بررسی اثرات فرانچایزینگ و چشم انداز راه آهن مسافری در بریتانیا : شواهد از دو دور اول فرانچایزینگ (1997-2008)

This paper provides an up-to-date review of the previous literature concerning the impact of passenger rail franchising on productivity and costs in Britain, and also presents important new evidence. In particular, the extension in time of previously-used datasets offers the first opportunity to study the impacts of re-franchising. The previous literature emphasised the failure of franchising to produce sustained productivity gains, with a sharp deterioration in productivity after 2000. The new evidence presented offers a somewhat more positive view of the British experience. It suggests that part of what was previously considered to be falling productivity may in fact be due to exogenous changes in diesel prices. Further, new data suggests that the recent increases in costs have resulted in higher quality of service. Finally, competitive re-franchising, and the associated unwinding of short-term management and re-negotiated contracts, seems to have led to improvements in productivity between 2006 and 2008. Nevertheless, it remains the case that passenger rail franchising in Britain has failed to reduce costs in the way experienced in many other industries and in rail in other European countries. The evidence is that somewhat larger franchises, avoiding overlapping and optimising train density and length, should reduce costs. We also speculate that the major increase in wages and conditions of staff might be moderated by longer franchises, although that remains to be proved. This re-appraisal of the British case is important in the context of the wider international interest in the use of franchising in passenger rail, and its relevance to the current review of ways of introducing competition into the domestic rail passenger market in Europe.

Franchising (or competitive tendering) has become an important method for introducing competition “for the market” where competition “in the market” may be undesirable. Starting with the 1991 European Commission Directive 91/440, Europe has embarked on a process of deregulation in the rail sector, progressively opening up rail markets to competition (both in and for the market). Though not currently required by legislation, a number of European countries have utilised competition for the market in order to improve the efficiency with which passenger rail services are delivered (see, for example, Alexandersson & Hulten, 2007). This method of procuring passenger rail services has also been adopted elsewhere, most notably Melbourne and South America. There is therefore considerable interest in understanding the impact of competitive tendering in passenger rail around the world, and learning from the varying experience in different countries. In Britain in particular, in late 2009 both the Office of Rail Regulation (ORR) and the Department for Transport (DfT) initiated reviews of the rail service delivery model in Britain, based on learning the lessons from passenger rail franchising around the world. Looking forward, the European Commission is reviewing options for the introduction of competition into the domestic rail passenger market (see Nash & Matthews, 2009). Thus, understanding the impact of passenger rail franchising, and the relative success of alternative approaches, will become even more important from a policy perspective. Against the background set out above, this paper aims to review and update the current state of knowledge concerning the impact of passenger rail franchising on costs, productivity and efficiency in Britain. Its contribution in the context of the previous literature is as follows. Firstly, the paper provides a comprehensive review of the previous literature, also bringing it up-to-date to incorporate recent studies. Secondly, new evidence is presented, including the addition of new data to cover the years 2007 and 2008 (the most recent study did not extend beyond 2006). The addition of two additional years of data is important because it gives us greater insight into the impact of the re-franchising process – which started in 2004 – during which most of the British franchises let at privatisation were again opened up to competition. Given the policy interest in Britain, and also internationally, with regard to rail franchising, its effects and optimal ways of implementation, it is important to have a comprehensive statement of the evidence on British rail franchising, based on the most up-to-date data. Further, given the problems experienced during the first round of franchising it is also particularly important to explore whether re-franchising has begun to turn the situation around. In competitive tendering experience more widely, it is often the case that the early rounds of franchising raise problems that are resolved in subsequent rounds (see, for example, Domberger, Meadowcroft, & Thompson, 1987). Overall, the new evidence presented in this paper indicates a more favourable interpretation of the impact of passenger rail franchising in Britain on costs and productivity than the previous literature suggests. The remainder of the paper is organised as follows. Section 2 sets out the scope, coverage and methods adopted in the previous literature, with the key findings with respect to the cost, productivity and efficiency impact of British rail franchising detailed in Section 3. Section 4 describes the data and presents the new evidence. Section 5 concludes.

