Year: 1998
Type: preprint
Abstract: This paper surveys four distinct approaches to frontier estimation of multi-output (and simultaneously multi-input) technologies, when nothing but noisy quantity data are available. Parametrized distributions for inefficiency and noise are necessary for identification of inefficiency, when only cross-sectional data are available. In other respects suitable techniques may differ widely, as is shown. A final technique presented rigorously exploits the possibilities from panel-data by dropping parametrization of distributions as well as functional forms. It is illustrated how this last technique can be coupled with the others to provide a state-of-the-art estimation procedure for this setting. (author's abstract)
Author: Gstach, Dieter
Institution: Vienna University of Economics and Business
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W2117459845