenger land transport modal split by Member State (2014) and change since 2009 (in percentage points)

Figure 11 –Passenger traffic volumes by Member State (p-km, 2014), proportion of international traffic (%) and average annual change of volumes since 2009 (%)

Figure 12 –Propensity to travel by rail (2014) and its average annual change since 2009 (p-km per year per inhabitant)

Figure 13 –Evolution of rail freight traffic volumes

Figure 14 –Freight land transport modal split (%)

Figure 15 –Freight land transport modal split by Member State (2014) and change since 2009 (in percentage points)

Figure 16 –Freight traffic volumes (t-km) by Member State (t-km, 2014), proportion of international traffic (%) and average annual change of volumes since 2009 (%)

Figure 17 –Stations' density in relation to network length (number of passenger stations per 1 000 line-km, 2014)

Figure 18 –Number of stations serving more than 10 000 travellers per day

Figure 19 –Number of stations serving less than 10 000 travellers per day

Figure 20 –Number of freight terminals (2014)

Figure 21 –Number of marshalling yards and train formation facilities (2014)

Figure 22 –Number of storage sidings (2014)

Figure 23 –Number of maintenance facilities (2014)

Figure 24 – Number of maritime and port facilities linked to rail activities (2014)

Figure 25 –Number of refuelling facilities (2014)

Figure 26 –Track access charges for different categories of trains (EUR per train-km, applicable 2016)

Figure 27 – Average revenue from the charges for the minimum access package (EUR per train-km, 2014)

Figure 28 –Evolution of track access charges (EUR per train-km, projected 2013 and 2016)

Figure 29 –Network utilisation rates (thousand train-km per line-km, 2014) and relative change since 2009

Figure 30 –Evolution of infrastructure expenditure ( billion EUR) and proportion of maintenance and renewal expenditure

Figure 31 –Total infrastructure expenditure in Member States (million EUR) and proportion of maintenance and renewal expenditure

Figure 32 –Maintenance and enhancement expenditure in Member States in relation to their network length (thousand EUR, per line-km)

Figure 33 –Distribution of allocated funding by Member States (2014-2020)

Figure 34 –Total allocated EU rail funding in Member States in relation to their network length (2014-2020, thousand EUR per line-km)

Figure 35 –Harmonised Index of Consumer Prices: rail transport/all transport

Figure 36 –Fare revenue per p-km (2012)

Figure 37 –Average annual change in revenue per p-km (2007-2012)

Figure 38 –Suburban fares: monthly or 30-day (PPP-adjusted fare EUR per km)

Figure 39 –Regional fares: peak single (PPP-adjusted fare EUR per km)

Figure 40 –Interurban fares under 300 km: peak single(PPP-adjusted fare EUR per km)

Figure 41 –Interurban fares above 300 km: peak single (PPP-adjusted fare EUR per km)

Figure 42 –High speed fare: peak single (PPP-adjusted fare per km)

Figure 43 –International fares: peak single (PPP-adjusted fare EUR per km)

Figure 44 –Comparison of rail and car travel: regional trips

Figure 45 –Comparison of rail and car travel: intercity trips under 300 km

Figure 46 –Comparison of rail and car travel: intercity trips over 300 km

Figure 47 –Comparison of rail and coach travel: interurban trips under 300 km

Figure 48 –Comparison of rail and coach travel: international trips

Figure 49 –Comparison of rail and air travel: interurban trips over 300 km (lowest observed fare)

Figure 50 –Comparison of rail and air travel: international trips (lowest observed fare)

Figure 51 –Punctuality of regional and local passenger services, percentage of services on time

Figure 52 –Punctuality of long distance passenger services, percentage of services on time

Figure 53 –Reliability of regional and local passenger services, percentage of services cancelled

Figure 54 –Reliability of long-distance passenger services, percentage of services cancelled

Figure 55 –Proportion of high and good satisfaction scores for railway stations and rail services

Figure 56 – Significant accidents and resulting causalities

Figure 57 – Significant accidents per type of accident (EU-28)

Figure 58 –PSO Services as % of total passenger services (2014)

Figure 59 – Split of PSO national, PSO international and non PSO services - EU (million p-km, 2014)

Figure 60 –Percentage of traffic under PSO for both type A and type B traffic (2014)

Figure 61 –PSO compensation per train-km (EUR/train-km, 2014)

Figure 62 –Proportion of fare-box revenue (2014)

Figure 63 –Trend in total number of active railway undertakings

Figure 64 – Legal liberalisation and entry of the first competitor in the freight market

Figure 65 – Legal liberalisation and entry of the first competitor in the passenger market

Figure 66 –Market share of competitors in the freight market (2014, % of t-km) and evolution 2011-2014 (in percentage points)

Figure 67 –Number of active railway undertakings in freight market (2014)

Figure 68 –Market share of competitors in the passenger market (2014, % of p-km) and evolution 2011-2014 (in percentage points)

Figure 69 –Number of active railway undertakings in passenger market (2014)

Figure 70 –Proportion of labour force between infrastructure managers and railway undertakings (2014)

Figure 71 –Staff employed in incumbent rail undertakings (2014, thousand)

Figure 72 –Staff employed in infrastructure managers (2014, thousand)

Figure 73 –Gender structure of railway staff (2012)

Figure 74 –Age pyramid of workers in rail (thousand employees, 2012)

Figure 75 –Railway staff structure per age

Figure 76 –Transposition monitoring: market, interoperability and safety Directives

Figure 77 –Monitoring of infringement proceedings: market, interoperability and safety Directives

Figure 78 – Cost and contribution of rail sector (billion EUR, 2012 in 2010 prices)

Figure 79 –Trends in input indicators (2007=100)

Figure 80 –Trends in output indicators (2007=100)

Figure 81 –Trends in financial indicators (2007=100, adjusted for HICP)

Figure 82 – Operating costs per train-km by Member State (EUR per train-km, 2012)

Figure 83 –Passenger track utilisation and freight track utilisation (2012) – absolute figures

Tables

Table1 –Utilisation high speed lines for high speed services (2014)

Table2 –Ownership (O) and management (M) of stations serving more than 25 000 travellers per day

Table3 –Ownership (O) and management (M) of stations serving less than 1000 travellers per day

Table 4 –Congested sections and nodes

Table 5 –Global distribution of allocated funds to transport and rail projects from EU funds under current financial framework (2014-2020, billion EUR)

Table 6 –Distribution of funds from CEF to railways (2014-2020)

Table 7 –Distribution of allocated funds to rail investment from European Structural and Investment funds (2014-2020, million EUR)

Table 8 –Quality: services defined as on time in the RMMS

Table 9 –Serious accidents in Europe (2014)

Table 10 –Fatality risk in different modes (2009-2013)

Table 11 –Proportion of international PSO in total PSO services (2014)

Table 12 –Market entry by domestic Open Access Operators

Boxes

Box 1 –Challenges with the quality of network data

Box 2 –Evaluation of the Train Drivers Directive

Box 3 –Turning around the negative trends in Estonia and Romania

Box 4 –Interoperability of rail networks

Box 5 –Inland freight transport modal split – adjusted methodology

Box 6 –Revitalising rail freight

Box 7 –Policy framework for tackling rail noise

Box 8 –Promoting combined transport

Box 9 –Service facilities and services - compliance with current rules and future developments

Box 10 –Accessibility of stations to persons with disabilities and persons with reduced mobility

Box 11 –Direct cost based charging

Box 12 –Measuring available capacity

Box 13 –Categories of infrastructure expenditure

Box 14 –European Fund for Strategic Investments

Box 15 –National fare data

Box 16 –Anomalies in fares

Box 17 –Monitoring the quality of rail freight services in future

Box 18 –Passenger Rights

Box 19 –The Fourth Railway Package and the Recast of the Railway Safety Directive

Box 20 –Provisions on competitive tendering of rail public service contracts in the Fourth Railway Package

Box 21 –Licensing data as available in ERADIS database

Box 22 –Training in the rail sector

Box 23 –Further harmonisation

Box 24 –Global efficiency gaps

Box 25 –Performance management in the existing EU law

Box 26 –PRIME and its work on KPIs and best practice benchmarking

Introduction

This Commission Staff Working Document accompanies the Fifth Report from the Commission to the Council and the European Parliament on monitoring development of the rail market. While the Fifth Report provides a very compact overview of the latest trends, the current documents develops each topic in more depth. In addition, on the DG MOVE website, the data and graphs used in this document have been made available in Excel format 1 . 

Coverage of the Report

This document presents a non-exhaustive report 2 covering the main developments in EU rail market along the lines of the topics listed in Article 15 (4) of Directive 2012/34/EU establishing a single European railway area 3 (hereinafter the 'Recast Directive'), according to which the European Commission has to report every two years to the European Parliament and the Council on:

1.The evolution of internal market in rail services;

2.Services to be supplied to railway undertakings (Annex II to the Recast Directive)

3.The framework conditions, including inter alia:

infrastructure charging

capacity allocation

investment made in infrastructure

developments as regards prices

quality of rail transport services

rail transport services covered by public service contracts

licensing

degree of market opening

harmonisation between Member States

development of employment and related social conditions

4.The state of the Union railway network

5.The utilisation of access rights

6.Barriers to more effective rail services

7.Infrastructure limitations

8.The need for legislation.

The main focus of this document is on developments between 2009 and 2014. Where available, 2015 data are incorporated. Depending on data availability, some comparisons are drawn with the 2011 (rather than the 2009) situation.

