The Scottish Government

« Previous | Contents | Next »

Listen

CHAPTER FOUR: MARKETS, COMPETITION AND FRANCHISING

Introduction

4.1. We subdivided services within, between and across networks into eight categories shown in Table 4.1.

TABLE 4.1 RAIL MARKET SECTORS

4.2. For the purposes of the study we have treated Scotland and Hamburg/Schleswig-Holstein as networks in their own right, so that services between Scotland and England, or between Hamburg/Schleswig-Holstein and the rest of Germany, are considered to be "cross-border".

4.3. In practice, transit flows of through rail passengers are also possible in the networks of Hamburg/Schleswig-Holstein, Denmark and Sweden, but we have found no data on these flows and would expect them to be immaterial compared with the flows within, and between adjacent, networks.

Evolution of passenger demand

4.4. To understand the recent evolution of passenger demand within, to and from Scotland we analysed Lennon industry data. We allocated passengers into the markets listed in Table 4.1, using the following definition:

• Cross-border: travel across the border between Scotland and England
• Point-to-point: travel between Edinburgh Waverley/Haymarket and Glasgow Central/Queen Street
• Urban/suburban: other travel wholly within the central belt
• Regional: other travel elsewhere within Scotland
• Long-distance: travel between the central belt and the rest of Scotland

4.5. Figure 4.1 shows the relative importance of these different market sectors in 2004/5, the latest year for which data are available.

FIGURE 4.1 RAIL TRAVEL IN SCOTLAND: MARKET SECTORS

4.6. Over 80% of passenger trips are "urban/suburban" within the central belt area including the Strathclyde PTE services, Dunblane, the Fife circle and North Berwick. These typically short journeys, however, only account for around a third of passenger-km, almost half of which are on Anglo-Scottish "cross-border" trips, although much of this travel takes place in England rather than Scotland.

4.7. Figure 4.2 shows the trends in passenger trips in each of these market sectors, expressed as a change relative to 1999/2000. Urban/suburban travel within the central belt dipped slightly during 2002/3 but has since recovered and is now above 1999/2000 levels. Long-distance, regional and point-to-point Edinburgh-Glasgow passenger trips have grown by an average of at least 5% per annum over this period.

4.8. Cross-border travel was in slight decline until 2004/5, probably as air captured most of the market, at least to and from London and South East England. However, rail services during this period were disrupted by the West Coast Route Modernisation ( WCRM) programme which is now coming to an end, and services linking Scotland with Birmingham and London will improve further in December 2008. Rail's competitive position relative to air may improve and it may recover market share.

FIGURE 4.2 RAIL TRAVEL IN SCOTLAND: PASSENGER TRIPS

4.9. Figure 4.3 shows the equivalent trends in average trip length for each sector. The point-to-point 75 km Edinburgh to Glasgow trip has not changed, and there has also been little change in the average lengths of other trips within Scotland. The greatest change has been that the average length of cross-border trips has fallen by 12%, consistent with the loss of longer-distance trips to air, but still averages over 400 km.

FIGURE 4.3 RAIL TRAVEL IN SCOTLAND: TRIP LENGTH

4.10. All the comparator railways have data on their ticket sales but may not be able to identify the number and length of rail trips made with the multimodal zonal tickets used in many large urban areas. Even where this underlying data is available, it is not always fully exploited as a source of management information or published as operating statistics.

4.11. We examined published data sources and also asked bodies responsible for collecting, analysing and publishing passenger statistics for whatever data they were prepared to make available. We focused on the key "workload" measure of passenger-km or, failing that, passenger numbers, and attempted to disaggregate urban/suburban flows, including any point-to-point travel, from regional and long-distance flows. In the event even this level of disaggregation was not readily supported by the data available.

4.12. The limited data available suggests that "workload" is growing on all the networks, but that the rates and patterns of growth vary widely and are inevitably dependent on local factors. In particular, the apparent relative growth on different networks proved to be highly sensitive to the base year chosen, and in many cases reflected definitional changes.