This paper set out to review and update the current state of knowledge concerning the impact of passenger rail franchising on costs, productivity and efficiency in Britain. This is a topic that is of considerable policy relevance both in Britain and in the context of the current review of ways of introducing competition into the domestic rail passenger market by the European Commission. Competitive tendering in passenger rail has also been adopted elsewhere in the world, most notably Melbourne and South America. The paper has first reviewed the literature and then provided some new evidence. The review of the literature showed that although productivity improved and costs came down initially after franchising, these cost reductions at the sector level were not sustained. The most up-to-date previous study showed productivity ending up in 2006 more or less exactly where it was at privatisation; and after taking account of changes in wages and terms and conditions of staff, a given set of passenger rail services in 2006 is found to cost 12% more in real terms than it did at privatisation. The review showed the drivers of the post-2000 cost rises to be as follows. Improved real wages and terms and conditions played a major role, and it has been argued that these increases cannot be viewed as entirely exogenous, in part being driven by the franchising structure adopted. Changes in average efficiency, driven largely by the management and re-negotiated contract arrangements put in place for distressed franchises were also found to have caused costs to rise. This finding suggests that re-negotiated or management contracts can be bad for efficiency, particularly if allowed to persist for longer than is necessary. However, based on previous analysis, the post-2000 deterioration in productivity even for the best performing firms (frontier shift) remained unexplained. The previous literature provides no consensus on the impact of franchise length on costs. It does however suggest evidence on scale and density that is broadly in line with the general literature on (vertically integrated) railways. That is, with modest scale economies and stronger economies of density, the literature suggests that reducing the number of franchises on its own may not reduce costs very much, but bigger gains are possible from merging franchises which overlap (so that the merged firm can, in principle, exploit economies of density). The new evidence reveals a somewhat more positive view of passenger rail franchising in Britain in terms of its impact on productivity and costs. In respect of enhancing our understanding of the cost trends over the period to 2006, we find that exogenous, sharp rises in diesel prices could explain around half of the unexplained upward shift in the cost frontier between 2000 and 2006 reported in Smith and Wheat (2009). Further, new data on TOC-caused delays shows that this more targeted measure of TOC performance fell by 11.4% on a per train-km basis between 2000 and 2006. In this respect it can now be said that higher costs have been associated with higher quality, at least in terms of the TOC contribution to performance. Turning to the updated analysis covering the period 2006–2008, our analysis shows that sector unit costs fell between 2006 and 2008. Competitive re-franchising therefore appears to be starting to bring costs down after a period of sharp cost growth. This fall is also not attributable to falling wages, as average wages have continued to rise in real terms and relative to the average earnings index, nor due to falling diesel prices, as the latter have also continued their upward rise in real terms (though less aggressively than during the earlier period). It should be noted that the post-2006 data may be less reliable than the earlier data, although there is no better dataset available. Furthermore, whilst there may be some uncertainty over the size of the fall in costs since 2006, the downward direction is nevertheless clear based on any reasonable interpretation of the data. The problem operators that ended up on management or re-negotiated contracts during the early period also saw their unit costs fall between 2006 and 2008; indicating that some of the inefficiency that accumulated in earlier years began to unwind as these operators were returned to more standard franchise agreements post-refranchising. However, these reductions were less than those achieved by the rest of the sector over this period, so the picture is not totally clear. An econometric analysis would be needed to explore this issue further. Overall, this paper has enhanced our understanding of TOC cost trends over the period since privatisation. The role that diesel prices have played in driving unexplained cost rises since 2000 has been quantified. Importantly, a potential link between rising TOC costs and improved TOC delays performance has been identified, though not proved statistically. The new, updated cost information also shows that between 2006 and 2008, costs started to fall across the whole cost sector, after many years of cost rises, suggesting that competitive re-franchising may now be starting to have a beneficial effect. Nevertheless, it remains the case that passenger rail franchising in Britain has failed to reduce costs in the way experienced in many other industries and in rail elsewhere in other European countries. The evidence is that somewhat larger franchises, avoiding overlapping and optimising train density and length, would reduce costs.