The sources of data include Rail Market Monitoring Survey (RMMS) responses, the Statistical pocketbook "EU Transport in Figures" 4 , Eurostat 5 , statistics collected by various sectoral organisations 6 and ad hoc presentations and studies. Contributions from the Member States, national regulators and stakeholders participating in the Working Group for Rail Market Monitoring in the framework of the Single European Railway Area Committee, have also been considered. In addition, the results of two studies commissioned by the EC in 2014-2015 – Cost and Contribution of the Rail Sector and Prices and Quality of Rail Passenger Services 7 - have formed the basis for the analysis presented in sections 4.4 and 4.5.

All EU Member States are covered, except Cyprus and Malta having no railways. In addition Norway participates in the Commission's Rail Market Monitoring exercise (hereinafter 'the RMMS') and is included in most parts of the report. However, EU total and average figures, where presented, do not include Norway or Switzerland. In addition, 2013 and 2014 RMMS responses were not received from Greece and Ireland did not respond to the 2014 survey. Croatian data are available only as from 2013.

The implementing act for rail market monitoring

This is the last report drawing on voluntary RMMS questionnaires to Member States. As from July 2015 the Commission Implementing Regulation (EU) 2015/1100 for rail market monitoring 8  (hereinafter 'RMMS Regulation') sets rules for mandatory data collection. The questionnaire annexed to the Regulation was developed in close cooperation with the Member States and stakeholders participating in the Working Group for Rail Market Monitoring. While mainly building on the existing RMMS, the new questionnaire includes also some new indicators, e.g. on revenues and traffic outputs, public service contracts, infrastructure charges and employment. With a better defined data requirements and a mandatory collection process, the new reporting arrangements are expected to lead to more consistent and coherent data. Member States' reports will be submitted electronically and, after validation, will be made publicly accessible.

1.The state of the Union railway network

1.1.Description

The rail network is the backbone of the EU's transport infrastructure. The importance attached to rail as a sustainable and clean mode of transport is reflected in the TEN-T Guidelines 9 , in the objectives of the Connecting Europe Facility 10 and the Cohesion Fund 11 priorities. National and European authorities are working together to ensure the necessary support for building new but also for improving existing rail infrastructure as a part of EU-wide multimodal network.

The total length of rail network in 2014 was about 220 thousand kilometres (km), which is about 2% more than in 2009. As shown in Figure 1 , in relative terms the rail network in use has increased the most in Spain and France and decreased in Greece, Portugal, Norway and Austria.

Box 1 –Challenges with the quality of network data

According to available statistics from Eurostat, about 60% of rail lines are single track. In Belgium, the United Kingdom and the Netherlands the share of multiple tracks (double or more) is more than 70%, while in sparsely populated Finland and Sweden less than 20% of the whole network has multiple tracks.

Since 2009, it is estimated that 2800 km of electrified lines has been added to the European rail network. The proportion of electrified lines has since 2009 increased 1.6 percentage points and was in 2014 52%. As shown in Figure 2 , there are marked differences between the Member States - in Luxembourg and Belgium more than 80% of lines are electrified, while in the Baltic States and Ireland electrification rates are below 15%. Most progress between 2009 and 2014 was made in absolute terms in Spain (+781 km) and in relative terms in Greece (+87%). For the coming years Banedanmark (Danish infrastructure manager) and Network Rail (the UK infrastructure manager) have launched large-scale projects for the electrification of major parts of their networks.

Travel speeds and high speed network

There are major differences between countries in terms of travel speeds. Despite the significant investment in the modernisation of the rail network in Eastern Europe, there are still regional networks where maximum permitted speed for passenger trains is 120 km per hour or even less. These are mainly in the Baltic States, Poland, Hungary, Romania, and Bulgaria 13 .

As regards the travel speed of freight trains, in some national networks and international rail freight corridors it is 50-60 km/h. However for the most of international freight trains, especially in central and eastern Europe, the average speed is between 20 and 30 km/h. On some international routes freight trains run at an average speed of only around 18 km/h 14 .

High speed lines 15 make up part of the rail networks of Belgium, Germany, Spain, France, Italy, the United Kingdom, the Netherlands, Austria (see Figure 3 ) and since 2015 also in Poland. In total 3.4% of the European rail network is high speed. Over the last six 16 years the high speed network has been expanded by 1 400 km (31%). More than 110 billion passenger-km (p-km) or 26% of all p-km in 2014 were run on high speed lines. The Spanish high speed network with its 2 871 km in operation and 1 200 km under construction is the second largest in the world after China 17 . In Denmark, Germany, France, Italy and Austria, another 1 200 km of high speed lines are under construction and new lines are planned in the United Kingdom and Sweden. In addition, large sections of the conventional rail network have been upgraded for use by high-speed trains. By 2030 the planned high-speed TEN-T should extend to over 30 000 km 18 .

Building high speed lines requires significant investment and, while there is always a societal dimension linked to railway developments, achieving high utilisation rates is nevertheless crucial for mitigating burdens. A recent study by the Feder Foundation 19 noted that only one high speed line in Europe – Paris/Lyon – generates profits while all other high speed lines are dependent on public subsidies. Table 1 presents the utilisation rates of high speed lines by high speed trains in 2014 20 . The best utilisation rate has been achieved by High Speed 1 in the United Kingdom, operating a dedicated line between London and the Channel Tunnel. Achieving high utilisation rates is more challenging for larger networks. In these terms the French high speed network clearly outperforms the others. The use of extensive high speed networks in Spain is 5 times lower than in in France and almost 4 times lower than in Germany. In the Netherlands, after the technical difficulties of launching the services with FYRA-trains, Thalys-trains are the only high speed trains operated on HLS-Zuid.

Table1 –Utilisation high speed lines for high speed services (2014)

1.2.Missing links and cross border vision

Cross-border projects to remove missing links and bottlenecks between national networks and creating the European single rail area are key components of the TEN-T policy supported by CEF financing. For instance – Öresund – the fixed rail and road link between Denmark and Sweden – has been completed and has brought positive socio-economic benefits to neighbouring regions. On-going projects include for example Evora-Merida (PT/ES), Gent-Terneuzen (BE/NL), Trieste-Divača (IT/SI), Karlsruhe-Basel (DE/CH) and Katowice-Žilina (PL/SK). The estimated cost of completing the TEN-T network is EUR 700 billion, a significant part of which goes to cross-border projects.

In addition, assuring connection between smaller cities in neighbouring countries can contribute to improved economic and social well-being. These links, even if not included in the TEN-T Network, can serve as feeder lines providing access to the TEN-T comprehensive network. Other initiatives, mostly stimulated through EU regional policy, are intended to identify and support such projects, if viable.

In response to the initiative of the EP Committee of Transport and Tourism, which aimed to map the cross-border rail sectors having been abandoned over the last decades, DG MOVE proceeded with an analysis of a selection of discontinued local connections. The aim was to identify necessary conditions for taking some of these projects forward and key factors for them being successful. The results are summarised in the brochure State of Play of Cross Border Railway Sections in Europe (February 2016).

The results of the analysis showed that apart from traditional financing issues, there are many additional barriers for implementing cross border rail projects. These include:

administrative and legal hurdles, such as different permission, concession and procurement rules in Member States;

political barriers, such as unaligned political priorities or opposition from local communities to building a line;

technical barriers in terms of different standards applicable to rail lines and rolling stock , variations in safety certification rules;

operational barriers, such as different languages, infrastructure charging approaches, issues with ticket sales and with the access to service facilities.

Several efforts are ongoing to facilitate and overcome these obstacles. The Fourth Railway Package has given a new role to the European Union Agency for Railways (the Agency) 21 to provide EU level safety certification and vehicle authorisation. Some operational barriers, as regards train drivers, will be addressed during the ongoing evaluation of the Train Drivers Directive 22 and implementing rules will be established to facilitate access to service facilities 23 . DG MOVE has also commissioned a study on 'Permitting and facilitating the preparation of the TEN-T Core Network Corridors' and the new Procurement Directive 24 contributes to the simplification of cross-border projects.

Box 2 –Evaluation of the Train Drivers Directive

1.3.Infrastructure management

Management of rail infrastructure is in the hands of national infrastructure managers. Each Member State has one 'main' (incumbent) infrastructure manager taking care of the core part of the network, and other smaller infrastructure managers (mostly few, but e.g. in Germany and Sweden couple of dozens). These smaller infrastructure managers are responsible e.g. for specific lines, for regional infrastructure or for lines linking railways and service facilities. Transparent and non-discriminatory access to rail infrastructure is of a paramount importance for attracting more operators to the market. Infrastructure manages are also working on removing bottlenecks, raising efficiency as well as improving service quality, punctuality and reliability.

The Fourth Railway Package has proposed further measures to ensure that infrastructure managers perform all the functions needed to run rail operations in an optimised, efficient and non-discriminatory manner. Network efficiency can be improved further by pursuing interoperability and encouraging cross-border cooperation.

An overview of various national governance structures of the main infrastructure managers is provided in the scheme below.

Figure 5 –Institutional setting in the Member States (end 2015)

Since 2012 the following changes have taken place:

In Belgium, the infrastructure manager is no longer part of a holding;

In Slovenia, a holding structure was created and infrastructure manager's functions were merged;

In Ireland, essential functions of infrastructure manager were separated, as required in case of integrated structures;

In France, a new holding structure was set up, and all infrastructure management functions were reattributed to one single infrastructure manager, which is a subsidiary of the holding;

Poland has a particular form of holding structure, where the State rather than the holding controls the infrastructure manager.