4.13. Nonetheless, a few specific points give some indication of the issues faced in the comparators:

• New Zealand has broadly stable numbers of passengers outside Auckland and Wellington, but the main service connecting these two cities is being withdrawn, suggesting that any growth in interurban travel lies elsewhere. The cities themselves are experiencing growth in passenger commuter demand and Auckland has invested in rolling stock and infrastructure to permit a citywide doubling of frequency and capacity.
• Ireland is seeing a fall in average trip length, with growth in the Dublin area but decline in longer-distance markets.
• Denmark's new links across the Storebælt, opened to rail traffic in 1997, and Øresund, opened in 2000, have contributed new demand, and the Storebælt link now carries one in eight domestic rail passengers. However, around 60% of domestic passenger journeys are in the Copenhagen area.

4.14. As we discuss below, one consistent feature across the networks is that the emerging capacity constraints tend to be on lines passing through the main or capital city, where it is typically most expensive to add new capacity and where there is the greatest need to coordinate the roles of rail and other modes.

Success factors

4.15. A number of factors contribute to the potential commercial viability of rail services:

• A dense flow, helping to fill trains and keep them utilised
• High speed, making rail both competitive with other travel options, and productive, with trains producing many seat-miles per day
• Even loads over the length of the trip, minimising unused space on each trip
• Even loads over the day, minimising unused space at quiet times of day
• Lack of fares regulation, allowing operators to set profit-maximising prices

4.16. Figure 4.4 summarises which of these factors apply to each type of service.

FIGURE 4.4 FACTORS CONTRIBUTING TO RAIL'S COMMERCIAL VIABILITY

4.17. Urban/suburban trains normally serve relatively dense flows and are likely to offer speeds competitive with car and bus travel, at least in Europe. This is less true in New Zealand, where speeds achievable on urban roads are generally higher, and on the rail infrastructure are generally lower, than in Europe. However, urban/suburban trains are slow in absolute terms, and hence have low productivity. They are also most loaded only at the "city" end of their trip and only twice a day, in the rush hours, and, despite demand which is inelastic to price, at least in the short to medium term, their fares are usually set at levels determined by local policy.

4.18. Regional trains rarely serve dense flows and, while faster than urban services, may also be slow, at least compared with car. Loadings may be more even at different parts of the trip and, with less reliance on commuter travel, at different times of day.

4.19. Long-distance trains, and many cross-border services, tend to serve dense flows at speeds which are both competitive and productive, but may still face variations in load over their trip or at times of day. Non-stop or point-to-point services, such as those between cities and their airports, have the same load throughout their trip and may also benefit from demand spread evenly throughout the day and week.

4.20. Freight services can benefit from all of these factors, but only in the form of "bulk" trains providing a service between two points to a timetable carefully matched to customer demand. The capacity is matched to a dense flow, and a non-stop fully-loaded train exploits the advantages of speed and loading, although trains may only be loaded in one direction. Point-to-point bulk freight, typically providing a dedicated service for a single customer, is often the only type of rail freight which can compete with road on overall cost, speed and quality of service.

Identifying commercial services

4.21. In practice, two factors make it more difficult to tell whether passengers services are, or could be, commercial.

4.22. First, fares are regulated, or at least subject to government approval, on all the urban/suburban services and on many of the regional and long-distance services we examined. European law allows Member States to set fares policy and hence, by implication, to maintain fares at unprofitable levels and prevent competitive entry on a commercial basis. At least some "non-commercial" or Public Service Obligation ( PSO) services may be potentially profitable and contestable, but there is no way of knowing this without removing restrictions on fares, which often cover only half of direct operating costs and would need to at least double before services could be operated commercially. However, to a potential entrant who believes that a service with unregulated fares could be profitable, a combination of regulated fares and PSO support has the same effect as state aid.

4.23. Second, the separation of commercial and non-commercial services required by the First Railway Package ( see Annex 4) is not imposed in practice and would probably be counter-productive if it was, for a number of reasons:

• Commercial and non-commercial elements cannot efficiently be split. For example, if a commercial operator does not find it worth stopping at a remote minor station, the only way to provide and maintain services may be to duplicate its trains.
• The pattern of services which can be operated commercially may vary by time of day, week and year and will change over time.
• Where services are marginal, minor changes in commercial operators' views of their viability may mean large and unpredictable changes in service levels.

4.24. In Sweden, SJ has freedom to select what services to operate on a commercial basis, but this does not prevent the PTAs from franchising services which may mix commercial and non-commercial elements or, on at least one occasion, SJ re-entering the market "commercially" in competition with a "non-commercial" franchisee. In Scotland and Denmark, there has been no attempt to separate franchises into commercial and non-commercial elements, and franchisees are free to add commercial services at the margin if they have, or can acquire, sufficient access rights, trains and crew to do so.

