Skip to main content

Advertisement

Log in

Regulation and cost efficiency in the European railways industry

Journal of Productivity Analysis Aims and scope Submit manuscript

Abstract

The objective of this article is to analyze European railways’ incentives to improve efficiency in the recent liberalization context. We build and estimate a structural model accounting for regulatory pressures faced by the firms. Our model includes demand equations, capacity constraints and a cost function, in which are specified an exogenous technical efficiency component and an endogenous cost reducing effort parameter. We find a significant positive effect of implementing the reforms on cost reducing activities, and a smaller cost of effort for firms choosing a more advanced separation of infrastructure from operation activities.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Similar content being viewed by others

Notes

  1. To ensure the effective enforcement of access rights, two complementary directives were created in 1995 and 2001. They clarify the conditions of access on several aspects such as licensing, capacity allocation and charging.

  2. Trans European Transport Networks.

  3. The European Rail Traffic Management System is aimed at enhancing European cross-border interoperability.

  4. See IDEI (2003) for a detailed description of railways regulatory regimes in European countries.

  5. We do not test for the effects of third party access or independent regulatory activity for two reasons. First, the debate mostly concerns how the separation of activities should be implemented, acknowledging that the two other reforms are the logical continuation of such a separation. Second, these two other reforms usually take place shortly after the separation is implemented, if not simultaneously, which would make it hard to identify separate effects.

  6. A step towards a deeper separation is observed for the Netherlands in 2003 and Spain in 2005.

  7. Note that for the European railways industry, Friebel et al. (2003) estimate a production function where the outputs aggregation process is determined endogenously in the model.

  8. We make the assumption that demand is independent of effort. Hence cost minimization is equivalent to profit maximization.

  9. The data we managed to gather for this variable were incomplete and introducing them in the estimations lead to unreasonable results.

  10. In our framework building such a model would require a description of the determination of prices on both markets.

  11. Note that series for Germany have to be treated and interpreted with caution as the two national railway companies DB and DR merged in 1994, simultaneously with the implementation of the European directive 91/440. .

  12. Note that cooperations of railways on certain high speed lines like Thalys or Eurostar are not included in our database.

  13. More appropriate measures for passenger and freight outputs are offered seat-kilometers and offered ton-kilometers, but available time series for these variables at Eurostat are too short.

  14. Wage per worker is computed as total wages divided by average operational staff.

  15. In locomotives here we include locomotives and automotives, both electrical and diesel.

  16. See “Appendix” for details on the calculations.

  17. Nevertheless at this stage we cannot say that the model with inefficiency only is preferred to the model without both inefficiency and effort. The test statistic is 0.34.

  18. According to Ritter and Simar (1997) this problem may be due to the small size of our sample.

  19. Note that an intermediary specification is the model with a distinction “before and after separation”, whatever the type of separation. We have estimated this model and we find it is indifferent to our reference model, as the separation is almost always the first reform to be implemented (see Table 1). This model is rejected against the one we are studying now.

  20. Operating expenses do not suffer major changes after the separation, as the operators have to pay access charges to the new infrastructure managers. Nevertheless, the management of remaining activities is simplified, leading to lower costs of implementing effort.

References

  • Affuso L, Angeriz A, Pollitt MG (2002) Measuring the efficiency of Britain’s privatised train operating companies. Regulation Initiative Discussion Paper Series, 48. London Business School

  • Battese GE, Coelli TJ (1992) Frontier production functions, technical efficiency and panel data: with application to paddy farmers in India. J Prod Anal 3:153–169

    Article  Google Scholar 

  • Cantos P (2001) Vertical relationships for the European railway industry. Transport Policy 8(2):77–83

    Google Scholar 

  • Cantos P, Maudos J (2001) Regulation and efficiency: the case of european railways. Transp Res Part A 35:459–472

    Google Scholar 

  • Cantos P, Pastor JM, Serrano L (1999) Productivity efficiency and technical change in the european railways: a non-parametric approach. Transportation 26:337–357

    Article  Google Scholar 

  • Caves D, Christensen LR, Tretheway MW (1998) Flexible cost functions for multiproduct firms. Rev Econ Stat 62:477–481

    Article  Google Scholar 

  • European Commission (2001), Whitepaper: European transport policy for 2010: time to decide, Brussels

  • Friebel G, Ivaldi M, Vibes C (2003) Railway (De)regulation: a European efficiency comparison. Economica 77(305):77–91