1.4.Infrastructure density

The Czech Republic, Belgium and Luxembourg with high population densities, have the most dense rail networks in terms of territorial coverage. Density of rail networks per inhabitant is the highest in sparsely populated Nordic countries, where ensuring connectivity of different regions requires relatively more km of lines. Interestingly, the Czech Republic and Hungary score high and Portugal and Greece low both in terms of lines per surface area as well as per inhabitant.

2.The evolution of the internal market in rail services

There are slight differences in total p-km and tonne-kilometres (t-km) as reported in the RMMS and Eurostat due to variations in the scope of reporting. The traffic volume indicators in this section are based on RMMS data (available as from 2007), because it provides break-downs between PSO and non-PSO services, domestic and international services and to some extent also traffic volumes in different market segments. Eurostat data (as presented in Statistical pocketbook 2016) are used to assess the modal split, which requires combining the data of different modes. Finally, for train-kilometres (train-km), the data of UIC, Eurostat and IRG-Rail have been combined in order to acquire a dataset as complete as possible. Data on train-km per type of traffic (freight and passengers) is unfortunately only partially available in Eurostat dataset and therefore is not used in the analysis.

2.1.Traffic volumes

Figure 7 looks at the evolution of traffic volumes for passengers and freight over the last 10 years. Despite an unfavourable economic climate across much of the EU since 2009, rail passenger outputs were almost not impacted by the crisis and have continued to grow on average 1% per year. Rail freight outputs in t-km in contrast dropped heavily in 2009, the low point of economic crisis and have not yet fully recovered. Total train-km (including both passenger and freight train movements) have effectively not increased indicating possibly certain productivity gains.

2.2.Passenger market

In 2014, about 9.4 billion rail trips were made in EU countries. According to RMMS data, over the last five years to 2014, passenger rail demand in the EU increased by 30 billion p-km to 429 billion. This represents an average growth rate of 1.5% per annum. About 6% of passenger traffic is international.

While domestic travel has constantly increased, the international traffic remained the same compared to 2014. An increase of 5% or more took place only in France and in Portugal. 

While volumes have grown, the modal share of passenger rail in land transport at the EU level has shifted only a half of percentage point from 7.1% to 7.5%.

It should however be noted that there are marked differences between Member States. Some Western European countries (the Netherlands, the United Kingdom) face the challenge of accommodating a modal shift towards rail on already saturated networks. Rail's modal share has increased also in the Czech Republic and Slovakia. At the same time some other Eastern European and peripheral countries (Latvia, Slovenia, Bulgaria, Croatia) cancel services due to financial constraints and decreased demand. The latter is often caused by low quality of services, where low frequency, long travel times and old fleet cause people to switch to other modes. This will in turn lead to a drop in revenue and to a vicious circle of service deterioration.

Figure 10 shows that rail's modal share in passenger transport is higher than the EU average (7.5%) in nine Member States - Austria, Denmark, Hungary, the Netherlands, France, Sweden, the United Kingdom, the Czech Republic, Germany and Belgium, while Spain, Italy and Slovakia are also catching up. In Croatia, Romania, Poland and Bulgaria rail's share has declined more than 1 percentage point.

According to Figure 11 the main markets for passenger rail travel are in large and high-income Western European Member States. The growth of rail use since 2009 varied significantly, the largest relative increases in p-km over the last 5 years being in the United Kingdom, the Czech Republic and Luxembourg. An average decrease of 4 % or more per year has taken place in Romania 26 , Bulgaria and Greece. Most dramatic reduction has taken place in Croatia, where according to available statistics passenger traffic has halved over the last 5 years 27 .

The divergence in growth rates in Member States reflects the impacts of a broad range of exogenous and endogenous factors. For example, increased access to car ownership and higher car use in EU13 Member States has suppressed rail demand in these countries. Conversely, the opening of new infrastructure or services, such as the West Coast Main Line upgrade works in the United Kingdom (2008) have supported rail demand and encouraged modal shift. Congestion on roads is also a strong factor supporting demand for rail services, in particular if facilitated by a suitable service offer and appropriate public policies (such as scarcity charge) 28 .

In Eastern Europe, upgrading rail infrastructure and rolling stock and thus enhancing service quality appears of being the lever to support modal shift towards rail.

Box 3 –Turning around the negative trends in Estonia and Romania

Rail travel per inhabitant in Member States varies by a factor of ten. In 2014, estimated rail travel per capita was almost 1 400 km per year in Austria and France, and more than 1000 in Sweden, Denmark, Germany and the Netherlands; it keeps growing in all these countries. At the same time the propensity to travel by rail was less than 100 km in Lithuania and Greece 29 .

2.3.Freight market

2.3.1.Evolution of volumes

In 2014, more than 1.7 billion tonnes of freight was carried on EU railways. Figure 13 shows that rail freight outputs (t-km), while still being 7% lower than the 2007 peak levels, have recovered 14% between 2009 and 2014 reaching according to RMMS data to 417.6 billion t-km and showing average increase of 3% per year.

In terms of t-km, more than 50% of traffic in 2014 was cross-border, giving to rail freight a much stronger European dimension than is the case for passenger traffic (where the proportion of cross-border traffic is only around 6%). The proportion of cross-border rail freight is expected to grow further, taking into account that the competitive advantage of rail freight vis-à-vis road tends to grow with distance.

At the same time this means that rail freight is particularly affected by the lack of interoperability and sometimes lack of cooperation between the national railway networks, or in other words – the absence of a Single European Rail Freight Area.

Box 4 –Interoperability of rail networks

As shown in Figure 14 rail's modal share in inland freight transport (including railways, roads and inland waterways) was 18% in 2014 and has been relatively constant for several years. However as with passenger transport, behind a stable EU level average there are widely diverging trends at the Member State level.

Figure 15 shows that the modal share of rail freight in total surface transport varies between 80% in Latvia and 1% in Ireland. Overall, rail freight is more dominant in the Baltic States and Scandinavia. Lately several Eastern European countries (Slovenia, Romania and Hungary) have managed to significantly improve the rail's share in freight transport. A drastic decrease has taken place in Estonia due to rapidly dropping transit volumes (see below).

Box 5 –Inland freight transport modal split – adjusted methodology

As shown in Figure 16 , countries with the highest absolute freight volumes are Germany, Poland and France. In small Member States, such as Estonia, Latvia, the Netherlands, Slovenia, Denmark and Luxembourg more than 90% of freight is international. Since 2009 (the low-point of economic crisis), in Spain, Luxembourg and Denmark 34 average annual increase has been close to 10% or more. Only in Croatia, Slovakia, Greece and Estonia there has been a further overall decline.

Estonia observed a decline in freight volumes of 51% since 2009. Rail freight traffic in the Baltic States consists mostly of bulk cargo originating in the Russian Federation and delivered to the ports of the Baltic Sea. These transit flows have been unstable over the recent years due to political instability, but also since the new Russian ports at the Baltic Sea have taken over a major part of this business. Of the three Baltic States, Estonia was most seriously hit and it appears that some of its freight traffic has been displaced to neighbouring Latvia, in part, due to lower access charges (both rail track access charges and port fees) in Latvia 35 .

2.3.1.Future of rail freight

Despite the positive developments listed above, it is clear that with the current pace it will not be possible to reach the objective of the 2011 White Paper 36 to shift 30% by 2030 and 50% by 2050 of long-distance road freight to more energy-efficient transport modes such as rail and inland waterways. The European Court of Auditors notes in its report Rail freight transport in the EU: still not on the right track that overall, despite the EU policy objectives and the EU funding available for rail infrastructure, the performance of rail freight transport in the EU remains unsatisfactory. Market opening has achieved uneven progress in Member States and a Single European Railway Area is still a long way from being achieved. As regards the travel speed of freight trains, in some national networks and international rail freight corridors it is 50-60 km/h. However for the most of international freight trains, especially in Central and Eastern Europe, the average speed is between 20 and 30 km/h. On some international routes freight trains run at an average speed of only around 18 km/h 37 . Different national rules govern path allocation, infrastructure management and pricing making it more difficult for rail to compete with other modes of transport (in addition to other factors like the lack of level playing field between the different modes, the type of goods transported etc). This is particularly true for road transport, whose infrastructure is easily accessible across borders. As a result, rail freight transport performs poorly in terms of volume and modal share. The Court therefore recommends that the Commission and the Member States should help the infrastructure managers and railway undertakings to increase further the competitiveness of rail freight transport, particularly in terms of reliability, frequency, flexibility, customer orientation, transport time and price.

Box 6 –Revitalising rail freight

2.3.2.Rail noise

Rail freight noise is the most sensitive environmental problem for the railway sector and a serious nuisance for citizens living close to railway lines. The European Environment Agency estimates that nearly 14 million Europeans are affected by rail noise.

A number of initiatives have been already adopted at the EU level in order to reduce noise exposure, including the Environmental Noise Directive 2002/49/EC, Technical Specification for Interoperability (TSI) on Noise 40 , financial assistance under the CEF and modalities for noise-differentiated track access charges 41 . However, despite the efforts of the Commission and Member States, progress in tackling rail noise is rather slow. There is a risk that excessive levels of railway noise can lead to uncoordinated unilateral actions by Member States along the most important European rail lines, in particular the Rhine-Alpine corridor, such as applying speed restrictions and restrictions on operating at night. Such restrictions would negatively impact the competitiveness of rail freight.