4.25. In practice, services are normally only accounted for separately where they are operated and managed separately, because they are naturally independent due to the geography of the railway or their use of distinct fleets of different types of rolling stock. Urban/suburban, regional and long-distance trains are normally designed for their specific markets, as are trains for point-to-point services such as airport rail links, which are often purpose-built for short journeys with large volumes of baggage.

4.26. Nonetheless, we examined the rail markets of Scotland and the comparators and attempted to assess, from the available information, where rail services could be profitable.

4.27. Figure 4.5 shows our initial analysis of where services would, in general, be capable of covering their direct operating costs and marginal infrastructure costs and hence, in principle at least, operate commercially.

4.28. In Hamburg/Schleswig-Holstein, Denmark and Sweden, regional services are split between two of these categories, which we have coloured as appropriate.

4.29. There are no cross-border freight services in New Zealand and no transit services there or in Ireland, Northern Ireland or Scotland. While transit freight may occur along the corridor through Norway, Sweden, Denmark, Hamburg/Schleswig-Holstein and the rest of Germany and beyond, we identified no data on transit passengers, who are likely to be few, and therefore do not distinguish them from cross-border passengers.

FIGURE 4.5 RAIL MARKETS IN SCOTLAND AND COMPARATORS: VIABILITY

4.30. Point-to-point passenger flows are only identified separately in Sweden, on Arlanda Express between Stockholm and Arlanda Airport, but for illustrative purposes Figure 4.5 also identifies services between Edinburgh and Glasgow.

4.31. Cross-border passenger flows do not exist in New Zealand but can be found in all the other networks:

• Ireland and Northern Ireland have the cross-border "Enterprise" service linking Dublin and Belfast
• Scotland has Anglo-Scottish services to many parts of England
• Hamburg/Schleswig-Holstein has services to Denmark and the "rest of Germany"
• Denmark has services to Hamburg/Schleswig-Holstein and Sweden
• Sweden has services to Demark and Norway

Commercial freight services

4.32. Freight services can generally be operated commercially among the comparator networks, but are most difficult to sustain commercially in Ireland, where distances are relatively short and there are few generators of dense flows. Northern Ireland, the smallest network under study, has no wholly internal freight flows, but Ireland's national operator CIÉ manages to sustain some domestic traffic. We understand, however, that these freight services are marginal, even without paying any access charges for the use of the infrastructure. With limited opportunities for rail freight, the Irish Department of Transport does not offer grants or subsidy to encourage mode shift, and contraction of freight services in favour of road transport appears likely to continue.

4.33. Commercial freight services operate on all the other networks. New Zealand is not subject to European law and the privatised former national operator has a profitable monopoly, but all the other networks have at least some competitive provision.

4.34. In the smaller networks only a relatively small proportion of freight traffic is internal. Denmark's small domestic market is, like Scotland's, complemented by long-distance cross-border flows into Schleswig-Holstein and Germany and is now, with the help of the Storebælt and Øresund links, building considerable transit traffic between there and Sweden. Recognising this market reality, in June 2001 Denmark's freight operations, DSB Gods, were sold to Railion Denmark, a subsidiary of the national railways of neighbouring Germany, Denmark's gateway to the rest of continental Europe. DSB continues to hold a 2% stake in Railion Denmark, but the formerly national industry has adapted its structure to reflect a predominantly international market. Denmark and Scotland both lack "domestic" freight operators and are instead served primarily by operators providing services within, or to and from, a larger neighbour.

Commercial passenger services

4.35. Commercial long-distance and regional services are only identifiable in Sweden, where SJ has chosen to operate them on relatively high speed lines and between relatively populous cities, particularly on the triangle between Stockholm, Malmö (for Copenhagen) and Göteborg (for Oslo). We understand that there has been an application to operate open access international services between Oslo, Göteborg and Stockholm, but in advance of being obliged to allow such services under the Third Railway Package ( see Annex 4) the government has turned down the application to protect SJ. There have also been applications by some of the PTAs to operate inter-regional services, but all such proposals to date have also been turned down.

4.36. Outside Sweden, all the other passenger services appear to be loss-making, either because of fares policy and regulation or, as we discuss below, because operators do not provide separate accounts for elements of the service, if any, which are profitable.

