Options for increasing competition in the Great Britain rail market: on-rail competition on the passenger rail market and contestability in rail infrastructure investment


The first part of this study reviews competitive options for on-rail passenger services. At present, 99% of passenger rail services are provided by franchised train operating companies, some of which operate competing services over either the same or parallel routes. The remainder are provided by ‘open access’ on-rail competitors, which have successfully applied to the ORR for track access. In authorising such services, the ORR has regard to their impact on passengers and on public subsidy to the railway sector. In practice this translates into a consideration of the proportions in which competitive services either generate new revenues or ‘abstract’ them from franchised services.

Available evidence on the impact of open access competition supports the view that passengers benefit from more frequent services which provide higher customer satisfaction and lower prices. Little evidence has been collected so far concerning the expected impact of competition on cost of service. This is likely to be limited by the degree to which train operating companies can in fact control their costs, but it is likely that exposing them to competition on some routes may improve overall productivity.

The report considers four incremental means of enhancing on-rail passenger competition and one more radical one. The options in the former category are:

  • amending the current basis for evaluating open access proposals, to focus more directly on estimates of the economic benefits such competition brings and its impact on public subsidy; this more direct method of considering costs and benefits also links with a proposal noted below;

  • allowing franchisees effectively to sub-contract or ‘sub-licence’ services to other providers; this would only happen if the latter could generate more revenue or provide the service more cheaply than the franchisee, but there may be (probably isolated) cases when this is so;

  • extend the scope of competition among franchisees. The evidence suggests that the extension of such competition has benefited passengers and that its withdrawal has harmed them. Such competition eliminates leakage of revenues from the franchised sectors, but lower fares might increase the demand for public subsidy. This might be mitigated by positive effects on productivity. The benefits would, however, fail to materialise if the operators colluded. This possibility would be reduced if competition between ‘symmetrical’ operators were avoided;

  • as in other sectors, competitive open access entrants offering profitable services might be asked to contribute to the costs of loss-making services by paying a levy on revenues or a surcharge on track access. This might be difficult to achieve directly, but a similar effect can result from conducting an auction for track access made available to open access competitors, the revenues from which would be set against the additional cost in public subsidy estimated to flow from the granting of any track rights.

We conclude that each of these methods has the potential to benefit passengers, while taking into account the other objective of the current regime to control public subsidy.

We consider in addition a more radical conceptual proposal to reduce considerably the scope of franchising and rely on competition for more services. This involves, on potentially competitive routes, inviting bids for long term rights to use the track in a combinatorial auction in which limits would be placed on what any operator can buy, in order to prevent the emergence of monopoly. After suitable changes to European legislation, such rights would be tradable, subject to caps, in the secondary market. The regime would be implemented under the supervision of a systems operator.

Only a sketch of this proposal is offered here, but we note that it could be implemented gradually and progressively over time.

Investment in the rail infrastructure is high - Network Rail’s capital spend is around £25bn for the current 5 year period. This is a large capital budget for one national monopoly provider to dispose of. Efficiency in delivery of this investment programme is dependent on the strength of regulatory incentives which are hampered by the lack of comparative data and an ownership structure that provides weak sanctions for poor performance.

The potential gains from increased contestability for Network Rail’s investment could be significant, arising from two sources. Firstly, gains might be made from more efficient delivery by new entrants. Secondly, contestability might lead to a greater transparency of Network Rail’s performance, creating sharper incentives for its own efficiency.

This report considers these issues in the light of the theoretical literature on separation and integration and the experience of other sectors, and has drawn some high level conclusions on the most promising potential options for introducing contestability – by identifying areas where the costs from separation are the least and are therefore most likely to be outweighed by the benefits. In reaching these conclusions, the authors have also considered how the incentives of alternative providers such as Train Operating Companies (TOCs) may differ from those of Network Rail. This is important in deciding whether outcomes are likely to represent an improvement on those resulting from the current provision by Network Rail.

Although there are some forms of investment where it is unlikely that contestability would be beneficial, there are others where there may be more potential and where scope for contestability should be tested.

Considerations of co-ordination are likely to mean that Network Rail will remain the most appropriate owner, decision-taker and delivery agent for enhancements that are embedded in the existing network such as track improvements. However, in the cases of customer-facing investments (such as in stations, platforms and associated infrastructure) and of significant investments in new track or extensions, the separation that contestability brings may be less costly. Chiltern Railways already represents an example of successful transfer of procurement responsibility to a third party for this type of investment.

A critical choice in determining the form of contestability is that of the stage of the investment process at which it is introduced; in particular, it is of key importance whether the incumbent remains responsible for tendering contracts to deliver capital projects or whether a third party takes on this responsibility . Evidence from other sectors and countries suggests that, whilst both approaches involve the creation of additional transactions boundaries, the costs will tend to be larger under the latter approach, as it will involve a new body (for instance ORR or the Department for Transport – DfT – and Transport Scotland - TS) running the tender and the
need for a a new infrastructure for the tendering process. This has sometimes led to significant delays where this model has been implemented.

On this basis several regulators (eg Ofgem and Ofwat) have concluded that this approach is likely to be appropriate only for very large projects, where timing of delivery is less urgent, and where the potential benefits may be of a scale to outweigh the additional costs. Such a mechanism already exists for very large rail infrastructure projects, with DfT or Transport Scotland taking the role of procurement body.

Were ORR to take on such a role for middle to large size projects, it would have to be sufficiently well resourced and skilled to run such a process effectively. In addition, this approach would need to fit with any reforms to the franchising process for TOCs. For many projects (where it is likely that TOCs would be the alternative provider) that franchising process might offer the vehicle for such contestability, although there are doubts about whether the current DfT proposals for reforming the franchising process are sufficiently flexible and conducive to eliciting good investment proposals, providing opportunities for attracting new investors/delivery agents other than TOCs, or for more radical options involving transfer of ownership of some assets. . The authors recommend an approach of carefully targeted experimentation to assess the merits in practice of a variety of approaches.

Whilst Network Rail already contracts out delivery of most enhancement projects, there should also be more scope for Network Rail to pass responsibility for procurement as well as delivery of investment to third parties such as TOCs, possibly working in consortia with parties such as developers and retailers. Network Rail should be encouraged to consider ceding responsibility for procurement at an earlier stage, such as the design stage, to ensure the maximum potential input from innovative third parties. Whilst there are good reasons for hesitating to recommend the imposition of mandatory ‘outsourcing’ requirements by ORR, the regulator should nonetheless consider enhancing its scrutiny of Network Rail’s procurement processes, and accept that compulsory outsourcing would remain a final sanction in cases of persistent poor efficiency by Network Rail.

In sum, the rail infrastructure comprises a number of different types of opportunity for investment, but contestability is most likely to deliver net benefits in the case of projects which are most ‘separable’ from the core network in terms of planning and operations and where there are few train operators to coordinate. Project scale will be an issue in deciding whether it is appropriate to introduce external tendering by ORR or DfT / TS, although in many of the ‘separable’ instances (for instance stations and track extensions) the tendering process may get absorbed into the franchising process. Finally, ORR should give consideration to increasing the regulatory scrutiny of Network Rail’s procurement of investment.

Prof. Martin Cave, Dr Janet Wright