Box 7 –Policy framework for tackling rail noise

2.3.3.Rail freight in multimodal context

Railways seldom provide door-to door delivery and therefore development of multimodal solutions is vital for its attractiveness. The study conducted by the Commission on the design features for support-programs for investment in last mile infrastructure (i.e. the movement of goods from a transportation hub to final destination) noted that while block trains and single wagon load transport still dominate the European rail freight market with 1.45 billion tonnes (82%), the intermodal transport accounts for 0.31 billion tonnes (18%) and is growing. The latter is triggered by rising maritime volumes, relevance for alpine transit and substitution of single wagon transport.

The above mentioned study also looked at the changes in the framework conditions for the rail freight market in Europe and concluded that over the last decades the opening of transport markets has led to increased intermodal competition for rail freight. There has been also a shift in the range of transported goods and with new cargo types the customers expect faster delivery and transparent transport chains with real-time information.

Therefore, to ensure that rail freight remains competitive, it is crucial to support its functioning in the intermodal context and to develop efficient solutions of combined transport 43 . According to UIRR, the Association of European Road-Rail Combined Transport, the performance or rail-road combined transport has over the last year grown by 12.2% in 2014 and 5.23% in 2015 in terms of t-km. Shorter distance domestic traffic has been decreasing, while cross-border and in particular intercontinental combined transport is increasing (+27% in 2015). Subsequently, the average distance travelled increased from 780 km in 2014 to 882 km in 2015, proving once again that combined rail-road transport in longer distance and cross-border services is more competitive vis-à-vis road-only services than short distance (domestic) transport.

Box 8 –Promoting combined transport

To support the development of combined transport, it is important that the clearance gauge of a line gives access to standard container trains. The Commission will publish a report providing the analysis of the state of the network in these terms and suggest measures for respective enhancements in early 2017.

3.The evolution of the internal market in services to be supplied to railway undertakings

The Recast Directive introduced a set of new rules for service facilities and rail related services. The new legal framework applies to a broad range of facilities including passenger stations, freight terminals, marshalling yards and train formation facilities, storage sidings, maintenance facilities, cleaning and washing facilities, maritime and inland port facilities and refuelling facilities. The provisions also cover the services provided in these facilities as well as additional and ancillary services such as traction current supply, pre-heating of trains, arrangements for transport of dangerous goods, access to telecommunication networks and ticketing services in stations.

Box 9 –Service facilities and services - compliance with current rules and future developments 

The main aim of the legislation is to increase the transparency of access conditions and charges applied and to ensure non-discriminatory access to facilities. Therefore, ownership and management of facilities needs to be monitored. RMMS information is rather complete for stations but still fragmented for other facilities and there is room for improvement as regards monitoring of this part of the rail market. In particular definitions of the various categories of facilities have not yet been harmonised. Therefore any data presented in this section must be interpreted with care. The Commission services are cooperating with national regulators and Member States to refine definitions and improve the quality of reporting.

3.1.Passenger rail stations

There were about 30 000 stations in the EU in 2014 47 , of which about 300 were large stations serving more than 25 000 travellers per day.

As shown in Figure 17 , Austria, Czech Republic and Slovakia have more than 250 stations per 1 000 line-km (i.e. less than 5 km between two stations) while Finland, Belgium and Ireland have, according to RMMS data, less than 50 stations per 1 000 line-km (i.e. more than 20 km on average between two stations).

Figure 18 and Figure 19 present the number of stations in different size clusters. Germany, France and the United Kingdom have the highest number of large stations serving more than 10 000 travellers per day. Bulgaria, Lithuania and Slovenia reported that they do not have such stations 48 . Germany, France, Poland and the Czech Republic have a high number of smaller stations, serving less than 10 000 travellers per day. Very small stations serving less than 1 000 travellers per day outnumber the stations serving between 1 000 and 10 000 travellers per day in all countries except in the Netherlands.

As shown in Table 2 , ownership/management of large stations is concentrated in the hands of a few players. Infrastructure managers or incumbent railway undertakings are managers and often also owners. Governments own large stations in 7 countries, leaving the management to infrastructure managers or integrated companies.

Table2 –Ownership (O) and management (M) of stations serving more than 25 000 travellers per day

Source: RMMS. FR data reflects the situation in 2014 before the 2015 national railway reform. In SE the main infrastructure manager manages platforms in stations, the state-owned Jernhusen owns and manage around 50 station-buildings (out of around 160), including the three largest; the rest is owned by various entities such as local municipalities and private companies. *= in ES the management of commuter stations is entrusted to Renfe Operadora (incumbent RU).

The ownership/management of small stations, presented in Table 3 , is more dispersed across the different players in all countries. In addition to government and main infrastructure managers, other railway undertakings and private operators are more involved in ownership and management of small stations.

Table3 –Ownership (O) and management (M) of stations serving less than 1000 travellers per day

Source: RMMS. FR data reflects the situation in 2014 before the 2015 national railway reform *= in ES the management of commuter stations is entrusted to Renfe Operadora.

Accessibility of stations for travellers with reduced mobility (including disabled people, people with temporary mobility restraints, elderly and accompanying persons such as parents with buggies) remains an EU wide challenge. Accessibility issues arise at different stages of a rail journey, not only at boarding a train and during the journey, but also in the preparatory stage. Given the societal dimension of rail transport, it is important for the sector to tackle this challenge making rail travel accessible to everybody and bringing new customers to rail.

Box 10 –Accessibility of stations to persons with disabilities and persons with reduced mobility

3.2.Other service facilities

Freight terminals

Based on RMMS data and as shown in Figure 20 , Poland has the highest number of freight terminals, followed by Romania, France and Slovakia (with more than 400 terminals each). The widely diverging definitions used by Member States for freight terminals explain the major variations in figures between similar countries. For instance, Poland included 978 tracks with the possibility to load and unload owned by PKP PLK, while the United Kingdom reported only intermodal terminals 49 .

Ownership/management of freight terminals is mixed. Non-incumbent undertakings own and manage terminals only in Germany and the United Kingdom. The government may own terminals (as in Bulgaria, Croatia, Italy, Luxembourg, Hungary Portugal, Slovakia, Sweden and the United Kingdom) but does not manage them. Terminals are mostly managed by infrastructure managers (in 12 out of 23 responding countries) or other private operators (in 11 countries).

Marshalling yards and train formation facilities

As shown in Figure 21 , France, Finland and the United Kingdom reported the highest number of marshalling yards and train formation facilities. However, again, there are divergences in definitions. For example, Germany's 71 marshalling yards and train formation facilities include only installations with gravity hill; including those without gravity hill would bring the figure to 236. Poland has reported 29 installations both with and without gravity hills (including 21 with gravity hill).

In terms of ownership and management, the infrastructure managers are owners of marshalling yards in 14 out of 22 responding countries. Governments own marshalling yards but they outsource the management, often to infrastructure managers (e.g. in Bulgaria, Croatia, Italy, Lithuania, Luxembourg, Hungary, Portugal and Slovakia).

Storage sidings

Storage sidings are sidings dedicated to temporary parking of railway vehicles between two assignments.

Again there are profound differences in definitions used in various countries as well as important reporting gaps (only 16 countries provided data in the RMMS). In most countries having provided data, infrastructure managers own and manage storage sidings. Government ownership is reported in Croatia, Luxembourg, Hungary and Slovakia, without associated management functions.

Maintenance facilities

Similar caveats in terms of definitions and comparability apply to maintenance facilities. Poland's data as reported in Figure 23 for example, includes five levels of maintenance from basic running checkouts to facilities for general overhauls.

The 8 maintenance facilities reported by Denmark cover at the same time only major maintenance and technical facilities on state lines.

Based on the data provided by 22 Member States, ownership and management of maintenance facilities is distributed across different players. The government is owner but never a manager of maintenance facilities. 'Other railway undertakings' and private operators often own and manage these facilities.

Maritime and port facilities linked to rail activities

As shown in Figure 24 , Italy, Germany, Romania, the United Kingdom and Sweden reported the highest number of these installations.

Based on the data reported by 21 Member States, most maritime and port facilities linked to rail activities are owned and managed by private operators. In Germany and Romania the government owns and manages the installations, whereas only in Germany, Slovenia and the United Kingdom the incumbent railway undertaking or other railway undertakings are the owners.

Refuelling facilities

According to IRG-Rail, the market of refuelling facilities appears to be more competitive than marshalling yards, given the high number of these facilities which are operated by independent infrastructure managers. Also, in some cases – as in Norway- railway undertakings make use of on-demand tank trucks in addition to refuelling facilities 50 .

According to the RMMS, ownership and management of refuelling facilities is spread among various market participants, including non-incumbent operators. No respondent indicated government at the same time as an owner and manager of refuelling facilities.

4.The evolution of framework conditions in the rail sector

4.1.Infrastructure charging

Infrastructure charges enable infrastructure managers to recover the cost they incur in providing infrastructure to train operators. The core principle is that the charges should cover at least "direct" costs.

Box 11 –Direct cost based charging 

In addition to direct costs there are other components of charging systems that infrastructure managers can use to enhance:

the effective use of infrastructure capacity (e.g. scarcity charge, reservation charge, discounts to specific traffic flows)

environmental performance (modulation of charges depending on noise emission and usage of diesel/electric locomotives);

cost recovery of specific investment projects (charges based on long-term costs); and

operational performance (penalties/rewards linked occurrence/avoidance of service disruptions).

In addition, mark-ups can be applied on top of the direct cost charges in market segments being able to pay such higher charges. The overall level of cost recovery through infrastructure charges affects the necessary level of government contribution and Member States may require different levels of cost recovery.