4.37. Sweden has, however, allowed direct competition on price, journey time and frequency between local services and the Arlanda Express airport concession, which was established as a private sector business and operates profitably, paying Banverket access charges based on marginal cost. Arlanda Express also allows other operators to use its infrastructure for negotiated charges which compensate it not only for marginal infrastructure costs but also for any abstraction of its passenger revenue.

4.38. Other examples of point-to-point airport rail links are Oslo's Flytoget and London's Heathrow Express and Gatwick Express, all of which operate commercially, in the case of Gatwick Express as a franchise paying a "premium" to the Department for Transport. We also tried to identify whether any interurban rail links could be operated commercially while paying the marginal costs they impose on the infrastructure and hence, in principle, be attractive to competitive operation or entry:

• In Ireland, we were told that the most likely candidate would be the comparatively dense interurban service between Dublin and Cork. However, IÉ is supported by a single national subvention and does not provide information on the revenues and costs of individual services.
• In Scotland, commercial operation might be possible on the principal link between Edinburgh and Glasgow, and possibly to other major cities, but the costs and revenues of these services are not currently separately identified.
• In Denmark, we were informed that it would probably be possible to operate a Copenhagen to Århus service across the Storebælt link on a commercial basis, but that this would mean a loss of cross-subsidy to other services currently provided under the same contract.
• Between Denmark and Sweden, commercial operations might also be possible on a Copenhagen to Malmö service across the Øresund link, which is shortly to be concessioned by the Danish and Swedish governments.
• In New Zealand, a "commercial" twice daily regional commuter service is currently offered between Wellington and Palmerston North, but revenues are insufficient to fund replacement rolling stock.

4.39. Where services are profitable, Figure 4.5 also shows how they are currently provided. These patterns of profitability reflect a gradual "retreat" by rail into markets in which it has a comparative advantage. Analyses of railways' comparative advantage generally agree that these are typically limited to:

• Transport of bulk freight, such as fuel, raw materials and agricultural produce, particularly over long distances
• High-speed interurban passenger services, at least at distances, or on routes, which do not face air competition
• Urban/suburban passenger services

4.40. Typically, bulk freight and high speed interurban or point-to-point passenger services can operate on a commercial basis after paying at least their marginal infrastructure costs, but support is required for urban/suburban services which are justified on other grounds such as congestion relief.

Non-commercial passenger services

4.41. Figure 4.5 showed that the majority of passenger rail services in Scotland and the comparators are probably not commercial. Figure 4.6 repeats Figure 4.5 but shows how these non-commercial services are provided.

4.42. In New Zealand the pattern of service provision is relatively simple: on long distance services the privatised incumbent, Toll Rail, has a monopoly where it is currently operating and also has first right of refusal to provide services on any new lines or new subsidised services. In practice, the only service which appears to be commercially viable operates over the 140 km between Palmerston North and Wellington and serves regional commuter and business travel.

4.43. Toll Rail also provides suburban services in Wellington under a contract negotiated with the regional authority, but in Auckland it has withdrawn and the regional authority has franchised the services to a new operator. A small number of trains used mainly by tourists also operate in New Zealand, but their primary role is leisure, rather than transport, and none of the comparators reported any scope for such services on their networks.

4.44. IÉ provides all of Ireland's internal services and Translink provides all of Northern Ireland's internal services. Cross-border passenger services, in the form of the Dublin-Belfast "Enterprise" service, are operated jointly by the two operators as an integrated service with a dedicated fleet.

4.45. In Scotland, both cross-border and domestic freight is provided commercially. Passenger services are all franchised, but point-to-point flows such as Edinburgh to Glasgow are served by two operators at the same price, and cross-border flows are served by up to 3 operators competing on journey time and price. All regional and urban/suburban services are provided, probably at a loss, by ScotRail.

FIGURE 4.6 RAIL MARKETS IN SCOTLAND AND COMPARATORS: PROVISION

4.46. In Schleswig-Holstein, many loss-making regional services have been franchised, but Hamburg's urban/suburban S-Bahn network is operated by national incumbent Deutsche Bahn. Deutsche Bahn also provides long distance services within Schleswig-Holstein, "cross-border" into the rest of Germany and, in cooperation with DSB, Denmark.

4.47. In Denmark some local networks have been exempted from the European requirements of separation of infrastructure and operations and open access and are managed on a self-contained but integrated basis. Urban/suburban services are provided either by incumbent DSB or under a franchise, as are some regional services, but DSB continues to operate all long-distance domestic services and cooperates to provide cross-border services.