    Article  Google Scholar 

  • Gagnepain P, Ivaldi M (2002) Incentive regulatory policies: the case of public transit systems in France. RAND J Econ 33:605–629

    Article  Google Scholar 

  • Gagnepain P, Marin PL (2006) Regulation and incentives in European aviation. J Law Econ 49:229–248

    Article  Google Scholar 

  • Gathon HJ, Perelman S (1992) Measuring technical efficiency in European railways: a panel data approach. J Prod Anal 3:135–151

    Article  Google Scholar 

  • Gathon HJ, Pestieau P (1995) Decomposing efficiency into its managerial and its regulator components: the case of European railways. Eur J Operat Res 80:500–507

    Article  Google Scholar 

  • IDEI (2003) The economics of passenger rail transport: a survey. http://www.idei.fr/doc/wp/2003/rapport2_db_2.pdf

  • Jondrow J, Lovell CAK, Materov IS, Schmidt P (1982) On the estimation of technical inefficiency in the stochastic frontier production function model. J Econom 19:233–238

    Article  Google Scholar 

  • Kumbhakar SC, Lovell CAK (2000) Stochastic frontier analysis. Cambridge University Press, Cambridge

    Book  Google Scholar 

  • Laffont JJ (1994) The new theory of regulation ten years after. Econometrica 62:507–537

    Article  Google Scholar 

  • Oum TH, Yu C (1994) Economic efficiency of railways and implications for public policy. J Transp Econ Policy 28:121–138

    Google Scholar 

  • Prognos (1998) European Centre for Economic Research and Strategy Consulting, Examination of the implementation and impact of directive 91/440/EC on the development of the community’s railways. Report to the Union Internationale des Chemins de Fer, Basel

  • Ritter C, Simar L (1997) Pitfalls of normal-gamma stochastic frontier models. J Prod Anal 8:167–182

    Article  Google Scholar 

  • Vuong QH (1989) Likelihood ratio tests for model selection and non-nested hypothesis. Econometrica 57:307–333

    Article  Google Scholar 

Download references

Acknowledgments

We benefited from the comments of participants at the fifth IIOC conference in Savannah, the fourth Annual Conference on Railroad Industry Structure, Competition and Investment, and the Econometric seminar at TSE. We are also grateful to the comments received by Christian Bontemps, Jan K. Brueckner, Philippe Gagnepain, Marc Ivaldi, Markus Ksoll, and Andre de Palma. All remaining errors are ours.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Miguel Urdánoz.

Appendix: Details on the calculations of Eq. (10)

Appendix: Details on the calculations of Eq. (10)

From Eq. 6:

$$ \begin{aligned} C_{it}^{S} & = \beta_{0} Q_{{pax_{it} }}^{{\beta_{p} }} Q_{{fr_{it} }}^{{\beta_{f} }} w_{{L_{it} }}^{{\beta_{L} }} w_{{M_{it} }}^{{\beta_{M} }} A_{it}^{{\beta_{A} }} \exp \left( {\varepsilon_{it} } \right), \\ \varepsilon_{it} & = \theta_{it} - e_{it}^{S} + u_{{c_{it} }} , \\ \end{aligned} $$
$$ C_{it}^{S} = \beta_{0} Q_{{pax_{it} }}^{{\beta_{p} }} Q_{{fr_{it} }}^{{\beta_{f} }} w_{{L_{it} }}^{{\beta_{L} }} w_{{M_{it} }}^{{\beta_{M} }} A_{it}^{{\beta_{A} }} \exp \left( {\theta_{it} - e_{it}^{R} - e_{it}^{D} + u_{{c_{it} }} } \right) $$
$$ \begin{aligned} C_{it} & = \xi_{it}^{R} \left[ {\beta_{0} Q_{{pax_{it} }}^{{\beta_{p} }} Q_{{fr_{it} }}^{{\beta_{f} }} w_{{L_{it} }}^{{\beta_{L} }} w_{{M_{it} }}^{{\beta_{M} }} A_{it}^{{\beta_{A} }} \exp \left( {\theta_{it} - e_{it}^{R} + u_{{c_{it} }} } \right)} \right] \\ & + \xi_{it}^{D} \left[ {\beta_{0} Q_{{pax_{it} }}^{{\beta_{p} }} Q_{{fr_{it} }}^{{\beta_{f} }} w_{{L_{it} }}^{{\beta_{L} }} w_{{M_{it} }}^{{\beta_{M} }} A_{it}^{{\beta_{A} }} \exp \left( {\theta_{it} - e_{it}^{D} + u_{{c_{it} }} } \right)} \right], \\ \end{aligned} $$

