Regulatory Design and Technical Efficiency: Public Transport in France

• Autores: Díaz, Guillermo, Vincent, Charles
• Formato: documento de trabajo
• Fecha de Publicación: 2015
• Institución: Pontificia Universidad Católica del Perú
• Repositorio: PUCP-Institucional
• Lenguaje: inglés
• OAI Identifier: oai:repositorio.pucp.edu.pe:20.500.14657/166770
• Enlace del recurso: https://repositorio.pucp.edu.pe/index/handle/123456789/166770; http://dx.doi.org/10.7835/ccwp-2015-11-0020
• Nivel de acceso: acceso abierto
• Materia: Data Envelopment Analysis (DEA); Conditional e ciency measures; Twostage e ciency analysis; Regulation; Public transport; http://purl.org/pe-repo/ocde/ford#5.02.04

Public transport systems are often subject to a close regulatory oversight because of their economic and social impacts. In the case of France, this has led to an institutional design that has involved the participation of private firms in the service provision, and the use of incentive contracts to regulate them, among other characteristics. We study the effect of these institutional features on the efficiency of the firms in the sector. For this, we use nonparametric Data Envelopment Analysis (DEA) techniques to estimate the input usage efficiency, and explore a few potential institutional and regulatory determinants. We apply a conditional DEA approach and fixed effects second stage regressions to control for potentially observed and unobserved sources of heterogeneity across different environments in which the firms operate. Our results point to a differential effect of private and mixed public-private companies. In particular, having the performance of public operators as the benchmark, efficiency is relatively higher for private firms, but lower when the service is delegated to a mixed public-private firm. Furthermore, the effects seem to diverge greatly by contract type when the firm is mixed so that, when the contract is of the cost reimbursement type, performance is lower than the public firm benchmark, while for other contract types there are no statistically significant differences.