4.48. In Sweden, incumbent SJ retains the monopoly on any inter-regional services it can operate profitably, but loss-making services and intra-regional operations are operated locally, often under a franchise.

Trends in commercial viability

4.49. The viability of rail services reflects market conditions in Scotland and the comparators, which may change over time. We therefore considered whether there are grounds to expect major changes in the range of services which will be commercially viable in the medium term of around ten years.

4.50. Rail freight services almost invariably face strong competition from road freight, with which they must remain cost-effective to compete. The principal drivers of competitiveness in road transport are:

• Continued efficiency gains through higher capacity vehicles and more effective fleet management
• Through the recent accessions to the EU, access to a lower-cost pool of drivers, at least for international freight
• Rising road congestion or, as and when this becomes controlled by road user charging, higher infrastructure access charges
• European employment legislation related to drivers' working hours

4.51. We concluded that there is unlikely to be a major shift in the commercial viability of rail freight services over the next ten years, and that commercial opportunities are likely to continue to be strongest over longer distances or when handling dense flows.

4.52. Rail passenger services face competition from air at longer distances, but the effect of changes in air competition is unlikely to be material to most of the networks we have studied. At one extreme, in Scotland, Ireland, Northern Ireland, Hamburg/Schleswig-Holstein and Denmark, railways have no material competition from air services. At the other, in New Zealand and Sweden, even where cities are linked by rail, journey times are so long that air is likely to remain the dominant mode unless its costs increase significantly.

4.53. Congestion on interurban roads is rising in densely populated parts of Europe stretching from the Ruhr, through the Netherlands and Belgium to South East England, but is less important in the comparators' networks which are largely rural. Road congestion is rising in the areas around their major cities and in Scotland's central belt. This suggests that rail's competitive advantage may grow most in urban/suburban areas, and in the central belt, but as we discussed above, these services are currently provided on a social, rather than a commercial basis. Over time, however, it is possible that more urban/suburban services will become commercially viable, at least where additional passengers can be carried on longer or more frequent trains without major investment in rolling stock or infrastructure.

4.54. In practice, the commercial viability of rail services would probably change most if fares regulation were removed, operators were allowed to set profit-maximising fares, and support was confined to services which could not operate profitably on even this basis. However, this would conflict with many governments' policies not only for transport but also for social inclusion and the environment. A number of interviewees commented that the concept of commercial viability was irrelevant in the context of current, and foreseeable, government policies. In Stockholm, for example, we were informed that there is a political consensus that local public transport fares remain at a level which covers less than half of their operating costs.

Operation of commercial services

4.55. In practice, while profitability is necessary for commercial or competitive provision of rail services, it is not sufficient, for several reasons:

• The limited capacity of the infrastructure to provide additional services
• The risk of competition policy allowing "cherry-picking" by other operators
• The need for service integration, for example, requiring operators to provide an integrated timetable and accept the same fares
• The availability of rolling stock
• Other barriers to entry

The capacity of the infrastructure

4.56. Sweden was the first of the networks examined to restructure its railway with a view to increasing competition. By separating infrastructure provider Banverket from operator SJ, and setting infrastructure access charges based on marginal costs, it hoped to stimulate market entry by competing operators.

4.57. In practice, this policy proved unsuccessful, at least partly because of the practicalities of operating competing services on a predominantly single track railway. Timetables on single track lines need to be constructed not only to separate trains travelling in the same direction but also to allow trains travelling in opposite directions to meet and cross in passing loops. Even if capacity was available for two operators, no workable means was found of letting them specify timetables which were mutually compatible, operationally efficient and commercially viable. Sweden now requires any operator wishing to provide services across county boundaries to seek government permission which, as described above, is normally refused if SJ objects. It is not clear how this timetabling issue would be resolved if and when other operators are permitted to provide services on single track sections of the network. The same issue would apply in the many parts of the comparator networks which are still predominantly single-track.

4.58. Even where the networks have multiple tracks, however, capacity may be fully used, particularly in and around the major cities. Germany's DB Netz infrastructure is, in principle, open to new entrants, and Connex and also Arriva have run some open access services, but capacity constraints are often a problem. Part of the difficulty is that DB has "grandfather rights" over its historic timetable. Connex's Nord-Ostsee-Bahn wanted to provide services linking Schleswig-Holstein with Köln and Berlin, but the best train paths it could obtain gave a journey time from Hamburg to Berlin of 5 hours compared with around 2 hours by DB, and it concluded that the services would not be competitive. Interviewees in other networks often commented that it would be difficult to offer new entrants competitive paths.