from Eqs. 8 and 9

$$ \begin{aligned} e_{it}^{R} & = 0, \\ e_{it}^{D} & =\frac{1}{\mu + 1}\left( {\ln \left( {\beta_{0} } \right) +\beta_{p} \ln \left( {Q_{{pax_{it} }} } \right) + \beta_{f} \ln\left( {Q_{{fr_{it} }} } \right)} \right. \\ & + \beta_{L} \ln\left( {w_{{L_{it} }} } \right) + \beta_{M} \ln \left( {w_{{M_{it}}} } \right) + \beta_{A} \ln \left( {A_{it} } \right)\left. { +\theta_{it} - \ln \left( \mu \right) + u_{{c_{it} }} } \right), \\\end{aligned} $$
$$ \begin{aligned} C_{it} & = \xi_{it}^{R} \left[ {\beta_{0} Q_{{pax_{it} }}^{{\beta_{p} }} Q_{{fr_{it} }}^{{\beta_{f} }} w_{{L_{it} }}^{{\beta_{L} }} w_{{M_{it} }}^{{\beta_{M} }} A_{it}^{{\beta_{A} }} \exp \left( {\theta_{it} + u_{{c_{it} }} } \right)} \right] \\ & + \xi_{it}^{D} \left[ {\beta_{0} Q_{{pax_{it} }}^{{\beta_{p} }} Q_{{fr_{it} }}^{{\beta_{f} }} w_{{L_{it} }}^{{\beta_{L} }} w_{{M_{it} }}^{{\beta_{M} }} A_{it}^{{\beta_{A} }} \exp \left( {\theta_{it} - e_{it}^{D} + u_{{c_{it} }} } \right)} \right], \\ e_{it}^{D} & = \frac{1}{1 + \mu }\left( {\ln \left( {\beta_{0} } \right) + \beta_{p} \ln \left( {Q_{{pax_{it} }} } \right) + \beta_{f} \ln \left( {Q_{{fr_{it} }} } \right)} \right. \\ & + \beta_{L} \ln \left( {w_{{L_{it} }} } \right) + \beta_{M} \ln \left( {w_{{M_{it} }} } \right) + \beta_{A} \ln \left( {A_{it} } \right)\left. { + \theta_{it} - \ln \left( \mu \right) + u_{{c_{it} }} } \right), \\ \end{aligned} $$
$$ \begin{aligned} C_{it} & = \xi_{it}^{R} \left[ {\beta_{0} Q_{{pax_{it} }}^{{\beta_{p} }} Q_{{fr_{it} }}^{{\beta_{f} }} w_{{L_{it} }}^{{\beta_{L} }} w_{{M_{it} }}^{{\beta_{M} }} A_{it}^{{\beta_{A} }} \exp \left( {\theta_{it} + u_{{c_{it} }} } \right)} \right] \\ & + \xi_{it}^{D} \left[ {\beta_{0} Q_{{pax_{it} }}^{{\beta_{p} }} Q_{{fr_{it} }}^{{\beta_{f} }} w_{{L_{it} }}^{{\beta_{L} }} w_{{M_{it} }}^{{\beta_{M} }} A_{it}^{{\beta_{A} }} \exp \left( \begin{gathered} \theta_{it} - \frac{1}{1 + \mu }\left( {\ln \left( {\beta_{0} } \right) + \beta_{p} \ln \left( {Q_{{pax_{it} }} } \right) + \beta_{f} \ln \left( {Q_{{fr_{it} }} } \right)} \right. \hfill \\ + \beta_{L} \ln \left( {w_{{L_{it} }} } \right) + \beta_{M} \ln \left( {w_{{M_{it} }} } \right) + \beta_{A} \ln \left( {A_{it} } \right)\left. { + \theta_{it} - \ln \left( \mu \right) + u_{{c_{it} }} } \right) + u_{{c_{it} }} \hfill \\ \end{gathered} \right)} \right], \\ \end{aligned} $$
$$ \begin{aligned} C_{it}^{{}} & = \xi_{it}^{R} \left[ {\beta_{0} Q_{{pax_{it} }}^{{\beta_{p} }} Q_{{fr_{it} }}^{{\beta_{f} }} w_{{L_{it} }}^{{\beta_{L} }} w_{{M_{it} }}^{{\beta_{M} }} A_{it}^{{\beta_{A} }} \exp \left( {\theta_{it} + u_{{c_{it} }} } \right)} \right] \\ + & \xi_{it}^{D} \left[ \begin{gathered} \beta_{0} Q_{{pax_{it} }}^{{\beta_{p} }} Q_{{fr_{it} }}^{{\beta_{f} }} w_{{L_{it} }}^{{\beta_{L} }} w_{{M_{it} }}^{{\beta_{M} }} A_{it}^{{\beta_{A} }} \exp \left( {\theta_{it} - \frac{1}{1 + \mu }\left( {\left. { + \theta_{it} + u_{{c_{it} }} } \right) + u_{{c_{it} }} } \right.} \right) \hfill \\ \exp \left( \begin{gathered} - \frac{1}{1 + \mu }\left( {\ln \left( {\beta_{0} } \right) + \beta_{p} \ln \left( {Q_{{pax_{it} }} } \right) + \beta_{f} \ln \left( {Q_{{fr_{it} }} } \right)} \right. \hfill \\ + \beta_{L} \ln \left( {w_{{L_{it} }} } \right) + \beta_{M} \ln \left( {w_{{M_{it} }} } \right) + \beta_{A} \ln \left( {A_{it} } \right)\left. { - \ln \left( \mu \right)} \right) \hfill \\ \end{gathered} \right) \hfill \\ \end{gathered} \right], \\ \end{aligned} $$
