A penalization approach for estimating inefficiency in stochastic frontier panel models

Firmin Doko Tchatoka, Magnus Söderberg and Mohammad Abbas Hakeem ()
Additional contact information
Mohammad Abbas Hakeem: The University of Adelaide
Authors registered in the RePEc Author Service: Mark Weder

Adelaide Economics Working Papers from Adelaide University, School of Economics

Abstract: Efficiency analysis is crucial for evaluating the performance of entities that provide essential and other homogenized services. The Jondow et al.’s (1982) estimator is widely used but has been criticized for biasing inefficiency toward its mean, distorting the distribution, and misrepresenting the conditional distribution of inefficiency, particularly in cross-sectional settings. Zeebari et al. (2023) propose a regularization approach that aligns sample and theoretical moments but it is primarily suited to cross-sectional data. This paper introduces a penalized mode estimator for unit inefficiency in panel data, accounting for heteroskedasticity in both inefficiency and idiosyncratic errors. A closed-form expression for this estimator is derived, and Monte Carlo simulations demonstrate its superior performance compared to existing methods. An empirical study of electricity providers in Australia, Canada, and New Zealand further underscores its advantages.

Keywords: Inefficiency; JLMS estimator; Heteroskedasticity; Restricted mode estimator (search for similar items in EconPapers)
Date: 2025-02
References: Add references at CitEc
Citations:

Downloads: (external link)
https://media.adelaide.edu.au/economics/papers/doc/wp2025-01.pdf (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:adl:wpaper:2025-01

Access Statistics for this paper

More papers in Adelaide Economics Working Papers from Adelaide University, School of Economics Contact information at EDIRC.
Bibliographic data for series maintained by Chansik Yoon () and Ruby Nguyen ().