4.59. We asked specifically about capacity constraints in the various networks studied and were told of the following constraints:

• In New Zealand, infrastructure capacity in the Auckland suburban network, requiring double-tracking of the Western Line and remodelling of key junctions, and a shortage of suitable rolling stock for urban/suburban services in the Wellington area.
• In Ireland, shortage of capacity through Dublin Connolly station, which could be relieved by a new tunnel for DART services. ( See also Chapter 6)
• In Northern Ireland, infrastructure capacity between Belfast and Coleraine, requiring some double-tracking, and a shortage of rolling stock
• In Hamburg/Schleswig-Holstein, shortage of capacity through Hamburg and in particularly at the Hauptbahnhof or main station.
• In Denmark, shortage of capacity through Copenhagen central station and on the main line west to Århus.
• In Sweden, shortage of capacity through Stockholm central station, which would be relieved by a new "Citybanan" tunnel. ( See also Chapter 6)

4.60.. The most common problem, shared by the networks of Ireland, Hamburg, Denmark and Sweden, is capacity through the central station of the main or capital city.

Competition policy on "cherry-picking"

4.61. A second issue faced by commercial operation is the extent to which it would allow new entrants to "cherry-pick" or "abstract" revenue, duplicating or displacing the incumbent's commercial services and leaving the incumbent with only loss-making services:

• If the incumbent is providing PSO services under a net cost contract, it may find that its operation is no longer viable and it may cease to operate
• If the incumbent is providing PSO services under a gross cost contract, it may continue to operate, but the public authority supporting the service will suffer a loss of revenue

4.62. Operators may also be unwilling to invest in developing open access services if any successful entry may immediately become subject to further abstractive open access. However, with current limited levels of open access in Scotland and the comparators, it is difficult to establish whether this is an issue in practice.

4.63. In New Zealand, there was a concern that open access to allow competition for freight business would lead to extensive abstraction, with competition focusing on key routes to the detriment of more marginal services. Full open access has therefore not been pursued and the privatised incumbent was granted a monopoly which it retained when the infrastructure was taken back into public ownership.

4.64. One approach to dealing with the problem of abstraction is to compensate the incumbent for its loss of profits. Sweden examined the idea of compensating SJ for any losses resulting from competitive entry, but could find no workable means of implementing this policy. Instead, as described above, SJ is still protected by a monopoly which is sustained as much by the need to support SJ as by considerations of economic efficiency or competition policy.

4.65. A more "rules-driven" approach was adopted when franchising began in Great Britain in 1996 and a need was identified to protect franchisees' uncertain levels of potential competition. ORR introduced a policy of "Moderation of Competition" whereby access rights would not be awarded to operators competing on a number of major passenger flows nominated by the franchisees. Although these arrangements were intended to be transitional, the need to protect the franchisees from uncertainty remained and the practical issues of permitting competitive access on an increasingly crowded network became apparent. In the event, ORR retains a policy of not awarding access rights for services which are "primarily abstractive" of existing operators' revenues, although award of access rights even on this basis has proved contentious as it can still have a material effect on the revenue of existing franchises.

4.66. The current position is that both Sweden and Scotland have policies that commercial entry should be restricted to protect existing operators. In Sweden, access effectively remains under political control, which may not prove to be consistent with future European liberalisation. In Scotland there is a rule-based approach which is overseen by an independent regulator but may not actually be necessary.

The need for integrated transport

4.67. Figure 4.6 shows that direct competition on both journey time and price exists on rail services such as:

• In Sweden, the point-to-point flow between Stockholm and Arlanda airport: competition of this type also exists to London's Heathrow and Gatwick airports
• In Anglo-Scottish markets, a range of longer-distance journeys between Scotland and England

4.68. Many flows in urban/suburban markets are also sufficiently dense to support more than one operator, but most urban transport authorities pursue policies of integrated transport, not only between rail operators but also between rail and other modes. Between Glasgow and Edinburgh, for example, ScotRail and GNER offer trains at different times and, in Glasgow, between different stations, but a common fare applies to all services. The fares of different operators are also fully interchangeable within the urban/suburban and regional networks of Hamburg/Schleswig-Holstein, Denmark and Sweden, although not within Ireland. Even where two operators serve a given passenger flow, they typically offer an integrated timetable and share revenues according to surveys of passenger usage of their services.