$$ \begin{aligned} C_{it}^{{}} & = \xi_{it}^{R} \left[ {\beta_{0} Q_{{pax_{it} }}^{{\beta_{p} }} Q_{{fr_{it} }}^{{\beta_{f} }} w_{{L_{it} }}^{{\beta_{L} }} w_{{M_{it} }}^{{\beta_{M} }} A_{it}^{{\beta_{A} }} \exp \left( {\theta_{it} + u_{{c_{it} }} } \right)} \right] \\ & + \xi_{it}^{D} \left[ \begin{gathered} \beta_{0} Q_{{pax_{it} }}^{{\beta_{p} }} Q_{{fr_{it} }}^{{\beta_{f} }} w_{{L_{it} }}^{{\beta_{L} }} w_{{M_{it} }}^{{\beta_{M} }} A_{it}^{{\beta_{A} }} \exp \left( {\frac{\mu }{1 + \mu }\left( {\left. {\theta_{it} + u_{{c_{it} }} } \right)} \right.} \right) \hfill \\ \exp \left( \begin{gathered} - \frac{1}{1 + \mu }\left( {\ln \left( {\beta_{0} } \right) + \beta_{p} \ln \left( {Q_{{pax_{it} }} } \right) + \beta_{f} \ln \left( {Q_{{fr_{it} }} } \right)} \right. \hfill \\ + \beta_{L} \ln \left( {w_{{L_{it} }} } \right) + \beta_{M} \ln \left( {w_{{M_{it} }} } \right) + \beta_{A} \ln \left( {A_{it} } \right)\left. { - \ln \left( \mu \right)} \right) \hfill \\ \end{gathered} \right) \hfill \\ \end{gathered} \right], \\ \end{aligned} $$
$$ \begin{aligned} C_{it} & = \xi_{it}^{R} \left[ {\beta_{0} Q_{{pax_{it} }}^{{\beta_{p} }} Q_{{fr_{it} }}^{{\beta_{f} }} w_{{L_{it} }}^{{\beta_{L} }} w_{{M_{it} }}^{{\beta_{M} }} A_{it}^{{\beta_{A} }} \exp \left( {\theta_{it} + u_{{c_{it} }} } \right)} \right] \\ & + \xi_{it}^{D} \left[ {\beta_{0} Q_{{pax_{it} }}^{{\beta_{p} }} Q_{{fr_{it} }}^{{\beta_{f} }} w_{{L_{it} }}^{{\beta_{L} }} w_{{M_{it} }}^{{\beta_{M} }} A_{it}^{{\beta_{A} }} \exp \left( {\frac{\mu }{1 + \mu }\left( {\left. {\theta_{it} + u_{{c_{it} }} } \right)} \right.} \right)\beta_{0}^{{ - \frac{1}{1 + \mu }}} Q_{{pax_{it} }}^{{ - \frac{{\beta_{p} }}{1 + \mu }}} Q_{{fr_{it} }}^{{ - \frac{{\beta_{f} }}{1 + \mu }}} w_{{L_{it} }}^{{ - \frac{{\beta_{L} }}{1 + \mu }}} w_{{M_{it} }}^{{ - \frac{{\beta_{M} }}{1 + \mu }}} A_{it}^{{ - \frac{{\beta_{A} }}{1 + \mu }}} \mu^{{\frac{1}{1 + \mu }}} } \right], \\ \end{aligned} $$
$$ \begin{aligned} C_{it} & = \xi_{it}^{R} \left[ {\beta_{0} Q_{{pax_{it} }}^{{\beta_{p} }} Q_{{fr_{it} }}^{{\beta_{f} }} w_{{L_{it} }}^{{\beta_{L} }} w_{{M_{it} }}^{{\beta_{M} }} A_{it}^{{\beta_{A} }} \exp \left( {\theta_{it} + u_{{c_{it} }} } \right)} \right] \\ & + \xi_{it}^{D} \left[ {\mu^{{\frac{1}{1 + \mu }}} \beta_{0}^{{\frac{\mu }{1 + \mu }}} Q_{{pax_{it} }}^{{\frac{{\mu \beta_{p} }}{1 + \mu }}} Q_{{fr_{it} }}^{{\frac{{\mu \beta_{f} }}{1 + \mu }}} w_{{L_{it} }}^{{\frac{{\mu \beta_{L} }}{1 + \mu }}} w_{{M_{it} }}^{{\frac{{\mu \beta_{M} }}{1 + \mu }}} A_{it}^{{\frac{{\mu \beta_{A} }}{1 + \mu }}} \exp \left( {\frac{\mu }{1 + \mu }\left( {\left. {\theta_{it} + u_{{c_{it} }} } \right)} \right.} \right)} \right], \\ \end{aligned} $$
$$ \begin{aligned} C_{it} & = \xi_{it}^{R} \left[ {\beta_{0} Q_{{pax_{it} }}^{{\beta_{p} }} Q_{{fr_{it} }}^{{\beta_{f} }} w_{{L_{it} }}^{{\beta_{L} }} w_{{M_{it} }}^{{\beta_{M} }} A_{it}^{{\beta_{A} }} \exp \left( {\theta_{it} + u_{{c_{it} }} } \right)} \right] \\ & + \xi_{it}^{D} \left[ {\mu^{{\frac{1}{\mu + 1}}} \beta_{0}^{{\frac{\mu }{\mu + 1}}} Q_{{pax_{it} }}^{{\gamma \beta_{p} }} Q_{{fr_{it} }}^{{\gamma \beta_{f} }} w_{{L_{it} }}^{{\gamma \beta_{L} }} w_{{M_{it} }}^{{\gamma \beta_{M} }} A_{it}^{{\gamma \beta_{A} }} \exp \left( {\gamma \left( {\theta_{it} + u_{{c_{it} }} } \right)} \right)} \right], \\ \end{aligned} $$