As a result of this and other factors, the level and approach to charging may vary both within the charging scheme of one infrastructure manager as well as between Member States. The current RMMS does not allow distinguishing between the various charging elements used by each Member State. Therefore, while comparing the level of charges as reported in the RMMS, the results need to be interpreted with caution.

Figure 26 illustrates the different situations in Member States. In most countries charges for freight trains are higher than for passenger trains, but in Germany, Spain, Belgium, France, Luxembourg and Portugal the situation is the other way around. In Austria, Italy, Sweden and Denmark there are no big differences or the results are mixed. Freight charges in the Baltic States are particularly high, which is to some extent justified due to higher permitted axle-loads. The intercity charges in Belgium, Germany, Spain and France are relatively higher because these include the charges for dedicated high speed lines. Suburban charges are most volatile (varying between EUR 0.17 in the United Kingdom and 11.50 in France) because their levels depend on national approaches to PSO contracts and rail financing (see section 5.1). In France, for example, the regions themselves (rather than railway undertakings) pay the so-called "redevance d'access" to the infrastructure manager for the rail services they have ordered under public service obligation. Norway does not apply charges to the major part of its network.

Better comparability of average charges between the Member States is possible by monitoring the average revenue (i.e. yield) from track charges per train-km. These data are available for most Member States in the IRG-Rail Annual Market Monitoring report and are presented in Figure 27 . Only charges for the minimum access package are included and differentiation is made between freight and passenger trains. It is important to remember that this is an average value, while in each country charges for specific types of trains and/or specific lines could be very different from this average. The extremes on Figure 26 and Figure 27 are similar – the highest freight charges are in the Baltic States and the lowest in Spain, while the highest passenger charges are in France. In Slovenia passenger trains which operate under public service contracts are exempt from track access charges; therefore the average passenger charge per train-km is very low. 

Figure 28 presents the evolution of the applicable track access charges between 2013 and 2016 in various market segments in each Member State, as reported in the RMMS. The figures show that the infrastructure managers with high charges and rapidly decreasing traffic levels may need to alter the level of their charges as to generate a constant level of revenues. This can trigger a shift from rail to road, unless the traffic is largely international. It should be also noted that the Polish and the Bulgarian infrastructure managers have reduced their direct cost based charges in response to an infringement procedure before the EU Court of Justice.

4.2.Capacity allocation and congestion

4.2.1.Network utilisation rates

Capacity allocation schemes should encourage infrastructure managers to optimise the use of their infrastructure, while ensuring fair and non-discriminatory access of all operators to tracks.

Figure 29 shows that rail infrastructure in some parts of Europe is increasingly busy. The Netherlands has by far the most saturated network running about 50 thousand train-km per each line-km per year and, according to available data, 70% of Dutch railway lines consist of multiple tracks. The next group of countries – the United Kingdom, Luxembourg, Denmark, Austria, Belgium and Germany run about 30 thousand train-km per each line-km per year, being still much higher than the EU average of 19 train-km per line-km. In all these Member States rail demand continues to increase setting high demands on infrastructure managers to accommodate the additional traffic, while at the same time maintaining the state of the network and its service quality. Over the last five years, the utilisation rates have in relative terms increased significantly also in Ireland, Denmark, Estonia and Sweden while declining in Greece, Croatia, Poland, Romania and Bulgaria.

Network utilisation rates provide a good basis for evaluation to what extent infrastructure managers are capable to recoup their costs from user charges. Network utilisation in Central Europe is three to five times that in the South East of the Europe, which means that the infrastructure managers in Central Europe can achieve better cost recovery rates. At the same time, many networks with already low utilisation rates have seen further strong decline of traffic levels, limiting their potential to recover the cost. This results in a widening financing gap, which can trigger a downward spiral for the rail system in that country as a whole, unless the State raises its subsidies.

4.2.2.Congested sections

Congested sections are of a particular concern in international corridors, given that saturated networks lead to degraded performance: rejected path requests, delays and longer recovery times in case of disruptions. Therefore, according to Article 47 of the Recast Directive, where after coordination of the requested train paths it is not possible to satisfy requests for infrastructure capacity, the relevant section has to be declared congested. The infrastructure manager has to carry out a capacity analysis to identify the reasons for the congestion and develop measures for easing the situation. According to the RMMS, in 2014 ten Member States had declared part of their network or certain nodes congested - in total more than 2000 km of tracks and 6 big passenger stations. In Sweden, the whole area of Stockholm has been declared congested.

Table 4 –Congested sections and nodes 53

Source: RMMS
* Indicates the whole Stockholm area

In addition to the sections which have been officially declared congested, there are many sections with highly saturated traffic levels. However, it is very challenging to capture the extent of this problem at the network level due to measurement difficulties.

Box 12 –Measuring available capacity 

4.2.3.Managing capacity shortage

Scarcity charges

To some extent, capacity problems can be managed by applying scarcity charges which reflect the scarcity of capacity of the identifiable section during periods of congestion. According to UIC, in 2012 Austria, Denmark, Germany, Luxembourg, the Netherlands, Sweden and the United Kingdom used scarcity charges 55 . EU law only allows such scarcity charges for a longer time if the infrastructure manager can demonstrate that it has exhausted all possible measures to do away with the causes of the problem. In order to ensure that applied scarcity charges were in line with EU rules and were not discriminatory, regulatory bodies need to develop the institutional capability to assess the capacity analyses and capacity enhancement plans of the infrastructure managers.

Framework agreements

Framework agreements are used for setting out the rights and obligations of an applicant and the infrastructure manager in relation to allocated infrastructure capacity and charges to be levied over a period longer than one working timetable period. Framework agreements are currently used by the main infrastructure managers in Germany, Austria, Italy, the United Kingdom, France and Greece. In conditions of limited capacity, it is important to set certain rules to framework agreements to ensure optimal use of infrastructure. The Commission accordingly adopted an Implementing Regulation on framework agreements 56 which established criteria for concluding and amending framework agreements in case other applicants are interested in the same capacity and no other solution can be found to fulfil their request.

Better planning capacity restrictions

Railway undertakings have underlined the importance of more timely and specific information on upcoming capacity restrictions in line with the guidelines of RailNetEurope 57 . The negative impact of planned capacity restrictions, such as maintaining a line, cannot be avoided but can be mitigated if infrastructure managers would consult railway undertakings beforehand and, in case of international services, coordinate among themselves to minimise the impact on users. Therefore, the Commission services are considering a review of scheduling rules as foreseen in Annex VII of the Recast Directive, to ensure the necessary lead times are respected.

Priority Rules

In congested lines the priority rules become important. In general, with few exceptions, passenger traffic has priority over freight and international over domestic. However, specific rules depend on transport strategies of each Member State.

In most Member States the rules for prioritisation of path allocation requests in case of conflicting interests are set in national legislation, but for example in France, Croatia and Sweden the priority rules have been provided only in the network statements of the infrastructure managers. In the Netherlands legislation provides both priority rules and minimum capacity allocations for each market segment. In the United Kingdom, track access contracts between the infrastructure manager and railway undertakings are pre-approved by the regulatory body.

In many countries (e.g. Hungary, Latvia, Poland, Norway) PSO services get the first priority due to their high value to society, while in some other countries (Austria, Estonia and Romania) international passenger traffic is prioritised. Often express trains get preferential treatment compared to commuter traffic.

As a result both the process of how the priorities are defined and to which market segments they apply, are very divergent. This can create obstacles for railway undertakings operating across borders, in particular to freight operators.

4.3.Infrastructure expenditure and funding

4.3.1.Infrastructure expenditure

Fostering railway infrastructure as part of the development of the TEN-T networks has been a prominent issue of European transport policy. The completion of the TEN-T network requires about 550 billion EUR until 2020. The total costs until 2030 for all transport modes are estimated by the Commission services at EUR 1.5 trillion 58 .

According to the RMMS, infrastructure expenditure has constantly increased over the last four years from EUR 29 billion in 2011 59 to more than EUR 45 billion in 2014. The maintenance expenditure has fluctuated, while investment into renewal and enhancements has continuously increased. In 2014 25% of infrastructure expenditure went on maintenance, 29% on renewals and 45% on enhancements, while in 2013 the respective proportions were 32%, 24% and 44%.

Box 13 –Categories of infrastructure expenditure 

As shown in Figure 31 , in 2014, the total infrastructure expenditure was highest in the United Kingdom and in France – much higher than in Germany, even if the German network is by far the largest in the EU. Both in France and the United Kingdom, the infrastructure managers have to catch up with the years of under-investment having at the same time significant investment into enhancements (including new infrastructure) ongoing. In Germany, the infrastructure expenditure in 2014 also increased rapidly (+57% compared to 2013) and additional 2 billion EUR was provided for maintenance and renewals.

Maintenance and renewals

In conditions of increasing traffic in many States and demanding performance targets agreed with public authorities and operators, maintaining the existing network in order to uphold its safety and operational performance and to ensure reliable service, presents a major challenge for infrastructure managers. Many infrastructure managers have suffered from chronic maintenance underinvestment due to a shortage of funds, but also due to the fact that over the last decade priority was often given to the investment into new lines. Liquidating accumulated maintenance backlogs is in general more expensive and disruptive to services than continuous routine maintenance. Therefore it is important to ensure a sufficient and stable level of maintenance and renewal expenditure.

In 2014 Infrastructure managers invested EUR 24 billion in maintenance and renewal of lines. A median proportion of this expenditure in total infrastructure expenditure was 52%. The extremes varied between 4% in Spain, where very high proportion of investment goes into new high speed lines and 100% in Croatia. Since 2011 spending on maintenance and renewal has increased on average 6% per year, most significantly in Romania, Norway, Latvia, Germany and Belgium. The countries with remarkable decrease were Spain and Poland.