4.69. In London, Europe's largest urban/suburban rail network, even this limited form of competition has gradually been removed, primarily because of the need to coordinate services to obtain the maximum effective capacity from the infrastructure and to balance peak passenger demand evenly between trains. Competition has been seen as irrelevant when neither operator can provide sufficient capacity for its passengers. Where possible, the policy is now to have a single operator into each London terminal.

The availability of rolling stock

4.70. Even if a new entrant can find infrastructure capacity, obtain the required access rights, and operate commercially at the fares it can charge or must accept, it will still not be able to enter the market unless it is able to obtain suitable rolling stock. Lack of rolling stock can be a critical barrier to entry for new operators.

4.71. Sweden recognised this problem in 1988, when Banverket and the principle of open access were first established, and has taken a number of steps to make rolling stock available:

• In 1988, legislation transferred the rolling stock used by SJ for local services to the PTAs, enabling them to make it available to other operators if they wished to do so.
• In 2001, when the "new" SJ was created, its rights over its rolling stock were restricted by an obligation to make it available to other operators if it decided to withdraw from services.
• Subsequently, a number of the PTAs, including Storstockholms Lokaltrafik, have established "Transitio", an organisation for buying and owning new rolling stock which can be made available to contracted operators.

4.72. Denmark has taken a similar approach to Sweden, with DSB holding rolling stock in a subsidiary company and required by law to make surplus stock available to the winners of competitive tenders for its services. Arriva leases trains from DSB, although it has since decided to buy its own stock, reducing this potential dependence on a competitor.

4.73. In New Zealand, regional authorities have purchased the incumbent's urban/suburban rolling stock and have established local companies to own it and to procure new stock for use by franchised operators. They also have access agreements with the incumbent which ensure reasonable access to its facilities when required.

4.74. In Scotland, ownership of all British Rail rolling stock was transferred in 1994 to three specially-created rolling stock companies " ROSCOs" which compete to procure and provide stock to franchisees.

4.75. Ireland and Northern Ireland have no provision for making the incumbents' rolling stock available to new entrants, and the uniqueness of the track gauge of their railways means that rolling stock from elsewhere cannot be used on the network without modification. As we discussed above, however, all services are at best marginal at current fares levels and there have been no applications for competitive entry.

4.76. In Schleswig-Holstein access to rolling stock remains a problem. There has been no political will to remove control of rolling stock from DB, making new entry difficult. New entrants must not only buy or lease new stock but also be able to get it in place ready to begin operations.

4.77. At first, the high cost of procuring a duplicate fleet, and the logistical difficulties of replacing one operator's fleet by another overnight, mean that franchising was confined to regional services operated with a small number of trains which could be either leased, purchased second hand or new, or provided by a relatively large franchise operator which can make them available from elsewhere in its operations. Subsequently, regional rail service franchises were awarded on the basis that the procuring authority would provide new or almost-new coaches to successful bidders who did not own rolling stock. More recently, however, a number of mechanisms for leasing rolling stock have emerged.

4.78. Some Länder have established rolling stock "pools", similar to Sweden's Transitio, to own rolling stock, particularly where unusual types of vehicle are needed, and make it available to franchisees. Schleswig-Holstein has decided against this approach but has introduced a "Wiedereinsetzungsgarantie" or re-use guarantee which guarantees the initial purchasers that the rolling stock will be used by subsequent operators.

4.79. Successful bidders have also been able to order stock to meet the requirements of the franchise but lease it from the main manufacturer or through a ROSCO such as Angel Trains or Deutsche Leasing. One difficulty remains the time required to award a franchise, order, manufacture and deliver rolling stock and start operations.

4.80. Testing new rolling stock can also be a problem. When Connex's Nord-Ostsee-Bahn took over services on the 237 km "Marschbahn" from Hamburg to the island of Sylt at the end of 2005, it suffered from a number of technical problems with newly-adapted locomotives.

4.81. Safety certification of new rolling stock is also a potential barrier to entry for operators who are forced, or choose, to provide their own rolling stock. This has been a major problem for some of the new rolling stock fleets introduced in Great Britain, but generally has been less of an issue among the comparators, and can be largely avoided in Sweden and Denmark where the incumbent's stock is made available to new operators of its services. In Germany we have not identified any particular difficulty in gaining acceptance of new stock. Nonetheless, the European Commission is aware that acceptance and safety certification can be problematic and the process is being addressed through some of the measures of the Second Railway Package ( see Annex 4).