with \( \gamma = \frac{\mu }{1 + \mu } \)

$$ \begin{aligned} C_{it} & = \xi_{it}^{R} \left[ {\beta_{0} Q_{{pax_{it} }}^{{\beta_{p} }} Q_{{fr_{it} }}^{{\beta_{f} }} w_{{L_{it} }}^{{\beta_{L} }} w_{{M_{it} }}^{{\beta_{M} }} A_{it}^{{\beta_{A} }} \exp \left( {\theta_{it} + u_{{c_{it} }} } \right)} \right] \\ & + \xi_{it}^{D} \left[ {c_{0} Q_{{pax_{it} }}^{{\gamma \beta_{p} }} Q_{{fr_{it} }}^{{\gamma \beta_{f} }} w_{{L_{it} }}^{{\gamma \beta_{L} }} w_{{M_{it} }}^{{\gamma \beta_{M} }} A_{it}^{{\gamma \beta_{A} }} \exp \left( {\gamma \left( {\theta_{it} + u_{{c_{it} }} } \right)} \right)} \right], \\ \end{aligned} $$

with \( \gamma = \frac{\mu }{1 + \mu } \) and \( c_{0} = \beta_{0} \exp \left( {\frac{{\ln \mu - \ln \beta_{0} }}{1 + \mu }} \right) \)

Rights and permissions

Reprints and permissions

About this article

Cite this article

Urdánoz, M., Vibes, C. Regulation and cost efficiency in the European railways industry. J Prod Anal 39, 217–230 (2013). https://doi.org/10.1007/s11123-012-0284-0

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11123-012-0284-0

Keywords

JEL Classification

Navigation