Using the Latent Class Approach to Cluster Firms in Benchmarking: An Application to the US Electricity Transmission Industry

Manuel Llorca*, Luis Orea, Michael Pollitt

*Corresponding author for this work

Research output: Contribution to journalJournal articleResearchpeer-review

29 Downloads (Pure)


In this paper we advocate using the latent class model (LCM) approach to control for technological differences in traditional efficiency analysis of regulated electricity networks. Our proposal relies on the fact that latent class models are designed to cluster firms by uncovering differences in technology parameters. Moreover, it can be viewed as a supervised method for clustering data that takes into account the same (production or cost) relationship that is analysed later, often using nonparametric frontier techniques. The simulation exercises show that the proposed approach outperforms other sample selection procedures. The proposed methodology is illustrated with an application to a sample of US electricity transmission firms for the period 2001–2009.
Original languageEnglish
JournalOperations Research Perspectives
Issue number1
Pages (from-to)6-17
Number of pages12
Publication statusPublished - 2014
Externally publishedYes


  • Electricity transmission
  • Utilities regulation
  • Latent class model approach
  • Nonparametric analysis

Cite this