Other barriers to entry

4.82. The European Commission anticipated in the First Railway Package that operators wishing to provide services would require access not only to the basic infrastructure but also to a range of other services, the "minimum access package", many of which might also be in the monopoly supply of the infrastructure manager. The Package therefore requires infrastructure managers to make available a range of ancillary services that are essential for the safe, efficient and competitive operation of rail services, such as traction current and access to passenger stations and train maintenance facilities.

I4.83. n 2005 we were appointed by the European Commission to carry out the Railimplement study to investigate and report on how the First Railway Package had been transposed into law, and implemented in practice, in the relevant European states. Our report is now available from the European Commission website and we do not repeat our findings here.

4.84. In 2006 we were appointed by the European Commission to carry out the Servrail study. The study's objective is to investigate and report on the current and likely future position of rail-related services, including those covered in the minimum access package, and to establish the degree to which non-discriminatory access to, and efficient pricing of, these services has been established. The Servrail report is not yet published, but we have discussed with the study team the draft final conclusions which have been provided to the European Commission.

4.85. A key finding of the study is that perceptions of the current effectiveness of rail access liberalisation measures vary between networks and between stakeholders. Across Europe as a whole, incumbents generally accepted the current situation as satisfactory, whereas new entrants to many networks reported a number of problems. In the networks of Great Britain and the comparators of Sweden, Denmark and Germany, stakeholders had greater experience not only of the problems associated with the implementation of the First Railway Package, but also of their actual and potential solutions. This variation in perception across Europe appears broadly consistent with that found in other liberalising industries.

4.86. Stakeholders reported that structural issues still remain to be resolved in most networks, but recognised that many of them are introductory or transitory problems arising particularly where the opening of the rail market, and provision for competitive entry, are still new and untested. In most cases, they considered that the problems would be resolved as more experience is gained and the procedures and processes are revised and developed to reflect lessons learned.

4.87. Many Servrail stakeholders and many of our interviewees expressed the view that a critical prerequisite for the effective liberalisation of markets is that the railways' regulatory functions are carried out by adequately skilled and well-informed staff.

Summary

4.88. While a number of barriers to entry remain where rail-related services are in limited or monopoly supply, the findings of the Railimplement and Servrail studies, and the opinions of our interviewees are that, at least in Great Britain, Sweden, Denmark and to a large extent in Germany, governments are committed to liberalisation and are working to remove technical and administrative barriers to entry.

4.89. Within the comparators, the principal issues for entry, and in particular to whether entry can be in the form of competition "in the market" or "for the market", are successively:

• Whether any additional services could be commercially viable
• Whether capacity could be made available for them alongside existing operators
• To what extent existing operators should be exposed to competition, depending on whether they are providing PSO services on a net cost basis or are themselves taking the commercial risk of open access
• Whether competition between operators would be consistent with wider objectives such as integration

4.90. Commercial viability is possible for freight services carrying bulk goods, at least over long distances. In practice, however, operators do not normally operate directly competing trains "in the market" in the hope of attracting customers to their service. Instead, operators compete to sign contracts "for the market" to provide a service tailored to the transport needs of a specific customer.

4.91. Commercial viability for passenger services is limited to a few interurban or point-to-point flows where the prevailing fares, and the regulatory environment, allow the recovery of costs from revenues. Unless fares are allowed to rise, in some cases to at least double their current levels, there is likely to be little scope for commercial operation in the comparator networks.

4.92. Given that commercial viability in Scotland and the comparator networks is likely to be so limited, at least at current fares levels, the potential difficulties of capacity, competition, and integration are of only limited importance. The principal scope for competition is likely to be "for the market", through franchises, contracts or concessions, which have successfully been introduced in Scotland, Sweden, Denmark, Schleswig-Holstein and New Zealand.

4.93. These networks which have introduced franchising generally consider it an effective means of providing services subject to a number of factors:

• The availability of suitable technical skills in the public sector specifiers, procurers and regulators of services
• The development of effective contracts and in particular a performance regime to incentivise service quality

4.94. We discuss some of these points further in the following chapters.

« Previous | Contents | Next »