Figure 32 gives an overview of the evolution of maintenance and renewal expenditure since 2011 (the earliest available data) in Member States per line-km 60 . It should be emphasised, that expenditure between Member States should not been benchmarked. An adequate level of expenditure has to be established for each network individually, given that this depends on many factors including the length of the network, its architecture (e.g. distance between nodes and switches, signalling system, geographical conditions), the number of tracks, the traffic intensity and current state of the network. For example, in France maintenance costs are predicted to increase, given that the country has to catch up with a maintenance backlog. In the Netherlands, the state of the network is good and maintenance expenditure is expected to decline due to efficiency gains. The high cost of the Luxembourg network is due to its specific architecture allowing trains to run either on left or right track in both directions and due to the relatively short distances between the nodes.

The most notable relative yearly increase in maintenance and renewal costs in 2014 took place in Germany (+72%); but also in the United Kingdom and Latvia. The countries with continuous increasing trend of maintenance/renewal expenditure per line-km were Austria, Norway, Belgium, France, Latvia, Sweden and Portugal.

Enhancements

At the EU level investment into network enhancements have since 2011 almost doubled reaching EUR 20.5 billion in 2014. At the Member State level, the year-to-year amounts are volatile, depending on project pipeline and availability of funding. The top 5 – France, Spain, the United Kingdom, Italy and Poland – accounted for more than 60% of total EU enhancement investment in 2014. In addition Norway (not included in EU total), invested EUR 1.1 billion to enhance its network.

High speed lines

Not all Member States (e.g. the United Kingdom and Germany, and France partially) have managed to distinguish between the expenditure to high speed and conventional network. Nevertheless, at least EUR 7 billion (or 16% of total expenditure) were in 2014 reported to have been spent on high speed lines, of that EUR 4 billion in France, EUR 2.5 billion in Spain and EUR 0.4 billion in Italy. In Spain, the investment into new high speed lines absorbed 90% of total infrastructure expenditure.

4.3.2.Funding and financing

Sources to cover infrastructure expenditure include:

own resources of infrastructure managers in the meaning of track access charges and other revenues; own resources are in general used for maintenance, but to a limited extent also for investment; 'other revenues' include income from renting facilities, such as shops in stations or selling land or structures no longer needed for railways;

national subsidies in the meaning of network grants; depending on country, subsidies can be used for investment only, or both for investment and maintenance expenditure;

EU funds (grants)- for investment only;

Bank loans – usually for investment only.

According to the study The Results and Efficiency of Railway Infrastructure Financing within the European Union Railway 61 more than 50% of railway infrastructure investment has previously been funded by national budgets. EU co-funding added an average of 12% and the remainder was financed by concessions, PPPs, loans, equity capital or, to a lesser extent, by track access charges.

National funding and contractual agreements

Regarding national funding, there is at the moment no comprehensive overview of the level of subsidies provided by national governments to rail infrastructure managers 62 . To ensure that the infrastructure manager has mid-term assurance on availability of sufficient funds, the Recast Directive obliges the Member States to conclude contractual agreements between the competent authority and the infrastructure manager covering a period of at least five years. Contractual agreements should contain performance targets the achievement of which conditions the agreed level of funding. 

By the end of 2014 63 at least five Member States (the Czech Republic, Estonia, Latvia, Portugal and Finland) 64 had no contractual agreement with any of their rail infrastructure managers. The duration of existing contracts varied between one year (for some smaller networks) and 30 years in case of the HighSpeed1 concession of in the United Kingdom. In general, in Eastern European countries duration of contracts was about 3-4 years, while in Western European countries 7-8 years. Most Member States had performance indicators attached to existing contracts, although the complexity varied widely from reliability indicators only to a range of 50+ indicators covering costs savings, energy performance, productivity, average speed, possessions, customer satisfaction etc.

EU funding

The EU can co-fund or support rail projects through the Cohesion Fund (CF), the European Regional Development Fund (ERDF) 65 , the Connecting Europe Facility (CEF), the European Investment Bank (EIB) (mainly loans) and through the European Fund for Strategic Investments (EFSI, guarantees). CEF is the main EU funding instrument for TEN-T investment, while CF and ERDF are mostly used by EU13. These funds play a major role in bridging the infrastructure investment gap in Europe, which is one of the Commission's top priorities.

In total more than EUR 33 billion in grants under the current financial framework (2014-2020) has been allocated to rail investment. As shown in Table 5 , almost three quarters of the CEF funding and 37% of total EU transport funding has been dedicated to rail. Table 6 and Table 7 provide an overview of distribution of funds between different types of projects.

Source: Innovation and Networks Executive Agency (INEA), DG REGIO

Table 6 –Distribution of funds from CEF to railways (2014-2020)

Table 7 –Distribution of allocated funds to rail investment from European Structural and Investment funds
(2014-2020, million EUR)

Figure 33 shows how EU rail funding is distributed among the Member States. Poland is by far the most prominent beneficiary receiving during this financial framework in total almost EUR 10 billion (equals to 30% of total EU funding). Overall, the distribution of funds under the CEF is more dispersed than under CF and ERDF.

If normalised by line-km ( Figure 34 ) Poland remains the main beneficiary with more than 500 thousand EUR per line-km.

Increasingly, the railways have to find ways to use other sources to finance their infrastructure than traditional public sector grants. This would allow them to gain further support for instance also through the European Fund of Strategic investments (EFSI).

Box 14 –European Fund for Strategic Investments

4.4.Developments as regards prices of passenger services

Sections 4.4 and 4.5 of this document are mostly based on the Commission Study on the prices and quality of rail passenger services 67   conducted by Steer Davies Gleave in 2015-2016. The study investigated how the fares and quality standards applicable to rail passenger services in Member States are set and have evolved. It differentiates between suburban, regional and intercity services and covers all Member States as well as Norway and Switzerland. The study also assessed the competitiveness of rail vis-à-vis air and road travel.

4.4.1. Overall evolution of fares and tickets

Each year millions of fares are calculated and marketed by a wide range of national, regional, local and urban authorities and operators. Therefore it is very challenging to assess the overall evolution of rail fares at EU level. Nevertheless, an attempt has been made to get an indication of historical trends by using Harmonised Index of Consumer Prices (HICP) and average yields.

HICP seems to be the only comparable tool available to track the trends of rail prices in all Member States. However it should be noted that at this level of disaggregation, the reliability of conclusions based on HICP data is limited 68 . Therefore the results presented on Figure 35 need to be interpreted with care.

In this chart, a value greater than zero suggests that rail travel is becoming more expensive than a basket of transport services (private and public) across all modes and vice versa.

In all but two Member States (Belgium and Sweden), rail travel appears to be becoming more expensive relative to other modes. While across Europe the divergence is modest (approximately 1% per year), Estonia and Latvia are notable outliers where rail travel prices have increased considerably quicker than the prices of other modes. In Estonia this may be due to the reduction in Russian freight transit traffic, which previously cross-subsidised passenger journeys.

However, there are examples where open access competition (e.g. in Austria, the Czech Republic, Germany, Italy and Sweden) has led to fare reduction on certain lines.

National average yield data

A further possible way to compare national rail fares is to use the average yields – i.e. the ratio between the total passenger revenue and p-km reported at national level. However such a ratio includes multiple ticket types and mixes the impacts of the changes in fares with the impacts of changes in passengers' travelling habits.

Figure 36 69 presents average yields per Member State in 2012. The highest average yields are found in high-income EU15 Member States with well-developed, high-quality passenger networks. Sweden is an exception to this general observation, having average fares broadly half those charged in neighbouring Denmark and Finland. The observed difference in average yield may reflect the proportion of operational costs covered, through necessity or design, by subsidies. As discussed later in section 4.6.2, the operating cost recovery from fare revenue in EU15 is typically higher than elsewhere in the EU. However, in some markets fares may be low, even if railway undertakings seek to maximise revenues, because of either low incomes or competition from other modes 70 .

Figure 37 shows the average annual change in average yield between 2007 and 2012 for a sample of Member States where data was available. In the majority of cases, average yields have risen, although the average rate of increase varied considerably. The very high increases in average yield in Greece may be a result of large reductions in the number of passenger services on offer and consequent reductions in rail use. Despite well-documented increases in the United Kingdom rail fares, average yields fell between 2007 and 2012. This is probably due to passengers shifting from First Class and unrestricted tickets to Standard Class and yield-managed advance purchase tickets 71 .

Box 15 –National fare data

4.4.2.Fares and tickets in different market segments

This section presents an overview of rail fares in various market segments in each Member State. The analysis is based on an illustrative single station-to-station journey, and fares per km for other station-to-station pairs might be considerably different. The fare data were collected during November 2015 and are converted at the market exchange rates and expressed in PPP-adjusted euros to reflect differences in purchasing power in different Member States. All fares per km have been calculated using the straight-line distance between the two cities identified. In some countries this will be closely related to the distance by rail but in others, such as Denmark, which has a large number of islands, this is not the case. Straight-line distances allow meaningful comparisons between modes and reflect the impact of direct versus indirect routing by different modes within this comparison.

Suburban fares

Suburban fares 73 are almost invariably in all Member States administered by the competent authority and may be common to rail and other modes. They are often "policy driven" aiming to increase train ridership, but some Member States consider also cost-recovery targets. However, no individual fare was set by calculating the “cost” of the service to which it applies. A range of single fares apply, but regular commuters will typically buy weekly, monthly or annual tickets, often valid on all modes, and offering increasing levels of discount relative to single tickets. Zonal fares structures are predominant in suburban markets as to simplify ticketing arrangements. There can be a wide variation between the suburban fares applied in cities in the same Member State.

