Measuring highway efficiency: A DEA approach and the Malquist index

Joaquim Miranda Sarmento, Luc Renneboog, Pedro Verga-Matos

Research output: Contribution to journalArticleScientificpeer-review

21 Citations (Scopus)
451 Downloads (Pure)

Abstract

A growing concern exists regarding the efficiency of public resources spent in transport infrastructures. In this paper, we measure the efficiency of seven highway projects in Portugal over the past decade by means of a data envelopment analysis and the Malmquist productivity and efficiency indices.
We distinguish between technical and technological efficiency and find that most highways face a reduction over time in both types of efficiency. This reduction is mainly due to an increase in operating and maintenance costs, follow-up investments, and a decline in traffic. Some highways only experience a reduction in technological efficiency after a decrease in traffic. They compensate with cost controls and stable investments. While controlling for scale efficiencies, we
find a lack of pure technical efficiency in highways that are not subject to a competitive environment, which produces a lack of incentives for better management. Not only does evidence exist of poor management due to a lack of competition, but the increased use of outsourcing also increases inefficiencies. The introduction of tolls and the outburst of the economic crisis in Portugal have substantially reduced traffic that further contributes to inefficiency. The local
context, such as highways in low-income areas and rural regions with a lower traffic density, also affects highway performance.
Original languageEnglish
Pages (from-to)530-551
JournalEuropean Journal of Transport and Infrastructure Research
Volume17
Issue number4
Publication statusPublished - Sept 2017

Keywords

  • DEA
  • Data Envelopment Analysis
  • highways
  • Malmquist index
  • productivity efficiency
  • transport

Fingerprint

Dive into the research topics of 'Measuring highway efficiency: A DEA approach and the Malquist index'. Together they form a unique fingerprint.

Cite this