Figure 38 outlines the results for selected station pairs. There are significant differences, but no clear-cut split between EU13 and EU15 as was observed in the analysis of average yields ( Figure 36 ). Most of the fares are between EUR 0.07 and EUR 0.20 per km (PPP adjusted), Prague and Dublin being at lower and higher extremes. In Dublin a monthly zonal ticket is poor value if used only for commuting by train between Dun Laoghaire and Connolly (EUR 0.31 per km after PPP adjustment) and it is cheaper to a buy single ticket (EUR 0.21 per km after). In Prague fare levels do not appear to have been changed since 2011 and remain the lowest among the observed countries.

Regional and interurban fares

Regional and interurban fares may be set by national, regional or local competent authorities. In Member States with opened rail markets some fares may be regulated or left to the market. For instance, Sweden has no regulated fares, but long-distance operators must accept local fares administered by County authorities. In the United Kingdom, a range of fares are regulated with a degree of flexibility varying by location, market segment and ticket type.

The analysis covered a range of fares on regional, interurban and high-speed routes including peak and off-peak single and return tickets and monthly passes. The overall conclusions were:

The fares vary widely between Member States, ticket-types and booking horizons.

On regional routes, in many Member States there is only one fare, and a return ticket costs the same as two single tickets.

For interurban trips under 300 km, fares are often lower if booked in advance or restricted to off-peak trains. Where season tickets are available they are often significantly cheaper than buying daily tickets.

For interurban trips over 300 km, fares per km tend to be lower than at shorter distances and may sometimes be constrained by competition from air services. Advance booking discounts are often available on more expensive Western European networks, with savings of up to 70% on full fares.

For journeys on domestic high speed routes, fares varied widely. Advance booking discounts and yield management systems are common.

For international journeys, fares per km are often higher than the domestic equivalent.

Below are the fares of regional and domestic and international interurban trips analysed in more detail.

Regional fares

Regional fares 74 varied from EUR 0.26 per km between Lausanne and Biel in Switzerland (where many citizens buy annual all-lines passes) to EUR 0.017 per km (for a monthly ticket) from the city of Ostrava to Prerov.

The most expensive fares were usually found on journeys in Western Europe, although some of the highest peak single fares were found in Slovenia and Croatia. Discounts for booking in advance or booking a round trip were only available in a few Member States and no distinction was made between peak and off-peak regional journeys. Finally, only in about half of all Member States was it readily possible to find a season ticket fare, although it is probable that fares may be available on request at a ticket office. For Member States where data was available, the saving per km 75 ranged from 2% in Sweden to 88% in Belgium.

Interurban fares

A distinction was made between interurban journeys below and above 300 km. For the former group, there were no effective domestic interurban services in either Luxembourg or Slovenia ( Figure 40 ). Interurban services over 300 km exist in only 16 of the 28 States with rail networks ( Figure 41 ).

The highest peak single fare found for trips under 300 km was an unregulated single fare of £109 from London Paddington to Cardiff, equivalent after PPP adjustment to nearly EUR 0.60 per km and twice the cost of the next highest fare in Germany. However, other, advance purchase fares for this corridor with the same operator were as low as EUR 0.20 per km. For those Member States where it was possible to find monthly fares, these were significantly cheaper than 20 peak return tickets: the effective discount ranged from 19% in Latvia to 90% in Belgium.

The highest fare found for trips over 300 km was an unregulated single fare of £140.50 from London Kings Cross to Edinburgh, equivalent after PPP adjustment to nearly EUR 0.30 per km. On both interurban routes under and above 300 km, discounts were available when booking in advance and/or having a round trip. Particularly the more expensive Western European networks often offered savings of over 50%. In Eastern Europe advance booking discounts were less common, but at the same time, the walk-up fare on these routes was usually lower. In the pool of routes under consideration, the route Prague-Ostrava shows the lowest fare. Three operators provide passenger transport services on that line (not under PSO).

High speed fares

Domestic services that operate mainly or wholly on high speed lines in seven Member States were considered in this category and Figure  42 summarises the findings.

The highest fare found was an unregulated off-peak single fare of £29.50 from Ashford to London, equivalent to over EUR 0.40 per km, assumingly reflecting the explicit premium placed upon high-speed domestic services to support capital cost recovery. The lowest fare found was EUR 0.05 per km in Italy, for a week-ahead fare using open access operator Italo. This low fare may be as a consequence of competition with the incumbent operator Trenitalia.

Advance booking discounts and yield management systems on high-speed services appear to be offered only in France, Germany, Italy and Spain. Nevertheless, the highest per km fare for both peak single and off-peak return fare types, in the United Kingdom and the Netherlands respectively, are on routes that do not appear to be yield managed.

International fares

Figure 43 summarises the average fares for a selection of international routes. The highest international fare found was a single fare of £170 from London St Pancras to Paris Gare du Nord, equivalent after PPP adjustment to nearly EUR 0.58 per km. The two routes served by Eurostar (Paris to London, Brussels to London) are the most expensive journeys within the sample, reflecting assumingly (1) the premium charged by HighSpeed1 in the United Kingdom to support capital cost recovery and (2) high access chargers applicable in the Channel Tunnel. The lowest peak fare found was EUR 0.031 per km from Bratislava to Prague.

Most international fares per km are roughly equal to, or higher than the fares for equivalent domestic journeys. At the same time, the advance booking discounts for international services are in general higher (up to 70%). This is likely to be because a greater proportion of international services are operated by non-incumbent operators who have the flexibility to yield manage.

Box 16 –Anomalies in fares

4.4.3.Intermodal competition

The Study on Prices and Quality of Rail Passenger Services also assessed how the travel costs and journey times in rail compare to equivalent journeys by car, bus or coach and air travel on regional and intercity routes 76 . The fares chosen were based on the cheapest peak single for travel on the day 77 . The suburban segment was not analysed, given that there was no detailed information or assumptions available on car parking charges, congestion levels and journey time reliability for selected routes.

It appeared that over very long distances airlines offer both the lowest costs and the fastest journey time, and are the dominant mode. For domestic or international journeys over 300 km rail operators may have to charge less than airlines unless they can offer a faster journey time between city centres (possible on high speed lines). Coach operators can compete with both air and rail by offsetting longer journey times with lower fares. Car is less attractive because of the cost and time penalties of entering and parking in large cities. Also, it is not possible to work or rest while driving.

For domestic interurban journeys under 300 km, air travel is not normally available, and coach and rail fares tend to depend on their relative speed and frequency, hence their market power. Car usage remains relatively unattractive due to the above-mentioned cost and time penalties of entering and parking in large cities.

On regional journeys, rail may face competition from coach but car may set an effective ceiling on their fares, unless there are extra high time and cost penalties linked to congestion and parking. Both rail and coach operators may be constrained to set fares, in some cases through PSO contracts, at levels low enough either to attract passengers from car or to be affordable to those with no car.

In every case the actual choice of mode may depend on the characteristics of the travelling party. Airlines can offer extremely low fares up to one year ahead, whereas rail can rarely confirm timetables this far ahead due to its reliance on engineering timetables issued by the infrastructure managers. This constrains the scope for rail operators to compete with airlines over longer booking horizons or to provide complementary travel services and through-ticketing.

Competition with car

The analysis on car competition considered one person per car and took into account the fleet structure, fuel cost and toll charges 78 . The result of the comparative analysis of costs and speeds of car 79 and rail travel for regional and interurban corridors are shown below.

For most regional ( Figure 44 ) and interurban ( Figure 45 ) trips less than 300 km, rail journeys appear more expensive on a fare per km basis than the equivalent journey by car. The largest disparities for regional and interurban journeys less than 300 km are in Switzerland between Lausanne and Biel, and the United Kingdom between London and Cardiff, where the estimated cost of a rail journey is more than five times the cost of a car journey.

While rail journeys are generally more expensive than car journeys, also among EU13 Member States, the disparity is less marked. Rail journeys are cheaper for both regional and interurban journeys less than 300 km in Bulgaria, the Czech Republic, Italy, Portugal and Romania.

Similar trends were found in the longer distance interurban market (over 300 km) with rail journeys only cheaper than the equivalent journey by car in a handful of cases as shown in Figure 46 .

The figures indicate that across all market segments, more expensive rail journeys also tend to be faster. Rail has a higher average speed than car for both regional and interurban trips under than 300 km in most EU15 countries. The fastest rail journeys, relative to the equivalent journey by car, are interurban trips in France between Paris and Reims and in Spain between Madrid and Cuenca, where rail has twice the average speed of car. Interurban trips over 300 km are almost always faster by rail.

In contrast, car travel is faster than rail for both regional and interurban trips in Bulgaria, Croatia and Portugal. The slowest rail service relative to car is the regional corridor in Bulgaria between Burgas and Zimnica where car is more than three times faster than rail.

Competition with coach

Barriers to entering the coach market can still exist in some European countries. The analysis in this segment focussed on comparison of prices and travel speed in interurban trips under 300 km and international destination pairs 80 . The findings for interurban trips under 300 km are presented in Figure 47 .  

Most interurban rail journeys appear to be more expensive than the equivalent journey by coach. This may be because rail offers superior journey time, comfort and reliability and coach is therefore perceived as an inferior mode. None of the investigated interurban coach fares was higher than 0.2 EUR per km; the highest coach fare found was 22 EUR in Austria between Vienna and Graz in Austria. The largest price difference is the journey in the United Kingdom between London and Cardiff, where rail is six times the cost of the journey by coach. Peak rail fares between these points are not regulated while the connecting M4 motorway is often congested.

For Bulgaria, the Czech Republic, Latvia and Lithuania coach fares are higher than the equivalent rail fare while coach trips are only marginally faster than the corresponding rail service. In all other Member States apart from Greece and Croatia, rail services are typically faster than the parallel coach service.

Findings for international trips are presented in Figure 48 Most international rail services are more expensive than the equivalent coach service. As with interurban domestic journeys, this may be because rail often offers faster journeys and can therefore operate as a market “price-maker”. However, in two corridors between Romania and Hungary, and Bulgaria and Greece, coach fares are between two and three times greater than the equivalent rail fare, despite average speeds being similar between modes. In this case it is likely that there are additional factors such as service frequency and quality which permit coach operators to charge a much higher fare.

The high speed rail services from Paris to London (FR/UK) and Frankfurt (FR/DE) are the fastest relative to the equivalent coach services. The percentage difference in fares is largest on the Paris to Frankfurt route, where the rail fare is over five times the cost of the equivalent coach fare. However, the absolute difference is largest on the Eurostar operated Paris to London route where the rail fare per km is EUR 0.42 higher than the equivalent journey by coach.

International coach journeys are more expensive than rail between Sofia and Thessaloniki (BG/EL), Timisoara and Budapest (RO/HU) and Bratislava and Prague (SK/CZ). Along these corridors, and as with domestic interurban services, coach can compete because it is almost as fast as, or faster than, the parallel rail service. Largely as a consequence of poor rail infrastructure, travel in Bulgaria is consistently cheaper and slower by rail than by car or coach, suggesting that rail is the inferior mode of land transport.

Competition with air

A sample of interurban (over 300 km) and international station pairs was assessed. Air travel assessment includes allowances for access and egress costs and time and check-in and border controls at the origin and destination airports 81 .

The costs of travelling by air were more than rail on all interurban routes except in Germany and Spain. The most expensive air costs, relative to rail fares, were in Finland between Helsinki and Vaasa and the Czech Republic between Prague and Ostrava. Both air routes seem to serve mainly business travellers with a relatively inelastic demand for travel.

In most cases, on routes where rail travel was faster than air travel, it was nevertheless less expensive. For example in France, between Paris and Lyon rail offered double the average speed, and half the price of air travel. The corridor in Spain between Madrid and Barcelona is the only route on which rail was both faster and slightly more expensive than the air, suggesting that rail is the superior mode on this route.

Rail was more expensive than air for more international journeys than domestic interurban journeys. The largest difference in fares was on Eurostar between Paris and London (FR/UK) which is twice as fast as the equivalent journey by air and 30% more expensive 82 . Air fares between Prague and Vienna (CZ/AT) were eight times higher than rail fares. This route is operated by a single airline and the main market may be last minute business travellers.

4.5.The quality of passenger rail services

This section assesses the quality of passenger rail services. Freight services are not covered, because there is so far no systematic and comparable data on quality available 83 .

Box 17 –Monitoring the quality of rail freight services in future

4.5.1.Punctuality and reliability

The information in this section builds on punctuality and reliability data from the RMMS dataset 85 . It should be noted that methods for calculating punctuality differ between Member States. In addition, data for some Member States or segments is missing.

According to the RMMS, a train should be considered as punctual if it is delayed by 5 minutes or less for regional services, and by 15 minutes or less for long-distance services. Member States that define on time services differently from this are reported in Table 8 .

Table 8 –Quality: services defined as on time in the RMMS

Source: RMMS

The comparability of statistics is further complicated by the fact that the selection of measurement points may vary 86 .

Figure 51 illustrates that the proportion of regional and local services arriving at their destination “on-time” ranges from 99% in Estonia to 78% in Hungary. However, given the range of exogenous and endogenous factors that might affect the level of punctuality, as well as variations in thresholds, this comparison between the Member States cannot be considered conclusive.

Nevertheless, it seems that the best performing Member States have small passenger rail networks, and Spain is the only large network recording punctuality over 95%. This can be due to the fact that utilisation rates of the Spanish network are relatively low. Three of the best performing regional and local networks are those of the Baltic States where the number of passenger services is limited on the infrastructure dominated by freight traffic.

The punctuality of long-distance services is shown in Figure 52 and tends to be worse than regional and local services. The number of on-time trains ranges from 99% in Estonia to 63% in Croatia.

Germany and Italy, two of the largest networks have some of the lowest long-distance punctuality scores, both with fewer than 75% of services being on time. While the punctuality threshold used in Germany is stricter than in Italy (5:59 and 15 minutes respectively) the punctuality of long-distance services in Germany is nevertheless significantly worse than in Austria, the Netherlands and Denmark which apply an even stricter 5-minute threshold. As with regional and local services, the relatively small networks of the three Baltic States (and Ireland) are the best performing. Punctuality may be hardest to maintain on busy lines, particularly if they carry a mix of long-distance and regional passenger services and freight services.

Reliability is defined as the proportion of scheduled passenger services that are cancelled. As can be seen in Figure 53 and Figure 54, comparable data is only available for some Member States. As a consequence it is difficult to make meaningful generalisations.

Based on data available, it can be said that a high level of cancellation is limited to operations in a few Eastern European Member States. This may, in part, explain Lithuania's relatively strong performance against punctuality metrics: if a train is cancelled it cannot be recorded as late. No Western or Central European Member States cancelled more than 3% of regional or 5% of long-distance services.

Customer satisfaction

The principal source of comparable pan-European data on customer satisfaction with rail services is the Eurobarometer survey of Europeans’ satisfaction with rail services, which was last undertaken in 2012-2013. Overall, at that time roughly half (51%) of respondents scored their level of satisfaction with railway stations and rail services as “high” or “good”, with the remainder (49%) recording “medium” or “low” satisfaction levels. Positive satisfaction scores were typically more prevalent in Western European Member States. However, Germany and Denmark underperformed compared to their Northern European peers, and Latvia outperformed the other Baltic States by a considerable margin. A slightly larger proportion (55%) of respondents scored their level of satisfaction with rail services as “high” or “good” compared to satisfaction with railway stations.

Box 18 –Passenger Rights 

4.5.2.Safety

A common regulatory framework for railway safety is set in Directive 2004/49/EU 88 (the Railway Safety Directive). With a view to establishing a Single European Rail Area, the Directive sets out a framework for harmonising national safety rules, safety certification of railway undertakings, the tasks and roles of the national safety authorities and the investigation of accidents. The Agency prepares a thorough report on the safety performance of European railways. 89  

Managing and monitoring safety

Rail safety in Europe is managed at operational, Member State and EU level. The railway undertakings and the infrastructure managers are expected to assess the risks relating to the safe operation of trains and establish a safety management system. At Member State level, the national safety authorities (NSAs) evaluate the quality of safety management systems when issuing safety certificates for operators and safety authorisations for infrastructure managers.

The NSAs have the legal obligation to report to the Agency all significant 90 accidents occurring on their territory. In case of serious 91 accidents, the National investigation Bodies must notify the Agency of the ongoing investigations and send to the Agency their final report once the investigation is closed.

Box 19 –The Fourth Railway Package and the Recast of the Railway Safety Directive

Safety Performance in 2014

Figure 56 and Figure 57 present the main indicators of safety performance, as collected by the Agency.

There were about 1 000 rail fatalities in 2014. Railway safety continued to improve between 2009 and 2014, with fatalities, serious injuries and significant accidents all decreasing.

Excluding suicides, more than two thirds were accidents to persons caused by rolling stock in motion and level crossing accidents, the latter not being related directly to the rail system.

In 2014 seven accidents, as reported in Table 9 , were categorised as serious.

Table 9 –Serious accidents in Europe (2014)

Source: "Railway Safety Performance in the European Union 2016", the Agency

Table 10 –Fatality risk in different modes (2009-2013)

Rail remains nevertheless one of the safest modes of transport.

4.6.Rail transport services covered by public service contracts

Public Service Obligations (PSO) mean requirements 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 commercial interests, would not assume or would not assume to the same extent or under the same conditions without reward 92 .

4.6.1.PSO scope

According to the RMMS, 68% of the p-km by rail in 2014 were made using PSO services (291 billion p-km out of 428 in EU28 93 ). Compared to 2012 the proportion of PSO services in total rail passenger services has increased 4 percentage points.

Like shown in Figure 58 , the proportion of PSO services in the total passenger traffic varies widely between the Member States. In Denmark, Ireland, Croatia, Greece and Luxembourg all passenger services are under PSO. In general, the larger Western-Europeans countries (except the United Kingdom) have relatively less PSO services.

Public service obligations target almost exclusively national lines; in international traffic the proportion is only 1%, Luxembourg having the highest share of international lines under PSO.

The RMMS asked the Member States to report also the proportion of the traffic under PSO for type A (regional, suburban) and type B (long distance) traffic. The results are summarised in Figure 60 . As expected, regional and suburban traffic is more often subject to PSO, while long distance traffic is more suitable for commercial services.

4.6.2.PSO compensation

Ensuring the provision of public rail services usually implies the need for compensating undertakings which, either after a tender or a direct award, commit to a certain level of service 95 . In most countries providing data via the RMMS, the PSO compensation per train-km is higher than EUR 5 96 and the total support provided was around EUR 20 billion..